Ischemic Optic Neuropathies

Ischemic Optic Neuropathies
Sohan Singh Hayreh
Ischemic
Optic Neuropathies
Dr. Sohan Singh Hayreh
Department of Ophthalmology
and Visual Science
University of Iowa Hospital and Clinics
College of Medicine
Hawkins Dr. 200
Iowa City, IA 522421091
USA
sohan-hayreh@uiowa.edu
ISBN 978-3-642-11849-4 e-ISBN 978-3-642-11852-4

DOI 10.1007/978-3-642-11852-4
Springer Heidelberg Dordrecht London New York
Library of Congress Control Number: 2011928229
Springer-Verlag Berlin Heidelberg 2011
This work is subject to copyright. All rights are reserved, whether the whole or part of the material is
concerned, specifically the rights of translation, reprinting, reuse of illustrations, recitation, broadcasting,
reproduction on microfilm or in any other way, and storage in data banks. Duplication of this publication
or parts thereof is permitted only under the provisions of the German Copyright Law of September 9, 1965,
in its current version, and permission for use must always be obtained from Springer. Violations are liable
to prosecution under the German Copyright Law.
The use of general descriptive names, registered names, trademarks, etc. in this publication does not imply,
even in the absence of a specific statement, that such names are exempt from the relevant protective laws
and regulations and therefore free for general use.
Product liability: The publishers cannot guarantee the accuracy of any information about dosage and appli-
cation contained in this book. In every individual case the user must check such information by consulting
the relevant literature.
Cover design: eStudioCalamar, Figueres/Berlin
Printed on acid-free paper
Springer is part of Springer Science+Business Media (www.springer.com)

Dedicated to all my patients with ischemic optic neuropathies

Preface
History teaches that the commencement of every branch of science is nothing more than a
series of observations and experiments which had no obvious connections with one another
Justus von Liebig (1846)
Ocular vascular occlusive disorders that is, optic nerve ischemic disorders, retinal isch-
emic disorders and retinal venous occlusive disorders collectively constitute the most
common cause of acute visual loss. Nevertheless, information on these disorders has until
recently been based primarily on clinical impressions, folklore, dogma, and conven-
tional wisdom taught by one generation of ophthalmologists to the next, rather than
based on systematic studies laying out basic scientific facts and pathogeneses. I have
investigated all these disorders systematically since 1955 by doing basic, experimental
and clinical studies. Based on the findings of my scientific studies, I have often chal-
lenged and contradicted dogmas and the conventional wisdom on these diseases.[1] As
Thomas Henry Huxley wrote: The great tragedy of Science the slaying of a beautiful
hypothesis by an ugly fact.[2] But the facts are what we need, even if they are sometimes
ugly and unwelcome. When dogmas and conventional wisdom are challenged, even
when new scientific information shows that they are no longer valid, the initial reaction is
almost always skepticism or even ridicule. I have experienced that so often. The objective
of this book is to clarify the understanding of ischemic optic neuropathies, in the light of
the best current scientific knowledge, and to put their management on a rational basis.
During my studies of ischemic optic neuropathies and other ocular vascular occlu-
sive disorders, I have found that one major reason for confusion and controversy about
these diseases has been lack of understanding of the basic scientific facts. The other
major reason has been the unthinking application of findings from superficially simi-
lar disorders in other organs to these disorders in the eye and optic nerve, even though
the two differ fundamentally in aspects such as morphology, physiology, pathology
and pathogenesis; a classical and highly prevalent example of that is the assumption
that findings about the pathogenesis and management of stroke can be applied to isch-
emic optic neuropathy.[3] As will be evident from the information in the book, that is
not true at all, and that has resulted in a confusion and controversy about the pathogen-
esis and management of non-arteritic anterior ischemic optic neuropathy. It seems
that, in the past, the lack of any hopeful treatment for these blinding disorders discour-
aged researchers from investing their valuable time in investigating them. Fortunately,
knowledge is constantly evolving and advancing. An in-depth understanding of the
morphology, blood supply and circulation of the optic nerve, the factors that influence
its blood flow, and the pathogenesis of various types of ischemic optic neuropathy is
crucial if we are to manage these diseases rationally; after all, the basic sciences are
the foundation of Medicine. In this book, I have tried to lay out in detail the basic
vii
viii Preface
scientific facts about various aspects of optic nerve ischemic disorders. Therefore, the
first section of the book covers the relevant basic scientific information related to isch-
emic optic neuropathies; the second section deals with the etiology, pathogenesis,
clinical features and management of the different types of ischemic optic neuropathy.
Some may comment that there is repetition of information among various chap-
ters. There are many common factors and issues between the various aspects of vari-
ous ischemic optic neuropathies, so that a certain amount of repetition among the
various chapters is unavoidable, for proper discussion and understanding; however, in
the interests of avoiding repetition, where appropriate, I have made references to
other chapters which contain detailed information on those particular topics, instead
of repeating the whole thing. Finally, since most persons tend to read only the parts
which interest them specifically, I have intentionally repeated some information in
various chapters in order to make each chapter self-explanatory.
Some clinical ophthalmologists may comment that I have gone to excessive length
in some places, in discussing the basic scientific facts, pathogeneses and controver-
sies. But since these diseases are ischemic in nature, an in-depth understanding of the
blood supply and blood flow of the optic nerve and other related issues is key to
understanding their pathogeneses. Also, when dealing with a subject such as ischemic
optic neuropathy, which has attracted ongoing controversy, one has to give a detailed
scientific background. Based on the most current scientific facts, I have discussed
controversial issues candidly and fully, sometimes discussing various issues raised in
those controversies point by point, to place them in proper perspective. Some may
consider that an unorthodox way of dealing with controversies in a book, but I feel
that is only way the reader can logically understand the controversial issues and the
subject. It is my sincere hope that the information in this book helps toward a wider
and better understanding of the ischemic optic neuropathies and the controversies.
I have given frequently my own studies while discussing the subject. Since ischemic dis-
orders of the optic nerve have been the focus of my research work since 1955, most of the
information in the book is based on my published studies. A search of Medline and other
indices of literature revealed thousands of articles by others, published on many aspects of
ischemic optic neuropathies. Their scientific credibility varies enormously. It is impossible to
cite all of them and one has to pick and choose. I have tried to cover as many relevant publica-
tions as possible. No doubt, I must have unintentionally missed some good material for one
reason or another, and I apologize for that. I have liberally used material from my own
previous publications where appropriate, when I do not have a better way to present or
phrase it than I did in those publications. In general, I have tried to avoid information from
anecdotal case reports or comments, which, in my opinion, do not necessarily provide valid
information. I have also avoided, as far as possible, using comments from review articles
written primarily by reviewers who have not themselves conducted scientific studies on the
subject. Unlike many authors, I have chosen to describe the findings of my research studies
in the first person rather than in the third person and passive voice; because I stand behind my
work. (I am grateful that my publishers, Springer Verlag, do not object to that preference).
References
1. Hayreh SS. Prevalent misconceptions about acute retinal vascular occlusive disorders. Prog
Retin Eye Res. 2005;24:493519.
2. Huxley TH. Presidential address at the British Association. Biogenesis and abiogenesis. Coll
Essays. 1870;8:229.
3. Hayreh SS. Ischemic optic neuropathy. Prog Retin Eye Res. 2009;28:3462.
Acknowledgements
The material in this book is mainly based on my research since 1955, but I cannot take
all the credit. During those 55 years, I have benefited mightily from the generosity
and kindness of many persons. I do not have space to name all those who have given
me advice, assistance and co-operation. I can make mention of only a few whose role
has been pivotal in this research.
When I left the Indian Army in 1955 to pursue an academic and research career,
the only opening available in India was in the anatomy department of a newly opened
medical college. Soon after I started there, I happened to meet my former teacher of
ophthalmology, late Professor Tulsi Dass, who was visiting that institution. During a
casual talk, I mentioned to him my interest in doing research, although I had never
done any research and had no idea what was involved. He told me that there had
recently been some publications in the British Journal of Ophthalmology dealing
with the vascularization of the optic pathway. That suggestion was the start of my
research career and impetus to explore that field. The other person, my mentor, to
whom I am forever grateful, is the late Sir Stewart Duke-Elder, Director of the
Institute of Ophthalmology, University of London, and the doyen of ophthalmology;
he started me on my ophthalmic career by agreeing to accept me in 1961 when I was
selected for the prestigious Beit Memorial Research Fellowship in Medical Sciences
of the University of London. The rest is history.
I have been helped in the conduct of my research by many people, not only col-
leagues but also Residents, Fellows, nurses and technicians over the years. Of all of
them, there were three persons whose help was critical. The expert help of Mr. John
Edwards, chief technician at the Department of Experimental Ophthalmology,
Institute of Ophthalmology, University of London, enabled me to learn and conduct
experimental studies, which provided important scientific information related to isch-
emic optic neuropathies. Mrs. Patricia Podhajsky, who has worked with me for over
36 years, first as a nurse and later as a research assistant, has been the backbone of my
clinical studies all along; without her expertise in handling, managing and auditing
the data in such a large cohort of patients, it would have been extremely difficult to
conduct my clinical studies. Finally, Professor M. Bridget Zimmerman has devoted
her extraordinary biostatistical expertise to the analysis and testing of my research
data, so that I could publish with complete confidence in the statistical soundness of
the findings. Without her help it would not have been possible to publish the studies
in large cohorts of cases in my studies.
In the preparation of this book, I have drawn heavily upon the help of several per-
sons. Professor Jost Jonas from Heidelberg University, Mannheim, Germany and
Professor Randy Kardon, Director of Neuro-ophthalmology in our department have
ix
x Acknowledgements
contributed one chapter each in their special fields of expertise. I am most grateful to
them. I am deeply grateful to Mrs. Patricia Duffel, our librarian, for her tireless help
with the literature search, and, above all, with the entire bibliography in this book
she has checked the accuracy and the style of every reference (vital in any scientific
publication) as well as helping with the preparation of the index. I am grateful to
Ms.Rita Gallo for her help in not only scanning most of the photographs given in this
book but also making sure that those are of high quality. I am grateful to my secretary,
Ms. Georgiane Perret, who has been with me for about 30 years now; she has played
an important role in many ways in this project over the years. Finally, I am grateful to
my wife Shelagh for her ungrudging support for my research over many years, and
her willingness (as a Classicist and linguist) to help me wrestle with the English
language. She has also been principal copy editor of this volume no small task.
Finally, I am grateful to Springer Verlag for agreeing to publish this book.
They also published my monograph entitled Anterior Ischemic Optic Neuropathy
in 1975.
Iowa City, Iowa, USA Sohan Singh Hayreh

Contents
1 Terminology and Types of Ischemic Optic Neuropathy . . . . . . . . . . . 1
Part I Basic Sciences
2 Structure of the Optic Nerve . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7
3 Blood Supply of the Optic Nerve . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35
4 Blood-Optic Nerve Barrier . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 79
5 Factors Influencing the Optic Nerve Head Blood Flow . . . . . . . . . . . . . 85
6 Measurement of the Optic Nerve Head Blood Flow . . . . . . . . . . . . . . . 111
Part II Clinical Sciences
7 Optic Disc . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 131
8 Clinical Evaluation of Patients with Ischemic Optic Neuropathies . . . 145
9 General Characteristics of Giant Cell Arteritis . . . . . . . . . . . . . . . . . . . 163
10 Clinical Features of Giant Cell Arteritis . . . . . . . . . . . . . . . . . . . . . . . . . 173
11 Diagnostic Criteria for Giant Cell Arteritis . . . . . . . . . . . . . . . . . . . . . . 179
12 Ophthalmic Manifestations of Giant Cell Arteritis . . . . . . . . . . . . . . . . 199
13 Management of Giant Cell Arteritis to Prevent Visual Loss . . . . . . . . 227
14 Pathogenesis of Classical Non-arteritic Anterior Ischemic Optic

Neuropathy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 265
15 Pathogenesis of Some Controversial Non-arteritic

Anterior Ischemic Optic Neuropathy Clinical Entities . . . . . . . . . . . . . 317
16 Clinical Features of Non-arteritic Anterior

Ischemic Optic Neuropathy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 337
17 Management of Non-arteritic Anterior Ischemic

Optic Neuropathy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 389
xi
xii Contents
18 Animal Model of Non-arteritic Anterior Ischemic

19 Classification and Incidence of Posterior Ischemic

20 Pathogenesis of Posterior Ischemic Optic Neuropathy . . . . . . . . . . . . . 427
21 Clinical Features of Posterior Ischemic Optic Neuropathy . . . . . . . . . 437
22 Management of Posterior Ischemic Optic Neuropathy . . . . . . . . . . . . . 445
Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 453
Terminology and Types of Ischemic Optic
Neuropathy 1
Terminology Infarction of the optic nerve [27]

Acute ischemia of the disc [28]
Ischemic edema of the disc [29, 30]
Ischemic optic neuropathy has been given different
Ischemic optic neuropathy [31, 32]
eponyms over the years because of differing opinions
Idiopathic ischemic optic neuropathy [33]
on its pathogenesis. Following is a list of the terms
which have been used to describe this condition.
Optic neuritis [1, 2]
Ischemic optic neuritis [3, 4] Types of Ischemic Optic Neuropathy
Ischemic retrobulbar neuritis [5, 6]
Acute optic neuritis of hypertension or arterioscle-
As will be discussed later (see Chap. 3), the optic nerve
rosis [710]
can be divided into two distinct regions based on blood
Ischemic neuritis of the papilla [11]
supply (Fig.1.1):
Arteriosclerotic papillitis [9, 10, 1215]
Senile papillitis or senile papillopathy [16] 1. The anterior part (optic nerve head), which is sup-
Pseudo-papillitis [14] plied primarily by the posterior ciliary artery circu-
Vascular pseudo-papillitis [1723] lation [3538].
Papillary apoplexy [24] 2. The rest of the optic nerve, which is not supplied by
Ischemic papillopathy [25] the posterior ciliary arteries, but is supplied from
Optico-malacia [26] multiple sources [36, 3941].
C
R S Col. Br.
PCA
D
A
Pia
Fig.1.1 Schematic
representation of blood
supply of the optic nerve.
A arachnoid, C choroid, CRA
central retinal artery, Col. Br. ON
Collateral branches, CRV
PR
central retinal vein, D dura,
LC lamina cribrosa, ON optic LC
nerve, PCA posterior ciliary
SAS
artery, PR prelaminar region,
R retina, S sclera, SAS
CRV
subarachnoid space (Modified CRA
from Hayreh [34]) PCA
S.S. Hayreh, Ischemic Optic Neuropathies, 1

DOI: 10.1007/978-3-642-11852-4_1, Springer-Verlag Berlin Heidelberg 2011
2 1 Terminology and Types of Ischemic Optic Neuropathy
Also, the optic nerve head is morphologically different osterior Ischemic Optic
P
from the rest of the optic nerve (see Chap. 2). Neuropathy (PION)
From my studies on the blood supply of the optic
nerve [35, 36, 3941], and my experimental [42] and
clinical [4345] studies, I found in the early 1970s Etiologically, this can be classified into:
that interference with the posterior ciliary artery cir- 1. Arteritic PION: This is almost invariably due to
culation resulted in the development of the clinical giant cell arteritis.
entity for which I coined the term anterior isch- 2. Non-arteritic PION: This is due to causes other
emic optic neuropathy AION [43]. From the sci- than giant cell arteritis.
entific point of view, this term represents the exact 3. Surgical PION: This is almost invariably attributable
site and ischemic nature of the lesion in the optic to a distant systemic surgical procedure [5052].
nerve. An accompanying editorial stated: The This entity has also been called postoperative [53]
abbreviation AION adopted by Hayreh almost of or perioperative [50]. I have used the term surgi-
necessity presupposes the existence of a posterior cal PION [51, 52] because it is more inclusive.
counterpart PION which might, perhaps, be
Thus, the clinical entity ischemic optic neuropathy
applied to the post-traumatic optic atrophy associ-
actually consists of five distinct clinical entities which
ated with damage to the optic nerve within its bony
differ etiologically, pathogenetically, clinically and from
canal. [46] I was later able to describe for the first
the management point of view. All are caused by acute
time the clinical entity of posterior ischemic optic
ischemia due to a variety of causes, as discussed later.
neuropathy (PION) [47]. Based on that informa-
Unfortunately, as is often the case in medical terminol-
tion, I therefore classified ischemic optic neuropathy
ogy, many of those older, misleading terms are still being
into two types:
used, particularly the term ischemic optic neuropathy
Anterior ischemic optic neuropathy (AION)
or optic neuritis for AION. The other terms which I
involving the anterior part of the optic nerve supplied
have found used not uncommonly are acute ischemic
by the posterior ciliary artery circulation, and
optic neuropathy and acute nonarteritic ischemic optic
Posterior ischemic optic neuropathy (PION)
neuropathy in many studies dealing with NA-AION;
involving the rest of the optic nerve not supplied by the
however, since there is no chronic ischemic optic neu-
posterior ciliary artery circulation [4749].
ropathy and it is always acute, there is little validity
Since then a colossal amount of literature has
in the use of the terms acute ischemic optic neuropa-
accumulated on both AION and PION. Neither
thy or acute nonarteritic ischemic optic neuropathy.
AION nor PION is a single clinical entity; each con-
The use of a correct, scientifically valid term to
sists of the following distinct clinical entities with
describe a clinical entity is essential. The correct term
different etiologies, pathogeneses, clinical features
reflects the true nature of the disease. It is exactly like
and managements.
describing different persons by their correct names, to
identify whom we are talking about.
nterior Ischemic Optic

A
Neuropathy (AION) elative Frequency of Various Types
R
of Ischemic Optic Neuropathies
This comprises:
1. Arteritic AION (A-AION): This is almost invari- So far there has been little information on this subject.
ably due to giant cell arteritis. I analyzed this in all patients with various types of isch-
2. Non-arteritic AION (NA-AION): This is due to emic optic neuropathies seen in my clinic from 1973 to
causes other than giant cell arteritis. 2008. There were about 1,400 patients. I analyzed the
References 3
relative frequency of the five types of ischemic optic 10. Desvignes P, Brun M. Nvrite optique aigu avec artrio-
neuropathy among this cohort. Therelative frequencies sclrose de lartre ophthalmique. [Acute optic neuritis with
arteriosclerosis of the optic artery]. Bull Mm Soc Fr
of AION and PION were 96% and 4% respectively. Ophtalmol. 1952;65:1103.
Since PION is a diagnosis of exclusion [47], however, 11. Hollenhorst RW, Wagener HP. The ocular fundi in relation to
it may sometimes have been missed, so that the fre- operations for hypertensive cardiovascular disease. Am J
quency I found may be a slight underestimate. Med Sci. 1949;218(2):22534.
12. Kurz O. ber Papillitis arteriosclerotica. Ophthalmologica.
Anterior ischemic optic neuropathy: The total 1948;116(45):2815.
number of patients seen in my clinic with this condi- 13. Larmande AM. Oedmes papillaires purs dorigine vascu-
tion was about 1,350. Of those about 90% had laire en dehors de lhypertension arterielle. Bull Soc
NA-AION and 10% arteritic AION. Ophtalmol Fr (Paris). 1948;8:5416.
14. Lasco F. Les affections vasculaires du nerf optique et leurs
Posterior ischemic optic neuropathy: There were manifestations cliniques. (Vascular disorders of the optic
about 50 patients. Of those, about 66% patients had nerve and their clinical manifestatios.). Ophthalmologica.
non-arteritic PION, 26% arteritic PION, and 7% surgi- 1961;142:42945. et seq.
cal PION. 15. Sanna M. Pseudopapilliti vascolari. Riv Otoneurooftalmol.
1958;33:44970.
Of course, the relative proportion of the various 16. Behrman S. Optic neuritis, papillitis, and neuronal retinopa-
types of ischemic optic neuropathy will vary among thy. Br J Ophthalmol. 1964;48:20917.
different clinics, depending upon the total number of 17. Franois J. Pseudo-papillites vasculaires (Vascular pseudo-
patients seen and the type of practice, but the above papillitis). Bull Soc Ophtalmol Fr. 1956;69:3657.
18. Francois J, Verriest G, Baron A. Pseudo-papillites vasculai-
data provides a useful general guide. res. (Vascular pseudo-papillitis.). Acta Ophthalmol (Copenh).
1957;35(1):3252.
19. Franois J, Verriest G, DeRouck AL. Llectro-oeulographie
en tant quexamen functionnel de la rtine. (Electro-
oculography in functional examination of the retina). Adv
References Ophthalmol. 1957;7:164.
20. Francois J, Verriest G, Neetens A, de R, Hanssens M. Psedo
nvrites optiques oedmateuses dorigine vasculaire. (Vascular
1. Carroll FD. Optic neuritis; a 15year study. Am J Ophthalmol. pseudopapillitis.). Ann Ocul (Paris). 1962;195:83085.
1952;35(1):7582. 21. Francois J, Verriest G, Neetens A, De Rouck A.
2. Reese AB, Carroll FD. Optic neuritis following cataract Pseudopapillites vasculaires. (Vascular pseudo-papillitis.).
extraction. Am J Ophthalmol. 1958;45(5):65962. Bull Soc Belge Ophtalmol. 1961;129:41327.
3. Birkhead NC, Wagener HP, Shick RM. Treatment of tempo- 22. Georgiads G, Konstas P, Stangos N. Rflexions issues de
ral arteritis with adrenal corticosteroids; results in fifty-five ltude de nombreux cas de pseudo-papillite vasculaire.
cases in which lesion was proved at biopsy. J Am Med (Study of numerous cases of vascular pseudopapillitis). Bull
Assoc. 1957;163(10):8217. Mm Soc Fr Ophtalmol. 1966;79:50639.
4. Hollenhorst RW, Brown JR, Wagener HP, Shick RM. 23. Moro F. Pseudo-papillite vascolare in arterite temporale
Neurologic aspects of temporal arteritis. Neurology. occulta. Atti Soc Oftal Lomb. 1963;18:7480.
1960;10:4908. 24. Rintelen F. On a study of conduction disorders of the fasicu-
5. Bruce GM. Temporal arteritis as a cause of blindness; review lus opticus, especially apoplexia papillae. Ophthalmologica.
of literature and report of a case. Trans Am Ophthalmol Soc. 1961;141:2839.
1949;47:30016. 25. Sanders MD. Ischaemic papillopathy. Trans Ophthalmol Soc
6. Bruce GM. Temporal arteritis as a cause of blindness, review UK. 1971;91:36986.
of the literature and report of a case. Am J Ophthalmol. 26. Kreibig W. Optikomalazie, die Folge eines Gefassverschlusses
1950;33(10):156873. im retrobulbren Abschnitt des Sehnerven. (Opticomalacia
7. Bonamour G. Les atrophies optiques chez les hypertendues caused by vascular occlusion in the retrobulbar portion of
arteriales. Bull Soc Ophtalmol Fr. 1954;1:1202. optic nerves.). Klin Monatsbl Augenheilkd Augenrztl
8. Bonamour G. Troubles vaso-moteurs oculaires dans les trau- Fortbild. 1953;122(6):71931.
matismes craniens fermes (presentation). Entretiens Annuels 27. Ellenberger Jr C, Netsky MG. Infarction in the optic nerve.
dOphtalmologie. 1955;n.p. J Neurol Neurosurg Psychiatry. 1968;31(6):60611.
9. Desvignes P, Brun M. Nvrite optique aigu avec artrio- 28. Siegert P. Akute Ischamie der Papille bei Arteriitis cranialis
sclrose de lartre ophthalmique. [Acute optic neuritis with (temporalis). (Acute ischemia of the papilla caused by arteri-
arteriosclerosis of the optic artery]. Bull Mem Soc Fr tis temporalis). Klin Monbl Augenheilkd Augenarztl
Ophtalmol. 1952;65:110113. Fortbild. 1952;120(3):25473.
4 1 Terminology and Types of Ischemic Optic Neuropathy
29. Bessiere E, Julien RG. Les manifestations papillaires dans le 41. Hayreh SS. Blood supply and vascular disorders of the optic
syndrome dit arterite temporale, frequence de loedeme nerve. An Inst Barraquer. 1963;4:7109.
papillaire ischemique. (Papillary manifestations in the syn- 42. Hayreh SS, Baines JA. Occlusion of the posterior ciliary
drome called temporal arteritis; frequency of ischemic papil- artery. III. Effects on the optic nerve head. Br J Ophthalmol.
lary edema.). Arch Ophtalmol Rev Gn Ophtalmol. 1972;56(10):75464.
1950;10(6):70114. 43. Hayreh SS. Anterior ischaemic optic neuropathy. I.
30. Kadlecova V. Ischemic papillary edema. (in Czech). Cesk Terminology and pathogenesis. Br J Ophthalmol.
Oftalmol. 1951;7(6):38894. 1974;58(12):95563.
31. Cogan DG. Neurology of the visual system. Springfield: 44. Hayreh SS. Anterior ischaemic optic neuropathy. II. Fundus
Charles C. Thomas; 1966. p. 137. 173, 1858. on ophthalmoscopy and fluorescein angiography. Br
32. Miller GR, Smith JL. Ischemic optic neuropathy. Am J J Ophthalmol. 1974;58(12):96480.
Ophthalmol. 1966;62(1):10315. 45. Hayreh SS. Anterior ischaemic optic neuropathy. III.
33. Boghen DR, Glaser JS. Ischaemic optic neuropathy. The Treatment, prophylaxis, and differential diagnosis. Br
clinical profile and history. Brain. 1975;98(4):689708. J Ophthalmol. 1974;58(12):9819.
34. Hayreh SS. Anatomy and physiology of the optic nerve 46. Editorial. Ischaemic optic neuropathy. Br J Ophthalmol.
head. Trans Am Acad Ophthalmol Otolaryngol. 1974;58(12):9534.
1974;78:OP24054. 47. Hayreh SS. Posterior ischemic optic neuropathy.
35. Hayreh SS. Blood supply of the optic nerve head and its role Ophthalmologica. 1981;182(1):2941.
in optic atrophy, glaucoma, and oedema of the optic disc. Br 48. Hayreh SS. Ischemic optic neuropathy. Int Ophthalmol.
J Ophthalmol. 1969;53(11):72148. 1978;1:918.
36. Hayreh SS. Pathogenesis of visual field defects. Role of the 49. Hayreh SS. Ischemic optic neuropathy. Acta XXIII Conc
ciliary circulation. Br J Ophthalmol. 1970;54(5):289311. Ophthalmol Kyoto. 1978;1:3138.
37. Hayreh SS. The 1994 Von Sallman lecture: the optic nerve 50. Sadda SR, Nee M, Miller NR, Biousse V, Newman NJ,
head circulation in health and disease. Exp Eye Res. Kouzis A. Clinical spectrum of posterior ischemic optic neu-
1995;61(3):25972. ropathy. Am J Ophthalmol. 2001;132(5):74350.
38. Hayreh SS. The blood supply of the optic nerve head and the 51. Hayreh SS. Neuropatas pticas Isqumicas. In: Ginebreda
evaluation of it: myth and reality. Prog Retin Eye Res. JA, Dalmau BS, editors. Neuropatas pticas diagnstico y
2001;20(5):56393. tratamiento. Barcelona: LXXVIII Ponencia Oficial de la
39. Singh S, Dass R. The central artery of the retina. II. A study Sociedad Espaola de Oftalmologa; 2002. p. 20937.
of its distribution and anastomoses. Br J Ophthalmol. 52. Hayreh SS. Posterior ischaemic optic neuropathy: clinical
1960;44:28099. features, pathogenesis, and management. Eye.
40. Hayreh SS. The blood supply of the optic nerve and its clinical 2004;18(11):1188206.
significance. Acta XIX Conc Ophthalmol New Delhi. 53. Roth S, Barach P. Postoperative visual loss: still no answers
1962;2:11949. yet. Anesthesiology. 2001;95(3):5757.
Part
I
Basic Sciences

Structure of the Optic Nerve
2
A good understanding of the structure of the optic shows considerable variations, all of which may be
nerve and topographic localization of the nerve fibers within physiological limits (see Chap. 7). I prefer to
in it from the various parts of the retina is essential for use the term optic nerve head; however, I have
comprehension of various aspects of ischemic optic retained the term optic disc, for two main areas of
neuropathy. usage: firstly, in reviews of work by other authors
The length of the optic nerve varies widely, even where I am not sure how much of the optic nerve
between the two eyes of the same person and is headhas been included by them under this term, and
3555mm from the eyeball to the chiasma (intraocular secondly, in ophthalmoscopic descriptions of optic
part 1 mm, intraorbital part 25 mm, intracanalicular nerve head lesions, where it is not possible to be defi-
part 410 mm and intracranial part 10 mm [1]). For nite as to how much of the nerve head is involved.
descriptive purposes, the optic nerve can be divided The optic nerve head is about 1mm long and about
into the following four parts: 1.5mm (1.181.75mm [3, 4]) in diameter, the vertical
diameter being slightly greater than the horizontal
1. Optic nerve head
Ishii [5] described the average horizontal diameter as
2. Intraorbital part
1.618mm and the vertical diameter as 1.796mm. The
3. Intracanalicular part
diameter of the optic disc depends upon the diameter
4. Intracranial part
of the chorioscleral canal at the level of Bruchs mem-
brane. The canal is usually conical in shape the pos-
terior part wider than the anterior part; but it may be
cylindrical, and very rarely it may either be of a
The Optic Nerve Head truncated (or triangular) type with its central part
narrowest [6] or an elbow extension type with the central
In the literature, the ophthalmoscopic term optic disc part widest [7]. The ophthalmoscopic configuration of
has been applied interchangeably either to the whole the optic disc and of the physiological cup depends
or a part of the anterior-most part of the optic nerve upon the size, shape and direction of the canal, as well
head (i.e., the surface nerve fiber layer and the prelami- as the diameter of the opening in the Bruchs mem-
nar region) or to the entire optic nerve head. Similarly brane and scleral canal. The shorter the canal diameter,
the term papilla has been used as a synonym for the the smaller is the cup and vice versa. This is because
optic disc or optic nerve head. The term papilla was the total optic nerve tissue volume (i.e., the nerve
coined by Briggs in 1676 [2] based on an erroneous fibers, glial tissue and blood vessels) in all probability
impression that the normal optic nerve head was ele- does not vary significantly in normal eyes; if the canal
vated like a papilla. Since the structure usually is not is narrow the available space may be just enough to
elevated above the level of the adjacent retina, lies in accommodate the tissue, with no physiological cup,
the same plane as retina and has in fact a central depres- but, on the other hand, with a wide canal the extra
sion (i.e., the physiological cup) or may even be flat, space may manifest as a physiological cup the wider
the term papilla is a misnomer. It must, however, be the canal, the larger the cup. The shape of the canal and
conceded that the appearance of the optic nerve head mode of insertion of the optic nerve into the eyeball

8 2 Structure of the Optic Nerve
may determine the shape of the cup. For example,

oblique insertion of the optic nerve results in a differ-
ent shape of the optic disc and size of the optic disc cup
(see Fig.7.8).
We [8] did histological studies on 12 eyes and
opticnerves of adult rhesus monkeys which provided
useful information. Most of the histological descrip-
tion in this chapter is based on our study [8], combined
with that provided by other studies. For descriptive
purposes, the structure of the optic nerve head can be
further subdivided into the following three regions
from front to back (Figs.2.12.4); however, there is no
sharp line of demarcation between these zones of the
optic nerve head.
1. The surface nerve fiber layer
2. The prelaminar region
3. The lamina cribrosa region
The Surface Nerve Fiber Layer Fig. 2.1 Longitudinal section of a half of the normal human
optic nerve showing the region of the optic nerve head and the
retrolaminar optic nerve. LCR Lamina cribrosa region, PLR
This is the most anterior layer of the optic nerve head,
Prelaminar region, RLR Retrolaminar region, SNFL Surface
containing the compact optic nerve fibers as they con- nerve fiber layer
verge here from all over the retina and bend to run into
the optic nerve (Figs.2.12.6). It is separated from the
vitreous by the inner limiting membrane of Elschnig The Prelaminar Region
[3, 4] which is composed of astrocytes [810]
(Fig.2.7). In the region of the physiological cup, this This part of the optic nerve head has been called the
membrane is often thick and is called the central glial, choroidal or more commonly anterior part of the
meniscus of Kuhnt, but is very thin when the cup is lamina cribrosa [8, 1117] which has created a certain
large. In addition to these tissues, this layer of the optic amount of confusion in communication. Sometimes
nerve head has a large number of blood vessels, even the very existence of this glial region as a dis-
consisting of not only a dense capillary network on its tinct entity has been ignored [11, 12]. In any discus-
surface but also the large retinal vessels and venous sion on pathologic changes in anterior ischemic optic
tributaries (Fig. 2.8). A conspicuous remnant of the neuropathy, optic disc edema and glaucoma, this is
hyaloid artery (Bergmeisters papilla) is very rare on perhaps one of the most important regions of the optic
the human disc (Fig. 2.9); however, in my study of nerve.
rhesus monkeys, I have found it in almost all the eyes, This is almost entirely built up by glial tissue and the
examined both histologically as well as ophthalmo- connective tissue fibers are seen only in connection with
scopically (Figs.2.4 and 2.10); it has a well-developed the vessels [8]. Glial fibers are much finer than the con-
muscular coat, with, of course, a superficial coating of nective tissue fibers and they still retain a direction per-
the glial tissue. pendicular to the nerve fiber bundles. Centrally, they
The region posterior to the surface nerve fiber layer remain attached to the connective tissue surrounding the
is subdivided into three parts from front backward [8]: central retinal vessels (Fig.2.11), and peripherally they
(1) pure glial part (prelaminar region), (2) mixed part are attached to the choroid as well as to the elastic mem-
a transition zone between the prelaminar and the lamina brane (Fig.2.12). The number of glial cells in this part
cribrosa regions, and (3) connective tissue part (lamina is enormous; they are markedly flattened in the antero-
cribrosa). posterior direction and are packed in dense transverse
The Optic Nerve Head 9
Fig.2.2 Longitudinal section

of a normal rhesus monkey
optic nerve showing the
region of the optic nerve head
and the retrolaminar optic
nerve. LCR Lamina cribrosa
region, PLR Prelaminar region,
RLR Retrolaminar region,
SNFL Surface nerve fiber
layer
Fig.2.3 Longitudinal section

of a normal rhesus monkey
optic nerve showing the
region of the optic nerve head
and the retrolaminar optic
nerve and central retinal
artery (A) and vein (V). LCR
Lamina cribrosa region, PLR
Prelaminar region, RLR
Retrolaminar region, SNFL
Surface nerve fiber layer
Fig.2.4 Longitudinal
section of a normal rhesus
monkey optic nerve showing
the region of the optic nerve
head and the retrolaminar
optic nerve and
Bergmeisters papilla (BP).
LCR Lamina cribrosa region,
PLR Prelaminar region, RLR
Retrolaminar region, SNFL
Fig.2.5 Bundles of the

optic-nerve fibers running
from the retina to the lamina
cribrosa in normal rhesus
monkey. (Golgis stain) LCR
Lamina cribrosa, PLR
Prelaminar region, SNFL
The prelaminar region consists of optic nerve fibers

arranged in bundles (Figs. 2.12.6), surrounded by
tube-like glial channels formed by specialized astro-
cytes, spider cells [11, 12, 15]. The glial tissue
between the nerve fiber bundles forms trabeculae
(Figs. 2.152.17), and capillaries are located within
the glial septa. A narrow, perivascular, connective
tissue space accompanies most of the capillaries [15].
Wolter [11, 12] described the presence of a shallow,
cap-like wicker basket, composed of the spider
cells lying in this part of the optic nerve head, closely
connected to the lamina cribrosa at its base, and with
its rounded convexity toward the vitreal surface of the
optic nerve head. According to him, the basket acts as
an important supporting, protective and nutritive organ
to the nerve fibers. In our histological study [8] on the
optic nerve head of rhesus monkeys, we did not observe
Fig.2.6 Bundles of the optic-nerve fibers running from the ret- an anterior limit of this basket. The only connective
ina to the lamina cribrosa in normal rhesus monkey (Golgis tissue seen in this part of the optic nerve head is that
stain). The nerve fibers are non-myelinated in the retina and
prelaminar and lamina cribrosa regions but myelinated in the
accompanying the capillaries [8, 15]. Wolter [11, 12]
retrolaminar region. LCR Lamina cribrosa, PLR Prelaminar described glial fibers surrounding both the nerve fiber
region, RLR Retrolaminar region, SNFL Surface nerve fiber bundles and the individual nerve fibers in the bundle.
layer (Reproduced from Hayreh and Vrabec [8]) Electron microscopic studies have shown that while
the bundles are surrounded by the astrocytes, their long
sheets (Figs.2.11 and 2.13). There may be many capil- processes extend into the fascicles at right angles to the
laries, surrounded by a glial limiting membrane, built up nerve fibers [18]; Anderson [15], however, found only
from the footplates of the glial cells (Fig.2.14). an occasional astrocyte process going to the nerve
Fig.2.7 Glial elements

at the bottom of the
physiological excavation in
rhesus monkey (Hortegas
stain) (Reproduced from
Hayreh and Vrabec [8])
Retinal vein
Retinal artery
OD
Cilio-retinal artery Fig.2.9 Fundus photograph of a human eye showing obliter-

ated remnant of the hyaloid artery (Bergmeisters papilla BP) at
Fig. 2.8 Diagrammatic representation of blood vessels in the the center of the optic disc
surface nerve fiber layer. OD Optic disc
fibers. The glial cells in the prelaminar portion are depression is separated from the vitreous by the inner
loose and the arrangement of the glial sheets is lost [8]. limiting membrane of Elschnig (Figs. 2.7, 2.20 and
It would seem probable that this loose arrangement of 2.21). When the depression does not reach the level of
the glial framework in the prelaminar portion might the lamina cribrosa, the central connective tissue sends
play a role in the pathologic swelling of the optic disc some anchoring fibers distally to the bottom of the
(Figs.2.18 and 2.19). excavation [8] (Fig. 2.20) In other cases the central
In the central part exists a central depression of excavation may reach the lamina cribrosa (Fig.2.21). In
varying degree corresponding to the physiological all these instances, the central connective tissue reaches
cup (Figs. 2.12, 2.20 and 2.21). The bottom of this the internal limiting membrane of Elschnig [8].
Fig.2.10 Obliterated
remnant of the hyaloid artery
(Bergmeisters papilla BP)
at the optic nerve head in a
rhesus monkey. (Hortegas
stain) LCR Lamina cribrosa,
Retrolaminar region
(Reproduced from Hayreh
and Vrabec [8])
Fig.2.11 Glial portion of

the lamina cribrosa, showing
its central attachment to the
connective-tissue sheath of
the central retinal vessels and
the arrangement of glial
fibers. (Hortegas stain) (A)
Artery (Reproduced from
section of a half of the optic Cup
nerve head, showing
peripheral anchorage of the
connective-tissue part of the
lamina cribrosa to the sclera,
and of the glial part to the
choroid and elastic lamina
(Bruchs membrane)
(Hortegas stain)
and Vrabec [8])
Fig.2.13 Prelaminar region

of the optic nerve head with
very fine fibers and a large
number of glial cells,
markedly flattened in the
anteroposterior direction and
packed in dense transverse
sheets (Hortegas stain)
and Vrabec [8])
Fig.2.14 Glial membrane

surrounding the capillaries
(arrow) in the prelaminar
region of the optic nerve
head (Hortegas stain)
and Vrabec [8])
The prelaminar region at its edge is separated by The optic nerve fibers coming from the retina make
several layers of astrocytes from the adjacent deeper a 90 bend in the surface layer of the optic nerve head
layers of the retina (Intermediate tissue of Kuhnt to run back in the optic nerve (Figs.2.12.6); and as
[6]) and from the adjacent choroid (Border tissue they make the bend their main support is the glial
of Jacoby [19]). These astrocytes not only form a tissue of the prelaminar region. The nerve fibers are
column around the circumference of the chorioretinal arranged in bundles. Within the bundles, the axons
canal and send processes internally among the nerve are separated either from each other by an intercellu-
bundles but also invest the connective tissue of all ves- lar space measuring 150 , or from an astrocyte
sels entering this portion of the optic nerve head [18]. process by a similar space [18]. Cohen [20] found the
Fig.2.15 Transverse
section of the glial part of the
prelaminar regions of the
optic nerve head showing the
arrangement of the septa
Fig.2.16 Transverse
section of the glial part of the
prelaminar regions of the
axons in this region to have no organized glial The Lamina Cribrosa Region
separation, often being contiguous with one another
for up to 200mm. The nerve fibers in this part, as in This region of the optic nerve head has been described
the surface nerve fiber layer and the retina, are unmy- as the scleral or posterior part of the lamina cribrosa.
elinated and vary in diameter, and are arranged in Ihave restricted the term lamina cribrosa to this part of
bundles; toward the retina the bundles become closely the optic nerve head in order to avoid unnecessary
packed (Figs.2.5 and 2.6). confusion.
Fig.2.17 Transverse section

of the mixed part of the
prelaminar region of the
and Vrabec [8])
Fig.2.18 Edema of the optic

nerve head in a rhesus
monkey due to the raised
intracranial pressure
(Hortegas stain) (Reproduced
from Hayreh and Vrabec [8])
The lamina cribrosa is usually convex posteriorly The Connective Tissue Part
and concave anteriorly (Figs.2.12.4 and 2.10). There
is no sharp transition between the anterior prelaminar This extends transversely across the entire thickness of
glial and posterior connective tissue parts, resulting in the optic nerve head, bridging the scleral canal
the transitional zone (mixed part) of varying size (Figs.2.22.4, 2.12 and 2.202.24). At the periphery,
between the two. the connective tissue part of the lamina cribrosa is
Fig.2.19 High-power view

of Fig.2.18 in the region
of the glial part of the
prelaminar region (Hortegas
section of the anterior part of
the optic nerve in rhesus
monkey showing a thick
horizontal fibrous band of
lamina cribrosa (LCR) and
the connective tissue sheath
of the central retinal vessels
centrally (*), which reaches
the bottom of the optic disc
cup. (Hortegas stain) LCR
Lamina cribrosa, PLR
Prelaminar region, RLR
Retrolaminar region
anchored to the surrounding sclera by thick columns of c entrally (Figs.2.20, 2.21 and 2.242.26), binding the
the connective tissue with enlarged bases (Figs.2.25 lamina cribrosa firmly to the connective tissue enve-
and 2.26); the same kind of anchorage is present lope of the central retinal vessels.
Fig.2.21 Optic nerve head

with a deep excavation (Cup).
(Hortegas stain) LCR Lamina
cribrosa, PLR Prelaminar
region, RLR Retrolaminar
region. *Connective tissue
around central retinal vessels
and Vrabec [8])
monkey, specifically showing
anchorage of the connective
tissue strands of the lamina
cribrosa to the sclera at the
periphery and to the
connective tissue sheath of
the central retinal vessels
centrally (*). (Hortegas
stain) LCR Lamina cribrosa
region, PLR Prelaminar region,
RPR Retrolaminar region.
and Vrabec [8])
The connective tissue fibers of the lamina cribrosa Wolff [17] and Hogan etal. [18] found a large amount
are tightly packed together so that, in a longitudinal of elastic tissue in the lamina cribrosa, but Anderson
section through this region, it is seen as a dense com- [15] found it to vary greatly from one eye to another.
pact band of the connective tissue bridging the scleral The posterior part of this is not sharply delimited from
canal (Figs.2.2, 2.4 and 2.202.24). In cross sections the septal system of the retrolaminar optic nerve, so
of the lamina cribrosa, this compact connective tissue that the connective tissue septa of the latter are attached
reveals many openings for the transmission of the optic to the posterior surface of the lamina cribrosa (Figs.2.4,
nerve fiber bundles (Figs.2.252.29). The lamina crib- and 2.202.24). Numerous vessels arising from the
rosa is of a lamellar nature with collagen bundles alter- circle of Haller and Zinn or paraoptic short poste-
nating with glial sheets [15, 17]; posteriorly the rior ciliary arteries (see Chap. 3) penetrate into the
collagen tissue sheets become increasingly prominent. optic nerve and large capillaries are seen within the
monkey showing transitional
strands between the
connective-tissue part
of the lamina cribrosa
and the retrolaminar
connective tissue. LCR
Lamina cribrosa region, PLR
Prelaminar region, RPR
Retrolaminar region
(Hortegas stain) (Reproduced
from Hayreh and Vrabec [8])
monkey showing anchoring
connective-tissue fibers
running from the central
connective-tissue strand
(surrounding the central
retinal vessels *) to the
bottom of the optic disc cup
(Hortegas stain)
connective tissue trabeculae. A continuous glial mem- connective tissue and glial cell structural elements
brane, similar to that seen in the retrolaminar part, were greater in the nasal-temporal region compared
separates the nerve fibers from the connective tissue in with the inferior and superior quadrants [22]. The
the region of the lamina cribrosa. openings are mostly oval or round and their diameter
It has been shown that there are regional differ- varies greatly (Figs. 2.252.29). Some of the larger
ences in the fine structure of the lamina cribrosa, openings of the lamina cribrosa are subdivided by the
with superior and inferior parts containing larger connective tissue. All the openings in the lamina cri-
pores and thinner connective tissue septa than in brosa are not only lined by astrocytes but are further
the nasal and temporal parts of the lamina cribrosa subdivided by glial trabeculae so that a large number
[21]. Another histological study showed that the of the glial fibers are seen crossing the opening. The
Fig.2.25 Transverse
section of the lamina
cribrosa, showing coarse
connective-tissue septa,
rounded openings for the
nerve bundles and, in the
periphery, glial fibers
expanded across the
openings. (A) Central retinal
artery, (V) Central retinal vein
(Hortegas stain)
and Vrabec [8])
Fig.2.26 Transverse section of the lamina

cribrosa, showing coarse connective-tissue
septa, rounded openings for the nerve
bundles and, in the periphery, glial fibers
expanded across the openings. (A) Central
retinal artery, (V) Central retinal vein.
(Hortegas stain.) (Reproduced from
glial tissue forms a continuous glial membrane Hernandez etal. [23], on immunofluorescent staining,
surrounding each nerve fiber bundle, as in the prelam- found that the lamina cribrosa consists of elastin fiber and
inar region; thus it separates the nerve fiber bundles collagen III and IV and laminin; at the insertion of the
from the adjacent connective tissue. In the middle lamina cribrosa in the sclera there are concentric circum-
of the opening many typical fibrous astrocytes can ferential elastin fibers surrounding the lamina cribrosa
beseen (Figs.2.28 and 2.29). As the myelin sheath is and astrocytic processes extend into the bundles of elastin
lacking here (Figs.2.5 and 2.6), the openings of the fibers. They [24] found that there was an age-related
lamina cribrosa are considerably smaller than the increase in apparent density of collagen types I and III
interseptal space in the retrolaminar part. The coarse and elastin and increase in density of collagen type IV.
connective tissue trabeculae (Figs.2.252.29), show The lamina cribrosa, throughout its entire thick-
mostly oval or round openings of very variable ness, is pierced centrally by the central retinal vessels
diameter as compared with the polygonal large spaces with their accompanying connective tissue (Figs.2.3,
of the retrolaminar septal system. 2.12, 2.202.22 and 2.242.26). The latter tissue forms

of the lamina cribrosa region
of the optic nerve head
showing the arrangement of
the septa (Reproduced from
Fig.2.28 Transverse
cribrosa, showing fibrous
astrocytes in the openings of
the lamina cribrosa
(Hortegas stain)
a cylindrical sheath surrounding the vessels, within Elschnig, much more strongly developed on the tem-
which numerous fine nerve fibers (autonomic) are seen poral than on the nasal side, separates the sclera from
running along the central vessels to the optic disc [8]. the nerve fibers and is composed of dense collagenous
Some of these are closely attached to the vessel wall. tissue with many glial and elastic fibers and some pig-
In our study [8], in some sections we were able to fol- ment [8, 25]. It continues forward to separate the chor-
low larger trunks of such nerve fibers, representing the oid from the prelaminar region. The glial framework,
so-called nerve of Tiedemann (Fig.2.30). formed by the astrocytes, extends throughout the entire
The nerve fibers in this region are similar to those optic nerve head and seems to account for more than
seen in the prelaminar region. The border tissue of half of the volume of the optic nerve head [15].
Intraorbital Part of the Optic Nerve 21
Fig.2.29 Transverse
cribrosa, showing fibrous
astrocytes in the openings of
the lamina cribrosa
(Hortegas stain)
Fig.2.30 Longitudinal section of the optic

nerve in rhesus monkey showing nerve of
Tiedemann (arrow) running parallel with
the wall of the central retinal artery (CRA)
in the optic nerve. (Gros-Schultzes stain)
(Reproduced from Hayreh and Vrabec [8])
Intraorbital Part of the Optic Nerve in the subarachnoid space (Figs.2.31 and 2.32). The
pia mater is closely related to the optic nerve.
This extends from the eyeball to the optic canal. The This part of the optic nerve has coarse connective-
diameter of the optic nerve in the retrolaminar region tissue septa (Figs. 2.33 and 2.34), containing blood
is about twice that of the optic disc (i.e., 34mm); this vessels. They run in all directions longitudinal and
is because the nerve fibers are myelinated in the orbital transverse septa in the optic nerve are all attached to
part and rest of optic nerve, but they are unmyelinated one another (Figs.2.3, 2.4, 2.202.24, 2.35 and 2.36).
in the lamina cribrosa and anterior to that (Figs. 2.5 In the central part of the optic nerve, near the central
and 2.6). This part of the optic nerve is surrounded by vessels, the main direction of the connective-tissue
the meningeal sheath which consists of the dura mater, fibers is longitudinal (Fig.2.37). These can easily be
arachnoid mater and pia mater, and cerebrospinal fluid differentiated from the glial fibers because of their
Fig.2.31 A diagrammatic C S
R
copy of a longitudinal section
of a normal human optic
nerve, showing the optic
nerve sheath in its different
parts. A arachnoid, C choroid,
D dura, OC Optic canal, ON
optic nerve, P Pia, R retina,
ON
S sclera (Reproduced from
Hayreh [26])
D P A
OC
Brain
EB
Fig.2.32 Schematic
diagram, showing various A
D
regions of the optic nerve
sheath. A arachnoid, D dura,
EB Eyeball, OC Optic canal, ON
ON optic nerve (Reproduced
from Hayreh [26]) OC
coarse and sinuous character. The longitudinal fibrous blood vessels by an astroglial layer, throughout the
septa of the retrolaminar part of the optic nerve are course of the optic nerve. The supporting glial frame-
firmly anchored to the posterior surface of the lamina work is built up by fibrous astrocytes lying within the
cribrosa (Figs. 2.4 and 2.202.24). The transverse nerve fiber bundles. The astrocytes are connected to
septa at the periphery are attached to the pia on the the connective tissue septa (Fig.2.38), and their capil-
surface, and to the fibrous envelope around the central laries by well-marked footplates. The course of all
retinal vessels centrally (Figs. 2.33 and 2.34). The their fibers is best seen in specimens stained by Golgis
septa form a rather complicated intercommunicating methods (Fig.2.39).
polygonal tubular framework, and enclose within them Within the nerve fiber bundles are situated longitu
the nerve fiber bundles with the accompanying glia dinal rows of angular elements of the oligodendroglia,
(Fig.2.34). The large connective tissue tubes are often with densely impregnated cytoplasm and short proto-
bridged by fine fibrous tissue bands (Fig.2.36). plasmic processes (Figs.2.402.42). Scattered among
As in the rest of the central nervous system, at the the nerve fibers are the thin, branched microglial cells,
neuroectodermal-mesodermal junction the nerve fibers with typical nuclei and processes (Figs.2.41 and 2.42)
are always separated from the collagenous tissue and they constitute a part of the reticuloendothelial
Intraorbital Part of the Optic Nerve 23
Fig. 2.33 Transverse section of a human retrolaminar optic

nerve, showing the arrangement of the septa. A Central retinal
artery, V Central retinal vein
Fig.2.34 Transverse section of the retrolaminar optic nerve in
rhesus monkey, showing the arrangement of the septa. A Central
retinal artery (Reproduced from Hayreh and Vrabec [8])
system. Impregnation of oligodendroglia, however, is
very difficult and requires practically intravital fixation
of the material. That is perhaps the reason why in the In the center of the anterior intraorbital part of the
normal human material most of the authors were not optic nerve lie the central retinal vessels enclosed in
able to stain these cells successfully [11, 12]. In our their envelope (Figs.2.33, 2.34, 2.44 and 2.45). In the
study [8] of rhesus monkeys, we did the intravital fixa- center of the posterior part lie some inconstant vessels
tion of the tissues. The rows of oligodendrocytes (see Chap. 3). The pia mater has a dense vascular plexus
(Fig. 2.43), as well as the myelin sheaths, stopped on the surface of the orbital optic nerve, formed by
shortly behind the lamina cribrosa. We did not find any branches from the various branches of the ophthalmic
typical oligodendroglia anterior to the retrolaminar artery (Fig. 2.46); this plexus forms the centripetal
region. Oligodendroglia are responsible for the forma- vascular system of blood supply to the optic nerve (see
tion of the myelin sheaths. Chap. 3). The dura mater also has vessels on it but much
The nerve fibers are arranged in bundles (Figs.2.1 less than on the pia mater. Most of the dural vessels
2.4 and 2.202.24), with frequent interchanging of penetrate the dura to supply the pial plexus (Fig.2.44).
some fibers between the neighboring bundles, thus The ophthalmic artery lies in close relationship to
forming a sort of network. The myelin sheaths reach this part of the optic nerve the first part of the
almost to the posterior part of the lamina cribrosa. ophthalmic artery is almost adherent to the inferolat-
Within the nerve fiber bundles are situated longitudinal eral part of the optic nerve (Fig.2.47), its second part
rows of angular elements of the oligodendroglia, with crosses over (in 83%) or under (17%) the optic nerve
densely impregnated cytoplasm and short protoplasmic in close relationship to the sheath, but the third part
processes (Fig.2.43). runs in a superior and medial direction away from the
Fig.2.36 A high-power view of a part of Fig. 2.35, showing

fine connective-tissue strands bridging over a large connective-
tissue tube (Reproduced from Hayreh and Vrabec [8])
Fig.2.35 Longitudinal section of the retrolaminar part of the

optic nerve in rhesus monkey, showing coarse connective-tissue rabbits [32]. The pattern was similar in the human and
septa. (Hortegas silver carbonate) LCR Lamina cribrosa, PLR monkeys. The sheath (i.e. dura mater and arachnoid
Prelaminar region, RLR Retrolaminar region (Reproduced from mater) normally is loose near the eyeball, with a much
bigger subarachnoid space between the optic nerve
and the sheath than elsewhere in its course, conse-
optic nerve [30] (Fig.2.48). The central retinal artery quently presenting a bulbous appearance just behind
and vein, after their entry/exit from the optic nerve the eyeball [32] (Figs.2.31 and 2.32). The space ends
(515mm, median 10mm behind the eyeball [31]) lie blindly at the junction of the sheath and the eyeball.
in close relationship or adherent to the inferior aspect The subarachnoid space between the optic nerve and
of the sheath (Fig.2.49). Near the eyeball, the nerve is the sheath is narrowest in the region of the optic canal.
surrounded by multiple short posterior ciliary arteries In the optic canal, the optic nerve is attached to the
and ciliary nerves (see Chap. 3). surrounding dura by thick fibrous bands (Figs. 2.32
and 2.50), which stretch from the dura to the pia, with
the arachnoid interrupted at the sites of the bands but
continuous in-between. The amount of these fibrous
Sheath of the Optic Nerve adhesions shows marked interindividual variation. The
characteristics of the optic nerve sheath in rabbits
I investigated the anatomy of the sheath of the optic differed from the human and monkeys in that there
nerve and its communication with the cranial cavity in were neither similar adhesions in the optic canal nor
80 human and 20 rhesus monkey optic nerves, and in looseness of the sheath behind the eyeball.
Sheath of the Optic Nerve 25
Fig.2.37 Longitudinal connective-tissue fibers near the central

retinal vessels. Fine glial fibers are seen running perpendicular
to the connective-tissue septa (Hortegas silver carbonate)
(Reproduced from Hayreh and Vrabec [8])
Fig.2.38 Transverse section of the optic nerve, showing fibrous

astrocytes and their processes within the nerve-fiber bundles
At the apex of the orbit, at the orbital opening of (Hortegas silver carbonate) (Reproduced from Hayreh and
Vrabec [8])
the optic canal, the optic nerve sheath is attached and
surrounded by the annular tendon to which are
attached the various recti muscles. In the optic canal
region, the two layers of the dura mater are joined
together but at the orbital end of the canal they split man, because the upper bony margin of its cranial
the outer layer forms the periosteum of the orbital cavity is prolonged by 0.50.6mm into the cranial by
bones while the inner layer forms the dural sheath of a falciform fold of dura [33] (Fig.2.51). Unlike the
the optic nerve (Fig. 2.32). In the optic canal, the orbital part of the sheath, the optic nerve in the optic
ophthalmic artery is closely related to the optic nerve, canal, as discussed above, is firmly bound down to
and lies partly in the subdural space and partly the dura by numerous thick fibrous bands which con-
between the two layers of the dura mater, usually the nect the dura to the pia (Figs.2.32 and 2.50). These
inferior and lateral aspect of the optic nerve [33] bands not only firmly hold the optic nerve in position
(Fig.2.47). in this region but also hold the dura and the optic
The anatomy of the sheath of the optic nerve in the nerve close to one another. In this region, the suba-
region of the optic canal deserves special consider- rachnoid space is reduced to almost a capillary size
ation because it seems to have great importance in (Fig. 2.50 SAS), which is interrupted by these
more than one way. The length of the bony optic canal bands. Therefore, the space assumes the character of
is smaller than the canal in situ in the normal body in a trabecular meshwork of closely knit fibers in the
Fig.2.39 Fibrous astrocytes

and their processes in the
optic nerve (Golgis stain)
and Vrabec [8])
stretching and distension of the sheath due to raised

cerebrospinal fluid pressure in it; however that is not
possible because the dura, being firm collagen fibers
(not elastic fibers), is incapable of distension with
any amount of raised cerebrospinal fluid seen in
patients. In my studies of orbital surgery in about
400 monkeys for experimental occlusion of various
ocular vessels, distension of the sheath of the optic
nerve, filled with cerebrospinal fluid, was always
present normally. The anatomy of the sheath in the
region of the optic canal plays a crucial role in the
dynamics of conveying the cerebrospinal fluid
pressure of the cranial cavity into the sheath of the
optic nerve. Communication between the subarach-
noid spaces of the cranial cavity and the optic nerve
sheath is almost always seen; however, the extent of
communication in the optic canal shows wide interin-
dividual variation [32]. To reach the orbital part of
the sheath, the cerebrospinal fluid has to percolate
Fig. 2.40 An oligodendrocyte (Penfields stain) (Reproduced
through the capillary meshed trabecular network
from Hayreh and Vrabec [8]) formed by the adhesions in the optic canal. The extent
of those adhesions determines the speed with which
the cerebrospinal fluid pressure can be transmitted
canal. It must, however, be added that in some cases from the cranial cavity to the orbital part of the sheath.
only scanty adhesions were seen all round the optic It is well-established now that it is the raised cerebro-
nerve in this region. spinal fluid pressure in the orbital part of the sheath
The normal bulbous appearance of the sheath has that plays a role in the development of optic disc
often erroneously been assumed to be caused by edema in raised intracranial pressure [26, 34, 35].
Intracranial Part of the Optic Nerve 27
Fig.2.41 A group of oligodendrocytes

together with a microglial element (arrow)
(Penfields stain) (Reproduced from Hayreh
and Vrabec [8])
Fig.2.42 Oligodendrocytes
and microglia (Penfields
Intracanalicular Part of the Optic Nerve Intracranial Part of the Optic Nerve
This part lies in the bony optic canal (Fig.2.32), sur- In this part, the optic nerve lies above the diaphragma
rounded by the meningeal sheath (see above). The sellae at first and then above the cavernous sinus, and is
basic structure of this part of the optic nerve is almost in close relationship to the ophthalmic artery infero
the same as that of the adjacent intraorbital part. laterally, internal carotid artery laterally (Fig.2.51) and
canal. The ophthalmic artery is always attached to the

under surface of the optic nerve by a loose meshwork
of vascular connective tissue [33]. Unfortunately, this
part of the optic nerve being a no-mans land between
neurologists and ophthalmologists, little information
about its structure is available.
Optic Nerve Fibers
There are about one million optic nerve fibers. They

are of 5 types: (1) visual afferent (serving visual func-
tion and going to the lateral geniculate body), (2)
pupillary afferent (serving the pupillary reflex and
going to the tectum), (3) efferent to the retina (unknown
function), (4) photostatic (running to the superior col-
liculus) and (5) autonomic fibers [1].
The nerve fibers vary in diameter (between 0.7 and
10mm, mostly 1mm [18]). According to Polyak [36]
the smaller axons come from the midget ganglion cells
of the central retina and the larger axons from the gan-
glion cells from the peripheral retina. In the prelaminar
region, the nerve fibers are unmyelinated (Figs.2.5 and
2.6) and arranged in bundles (Figs. 2.12.6). Toward
Fig.2.43 Oligodendrocytes (arrow) in the retrolaminar portion the retina the bundles become closely packed. In the
of the optic nerve (Penfields stain) (Reproduced from Hayreh lamina cribrosa region the nerve fiber bundles lie in its
and Vrabec [8])
pores and are unmyelinated (Figs.2.5 and 2.6). In the
retrolaminar and orbital parts of the optic nerve, the
anterior cerebral artery superiorly. The nerve is nerve fibers are arranged in bundles (8001,200 bun-
covered by the pia mater and there is no meningeal dles [18]). They lie in polygonal spaces formed by the
sheath comparable to that in the orbit and the optic fibrovascular septa in the orbital part of the optic nerve.
C
R S Col. Br.
PCA
D
A
Pia
Fig.2.44 Schematic
A arachnoid, C choroid, CRA ON
central retinal artery, Col. Br.
PR
collateral branches, CRV
central retinal vein, D dura, LC
LC lamina cribrosa, ON optic
SAS
arteries, PR prelaminar
region, R retina, S sclera, SAS CRV
PCA CRA
subarachnoid space
Optic Nerve Fibers 29

of the central part of the
retrolaminar part of the optic
nerve, showing central
retinal artery (CRA) and vein
(CRV) enclosed by a
common fibrous tissue
envelope (FTE) (Masons
trichrome staining)
etal. [27])
The nerve fiber bundles lie in the complicated inter-

communicating tubular framework formed by the septa,
accompanied by glia. There is a frequent interchange
of fibers between the neighboring bundles, thus form-
ing a sort of network. Within the bundles, the nerve
fibers are insulated from their neighbors by neuroglia.
The myelin sheaths reach almost to the posterior part of
the lamina cribrosa. Within the nerve fiber bundles are
rows of supporting astrocytes, oligodendrocytes and
some microglial cells. A continuous glial membrane
formed by the astrocytes separates the nerve fibers from
the connective tissue in the lamina cribrosa and the ret-
rolaminar region.
opographic Localization of the Nerve

T
Fibers from the Various Parts of the
Retina in the Optic Nerve
This subject is of tremendous significance to any

discussion and understanding of the type, location and
pathogenesis of visual field defect in optic neuropathies.
In the retina there is no functional overlap between
the superior and inferior temporal halves and the two
Fig.2.46 Inferior surface of the human optic nerve (intraorbital
part) after removal of the sheath, showing pial vessels arising meet at a sharp horizontal temporal raphe, situated
from different sources. C.A.R. Central artery of the retina, Col. between the temporal periphery and the macula
Br. Of O.A. Collateral branches of ophthalmic artery, CZ Circle (Fig. 2.52). One could compare that to unfolding a
of Zinn and Haller, Pial br. of C.A.R. Pial branches of central
Japanese fan, with its central axis corresponding to the
artery of the retina, Rec. pial br. of CZ Recurrent branches of
Circle of Zinn and Haller (Reproduced from Singh (Hayreh) and optic disc and the two free borders meeting in the
Dass [28]) temporal part. The nerve fibers diverge from the raphe
Fig.2.47 Intraorbital course III

of the human ophthalmic OA Bend
artery, as seen from the
lateral side of the optic nerve.
OA Ophthalmic artery, OC
Optic canal, ON Optic nerve, ON
I,II and II Three parts of the
intraorbital part of the
ophthalmic artery
(Reproduced from OC
Hayreh[29]) II I
Angle
17.4% 82.6%
ON
III
Bend
II
Angle
OA
ICA
Fig.2.48 Course of the human ophthalmic artery left figure

Fig.2.49 Origin and intraorbital course of the human central
when it crosses under the optic nerve (in 17.4%) and right figure
retinal artery as seen from below. CAR Central artery of the
when it crosses over the optic nerve (in 82.6%). ICA Internal
retina, LPCA Lateral posterior ciliary artery, OA Ophthalmic
carotid artery, OA Ophthalmic artery, ON Optic nerve, I,II and II
artery, PPS Point of penetration of the sheath by the central
Three parts of the intraorbital part of the ophthalmic artery
artery of the retina (Reproduced from Hayreh and Dass [31])
to go to the optic disc. This arrangement is most easily

understood if the raphe is looked upon as a fold in the the peripheral retina maintain a deep position in the
temporal margin of the retina drawn in toward the nerve fiber layer, while those which arise closer to the
macula with its edges falling into apposition [38]. This optic disc course forward through the nerve fiber layer
gave an erroneous impression to some authors that the to lie more superficially in the nerve fiber layer
optic nerve fibers in the optic nerve head had a macu- (Fig. 2.53). Thus, the longer and more peripheral
locentric arrangement. In fact the fibers are centered at nerve fibers lie deeper in the nerve fiber layer in
the optic disc. comparison with the shorter fibers, which arise nearer
In the retina the nerve fibers are arranged in layers, the disc [3941]. When these fibers reach the optic
so that the fibers which arise from the ganglion cells in nerve head, the more peripheral fibers come to lie in
Optic Nerve Fibers 31
section of the optic nerve in
the region of the optic canal,
showing a capillary
subarachnoid space (SAS)
and fibrous bands connecting
the optic nerve with the
surrounding sheath (Masons
trichrome staining)
Fig.2.52 Schematic representation of the optic nerve fibers in

the nerve fiber layer of the retina: 1 and 2 from the superior and
inferior temporal quadrants of the retina respectively, 4 and 5
Fig.2.51 Human ophthalmic (O.A.) and internal carotid (I.C.A.)
from the superior and inferior nasal quadrants of the retina
arteries and optic nerve (O.N.), as seen intracranially, with the
respectively (Reproduced from Hayreh [37])
intracranial end of the optic canal and optic nerve in it
(Reproduced from Hayreh [33])
peripheral part of the nerve head, while the fibers aris- (Fig. 2.53). However, autoradiographic studies of the
ing from the retinal ganglion cells near the optic nerve arcuate fibers in the retina in rhesus monkeys revealed
head pass across the surface of the optic disc to leave that the nerve fibers intermingle freely along their
the eyeball near the center of the nerve head [36, 4147] intraretinal course; the segregation of the nerve fibers
Fig.2.53 Schematic
representation of arrange- Nerve
ment of optic nerve fibers in fiber
the nerve fiber layer of the layer
retina and the optic nerve
head (Reproduced from
Hayreh [37]) Ganglion cells
Optic nerve head
takes place at or immediately after the fibers cross the 1. Lesions in the Upper and Lower Temporal Sectors
edge of the disc, so the fibers from the far periphery of of the Retina Temporal to the Fovea
the retina come to lie in the outer part of the disc, while This resulted in interruption of most of the peripheral
those from nearer the disc go to the center of the nerve and all paracentral axons (sparing the macula) from
head [48]. In the optic nerve the fibers rearrange them- the upper and lower temporal retinal quadrants [50].
selves as they travel posteriorly in the nerve. This caused degeneration of the optic nerve fibers in
The studies by Hoyt and colleagues [4952] have upper and lower temporal sectors of the optic nerve up
contributed significantly to our present concept of to 2 to 3mm behind the eyeball (Fig.2.54) and these
the topographic localization of the nerve fibers from sectors were joined behind that. The two groups of
the various parts of the retina in the anterior part of the fibers were separated by a temporal horizontal area
optic nerve. They investigated the topographic local- and they maintained their relative position through the
ization of the optic nerve fibers in various parts of the optic nerve. They extended to the central retinal ves-
optic nerve in monkeys by producing lesions in the sels and became more crescentic in the middle and
retinal nerve fibers by photocoagulating different proximal thirds of the optic nerve. The apices of the
regions of the retina and histologically localizing the two sectors did not reach as far as the center of the
nerve fiber degeneration in different part of the optic optic nerve (Fig.2.54). There was incomplete sector-
nerve. They found the following. shaped degeneration in the superior and inferior tem-
poral part of the optic nerve in the distal one third of
the nerve. The sectors receded farther away from the
Superior center of the optic nerve, as the chiasma was
approached. Both the upper and lower sectors extended
1 slightly into the nasal half of the optic nerve.
Superior In another study, Hoyt [49] produced photocoagula-
4
temporal tion lesions in monkeys, in the retina at or near the
Superior optic disc, to investigate the location of arcuate nerve
nasal fibers bundles in the optic nerve. This showed that
Macular Temporal arcuate bundles have a stable and predictable arrange-
Nasal
3 ment within the optic nerve. These fibers are located in
Inferior
nasal the temporal half of the optic nerve, corresponding to
Inferior their site in the retina. In the anterior part of the optic
5 temporal
nerve, the apex of the sector extends to the central core
2
of the nerve adjacent to the central retinal vessels. In
the middle and posterior parts of the nerve, the apices
Inferior
of the sectors are blunted and compressed toward the
outer portion of the nerve. The density of the sector
Fig.2.54 Schematic representation of the arrangement of optic varies directly with the nearness of retinal lesion to the
nerve fibers in the optic nerve head, based on the studies of Hoyt optic disc and its width varies with the number of arcu-
and Luis [50]: 1 and 2 from the superior and inferior temporal
quadrants of the retina respectively, 4 and 5 from the superior
ate bundles cut. The closer the lesion to the optic disc,
and inferior nasal quadrants of the retina respectively, 3 from the the closer the apex of the degenerating sector
macular region (Reproduced from Hayreh [37]) approaches the core of the nerve.
References 33
2. Lesions in the Retina Nasal to the Optic Disc A study by Hoyt and Kommerell [52] in patients
This resulted in very discrete sector degeneration of with homonymous hemianopia showed that near the
nerve fibers in the distal portions of the optic nerve vertical poles of the optic disc, the retina has far more
[50]. Both upper and lower quadrant sectors extended nerve fibers coming from the temporal peripheral retina
to the central retinal vessels (Fig. 2.51) and laterally than from the nasal proximal retina. This may be due to
were adjacent to each other along the horizontal merid- the fact that the fibers from the nasal retina have wide-
ian. That sector location persisted throughout the optic spread access to the nasal part of the optic disc, but the
nerve. Retinal lesions in the far nasal periphery pro- fibers arising from the temporal peripheral retina have
duced degeneration only in the nasal periphery of the no similar access to the lateral aspects of the optic disc,
optic nerve. which is already occupied by the macular fibers, so that
3. Lesion in the Macular Region the temporal peripheral fibers have to crowd near the
Hoyt and Luis [50] found that following macular poles to enter the optic nerve head (Fig.2.49).
lesion, the macular fibers were scattered over a tem- The information provided by these studies has tre-
poral sector occupying at least one third of the cross mendous practical implications. Lesions in different
sectional area of the nerve in the most anterior part parts of the optic nerve are likely to produce very
of the optic nerve. In the proximal part of the optic different types of visual field defects. This topographic
nerve, the macular fibers spread out rapidly toward localization of the optic nerve fibers may help in a
the nasal side of the nerve, distributed diffusely better understanding of the visual field defects in vari-
throughout the nerve, and were found peripherally ous types of ischemic optic neuropathy.
as well as centrally. Before reaching the middle third
of the nerve, some macular fibers had already
migrated nasally. Unlike the peripheral axons, which
are closely packed and run directly posteriorly, the References
macular fibers constantly wandered throughout the
nerve and seemed to be crossing through the bundles 1. Duke-Elder S, Wybar KC. System of ophthalmology, The
of the optic nerve. This resulted in widespread anatomy of the visual system, vol. 2. London: Kimpton;
1961. p. 277.
mixing of these fibers throughout the area, except in 2. Briggs G. Ophthalmographia, sive Oculi ejusque partium
the most distal part of the nerve. But, according to descriptio Anatomica. Foan Hayes, Celeberrimae, Cantab.
others, they lie in the central part of the optic nerve 1676. p. 28.
posteriorly [53]. 3. Elschnig A. Der normale Sehnerveneintritt des menschli-
chen. Auges. Denkschriften mathematisch naturwissen-
4. Lesions in the Region between the Fovea and the schaftlichen Cl kaiserl Akad Wiss Wein. 1899;70:
Optic Disc 219303.
The nerve fibers from this area are limited to an area 4. Elschnig A. Normale Anatomie des Sehnerveneintrittes. In:
ofthe optic nerve on the horizontal meridian temporal Magnus H, editor. Unterrichtstafeln1899;16. p. 124.
5. Ishii K. The diameter of the optic disc in the Japanese (In
tothe central retinal vessels, and the nerve fibers are of Japanese). Acta Soc Ophthalmol Jpn. 1951;55:2423.
moderately large size. 6. Kuhnt H. Zur Kenntniss des Sehnerven und der Netzhaut.
5. Lesions in the Region of the Arcuate Nerve Albrecht Von Graefes Arch Ophthalmol. 1879;25(3):
Fibers 179288.
7. von Hippel E. Sind die markhaltigen Nervenfasern der
In this study the lesions were placed in the retina tem- Retina eine angeborene Anomalie? Albrecht Von Graefes
poral to the optic disc in the region occupied by the Arch Ophthalmol. 1900;49:5918.
arcuate nerve fiber bundles. The result was well demar- 8. Hayreh SS, Vrabec F. The structure of the head of the optic
cated sector shaped areas of degeneration in the optic nerve in rhesus monkey. Am J Ophthalmol. 1966;61(1):
13650.
nerve just behind the lamina cribrosa, in the temporal 9. Anderson DR. Ultrastructure of the optic nerve head. Arch
half of the optic nerve. The apices of the sectors in the Ophthalmol. 1970;83(1):6373.
anterior part of the optic nerve extended to the central 10. Roth AM, Foos RY. Surface structure of the optic nerve
core of the nerve adjacent to the central retinal vessels. head. 1. Epipapillary membranes. Am J Ophthalmol.
1972;74(5):97785.
In the middle and posterior parts of the nerve the 11. Wolter JR. Die Struktur der Papille des menschlichen Auges.
apices of the sectors were blunted and compressed [Structure of the papilla of the human eye. ]. Albrecht Von
toward the outer portion of the nerve. Graefes Arch Ophthalmol. 1956;158(3):26876.
12. Wolter JR. The human optic papilla; a demonstration of new 36. Polyak SL. The vertebrate visual system; its origin, struc-
anatomic and pathologic findings. Am J Ophthalmol. ture, and function and its manifestations in disease. Chicago:
1957;44(4, part 2):4865. University of Chicago Press; 1957.
13. Duke-Elder S, Wybar KC. System of ophthalmology, The 37. Hayreh SS. Structure and blood supply of the optic nerve.
anatomy of the visual system, vol. 2. London: Kimpton; In: Heilmann K, Richardson KT, editors. Glaucoma: con-
1961. p. 28693. ception of a disease. Stuttgart: Georg Thieme; 1978.
14. Bonamour GP, Bregeat M, Bonnet P. Juge: La Papille p. 7896.
Optique. Paris: Masson; 1968. p. 8. 164283, 3623, 426. 38. Rnne H. The different types of defects of the field of vision.
15. Anderson DR. Ultrastructure of human a, d monkey lamina J Am Med Assoc. 1927;89(22):18605.
cribrosa and optic nerve head. Arch Ophthalmol. 1969;82(6): 39. Sjaaff M, Zeeman WPC. ber den Faserverlauf in der
80014. Netzhaut und im Sehnerven beim Kaninchen. Albrecht Von
16. Levitzky M, Henkind P. Angioarchitecture of the optic nerve. Graefes Arch Ophthalmol. 1924;114:192211.
II. Lamina cribrosa. Am J Ophthalmol. 1969;68(6):98696. 40. Magitot A. Physiologie oculaire clinique. Paris: Masson;
17. Bron AJ, Tripathi RC, Tripathy BJ. Optic nerve, Section15.1. 1946.
Wolffs anatomy of the eye and orbit. 8th ed. London: 41. Wolff E, Penman GG. The position occupied by the periph-
Chapman & Hall; 1997. p. 489535. eral retinal fibers in the nerve fiber layer and the nerve head.
18. Hogan MJ, Alvarado JA, Weddell JE. Histology of the human Acta XVI Conc Ophthalmol Lond. 1950;1:62535.
eye. Philadelphia: Saunders; 1971. p. 523. 537, 544, 561, 606. 42. Dean G, Usher CH. Experimental research on the course of
19. Jacoby E. ber die Neuroglia des Sehnerven. Klin Mbl the optic fibres. Trans Ophthalmol Soc UK. 1896;16:
Augenheilk. 1905;43:12937. 24876.
20. Cohen AI. Ultrastructural aspects of the human optic nerve. 43. Dean G, Usher CH. Experimental research on the course of
Invest Ophthalmol. 1967;6(3):294308. the optic fibres. Brain. 1903;26:52442.
21. Quigley HA, Addicks EM. Regional differences in the struc- 44. Seidel E. Experimentelle Untersuchung ber die Lage der
ture of the lamina cribrosa and their relation to glaucomatous Versorgungs-gebiete der Nervenfasern des Shenervenstam
optic nerve damage. Arch Ophthalmol. 1981;99(1):13743. mes in der Netzhaut des Menschen. Albrecht Von Graefes
22. Radius RL, Gonzales M. Anatomy of the lamina cribrosa in Arch Ophthalmol. 1919;100:16878.
human eyes. Arch Ophthalmol. 1981;99(12):215962. 45. Loddoni G. Sui decorso delle fibre nervose nei nervo ottico
23. Hernandez MR, Luo XX, Igoe F, Neufeld AH. Extracellular e nella retina. Ann Ottal. 1930;58:46890.
matrix of the human lamina cribrosa. Am J Ophthalmol. 46. Peyman GA, Apple D. Peroxidase diffusion processes in the
1987;104(6):56776. optic nerve. Arch Ophthalmol. 1972;88(6):6504.
24. Hernandez MR, Luo XX, Andrzejewska W, Neufeld AH. 47. Leber T. Die Krankheiten der Netzhaut und des Sehnerven.
Age-related changes in the extracellular matrix of the human III. Die Sehstrungen durch erkrankung der opticuscentren,
optic nerve head. Am J Ophthalmol. 1989;107:47684. der tractus optici und des chiasma. Die Hemianopsie. In:
25. Salzmann M. The anatomy and histology of the human eye- Graefe A, Saemisch T, editors. Handbuch der gesamten
ball in the normal state; its development and senescence. augenheilkunde, vol. 5. Leipzig: Wilhelm Engelmann; 1877.
Chicago: University of Chicago Press; 1912. p. 94. p. 92939.
26. Hayreh SS. Pathogenesis of oedema of the optic disc (papil- 48. Ogden TE. The nerve-fiber layer of the primate retina: an
loedema) A preliminary report. Br J Ophthalmol. 1964; autoradiographic study. Invest Ophthalmol. 1974;13(2):
48:52243. 95100.
27. Hayreh SS et al. Morphologic changes in chronic 49. Hoyt WF. Anatomic considerations of arcuate scotomas
high-pressure experimental glaucoma in rhesus monkeys. associated with lesions of the optic nerve and chiasm.
J Glaucoma. 1999;8:5671. A nauta axon degeneration study in the monkey. Bull Johns
28. Singh (Hayreh) S, Das R. The central artery of the retina. II. Hopkins Hosp. 1962;111:5771.
A study of its distribution and anastomoses. Br J Ophthalmol. 50. Hoyt WF, Luis O. Visual fiber anatomy in the infragenicu-
1960;44:28099. late pathway of the primate. Arch Ophthalmol. 1962;
29. Hayreh SS. Arteries of the orbit in the human being. Br 68:94106.
J Surg. 1963;50:93853. 51. Hoyt WF, Tudor RC. The course of parapapillary temporal
30. Singh (Hayreh) S, Das R. The ophthalmic artery II. Intra- retinal axons through the anterior optic nerve. A Nauta
orbital course. Br J Ophthalmol. 1962;46:16585. degeneration study in the primate. Arch Ophthalmol. 1963;
31. Singh (Hayreh) S, Das R. The central artery of the retina I. 69:5037.
Origin and course. Br J Ophthalmol. 1960;44:193212. 52. Hoyt WF, Kommerell G. Der Fundus oculi bei homonymer
32. Hayreh SS. The sheath of the optic nerve. Ophthalmologica. Hemianopie. (Fundus oculi in homonymous hemianopia).
1984;189:5463. Klin Monatsbl Augenheilkd. 1973;162(4):45664.
33. Singh (Hayreh) S, Das R. The ophthalmic artery I. Origin 53. Duke-Elder S, Wybar KC. System of ophthalmology, vol. 2.
and intra-cranial and intra-canalicular course. Br J London: Kimpton; 1961. p. 6435.
Ophthalmol. 1962;46:6598.
34. Hayreh SS. Pathogenesis of oedema of the optic disc (papil-
loedema) [Thesis for PhD]. London: London University;
1965. For a complete bibliography and detailed review of previously
35. Hayreh SS. Pathogenesis of oedema of the optic disc. Doc published studies in the literature on the subject, please refer to
Ophthalmol. 1968;24:289411. bibliography in previous publications listed here.
Blood Supply of the Optic Nerve
3
Since ischemic optic neuropathies are vascular in into a topic of great interest for researchers working in
origin, it is of paramount importance to have an in- both basic and clinical fields, and an enormous amount
depth understanding of the circulation of the optic of literature has accumulated, generating light on some
nerve, to understand their pathogenesis and manage- aspects and a great deal of heat on others.
ment. This is particularly necessary due to the highly I developed an interest in the field of optic nerve
complex nature of the invivo circulation of the optic circulation in 1955, prompted by the two papers pub-
nerve head (ONH), which plays a decisive role in the lished at that time by Franois and colleagues [1, 2],
development of anterior ischemic optic neuropathy claiming the discovery of a central artery of the optic
(AION). It is the lack of this information which is nerve in man, which (according to them) supplied the
responsible for the current controversy and confusion ONH (Fig. 3.1). Since then, I have investigated the
on various aspects of ischemic optic neuropathies. blood supply and circulation of the optic nerve using a
Historically, it is interesting to note that Franois variety of techniques, including:
and Neetens in 1954 [1] characterized the subject of the
blood supply of the ONH as of only moderate interest 1. Anatomical Studies
to research workers. However, at that time ischemic In these studies the vascular pattern was studied
optic neuropathy as a distinct clinical entity was after postmortem intra-arterial injection of liquid
unknown. As our knowledge of the vital importance of neoprene latex in 100 human and 8 rhesus monkey
the ONH circulation has increased, it has transformed specimens [38].
Fig.3.1 Central artery of the

optic nerve, described by
Francois and Neetens [1].
CAON Central artery of the
optic nerve, CRA Central CRA
retinal artery (Modified from CAON
Fig.3.9 by Francois and
Neetens [1])

36 3 Blood Supply of the Optic Nerve
2. Histological Studies the chapter dealing with the structure of the optic nerve
Serial sections (with and without reconstructions) of (see Chap. 2). Following are the blood supply patterns
the anterior part of the optic nerve were studied in 12 for the four parts.
human and 8 rhesus monkeys [38].
1. The Surface Nerve Fiber Layer
3. Experimental Studies
This most anterior part of the ONH (surface of the
In over 400 rhesus monkey eyes, I studied the subject
optic disc) is typically supplied by the retinal arteri-
using fluorescein fundus angiography following (1)
oles (Figs.3.2 and 3.3). In some cases, however, its
experimental occlusion of main/short/long posterior
temporal region may instead be supplied by the poste-
ciliary arteries (PCAs), vortex veins or central retinal
rior ciliary artery (PCA) circulation from the deeper
vessels, (2) after raising intraocular pressure or alter-
prelaminar region. The cilioretinal artery (rarely a
ing perfusion pressure, (3) after production of malig-
tiny cilio-papillary artery), when present, usually sup-
nant arterial hypertension, and (4) after other relevant
plies the corresponding sector of the surface layer
studies [340].
(Figs.3.33.5).
4. Clinical Studies
2. The Prelaminar Region
Since 1964, when fluorescein fundus angiography was
This region lies between the surface nerve fiber layer and
first used, I have studied the ONH circulation in a large
the lamina cribrosa (Figs.3.2 and 3.3). It is supplied by
population of normal persons, in over 1,000 patients
the fine centripetal branches from the peripapillary
with AION [12, 13, 1821], in over 200 patients with
choroid, which gives out branches to the corresponding
glaucomatous optic neuropathy [9, 24, 25, 41], and in
part of the prelaminar region, as shown by a large num-
many other ONH and ocular disorders.
ber of fluorescein angiographic (Figs. 3.63.11) and
The pattern of blood supply of the optic nerve morphological studies, the former providing the most
Idescribe below is based on the cumulative results of convincing information [913, 1721, 2528, 32, 4262].
my multifaceted studies, each of which has provided However, a few papers based purely on postmortem
information on different aspects of the blood supply,
and has been confirmed by many other studies on the
subject.
Based on the blood supply of the optic nerve, for
descriptive purposes, the entire length of the optic
NFL
nerve can be divided into the following parts:
R
A. Optic nerve head.
C RA
B. Intraorbital part.
C. Intracanalicular part.
PLR
D. Intracranial part. S
LC
Blood Supply of the Optic Nerve Head
This is an extremely important, complex and contro-

Branch,
versial subject, and a comprehensive understanding of CRA
it is crucial to a good grasp of the pathogenesis of
PCA P ON
AION.
Based on the blood supply, the ONH consists of the Fig.3.2 Schematic representation of blood supply of the optic
following 4 regions from anterior to posterior aspects: nerve head. C choroid, CRA central retinal artery, LC lamina
cribrosa, NFLsurface nerve fiber layer of the disc, ON optic
(1) the surface nerve fiber layer, (2) prelaminar region,
nerve, P pia, PCA posterior ciliary artery, PLR prelaminar
(3) lamina cribrosa region and (4) retrolaminar region. region, R retina, RA retinal arteriole, S sclera, (Reproduced from
This classification is different from that described in Hayreh [44])
Blood Supply of the Optic Nerve Head 37
CZ
Retinal vein C
Retinal artery R S Col. Br.
PCA
D
A
Pia
OD ON
PR
LC
SAS
CRV
PCA CRA
Cilio-retinal artery
Fig.3.3 Schematic representation of blood supply of the optic ON optic nerve, PCA posterior ciliary artery, PR prelaminar
nerve. A arachnoid, C choroid, CRA central retinal artery, Col. region, R retina, S sclera, SAS subarachnoid space (Modified
Br. Collateral branches, CRV central retinal vein, CZ circle of from Hayreh [145])
Zinn and Haller, D dura, LC lamina cribrosa, OD optic disc,
a b
Fig. 3.4 Fluorescein fundus angiograms of a rhesus monkey pigment at disc margin. (b) During the retinal arteriovenous
eye after experimental occlusion of the lateral PCA. (a) During phase: The retinal capillary bed has filled completely except in
the early retinal arterial phase: There is no filling of the tempo- the area of supply by a cilioretinal artery which did not fill
ral half of the choroid and optic disc, but normal filling of the because it was a branch of the lateral PCA. Note that the retinal
medial PCA, central retinal artery and nasal part of prelaminar capillary filling has completely masked the underlying infor
region; also the superficial retinal capillaries fill in the inferior mation about the optic disc and choroid, except in the area
temporal part of the optic disc. Note: vessels in the nasal part wherethe retinal capillaries have not filled; that shows absence
of the prelaminar region fill from the adjacent peripapillary of filling of the choroid and optic disc (Reproduced from
choroid, and the dark nasal crescent corresponds to superficial Hayreh [11])
morphological/histological studies have claimed that this artery play no role in its supply. My clinical and experi-
is not so [6267]; this is because of limitations of those mental fluorescein angiographic studies have shown a
methods to provide that information. The peripapillary sectoral blood supply in this region (Figs.3.73.10, also see
choriocapillaris (Figs.3.12 and 3.13) and central retinal many more angiograms below in the description dealing
Fig. 3.5 Fluorescein fundus angiogram of right eye with old

central retinal artery occlusion but patent cilioretinal arteries. Fig.3.7 Fluorescein angiogram in a rhesus monkey during the
Note filling of the choroid on the temporal and nasal sides, with preretinal-arterial phase, showing early filling of the peripapillary
watershed zone (arrow) between them still not filled. There are choroid and its radial branches to the prelaminar region of the
two cilioretinal arteries on the temporal side and the upper optic nerve head. Note inferior temporal peripapillary choroid and
cilioretinal artery feeds the capillaries in the surface nerve fiber corresponding part of the prelaminar region show no filling as yet
layer (Reproduced from Hayreh [17])
Peri-papillary
choroid Choroidal
vessels
Pre-laminar
region of O.D.
Fig.3.8 Fluorescein angiogram in a rhesus monkey during the

preretinal-arterial phase, showing early filling of the peripapillary
Fig.3.6 Schematic representation of the choroidal and peripap- choroid and radiating branches from it to the prelaminar region of
illary arteries and their centripetal contribution to the prelaminar the optic nerve head. Note the presence of watershed zone (arrow)
region of the optic nerve head. O.D. Optic disc (Reproduced with no filling of the peripapillary choroid as well as that part of
from Hayreh [9]) the prelaminar region (Reproduced from Hayreh [9])
Fig. 3.10 Fluorescein fundus angiogram of right human eye.

Note complete filling of the choroid on the temporal and nasal
sides, with watershed zone between them still not filled (arrow).
Nasal peripapillary choroid filled fully along with the corre-
sponding prelaminar region and only very early filling of the
Fig. 3.9 Fluorescein angiogram in a monkey during the prelaminar region in the inferior-temporal region (Reproduced
preretinal-arterial phase, showing early filling of the temporal from Hayreh [20])
half of the choroid and peripapillary choroid and the corres
ponding temporal part of the prelaminar region of the optic
nerve head. Note the presence of watershed zone (arrow) with
no filling of the peripapillary choroid as well as in that part of the
prelaminar region (Reproduced from Hayreh [10])
with the watershed zone and posterior ciliary arteries),

which agrees with the overall segmental distribution of
the posterior ciliary artery (PCA) circulation [17] as
well as the segmental nature of visual loss in ONH isch-
emic disorders, e.g., AION. Many experimental and
clinical studies have demonstrated selective filling delay
or non-filling of the peripapillary choroid in AION (see
angiograms in the description of the watershed zone and
PCAs), glaucomatous optic neuropathy and ocular
hypertension [9, 10, 12, 13, 18, 20, 22, 26, 27, 38, 39,
41, 43, 48, 58, 6879].
3. Lamina Cribrosa Region
This part of the ONH is entirely supplied by centripetal
branches from the short PCAs, either directly or from
the circle of Haller and Zinn formed by the short PCAs Fig.3.11 Fluorescein fundus angiogram of right eye of a patient
with retinitis pigmentosa shows filling of the peripapillary chor-
(Figs. 3.2, 3.3, 3.14 and 3.15), when that is present.
oid and the prelaminar region of the optic nerve head but no
Ducournau [80] designated these short PCAs as filling of the retinal vessels at this stage (Reproduced from
paraoptic short PCAs and reported branches from Hayreh [9])
Fig.3.14 A neoprene latex cast of Circle of Haller and Zinn

(CZ), exposed in the sclera around the ONH, with its feeding
arteries, i.e. medial and lateral PCAs. Optic nerve (ON) had
been cut close to the eyeball (Reproduced from Hayreh [7])
Fig. 3.12 Fluorescein angiogram in a monkey during the

preretinal-arterial phase, showing early filling of the peripapil-
lary choriocapillaris without any filling of the optic nerve head them to the peripapillary choroid, the circle of Haller
vessels and Zinn and recurrent branches to the pial vascular
plexus on the retrolaminar part. The central retinal
artery gives no branches in this region [13, 58, 61,
62, 81, 82]; Wolff [83] and Lieberman etal. [64], how-
ever, are the only authors who have claimed to have
traced a capillary branch (1020 mm in diameter) in
this region from the central retinal artery in three eyes
on histological studies. The blood vessels in the lamina
cribrosa lie in the fibrous septa and form a dense capil-
lary plexus which makes this part of the ONH a highly
vascular structure [46].
4. Retrolaminar Region
This part of the ONH lies immediately behind the lam-
ina cribrosa. It may have a dual source of blood supply
(Figs.3.2 and 3.3).
(i) Peripheral Centripetal Vascular Supply
This is always present and is formed by the pial
vascular plexus (Figs. 3.2, 3.3 and 3.153.17). The
major sources of supply to the latter are the recurrent
pial branches arising from the circle of Haller and
Zinn or the short PCAs in that area, and the peripapil-
lary choroid (Figs.3.143.18); the recurrent branches
run backward in the pia of the retrolaminar region.
In addition to those, pial branches from the central
Fig. 3.13 Fluorescein angiogram in a monkey during the very
early arterial phase, showing complete filling of the peripapillary retinal artery and other orbital arteries may also con-
choriocapillaris without any filling of the optic nerve head vessels tribute [3, 5, 7] (Figs.3.3 and 3.153.17).
Rec.Br.CZ
lera
Sc CZ
MPCA LPCA
Col. Br. Col. Br.

Med. Mus.
CAR
OA
Col. Br.
Optic Canal
Col. Br.
Col. Br.
Optic Nerve
ICA
Ant.Sup.Hyp.
Art.
Optic Chiasma
Optic Tract Fig.3.16 Shows branches arising from the intravaginal part of
the central retinal artery (C.A.R.) and anastomoses established
by them, as seen from below after removing the dural sheath.
Fig. 3.15 Diagrammatic representation of blood supply of the There are anastomoses with the pial collateral branches from the
various parts of the optic nerve, and location of the Circle of Haller ophthalmic artery (Col. br. of O.A.) and with recurrent pial
and Zinn (CZ), as seen from above. CAR central retinal artery, branches of the circle of Haller and Zinn (CZ). Branches of the
Col. Br. Collateral branches, CZ circle of Zinn and Haller, ICA central retinal artery are slightly retouched to show them clearly
internal carotid artery, LPCA lateral posterior ciliary artery, Med. (Reproduced from Singh and Dass [5])
Mus. medial muscular artery, MPCA medial posterior ciliary
artery, OA ophthalmic artery, Rec. Br. CZ recurrent pial branches
from peripapillary choroid/CZ, (Reproduced from Hayreh [6])
Central artery of the optic nerve: Francois etal.
[1, 2, 63, 84] claimed to have discovered this artery
(Fig. 3.1). According to them, it supplied the ONH.
(ii) Axial Centrifugal Vascular Supply However, in my study of 100 human eyes, this artery
This is present in 75% of the cases. This is constituted was never found; nor was it ever found by any other
by branches (from 1 to 8) from the central retinal investigators.
artery, during its course in the center of the optic nerve From this description of the blood supply of the
[3, 5, 7] (Fig.3.3). Thus, the axial centrifugal vascular ONH, it is evident that PCAs, via either the peripapil-
supply is not as constant and extensive as the periph- lary choroid or branches of the short PCAs (or the
eral centripetal vascular supply. circle of Haller and Zinn when present), are the only
The centripetal pial vessels as well as the centrifu- source of blood to the lamina cribrosa and prelami-
gal vessels (when present) run in the septa of the nerve nar region, and the main (if not the only) source to
(Fig. 3.19), so that the optic nerve septa are in fact the retrolaminar region; in some cases they may sup-
fibrovascular in nature. ply the temporal part of the surface nerve fiber layer
Fig. 3.17 This shows intravaginal course of central artery

of the retinal (CAR) and multiple pial branches arising from
this part with their anastomoses with recurrent pial branches
of the circle of Haller and Zinn (Rec. br. of CZ) anteriorly,
as seen from below on removal of the dural sheath.
MPCA Medial posterior ciliary artery, OA Ophthalmic
artery, ON Optic nerve, PPS Point of penetration of the
sheath by the central retinal artery (Reproduced from Singh
and Dass [5])
Fig.3.18 A longitudinal section of the anterior segment of the

(Figs. 3.23.4 and 3.19, also see many more angio- intraorbital part of the optic nerve and the adjoining part of the
eyeball, showing fine angioarchitecture. A recurrent pial branch
grams below in the description of the watershed zone (*) from the choroidal (C) vessels is seen running backwards in
and PCAs). This fact is of tremendous significance in the pia
any consideration of ischemic disorders of the ONH.
the controversy surrounding their pathogeneses.

Posterior Ciliary Arteries There is much misunderstanding among ophthal-
mologists on the anatomical description and
Since the PCA circulation is the main source of distribution of the various types of PCAs. For a
blood supply to the ONH, a detailed description of proper understanding of the subject, it is essential
the PCAs and their distribution is highly relevant to to clarify their nomenclature, origin, number and
an understanding of ONH ischemic disorders and distribution.
Posterior Ciliary Arteries 43

of the lamina cribrosa region
of the optic nerve head
showing the arrangement of
the septa which are fibrovas-
cular in nature (Reproduced
from Hayreh [146])
Terminology of PCAs and its Implications
Most investigators have used the term PCA loosely

and as a generic term, covering all types of PCAs and
their branches. This is misleading. The main PCAs arise
from the ophthalmic artery [8588] (Figs. 3.15 and
3.203.22). In my series of 59 orbits studied following
postmortem intra-arterial injection of neoprene latex, the
ophthalmic artery was found to give out 15 PCAs (1 in
3%, 2 in 48%, 3 in 39%, 4 in 8% and 5 in 2%) [8588].
Usually there are 2 or 3 PCAs that supply an eye
(Figs.3.21 and 3.22). Ducournau [80], in his series of 24
eyes, found 25 PCAs per eye (2 in 25%, 3 in 50%, 4 in
17% and 5 in 8%). I designated the PCAs as medial (1 in
71% and 2 in 29%), lateral (1 in 75%, 2 in 20%, 3 in 2%
and absent in 3%), and superior (small and inconstant 1
in 7% and 2 in 2%), depending upon their site of entry into
the sclera in relation to the optic nerve [3, 4, 85].
The PCAs, during their intraorbital course and atvari-
able distances from the eyeball, start to divideinto multiple
branches before entering the eyeball (Figs. 3.203.22) and
are named as follows:
a. Long PCAs: There are two long PCAs one on the
medial and the other on the lateral side (Fig.3.23).
They mainly supply a segment of the iris, ciliary Fig.3.20 Origin and intraorbital course of central retinal artery
as seen from below. Abbreviations: CAR Central artery of the
body and peripheral choroid on the medial and
retina, LPCA Lateral posterior ciliary artery, OA Ophthalmic
lateral sides respectively, and play no role in the artery, PPS Point of penetration of the sheath by the central reti-
ONH blood supply [16]. nal artery
Available evidence suggests that the ONH is mostly

supplied by the paraoptic short PCAs.
(ii) Distal SPCAs: The majority of the short PCAs be-
long to this group, and enter the eyeball midway
between the paraoptic SPCAs and long PCAs
(Fig.3.22). They mainly supply the choroid.
In Vivo Supply by the PCAs
Knowledge of the vascular pattern and supply of the PCAs

and their branches was originally drawn entirely from
postmortem cast studies, since the first description by
Frederick Ruysch [89] in 1737 A.D. All the postmortem
cast studies appeared to show that: (1) PCAs have no seg-
mental distribution, (2) they anastomose freely with one
another as well as with the anterior ciliary arteries, (3)
Fig.3.21 Origin and intraorbital course of central retinal artery there are inter-arterial and arteriovenous anastomoses in
as seen from below. Abbreviations: CAR Central artery of the the choroid, and (4) choriocapillaris form a freely com-
retina, LPCA Lateral posterior ciliary artery, MPCA Medial pos-
terior ciliary artery, OA Ophthalmic artery, ON Optic nerve PPS
municating and uninterrupted vascular bed in the entire
Point of penetration of the sheath by the central retinal artery choroid [60, 90]. However, clinically, inflammatory, isch-
emic, metastatic and degenerative choroidal lesions are
usually localized. This puzzling discrepancy was well
summarized by Duke-Elder and Wybar in 1961 [91] when
b. Short PCAs (SPCAs): There may be any number
they stated: The tendency for inflammatory and degen-
up to 20 or so of these (Fig.3.22). They are further
erative diseases of the choroid to show a considerable
subdivided into two subgroups [80, 86, 87].
degree of selective localization, despite the fact that ana-
(i) Paraoptic SPCAs: These are a few SPCAs which en- tomically the vessels would appear to form a continuous
ter the eyeball closest to the optic nerve (Fig.3.22). network, has given rise to speculations regarding the
a b c
Long PCA
Paraoptic
Long PCA SPCAs Long PCA
Optic
Distal SPCAs Nerve Distal SPCAs
OPTIC NERVE
MPCA
CA
MP
LP
CA LPCA
Fig.3.22 Diagrammatic representation of actual branching pattern of medial (MPCA) and lateral (LPCA) PCAs in two eyes (a, b)
and in (c) of the various long and short PCAs as seen on the back of the eyeball (Reproduced from Hayreh [147])
Fig. 3.24 Fluorescein fundus angiogram of an eye with

arteritic AION, showing the sites of entry of the nasal short
PCAs and their course in the choroid. Note no filling of the
Fig. 3.23 Fluorescein fundus angiogram of a normal human lateral PCA
eye, showing the sites of entry of the temporal short PCAs
and their course in the choroid. Arrow marks the center
of themacular region. Note that no artery lies in the center of
themacular region (Reproduced by the kind courtesy of
spectacular appearance. The reason for this discrepancy
Dr. P. Amalric [148]) is that in the postmortem studies, when the cast material
is injected under pressure, the vessels fill from all
sources, irrespective of the normal blood flow pattern,
a natomical isolation of specific choroidal areas. To inves- and therefore give information about the morphological
tigate this discrepancy, I started to research the subject conduits only, while the invivo studies with fluorescein
invivo experimentally as well as clinically in 1967. These angiography reveal the actual pattern of the blood flow
invivo studies on the distribution of PCA and its subdivi- in those channels. It is also possible that in the living eye
sions [1417, 23, 28, 92] discussed below contradicted the the neural supply to the potential anastomotic channels
centuries-old belief (all based on postmortem morphologi- seen on postmortem injection studies may influence the
cal studies) that the entire choroid is one continuous vas- pattern of blood flow; however, as yet, we have no defi-
cular bed, with extensive anastomoses between the various nite explanation for this very important disparity. In any
branches of all the PCAs in the choroid, with no evidence case, what matters clinically in understanding different
of segmental distribution [88]. In sharp contrast to that, my vascular disorders is the invivo circulatory pattern.
invivo experimental [10, 11, 1416, 28, 32, 43] and clini-
cal [1113, 1720, 22, 23, 88, 92] studies showed that
PCAs have a segmental distribution invivo, inthe choroid
In Vivo Segmental Supply by the PCAs
as well as in the ONH. This segmental distribution pattern
exists right down to the terminal arterioles and choriocap-
illaris. These studies clearly demonstrated that invivo the Fluorescein fundus angiograms shown in (Figs.3.27
PCAs are end-arteries. That information is of crucial 3.59) demonstrate the variable number of PCAs sup-
importance in understanding the vascular disorders involv- plying an eye and marked regional variation in the
ing both the choroid and the ONH. choroid and the ONH supplied by different PCAs.
The impressive and fascinating scanning microscopic The PCAs radiate out from their site of entry into the
pictures of casts published over recent years are also, eyeball (Figs. 3.233.25). There may be one PCA
unfortunately, equally misleading in spite of their (Fig. 3.26), two PCAs (Figs. 3.273.42) three PCAs
Fig.3.25 Fluorescein fundus angiogram of an eye with arteritic

AION, showing the sites of entry of two superior short PCAs
and their course in the choroid but not of any other PCA
Fig. 3.27 Fluorescein fundus angiogram of right eye with

embolic NA-arteritic AION and carotid artery disease (negative
temporal artery biopsy for giant cell arteritis), showing normal
filling of the area supplied by the lateral PCA (including the
temporal half of optic disc), but no filling of the area supplied by
the medial PCA (including the nasal half of optic disc)
Fig.3.26 Fluorescein fundus angiogram of an eye with arteritic

AION during the retinal arterial phase showing no filling of the
choroid and the optic nerve head (Reproduced from Hayreh [13])
(Figs. 3.433.45, 3.48, 3.49, 3.513.53 and 3.57) or

four PCAs (Figs. 3.47 and 3.50) in an eye. More
importantly, these angiograms show that there is a
strict segmental distribution by the PCAs in vivo, in
the choroid as well as in the ONH. This segmental Fig.3.28 Fluorescein fundus angiogram of right eye with arter-
distribution pattern also exists in branches of the itic AION showing normal filling of the area supplied by the
lateral PCA but no filling of the area supplied by the medial PCA
PCAsright down to the terminal arterioles and chorio- including the entire optic disc and temporal peripapillary chor-
capillaris. Thus, invivo experimental [10, 17, 28, 92] oid. There was occlusion of a tiny cilioretinal artery (arrow) dur-
and clinical [12, 13, 1820, 22, 39] studies have ing the initial stages (Reproduced from Hayreh [149])
Fig.3.29 Fluorescein fundus angiogram of right eye with arter-

itic AION, showing normal filling of the area supplied by the
Fig.3.31 Fluorescein fundus angiogram of left eye with arter-
lateral PCA, but no filling of the area supplied by the medial
itic AION and associated cilioretinal artery occlusion, showing
PCA including the entire optic disc and temporal peripapillary
normal filling of the central retinal artery and the area supplied
choroid
by the lateral PCA temporal to the fovea only, but no filling of
the choroid supplied by the medial PCA up to the fovea and the
entire optic disc and the area supplied by the cilioretinal artery
(arrow) (Reproduced from Hayreh [19])

lateral PCA and a segment of the superior temporal peripapillary
lateral PCA (including the temporal of the optic disc) but no
choroid and adjacent optic disc, but no filling of the area sup-
filling of the area supplied by the medial PCA (including the
plied by the medial PCA including the most of the optic disc and
nasal of the disc) (Reproduced from Hayreh [19])
temporal peripapillary choroid
demonstrated that the PCAs and their branches a. Interindividual Variation in Area of Supply in the
haveasegmental distribution. This has the following ONH by the Various PCAs
important implications for the evaluation of circula- Fluorescein fundus angiography provides the neces-
tion in the ONH: sary information about this. Figures3.263.61 show

embolic non-arteritic AION (temporal artery biopsy negative for
Fig. 3.33 Fluorescein fundus angiogram of right eye with giant cell arteritis on both sides), shows normal filling of the area
embolic non-arteritic AION in a diabetic, showing normal filling supplied by the medial PCA in the choroid and optic disc but no
of the area supplied by the medial PCA (including the nasal half filling of the area supplied by the lateral PCA (including a small
of disc), but no filling of the area supplied by the lateral PCA sector of the papillomacular bundle in the disc) (Reproduced
including the temporal half of the optic disc from Hayreh [20])
Fig. 3.36 Fluorescein fundus angiogram, of right eye with

arteritic AION, shows normal filling of the area supplied by the
medial PCA in the choroid and optic disc but no filling of the
Fig.3.34 Fluorescein fundus angiogram of left eye with arter- area supplied by the lateral PCA (including a small superior
itic AION, shows normal filling of the area supplied by the temporal sector of the optic disc)
medial PCA, including the nasal half of disc, but no filling of the
area supplied by the lateral PCA, including the temporal half of
disc (Reproduced from Hayreh [12])
3.10, 3.273.41, 3.46 and 3.60). This shows that there
is a marked interindividual variation in the area sup-
the various types of supply in the ONH by the various plied by the various PCAs in man, both in the choroid
PCAs. The lateral and medial PCAs supply the cor- and in the ONH [12, 19, 20, 22, 23, 39, 92]. The
responding parts of the choroid and ONH (Figs.3.9, medial PCA may supply the entire nasal choroid up to
Fig.3.39 Fluorescein fundus angiogram of right eye with non-

arteritic AION showing non-filling of the choroidal watershed
Fig. 3.37 Fluorescein fundus angiogram, of right eye with zone (arrows) between the lateral and medial PCAs and of the
arteritic AION, shows normal filling of the area supplied by the temporal part of optic disc (Reproduced from Hayreh [20])
medial PCA in the choroid up to the fovea, including the entire
optic disc and temporal peripapillary choroid, but no filling of
the area supplied by the lateral PCA from fovea temporally
Fig.3.40 Fluorescein fundus angiogram of left eye with arter-

itic AION with the optic disc lying in the center of the watershed
zone (arrows), during the late retinal arteriovenous phase
the level of the fovea, including the entire ONH

(Figs. 3.31 and 3.37), or its supply may stop short
nasal to the nasal peripapillary choroid (Fig.3.41), so
that it may take no part at all in the blood supply of the
Fig.3.38 Fluorescein fundus angiogram of right eye with mild ONH; or there may be any number of variations
arteritic AION, shows non-filling of the watershed zone (indi-
cated by arrows) between lateral and medial PCAs (Reproduced between these two extremes (Figs. 3.273.41). The
from Hayreh [23]) lateral PCA supplies the area of the choroid and ONH
Fig.3.43 Fluorescein fundus angiogram of left eye with arteritic

AION, shows normal filling of the choroid supplied by the
lateral PCA (including superior and temporal parts of the optic
disc) but no filling of the inferior nasal area of the choroid
Fig.3.41 Fluorescein fundus angiogram, of left eye with non- supplied by the medial PCA (including the inferior nasal part of
arteritic AION, shows non-filling of the watershed zone (indi- the disc) (Reproduced from Hayreh [113])
cated by arrows) between the lateral and medial PCAs passing
through the nasal part of the disc and adjacent nasal peripapil-
lary choroid (Reproduced from Hayreh [20])
Fig.3.42 Fluorescein fundus angiograms of a normal right eye

of a healthy cynomolgus monkey (after experimental occlusion Fig. 3.44 Fluorescein fundus angiogram of a normal human
of the central retinal artery to unmask vessels from the PCA- right eye, showing superior sectoral filling of the choroid and
circulation in the prelaminar region of the optic disc). It shows adjacent prelaminar region of the optic nerve head, and a Y-shaped
filling of upper half of the choroid and optic disc with a well- watershed zone (indicated by arrows) between the superior,
demarcated horizontal border (Reproduced from Hayreh [10]) lateral and medial PCAs (Reproduced from Hayreh [12])
not supplied by the medial PCA or vice versa. When present, it accordingly may supply a superior sector
there is more than one medial or lateral PCA, the area (Figs. 3.42, 3.44, 3.46 and 3.47). Therefore, these
supplied by each of them may be highly variable examples show a high interindividual variation in
(Figs.3.433.51 and 3.61). When a superior PCA is number and distribution by the various PCAs. This
Fig.3.45 Fluorescein fundus angiogram of right eye with non-

arteritic AION, shows a markedly delayed filling of a Y-shaped
watershed zone (indicated by arrows) between superior, medial
and lateral PCAs (Reproduced from Hayreh [20])
Fig.3.47 Fluorescein fundus angiogram of a normal right eye

of a healthy cynomolgus monkey (after experimental occlusion
of the central retinal artery to unmask PCA supply) with four
PCAs (similar to Fig.3.60e). It shows filling of the PCA supply-
ing the choroid in the superior nasal quadrant only. The marked
fluorescence of the optic disc was due to this being the second
injection of fluorescein (Reproduced from Hayreh [10])
common practice of measuring the blood flow in a

PCA may be irrelevant to the rest of the ONH, if that
PCA does not supply the ONH or supplies only a
small segment of it (see Chap. 7).
b. Segmental Supply in the Peripapillary Choroid
As discussed above, peripapillary choroid is an
important source of blood supply to the ONH.
Experimental [15, 18, 29] and clinical studies on
AION and other ischemic disorders of the ONH have
revealed that the peripapillary choroid has a seg-
mental blood supply in vivo (Figs. 3.83.10, 3.27,
Fig. 3.46 Fluorescein fundus angiogram of right eye shows 3.30, 3.323.36, 3.38, 3.39, 3.413.46, 3.48, 3.51
filling mainly of the superior nasal quadrant of the choroid and 3.55 and 3.573.59). This is responsible for the seg-
adjacent optic disc and watershed zone (big arrow). Note early
filling of a small cilioretinal artery temporally and that part of the mental pattern of blood supply in the ONH, and
optic disc shows filling by the cilioretinal artery (small arrow) consequently for the well-known sectoral nature of
ischemic lesions seen in AION and other ischemic
shows that the pattern of distribution by the PCAs in disorders of the ONH. Thus, this segmental blood
the choroid and ONH is extremely variable the supply in the peripapillary choroid plays an impor-
most important fact to be borne in mind in any consid- tant role in the ONH ischemic disorders [9, 10, 18,
eration of ischemic disorders of the ONH. The 20, 68, 92].
Fig. 3.48 Fluorescein fundus angiogram of left eye with

normal-tension glaucoma during the early retinal arterial phase,
showing delayed filling of the inferior temporal PCA, particu-
larly in the peripapillary region. Note filling of the prelaminar Fig.3.50 Fluorescein fundus angiogram of right eye with arter-
region in the nasal part where peripapillary choroid is also filled. itic AION with 4 PCAs (1, 2, 3, 4), showing no filling of the
Arrows shows the watershed zone inferior nasal PCA (number 3), the entire peripapillary choroid,
and watershed zones (arrows), during the late retinal venous
phase of angiogram
Fig.3.51 Fluorescein fundus angiograms, of right eye with low-

tension glaucoma and cupping of the lower half of the optic disc
with corresponding visual field defect. (a) shows a filling defect
in the inferior temporal quadrant of the choroid and lower half of
the optic disc. (b) Angiogram 5s after (a) shows non-filling of
Fig.3.49 Fluorescein fundus angiograms of left eye with arter- lower half of the watershed zone (arrow) and the involved lower
itic AION, showing no filling of the inferior temporal PCA half of the optic disc (Reproduced from Hayreh [20])

normal tension glaucoma. It shows no filling of the lower part
of the optic disc and adjacent peripapillary choroid and patchy
filling of the watershed zones (arrows) (Reproduced from
Hayreh [20])
Fig.3.51 (continued)

non-arteritic AION showing non-filling of temporal part of Fig.3.54 Fluorescein fundus angiogram of right eye with non-
the peripapillary choroid and adjacent optic disc and of upper arteritic AION, showing non-filling of the peripapillary choroid,
half of the choroidal watershed zone (arrow) (Reproduced from upper half of the watershed zone (arrow) and most of the optic
Hayreh [20]) disc (Reproduced from Hayreh [20])
Fig.3.57 Fluorescein fundus angiogram of left eye with non-

arteritic AION shows a large area of non-filling of the inferior
Fig.3.55 Fluorescein fundus angiogram of right eye with arter- temporal peripapillary choroid and adjacent part of the optic
itic AION shows a non-filling of the upper half of the optic disc disc and lower half of the watershed zone (arrow) (Reproduced
and upper half of the watershed zone (arrow) (Reproduced from from Hayreh [113])
Hayreh [13])
Fig.3.58 Fluorescein fundus angiogram of right eye of a patient

Fig. 3.56 Fluorescein fundus angiogram of left eye shows a with non-arteritic AION and internal carotid artery occlusion. It
non-filling of the upper half of the watershed zone (arrow) and shows non-filling of lower half of the watershed zone (arrow)
temporal peripapillary choroid in a person with carotid artery and temporal part of the optic disc and adjacent peripapillary
disease choroid (Reproduced from Hayreh [20])
Watershed Zones Between the Various PCAs 55
one or more of the end-arteries, the watershed zone,

being an area of comparatively poor vascularity, is
most vulnerable to ischemia. Watershed zones play an
important role in ischemic disorders of the ONH.
As discussed above, fluorescein fundus angio-
graphic studies, both in health and disease, have clearly
shown that the PCAs and their subdivisions are end-
arteries invivo, right down to the choroidal arterioles
[10, 1220, 22, 28, 29, 32, 68, 92]. Thus, there are
watershed zones between the distribution of the vari-
ous PCAs (Figs.3.10, 3.383.41, 3.44, 3.45, 3.51b and
3.523.59). My studies have shown that the watershed
zones are arranged as follows.
L ocations of Watershed Zones Between

PCAs [17, 19, 20, 68, 92]
The location of the watershed zones varies according

Fig.3.59 Fluorescein fundus angiogram of right eye of a man
with non-arteritic AION in the left eye and multiple recurrent to the number of the PCAs and their locations.
episodes of transient visual loss in the right eye due to excessive
arterial hypotensive therapy for white-coat hypertension, show-
ing non-filling of the inferior half of the watershed zone (arrow) When There Are Two (Medial and Lateral) PCAs
c. PCA Watershed Zones Distribution and Their Since the area of the choroid supplied by the medial
Location and lateral PCAs shows a marked inter-individual vari-
My fluorescein angiographic studies of distribution of ation (Figs.3.10 and 3.273.41), this must result in a
the PCAs in the choroid revealed that there are watershed wide variation in the location of the watershed zone
zones between the distribution of the various PCAs and between the two. Figure3.60 is a diagrammatic repre-
their branches [17, 20, 23]. Studies in eyes with AION sentation of some of the locations of the watershed
and glaucomatous optic neuropathy have indicated that zone between the medial and lateral PCAs and shows
the location of the watershed zone determines the vulner- that the watershed zone may:
ability of the corresponding part of the ONH to ischemia
[20, 23]. The importance of these watershed zones in the (a) be situated temporal to the peripapillary choroid
ONH ischemic disorders is discussed below. (Fig. 3.38),
(b) pass through the temporal peripapillary choroid
(Fig.3.10, 3.39),
atershed Zones Between
W (c) pass through the medial peripapillary choroid
(Fig.3.41),
the Various PCAs
(d) pass through one or the other part of the optic disc,
or the entire optic disc may lie in the watershed
When a tissue is supplied by two or more end-arteries, zone (Fig.3.40), or
the border between the territories of distribution of any (e) various combinations of the above.
two end-arteries is called a watershed zone. The
occurrence of such watershed zones between the vari-
ous cerebral arteries is well-known, as is also the case When There Are Three or More PCAs
in some other organs having end-arterial systems. The
significance of the watershed zones is that in the event When three or more PCAs supply an eye, the loca-
of a fall in the perfusion pressure in the vascular bed of tions of the watershed zones vary accordingly, as
shown diagrammatically in Fig.3.61 and on fluores- (Figs.3.44 and 3.45). In almost all of the patterns, the
cein fundus angiography in Figs. 3.44, 3.45, 3.51b watershed zone passes through the ONH [20, 68]; that
and 3.57. makes the watershed zones play a very important role
in ONH ischemic disorders, because the part of the
ONH that is located in the watershed zone is most vul-
I mportance of Location of the Watershed Zone nerable to ischemia, and the most vulnerable position
in Ischemic Disorders of the Optic Nerve Head is when the entire ONH lies the center of a watershed
zone (Figs.3.40 and 3.45). If one or more of the PCAs
The location of a watershed zone in relation to the in such circumstances has low perfusion pressure, the
ONH is key to any discussion of ischemic disorders of eye will show a filling defect in the corresponding
the ONH. As noted above, the watershed zone may watershed zone, which may extend vertically in the
pass temporal or nasal to the ONH, or pass through entire length (when there are only two PCAs) or
one or the other part of the ONH, or cover the entire obliquely oriented or involve only upper/lower half of
ONH. It may extend vertically along the entire length the watershed zone [20, 68].
(Figs. 3.10, 3.383.41 and 3.60), may be only in the Shimizu etal. [54] on fluorescein fundus angiogra-
upper vertical half region (Figs.3.523.56 and 3.61) or phy of 172 eyes found filling defects between medial
only in the lower vertical half (Figs.3.51b and 3.57 and lateral PCAs at the following 3 locations: (1)
3.61), or it may assume the shape of the letter Y Involving the entire optic disc and temporal and nasal
Fig.3.60 Diagrammatic representation of some of the locations of the watershed zones (shaded area) between the medial and
lateral PCAs in human eyes (Reproduced from Hayreh [150])
Watershed Zones Between the Various PCAs 57
Fig.3.61 Diagrammatic representations of some examples of one medial and two lateral PCAs, and (d) one lateral and two
the locations of the borders between the territories of the various medial; and (e) with four PCAs: two medial and two lateral
PCAs: (a) with two PCAs: one medial and the other lateral; (Reproduced from Hayreh [151])
(bd) with three PCAs in different combinations: (b and c) have
peripapillary regions in 20%, (2) involving the tempo- the choroid along nasal and temporal edges of the disc
ral part of the disc in 27%, and (3) randomly distrib- (20.2%), Type 2c optic disc and a band of choroid
uted non-fluorescent patches in a geographic pattern in temporal to the disc (21.3%), Type 3 temporal half of
52%. They found no filling defect extending beyond the disc and a band of choroid temporal to the disc
the temporal peripapillary region on to the foveal (45.4%), and Type 4 between optic disc and fovea
region. The filling defects reported by them in location (4.8%). These locations are similar to ones seen in my
(1) and (2) above corresponded to the watershed zones fluorescein angiographic studies described above.
in the choroidal vascular bed, although they did not In my fluorescein angiographic studies in eyes with
describe these filling defects as watershed zones. primary open angle glaucoma and normal-tension
Giuffre [93] reported the prevalence and position of glaucoma, where I could outline the watershed zone, I
the choroidal watershed zones between the medial and found the incidence of the various locations of the
lateral PCAs on fluorescein angiography of 800 nor- watershed zones as shown in Fig.3.62 [23, 68]. Sato
mal subjects. He could not outline the watershed zone et al. [94], based on Indocyanine green angiography,
in 33% due to technical problems and in 22.3% due to reported the location of watershed zones in 54 eyes (27
the simultaneous filling of the various PCAs. In the patients) with normal-tension glaucoma. In 8 eyes
remaining 44.6% (357 eyes) the watershed zone was (14.8%) of 7 patients the watershed zone did not
well outlined. He classified the watershed zone into 6 include the ONH, in 32 eyes of 20 patients (59.3%) the
types, involving the following regions: Type 1 nasal watershed zone partially included the ONH, and in 10
half of the optic disc and a band of choroid nasal to the eyes of 14 patients (18.5%) the watershed zone
disc (3.1%), Type 2a optic disc and choroid nasal to included the ONH. Of the 27 normal-tension glaucoma
the disc (5.3%), Type 2b optic disc and some part of patients, 10 (37.0%) had different locations of
Fig.3.62 Diagrammatic
representation of some of the 60%
locations of the watershed
zones (shaded area) between
the medial and lateral PCAs
seen in glaucomatous optic
neuropathy (Reproduced
from Hayreh [23])
16% 10%
watershed zone in each eye. The mean deviation of studies may simply represent the area of supply by the
visual fields was greater in the eye with watershed paraoptic short PCA instead of being a true watershed
zone which included a larger part of the ONH than in zone. My studies indicate that that is not the case at all,
the contralateral eye. Conversely, the eye with the as is evident from the following:
greater mean deviation had a watershed zone that
included a larger part of the ONH. They concluded that (a) I investigated the in vivo pattern of supply by the
the location of the watershed zone appeared to influ- various short PCAs in 47 rhesus monkey eyes by
ence the progress of visual field defect. cutting one of them in an eye and outlining the
Ducournau [80] and Olver [95], based on their post- immediate post-occlusion filling defect in the chor-
mortem cast studies, have claimed that the choroidal oid by fluorescein angiography [15]. I found that in
branches from the paraoptic short PCAs supply a verti- rhesus monkeys, as in man, the short PCAs fall into
cal meridional region of the choroid. There is some two subgroups at the site of entry into the sclera
implication in this observation that the watershed zone one entering close to the optic nerve (paraoptic
shown by various studies in fluorescein angiographic group) and the other entering between the paraoptic
Circle of Haller and Zinn 59
group and the long PCA (distal group). In 12 eyes I Circle of Haller and Zinn
cut a paraoptic short PCA and in 35 a distal short
PCA. On cutting the paraoptic short PCA, there was Since Haller in 1754 [96] and Zinn in 1755 described
usually a roughly triangular wedge-shaped choroidal this arterial circle, situated in the peripapillary sclera
filling defect, with its apex towards the optic disc and and formed by anastomoses between the PCAs
the base towards the equator of the eye; in 6 of the 12 (Figs.3.3 and 3.143.16), its existence has been often
eyes the filling defect involved the corresponding reported [5, 7, 46, 57, 62, 65, 81, 82, 9799]. However,
segment of the peripapillary choroid, but not in the its prevalence, what it supplies and how often it is a
rest. The size, shape and distribution pattern of complete or incomplete circle have been controver-
thefilling defects bore no resemblance whatever to sial. More recently scanning electron microscopic
the watershed zone. The choroidal filling defect seen examination of plastic microvascular corrosion casts
on cutting the distal short PCA was usually much of the human ONH vasculature has provided detailed
larger and of greatly varied shape, and it also included information on the circle of Haller and Zinn. Reports
the peripapillary choroid in 21 of the 35 eyes. The of its prevalence in human eyes have varied; e.g. Olver
overall distribution pattern of the paraoptic and distal etal. [97] found it in more than 75% of 18 casts, Onda
short PCAs had so many common features that I did etal. [62] in 61% of 18 eyes, and Gauntt etal. [99] in
not feel it justified to divide the short PCAs into these 83% of 29 eyes. It is generally agreed that the circle of
two subgroups from the functional or anatomical Haller and Zinn gives three sets of branches: to the
point of view; it was more informative to divide them lamina cribrosa, to the peripapillary choroid and recur-
into superior, middle and inferior groups, depending rent pial branches to the retrolaminar region
upon the location at their site of entry into the eyeball (Figs. 3.153.17) [62, 81, 82, 97, 100]. Olver et al.
and the area supplied [15]. It has been shown that in [82, 97] described the circle of Haller and Zinn as a
nonhuman primates there is no circle of Haller and microscopic, intrascleral, elliptical, microvascular
Zinn [8], as is also the case with many humans; in anastomosis formed by branches of the medial and
such a situation, direct branches from the paraoptic lateral paraoptic short PCAs; they preferred to call it
short PCAs to the ONH behave like those from the perioptic nerve arteriolar anastomoses [97].
circle of Haller and Zinn. My findings suggest that: According to them, the complete or incomplete ellipse
(1) every paraoptic artery does not supply the ONH; is divided into superior and inferior parts by the entry
(2) once they are in the choroid, they have wedge- points of these branches into the eye and shows mor-
shaped supply similar to that by the other short phological variations (interindividual as well as
PCAs; and (3) their choroidal distribution may not interocular in the same person) in terms of its form,
necessarily correspond to that in the ONH. position and branches [82]. Of the 18 casts of the cir-
(b) On routine fluorescein angiography in man, I have cle of Haller and Zinn they examined, they found
found watershed zones much more common in complete anastomoses in 8 (44%), complete with nar-
eyes with low perfusion pressure in the choroid, rowed sections in 6 (33%) and incomplete anastomo-
AION and glaucomatous optic neuropathy than in ses in 4 (23%). The Circle was supplied by branches
normal healthy individuals. of medial and lateral paraoptic short PCAs in 15
(c) I have seen horizontal watershed zones as well (83%), lateral paraoptic short PCAs in 2 (11%), and
(Figs. 3.44, 3.45, 3.53 and 3.61) [20, 23] which medial, lateral and superior short PCAs in 1 (6%) [97].
cannot be explained by the paraoptic short PCA They concluded that it lies in different planes antero-
supply pattern reported by the two authors and posteriorly (like a hammock) and has both extrascleral
seen in my studies. and intrascleral portions. They found that branches
from the circle of Haller and Zinn also formed arteri-
From the above account it is clear that understanding the olar-arteriolar anastomoses. Onda et al. [62] on
blood supply of the ONH is central to a logical under- examination of 18 casts found that the circle is formed
standing of the mechanisms responsible for the various by branches of the short PCAs; it was a complete (in
ischemic disorders of the ONH. It is also evident that it 2 of 11 eyes) or a well-developed (9 of 11 eyes)
is only the PCA circulation in the ONH that is relevant anastomotic arterial circle surrounding the ONH,
to the various ischemic disorders of the ONH. located approximately 200300 mm posterior to the
suprachoroidal space within the perineural sclera. standard vascular pattern. My multifaceted studies on
Gauntt etal. [99] on serial sectioning of 29 eyes found the blood supply of the ONH have revealed that there
the circle of Haller and Zinn incomplete in 3 and nar- are tremendous inter-individual variations in the blood
row in 8. They described two types of position of the supply pattern of the ONH [20, 8688]. A number of
circle in the sclera around the optic nerve head in factors are responsible for that.
Type 1 it was located lateral to the site of attachment
(1) As discussed above, there is marked interindivid-
of the dural sheath to the sclera (20 eyes 69.0%), and
ual variation in the number and pattern of supply
in Type 2 medial to the sheath (4 eyes 13.8%); the
by the PCAs the main source of blood supply to
circle of Haller and Zinn was more medially located in
the ONH. Therefore, the pattern of distribution
small optic discs. They postulated that a combination
by the PCAs in the choroid and ONH is extremely
of small disc, medial displacement of the circle, and
variable a most important fact to be borne in
anatomical variation in the vascular pattern may pre-
mind in any consideration of ischemic disorders of
dispose an ONH to ischemic events. Strek etal. [100]
the ONH.
found considerable variations in the position of the
(2) I studied the anatomical pattern in 100 human
circle in 15 human fetuses aged 1620 weeks. Heimann
specimens and no two specimens had identical
[101, 102] also described the beginning of the circle
patterns, not even the two eyes of the same person
formation in fetal studies.
[3, 5].The anatomical studies revealed marked
Some authors claimed to have outlined the circle
interindividual variation in the sources of branches
of Haller and Zinn on fluorescein angiography [65,
which supply the ONH [6, 7]. The ONH, in addi-
103]. In my fluorescein angiographic studies, I have
tion to supply from the PCAs, may have in the ret-
found that various peripapillary choroidal arteries
rolaminar region an infinitely variable amount of
usually join together to form a peripapillary choroidal
supply from the central retinal artery [5] and pial
arterial arcade around the optic disc, situated in front
branches from sources other than PCA and central
of the sclera, in the choroid and NOT in the sclera. It
retinal artery (Figs.3.3 and 3.153.17).
can be easily seen when there is atrophy of the peri-
papillary retinal pigment epithelium. To visualize This inter- and intra-individual variation has impor-
blood vessels lying deep in the opaque scleral tissue tant implications both from the clinical aspect and in
by fluorescein angiography does not seem feasible explaining the disagreements on the subject. Clinically,
[21, 98]. The same objection applies to claims of see- for example, it explains why occlusion of the same
ing the circle of Haller and Zinn on Indocyanine green PCA gives rise to widely varying patterns and mani-
angiography in high myopia with peripapillary degen- festations of ischemic disorders of the ONH. Thus, it
eration [104, 105]. Shimizu and Ujiie [59] in their is a mistake to attempt to explain various ONH isch-
casts of the peripapillary choroid have very nicely emic disorders on one standard vascular pattern.
demonstrated the existence of the peripapillary chor-
oidal arterial arcade. It seems claims of seeing the
circle of Haller and Zinn on fluorescein or indocya-
nine angiography arise from the authors confusing ontroversies on the Blood Supply
C
the peripapillary choroidal arcade with the circle of of the Optic Nerve Head
Haller and Zinn.
The subject of the blood supply of the ONH has been
bedeviled by disagreements for four decades [1, 2,
6267, 82, 84, 106]. That in turn has produced con-
I nter-Individual Variation in the Blood troversies not only on the pattern of blood supply of
Supply of the Optic Nerve Head the ONH but also on ONH ischemic disorders and
how they are to be understood. In view of this, it is
There is a general impression that the pattern of blood essential to distinguish myth from reality. Briefly,
supply of the ONH is similar in all eyes, and therefore following are the major reasons and areas of
all ischemic lesions of the ONH are explainable by one controversy.
Controversies on the Blood Supply of the Optic Nerve Head 61
Reasons for the Controversies not PCAs any more. Thus, it is simply semantics
which cause the confusing claims.
1. Investigation Techniques 4. Extrapolation from Postmortem Morphological
Different investigators have used different techniques Studies to the In Vivo Blood Supply Pattern in
to study the subject without realizing their limitations the ONH
[86, 87, 107]. Different techniques give different As discussed above, almost all of the studies on the
types of information on the vascular bed. Over the blood supply of the ONH in the literature are based
years I have collected, bit by bit, information on the on postmortem morphological studies rather than
blood supply and circulation of the ONH by using a in vivo studies. I feel that is the primary cause of con-
variety of differing techniques, including: anatomical fusion and controversy. For example, we know now
studies using postmortem neoprene casts, histologi- that for almost two and half centuries, since the first
cal studies using serial sections and reconstruction of description by Ruysch in 1737 [89], postmortem cast
the anterior part of the optic nerve; fluorescein fun- studies on the PCAs and the choroidal vascular bed
dus angiography in experimental ocular and ONH misled us badly into believing that that was a freely
vascular studies (in rhesus monkeys), and in clinical communicating vascular system; whereas in vivo
studies dealing with the ONH, retinal and choroidal fluorescein fundus angiographic studies have shown
vascular disorders, as listed above [329, 3139, 40, conclusively that the PCAs have a strictly segmental
86, 108]. I have found that all techniques do not distribution in the choroid and ONH (see above), and
always give identical information, and the histologi- the PCA vascular bed is an end-arterial system [10,
cal studies are the least reliable. This is rather like the 12, 1520, 22, 23, 2830, 32, 39, 92]. This informa-
old story of the blind men describing an elephant; tion is highly relevant to the circulation in the ONH
each found something different but none could since the PCAs and peripapillary choroid are the
describe the whole animal. Only a multifaceted study main source of its blood supply. As discussed above,
conducted on large series and using a variety of tech- scanning microscopic pictures of casts recently pub-
niques, including in vivo clinical and experimental lished are unfortunately equally misleading in spite
investigations, can give a comprehensive and accu- of their spectacular appearance, with the additional
rate account. problems caused by lack of exact orientation of the
2. Sample Size casts to the soft tissues. Another important reason for
Because of the considerable interindividual varia- the discrepancy between the postmortem morpho-
tions in the blood supply pattern of the ONH, stud- logic and invivo patterns is that when the cast mate-
ies based on only a few eyes generally give skewed rial is injected under pressure the vessels fill from all
information, depending upon what method of inves- sources and directions, irrespective of the normal
tigation was used and what patterns were revealed in blood flow pattern; whereas fluorescein angiography,
those few specimens with that particular technique. as a marker for the blood flow, reveals the actual pat-
3. Semantics tern of the blood flow in those channels. What mat-
This can also be a cause of confusion. For example, ters clinically, in explaining different vascular
Cioffi and Van Buskirk [67] have stressed that the disorders, is the in vivo circulatory pattern. Having
peripapillary choroid does not contribute to the myself done histological studies on the blood supply
prelaminar region of the ONH; however, they have of the optic nerve, I have found such studies are even
stated that some branches of the circle of Haller and more fraught with problems and may provide mis-
Zinn and of the short PCAs course through the chor- leading information about the vascular pattern and
oid and ultimately supply the prelaminar region; blood flow in the ONH. Clearly caution must be used
once those branches are lying in the peripapillary when trying to make inferences about physiological
choroid, they are in fact a part of the peripapillary ONH blood flow purely from postmortem anatomic
choroid. From the conventional semantic point of studies.
view, once short and long PCAs and branches of cir- 5. Miscellaneous
cle of Haller and Zinn have pierced the sclera, then Confusion can also be caused by differences in inter-
they and their branches are a part of the choroid and pretation and by unconscious personal bias.
ajor Areas of Controversy in Blood

M 2. Role of Circle of Haller and Zinn in Blood
Supply of the Optic Nerve Head Supply of the ONH
It is discussed above.
3. Nature of the Axial Vascular System in the ONH
1. Role of Peripapillary Choroid in Blood Supply Franois and colleagues [1, 2, 63, 84] during the mid-
of the ONH fifties claimed that the central artery of the optic
There is overwhelming evidence that the peripapillary nerve a branch of the ophthalmic artery (Fig. 3.1),
choroid supplies the prelaminar region [1, 2, 6267, formed the axial vascular system, with the central reti-
82, 84, 106], and the most convincing evidence comes nal artery playing no role in the blood supply of the
from the fluorescein angiographic studies (Figs.3.7 optic nerve. They claimed that the anterior branch of
3.11, 3.26, 3.27, 3.28, 3.30, 3.323.36, 3.38, 3.39, the central artery of the optic nerve ran in the center
3.41, 3.423.48, 3.503.55, 3.57, 3.58 and 3.63). of the anterior part of the optic nerve and that it was the
However, a few papers based purely on postmortem main source of supply to the ONH. They stressed that
morphological studies have claimed that this is not so a central artery of the optic nerve undoubtedly exists
[6267]. It is interesting that while Onda et al. [62] [63] and that the existence of such an artery in the
stated that the peripapillary choroid did not supply the optic nerve is beyond all doubt [84]. Yet neither I
prelaminar region, they also stated in their paper that: [37] nor any other investigator in the field could find
The peripapillary choroid provided several small cen- any such artery despite careful, detailed studies in
tripetal arteries to the prelaminar region. As men- much larger samples. In view of this controversy,
tioned above, these authors have also stated that some Franois group later re-investigated the subject
branches of the circle of Haller and Zinn and of the [109112] and they also did not find any central artery
short PCAs course through the choroid and ultimately of the optic nerve in any specimen; that was the
supply the prelaminar region [67]. I have discussed the demise of the mythical central artery of the optic
reasons for the discrepancy at length elsewhere [21, nerve. Lieberman etal. [64], based on their histologi-
45, 68, 8688]. cal study, nevertheless tried to revive the concept by
describing so-called central optic nerve artery equiva-
lents when they stated that the pial-derived longitu-
dinal arterioles in the retrolaminar nerve qualify as
central optic nerve artery equivalents. This statement
is totally misleading, because the pial feeders to the
optic nerve run in all directions and lie throughout the
thickness of the optic nerve in its intraorbital part,
without any invivo functional anastomoses; thus these
pial-derived vessels in no way deserve the designation
of central optic nerve artery equivalents. Moreover,
the pial branches do not form any axial vascular sys-
tem of the optic nerve (Fig. 3.3); that is mostly sup-
plied by the intraneural branches of the central retinal
artery (Figs. 3.3 and 3.643.66), and most probably
Lieberman etal. [64] misinterpreted those branches of
the central retinal artery in their histological study. In
my studies with liquid neoprene latex [3, 5, 7], the cen-
tral retinal artery gave 18 prominent branches during
its course within the optic nerve (Figs.3.643.66) in
75% of the specimens. I feel the primary reason for
misinterpretation by Lieberman etal. [64] is that their
Fig.3.63 Fluorescein fundus angiogram, during retinal venous
phase, of right eye with arteritic AION. It shows non-filling of
conclusions were based purely on a histological study,
the peripapillary choroid and the optic disc (Reproduced from which is not the ideal way to examine the optic nerve
Hayreh [12]) vascular supply.
Blood Supply of the Intraorbital Part of the Optic Nerve 63
clinical studies using fluorescein fundus angiography

have demonstrated the sectoral nature of the blood
supply in the ONH over and over again [9, 10, 12, 13,
1720, 22, 2528, 32, 44, 68]. Similarly, ischemic
lesions of the ONH, as in AION, almost always have a
sectoral distribution and visual field defects [113, 114],
which gives definite evidence of the sectoral nature of
blood supply in the ONH.
5. Watershed Zones in the ONH
As discussed above, the presence of watershed zones
among the distribution of various PCAs is now well
established. From time to time, based purely on specu-
lation, some authors have suggested the presence of
other types of watershed zone in the ONH. For exam-
ple, Rootman and Butler [115] speculated about the
presence of a watershed area in the retrolaminar optic
nerve. More recently Feldon [116] postulated a water-
shed zone between the retinal and PCA circulation in
the prelaminar region. These speculations were put
forward to explain pet theories, but without any actual
scientifically valid studies or evidence.
enous Drainage of the Optic

V
Nerve Head
Fig.3.64 Course of the central artery of the retina (CAR) as seen
from below, before its site of penetration of the sheath (PPS) of The ONH is primarily drained by the central retinal
the optic nerve (ON), and after its site of penetration (PPN) into vein (Fig.3.3). The prelaminar region also drains into
the optic nerve. Its intraneural (I/N) course is seen on splitting the peripapillary choroidal veins because in this part the
open the optic nerve anterior to the site of penetration. It shows
central retinal vein communicates with the choroidal
branches of the central artery of the retina one (1) before its
penetration into the optic nerve, and three (1, 2, 3) during its intra- circulation [9, 25, 47, 50, 53, 55]. This communication
neural course [Reproduced from Singh (Hayreh) and Dass [5]] assumes significance in acting as retino-ciliary collater-
als (erroneously called opto-ciliary shunts, because
there is no optic vein and these are not shunts but
4. Sectoral Distribution of Blood Supply collaterals) to drain away the blood in the event of a
in the ONH central retinal vein occlusion behind the lamina crib-
Lieberman etal. [64], based only on histological stud- rosa (Fig. 3.68). There is no venous channel corre-
ies, stated that there is no sectoral distribution of blood sponding to the circle of Haller and Zinn.
supply in the ONH because there is a continuous capil-
lary network. One very fundamental fact of circulation
to be borne in mind in this respect is that capillaries
lood Supply of the Intraorbital
B
in vivo DO NOT act as collaterals in the event of
occlusion of their feeding artery/arteriole the devel- Part of the Optic Nerve
opment of retinal cotton-wool spots (inner retinal isch-
emic spots) is a good example of that: in spite of a This is further subdivided by point of entry of the cen-
continuous retinal capillary bed, occlusion of a termi- tral retinal artery in the optic nerve into (a) anterior and
nal retinal arteriole produces a focal retinal infarct (b) posterior segments (Fig.3.3). I have discussed the
(Fig.3.67). As discussed above, my experimental and subject in detail elsewhere [5, 7, 10].
Fig.3.65 Course of the

a b
central artery of the retina
(CAR) as seen from below,
before (*) and after its site of 8
penetration into the optic
nerve (ON). Its intraneural 7 6
(I/N) course is seen on
splitting open the optic nerve
anterior to the site of
5
penetration. It shows branches
of the central artery of the
retina two (1, 2) before its ON
penetration (PPN) into the split open
optic nerve, and six (38) 1/N
during its intraneural course.
Figure (b) is Camera Lucida 4
drawing of (a) for better
clarity (Reproduced from
Singh (Hayreh) and Dass [5])
3 1
2
CAR
ON
unsplit
Anterior Segment supplied by multiple small collateral arteries usually

arising directly from the ophthalmic artery and less
This is between the ONH and the site of entry of the often from other orbital arteries (Figs. 3.3 and 3.15
central retinal artery into the nerve. This part has two 3.17) [7, 85]. In about 10% of optic nerves there may
vascular systems for its supply. be an axial centrifugal vascular system extending
backward for a variable distance, formed by intraneu-
(i) A Peripheral Centripetal Vascular System
ral branches of the central retinal artery (Fig. 3.69)
This is present in all cases and is formed by a pial vascu-
[3, 5, 6]. Isayama etal. [117], on cast studies in mon-
lar plexus, supplied by multiple pial branches originating
keys found that the posterior intraorbital part was sup-
from the peripapillary choroid, circle of Haller and Zinn,
plied by a centripetal vascular system, formed by the
central retinal artery, ophthalmic artery and other orbital
pial vessels arising from the first branches of the oph-
arteries (Figs.3.2, 3.3 and 3.153.18) [3, 5, 7, 85].
thalmic artery.
(ii) An Axial Centrifugal Vascular System
This is present in 75% of nerves and supplied by 18
intraneural branches of the central retinal artery
(Figs.3.2, 3.3 and 3.643.66) [3, 5, 6], and the supply enous Drainage of the Intraorbital
V
by the central retinal artery may extend for 14 mm Part of the Optic Nerve
behind the site of its penetration into the optic nerve
and give centrifugal branches. The intraorbital part of the optic nerve has pial veins;
however, the central retinal vein is the primary vein
draining this part of the optic nerve (Figs. 3.3 and
Posterior Segment 3.70). Pial veins lie in the pia mater and drain mainly
into the central retinal vein at the site where the latter
This is primarily supplied by the peripheral centripetal exits the optic nerve. The central retinal vein is usually
vascular system formed by the pial vascular plexus, formed at the optic disc by the union of various retinal
Fig. 3.68 Fundus photograph of left eye with central retinal

vein and retinociliary collaterals (arrows) on the optic disc
veins and runs back in the center of the optic nerve,

lying on the temporal side of the central retinal artery
in the optic nerve, enclosed in a fibrous envelope
(Fig. 3.71). During embryonic life, there are two
venous trunks lying on either side of the central retinal
Fig.3.66 A neoprene cast of the central retinal artery showing artery (Fig.3.72), which coalesce during the third tri-
the intraneural part and its retinal branches. Two very small mester to form the single trunk of the central retinal
branches posterior to the lamina cribrosa are shown from the
intraneural part (Reproduced from Hayreh [8])
vein [118]; however, in 20% that congenital anomaly
may persist during adult life [119]. In my anatomical
studies I found the two trunks of the central retinal
vein in some adult eyes the two lying on either side
of the central retinal artery in the anterior part of the
optic nerve [3]. Hughes [120] described two veins
passing backwards beside the central retinal artery in
the nerve, one emerging out of the nerve proximal to
the central retinal artery, the other one leaving the
nerve in the optic canal. Both pierce the sheath of the
optic nerve to join the veins of the orbit. When a dural
trunk of the central retinal vein is present during adult
life, occlusion of one of the two trunks results in devel-
opment of hemi-central retinal vein occlusion [121].
Originally the hyaloid artery is not accompanied by
any vein [118].
The site of exit of the vein from the optic nerve sub-
stance is very variable. The exit of the vein and the
entry of the central retinal artery may be situated on the
Fig.3.67 Fundus photograph of right eye of a rhesus monkey, same aspect [122126], with both at the same point
with malignant arterial hypertension, shows old cotton wool
spots, with development of multiple nerve fiber bundle loss in (Fig. 3.73) [124], or the artery anterior to the vein
their location (seen as dark semicircular bands arrows indicat- [124126] or vice versa [127]; these may be at different
ing some of them) (Reproduced from Hayreh etal. [152]) aspects with the vein usually on the lateral side [124]
Fig.3.69 Diagram (based

on Camera Lucida drawing),
from one of the specimens
in my anatomical study of
the central retinal artery
(CRA) in humans, shows
one of its intraneural
branches running backward
in the axial part of the optic
nerve posterior the central
retinal artery (*). CRA
central retinal artery, CZ
circle of Zinn and Haller,
PCA posterior ciliary
artery, (Reproduced from
Hayreh [153])
Fig.3.70 Neoprene cast of the central retinal vein (CRV) in my optic nerve and none in the optic nerve head region in this speci-
study, shows its entire course from the optic nerve head (ONH) men. At the left end are its retinal branches (Reproduced from
to its exit from the optic nerve (ON) sheath. Note the presence Hayreh etal. [154])
of a large number of prominent venous tributaries within the
having the latter anterior to the artery [124]. Charpy anterior to the artery in one, and posterior in two). In
and Hovelacque [122] found the site of exit of the vein the latter case, the artery and the vein diverge at right
usually 10mm. from the eyeball. Vossius [128] in 1883 angles in the centre of the nerve, to reach the pia differ-
demonstrated that the vessels cross each other at the ently. In 11% Fry [124] found the vein emerging on the
entrance/exit into the nerve. In nine specimens, where I lateral side of the optic nerve and then making an arc of
happened to study the vein, its exit was at the same a quarter of a circle in the pia on the nerve to come to
point as the artery in four (the vein anterior to the artery lie at the same aspect as the central retinal artery.
in three, and posterior in one), while in five the vein The vein runs on the optic nerve for some distance
exit was on the lateral side of the nerve with the entry before it pierces the sheath to come to lie in the orbit.
of the artery on the inferior aspect (both these sites at This course of the vein in the pia is usually longer than
the same distance from the eye in two, the vein exit that of the artery there; Fry [124] found it 3.88mm.
Fig.3.71 A transverse
section through the center of
the intraorbital part of the
optic nerve showing central
retinal artery (CRA) and vein
(CRV) lying side by side in
close apposition, enclosed by
a common fibrous tissue
envelope (FTE) (Reproduced
from Hayreh etal. [155])
long. It has been noticed that the vein pierces the dural the nerve are extensive and are big in size, going as far
sheath very gradually, so as to lie for a longer distance as the pia on the periphery (Fig.3.70).
in the sheath. The point of emergence of the vein from Cone and MacMillan (1932) described in the centre
the sheath has been described as always lying a consid- of the optic nerve a posterior central vein which,
erable distance posterior to the entrance of the artery, according to them, is formed by the union of a large
irrespective of which vessel pierced the nerve anteri- number of venules from the orbital part of the nerve.
orly. In my series the vein mostly pierced the dura pos- This vein leaves the optic nerve inferiorly, and runs
terior to the artery, though rarely it did so anteriorly. backwards on the lateral side of the nerve to open in the
When an optic nerve sheath meningioma compresses cavernous sinus [132]. According to Cone and
the central retinal vein, that results in the development MacMillan [133], this vein is constantly present and is
of well-known retinociliary collaterals (erroneously of the same size as the central retinal vein. They added
called optociliary shunts see above). that this vein plays an important role, not only in injury
On emerging from the optic nerve sheath, the vein to the optic nerve in the optic canal and causing hemor-
usually joins the orbital plexus of veins in the fat and rhages, but also by sudden compression in the sponta-
communicates with the superior and/ or inferior oph- neous subarachnoid hemorrhage, when it may lead to
thalmic veins or cavernous sinus directly, or there may marked remote changes in the retina. In my study, I
be a combination of all these [129]. Beauvieux and quite often saw a vein lying in the posterior segment of
Ristitch [123] found it communicating with the plexus the intraorbital part of the optic nerve behind the cen-
of veins on the optic nerve; this was also confirmed in tral retinal vein; it was very variable in size, and it
my study. The vein, therefore, has multiple connections joined the central retinal vein near about its bend in the
with the adjoining veins and a blockage of the cavern- center of the optic nerve. Its size was biggest where it
ous sinus is not likely to cause blockage of the blood joined the central retinal vein and diminished as it was
flow in the central vein as was originally thought by von traced backwards. Unlike Cone and MacMillan [133],
Graefe [130] in the pathogenesis of optic disc edema in I found that it ran from behind forward to join the cen-
raised intracranial pressure (refuted later by Sesemann tral retinal vein. Damel [134] described a similar vein
[131] on the basis of these multiple communications). which, according to him, drains the orbital portion of
The central retinal vein, in addition to drainage of the optic nerve and becomes superficial in the optic
the retina, also receives multiple tributaries from the canal to pass backwards to the cavernous sinus; no
ONH (Figs.3.3 and 3.70), and all along its course in other worker has described a similar vein and I have
the optic nerve. Its tributaries in the intraorbital part of never observed it.
C
a R S
D
A
Pia
PR
OD ON
LC
SAS
CRV
CAR
Optic nerve
Retina
CRV CRA
Dural Sheath
Sclera
Fig. 3.72 (a) Schematic representation of two trunks of the [121]) (b) Diagrammatic reconstruction from serial sections
c entral retinal vein in the optic nerve. (For abbreviations see (10 mm thick) of anterior part of the optic nerve in my study,
Fig.3.3.) A arachnoid, C choroid, CAR and CRA central retinal shows intraneural course of the central retinal vessels. Note
artery, CRV central retinal vein, D dura, LC lamina cribrosa, OD duplicate central retinal vein (CRV) in the anterior part of the
optic disc, ON optic nerve, PR and PLR prelaminar region, S optic nerve and the two trunks separating in the retina. CRA
sclera, SAS subarachnoid space (Reproduced from Hayreh etal. Central retinal artery
lood Supply of the Intracanalicular

B from the internal carotid, anterior cerebral and anterior
Part of the Optic Nerve communicating arteries, in common with the supply to
the intracranial part. Isayama etal. [117] in cast studies
According to Wolff [125], the ophthalmic artery supplies of monkeys found that the intracanalicular optic nerve
this part of the optic nerve. Behr [135] and Steele and was supplied by a centripetal vascular system, formed by
Blunt [126] also found small branches from the ophthal- the pial vessels arising from the first branches of the oph-
mic artery, arising just distal to the optic canal. However, thalmic artery. Superior and inferior vascular semicircles
Magitot [136] and Francois etal. [1] described its supply were detected in the intracanalicular optic nerve.
Blood Supply of the Intracranial Part of the Optic Nerve 69
Fig.3.74 Intraorbital (upper) and intracanalicular (lower) parts

of the optic nerve (ON) showing ophthalmic artery (OA) and
two collateral branches from the ophthalmic artery going to the
intracanalicular part of the optic nerve, as seen from below
lood Supply of the Intracranial Part

B
of the Optic Nerve
There is a great diversity of opinion about the source of

blood supply to this part. According to Behr [135] and
Francois etal. [1], the intracranial part is supplied by pial
branches from the internal carotid, anterior cerebral and
anterior communicating arteries. These nutrient vessels
Fig.3.73 Inferior surface of the intraorbital part of the optic
nerve (ON), showing the central retinal artery (CRA) and
supply the nerve centripetally and are end-arteries.
vein (CRV) and their site of penetration into the sheath of the Magitot [136] described a small artery, usually from the
optic nerve in rhesus monkey in my study (Reproduced from internal carotid artery, as the principal vascular supply to
Hayreh [156]) the intracranial portion of the nerve, and, according to
him, this artery terminates in the chiasma. He also
described a branch of the ophthalmic artery, dividing into
In my study [7, 85], I found that this part of the ascending and recurrent branches which supplied the
optic nerve is supplied by the ophthalmic artery upper surface of the nerve. Wolff [125] described ante-
entirely (Fig.3.15). Collateral branches, usually 13 rior cerebral and internal carotid arteries supplying supe-
and rarely even more, arise from the ophthalmic rior and inferior aspects of the nerve respectively, with
artery in the optic canal or occasionally near the apex the help of the anterior communicating and ophthalmic
of the orbit (Fig.3.74), and at times from the distal arteries. According to Dawson [137, 138] the prechias-
intraorbital part of the artery or its intraorbital mal artery, a branch from the ophthalmic artery near its
branches on the inferior aspect. These branches usu- origin, is the main supply, and it runs backwards on the
ally reach the optic nerve by traversing through the medial side of this part of the nerve to join its fellow of
connective tissue bands binding the optic nerve in the opposite side, along the anterior border of the chiasm
this part to the surrounding dural sheath in the optic and supplies the two. But, neither I [85] nor Steele and
canal (Fig. 3.75). Therefore, a fracture of the skull Blunt [126] could find any such vessel from the ophthal-
involving the optic canal is likely to tear away these mic artery. I [7] found a branch from the anterior superior
bands and interfere with the blood supply of the hypophysial branch of the internal carotid, which closely
nerve. No axial vascular system was seen in this and resembled the description of the prechiasmal artery of
the intracranial parts. Dawson [137, 138]. I [7] found that it runs forward on
section of the optic nerve in
the region of the optic canal,
showing a capillary
subarachnoid space (SAS)
and fibrous bands connecting
the optic nerve with the
surrounding sheath (Masson
trichrome staining)
the infero-medial aspect of the intracranial part of the brain, of which it is, in effect, an outgrowth; that is, all
nerve from the chiasma side, even winding to go on to the arteries which supply the optic nerve do so through
the medial side of the nerve, and ramifying in the pia the pial vascular plexus which acts like a distributing
(Fig.3.15). Steele and Blunt [126] found additional small centre for the vessels (Figs. 3.153.18). The vessels
recurrent branches from the ophthalmic artery to this part entering the nerve from the pia take a covering of the
of the nerve. Hughes [120] reported having frequently pia and glia with them, which forms the septal system
noted at operations branches of considerable size to the of the nerve (Fig.3.19), so that the distribution of the
superior aspect of the nerve by the anterior cerebral and septa is really a distribution of the blood vessels. The
ophthalmic arteries; however, he was not sure of their thickness of various septa depends upon the size of the
role in the blood supply. Isayama etal. [117], in their cast vessels they contain [125, 143]. The nerve is not only
studies on monkeys, found that the intracranial optic supplied by the pial vessels, but also to some extent by
nerve was supplied by branches of the internal carotid the intraneural branches of the central retinal artery
artery, anterior cerebral artery and/or anterior communi- anteriorly (Figs.3.2, 3.3, 3.64 and 3.65). Thus the fine
cating artery, and ophthalmic artery. In my studies [7], vascular system of the optic nerve is fed by both the
I found small collateral branches of variable numbers pial and the central retinal vessels, though mainly by
arising from the ophthalmic artery from its origin the former.
onwards [85], which ran backward on the inferior aspect The angioarchitecture in the region of the lamina
of the optic nerve and wound round the margins of the cribrosa has been described as comprising numerous,
optic nerve to go on to its superior aspect (Fig.3.15), so small, very close and transversely elongated meshes
as to supply the optic nerve all round [7]. However, the [139, 140], a very dense capillary plexus 1020mm in
branch from the anterior superior hypophysial artery, diameter [2, 46, 63, 126], which makes this part of the
mentioned above, was the biggest of all. ONH a highly vascular structure [46]. In this region,
according to Leber [139] and Parsons [140], the ves-
sels become smaller, with a more closely woven net-
work. Francois et al. [2, 63] found that capillaries
Angioarchitecture of the Optic Nerve become tortuous, corkscrewed, curly and wind spirally
in the optic nerve near the lamina cribrosa. In the lam-
Many investigators since 1903 have described this, on ina cribrosa, they found a very dense and tortuous
light as well as scanning electron microscopy [2, 7, 46, plexus with no difference between the periphery and
59, 60, 63, 64, 81, 125, 126, 139142]. The blood sup- centre in density. Steele and Blunt [126] described
ply of the optic nerve is planned similar to that of the small branches from the PCAs and choroid. According
Angioarchitecture of the Optic Nerve 71
to them, the capillary plexus is very dense here, and is intraneural branches within the optic nerve: no end-
continuous anteriorly with capillaries of the choroid artery appeared to exist.
and retina, and posteriorly with those of the adjoining In the intracranial part of the nerve, Francois etal.
optic nerve. In the region of the optic disc, Francois [2, 63] found fine independent arterioles entering the
etal. [2, 63] found disappearance of capillaries in the nerve from the periphery. These run transversely for
physiological cup while at the periphery the capillaries a short length and then longitudinally to divide in
were large, forming a lace like pattern winding in mul- branches. The medial side has denser visualization
tiple anastomoses. than the lateral side. The course of the capillaries is
In the intra-orbital and intra-canalicular part, straight in this part of the nerve, with irregular
Francois et al. [2, 63] found capillaries of different branching resulting in a complicated pattern. Hairpin-
sizes between the nerve bundles and described the like bends are seen on the lateral side. In this part
large ones as venous, medium arterial, and the small there is no association between myelo- and angio-
ones as serving a nutritional purpose for the nerve architecture, as seen elsewhere. Steele and Blunt
bundles. They described the existence of two basic [126] described an irregular capillary arrangement in
vascular patterns in this part of the nerve longitudi- this part, with frequent spiral formations. In contrast
nal and transverse. The longitudinal run in between to Francois etal. [2, 63] they described here a marked
the bundles from before backwards, while the trans- correspondence between the vessels and the glial
verse surround single nerve bundles at regular dis- septal arrangement. Fazio and Farina [141] also
tances and may be complete or incomplete pentagonal found a large number of very small parallel vessels
shaped. Oval capillary meshes surround nerve fibers, in this part.
going from one bundle to another. They [2, 63] stated In my rather limited study [7] of the fine angio-
that the most typical vascular pattern is seen in the architecture of the optic nerve, I found that in longi-
middle third of the intraorbital part of the nerve. tudinal sections of the nerve, the pattern varies in
Collaterals from both longitudinal and transverse different parts. Moreover, the network formed by
units join to form a complicated capillary plexus. The capillaries at places is so complicated that it is dif-
plexus is denser at the centre than at the periphery of ficult to work out a definite descriptive pattern.
the nerve. Fazio and Farina [141], on the other hand, Hence their interpretation varies from author to
found no such difference but found an irregular quad- author, and that may be responsible for the differing
rangular network of very small capillaries. In the pos- descriptions by various authors. However, in trans-
terior part of the optic nerve, Francois etal. [2, 63] verse sections of the nerve, the variation and confu-
found meshes wider and less tortuous than the middle sion is comparatively less.
third with no difference in their size at the centre and The angioarchitecture, like the gross blood supply,
periphery of the nerve. From the pia, larger arterioles is different in different parts of the nerve. The prelami-
penetrate into the nerve, and there are few collaterals nar region and surface nerve fiber layer of the ONH
from the longitudinal units. Steele and Blunt [126] have an irregular network, like the nerve posterior to
also described constant regional variation in the the lamina cribrosa (Fig.3.76). The peripapillary cho-
intrinsic capillary plexus pattern of the nerve. Awai riocapillaris do not anastomose with the capillary
[144] investigated the angioarchitecture of the mon- plexus in the prelaminar region (Figs.3.12 and 3.13);
key intracanalicular optic nerve by cast preparation. however, capillaries in the disc are continuous with
In the intracanalicular optic nerve, the vascular archi- those of the adjoining retina (Fig.3.76). In the lamina
tecture consisted of branches from the ophthalmic cribrosa, a horizontal section reveals the capillaries
artery. Some branches derived from the chiasmal part arranged mostly transversely, like the connective tis-
of the ophthalmic artery, some from its intracanalicu- sue arrangement in the lamina cribrosa, with a forward
lar part, and others from its intraorbital part. On the concavity (Fig.3.76).
surface of the intracanalicular optic nerve and in its In the intraorbital part of the optic nerve, the pattern
inner portion, vascular distribution was sparse. A is comparatively simpler, with longitudinal vessels con-
clear pial plexus was not observed. However, it was nected irregularly by irregular transverse vessels run-
ascertained that there existed a few anastomoses ning in various directions, forming irregular
between the branches from the ophthalmic artery on quadrangular, oval and multishaped meshes in a longi-
the surface of the optic nerve, and between the tudinal section (Figs.3.77 and 3.78). On the transverse
Fig.3.76 A longitudinal
the optic nerve and the
adjoining part of the eyeball,
showing fine angioarchitec-
ture. LC Lamina cribrosa,
Retrolaminar region
sections of the posterior
segment of the intraorbital
part of the optic nerve,
showing fine angioarchitec-
ture. The arrow indicates the
anterior end of the optic
nerve
Angioarchitecture of the Optic Nerve 73
Fig.3.78 A longitudinal
section of the anterior half of
the intracanalicular (ICON)
part and posterior segment
of the intraorbital part of the
optic nerve, showing the fine
angioarchitecture
almost similar to the adjoining intraorbital part, though

it becomes more and more complicated as traced
towards the cranial side, thus losing the more evident
differentiation into longitudinal and transverse
vessels.
In the intracranial region of the nerve (Figs.3.80
3.82), the distal part has a pattern similar to the
adjoining intracanalicular part, but as it is traced
backwards, longitudinal vessels become prominent
and transverse become scarce, so that in the posterior
part on the medial side, there are straight longitudinal
vessels lying nearly parallel with each other, with less
marked irregular collaterals (Figs. 3.80 and 3.81).
These prominent longitudinal vessels are not seen on
the lateral side, and become continuous posteriorly
with similar types of vessels in the anterior part of the
chiasm (Fig. 3.80), which are connected anteriorly
with big vessels in the optic nerve lying both in the
medial and lateral parts of the nerve. The vessels in
the proximal portion of the intracranial part are less
Fig. 3.79 A longitudinal sections of the posterior part of the dense, more so on the lateral side than on the medial
intracanalicular part of the optic nerve, showing fine angioarchi- side. The longitudinal vessels mentioned above are
tecture. Arrow indicates the anterior end of the nerve
not seen throughout the thickness of the nerve
(Fig. 3.82). Big vessels are seen frequently in the
section, complete or incomplete polygonal, shaped intracranial part and at times even in the intracanalicu-
meshes, surrounding usually a single nerve bundle but lar part of the nerve.
at times more than one, are seen. The longitudinal chan- In the optic nerve, the capillary vessels are
nels are usually more prominent than the transverse continuous all along, without any watershed or water-
ones; the latter invariably seem to connect the former. In tight compartment formation at any level. At the
the anterior part, on cross section the picture is not very capillary level, anastomoses exist within the optic
different, but a horizontal section reveals a more nerve throughout its entire length, and anteriorly
complicated pattern and the longitudinal and transverse with retinal capillaries; however, these anastomoses
vessels are not clearly made out, as in the previous one have no functional importance in the event of block-
(Fig.3.18). age of their main feeding vessels. This fact is
In the intracanalicular part of the optic nerve extremely important in understanding ischemic
(Figs. 3.78 and 3.79), the pattern of capillaries is optic neuropathy.
Fig.3.80 Fine angioarchi-

tecture of the whole of the
intracranial part of the optic
nerve (ON) with the
adjoining intracanalicular
part (ICON)
Fig.3.81 Fine angioarchi-

tecture of the intracranial part
of the optic nerve is seen with
well-marked longitudinal
straight capillaries. The
arrows indicates the anterior
end of the optic nerve
Conclusion
The above description about the blood supply of the

optic nerve shows that it is different in various
regions of the optic nerve, is highly complex, and
shows marked interindividual variation. It is a sub-
ject of paramount importance to understand the
pathogeneses and management of ischemic optic
neuropathies.
References
1. Franois J, Neetens A. Vascularization of the optic pathway.

I. Lamina cribrosa and optic nerve. Br J Ophthalmol. 1954;
38(8):47288.
2. Franois J, Neetens A, Collette JM. Vascular supply of the
optic pathway. II. Further studies by micro-arteriography of
the optic nerve. Br J Ophthalmol. 1955;39(4):22032.
3. Hayreh SS. A study of the central artery of the retina in
human beings. [Thesis for the Master of Surgery]. Panjab
University (India) 1958.
4. Singh (Hayreh) S, Dass R. The central artery of the retina. I.
Fig. 3.82 This section is from the same specimen as that of Origin and course. Br J Ophthalmol. 1960;44:193212.
Fig.3.81 but at a different level. Unlike the previous section, no 5. Singh (Hayreh) S, Dass R. The central artery of the retina. II.
longitudinal capillaries are seen here. The arrow indicates the A study of its distribution and anastomoses. Br J Ophthalmol.
anterior end of the optic nerve 1960;44:28099.
References 75
6. Hayreh SS. The central artery of the retina. Its role in the 29. Hayreh SS, Baines JA. Occlusion of the posterior ciliary
blood supply of the optic nerve. Br J Ophthalmol. 1963;47: artery. III. Effects on the optic nerve head. Br J Ophthalmol.
65163. 1972;56(10):75464.
7. Hayreh SS. Blood supply and vascular disorders of the optic 30. Hayreh SS, Baines JA. Occlusion of the vortex veins: an
nerve. An Inst Barraquer. 1963;4:7109. experimental study. Br J Ophthalmol. 1973;57(4):21738.
8. Hayreh SS. The orbital vessels of rhesus monkeys. Exp Eye 31. Hayreh SS, Weingeist TA. Experimental occlusion of the
Res. 1964;3:1630. central artery of the retina I. Ophthalmoscopic and fluores-
9. Hayreh SS. Blood supply of the optic nerve head and its role cein fundus angiographic studies. Br J Ophthalmol.
in optic atrophy, glaucoma, and oedema of the optic disc. Br 1980;64(12):896912.
J Ophthalmol. 1969;53(11):72148. 32. Hayreh SS, Chopdar A. Occlusion of the posterior ciliary
10. Hayreh SS. Pathogenesis of visual field defects. Role of artery. V. Protective influence of simultaneous vortex vein
the ciliary circulation. Br J Ophthalmol. 1970;54(5): occlusion. Arch Ophthalmol. 1982;100(9):148191.
289311. 33. Kishi S, Tso MO, Hayreh SS. Fundus lesions in malig-
11. Hayreh SS. Colour and fluorescence of the optic disc. nant hypertension. II. A pathologic study of experimen-
Ophthalmologica. 1972;165(2):1008. tal hypertensive optic neuropathy. Arch Ophthalmol.
12. Hayreh SS. Anterior ischaemic optic neuropathy. I. 1985;103(8):1198206.
Terminology and pathogenesis. Br J Ophthalmol. 1974; 34. Hayreh SS, Servais GE, Virdi PS. Fundus lesions in malig-
58(12):95563. nant hypertension. V. Hypertensive optic neuropathy.
13. Hayreh SS. Anterior ischaemic optic neuropathy. II. Fundus Ophthalmology. 1986;93(1):7487.
on ophthalmoscopy and fluorescein angiography. Br J 35. Hayreh SS, Servais GE, Virdi PS. Fundus lesions in malig-
Ophthalmol. 1974;58(12):96480. nant hypertension. VI. Hypertensive choroidopathy.
14. Hayreh SS. The choriocapillaris. Albrecht Von Graefes Arch Ophthalmology. 1986;93(11):1383400.
Klin Exp Ophthalmol. 1974;192(3):16579. 36. Hayreh SS, Bill A, Sperber GO. Metabolic effects of high
15. Hayreh SS. Submacular choroidal vascular pattern. intraocular pressure in old arteriosclerotic monkeys. Invest
Experimental fluorescein fundus angiographic studies. Ophthalmol Vis Sci. 1991;32(4):810.
Albrecht Von Graefes Arch Klin Exp Ophthalmol. 37. Hayreh SS, Bill A, Sperber GO. Effects of high intraocular
1974;192(3):18196. pressure on the glucose metabolism in the retina and optic
16. Hayreh SS. The long posterior ciliary arteries. An experi- nerve in old atherosclerotic monkeys. Graefes Arch Clin
mental study. Albrecht Von Graefes Arch Klin Exp Exp Ophthalmol. 1994;232(12):74552.
Ophthalmol. 1974;192(3):197213. 38. Hayreh SS, Piegors DJ, Heistad DD. Serotonin-induced
17. Hayreh SS. Segmental nature of the choroidal vasculature. constriction of ocular arteries in atherosclerotic monkeys:
Br J Ophthalmol. 1975;59:63148. implications for ischemic disorders of the retina and optic
18. Hayreh SS. Anterior ischemic optic neuropathy. New York: nerve head. Arch Ophthalmol. 1997;115(2):2208.
Springer; 1975. 39. Hayreh SS, Podhajsky PA, Zimmerman B. Ocular manifes-
19. Hayreh SS. Anterior ischemic optic neuropathy. Arch tations of giant cell arteritis. Am J Ophthalmol. 1998;125(4):
Neurol. 1981;38:6758. 50920.
20. Hayreh SS. Inter-individual variation in blood supply of the 40. Hayreh SS, Jonas JB. Optic disk and retinal nerve fiber layer
optic nerve head. Doc Ophthalmol. 1985;59:21746. damage after transient central retinal artery occlusion: an
21. Hayreh SS. Microangioarchitecture of optic papilla [com- experimental study in rhesus monkeys. Am J Ophthalmol.
ment]. Jpn J Ophthalmol. 1989;33(4):51925. 2000;129(6):78695.
22. Hayreh SS. Anterior ischaemic optic neuropathy. 41. Hayreh SS. Pathogenesis of optic nerve damage and visual
Differentiation of arteritic from non-arteritic type and its field defects. In: Heilmann K, Richardson KT, editors.
management. Eye. 1990;4(Pt 1):2541. Glaucoma: conceptions of a disease. Stuttgart: Georg Theime
23. Hayreh SS. In vivo choroidal circulation and its watershed Publishers; 1978. p. 10437.
zones. Eye. 1990;4(Pt 2):27389. 42. Ernest JT, Potts AM. Pathophysiology of the distal portion
24. Hayreh SS, Walker WM. Fluorescent fundus photography in of the optic nerve. II. Vascular relationships. Am J
glaucoma. Am J Ophthalmol. 1967;63(5):9829. Ophthalmol. 1968;66(3):3807.
25. Hayreh SS, Perkins ES. Clinical and experimental studies on 43. Hayreh SS. Optic disc changes in glaucoma. Br J Ophthalmol.
the circulation at the optic nerve head. Proceedings of the 1972;56(3):17585.
Wm MacKenzie Centenary Symposium on the Ocular 44. Hayreh SS. Structure and blood supply of the optic nerve. In:
Circulation in Health and Disease, Glasgow. London, Heilmann K, Richardson KT, editors. Glaucoma: conceptions of
Kimpton; 1968. p. 7186. a disease. Stuttgart: Georg Theime Publishers; 1978. p. 7896.
26. Hayreh SS, Perkins ES. The effects of raised intraocular 45. Henkind P, Levitzky M. Angioarchitecture of the optic nerve.
pressure on the blood vessels of the retina and optic disc. I. The papilla. Am J Ophthalmol. 1969;68(6):97986.
Proceedings of the International Symposium Fluorescein 46. Levitzky M, Henkind P. Angioarchitecture of the optic nerve.
Angiography, Albi. Basel: Karger; 1969. p. 3238. II. Lamina cribrosa. Am J Ophthalmol. 1969;68(6):98696.
27. Hayreh SS, Revie IH, Edwards J. Vasogenic origin of visual 47. Anderson DR. Vascular supply to the optic nerve of pri-
field defects and optic nerve changes in glaucoma. Br J mates. Am J Ophthalmol. 1970;70(3):34151.
Ophthalmol. 1970;54(7):46172. 48. Swietliczko I, David NJ. Fluorescein angiography in experimental
28. Hayreh SS, Baines JA. Occlusion of the posterior ciliary ocular hypertension. Am J Ophthalmol. 1970;70(3):35163.
artery. I. Effects on choroidal circulation. Br J Ophthalmol. 49. Theodossiadis GP. Uber die Vaskularisation in der Regio
1972;56(10):71935. praelaminaris der Papilla optica. [Vascularization in the
prelaminar region of the optic papilla]. Klin Monatsbl nervehead in relation to the arcuate scotoma in glaucoma.
Augenheilkd. 1971;158(5):64652. Can J Ophthalmol. 1972;7(1):6874.
50. Araki M, Honmura S. The collateral communications of the 71. Best M, Blumenthal M, Galin MA, Toyofuku H. Fluorescein
retinal circulation with the choroidal circulation at the optic angiography during induced ocular hypertension in glau-
nerve head (in Japanese). Nippon Ganka Gakkai Zasshi. coma. Br J Ophthalmol. 1972;56(1):612.
1973;77(10):155766. 72. Best M, Toyofuku H. Ocular hemodynamics during induced
51. Yokochi K, Maruyama H, Sodeno Y. Microcirculation of the ocular hypertension in man. Am J Ophthalmol. 1972;
disc and peripapillary choroid (in Japanese). Nippon Ganka 74(5):9329.
Gakkai Zasshi. 1973;77(10):153442. 73. Blumenthal M, Gitter KA, Best M, Galin MA. Fluorescein
52. Evans PY, Shimizu K, Limaye S, Deglin E, Wruck J. angiography during induced ocular hypertension in man.
Fluorescein cineangiography of the optic nerve head. Trans Am J Ophthalmol. 1970;69(1):3943.
Am Acad Ophthalmol Otolaryngol. 1973;77(3):OP26073. 74. Blumenthal M, Best M, Galin MA, Toyofuku H. Peripapillary
53. Ito K. Fluorescein angiographic findings of normal optic choroidal circulation in glaucoma. Arch Ophthalmol.
disc (in Japanese). Nippon Ganka Gakkai Zasshi. 1971;86(1):318.
1973;77(10):154356. 75. Boyd TA, Rosen ES. A new method of clinical assessment of
54. Shimizu K, Yokochi K, Okano T. Fluorescein angiography an intraocular pressure sensitive ischaemic mechanism in
of the choroid. Jpn J Ophthalmol. 1974;18:97108. glaucoma. Can J Ophthalmol. 1970;5(1):125.
55. Araki M. Anatomical study of the vascularization of the 76. Franois J, de Laey JJ. Fluorescein angiography of the glau-
optic nerve (in Japanese). Nippon Ganka Gakkai Zasshi. comatous disc. Ophthalmologica. 1974;168(4):28898.
1975;79(2):1019. 77. De Freitas F, Morin JD. The changes in the blood supply
56. Ernest JT. Pathogenesis of glaucomatous optic nerve dis- of the posterior pole of rabbits with ocular hypertension.
ease. Trans Am Ophthalmol Soc. 1975;73:36688. Can J Ophthalmol. 1971;6(2):13942.
57. Anderson DR, Braverman S. Reevaluation of the optic disk 78. Kalvin HN, Hamasaki DI, Gass JD. Experimental glaucoma
vasculature. Am J Ophthalmol. 1976;82(2):16574. in monkeys. II. Studies of intraocular vascularity during
58. Schwartz B, Rieser JC, Fishbein SL. Fluorescein angio- glaucoma. Arch Ophthalmol. 1966;76(1):94103.
graphic defects of the optic disc in glaucoma. Arch 79. Laatikainen L. Fluorescein angiographic studies of the peri-
Ophthalmol. 1977;95(11):196174. papillary and perilimbal regions in simple, capsular and low-
59. Shimizu K, Ujiie K. Structure of ocular vessels. New York: tension glaucoma. Acta Ophthalmol Suppl. 1971;111:383.
Igaku-Shoin; 1978. p. 54. 709, 10824. 80. Ducournau D. Systematisation vasculaire de la choroide.
60. Risco JM, Grimson BS, Johnson PT. Angioarchitecture of Lyon: Association Corporative des Etudiants en Medecine
the ciliary artery circulation of the posterior pole. Arch de Lyon; 1979. p. 1722.
Ophthalmol. 1981;99(5):8648. 81. Fryczkowski AW, Grimson BS, Peiffer Jr RL. Scanning
61. Fryczkowski AW. Mikrokrzenie w nerwie wzrokowym. II. electron microscopy of vascular casts of the human scleral
Blaszka sitowa [Optic nerve microcirculation. III. Lamina lamina cribrosa. Int Ophthalmol. 1984;7(2):95100.
cribrosa]. Klin Oczna. 1991;93(23):750. 82. Olver JM, Spalton DJ, McCartney AC. Microvascular study
62. Onda E, Cioffi GA, Bacon DR, Van Buskirk EM. of the retrolaminar optic nerve in man: the possible signifi-
Microvasculature of the human optic nerve. Am J Oph cance in anterior ischaemic optic neuropathy. Eye. 1990;4
thalmol. 1995;120(1):92102. (Pt 1):724.
63. Franois J, Neetens A. Vascularization of the optic pathway. 83. Wolff E. The blood supply to the lamina cribrosa (a correc-
III. Study of intra-orbital and intracranial optic nerve by tion). Trans Ophthalmol Soc UK. 1940;60:69.
serial sections. Br J Ophthalmol. 1956;40(1):4552. 84. Franois J, Neetens A. Central retinal artery and central
64. Lieberman MF, Maumenee AE, Green WR. Histologic stud- optic nerve artery. Br J Ophthalmol. 1963;47(1):2130.
ies of the vasculature of the anterior optic nerve. Am J Oph 85. Hayreh SS. The ophthalmic artery: III. Branches. Br
thalmol. 1976;82(3):40523. J Ophthalmol. 1962;46(4):21247.
65. Zhao Y, Li FM. Microangioarchitecture of optic papilla. Jpn 86. Hayreh SS. The 1994 Von Sallman lecture: the optic nerve
J Ophthalmol. 1987;31(1):14759. head circulation in health and disease. Exp Eye Res.
66. Cioffi GA, Van Buskirk EM. Microvasculature of the ante- 1995;61(3):25972.
rior optic nerve. Surv Ophthalmol. 1994;38(Suppl):S10716. 87. Hayreh SS. The blood supply of the optic nerve head and the
discussion S11617. evaluation of it Myth and Reality. Prog Retin Eye Res.
67. Cioffi GA, Van Buskirk EM. Vasculature of the anterior 2001;20:56393.
optic nerve and peripapillary choroid. In: Ritch R, Shields 88. Hayreh SS. Posterior ciliary artery circulation in health and
MB, Krupin T, editors. The Glaucomas. 2nd ed. St. Louis: disease: the Weisenfeld Lecture. Invest Ophthalmol Vis Sci.
Mosby; 1996. p. 17788. 2004;45:74957.
68. Hayreh SS. Blood supply of the optic nerve head in health and 89. Ruysch F. De oculorum tunicis. Amsterdam: Janssonio-
disease. In: Lambrou GN, Greve EL, editors. Ocular blood flow Waesbergios; 1737.
in glaucoma. Amsterdam: Kugler & Ghedini; 1989. p. 354. 90. Wybar KC. Vascular anatomy of the choroid in relation to
69. Hayreh SS. Retinal and optic nerve head ischemic disorders selective localization of ocular disease. Br J Ophthalmol.
and atherosclerosis: the role of serotonin. Prog Retin Eye 1954;38:51327.
Res. 1999;18(2):191221. 91. Duke-Elder S, Wybar KC. In: System of ophthalmology.
70. Begg IS, Drance SM, Goldman H. Fluorescein angiography vol. 2, The anatomy of the visual system. London:
in the evaluation of focal circulatory ischaemia of the optic Kimpton; 1961. p. 351.
References 77
92. Hayreh SS. Physiological anatomy of the choroidal vascu- 113. Hayreh SS. Acute ischemic disorders of the optic nerve:
lar bed. Int Ophthalmol. 1983;6:8593. pathogenesis, clinical manifestations, and management.
93. Giuffre G. Main posterior watershed zone of the choroid. Ophthalmol Clin North Am. 1996;9(3):40742.
Variations of its position in normal subjects. Doc 114. Hayreh SS, Zimmerman B. Visual field abnormalities in
Ophthalmol. 1989;72(2):17580. nonarteritic anterior ischemic optic neuropathy: their
94. Sato Y, Tomita G, Onda E, Goto Y, Oguri A, Kitazawa Y. pattern and prevalence at the initial presentation. Arch
Association between watershed zone and visual field defect Ophthalmol. 2005;123:155462.
in normal tension glaucoma. Jpn J Ophthalmol. 2000; 115. Rootman J, Butler D. Ischaemic optic neuropathy: a
44(1):3945. combined mechanism. Br J Ophthalmol. 1980;64(11):
95. Olver JM. Functional anatomy of the choroidal circulation: 82631.
methyl methacrylate casting of human choroid. Eye Lond. 116. Feldon SE. Anterior ischemic optic neuropathy: trouble
1990;4(Pt 2):26272. waiting to happen. Ophthalmology. 1999;106(4):
96. Haller A. Arteriarum oculi historia et tabulae arteriarum 6512.
oculi. Gottingen; 1754. 117. Isayama Y, Hiramatsu K, Asakura S, Takahashi T. Posterior
97. Olver JM, Spalton DJ, McCartney AC. Quantitative ischemic optic neuropathy. I. Blood supply of the optic
morphology of human retrolaminar optic nerve vascula- nerve. Ophthalmologica. 1983;186(4):197203.
ture. Invest Ophthalmol Vis Sci. 1994;35(11):385866. 118. Mann IC. Development of the human eye. Cambridge:
98. Gauntt CD. Peripapillary circle of Zinn-Haller. Br J Cambridge University Press, Br J Ophthalmol. 1928;
Ophthalmol. 1998;82(7):849. p. 230234.
99. Gauntt CD, Williamson TH, Sanders MD. Relationship of 119. Chopdar A. Dual trunk central retinal vein incidence in
the distal optic nerve sheath to the circle of Zinn. Graefes clinical practice. Arch Ophthalmol. 1986;104(8):112830.
Arch Clin Exp Ophthalmol. 1999;237(8):6427. 120. Hughes B. Blood supply of the optic nerves and chiasma
100. Strek P, Strek W, Nowogrodzka-Zagorska M, Pitynski K. and its clinical significance. Br J Ophthalmol. 1958;42
Angiomorphology of the retrobulbar part of the optic nerve. (2):10625.
A scanning electron microscopy of vascular casts. Folia 121. Hayreh SS, Hayreh MS. Hemi-central retinal vein occlu-
Morphol Warsz. 1996;55(3):14350. sion. Pathogenesis, clinical features, and natural history.
101. Heimann K. The development of the choroid in man. Arch Ophthalmol. 1980;98:16009.
Ophthalmic Res. 1972;3:25773. 122. Charpy A, Hovelacque A. Veines. In: Nicolas A, editor.
102. Heimann K. Enterschunger zur Entwicklung der menschli- Traite dAnatomie Humaine de Poirier et Charpy. 3rd ed.
chen Aderhaut [Development of the human choroid]. Adv Paris: Masson; 1920;2. pp. 53278.
Ophthalmol. 1974;28:3077. 123. Beauvieux J, Ristitch K. Les vaisseaux centraux du nerf
103. Ko MK, Kim DS, Ahn YK. Peripapillary circle of Zinn- optique. Etude anatomique. Arch Ophtal Paris. 1924;41:
Haller revealed by fundus fluorescein angiography. Br J 35269.
Ophthalmol. 1997;81(8):6637. 124. Fry WE. Variations in the intra-neural course of the central
104. Park KH, Tomita G, Onda E, Kitazawa Y, Cioffi GA. In vivo vein of the retina. Arch Ophthalmol. 1930;4(2):1807.
detection of perineural circular arterial anastomosis (circle of 125. Wolff E. Some aspects of the blood supply of the optic
Zinn-Haller) in a patient with large peripapillary chorioretinal nerve. Trans Ophthalmol Soc UK. 1939;59:15762.
atrophy. Am J Ophthalmol. 1996;122(6):9057. 126. Steele EJ, Blunt MJ. The blood supply of the optic nerve
105. Ohno-Matsui K, Morishima N, Ito M, Yamashita S, and chiasma in man. J Anat. 1956;90(4):48693.
Futagami S, Tokoro T, etal. Indocyanine green angiogra- 127. Paton L, Holmes G. The pathology of papilloedema. Brain.
phy of retrobulbar vascular structures in severe myopia. 1911;33:389432.
Am J Ophthalmol. 1997;123(4):494505. 128. Vossius A. Beitrge zur Anatomie des Nervus Opticus.
106. Araki M. The role of blood circulation of prelaminar capil- Albrecht Von Graefes Arch Ophthalmol. 1883;29(4):
laries in producing glaucomatous cupping (in Japanese). 11950.
Nippon Ganka Gakkai Zasshi. 1976;80(4):2017. 129. Hayreh SS. Orbital vascular anatomy. Eye. 2006;20:
107. Hayreh SS. Factors determining the glaucomatous optic 113044.
nerve head damage. In: Krieglstein GK, editor. Glaucoma 130. Graefe AV. Ueber Complication von Sehnervenentzndung
Update III. Heidelberg: Springer; 1987. p. 406. mit Gehirnkranheiten. Albrecht Von Graefes Arch Klin
108. Petrig BL, Riva CE, Hayreh SS. Laser Doppler flowmetry Exp Ophthalmol. 1860;7(2):5871.
and optic nerve head blood flow. Am J Ophthalmol. 131. Sesemann E. Die Orbitalvenen des Menschen und ihr
1999;127(4):41325. Zusammenhang mit den oberflchlichen Venen des Kopfes.
109. Franois J. Vascularization of the optic nerve. Arch Arch Anat Physiol Wiss Med. 1869;2:15473.
Ophthalmol. 1977;95(3):520. 132. Greeff R. Die microskopische, Anatomie des Sehnerven
110. Franois J, Fryczkowski A. Microcirculation of the anterior und der Netzhaut. In: Graefe-Saemisch Handbuch der
part of the optic nerve. Ophthalmologica. 1977;175(4): Auenheilk. Bd. 1, Kapitel V. 1900.
2229. 133. Cone W, Mac Millan JA. The optic nerve and papilla. In:
111. Franois J, Fryczkowski A. The blood supply of the optic Penfield W, editor. Cytology and cellular pathology of the
nerve. Adv Ophthalmol. 1978;36:16473. nervous system, vol. 2. New York: Hoeber; 1932.
112. Franois J, Fryczkowski A. Functional importance of cen- 134. Damel CS. Anatomia de la arteria y vena central de la ret-
tral retinal artery anastomoses in the anterior part of the ina en su porcion orbitaria y neural. Arch de Oft de Buenos
optic nerve. Ophthalmologica. 1982;185(1):1525. Aires. 1936;11(3):5782.
135. Behr C. Beitrag zur Anatomie und Klinik des septalen 147. Hayreh SS. Blood supply of the optic nerve head: a reality
Gewebes und des Arterieneinbaus im Sehnervenstamm. check. In: Pillunat LE etal., editors. Current concepts on
Arch f Ophthal. 1935;134:22767. ocular blood flow in glaucoma. The Hague: Kugler; 1999.
136. Magitot A. Sur lorigine intra-crnienne de latrophie p. 331.
optique glaucomateuse. Ann dOcul. 1947;180:32141. 148. Amalric P. The choriocapillaris in the macular area. Clinical
cited in Ophthal. Lit. 1, 291292, 1947. and angiographic study. Int Ophthalmol. 1983;6:14953.
137. Dawson BH. On the blood vessels of the human optic chi- 149. Hayreh SS. Ischemic optic neuropathy. Prog Retin Eye
asma, hypophysis and hypothalamus. M. D. Thesis, Res. 2009;28:3462.
Manchester; 1948. 150. Hayreh SS. Acute ischemia of the optic nerve. In: Bernstein
138. Dawson BH. The blood vessels of the human optic chiasma EF, editor. Amaurosis Fugax. New York: Springer-Verlag;
and their relation to those of the hypophysis and hypothala- 1988. p. 93113.
mus. Brain. 1958;81(2):20717. 151. Hayreh SS. Blood supply of the optic nerve head in health
139. Leber T. Die Circulations- und Ernhrungsverhltnisse and disease. In: Lambrou GN, Greve EL, editors. Ocular
des Auges. Graefe-Saemisch Handbuch der Gesamten blood flow measurements in glaucoma. Amsterdam:
Augenheilkunde, vol. 2(2). 2nd ed. Leipzig: Wilhelm Kugler& Ghedini Publications; 1989. p. 146.
Engelmann; 1903. 152. Hayreh SS etal. Cotton-wool spots (inner retinal ischemic
140. Parsons JH. The pathology of the eye, vol. 3. London: spots) in malignant arterial hypertension. Ophthalmologica.
Hodder and Stoughton; 1906. p. 94950. 344. 1989;198:197215.
141. Fazio C, Farina P. Sull angioarchitettonica del nervo ottica, 153. Hayreh SS. Posterior ischaemic optic neuropathy: clinical
del chiasma, della bandelleta. Rev Oto-neuro-oftal. 1940; features, pathogenesis, and management. Eye. 2004;
17:3854. 18:1188206.
142. Franois J, Neetens A. The fine angio-architecture of the 154. Hayreh SS, Podhajsky PA, Zimmerman MB. Natural his-
anterior optic nerve. In: Kant JS, editor. The optic nerve: tory of visual outcome in central retinal vein occlusion.
proceedings second Mackenzie symposium. London: Ophthalmology. 2011;118(1):11933.
Kimpton; 1972. p. 2839. 155. Hayreh SS et al. Morphologic changes in chronic high-
143. Henderson JW. Anatomy and physiology; symptoms and pressure experimental glaucoma in rhesus monkeys.
signs (Symposium. Diseases of the optic nerve). Trans Am JGlaucoma. 1999;8:5671.
Acad Ophthalmol Otolaryngol. 1956;60(1):813. 156. Hayreh SS. Occlusion of the central retinal vessels. Br J
144. Awai T. Angioarchitecture of intracanalicular part of mon- Ophthalmol. 1965;49:62645.
key optic nerve. Jpn J Ophthalmol. 1986;30(1):6373.
145. Hayreh SS. Anatomy and physiology of the optic nerve
head. Trans Am Acad Ophthalmol Otolaryngol. 1974;
78:OP24054. For a complete bibliography and detailed review of previously
146. Hayreh SS, Vrabec F. The structure of the head of the optic nerve published studies in the literature on the subject, please refer to
in rhesus monkey. Am J Ophthalmol. 1966;61(1):13650. bibliography in previous publications listed here.
Blood-Optic Nerve Barrier
4
The blood-optic nerve barrier is an important subject in Blood-Optic Nerve Barrier

any consideration of optic nerve and optic nerve head
(ONH) vascular disorders. The capillaries in the ONH
and the rest of the optic nerve, as in the retina [15] and The available studies have firmly established that blood
in the central nervous system [610] have tight cell junc- vessels in the optic nerve have a blood-optic nerve
tions between adjacent endothelial cells, without fenes- barrier [1219, 21, 24, 2628]. If we assume that the
tration [1113]. These tight junctions are responsible for blood-optic nerve barrier and blood-brain barrier have
the blood-optic nerve barrier, which has been shown by identical properties, then the barrier exercises a physi-
intravenous injection of tracer substances, such as horse- cal restraint on the passage of material from the vascu-
radish peroxidase [1416] and diaminoacridines [17]. lar system of the optic nerve into the extracellular
space surrounding it [29]. It allows a highly selective
passage of solutes from the blood into the extracellular
fluid of the optic nerve. This selectivity ensures that
Anatomical Considerations certain substances exchange with plasma only very
slowly while others, either because of their lipid solu-
Capillaries of certain tissues have tight cell junctions bility or their affinity for certain carriers in the mem-
with zonulae occludentes between adjacent cells; they brane, exchange more rapidly. Associated with these
control fluid and molecular movement and provide a two exchanges with blood, there is a relatively unre-
barrier effect against leakage of materials from the stricted and unselected exchange between the extracel-
capillaries, e.g., fluorescein, horseradish peroxidase, lular fluid and cerebrospinal fluid [29]. Some materials
diaminoacridine dyes, proteins, etc. The following tis- with large molecules are probably transported across
sues in the eye possess this barrier property: capillaries the endothelial cell from the capillary lumen by intra-
of the retina and optic nerve, retinal pigment epithe- endothelial vesicular transport [3036] or, under path-
lium, and intermediary tissue of Kuhnt, which lies ological circumstances, by intraendothelial channels
between the outer layers of the retina and the prelami- [32, 3742]. Respiratory gases diffuse easily across
nar region of the ONH and is composed of glial cells. the cells. The barrier probably protectively excludes
From the point of view of the discussion in this chap- unwanted substances from the neural tissue and, by
ter, of interest here is the presence of this barrier in the keeping large molecules out of the tissue fluid, helps
optic nerve capillaries [1114, 1619], intermediary maintain normal osmotic pressure. In the retina a simi-
tissue of Kuhnt [15, 16, 20] and pial vessels [11]. The lar barrier, by keeping large molecules and proteins out
presence of blood-brain barrier is well-established. of the tissue, maintains a constant osmotic pressure
In contrast to that, certain other ocular tissues leak and prevents development of retinal edema and loss of
materials freely across them, e.g., choriocapillaris, retinal transparency. Rodriguez Peralta [17] stated that
which have fenestrations in their walls [1, 2123] and although the blood-optic nerve barrier makes the vessel
the border tissue of Elschnig (also known as the bor- wall impermeable to certain materials in the blood-
der tissue of Jacoby), which lies between the choroid stream, it permits the material to enter the bloodstream
and the ONH [15, 21, 24, 25]. from the extravascular compartment of the optic nerve

80 4 Blood-Optic Nerve Barrier
tissues. However, as discussed below, evidence from cells and optic nerve fibers of the prelaminar region
later studies does not support the one-way traffic and lamina cribrosa in 810min; Okinami etal. [20]
view of Rodriguez-Peralta [17]. confirmed their findings. Horseradish peroxidase and
other tracers did not enter the capillaries of the optic
nerve in any of these studies; it surrounded the blood
vessels. This seems to contradict the conclusions of
Blood-Optic Nerve Head Barrier Rodriguez-Peralta [17] that the blood-optic nerve bar-
rier of capillaries is only a one-way impermeability
The blood vessels in the ONH have a blood-optic system and permits materials from the optic nerve tis-
nerve barrier by virtue of having tight interendothe- sues to enter the bloodstream.
lial cell junctions in the blood vessels. However, the Thus, while blood vessels in the optic nerve have
Border Tissue of Elschnig, which separates the chor- the property of a blood-optic nerve barrier, diffusion of
oid from the ONH, lacks such a property. This has the fluid from the peripapillary choroid into the ONH
been demonstrated by a large number of workers, using would nullify all the advantages attributed to the barri-
a variety of techniques [15, 16, 20, 21, 2428, 43, 44]. ers in the ONH blood vessels. The only other place in
Through large pores in the choriocapillaris, much the central nervous system where there is no blood-
material leaks out and permeates freely through the brain barrier is the area postrema (a small tongue-
border tissue of Elschnig to enter the prelaminar region shaped area on the lateral wall of the fourth ventricle).
of the ONH. The most informative studies about the It is puzzling that the ONH, interposed between the
distribution of the extravascular tracer in the ONH are blood-retinal and chorio-retinal barrier (at the level of
those based on intravenous injection of horseradish retinal pigment epithelium) systems anteriorly and the
peroxidase and electronmicroscopy. Thirty seconds blood-optic nerve barrier system posteriorly in the
after injection in mice, Yamashita et al. [44] found orbital part of the optic nerve, does not have a barrier.
horseradish peroxidase in the lamina cribrosa and This unorthodoxy makes the ONH and area postrema
adjacent retrolaminar optic nerve, most intensely col- unique and the significance of lack of blood-brain bar-
lected 8min after the injection in the prelaminar region rier remains an enigma. However, the lack of blood-
and the lamina cribrosa. Tso etal. [15] enucleated the optic nerve barrier in the prelaminar region of the
eyes of rhesus monkeys 230min after injection and ONH plays an important role in the development of
found the following distribution of the tracer: The sur- some optic neuropathies.
face nerve fiber layer of the optic disc showed no tracer
outside its capillaries. There was no diffusion from the
peripapillary outer retina into the ONH because of
tight junctions between the astrocytes that make up the Clinical Importance of Absence of Blood-
marginal tissue of Kuhnt. The prelaminar region Optic Nerve Barrier in the Prelaminar
showed a diffuse distribution of the tracer in its extra- Region
cellular spaces, with a gradient from Bruchs mem-
brane into the adjacent ONH (minimal at the center).
The tracer was most thickly concentrated in the poste- There is evidence that absence of blood-optic nerve
rior part of the prelaminar region and aggregated barrier in the prelaminar region plays a role in the
densely in the glial columns and around the capillaries development of the following optic neuropathies.
in the ONH. The border tissue of Elschnig was densely
infiltrated. Compared to the prelaminar region, the dif-
fusion of the tracer was less extensive in the lamina
cribrosa and it diffused among collagen fibers. In the Hypertensive Optic Neuropathy
retrolaminar optic nerve, the tracer lay in the extracel-
lular spaces, with decreasing concentration toward the Patients with malignant arterial hypertension, toxemia
orbital part of the optic nerve. The findings of Tsukahara of pregnancy and renal disease can develop non-
and Yamashita [16] agree with the above observations; arteritic anterior ischemic optic neuropathy. Its patho-
they found the tracer in the intercellular spaces of glial genesis is discussed at length elsewhere [45]. Briefly,
Clinical Importance of Absence of Blood-Optic Nerve Barrier in the Prelaminar Region 81
from the available evidence, there is a strong indica- neuropathy (see Chap. 16). This has also been sug-
tion that the renin-angiotensin-aldosterone system gested by other authors [5254].
plays an important role in the development and main-
tenance of malignant arterial hypertension. Angiotensin
II is a powerful vasopressor agent. Since choriocapil-
Toxic Optic Neuropathy Due to Methanol
laris are very leaky, the plasma, along with angiotensin
and other vasoconstrictor agents in it, leaks freely into Intoxication
the choroidal interstitial fluid; there it causes vasocon-
striction and/or occlusion of the choroidal vessels as The pathogenesis of this is discussed at length elsewhere
was seen on fluorescein fundus angiography [46] and [55]. Briefly, in methanol intoxication, methanol is con-
in pathological studies [47] in all these eyes. The leak- verted into the toxic metabolite formate. Formate leaks
age of the angiotensin into the choroidal interstitial out from the choriocapillaris into the choroidal intersti-
fluid produces ischemia of the ONH by the following tial fluid, and then diffuses into the ONH through the
two mechanisms: Border Tissue of Elschnig (due to absence of blood-
optic nerve barrier in the prelaminar region of the ONH).
(a) As discussed above, there is no blood-optic nerve
Formate leaks out from the choriocapillaris into the
barrier in the prelaminar region of the ONH.
choroidal interstitial fluid, and then diffuses into the
The angiotensin and other vasoconstrictors from
ONH through the Border Tissue of Elschnig (due to
the choroidal interstitial fluid would diffuse
absence of blood-optic nerve barrier in the prelaminar
into the ONH through the Border Tissue of
region of the ONH). Formate inhibits the cytochrome
Elschnig which is freely permeable. The vasocon-
oxidase enzyme mechanism in the tissue inhibition
strictors in the tissues of the optic nerve would
of ATP formation inhibition of electrical conduction
produce vasoconstriction and occlusion by direct
and visual dysfunction as well as axoplasmic flow sta-
action on the capillaries and other vessels in the
sis, by the mechanism discussed elsewhere [55].
ONH; this was also suggested by the experimental
Axoplasmic flow stasis results in optic disc swelling,
studies of Sossi and Anderson [48]. Pathologic
seen in this condition. The primary cause of visual loss
studies in these eyes showed swollen endothelial
in methanol intoxication is toxic optic neuropathy.
cells and degenerated pericytes with obliterated
lumen of vessels in the optic nerve [49].
(b) In addition, as a part of the constriction of the chor-
oidal vascular bed, caused by leaked vasoconstric- Non-arteritic Anterior Ischemic Optic
tors from the choriocapillaris, there is vasoconstriction Neuropathy
of the peripapillary choroid. As discussed in Chap.
3, the peripapillary choroid is the main source of
It is quite possible that in some cases of non-arteritic ante-
blood supply to the ONH. Vasoconstriction and
rior ischemic optic neuropathy, absence of a blood-optic
occlusion of the peripapillary choroid would sec-
nerve barrier in the prelaminar region may also be play-
ondarily cause ischemia of the ONH.
ing a role. For example, from time to time I have come
Ischemia of the axons results in axoplasmic flow sta- across patients (particularly persons with type A person-
sis which accumulates proximal to the ischemic site ality [56]) who complained of development of blurring of
and results in the axonal swelling at the optic disc vision and even development of non-arteritic anterior
[4951]. Pathological studies in these eyes revealed ischemic optic neuropathy when exposed to a stressful
that optic disc swelling was essentially due to isch- situation. There is a huge volume of literature showing
emic hydropic swelling of the axons [49]. Thus, stud- association of stress and release of catecholamines (nor-
ies strongly indicate that the optic disc edema epinephrine and epinephrine). Norepinephrine and epi-
(swelling) in malignant arterial hypertension is due to nephrine are vasoconstrictors, which can leak from the
ischemia of axons in the ONH, and hence represents choriocapillaris into the choroidal interstitial tissue, and
non-arteritic anterior ischemic optic neuropathy. The through the Border Tissue of Elschnig into the prelaminar
changes in the optic disc in these eyes follow the region, and result in ONH ischemia, by the mechanism
course seen in non-arteritic anterior ischemic optic discussed above in hypertensive optic neuropathy.
Effect of Hypoxia/Ischemia - VEGFR1 (Flt-1) and VEGFR2 (KDR/Flk-1) [69].

on Blood-Optic Nerve Barrier VEGFR1 and VEGFR2 activation is known to result in
disruption of tight endothelial junctions and may be
important in the pathogenesis of breakdown of the
Ischemic optic neuropathy is an ischemic disorder. blood-retinal barrier [70]. Inhibition of VEGF produc-
Therefore, it is relevant to know the effect of ischemia/ tion has been reported to reduce the blood-retinal bar-
hypoxia on the blood-optic nerve barrier. A recent rier permeability [58]. Clinically, the use of magnetic
review by Kaur etal. [57], of the effect of ischemia/ resonance imaging with gadolinium-diethylenetri-
hypoxia on the blood-retinal barrier has information aminepentaacetic acid may be a useful neurodiagnostic
which is also relevant to the blood-optic nerve barrier. procedure for visualization of increased permeability
There is disruption of the blood-retinal barrier in of the breakdown in blood-optic nerve barrier.
hypoxia/ischemia, which may be due to several mech- Corticosteroids have a specific effect in reducing
anisms, including increased production of vascular vascular permeability [71], in addition to having anti-
endothelial growth factor (VEGF), nitric oxide, free angiogenic, anti-inflammatory, anti-apoptotic, anti-
radicals, inflammatory mediators and hemodynamic proliferative and anti-edematous activity [72]. All this
alterations. Kaur et al. [57] reported an increase in is applicable to the ischemia of the optic nerve also.
AQ4 mRNA and protein specifically localized in astro- For example, in non-arteritic anterior ischemic optic
cytes in hypoxic rat retina in comparison to controls neuropathy corticosteroid therapy has been shown to
[58], supporting the view that these cells are involved result in quicker resolution of optic disc edema than in
in the transport of water from blood vessels to the reti- an untreated group, by reducing vascular permeability
nal tissues and, hence, may contribute to edema forma- [73]. Non-arteritic anterior ischemic optic neuropathy
tion in hypoxic conditions. Disruption of ionic eyes treated with corticosteroid therapy, because of
homeostasis, free radical production and energy fail- this faster resolution of optic disc edema, have a sig-
ure during hypoxia/ischemia are some of the factors nificantly better visual outcome than the untreated eyes
responsible for swelling of astrocytes. The optic nerve (see Chap. 17) [74]. Corticosteroid therapy in non-
has astrocytes (see Chap. 2) and that may also be hap- arteritic posterior ischemic optic neuropathy has also
pening with the astrocytes in the optic nerve. been shown to result in a significantly better visual
VEGF is known as a vascular permeability enhanc- outcome than the untreated eyes (see Chap. 22) [75].
ing factor [59, 60]. It promotes leakage of plasma pro-
teins from blood vessels [61]. It has been shown to
induce increased vascular permeability via synthesis or
release of nitric oxide [62]. VEGF also induces the
References
appearance of interendothelial gaps, fragmentation of
the endothelium and degenerative changes in endothe-
1. Cunha-Vaz JG, Shakib M, Ashton N. Studies on the perme-
lial basement membranes affecting the structural integ- ability of the blood-retinal barrier. I. On the existence, devel-
rity of the microvessels leading to extravasation of opment, and site of a blood-retinal barrier. Br J Ophthalmol.
blood plasma proteins, including albumin [59, 63]. 1966;50(8):44153.
VEGF is an important pathogenic factor for the devel- 2. Ishikawa T. Fine structure of retinal vessels in man and the
Macaque monkey. Invest Ophthalmol. 1963;2:115.
opment of vascular leakage, which is increased by 3. Lasansky A, Wald F. The extracellular space in the toad ret-
hypoxia, leading in the brain to the development of ina as defined by the distribution of ferrocyanide: a light and
cerebral edema [62, 64]. Increased expression of VEGF electron microscope study. J Cell Biol. 1962;15(3):46379.
has been reported in hypoxic retinal astrocytes [65]. In 4. Lasansky A. The pathway between hyaloid blood and retinal
neurons in the toad. Structural observations and permeabil-
the brain, breakdown of the blood-brain barrier has ity to tracer substances. J Cell Biol. 1967;34(2):61726.
been shown to trigger theactivation of astrocytes [66, 5. Shakib M, Cunha-Vaz JG. Studies on the permeability of the
67]. In the brain, up-regulation of endogenous VEGF blood-retinal barrier: IV. Junctional complexes of the retinal
in ischemia is believed to interact with receptors for vessels and their role in the permeability of the blood-retinal
barrier. Exp Eye Res. 1966;5(3):22934.
VEGF on ischemic vessels and contributes to disrup- 6. Barber VC, Graziadei P. The fine structure of cephalopod
tion and leakage of blood-brain barrier [68]. VEGF blood vessels. 3. Vessel innervation. Z Zellforsch Mikrosk
binds to, and activates, two tyrosine kinase receptors Anat. 1967;77(2):16274.
References 83
7. Mugnaini E, Walberg F. The fine structure of the capillaries 27. Laties AM, Rapoport S. The blood-ocular barriers under
and their surroundings in the cerebral hemispheres of Myxine osmotic stress. Studies on the freeze-dried eye. Arch
Glutinosa. Z Zellforsch Mikrosk Anat. 1965;66(3):33351. Ophthalmol. 1976;94(7):108691.
8. Reese TS, Karnovsky MJ. Fine structural localization of a 28. McMahon RT, Tso MO, McLean IW. Histologic localization
blood-brain barrier to exogenous peroxidase. J Cell Biol. of sodium fluorescein in human ocular tissues. Am J Oph
1967;34(1):20717. thalmol. 1975;80(6):105865.
9. Stensaas LJ, Stensaas SS. Astrocytic neuroglial cells, oligo- 29. Davson H. The blood-brain barrier. In: Bourne GH, editor.
dendrocytes and microgliacytes in the spinal cord of the toad The structure and function of nervous tissue, vol. IV. London:
II. Electron microscopy. Z Zellforsch Mikrosk Anat. 1968; Academic; 1972. p. 321445.
86(2):184213. 30. Beggs JL, Waggener JD. Transendothelial vesicular trans-
10. Wolff E. Beitrge zur Ultrastruktur der Kapillaren ion der port of protein following compression injury to the spinal
normalen Grosshirnrinde. Z Zellforsch Mikrosk Anat. 1963; cord. Lab Invest. 1976;34(4):42839.
60:40931. 31. Beggs JL, Waggener JD, Miller W. Vasogenic edema in the
11. Anderson DR. Ultrastructure of meningeal sheaths. Normal injured spinal cord: possible role of microtubules in endothe-
human and monkey optic nerves. Arch Ophthalmol. 1969; lial vasicular transport of HRP. In: Bailey GW, editor. Thirty-
82(5):65974. third Annual Proceedings of the Electron Microscopy
12. Anderson Dr. Ultrastructure of human and monkey lamina Society of America. Baton Rouge: Claitors Publishing
cribrosa and optic nerve head. Arch Ophthalmol. Division; 1975. p. 3989.
1979;82(6):80014. 32. Eto T, Omae T, Yamamoto T. An electron microscope study
13. Anderson DR, Hoyt WF. Ultrastructure of intraorbital por- of hypertensive encephalopathy in the rat with renal hyper-
tion of human and monkey optic nerve. Arch Ophthalmol. tension. Arch Histol Jpn. 1971;33(2):13343.
1969;82(4):50630. 33. Reyners H, de Reyners EG, Jadin JM, Maisin JR. An ultra-
14. Peyman CA, Apple D. Peroxidase diffusion processes in the structural quantitative method for the evaluation of the perme-
optic nerve. Arch Ophthalmol. 1972;88(6):6504. ability to horseradish peroxidase of cerebral cortex endothelial
15. Tso MOM, Shih C-Y, McLean IW. Is there a blood-brain cells of the rat. Cell Tissue Res. 1975;157(1):939.
barrier at the optic nerve head? Arch Ophthalmol. 34. Westergaard E. Enhanced vesicular transport of exogenous
1975;93(9):81525. peroxidase across cerebral vessels, induced by serotonin.
16. Tsukahara I, Yamashita H. An electron microscopic study Acta Neuropathol. 1975;32(1):2742.
on the blood-optic nerve and fluid-optic nerve barrier. 35. Westergaard E. The effect of serotonin, norepinephrine and
Albrecht Von Graefes Arch Klin Exp Ophthalmol. 1975; cyclic AMP on the blood-brain barrier. J Ultrastruct Res.
196(3):23946. 1975;50:383.
17. Rodriguez-Peralta LA. Hematic and fluid barriers in the 36. Westergaard E, Brightman MW. Transport of proteins across
optic nerve. J Comp Neurol. 1966;126(1):10921. normal cerebral arterioles. J Comp Neurol. 1973;152(1):1744.
18. Olsson Y, Kristensson K. Permeability of blood vessels and 37. Baumbach GL, Cancilla PA, Hayreh MS, Hayreh SS.
connective tissue sheaths in retina and optic nerve. Acta Experimental injury of the optic nerve with optic disc swell-
Neuropathol. 1973;26(2):14756. ing. Lab Invest. 1978;39:5060.
19. Rodriguez-Peralta LA. The blood-optic nerve barrier. Anat 38. Bruns RR, Palade GE. Studies on blood capillaries. II.
Rec. 1963;145(2):277. Transport of ferritin molecules across the wall of muscle
20. Okinami S, Ohkuma M, Tsukahara I. Kuhnt intermediary capillaries. J Cell Biol. 1968;37(2):27799.
tissue as a barrier between the optic nerve and retina. 39. Hashimoto PH. Intracellular channels as a route for protein
Albrecht Von Graefes Arch Klin Exp Ophthalmol. 1976; passage in the capillary endothelium of the shark brain. Am
201(1):5767. J Anat. 1972;134(1):4157.
21. Cohen AI. Is there a potential defect in the blood-retinal bar- 40. Hashimoto PH, Takaesu S, Chazono M, Amano T. Vascular
rier at the choroidal level of the optic nerve canal? Invest leakage through intraendothelial channels induced by cholera
Ophthalmol. 1973;12(7):5139. toxin in the skin of guinea pigs. Am J Pathol. 1974;75(1):
22. Missotten L. Study of the capillaries of the retina and the 17180.
choroid capillaries with the electron microscope. Oph 41. Hirano A, Dembitzer HM, Becker NH, Levine S, Zimmerman
thalmologica. 1962;144:112. HM. Fine structural alterations of the blood-brain barrier in
23. Spitznas M. Anatomical correlation of the normal fluoroangiog- experimental allergic encephalomyelitis. J Neuropathol Exp
raphy of the fundus. Doc Ophthalmol Proc Ser. 1976;9: Neurol. 1970;29(3):43240.
17780. 42. Simionescu N, Siminoescu M, Palade GE. Permeability of
24. Grayson MC, Laties AM. Ocular localization of sodium muscle capillaries to small heme-peptides. Evidence for the
fluorescein. Effects of administration in rabbit and monkey. existence of patent transendothelial channels. J Cell Biol.
Arch Ophthalmol. 1971;85(5):6003 passim. 1975;64(3):586607.
25. Tsukahara I, Ota M. Histological study on the distribution of 43. Machemer R. Angiographic-histologic correlation of eye
sodium fluorescein in the fundus. Proceedings of the 4th vessel permeability with protein-bound fluorescent dye. Am
Congress European Society of Ophthalmology. Budapest; J Ophthalmol. 1970;69(1):2738.
1972. Basel: Karger; 1974. p. 30711. 44. Yamashita H, Miki H, Tsukahara I, Ogawa K. A histochemi-
26. Ben-Sira I, Riva CE. Fluorescein diffusion in the human cal study on the blood-optic nerve and fluid-optic nerve bar-
optic disc. Invest Ophthalmol. 1975;14(3):20511. rier. Acta Histochem Cytochem. 1973;6:1639.
45. Hayreh SS, Servais GE, Virdi PS. Fundus lesions in malig- 62. Mayhan WG. VEGF increases permeability of the blood
nant hypertension V. Hypertensive optic neuropathy. brain barrier via a nitric oxide synthase/cGMP-dependent
Ophthalmology. 1986;93(1):7487. pathway. Am J Physiol. 1999;276(5 Pt 1):C114853.
46. Hayreh SS, Servais GE, Virdi PS. Fundus lesions in malig- 63. Fischer S, Wobben M, Marti HH, Renz D, Schaper W.
nant hypertension VI. Hypertensive choroidopathy. Hypoxia-induced hyperpermeability in brain microvessel
Ophthalmology. 1986;93(11):1383400. endothelial cells involves VEGF-mediated changes in the
47. Kishi S, Tso MOM, Hayreh SS. Fundus lesions in malignant expression of zonula occludens-1. Microvasc Res. 2002;
hypertension I. A pathologic study of experimental hyper- 63(1):7080.
tensive choroidopathy. Arch Ophthalmol. 1985;103(8): 64. Schoch HJ, Fischer S, Marti HH. Hypoxia-induced vascular
118997. endothelial growth factor expression causes vascular leak-
48. Sossi N, Anderson DR. Blockage of axonal transport in age in the brain. Brain. 2002;125(Pt 11):254957.
optic nerve induced by elevation of intraocular pressure; 65. Kaur C, Sivakumar V, Foulds WS. Early response of neurons
effect of arterial hypertension induced by angiotensin I. and glial cells to hypoxia in the retina. Invest Ophthalmol
Arch OphthalmoI. 1983;101(1):947. Vis Sci. 2006;47(3):112641.
49. Kishi S, Tso MOM, Hayreh SS. Fundus lesions in malignant 66. Dietrich WD, Alonso O, Halley M. Early microvascular and
hypertension II. A pathologic study of experimental hyper- neuronal consequences of traumatic brain injury: a light and
tensive optic neuropathy. Arch Ophthalmol. 1985;103(8): electron microscopic study in rats. J Neurotrauma.
1198206. 1994;11(3):289301.
50. Hayreh SS. Fluids in the anterior part of the optic nerve in 67. Hirano A, Kawanami T, Llena JF. Electron microscopy of
health and disease. Surv Ophthalmol. 1978;23(1):125. the bloodbrain barrier in disease. Microsc Res Tech.
51. McLeod O, Marshall J, Kohner EM. Role of axoplasmic 1994;27(6):54356.
transport in the pathophysiology of ischaemic disc swelling. 68. Zhang ZG, Zhang L, Tsang W, Soltanian-Zadeh H, Morris
Br J OphthalmoI. 1980;64(4):24761. D, Zhang R, et al. Correlation of VEGF and angiopoietin
52. Ashton N. The eye in malignant hypertension. Trans Am expression with disruption of bloodbrain barrier and angio-
Acad Ophthalmol Otolaryngol. 1972;76(1):1740. genesis after focal cerebral ischemia. J Cereb Blood Flow
53. Garner A, Ashton N, Tripathi R, etal. Pathogenesis of hyper- Metab. 2002;22(4):37992.
tensive retinopathy. An experimental study in the monkey. 69. Shibuya M. Vascular endothelial growth factor receptor-1
Br J Ophthalmol. 1975;59(1):344. (VEGFR-1/Flt-1): a dual regulator for angiogenesis.
54. Dollery CT. Hypertensive retinopathy. In: Genest J, Kuchel Angiogenesis. 2006;9(4):22530.
O, Hamet P, Cantin M, editors. Hypertension: physiopathol- 70. Spoerri PE, Afzal A, Li Calzi S, Shaw LC, Cai J, Pan H,
ogy and treatment. 2nd ed. New York: McGraw-Hill; 1983. etal. Effects of VEGFR-1, VEGFR-2, and IGF-IR hammer-
p. 72332. head ribozymes on glucose-mediated tight junction expres-
55. Hayreh MS, Hayreh SS, Baumbach GL, Cancilla P, Martin- sion in cultured human retinal endothelial cells. Mol Vis.
Amat G, Tephly TR, et al. Methyl alcohol poisoning III. 2006;12:3242.
Ocular toxicity. Arch Ophthalmol. 1977;95(10):18518. 71. Nussenblatt RB, Whitcup SM, Palestine AG. Uveitis: funda-
56. Hayreh SS. Acute ischemic disorders of the optic nerve: mentals and clinical practice. St. Louis: Mosby; 1996.
pathogenesis, clinical manifestations and management. 72. Cunningham Jr ET, Adamis AP, Altaweel M, Aiello LP,
Ophthalmol Clin North Am. 1996;9(3):40742. Bressler NM, DAmico DJ, et al. Macugen Diabetic
57. Kaur C, Foulds WS, Ling EA. Bloodretinal barrier in hypoxic Retinopathy Study Group. A phase II randomised double-
ischaemic conditions: basic concepts, clinical features and masked trial of pegaptanib, an anti-vascular endothelial
management. Prog Retin Eye Res. 2008;27(6):62247. growth factor aptamer, for diabetic macular edema.
58. Kaur C, Sivakumar V, Yong Z, Lu J, Foulds WS, Ling EA. Ophthalmology. 2005;112(10):174757.
Bloodretinal barrier disruption and ultrastructural changes 73. Hayreh SS, Zimmerman MB. Optic disc edema in non-arter-
in the hypoxic retina in adult rats: the beneficial effect of itic anterior ischemic optic neuropathy. Graefes Arch Clin
melatonin administration. J Pathol. 2007;212(4):42939. Exp Ophthalmol. 2007;245(8):110721.
59. Dobrogowska DH, Lossinsky AS, Tarnawski M, Vorbrodt 74. Hayreh SS, Zimmerman MB. Non-arteritic anterior
AW. Increased bloodbrain barrier permeability and endothe- ischemic optic neuropathy: role of systemic corticosteroid
lial abnormalities induced by vascular endothelial growth therapy. Graefes Arch Clin Exp Ophthalmol. 2008;
factor. J Neurocytol. 1998;27(3):16373. 246(7):102946.
60. Croll SD, Ransohoff RM, Cai N, Zhang Q, Martin FJ, Wei T, 75. Hayreh SS. Posterior ischaemic optic neuropathy: clinical
etal. VEGF-mediated inflammation precedes angiogenesis features, pathogenesis, and management. Eye. 2004;18(11):
in adult brain. Exp Neurol. 2004;187(2):388402. 1188206.
61. Mark KS, Burroughs AR, Brown RC, Huber JD, Davis TP.
Nitric oxide mediates hypoxia-induced changes in paracel- For a complete bibliography and detailed review of previously
lular permeability of cerebral microvasculature. Am J published studies in the literature on the subject, please refer to
Physiol Heart Circ Physiol. 2004;286(1):H17480. bibliography in previous publications listed here.
Factors Influencing the Optic Nerve Head
Blood Flow 5
In ischemic optic neuropathies, ischemia is due to Thus, the ONH blood flow depends upon three
deranged blood flow in the optic nerve. To understand parameters: (a) BP, (b) IOP and (c) resistance to blood
their pathogenesis and management a firm grasp of flow. Before discussing the pathophysiology of these
the following is crucial: (i) the blood supply of the three factors at some length, it is essential to discuss
optic nerve (as discussed before see Chap. 3), and first the role of (1) autoregulation of blood flow and (2)
(ii) the factors that influence the blood flow of the endothelial derived vasoactive agents, both of which
optic nerve head (ONH). It is the lack of in-depth affect BP and resistance to blood flow.
knowledge of these two factors which is responsible
for most of the controversy and confusion on isch-
emicoptic neuropathies. The purpose of this chapter is
an attempt to present a brief overview of the current Autoregulation of Blood Flow
state of our knowledge about the factors that finally
determine the state of the circulation in the ONH, as Autoregulation plays a very important role in the con-
discussed previously [1, 2]. trol of blood flow in a tissue. Therefore, I will discuss
that first.
Pathophysiology of Factors Influencing

the ONH Blood Flow What Is Autoregulation of Blood Flow?
The following formula is used to calculate the ONH The goal of autoregulation in a tissue is to maintain
blood flow: relatively constant blood flow, capillary pressure and
nutrient supply, in spite of changes in perfusion pres-
sure. Autoregulation is due to alteration in resistance
Perfusion pressure
Flow = to blood flow, and that in turn is due to changes in the
Resistance to flow
tone of the blood vessels. It is generally thought that
the terminal arterioles regulate the resistance to flow,
Perfusion pressure=Mean arterial blood pressure that is, they dilate to increase the blood flow when the
(BP) minus venous BP in a vascular bed. Normally the perfusion pressure falls and constrict to reduce the
pressure in the central retinal vein at the optic disc is blood flow in arterial hypertension. But there is a limit
slightly higher than the intraocular pressure (IOP), so to how far the terminal arterioles can constrict or dilate,
that for all practical purposes, IOP is usually a good so autoregulation operates only within a certain critical
index of the ocular venous BP. In view of that, perfu- range of perfusion pressure, and breaks down when the
sion pressure is also equal to mean arterial BP in the perfusion pressure goes above or below this critical
ONH vessels minus the IOP. range (Fig.5.1). Thus, contrary to the general belief,
Mean BP=Diastolic BP+1/3 (systolic minus the presence of autoregulation does not automatically
diastolic BP). ensure consistent blood flow in a tissue at all times.

86 5 Factors Influencing the Optic Nerve Head Blood Flow
the capillary bed through the contractile properties of

Blood Flow
pericytes. The local pCO2 is one of several potential
Autoregulation Range
indicators of whether or not local blood flow is ade-
NORMAL quate for local tissue needs.
(b) Myogenic Hypothesis
According to this, the rise of intra-vascular pressure
causes vasoconstriction [17, 18] because stretching of
Absent Present Absent the vessel wall is counteracted by vasoconstriction in
the arterioles. The myogenic response is inhibited by
Perfusion Pressure calcium-channel inhibitors [19]. Endothelial derived
vasoactive agents (discussed below) are thought to be
Fig.5.1 A diagrammatic representation of blood flow autoregu- most probably responsible for initiating the myogenic
lation range at different perfusion pressures in normal persons. contraction [2022].
Absent and present denote absence or presence of the auto- (c) Neurogenic Hypothesis
regulation (Reproduced from Hayreh [3])
We do not have much evidence of this in the ONH,
since vessels in the retina and ONH have no auto-
Anderson and his colleagues have discussed the role nomicnerve supply but still have autoregulation. The
of pericytes in the capillaries in regulation of the choroid, by contrast, is richly supplied by the auto-
bloodflow autoregulation by virtue of the presence of nomic nerves [2339] and yet has no appreciable
the contractile proteins actin and myosin [4, 5]. Reti autoregulation.
nal capillary pericytes, by their endothelial-relaxing
factornitric oxide, can modulate the tone of pericytes
and thus have the potential to influence blood flow in
capillaries [6].
Blood Flow Autoregulation in the ONH
Early studies found no evidence of it [40, 41], but sub-

Mechanism(s) Responsible for Blood sequent experimental and clinical studies have demon-
Flow Autoregulation strated efficient autoregulation in the normal ONH and
its derangement in different conditions [26, 4257]. It
This is still not clear. It has been discussed at length by is generally believed that the choroidal vascular bed
Orgl etal. [7]. Briefly, there are three hypotheses: (which, like the ONH, is also supplied by the posterior
ciliary arteries) has no autoregulation [25, 29, 33, 40,
(a) Metabolic Hypothesis
5861]; however, some authors have claimed that
According to this, local arteriolar smooth muscle tone
choroidal autoregulation was found in rabbits [62],
is regulated by local concentration of metabolic prod-
pigeons [63] and rats [64]. It is well-established that
ucts, pO2 and pCO2, and plays a role in maintaining
the retinal blood flow has an efficient autoregulation
autoregulation. Accumulation of metabolites is due to
[26, 29, 33, 34, 40, 59, 60, 6567]. Thus the evidence
hypoxia or lack of ability to wash out such metabo-
shows that blood flow in both the retina and ONH has
lites, and that causes vasodilation [810]. Elevation of
autoregulation.
pCO2 does seem to be one of the factors in vasodilation
of arterioles [11, 12]. Changes in pCO2 markedly influ-
ence the peripheral resistance [13, 14] and the effect is
most marked in the arterioles [15]. Anderson and Davis
The Range of Perfusion Pressure over
[16] found that an increase in pCO2 caused acidosis of
the medium and relaxation of the pericytes in a revers- Which ONH Autoregulation Operates
ible, concentration-dependent manner. These findings
suggest that the regulation of blood flow to meet local As discussed above, autoregulation operates only
metabolic needs does not reside entirely in the arteri- over a critical range of perfusion pressure, so that
oles and precapillary sphincters, but at least in part in with a rise or fall of perfusion pressure beyond the
Blood Flow Autoregulation in the ONH 87
critical range, the autoregulation becomes ineffective

Blood Flow
and breaks down (Fig.5.1); under those circumstances
Autoregulation Range
the blood flow is directly proportional to the perfusion
pressure. Thus, it is essential to remember that auto-
regulation does not protect the ONH blood flow at all Hyperten Absent Present Absent
times. The range has been investigated by a number of
studies [26, 43, 46, 47, 52, 68]. In normal monkeys,
Normal Absent Present Absent
autoregulation in the ONH has been reported to be nor-
mal at perfusion pressure of 30 mmHg by Bill and
co-workers [46, 52] and >50 mmHg by Ernest [43], Perfusion Pressure
but it definitely breaks down below 30mmHg [68]; in
Fig.5.2 A diagrammatic representation of blood flow autoregu-
normal cats it has been reported to break down at lation range at different perfusion pressures in normal persons
25mmHg perfusion pressure [47]. In old, atheroscle- and in hypertensives (Hyperten). Absent and present denote
rotic rhesus monkeys ONH autoregulation was already absence or presence of the autoregulation (Reproduced from
defective at 3035mmHg perfusion pressure [69]. Hayreh [3])
(Fig.5.2). The subject is discussed in detail in the

section dealing with arterial hypertension below.
(b) Vascular Endothelial Vasoactive Agents: Recent
Breakdown of Blood Flow Autoregulation studies have shown that these agents play an
in the ONH important role in modulating the local vascular
tone and most probably also in blood flow auto-
(i) Level of Perfusion Pressure at which Auto regulation (see below).
regulation Breaks Down (c) Other Causes: There may well be other, undiscov-
The level of perfusion pressure below which autoregu- ered causes of derangement of the autoregulation
lation in the ONH breaks down has been reported vari- in the ONH. We still have much to learn.
ably in different experimental studies, varying from 15
to 30mmHg in young healthy animals [26, 43, 44, 46,
47, 52, 56, 57]. But this level may be very different in
Role of Deranged ONH Blood Flow
persons with various cardiovascular disorders, from
that in normal, healthy individuals. As mentioned Autoregulation in ONH Ischemic Disorders
above, in old, atherosclerotic rhesus monkeys ONH
autoregulation was already defective at 3035mmHg Autoregulation of blood flow in the optic nerve plays
perfusion pressure [69]. an important part in the development of ONH isch-
(ii) Causes of Derangement of the ONH Auto emia, including non-arteritic anterior ischemic optic
regulation neuropathy and glaucomatous optic neuropathy. In
This may occur due to many systemic and local causes, non-arteritic anterior ischemic optic neuropathy, in
including the aging process, arterial hypertension spite of the presence of autoregulation normally, the
(Fig.5.2), diabetes mellitus, marked arterial hypoten- ONH develops it because of a transient fall in perfu-
sion from any cause, arteriosclerosis, atherosclerosis, sion pressure during sleep [73]. This suggests that
hypercholesterolemia, vasospasm and probably defective autoregulation of the blood flow in the ONH,
regional vascular endothelial disorders [6972]. It is caused by a fall of blood pressure below a critical level,
also possible that many more, as yet unknown, causes may be partly responsible for the development of non-
can derange the autoregulation, and possibly some arteritic anterior ischemic optic neuropathy. Similarly,
persons are born with defective autoregulation like, it is well-established that the vascular bed in the cen-
perhaps, those patients who suffer from orthostatic tral nervous system normally has very efficient auto-
arterial hypotension for no apparent reason. regulation; in spite of that, in the brain, development of
(a) Arterial Hypertension: Arterial hypertension transient ischemic attacks or ischemic lesions in the
causes derangement of blood flow autoregulation event of a fall of perfusion pressure is well-known.
Thus, evidence is accumulating from both clinical (d) Renin-angiotensin System: has an important role
and experimental studies that although autoregulation in the control of arterial BP. It has been shown that
normally controls blood flow in the ONH, it can become this system also exists in the vessel wall and plays
defective under certain conditions. An ONH with poor a significant part in vasomotor control [7882].
autoregulation is at much greater risk of developing Angiotensin-1 can be synthesized by endothelial
ischemic damage than one with efficient autoregula- cells and is converted to angiotensin-2 (a power-
tion. Therefore, it would seem reasonable that derange- ful vasoconstrictor) by angiotensin converting
ment in the autoregulation of the ONH may play an enzyme (ACE) in the endothelial cells. ACE
important role in the ischemic damage to the ONH seen also inactivates bradykinin which stimulates the
in non-arteritic anterior ischemic optic neuropathy as release of nitric oxide and prostanoids. Thus, an
well as in glaucomatous optic neuropathy. There are, ACE-inhibitor would decrease angiotensin-2 as
however, still many unanswered questions on this topic. well as increase bradykinin activity, the latter
Unfortunately, so far, we do not have a clinical method resulting in increased formation of endothelial
to evaluate in man ONH autoregulation of blood flow. derived relaxing factors both mechanisms caus-
ing vasodilatation.
(e) Vascular Endothelial Growth Factor (VEGF):
A tremendous amount of literature has accumu-
Endothelial Derived Vasoactive Agents lated on this over the recent past. Its role in the
development of neovascularization is well known.
In ischemia/hypoxia, there is increased production
From the vasomotor and blood flow point of view, one
of VEGF. There is evidence to suggest that VEGF
of the most important discoveries has been the charac-
terization of the role of various endothelial derived causes increased vascular permeability. This sub-
vasoactive agents. These agents are formed by vascu- ject is discussed briefly in Chap. 4.
lar endothelium. (f) Mechanosensors: Endothelial cells not only regu-
late vasomotor function by the various endothelial
derived vasoactive agents but also can function as
mechanosensors and can transduce the mechani-
cal signals produced by the physical force of blood
The Various Known Endothelial Derived
flow into a biochemical signal to which the vessel
Vasoactive Agents can respond [22]; this may be responsible for flow
dependent changes in circulation.
These are prostanoids, nitric oxide, endothelins, angio-
tensins, oxygen free radicals, smooth muscle cell
hyperpolarization, thromboxane A2 and other agents
[7477]. Causes of Abnormalities of Endothelial
(a) Prostanoids: consist of prostacyclin (a vasodila- Derived Vasoactive Agents
tor), and antiplatelet aggregator and profibrinolytic
agents. Pathophysiologic alterations in endothelial cell struc-
(b) Nitric oxide: is the endothelial derived relaxing ture and/or function occur in most major cardiovascu-
factor and is synthesized from L-arginine by nitric lar diseases [77]. For example, endothelial changes are
oxide synthetase; the latters activity is modulated seen in atherosclerosis (see below), arterial hyperten-
by Ca2+ influx into the cell. sion [8385], diabetes mellitus [86, 87], ischemia [88
(c) Endothelins: are the most potent vasoconstrictors 90], and cardiovascular diseases [77]. Available data
and comprise endothelin-1, -2 and -3, which act suggest that essential hypertension decreases produc-
via the endothelin receptors A and B. Endothelin tion of whole-body nitric oxide below the basal condi-
receptor A is found on the vascular smooth muscle tions [85]. Ischemia impairs production of nitric oxide
and B on the vascular endothelial cells. One of the by the endothelial cells [8892] and may also bring on
ways the endothelin receptor A mediates vasocon- vasospasm by increased formation and binding of
striction is by increased influx of calcium ions. endothelin [9395]. It has been shown that when the
Endothelial Derived Vasoactive Agents 89
arterial oxygen falls to a critical level, the endothelial Endothelial dysfunction appears to have several causes
cells stop production of endothelial-derived relaxing and effects, including the following:
factor, resulting in vasoconstriction [96, 97].
(a) There is defective dilation of the atherosclerotic
arteries [102105], even though the endothelium
is intact, which may contribute to increases in
Role of Endothelial Derived Vasoactive vasoconstrictor response.
Agents in Regulation of Blood Flow (b) Vasodilatation in response to vasoactive products
released by activated platelets is attenuated or
reversed to vasoconstriction [106]; impaired
Vascular endothelium plays an important and dynamic
endothelium-dependent relaxation of atheroscle-
role in vasomotor function of both macro- and micro-
rotic arteries may contribute to augmented vaso-
vasculatures, including maintenance of vascular tone
constriction by serotonin released by platelet
and regulation of blood flow, and it also regulates plate-
aggregation on atherosclerotic plaque.
let function, coagulation and vascular growth [77].
(c) Increased destruction of nitric oxide may play an
Vascular tone depends upon a balance between the
important role in impairment of endothelium-
endothelial vasodilators (e.g., nitric oxide) and vaso-
dependent relaxation in atherosclerotic arteries
constrictors (e.g., endothelin) (Fig.5.3), so that reduced
[107]; thus, in atherosclerosis nitric oxide forma-
formation of vasodilators results in vasoconstriction
tion may be normal or even increased [107a] but
and vice versa. Therefore, endothelial cells play an
increased degradation of nitric oxide may result in
important role in modulating the microvascular tone
impaired vasodilation.
and blood flow autoregulation. In addition to vasomo-
(d) Although atherosclerotic lesions are confined to
tor function, endothelium also plays a role in regula-
arteries, abnormal endothelial function is also seen
tion of fibrinolysis, because plasminogen activators
in microcirculation in atherosclerotics [108].
and inhibitors are synthesized by the endothelium [98,
(e) Nitric oxide reduces adherence of platelets and
99], and in thrombus formation by affecting the plate-
leukocytes to endothelium [109] but, in athero-
let aggregation, adhesion and other properties through
sclerotic arteries, reduced formation or increased
prostacyclin, nitric oxide and other agents [77].
destruction of nitric oxide by the dysfunctional
endothelium may increase adherence of platelets
and consequently increase release of serotonin.
Atherosclerosis and Endothelial Derived
There is also evidence that changes in the smooth mus-
Vasoactive Agents
cles of atherosclerotic arteries play a role in vasomotor
dysfunction.
Atherosclerosis is a common cause of ischemic disor-
ders in the body, including the eye and ONH. In ath- (a) They make them less responsive to nitric oxide,
erosclerosis, endothelial dysfunction probably plays a which increases vasoconstriction [107].
critical role in vasomotor dysfunction [100, 101]. (b) They cause a modest nonspecific increase in
responses to several vasoconstrictor agents, includ-
ing thromboxane and endothelin-1.
Vasodilator Vasoconstrict (c) They produce a marked increase in the vasocon-
strictor effect of serotonin [101, 110, 111].
Nitric oxide Endothelin-1
Role of Endothelial Derived Vasoactive

Agents in ONH and Ocular Ischemic
Disorders
Fig.5.3 A diagrammatic representation of a balance between
the endothelial vasodilators (nitric oxide) and vasoconstrictors As far the ocular vascular bed is concerned, Bill [24]
(endothelin) modulating the local vascular tone recognized that nitrovasodilators act as vasodilators in
the eye. Nitric oxide is released in the ONH and retina. studies in glaucomatous optic neuropathy have shown
It is released from the endothelial cells of the blood the presence of many neurochemical agents; however,
vessels [112, 113], and in the retina by bipolar cells whether all those are the cause or effect of the neural
[114], amacrine cells [115] and ganglion cells [116]. A damage in the ONH still remains to be seen.
number of studies have reported the important role It has been postulated that an imbalance of endo
played by endothelin-1 and nitric oxide in modulating thelium-derived relaxing (nitric oxide) and contract-
the local vascular tone and regulating blood flow ing(endothelin-1) factors (Fig.5.3) may be important
(Fig.5.3) in the ophthalmic, posterior ciliary and reti- for the development of ocular vascular complications
nal arteries and retinal and ONH vessels [39, 70, 117 seen in diabetes mellitus, arterial hypertension, hyper-
126]; angiotensin-2 has been reported to cause lipidemia, arteriosclerosis and ischemia. Our studies on
contraction of isolated human posterior ciliary arteries hypertensive optic neuropathy [71] and choroidopa-
[127]. In larger ophthalmic vessels this is due to their thy [130] support the finding that vascular endothe-
effect on vascular smooth muscle cells and in small lium damage does occur in arterial hypertension.
vessels by their effect on pericytes. Meyer etal. [122] In our studies in atherosclerotic monkeys, serotonin
reported that in isolated ciliary arteries of pig local (5-hydroxytryptamine) caused vasoconstriction result-
anesthetics impair endothelial formation of nitric oxide ing in transient occlusion or impaired blood flow in the
from l-arginine after stimulation with bradykinin; the central retinal artery and/or posterior ciliary artery [131,
authors postulated that local anesthetic drugs interact 132]. We speculated that it is likely that platelets adhere
with vascular endothelium and cause endothelial dys- to and aggregate on atherosclerotic plaques and release
function. Ferrari-Dileo et al. [128] showed the pres- their vasoactive materials, including serotonin, and that
ence of angiotensin-2 binding receptors in cat retinal the released serotonin triggers vasoconstriction by its
arterioles and ONH capillaries and hypothesized that action on the smooth muscles of atherosclerotic arter-
microvascular tone and perhaps autoregulatory ies; this serotonin-induced vasoconstriction may play
responses of ONH capillaries might be influenced by an important role in the development of ischemic disor-
vasoactive substances, such as angiotensin-2, either ders of the retina and ONH in atherosclerosis. In our
leaking from the choroid or locally synthesized. atherosclerotic study, stopping the atherogenic diet
Studies suggest that nitric oxide in the choroid may abolished or markedly improved theserotonin induced
maintain basal blood flow [39, 125]. Arteriosclerosis, vasoconstriction within a few months.
hypercholesterolemia, aging, arterial hypertension,
diabetes mellitus, ischemia and other so far unknown
causes in the regional ocular and ONH vessels may be
associated with abnormalities in production of endothe- Experimental Endothelin-1 Studies
lial derived vasoactive agents, thereby influencing the
vascular resistance and blood flow in the ONH [70]. Cioffi and co-workers have investigated the effect of
Neufeld etal. [129] reported their findings on the dis- perineural chronic infusion of endothelin-1 on the
tribution of three types of nitric oxide synthase (an anterior optic nerve in rabbits and rhesus monkeys
enzyme involved in synthesis of nitric oxide) in nor- [133136]. They found that there was a dose-depen-
mal and moderately to markedly glaucomatous ONHs. dent vasoconstriction with chronic local application of
All three isoforms of nitric oxide synthase were pres- endothelin-1, associated with a significant (p=0.015)
ent in increased amounts in the glaucomatous ONH. decrease in blood flow, increased optic disc cupping in
The authors postulated that the increased presence of rabbits, and diffuse loss of axons in monkeys. Oku
nitric oxide synthase-1 and the induction of nitric oxide etal. [137] injected endothelin-1 into the posterior vit-
synthase-2 in astrocytes of the lamina cribrosa suggest reous twice weekly for 4 weeks in seven rabbit eyes
that the glaucomatous ONH is exposed to excessive and found reduction of blood flow, enlargement and
levels of nitric oxide which may be neurodestructive to excavation of the optic disc cup, and significant reduc-
the axons or, conversely, the increased amount of tion of visual evoked potential. The IOP was normal in
nitric oxide synthase-3 in vascular endothelial cells all these studies, showing that optic disc cupping can
may be neuroprotective, causing vasodilatation and develop at normal IOP when the ONH is subjected to
increased blood flow in the tissue. Histochemical chronic ischemia. Ota and Oku [138], in acute studies
Arterial Hypertension 91
in rabbits, similarly injected endothelin-1 in the vitre- very much over-estimates the actual BP in the ONH
ous and found that that induced reduction of visual vessels and capillaries. There is a fast drop in BP along
evoked potential which was reversed by intravenous the small arteries and arterioles, and a continued decrease
nicardipine (a calcium-channel blocker). They con- as the blood moves through the capillaries.
cluded that endothelin -1 regulates microcirculation by Both arterial hypertension and hypotension can
modulating Ca2+ channels and nicardipine blocks the influence the ONH blood flow in a number of ways.
effect of endothelin-1. Nishimura etal. [139] injected
endothelin-1 into the vitreous of cats and reported a
dose-related, sustained decrease in the ONH blood
flow; endothelin-1 injected intravenously, however, Arterial Hypertension
produced a brief dose-related increase in the blood
flow. They concluded that different populations of Both chronic hypertension and malignant hyperten-
endothelin receptors on vascular smooth muscles or sion can interfere with the ONH blood flow. Blood
endothelial cells would reflect the opposing effects. flow in the ONH may be altered by a number of mech-
Thus it seems that unless endothelin-1 reaches the vas- anisms in these patients, including the following.
cular smooth muscles of the ONH vessels, it cannot
1. Secondary Changes in Blood Flow Auto
produce the vasoconstriction attributed to it.
regulation
Arterial hypertension can interfere with the normal
autoregulation of the blood flow. This may be due to
Arterial Blood Pressure the following mechanisms:
(a) Due to Abnormalities in Endothelial Derived
As mentioned earlier, one of the factors which deter- Vasoactive Agents
mine the ONH blood flow is the perfusion pressure in Arterial hypertension produces endothelial abnormali-
the ONH vascular bed, and perfusion pressure in turn ties, and those can cause abnormalities in the produc-
depends primarily upon the arterial BP. Clinically, the tion of endothelial derived vasoactive agents,
BP is measured in the brachial artery; however, it is the particularly reduced production of nitric oxide. Vascular
BP in the ONH vessels (particularly the capillaries) that tone depends upon the balance between the vasocon-
is relevant for ONH blood flow. We [140] measured, by strictor (endothelin) and vasodilator (nitric oxide) (see
direct cannulation of the vessels, the BP in the ophthal- above). Decreased production of nitric oxide upsets the
mic artery and in the aorta in normal, healthy, adult rhe- balance, so that the vasoconstrictor endothelin becomes
sus monkeys simultaneously, and regression equations overactive and results in vasoconstriction (Fig. 5.3).
were computed to calculate the ophthalmic artery pres- That in turn would result in increased vascular resis-
sure from the aortic BP. These were as follows: tance and deranged autoregulation of blood flow.
(b) Due to Hypertension-Induced Adaptive Changes
Systolic ophthalmic artery pressure=0.80 systolic
in Blood Flow Autoregulation
aortic BP 8.63 (SE 3.8)mmHg
In accelerated or malignant arterial hypertension, when
Diastolic ophthalmic artery pressure=0.80 diastolic
the BP suddenly goes above the autoregulation range, it
aortic BP+6.95 (SE 3.4)mmHg
produces vascular damage, e.g., hypertensive fundus
If there are vascular changes in the arteries feeding the changes [141, 142] and hypertensive encephalopathy.
ONH circulation, i.e. in the internal carotid artery, oph- However, if the BP rises less rapidly, there is ample evi-
thalmic artery and posterior ciliary arteries and/or smaller dence that the various circulatory systems in the body
vessels in the ONH itself (produced by vasospasm, arte- with autoregulatory mechanisms, e.g., cerebral circula-
riosclerosis, atherosclerosis, vasculitis, drug-induced tion [143151a] and coronary circulation [152], show an
vasoconstriction or dilation, or other systemic and car- adaptive phenomenon, wherein the range of autoregula-
diovascular diseases), then the BP in the ONH capillaries tion shifts to a higher level than in normal persons
would be much lower than that in the ophthalmic artery (Fig.5.2). This adaptation improves a persons tolerance
in normal healthy monkeys. Thus, it is important to real- of high BP but at the same time makes theindividual less
ize that clinical BP measurement in the brachial artery tolerant of low BP. There are a number of reports in
which patients with malignant hypertension, who had adjacent ONH and cause vasoconstriction by their
their BP lowered precipitously to prevent severe neuro- direct action on the ONH vessels (discussed in
logical and cardiovascular complications, suffered detail elsewhere [71]).
immediate and permanent blindness or severe visual loss 2. Increased Vascular Resistance in Terminal
from non-arteritic anterior ischemic optic neuropathy Arterioles
[153157]. This is fairly good proof that the autoregula- This is the basic pathology in arterial hypertension
tory mechanism in the ONH is deranged in hypertension and it reduces the blood flow in the capillary bed.
[71]. The same holds true for the coronary circulation; a 3. Secondary Hypertensive Vascular Changes
marked fall of BP in hypertensives during sleep can pre- These changes make the ONH vascular bed more vul-
cipitate myocardial ischemia, particularly in patients nerable to circulatory disorders. As discussed above,
with obstructive coronary artery disease [152]. Alderman arterial hypertension has also been shown to produce
etal. [158] found that a large reduction in BP (18mmHg) vascular endothelial abnormalities and consequently
induced by treatment of hypertension showed a relation- abnormalities in the formation of vascular endothelial
ship with development of myocardial ischemia. This is vasoactive agents.
further suggested by recent studies which have shown
that when arterial hypertensives are treated with antihy-
pertensive agents, lowering BP below a critical level Arterial Hypotension
increases cardiac mortality and morbidity. For example,
the rate of cardiac events at 75 mmHg diastolic BP is In an ONH with defective autoregulation, a fall of BP
twice than that at 85mmHg [159]. Similarly, in our 24-h below a critical level must decrease its blood flow
ambulatory BP monitoring studies we found that hyper- (Fig.5.1). A fall of BP in the ONH may be due to sys-
tensives on oral arterial hypotensive therapy (with beta- temic or local hypotension.
blockers, calcium-channel blockers, ACE inhibitors and
1. Systemic Arterial Hypotension
other hypotensive drugs) showed a significant associa-
Common causes of this are physiological arterial
tion between progressive visual deterioration and noc-
hypotension during sleep (i.e. nocturnal arterial
turnal hypotension, particularly in patients with
hypotension) [3, 160162] and intensive antihyperten-
non-arteritic anterior ischemic optic neuropathy [160,
sive medication in hypertensives (with beta-blockers,
161]. Thus, nocturnal arterial hypotension in hyperten-
calcium-channel blockers and/or ACE inhibitors [3,
sives could contribute to cardiac, cerebral and ONH
161, 163, 164], and other arterial hypotensive drugs);
ischemia because of defective autoregulation produced
less common causes include massive blood loss or
by hypertension in these organs.
shock [165]. Recent studies have shown an association
(c) Due to Direct Effect of Vasoconstrictor Agents
between nocturnal arterial hypotension and glaucoma-
on the ONH Blood flow
tous visual loss, an ONH ischemic disorder [3, 160
Malignant arterial hypertension can interfere directly
162, 166170]. Nocturnal arterial hypotension also
with the autoregulation of the ONH blood flow by the
plays a role in the development of non-arteritic ante-
following mechanism. As discussed in Chap. 4 and 14,
rior ischemic optic neuropathy; at least 73% of these
angiotensin and other circulating vasoconstrictor
patients discover visual loss on waking from sleep or
agents in malignant arterial hypertension leak out of
first use of critical vision, and the rest were not sure of
the choriocapillaris into the choroidal extravascular
the time of onset [73, 132, 171, 172].
tissue fluid. These vasoconstrictor agents can interfere
2. Local Arterial Hypotension
with the ONH blood flow by two mechanisms.
Instead of or in addition to the systemic arterial
(i) These agents, by their direct action on the peripap- hypotension, there may be a regional fall in the mean
illary choroidal vessels, produce vasoconstriction BP in the ocular blood vessels. This may be due to nar-
in those vessels [130]. As previously emphasized rowing of the regional arteries, such as the internal
(see Chap. 3), the peripapillary choroid is an carotid, ophthalmic or one or more of the posterior
important source of blood supply to the ONH, so ciliary arteries or arterioles supplying the ONH, or a
that vasoconstriction of the peripapillary choroidal fall of perfusion pressure locally in the peripapillary
vessels would reduce the ONH blood flow. choroid which supplies the ONH. Under such circum-
(ii) The vasoconstrictor agents also seep directly from stances, brachial BP measurement would grossly over-
the peripapillary choroidal tissue fluid into the estimate the BP in the ONH vessels.
Arterial Hypotension 93
Role of Calcium-Channel Blockers only I have not seen any beneficial effect, I have
in Glaucomatous Optic Neuropathy actually seen marked visual field deterioration with
this therapy. Thus, from my experience, I do not con-
sider this therapy safe and effective for treatment of
It has been postulated that, in glaucomatous optic neu- glaucomatous optic neuropathy.
ropathy, vascular insufficiency in the ONH may be due (e) Do Calcium-Channel Blockers Have Any
to vasospasm [173180] and that calcium-channel block- Serious Systemic Side-Effects?
ers, because of their vasodilatory property, may help to Nifedipine, a calcium-channel blocker, has been
relieve the vasospasm and thereby help to improve the advocated for glaucomatous optic neuropathy. Studies
ONH blood flow [181, 182]. Their use has been advo- have raised concerns about the safety of calcium-
cated in glaucomatous optic neuropathy, particularly in channel blockers, particularly short-acting dihydro-
normal tension glaucoma, and beneficial effects claimed pyridine derivatives, e.g., nifedipine and isradipine
[183190]. In this connection there are issues and ques- [195, 196]. These concerns include the following:
tions which are relevant to the present discussion and
require careful consideration, including the following: (i) In a large case-control study of hypertensive
patients treated with nifedipine, verapamil or
(a) Is There Evidence of Vasospasm in the ONH in
diltiazem there was an increased incidence of
Normal Tension Glaucoma?
myocardial infarction as compared to those
There is little direct evidence of that. The evidence so
treated with either beta-blockers or diuretics
far is essentially based on findings of the vasospastic
[197]. Similarly, Borhani et al. [198] in a large
response of the finger nail vascular bed to cold or the
randomized controlled clinical trial found a
presence of other systemic vasospastic disorders (e.g.,
higher incidence of major vascular events (e.g.,
migraine) [173180].
myocardial infarction, stroke, congestive heart
(b) Is There Definite Evidence that Calcium-
failure, angina, and sudden death) with the cal-
Channel Blockers Increase ONH Blood Flow?
cium-channel blocker isradipine (5.65%) as com-
Since we currently have no reliable method to measure
pared to hydrochlorothiazide (3.17%) and a
the blood flow in the ONH accurately, in spite of claims
significant increase in non-major vascular events
to the contrary [191] (see Chap. 6), we have no credi-
and procedures (e.g., transient ischemic attacks,
ble evidence that calcium-channel blockers actually
dysrhythmia, aortic valve replacement and femo-
increase blood flow in the ONH.
ral popliteal bypass graft) 9.05% vs. 5.22%.
(c) Does Associated Arterial Hypotension Have
(ii)Calcium-channel blockers have also been shown
Any Adverse Effects on ONH Blood Flow?
to be associated with increased incidence of peri-
It is a well-established fact that calcium-channel block-
operative [199] and gastrointestinal bleeding
ers lower systemic arterial blood pressure and can also
[200202] because of their antiplatelet aggrega-
produce marked nocturnal arterial hypotension [3, 160,
tion effect.
161, 164] which adversely affects the blood flow in the
(iii)In elderly patients treated with calcium-channel
ONH. Nocturnal arterial hypotension is a risk factor in
blockers (particularly nifedipine), the rate of
the development of glaucomatous optic neuropathy
development of cancer has been reported as sig-
and non-arteritic anterior ischemic optic neuropathy,
nificantly greater than with beta-blockers or ACE-
as discussed above.
inhibitors [201203].
(d) Is There Definite Evidence that Calcium-
(iv)Proptosis and periorbital edema have also been
Channel Blockers Help Glaucomatous Optic
reported as side-effects of calcium antagonists
Neuropathy?
[204], and these could adversely affect the ONH
There are reports in the literature claiming that [183
circulation by direct compression.
190]. At the same time, there are studies which have
found no significant beneficial effect of this therapy in In conclusion, in view of all these considerations, con-
glaucomatous optic neuropathy [192194]. I have trary to the claims made by some authors [183190],
tried calcium-channel blocker therapy in normal there seems no definite evidence to support the conten-
tension glaucoma and frequently have had patients tion that calcium-channel blockers are a safe and effec-
referred to me who had been prescribed calcium- tive therapy for the management of glaucomatous optic
channel blockers by other ophthalmologists; but not neuropathy.
Intraocular Pressure more than 2,000 persons over the years, recording IOP
every 3h between 7 a.m. and 10 p.m. with a Goldmann
tonometer, and have found that it is generally highest
As discussed above, perfusion pressure plays an impor-
early in the morning and lowest late in the evening,
tant role in determining the ONH blood flow; the per-
although infrequently other variations are seen. I have
fusion pressure is equal to mean BP minus IOP. Thus,
seen patients with IOP almost 100% higher early in the
there is an inverse relationship between IOP and perfu-
morning compared to that later on, in the afternoon or
sion pressure in the ONH the higher the IOP, the
evening. Such a marked circadian change of IOP may
lower the perfusion pressure and consequently the
be completely missed on routine clinic visits.
lower the blood flow in the ONH, if the autoregula-
3. Sleep
tion is defective. Normally the magnitude of fluctua-
A number of studies have investigated the effect of sleep
tions in mean BP is many times greater than of those in
on the IOP and have found IOP higher during sleep than
the IOP; hence, for any significant change to occur in
during waking hours [213216]. Brown et al. [214]
the perfusion pressure and the ONH blood flow, mean
found that IOP was significantly higher during sleep
BP is comparatively a much more important factor
than during wakefulness; it was 3.45 mmHg higher
than the IOP. A much larger rise in IOP would be
after 30min sleep, increasing thereafter to 6.41mmHg
required to interfere significantly with the ONH blood
above baseline after sleep up to 4h. Buguet and Romanet
flow in healthy persons (with normal BP and autoregu-
[216] also found that IOP was significantly lower dur-
lation) than in persons with arterial hypotension, defec-
ing the day than at night the rise was sharp at onset of
tive autoregulation or other vascular risk factors; in the
sleep in young persons but in the elderly population the
latter, even normal IOP may interfere with the ONH
increase in IOP was more gradual throughout the night.
blood flow. That may be of importance in the patho-
Day and night variation in IOP was significantly related
genesis of glaucomatous optic neuropathy, particularly
to the state of sleep/wakefulness. IOP was lower during
normal tension glaucoma [160, 161].
wakefulness than during light sleep, slow-wave sleep
A number of factors produce a short-term rise in
and rapid eye movement (REM) sleep. There was a sig-
IOP. Those include the following:
nificant difference between slow-sleep IOP and REM
1. Body Position sleep IOP. IOP was highest when subjects were awak-
IOP is elevated by a supine position [205211], usu- ened from slow-wave sleep which is predominant at the
ally 34mmHg [206, 207, 209]. Anderson and Grant beginning of the night. IOP during REM sleep was
[208] found considerable inter-individual variation in lower than slow-wave sleep and not related to the eye
the rise of IOP in a recumbent position compared to movements but most probably to miosis occurring at
sitting; there was no change in 30%, 1mmHg rise in that time. Lowest IOP occurred upon awakening from
22.6%, 2 mmHg rise or fall in 87% and more than REM sleep. Wildsoet etal. [215] found that the amount
2 mmHg in 13%. The maximum rise was 11mmHg of rise in IOP during sleep was reduced by sleeping in
and maximum fall 8mmHg; IOP changes were greater bright light as compared to sleeping in the dark.
in glaucoma patients on therapy than in those not on Brubaker [217] found that the rate of aqueous flow
therapy. Tsukahara and Sasaki [210] found a mean through the anterior chamber is highest during morning
difference in IOP between sitting and supine positions hours, slightly lower during the afternoon, and the rate
in normal persons of 5.61.71(SD)mmHg, in glau- during sleep is approximately one half of that during the
comatous optic neuropathy 6.52.66 mmHg, and in morning. The exact mechanism of the circadian varia-
normal tension glaucoma 8.63.5mmHg. tion in the IOP still remains unclear, but it has been
2. Circadian Variation in IOP found to be independent of the rise due to a supine
This is a well established phenomenon; the literature position; it has been speculated that it is due to plasma
on the subject has been well reviewed by Zeimer [212]. cortisol level, melatonin level or other autonomic or
Available evidence in the literature indicates that humoral mechanisms.
although there is a marked variability in the time of the The recurrent spikes of raised IOP above normal
day when the IOP peaks, in normal persons it most levels during sleep (due to supine position and also sleep)
often peaks early in the morning and decreases as the may be completely missed during routine clinic visits.
day progresses. I have done the diurnal curve of IOP in During sleep, this rise of IOP and concurrent
Resistance to Blood Flow 95
development of nocturnal arterial hypotension is a case) regulate the blood flow to the capillary bed by
dangerous combination, resulting in marked fall of altering their caliber. Anderson and his colleagues
perfusion pressure in the ONH. This may constitute an have discussed the role of pericytes in regulation of
important hidden risk factor for ONH ischemia in vul- the blood flow in the retina [4, 5]. There are many
nerable ONHs. This is further suggested by the develop- factors that influence the caliber of the feeding vessels,
ment of non-arteritic anterior ischemic optic neuropathy including the following:
during sleep in at least 75% of cases [73]. This mecha-
(a) Blood Flow Autoregulation: Its role in influenc-
nism may also play an important role in the pathogenesis
ing the caliber of the blood vessels in the ONH is
of glaucomatous optic neuropathy and its progression.
discussed above in detail.
From time to time, it has been claimed or implied
(b) Endothelial Derived Vasoactive Agents: From the
that some of the topically applied ocular hypotensive
vasomotor and blood flow point of view these
drugs (e.g., selective or non-selective beta-blockers,
agents play a very important role. This subject has
alpha agonists, carbonic anhydrase inhibitors, etc.)
been discussed in detail above.
influence the ONH blood flow by direct ocular penetra-
(c) Vascular Changes in the Arteries Feeding the
tion and effect on the ONH blood vessels. However, it is
ONH Circulation: Although the size of the lumen
important to bear in mind that drugs instilled in the con-
of the pre-capillary arterioles the primary resis-
junctival sac cannot reach the ONH by direct penetra-
tance vessels is thought to be the main factor nor-
tion (despite claims to that effect in the literature), and
mally determining the vascular resistance to blood
therefore there is no scientific basis for the assumption
flow, arterial changes altering the vascular resis-
that they have a direct action on the ONH or its circula-
tance in the internal carotid artery, ophthalmic
tion (see below). Some of the drug may be absorbed
artery and posterior ciliary arteries and/or smaller
into the systemic circulation and indirectly influence the
vessels in the ONH itself may be produced by
ONH circulation; for example, our study showed that
vasospasm, arteriosclerosis, atherosclerosis, vascu-
persons using topical beta-blocker eye drops developed
litis, drug-induced vasoconstriction or dilation, and
a significantly greater percentage drop in systemic dia-
a host of other systemic and cardiovascular dis-
stolic BP at night (p=0.028), lower minimum nighttime
eases. As discussed above, recent evidence sug-
diastolic BP (p=0.072), and lower minimum nighttime
gests that pericytes in the capillaries may also play
heart rate (p=0.002) than those not using them [218].
a role in regulation of blood flow, so pericyte abnor-
malities may likewise influence the blood flow.
2. Rheological Properties of the Blood
Resistance to Blood Flow These are influenced by a large variety of hematologic
disorders, particularly those causing increased blood
This is perhaps the most important of the three factors viscosity. Many studies have shown an association
that influence the blood flow in the ONH, according to between glaucomatous optic neuropathy and abnor-
the available information. According to Poiseuilles malities in blood rheology [219224].
Law, the resistance to flow is directly proportional to 3. Miscellaneous Experimental Vasomotor
blood viscosity, the length of vessel, and inversely pro- Studies
portional to the fourth power of the radius of the vessel. Flicker has been shown to increase the release of
The radius is a critical factor in determining resistance nitric oxide in the retina [225]. In monkeys, flicker
because of its geometrically disproportionate influ- increased glucose consumption in the inner retina and
ence. Therefore, the factors that can influence resis- optic nerve, most likely by increasing neuronal activ-
tance to blood flow in the ONH include the following: ity [33, 68, 226] Kondo et al. [226] found that, in
1. The State and Caliber of the Vessels Feeding anesthetized cats, flickering light increased blood
the ONH Circulation flow in the retina and ONH by 39% and 256% respec-
As discussed in Chap. 3, the ONH capillary bed is sup- tively, and the authors concluded that nitric oxide
plied by the posterior ciliary artery circulation. release may mediate much of the vasodilation effect
Available evidence indicates that in general terminal of flicker and play a role in maintaining normal
arterioles (of short posterior ciliary arteries in this vascular tone in the ONH.
Anderson and his group [227, 228] studied, by configuration of the venous outflow channel through the
invitro testing, the possible role of pericytes in vaso- lamina cribrosa is a primary determinant of the extent of
motor function. Ferrari-Dileo etal. [227] tested in cul- vascular and ultimately neuronal damage that occurs at
tured bovine retinal vascular pericytes the potential for a given IOP level in normal-tension and hypertensive
vasoactive neuropeptide receptors to affect capillary primary open-angle glaucoma. His entire concept,
resistance. They found that circulating or locally pro- however, is purely speculative. For example, the author
duced vasoactive neuropeptides might affect pericyte states that IOP rise of only a few mmHg in magnitude or
contractile tone, with the net effect on local blood flow IOP spikes of only a fraction of a minute in duration are
resulting from the effects on arteries and veins as well enough to cause permanent axonal damage in the ONH.
as capillaries. Zschauer etal. [228] reported that iso- As far as I am aware, no scientific evidence supports
proterenol produced relaxation of pericytes in a dose- that theory, because:
dependent manner, and low concentrations of the
(a) An IOP rise of a few mmHg cannot compromise
nonselective beta-blockers Propranolol and Timolol
the ONH circulation and cause irreversible axonal
blocked the relaxation produced by isoproterenol. By
damage unless the mean BP is extremely low,
contrast, Atenolol and Betaxolol, as relatively selec-
which is never the case in normal tension glau-
tive beta 1-adrenergic blockers, had no effect on the
coma and primary open angle glaucoma.
isoproterenol-induced relaxation.
(b) In our experimental studies on transient acute reti-
The ONH vasomotor properties of various eye drops
nal ischemia in rhesus monkeys, we have shown
used to lower IOP in glaucoma patients have been inves-
that acute ischemia of less than 100min produces
tigated. Orgl etal. [229] evaluated the invivo effects of
no demonstrable retinal damage; it takes about 4h
chronic topical 0.5% apraclonidine hydrochloride (an
of ischemia for the retina to suffer total or almost
alpha-adrenoreceptor agonist) in rabbits, and found no
total irreversible damage [233, 234]. Similarly gan-
observable ONH vasomotor effects. Bhandari et al.
glion cells and other neural elements in the brain
[230] studied the invivo effects of long-term applica-
can recover function in many patients with fairly
tion of brimonidine (an alpha-2-adrenoreceptor agonist)
marked cerebrovascular accidents. It is also known
on the ONH blood flow in rabbits, and found no effect
that axons can tolerate much longer and severer
on blood flow or vasomotor activity in the ONH. Tamaki
ischemia than ganglion cells in the entire nervous
etal. [231], after a 3-week application of topical 0.5%
system. Thus, permanent axonal damage in the
Betaxolol (a beta1-adrenoceptor antagonist) in the
ONH cannot occur with ischemia of only a fraction
human, found a clinically insignificant increase in blood
of a minute in duration, as the author speculated.
velocity in the ONH. This seems to indicate that none of
these eye drops have any direct effect on the blood flow Cioffi and Van Buskirk [235] put forward an intriguing
in the ONH, and consequently no direct beneficial effect concept: that the venous drainage in the prelaminar
on the ONH circulation. region and the peripapillary choroid are in two oppo-
site directions, and in view of that, the unique solitary
(venous) outflow system in the ONH may have a role
to play in the ischemic disorders of the ONH. There is
evidence for the first part of the concept, because the
Other Factors Influencing the Perfusion
venous drainage of the prelaminar region is centripetal,
Pressure in the ONH into the CRV, while that of the adjacent peripapillary
choroid is centrifugal, into the vortex veins. However,
1. Central Retinal Venous Pressure that is not always the case. There are eyes with a big
The perfusion pressure also depends upon the difference peripapillary choroidal vein draining the choroid pos-
between the mean arterial BP and the venous pressure in teriorly and exiting the eyeball in the peripapillary
the ONH vascular bed. The central retinal vein (CRV) is region (called the chorio-vaginal veins [236239])
the main venous channel draining the blood from the (Fig. 5.4). Also there is communication between the
ONH (see Chap. 3). Therefore, theoretically the pres- CRV and peripapillary choroidal veins in the prelami-
sure in the CRV could influence the perfusion pressure nar region (Fig.5.5), which is responsible for develop-
in the ONH. Bito [232] hypothesized that the size and ment of the commonly seen retinociliary collaterals
Other Factors Influencing the Perfusion Pressure in the ONH 97
(erroneously called opto-ciliary shunts) in eyes with developed ONH ischemic damage because of the
CRV occlusion (CRVO) (Fig.5.6). invariable rise in venous pressure in the prelami-
There are problems with the second part of the con- nar region in these two conditions; however, in
cept, i.e. that the two opposite directional forms of non-ischemic CRVO and hemi-CRVO, there is
venous drainage in the prelaminar region and peripap- never any evidence of ONH ischemia. In the isch-
illary choroid play a role in the ischemic disorders of emic varieties of CRVO and hemi-CRVO, only a
the ONH; these include the following: few of the eyes develop optic disc pallor later in
(a) I have prospectively studied in detail on a long-term the course of the disease, and those are the eyes
basis about 1,000 patients with CRVO and about with extensive retinal ischemic damage with
200 patients with hemi-CRVO (pathogenetically a almost total obliteration of retinal capillaries. In
variant of CRVO [241]). If the proposed hypothe-
sis were valid, then most of them should have
Fig.5.6 Fundus photograph of an eye with non-ischemic cen-

Fig.5.4 Fundus photograph showing the presence of a chorio- tral retinal vein occlusion showing retinociliary collaterals
vaginal vein (arrows) (arrows)
C
R S Col. Br.
PCA
D
A
Pia
Fig.5.5 Schematic
A arachnoid, C choroid, CRA
central retinal artery, Col. Br. ON
Collateral branches, CRV
PR
central retinal vein, D dura,
LC lamina cribrosa, ON optic LC
SAS
R retina, S sclera, SAS
CRV
subarachnoid space (Modified CRA
from Hayreh [240]) PCA
those eyes the optic disc pallor is purely secondary end-arteries is called a watershed zone. Watershed
to the retinal nerve fiber damage from retinal isch- zones between the various cerebral arteries are well-
emia, and NOT due to primary ONH ischemia. known, and they are also found in some other organs
(b) Similarly, I have studied prospectively in detail over having end-arterial systems. The significance of the
1,000 patients with non-arteritic anterior ischemic watershed zones is that in the event of a fall in the
optic neuropathy, and none of them showed any perfusion pressure in the vascular bed of one or more
evidence of compromise of the CRV circulation. of the end-arteries, the watershed zone, being an area
(c) I have also studied a large number of patients with of comparatively poor vascularity, is most vulnerable
optic disc edema due to a variety of causes, some to ischemia; this phenomenon is well illustrated by
lasting for months, but without any ONH ischemic the development of watershed zone infarcts in the
damage, in spite of some retinal venous engorge- cerebral cortex.
ment secondary to compression of the CRV in the Since the posterior ciliary arteries and their subdi-
ONH region by marked edema. visions, right down to the terminal choroidal arteri-
(d) I have performed fluorescein fundus angiography oles, are end-arteries in vivo [132, 171, 172, 245,
in all my non-arteritic anterior ischemic optic neu- 247255], they have watershed zones between them
ropathy patients, as well as in a large group of in both the choroid and the ONH (Figs.5.75.11).
patients with glaucomatous optic neuropathy, and When there are two posterior ciliary arteries (medial
they have never shown any evidence of compro- and lateral Figs.5.7 and 5.8), the area of the choroid
mise of the CRV circulation. and ONH supplied by the two shows a marked interin-
(e) The authors mentioned a high association between dividual variation which results in wide variation in
ocular hypertension/glaucoma and CRVO. That is the location of the watershed zone between the two.
correct. In our studies on CRVO, the prevalence of This is particularly important for the ONH because
glaucoma/ocular hypertension was found to be the watershed zone may pass temporal (Figs.5.7a and
significantly (P<0.0001) higher than in the gen- 5.8a) or nasal (Figs. 5.7d and 5.8e) to the ONH, or
eral population [242, 243]. But that is an entirely pass through only one or the other part of the ONH, or
different phenomenon, as discussed elsewhere cover the entire ONH (Figs. 5.7c and 5.8c). When
[244], and has nothing to do with the direction of there are three or more posterior ciliary arteries sup-
the venous drainage or unique solitary (venous) plying an eye, the locations of the watershed zones
outflow system in the ONH. vary according to the number of the posterior ciliary
arteries and their locations, as shown diagrammati-
Thus, there is little scientifically valid evidence so far
cally in Fig.5.8; this may result in a watershed-zone-
that CRV pressure normally plays any significant role
filling-defect in the upper or lower half only, depending
in the ONH blood flow and in ischemia of the ONH.
upon the artery or arteries involved. The part of the
This may be because the magnitude of the change in
ONH situated in the watershed zones is most vulner-
pressure in the venous circulation is not significant
able to ischemic disorders.
enough, compared to that in the arterial circulation, to
make any appreciable difference to the perfusion pres-
sure in the ONH.
2. Location of Posterior Ciliary Artery
Significance of Watershed Zones
Watershed Zone in Relation to the ONH
The presence, location and importance of watershed in Ischemic Disorders of the ONH
zones in the posterior ciliary artery circulation in
relation to the ONH circulation is discussed at length A filling delay or non-filling represents poor perfusion
in Chap. 3 and elsewhere [245, 246]. This an pressure in the posterior ciliary arteries supplying that
extremely important subject in any discussion of the part of the choroid and the ONH. My studies in eyes
factors influencing the blood flow in the ONH and with anterior ischemic optic neuropathy and glaucoma-
ischemic disorders of the ONH. When a tissue is tous optic neuropathy have indicated that the location
supplied by two or more end-arteries, the border of the watershed zone determines the vulnerability or
between the territories of distribution of any two involvement of the corresponding part of the ONH to
Other Factors Influencing the Perfusion Pressure in the ONH 99
Fig. 5.7 Fluorescein fundus angiograms of four eyes with choroid. (c) Left eye with the optic disc lying in the center of the
AION showing some examples of locations of the watershed watershed zone. (d) Left eye with the watershed zone passing
zone (vertical dark bands) in relation to the optic disc. (a) Right through the nasal part of the disc and adjacent nasal peripapil-
eye with the watershed zone lying temporal to the optic disc. lary choroid. See Fig.5.8 for a diagrammatic representation of
(b) Right eye with the watershed zone passing through the the patterns (Reproduced from Hayreh [245])
temporal part of the disc and adjacent temporal peripapillary
ischemia in the event of a fall of perfusion pressure in glaucoma, showing delayed filing of the lower half of
the posterior ciliary arteries or their branches [245, 246]. the choroidal watershed zone, with a corresponding cup-
Chapter 3 gives many fluorescein angiograms to dem- ping of the lower half of the optic disc. Fig.5.11 shows
onstrate this. For example, Fig.5.10 represents fluores- a filling defect in the upper half of the watershed zone
cein fundus angiograms of an eye with normal tension and upper half of the optic disc in an eye with arteritic
Fig.5.8 Diagrammatic representations of some examples of the locations of the watershed zones between the territories of the usual
two PCAs (one medial and the other lateral) (Reproduced from Hayreh [256])
Fig.5.9 Diagrammatic representations of some examples of the locations of the borders between the territories of three PCAs (ad)
in different combinations and of four PCAs (e) (Reproduced from Hayreh [261])
Conclusions 101
Fig.5.9 (continued)
Fig.5.10 Fluorescein fundus angiogram of right eye with nor-

mal tension glaucoma and cupping of the lower half of the optic
disc with corresponding visual field defect. It shows non-filling
of lower half of the watershed zone (arrow) and lower half of the
optic disc. See Fig.5.9 for a diagrammatic representation of this
type of pattern (Reproduced from Hayreh [245])
arteritic anterior ischemic optic neuropathy shows a non-filling
of the upper half of the watershed zone (arrow) as well as upper
anterior ischemic optic neuropathy involving the upper half of the optic disc (Reproduced from Hayreh [250])
half of the ONH. Sato etal. [257], based on Indocyanine
green angiography, found that the mean deviation of
visual fields was greater in the eye with a watershed
zone which included a larger part of the ONH than in Conclusions
the contralateral eye. The opposite was also true; the eye
with the greater mean deviation had a watershed zone This brief review gives some idea of the great com-
that included a larger part of the ONH. They concluded plexity of the blood flow in the ONH and the whole
that the location of the watershed zone appeared to influ- host of factors influencing it. In ischemic disorders
ence the progress of visual field defect. of the ONH, a large number of systemic and local
factors acting in different combinations and to differ- 7. Orgl S, Meyer P, Cioffi GA. Physiology of blood flow regu-
ent extents may derange the circulation in the ONH. lation and mechanisms involved in optic nerve perfusion. J
Glaucoma. 1995;4:42743.
These risk factors may be classified into two catego- 8. Sullivan SM, Johnson PC. Effect of oxygen on blood flow
ries: (1) predisposing risk factors which make the autoregulation in cat sartorius muscle. Am J Physiol.
ONH susceptible to ischemia, and (2) precipitating 1981;241(6):H80715.
risk factors which act as the final insult [160, 161, 9. Johnson PC. Autoregulation of blood flow. Circ Res.
1986;59(5):48395.
258]. Available evidence strongly suggests that ONH 10. Cowley AW, Hinojosa-Laborde C, Barber BJ, Harder DR,
ischemic disorders, including glaucomatous optic Lombard JH, Greene AS. Short-term autoregulation of sys-
neuropathy and non-arteritic anterior ischemic temic blood flow and cardiac output. News Physiol Sci.
optic neuropathy, are multifactorial in nature [160, 1989;4:21925.
11. Siegel G, Kmpe C, Ebeling BJ. pH-dependent myogenic
163, 258260]. In such a multifactorial scenario, one control in cerebral vascular smooth muscle. In: Cervos-
set of risk factors may be responsible for ONH isch- Navarro J, Fritschka E, editors. Cerebral microcirculation
emia in one case and a totally different set in another, and metabolism. New York: Raven; 1981. p. 21326.
and a particular risk factor may be present or criti- 12. Harder DR, Madden JA. Cellular mechanism of force devel-
opment in cat middle cerebral artery by reduced pCO2.
cal in one case and not inanother. Thus, each patient Pflugers Arch. 1985;403(4):4026.
with glaucomatous optic neuropathy, non-arteritic 13. Wray S. Smooth muscle intracellular pH: measurement,
anterior ischemic optic neuropathy or other ONH regulation, and function. Am J Physiol. 1988;254(2 Pt 1):
ischemic disorder may have a unique combination C21325.
14. Aalkjaer C. Regulation of intracellular pH and its role in
of systemic and local factors which together produce vascular smooth muscle function. J Hypertens.
ONH ischemic damage; no stereotyping is possible. 1990;8(3):197206.
Unfortunately, the widespread lack of understand- 15. Wei EP, Ellis EF, Kontos HA. Role of prostaglandins in pial
ing of the great complexity of ONH blood flow in arteriolar response to CO2 and hypoxia. Am J Physiol.
1980;238(2):H22630.
health and disease, along with our inability to eval- 16. Anderson DR, Davis EB. Glaucoma, capillaries and
uate the ONH blood flow reliably invivo (see Chap. pericytes. 5. Preliminary evidence that carbon dioxide
6), has caused much controversy on the subject of relaxes pericyte contractile tone. Ophthalmologica. 1996;
the role of vascular insufficiency in the ONH in 210(5):2804.
17. Harder DR. Pressure-dependent membrane depolarization in
glaucomatous optic neuropathy and with non-arter- cat middle cerebral artery. Circ Res. 1984;55(2):
itic anterior ischemic optic neuropathy. 197202.
18. Osol G, Halpern W. Myogenic properties of cerebral blood
vessels from normotensive and hypertensive rats. Am J
Acknowledgement A part of this review contains material
Physiol. 1985;249(5 Pt 2):H91421.
from my previous papers on the subject [1, 2].
19. Harder DR, Gilbert R, Lombard JH. Vascular muscle cell
depolarization and activation in renal arteries on elevation of
transmural pressure. Am J Physiol. 1987;253(4 Pt 2):
F77881.
References 20. Harder DR. Pressure-induced myogenic activation of cat
cerebral arteries is dependent on intact endothelium. Circ
Res. 1987;60(1):1027.
1. Hayreh SS. Factors influencing blood flow in the optic nerve 21. Harder DR, Kauser K, Roman RJ, Lombard JH. Mechanisms
head. J Glaucoma. 1997;6(6):41225. 1998;7:71. of pressure-induced myogenic activation of cerebral and
2. Hayreh SS. Blood flow in the optic nerve head and factors renal arteries: role of the endothelium. J Hypertens Suppl.
that may influence it. Prog Retin Eye Res. 2001;20: 1989;7(4 Suppl):S115.
595624. 22. Davies PF, Tripathi SC. Mechanical stress mechanisms and
3. Hayreh SS. Duke-Elder lecture. Systemic arterial blood the cell. An endothelial paradigm. Circ Res. 1993;72(2):
pressure and the eye. Eye. 1996;10(Pt 1):528. 23945.
4. Anderson DR. Glaucoma, capillaries and pericytes. 1. Blood 23. Bill A. Autonomic nervous control of uveal blood flow. Acta
flow regulation. Ophthalmologica. 1996;210(5):25762. Physiol Scand. 1962;56:7081.
5. Anderson DR, Davis EB. Glaucoma, capillaries and peri- 24. Bill A. Aspects of physiological and pharmacological regu-
cytes. 2. Identification and characterization of retinal peri- lation of uveal blood flow. Acta Soc Med Ups.
cytes in culture. Ophthalmologica. 1996;210(5):2638. 1962;67:12234.
6. Haefliger IO, Zschauer A, Anderson DR. Relaxation of reti- 25. Bill A. Blood circulation and fluid dynamics in the eye.
nal pericyte contractile tone through the nitric oxide-cyclic Physiol Rev. 1975;55(3):383417.
guanosine monophosphate pathway. Invest Ophthalmol Vis 26. Bill A. Some aspects of the ocular circulation. Friedenwald
Sci. 1994;35(3):9917. lecture. Invest Ophthalmol Vis Sci. 1985;26(4):41024.
References 103
27. Ehinger B. Adrenergic nerves to the eye and to related struc- 47. Sossi N, Anderson DR. Effect of elevated intraocular
tures in man and in the cynomolgus monkey (Macaca irus). pressure on blood flow. Occurrence in cat optic nerve head
Invest Ophthalmol. 1966;5:4252. studied with iodoantipyrine I 125. Arch Ophthalmol. 1983;
28. Ruskell GL. Facial parasympathetic innervation of the chor- 101(1):98101.
oidal blood vessels in monkeys. Exp Eye Res. 1971; 48. Weinstein JM, Duckrow RB, Beard D, Brennan RW.
12(2):16672. Regional optic nerve blood flow and its autoregulation.
29. Alm A, Bill A. The effect of stimulation of the cervical sym- Invest Ophthalmol Vis Sci. 1983;24(12):155965.
pathetic chain on retinal oxygen tension and on uveal, retinal 49. Robert Y, Maurer W. Pallor of the optic disc in glaucoma
and cerebral blood flow in cats. Acta Physiol Scand. patients with artificial hypertension. Doc Ophthalmol.
1973;88(1):8494. 1984;57(3):20314.
30. Alm A. The effect of sympathetic stimulation on blood flow 50. Robert Y, Grauwiller T, Hendrickson P, Brunner HR. Die
through the uvea, retina and optic nerve in monkeys (Macaca Autoregulation der Papillengefasse und ihr Verhalten unter
irus). Exp Eye Res. 1977;25(1):1924. Halothan. [Autoregulation of the papillary vessels and their
31. Matsusaka T. An evidence for adrenergic involvement in the behavior as affected by halothane]. Klin Monatsbl
choroidal circulation. Albrecht Von Graefes Arch Klin Exp Augenheilkd. 1988;192(2):11721.
Ophthalmol. 1981;216(1):1721. 51. Robert Y, Steiner D, Hendrickson P. Papillary circulation
32. Morgan TR, Green K, Bowman K. Effects of adrenergic dynamics in glaucoma. Graefes Arch Clin Exp Ophthalmol.
agonists upon regional ocular blood flow in normal and gan- 1989;227(5):4369.
glionectomized rabbits. Exp Eye Res. 1981;32(6):6917. 52. Sperber GO, Bill A. Blood flow and glucose consumption in
33. Bill A, Sperber GO. Aspects of oxygen and glucose con- the optic nerve, retina and brain: effects of high intraocular
sumption in the retina: effects of high intraocular pressure pressure. Exp Eye Res. 1985;41(5):63953.
and light. Graefes Arch Clin Exp Ophthalmol. 53. Pillunat LE, Stodtmeister R, Wilmanns I, Christ T.
1990;228(2):1247. Autoregulation of ocular blood flow during changes in
34. Bill A, Sperber GO. Control of retinal and choroidal blood intraocular pressure. Preliminary results. Graefes Arch Clin
flow. Eye. 1990;4(Pt 2):31925. Exp Ophthalmol. 1985;223(4):21923.
35. Bill A, Sperber G, Ujiie K. Physiology of the choroidal vas- 54. Pillunat LE, Stodtmeister R, Wilmanns I, Christ T.
cular bed. Int Ophthalmol. 1983;6(2):1017. Drucktoleranztest des Sehnervenkopfes bei okularer
36. Grajewski AL, Ferrari-Dileo G, Feuer WJ, Anderson DR. Hypertension. [A test of pressure tolerance of the optic nerve
Beta-adrenergic responsiveness of choroidal vasculature. head in ocular hypertension]. Klin Monatsbl Augenheilkd.
Ophthalmology. 1991;98(6):98995. 1986;188(1):3944.
37. Riva CE, Cranstoun SD, Mann RM, Barnes GE. Local chor- 55. Pillunat LE, Stodtmeister R, Wilmanns I. Pressure com
oidal blood flow in the cat by laser Doppler flowmetry. Invest pliance of the optic nerve head in low tension glaucoma. Br
Ophthalmol Vis Sci. 1994;35(2):60818. J Ophthalmol. 1987;71(3):1817.
38. Abe S, Karita K, Izumi H, Tamai M. Increased and decreased 56. Bill A, Sperber GO. Blood flow and glucose consumption in
choroidal blood flow elicited by cervical sympathetic nerve the optic nerve: effects of high intraocular pressure. In:
stimulation in the cat. Jpn J Physiol. 1995;45(2):34753. Krieglstein GK, editor. Glaucoma update III. Heidelberg:
39. Zagvazdin YS, Fitzgerald ME, Sancesario G, Reiner A. Springer; 1987. p. 517.
Neural nitric oxide mediates Edinger-Westphal nucleus 57. Novack RL, Stefansson E, Hatchell DL. Intraocular pressure
evoked increase in choroidal blood flow in the pigeon. Invest effects on optic nerve-head oxidative metabolism measured
Ophthalmol Vis Sci. 1996;37(4):66672. in vivo. Graefes Arch Clin Exp Ophthalmol.
40. Alm A, Bill A. Ocular and optic nerve blood flow at normal 1990;228(2):12833.
and increased intraocular pressures in monkeys (Macaca 58. Alm A, Bill A. Blood flow and oxygen extraction in the cat
irus): a study with radioactively labelled microspheres uvea at normal and high intraocular pressures. Acta Physiol
including flow determinations in brain and some other tis- Scand. 1970;80(1):1928.
sues. Exp Eye Res. 1973;15(1):1529. 59. Alm A, Bill A. The oxygen supply to the retina. I. Effects of
41. Ernest JT, Potts AM. Pathophysiology of the distal portion changes in intraocular and arterial blood pressures, and in
of the optic nerve. IV. Local temperature as a measure of arterial P O2 and P CO2 on the oxygen tension in the vitreous
blood flow. Am J Ophthalmol. 1971;72(2):43344. body of the cat. Acta Physiol Scand. 1972;84(2):
42. Ernest JT. Autoregulation of optic-disk oxygen tension. 26174.
Invest Ophthalmol. 1974;13(2):1016. 60. Alm A, Bill A. The oxygen supply to the retina. II. Effects of
43. Ernest JT. Optic disc blood flow. Trans Ophthalmol Soc UK. high intraocular pressure and of increased arterial carbon
1976;96(3):34851. dioxide tension on uveal and retinal blood flow in cats.
44. Ernest JT. Optic disk oxygen tension. Exp Eye Res. A study with radioactively labelled microspheres including
1977;24(3):2718. flow determinations in brain and some other tissues. Acta
45. Ernest JT. Autoregulation of blood flow in the distal segment Physiol Scand. 1972;84(3):30619.
of the optic nerve. In: Krieglstein GK, Leydhecker W, editors. 61. Takats I, Leiszter F. Relationship between blood flow
Glaucoma update. Heidelberg: Springer; 1979. p. 93100. velocity in the choroid and intraocular pressure in rabbits.
46. Geijer C, Bill A. Effects of raised intraocular pressure on Acta Ophthalmol (Copenh). 1979;57(1):4854.
retinal, prelaminar, laminar, and retrolaminar optic nerve 62. Kiel JW, van Heuven WA. Ocular perfusion pressure and
blood flow in monkeys. Invest Ophthalmol Vis Sci. choroidal blood flow in the rabbit. Invest Ophthalmol Vis
1979;18(10):103042. Sci. 1995;36(3):57985.
63. Reiner A, Zagvazdin Y, Fitzgerald ME. Choroidal blood 83. Lscher TF. Imbalance of endothelium-derived relaxing
flow in pigeons compensates for decreases in arterial blood and contracting factors. A new concept in hypertension?
pressure. Exp Eye Res. 2003;76(3):27382. Am J Hypertens. 1990;3(4):31730.
64. Reiner A, Li C, Del Mar N, Fitzgerald ME. Choroidal blood 84. Lscher TF, Raij L, Vanhoutte PM. Endothelium-dependent
flow compensation in rats for arterial blood pressure vascular responses in normotensive and hypertensive Dahl
decreases is neuronal nitric oxide-dependent but compensa- rats. Hypertension. 1987;9(2):15763.
tion for arterial blood pressure increases is not. Exp Eye Res. 85. Forte P, Copland M, Smith LM, Milne E, Sutherland J,
2010;90(6):73441. Benjamin N. Basal nitric oxide synthesis in essential hyper-
65. Porsaa K. Experimental studies on the vasomotor innervation tension. Lancet. 1997;349(9055):83742.
of the retinal arteries. Acta Ophthalmol (Copenh). 1941; 86. de Saenz TI, Goldstein I, Azadzoi K, Krane RJ, Cohen RA.
Suppl. 18. Impaired neurogenic and endothelium-mediated relaxation
66. Russell RW. Observations on intracerebral aneurysms. of penile smooth muscle from diabetic men with impo-
Brain. 1963;86:42542. tence. N Engl J Med. 1989;320(16):102530.
67. Dollery CT, Hill DW, Hodge JV. The response of normal 87. Tesfamariam B, Jakubowski JA, Cohen RA. Contraction of
retinal blood vessels to angiotensin and noradrenaline. diabetic rabbit aorta caused by endothelium-derived PGH2-
JPhysiol Lond. 1963;165:5007. TxA2. Am J Physiol. 1989;257(5 Pt 2):H132733.
68. Sperber GO, Bill A. The 2-deoxyglucose method and ocular 88. Ku DD. Coronary vascular reactivity after acute myocar-
blood flow. In: Lambrou GN, Greve E, editors. Ocular blood dial ischemia. Science. 1982;218(4572):5768.
flow in glaucoma. Amsterdam: Kugler & Ghedini; 1989. p. 89. VanBenthuysen KM, McMurtry IF, Horwitz LD. Reperfusion
7380. after acute coronary occlusion in dogs impairs endothelium-
69. Hayreh SS, Bill A, Sperber GO. Effects of high intraocular dependent relaxation to acetylcholine and augments contrac-
pressure on the glucose metabolism in the retina and optic tile reactivity invitro. J Clin Invest. 1987;79(1):26574.
nerve in old atherosclerotic monkeys. Graefes Arch Clin 90. Pearson PJ, Schaff HV, Vanhoutte PM. Long-term impair-
Exp Ophthalmol. 1994;232(12):74552. ment of endothelium-dependent relaxations to aggregating
70. Haefliger IO, Meyer P, Flammer J, Lscher TF. The vascular platelets after reperfusion injury in canine coronary arter-
endothelium as a regulator of the ocular circulation: a new ies. Circulation. 1990;81(6):19217.
concept in ophthalmology? Surv Ophthalmol. 1994;39(2): 91. Clavier N, Kirsch JR, Hurn PD, Traystman RJ. Effect of
12332. postischemic hypoperfusion on vasodilatory mechanisms
71. Hayreh SS, Servais GE, Virdi PS. Fundus lesions in malig- in cats. Am J Physiol. 1994;267(5 Pt 2):H20128.
nant hypertension. V. Hypertensive optic neuropathy. 92. Myers PR, Muller JM, Tanner MA. Effects of oxygen
Ophthalmology. 1986;93(1):7487. tension on endothelium dependent responses in canine cor-
72. Hayreh SS, Bill A, Sperber GO. Metabolic effects of high onary microvessels. Cardiovasc Res. 1991;25(11):88594.
intraocular pressure in old arteriosclerotic monkeys. Invest 93. Liu J, Chen R, Casley DJ, Nayler WG. Ischemia and reper-
Ophthalmol Vis Sci. 1991;32(4):810. fusion increase 125I-labeled endothelin-1 binding in rat
73. Hayreh SS, Podhajsky PA, Zimmerman B. Nonarteritic cardiac membranes. Am J Physiol. 1990;258(3 Pt
anterior ischemic optic neuropathy: time of onset of visual 2):H82935.
loss. Am J Ophthalmol. 1997;124(5):6417. 94. Miyauchi T, Yanagisawa M, Tomizawa T, Sugishita Y,
74. Rubanyi GM. Cardiovascular significance of endothelium- Suzuki N, Fujino M, etal. Increased plasma concentrations
derived vasoactive factors. 1st ed. Mount Kisco: Futura; of endothelin-1 and big endothelin-1 in acute myocardial
1991. infarction. Lancet. 1989;2(8653):534.
75. Ryan US, Rubanyi GM. Endothelial regulation of vascular 95. Watanabe T, Suzuki N, Shimamoto N, Fujino M, Imada A.
tone. 1st ed. New York: Marcel Dekker; 1992. Endothelin in myocardial infarction. Nature. 1990;344
76. Vanhoutte PM. The endothelium: modulator of vascular (6262):114.
smooth-muscle tone. N Engl J Med. 1988;319(8):5123. 96. Rubanyi GM, Vanhoutte PM. Hypoxia releases a vasocon-
77. Davies MG, Hagen PO. The vascular endothelium. A new stricting substance from the canine vascular endothelium.
horizon. Ann Surg. 1993;218(5):593609. J Physiol Lond. 1985;364:4556.
78. Suzuki M, Ohyama Y, Satoh S. Conversion of angiotensin I 97. Kourembanas S, Marsden PA, McQuillan LP, Faller DV.
to angiotensin II and inactivation of bradykinin in canine Hypoxia induces endothelin gene expression and secretion
peripheral vascular beds. J Cardiovasc Pharmacol. in cultured human endothelium. J Clin Invest. 1991;88(3):
1984;6(2):24450. 10547.
79. Dzau VJ. Vascular renin-angiotensin: a possible autocrine or 98. Todd AS. The histological localisation of fibrinolysin acti-
paracrine system in control of vascular function. J Cardiovasc vator. J Pathol Bacteriol. 1959;78:2813.
Pharmacol. 1984;6 Suppl 2:S37782. 99. Dosne AM, Dupuy E, Bodevin E. Production of a fibrin-
80. Dzau VJ. Significance of the vascular renin-angiotensin olytic inhibitor by cultured endothelial cells derived from
pathway. Hypertension. 1986;8(7):5539. human umbilical vein. Thromb Res. 1978;12(3):37787.
81. Dzau VJ. Implications of local angiotensin production in 100. Vanhoutte PM, Houston DS. Platelets, endothelium, and
cardiovascular physiology and pharmacology. Am J Cardiol. vasospasm. Circulation. 1985;72(4):72834.
1987;59(2):59A65A. 101. Heistad DD, Armstrong ML, Marcus ML, Piegors DJ,
82. Lindsey CJ, Bendhack LM, Paiva AC. Effects of teprotide, cap- Mark AL. Augmented responses to vasoconstrictor stimuli
topril and enalaprilat on arterial wall kininase and angiotensin in hypercholesterolemic and atherosclerotic monkeys. Circ
converting activity. J Hypertens Suppl. 1987;5(2):S716. Res. 1984;54(6):7118.
References 105
102. Jayakody RL, Senaratne MP, Thomson AB, Kappagoda 116. Dawson TM, Bredt DS, Fotuhi M, Hwang PM, Snyder SH.
CT. Cholesterol feeding impairs endothelium-dependent Nitric oxide synthase and neuronal NADPH diaphorase are
relaxation of rabbit aorta. Can J Physiol Pharmacol. identical in brain and peripheral tissues. Proc Natl Acad Sci
1985;63(9):12069. USA. 1991;88(17):7797801.
103. Freiman PC, Mitchell GG, Heistad DD, Armstrong ML, 117. Nyborg NC, Nielsen PJ. The level of spontaneous myo-
Harrison DG. Atherosclerosis impairs endothelium- genic tone in isolated human posterior ciliary arteries
dependent vascular relaxation to acetylcholine and throm- decreases with age. Exp Eye Res. 1990;51(6):7115.
bin in primates. Circ Res. 1986;58(6):7839. 118. Nyborg NC, Prieto D, Benedito S, Nielsen PJ. Endothelin-
104. Habib JB, Bossaller C, Wells S, Williams C, Morrisett 1-induced contraction of bovine retinal small arteries is
JD, Henry PD. Preservation of endothelium-dependent reversible and abolished by nitrendipine. Invest Ophthalmol
vascular relaxation in cholesterol-fed rabbit by treatment Vis Sci. 1991;32(1):2731.
with the calcium blocker PN 200110. Circ Res. 1986; 119. Yao K, Tschudi M, Flammer J, Lscher TF. Endothelium-
58(2):3059. dependent regulation of vascular tone of the porcine
105. Bossaller C, Habib GB, Yamamoto H, Williams C, Wells ophthalmic artery. Invest Ophthalmol Vis Sci. 1991;
S, Henry PD. Impaired muscarinic endothelium-depen- 32(6):17918.
dent relaxation and cyclic guanosine 5-monophosphate 120. Haefliger IO, Flammer J, Luscher TF. Nitric oxide and
formation in atherosclerotic human coronary artery and endothelin-1 are important regulators of human ophthalmic
rabbit aorta. J Clin Invest. 1987;79(1):1704. artery. Invest Ophthalmol Vis Sci. 1992;33(7):23403.
106. Faraci FM, Lopez AG, Breese K, Armstrong ML, Heistad 121. Haefliger IO, Flammer J, Luscher TF. Heterogeneity of
DD. Effect of atherosclerosis on cerebral vascular endothelium-dependent regulation in ophthalmic and ciliary
responses to activation of leukocytes and platelets in arteries. Invest Ophthalmol Vis Sci. 1993;34(5):172230.
monkeys. Stroke. 1991;22(6):7906. 122. Meyer P, Flammer J, Lscher TF. Local anesthetic drugs
107. Tagawa H, Tomoike H, Nakamura M. Putative mecha- reduce endothelium-dependent relaxations of porcine ciliary
nisms of the impairment of endothelium-dependent relax- arteries. Invest Ophthalmol Vis Sci. 1993;34(9):27306.
ation of the aorta with atheromatous plaque in heritable 123. Meyer P, Flammer J, Lscher TF. Endothelium-dependent
hyperlipidemic rabbits. Circ Res. 1991;68(2):3307. regulation of the ophthalmic microcirculation in the per-
107a. Minor Jr RL, Myers PR, Guerra Jr R, Bates JN, Harrison fused porcine eye: role of nitric oxide and endothelins.
DG. Diet-induced atherosclerosis increases the release of Invest Ophthalmol Vis Sci. 1993;34(13):361421.
nitrogen oxides from rabbit aorta. J Clin Invest. 124. Meyer P, Flammer J, Lscher TF. Local action of the renin
1990;86(6):210916. angiotensin system in the porcine ophthalmic circulation:
108. Yamamoto H, Bossaller C, Cartwright Jr J, Henry PD. effects of ACE-inhibitors and angiotensin receptor antago-
Videomicroscopic demonstration of defective cholinergic nists. Invest Ophthalmol Vis Sci. 1995;36(3):55562.
arteriolar vasodilation in atherosclerotic rabbit. J Clin 125. Mann RM, Riva CE, Stone RA, Barnes GE, Cranstoun SD.
Invest. 1988;81(6):17528. Nitric oxide and choroidal blood flow regulation. Invest
109. Moncada S, Higgs EA. Endogenous nitric oxide: physiol- Ophthalmol Vis Sci. 1995;36(5):92530.
ogy, pathology and clinical relevance. Eur J Clin Invest. 126. Zhu P, Bny JL, Flammer J, Lscher TF, Haefliger IO.
1991;21(4):36174. Relaxation by bradykinin in porcine ciliary artery. Role of
110. Golino P, Piscione F, Willerson JT, Cappelli-Bigazzi M, nitric oxide and K(+)-channels. Invest Ophthalmol Vis Sci.
Focaccio A, Villari B, etal. Divergent effects of serotonin 1997;38(9):17617.
on coronary-artery dimensions and blood flow in patients 127. Nyborg NC, Nielsen PJ. Angiotensin-II contracts isolated
with coronary atherosclerosis and control patients. N human posterior ciliary arteries. Invest Ophthalmol Vis Sci.
Engl J Med. 1991;324(10):6418. 1990;31(11):24713.
111. McFadden EP, Clarke JG, Davies GJ, Kaski JC, Haider 128. Ferrari-Dileo G, Davis EB, Anderson DR. Angiotensin II
AW, Maseri A. Effect of intracoronary serotonin on coro- binding receptors in retinal and optic nerve head blood ves-
nary vessels in patients with stable angina and patients sels. An autoradiographic approach. Invest Ophthalmol Vis
with variant angina. N Engl J Med. 1991;324(10): Sci. 1991;32(1):216.
64854. 129. Neufeld AH, Hernandez MR, Gonzalez M. Nitric oxide
112. Bergua A. Nitrergische Reaktivitat in den Endothelzellen synthase in the human glaucomatous optic nerve head.
der menschlichen uvealen Gefasse. [Nitrergic reactivity Arch Ophthalmol. 1997;115(4):497503.
in endothelial cells of human uveal blood vessels]. Klin 130. Hayreh SS, Servais GE, Virdi PS. Fundus lesions in malig-
Monatsbl Augenheilkd. 1995;206(2):11521. nant hypertension. VI. Hypertensive choroidopathy.
113. Chakravarthy U, Stitt AW, McNally J, Bailie JR, Hoey Ophthalmology. 1986;93(11):1383400.
EM, Duprex P. Nitric oxide synthase activity and expres- 131. Hayreh SS, Piegors DJ, Heistad DD. Serotonin-induced
sion in retinal capillary endothelial cells and pericytes. constriction of ocular arteries in atherosclerotic monkeys.
Curr Eye Res. 1995;14(4):28594. Implications for ischemic disorders of the retina and optic
114. Shiells R, Falk G. Retinal on-bipolar cells contain a nitric nerve head. Arch Ophthalmol. 1997;115(2):2208.
oxide-sensitive guanylate cyclase. NeuroReport. 1992;3 132. Hayreh SS. Anterior ischaemic optic neuropathy.
(10):8458. Differentiation of arteritic from non-arteritic type and its
115. Lau KC, So KF, Tay D, Leung MC. NADPH-diaphorase management. Eye. 1990;4(Pt 1):2541.
neurons in the retina of the hamster. J Comp Neurol. 133. Cioffi GA, Orgl S, Onda E, Bacon DR, Van Buskirk EM.
1994;350(4):5508. An in vivo model of chronic optic nerve ischemia: the
dose-dependent effects of endothelin-1 on the optic nerve 151a. MacKenzie ET, McGeorge AP, Graham DI, Fitch W,
microvasculature. Curr Eye Res. 1995;14(12):114753. Edvinsson L, Harper AM. Breakthrough of cerebral auto-
134. Cioffi GA, Orgl S. The effects of chronic optic nerve regulation and the sympathetic nervous system. Acta
ischemia in the rabbit. In: Drance SM, editor. Vascular risk Neurol Scand Suppl. 1977;64:489.
factors and neuroprotection in glaucoma - update. Amsterdam: 152. Floras JS. Antihypertensive treatment, myocardial infarc-
Kugler Pub; 1996. p. 11522. tion, and nocturnal myocardial ischaemia. Lancet. 1988;
135. Cioffi GA, Sullivan P. The effect of chronic ischemia on the 2(8618):9946.
primate optic nerve. Eur J Ophthalmol. 1999;9 Suppl 153. Cove DH, Seddon M, Fletcher RF, Dukes DC. Blindness
1:S346. after treatment for malignant hypertension. Br Med J.
136. Orgl S, Cioffi GA, Bacon DR, Van Buskirk EM. An 1979;2(6184):2456.
endothelin-1-induced model of chronic optic nerve isch- 154. Hulse JA, Taylor DS, Dillon MJ. Blindness and paraple-
emia in rhesus monkeys. J Glaucoma. 1996;5(2):1358. gia in severe childhood hypertension. Lancet. 1979;
137. Oku H, Sugiyama T, Kojima S, Watanabe T, Azuma I. 2(8142):5536.
Experimental optic cup enlargement caused by endothelin- 155. Pryor JS, Davies PD, Hamilton DV. Blindness and malig-
1-induced chronic optic nerve head ischemia. Surv nant hypertension. Lancet. 1979;2(8146):803.
Ophthalmol. 1999;44 Suppl 1:S7484. 156. Wetherill JH. Blindness after treatment for malignant
138. Ota M, Oku H. Nicardipine modification of endothelin-1 hypertension. Br Med J. 1979;2(6189):550.
effects on visual evoked potential. Jpn J Ophthalmol. 157. Taylor D, Ramsay J, Day S, Dillon M. Infarction of the
1997;41(1):3842. optic nerve head in children with accelerated hyperten-
139. Nishimura K, Riva CE, Harino S, Reinach P, Cranstoun SD, sion. Br J Ophthalmol. 1981;65(3):15360.
Mita S. Effects of endothelin-1 on optic nerve head blood 158. Alderman MH, Ooi WL, Madhavan S, Cohen H.
flow in cats. J Ocul Pharmacol Ther. 1996;12(1):7583. Treatment-induced blood pressure reduction and the risk
140. Hayreh SS, Edwards J. Ophthalmic arterial and venous of myocardial infarction. JAMA. 1989;262(7):9204.
pressures. Effects of acute intracranial hypertension. Br J 159. Farnett L, Mulrow CD, Linn WD, Lucey CR, Tuley MR.
Ophthalmol. 1971;55(10):64963. The J-curve phenomenon and the treatment of hyperten-
141. Hayreh SS. Malignant arterial hypertension. Ophthalmol sion. Is there a point beyond which pressure reduction is
Clin North Am. 1992;5:44573. dangerous? JAMA. 1991;265(4):48995.
142. Hayreh SS. Duke-Elder Lecture: Systemic arterial blood 160. Hayreh SS, Zimmerman MB, Podhajsky P, Alward WL.
pressure and the eye. Eye 1996; 10:528. Nocturnal arterial hypotension and its role in optic nerve
143. Strandgaard S, Olesen J, Skinhoj E, Lassen NA. head and ocular ischemic disorders. Am J Ophthalmol.
Autoregulation of brain circulation in severe arterial hyper- 1994;117(5):60324.
tension. Br Med J. 1973;1(5852):50710. 161. Hayreh SS, Podhajsky P, Zimmerman MB. Role of noc-
144. Strandgaard S, Jones JV, MacKenzie ET, Harper AM. turnal arterial hypotension in optic nerve head ischemic
Upper limit of cerebral blood flow autoregulation in experi- disorders. Ophthalmologica. 1999;213(2):7696.
mental renovascular hypertension in the baboon. Circ Res. 162. Hayreh SS. Vascular factors in the pathogenesis of glau-
1975;37(2):1647. comatous optic neuropathy. In: Drance SM, editor.
145. Fitch W, MacKenzie ET, Harper AM. Effects of decreasing International symposium on glaucoma, ocular blood flow,
arterial blood pressure on cerebral blood flow in the baboon. and drug treatment, 1990, Seville, Spain. Baltimore:
Influence of the sympathetic nervous system. Circ Res. Williams & Wilkins; 1992. p. 3341.
1975;37(5):5507. 163. Hayreh SS. Acute ischemic disorders of the optic nerve:
146. Strandgaard S. Autoregulation of cerebral blood flow in pathogenesis, clinical manifestations, and management.
hypertensive patients. The modifying influence of pro- Ophthalmol Clin North Am. 1996;9:40742.
longed antihypertensive treatment on the tolerance to acute, 164. Hayreh SS. The role of age and cardiovascular disease in
drug-induced hypotension. Circulation. 1976;53(4):7207. glaucomatous optic neuropathy. Surv Ophthalmol.
147. Strandgaard S, MacKenzie ET, Jones JV, Harper AM. 1999;43 Suppl 1:S2742.
Studies on the cerebral circulation of the baboon in acutely 165. Hayreh SS. Anterior ischemic optic neuropathy. VIII.
induced hypertension. Stroke. 1976;7(3):28790. Clinical features and pathogenesis of post-hemorrhagic
148. Jones JV, Fitch W, MacKenzie ET, Strandgaard S, Harper amaurosis. Ophthalmology. 1987;94(11):1488502.
AM. Lower limit of cerebral blood flow autoregulation in 166. Bchetoille A, Bresson-Dumont H. Diurnal and noctur-
experimental renovascular hypertension in the baboon. nal blood pressure drops in patients with focal ischemic
Circ Res. 1976;39(4):5557. glaucoma. Graefes Arch Clin Exp Ophthalmol.
149. Bill A, Linder J. Sympathetic control of cerebral blood flow 1994;232(11):6759.
in acute arterial hypertension. Acta Physiol Scand. 1976; 167. Graham SL, Drance SM, Wijsman K, Douglas GR,
96(1):11421. Mikelberg FS. Ambulatory blood pressure monitoring in
150. Edvinsson L, Owman C, Siesjo B. Physiological role of glaucoma. The nocturnal dip. Ophthalmology. 1995;
cerebrovascular sympathetic nerves in the autoregulation 102(1):619.
of cerebral blood flow. Brain Res. 1976;117(3):51923. 168. Bresson-Dumont H, Bechetoille A. Rle de la tension
151. Boisvert DP, Jones JV, Harper AM. Cerebral blood flow artrielle dans lvolutivit des lsions glaucomateuses.
autoregulation to acutely increasing blood pressure during [Role of arterial blood pressure in the development of
sympathetic stimulation. Acta Neurol Scand Suppl. 1977; glaucomatous lesions]. J Fr Ophtalmol. 1996;19(67):
64:467. 43542.
References 107
169. Meyer JH, Brandi-Dohrn J, Funk J. Twenty four hour blood camine in eyes with normal-tension glaucoma. Oph
pressure monitoring in normal tension glaucoma. Br J thalmology. 1996;103(2):2838.
Ophthalmol. 1996;80(10):8647. 189. Kanellopoulos AJ, Erickson KA, Netland PA. Systemic
170. Detry M, Boschi A, Ellinghaus G, De Plaen JF. Simulta calcium channel blockers and glaucoma. J Glaucoma.
neous 24-hour monitoring of intraocular pressure and 1996;5(5):35762.
arterial blood pressure in patients with progressive and 190. Piltz JR, Bose S, Lanchoney D. The effect of nimodipine, a
non-progressive primary open-angle glaucoma. Eur J centrally active calcium antagonist, on visual function and
Ophthalmol. 1996;6(3):2738. macular blood flow in patients with normal-tension
171. Hayreh SS. Anterior ischaemic optic neuropathy. I. glaucoma and control subjects. J Glaucoma. 1996;7:
Terminology and pathogenesis. Br J Ophthalmol. 1974; 33642.
58(12):95563. 191. Hayreh SS. Evaluation of optic nerve head circulation: review
172. Hayreh SS. Anterior ischemic optic neuropathy. Arch of the methods used. J Glaucoma. 1997;6(5):31930.
Neurol. 1981;38(11):6758. 192. Lumme P, Tuulonen A, Airaksinen PJ, Alanko HI.
173. Phelps CD, Corbett JJ. Migraine and low-tension glau- Neuroretinal rim area in low tension glaucoma: effect of
coma. A case-control study. Invest Ophthalmol Vis Sci. nifedipine and acetazolamide compared to no treatment.
1985;26(8):11058. Acta Ophthalmol Copenh. 1991;69(3):2938.
174. Gasser P, Flammer J, Guthauser U, Niesel P. Bedeutung des 193. Liu S, Araujo SV, Spaeth GL, Katz LJ, Smith M. Lack of
vasospastischen syndroms in der Augenheilkunde. Klin effect of calcium channel blockers on open-angle glau-
Monatsbl Augenheilkd. 1986;188:3989. coma. J Glaucoma. 1996;5(3):18790.
175. Gasser P, Flammer J. Influence of vasospasm on visual 194. Harris A, Evans DW, Cantor LB, Martin B. Hemodynamic
function. Doc Ophthalmol. 1987;66(1):318. and visual function effects of oral nifedipine in patients
176. Guthauser U, Flammer J, Mahler F. The relationship with normal-tension glaucoma. Am J Ophthalmol.
between digital and ocular vasospasm. Graefes Arch Clin 1997;124(3):296302.
Exp Ophthalmol. 1988;226(3):2246. 195. Furberg CD, Psaty BM, Meyer JV. Nifedipine: dose-related
177. Drance SM, Douglas GR, Wijsman K, Schulzer M, Britton increase in mortality in patients with coronary heart dis-
RJ. Response of blood flow to warm and cold in normal and ease. Circulation. 1995;92(5):132631.
low-tension glaucoma patients. Am J Ophthalmol. 1988; 196. Chobanian AV. Calcium channel blockers. Lessons learned
105(1):359. from MIDAS and other clinical trials. JAMA. 1996;276
178. Lewis RA, Vijayan N, Watson C, Keltner J, Johnson CA. Visual (10):82930.
field loss in migraine. Ophthalmology. 1989;96(3):3216. 197. Psaty BM, Heckbert SR, Koepsell TD, Siscovick DS,
179. Gasser P, Flammer J, Guthauser U, Mahler F. Do vasos- Raghunathan TE, Weiss NS, etal. The risk of myocardial
pasms provoke ocular diseases? Angiology. 1990;41(3): infarction associated with antihypertensive drug therapies.
21320. JAMA. 1995;274(8):6205.
180. Wang L, Cioffi GA, Van Buskirk EM, Zhao D-Y, Bacon 198. Borhani NO, Mercuri M, Borhani PA, Buckalew VM,
DR. Comparison of optic nerve blood flow measured with Canossa-Terris M, Carr AA, etal. Final outcome results of
laser Doppler flowmetry and microspheres. Invest the Multicenter Isradipine Diuretic Atherosclerosis Study
Ophthalmol Vis Sci. 1999;47(4 Suppl):S276. (MIDAS). A randomized controlled trial. JAMA. 1996;
181. Harino S, Riva CE, Petrig BL. Intravenous nicardipine in 276(10):78591.
cats increases optic nerve head but not retinal blood flow. 199. Wagenknecht LE, Furberg CD, Hammon JW, Legault C,
Invest Ophthalmol Vis Sci. 1992;33(10):288590. Troost BT. Surgical bleeding: unexpected effect of a
182. Yu DY, Alder VA, Su EN, Cringle SJ. Relaxation effects of calcium antagonist. BMJ. 1995;310(6982):7767.
diltiazem, verapamil, and tolazoline on isolated cat oph- 200. Kario K, Imiya M, Ohta Y, Shimada K. Gastrointestinal
thalmociliary artery. Exp Eye Res. 1992;55(5):75766. bleeding induced by amlodipine. J Hum Hypertens. 1995;
183. Kitazawa Y, Shirai H, Go FJ. The effect of Ca2+-antagonist 9(3):2067.
on visual field in low-tension glaucoma. Graefes Arch Clin 201. Pahor M, Guralnik JM, Furberg CD, Carbonin P, Havlik R.
Exp Ophthalmol. 1989;227(5):40812. Risk of gastrointestinal haemorrhage with calcium antago-
184. Netland PA, Chaturvedi N, Dreyer EB. Calcium channel nists in hypertensive persons over 67 years old. Lancet.
blockers in the management of low-tension and open-angle 1996;347(9008):10615.
glaucoma. Am J Ophthalmol. 1993;115(5):60813. 202. Pahor M, Guralnik JM, Salive ME, Corti MC, Carbonin P,
185. Gaspar AZ, Flammer J, Hendrickson P. Influence of nife- Havlik RJ. Do calcium channel blockers increase the risk
dipine on the visual fields of patients with optic-nerve-head of cancer? Am J Hypertens. 1996;9(7):6959.
diseases. Eur J Ophthalmol. 1994;4(1):248. 203. Daling JR. Calcium channel blockers and cancer: is an
186. Gaspar AZ, Gasser P, Flammer J. The influence of magne- association biologically plausible? Am J Hypertens.
sium on visual field and peripheral vasospasm in glaucoma. 1996;9(7):7134.
Ophthalmologica. 1995;209(1):113. 204. Friedland S, Kaplan S, Lahav M, Shapiro A. Proptosis and
187. Bose S, Piltz JR, Breton ME. Nimodipine, a centrally active periorbital edema due to diltiazem treatment. Arch
calcium antagonist, exerts a beneficial effect on contrast Ophthalmol. 1993;111(8):10278.
sensitivity in patients with normal-tension glaucoma and in 205. Galin MA, McIvor JW, Magruder GB. Influence of position
control subjects. Ophthalmology. 1995;102(8):123641. on intraocular pressure. Am J Ophthalmol. 1963;55:7203.
188. Sawada A, Kitazawa Y, Yamamoto T, Okabe I, Ichien K. 206. Pathak A, Raoul V, Montastruc JL, Senard JM. Adverse drug
Prevention of visual field defect progression with brovin- reactions related to drugs used in orthostatic hypotension:
a prospective and systematic pharmacovigilance study in 227. Ferrari-Dileo G, Davis EB, Anderson DR. Glaucoma, cap-
France. Eur J Clin Pharmacol. 2005;61(56):4714. illaries and pericytes. 3. Peptide hormone binding and influ-
207. Armaly MF, Salamoun SG. Schiotz and applanation tonom- ence on pericytes. Ophthalmologica. 1996;210(5):26975.
etry. Arch Ophthalmol. 1963;70:6039. 228. Zschauer AO, Davis EB, Anderson DR. Glaucoma, capillaries
208. Anderson DR, Grant WM. The influence of position on and pericytes. 4. Beta-adrenergic activation of cultured reti-
intraocular pressure. Invest Ophthalmol. 1973;12(3):20412. nal pericytes. Ophthalmologica. 1996;210(5):2769.
209. Wthrich UW. Postural change and intraocular pressure in 229. Orgl S, Bacon DR, Van Buskirk EM, Cioffi GA. Optic
glaucomatous eyes. Br J Ophthalmol. 1976;60(2):1114. nerve vasomotor effects of topical apraclonidine hydro-
210. Tsukahara S, Sasaki T. Postural change of IOP in normal chloride. Br J Ophthalmol. 1996;80(1):824.
persons and in patients with primary wide open-angle glau- 230. Bhandari A, Cioffi GA, Van Buskirk EM, Orgul S, Wang L.
coma and low-tension glaucoma. Br J Ophthalmol. 1984; Effect of brimonidine on optic nerve blood flow in rabbits.
68(6):38992. Am J Ophthalmol. 1999;128(5):6015.
211. Kothe AC. The effect of posture on intraocular pressure and 231. Tamaki Y, Araie M, Tomita K, Nagahara M. Effect of topi-
pulsatile ocular blood flow in normal and glaucomatous cal betaxolol on tissue circulation in the human optic nerve
eyes. Surv Ophthalmol. 1994;38(Suppl):S1917. head. J Ocul Pharmacol Ther. 1999;15(4):31321.
212. Zeimer RC. Circadian variations in intraocular pressure. In: 232. Bito LZ. Impact of intraocular pressure on venous outflow
Ritch R, Shields MB, Krupin T, editors. The glaucomas. from the globe: a hypothesis regarding IOP-dependent vas-
2nd ed. St. Louis: Mosby; 1996. p. 42945. cular damage in normal-tension and hypertensive glau-
213. Frampton P, Da Rin D, Brown B. Diurnal variation of coma. J Glaucoma. 1996;5(2):12734.
intraocular pressure and the overriding effects of sleep. Am 233. Hayreh SS, Jonas JB. Optic disk and retinal nerve fiber
J Optom Physiol Opt. 1987;64(1):5461. layer damage after transient central retinal artery occlusion:
214. Brown B, Morris P, Muller C, Brady A, Swann PG. an experimental study in rhesus monkeys. Am J Ophthalmol.
Fluctuations in intra-ocular pressure with sleep: I. Time 2000;129(6):78695.
course of IOP increase after the onset of sleep. Ophthalmic 234. Hayreh SS, Zimmerman MB, Kimura A, Sanon A. Central
Physiol Opt. 1988;8(3):2468. retinal artery occlusion. Retinal survival time. Exp Eye
215. Wildsoet C, Eyeson-Annan M, Brown B, Swann PG, Res. 2004;78:72336.
Fletcher T. Investigation of parameters influencing intraoc- 235. Cioffi GA, Van Buskirk ME. Vasculature of the anterior
ular pressure increases during sleep. Ophthalmic Physiol optic nerve and peripapillary choroid. In: Ritch R, Shields
Opt. 1993;13(4):35765. MB, Krupin T, editors. The glaucomas. 2nd ed. St Louis:
216. Buguet A, Py P, Romanet JP. 24-hour (nyctohemeral) and Mosby; 1996. p. 17788.
sleep-related variations of intraocular pressure in healthy 236. Schoute GJ. Vena Vorticosa im hinteren Bulbustheile. Von
white individuals. Am J Ophthalmol. 1994;117(3):3427. Albrecht Graefes Arch Ophthalmol. 1898;46:3579.
217. Brubaker RF. Flow of aqueous humor in humans [The 237. Coats G. On the pathology of chorio-vaginal (posterior vor-
Friedenwald Lecture]. Invest Ophthalmol Vis Sci. 1991; tex) veins. Ophthalmol Rev. 1906;25:99112.
32(13):314566. 238. Duke-Elder S. System of ophthalmology. 2nd ed. London:
218. Hayreh SS, Podhajsky P, Zimmerman MB. Beta-blocker Kimpton; 1961. p. 351, 8.
eyedrops and nocturnal arterial hypotension. Am J 239. Hayreh SS. Physiological anatomy of the choroidal vascu-
Ophthalmol. 1999;128(3):3019. lar bed. Int Ophthalmol. 1983;6(2):8593.
219. Klaver JH, Greve EL, Goslinga H, Geijssen HC, Heuvelmans 240. Hayreh SS. Anatomy and physiology of the optic nerve
JH. Blood and plasma viscosity measurements in patients head. Trans Am Acad Ophthalmol Otolaryngol. 1974;78:
with glaucoma. Br J Ophthalmol. 1985;69(10):76570. OP24054.
220. Foulds WS. 50th Bowman lecture. Blood is thicker than 241. Hayreh SS, Hayreh MS. Hemi-central retinal vein occul-
water. Some haemorheological aspects of ocular disease. sion. Pathogenesis, clinical features, and natural history.
Eye. 1987;1(Pt 3):34363. Arch Ophthalmol. 1980;98(9):16009.
221. Trope GE, Salinas RG, Glynn M. Blood viscosity in primary 242. Hayreh SS, March W, Phelps CD. Ocular hypotony follow-
open-angle glaucoma. Can J Ophthalmol. 1987;22(4):2024. ing retinal vein occlusion. Arch Ophthalmol. 1978;96(5):
222. Carter CJ, Brooks DE, Doyle DL, Drance SM. Investigations 82733.
into a vascular etiology for low-tension glaucoma. Oph 243. Hayreh SS, Zimmerman MB, Beri M, Podhajsky P.
thalmology. 1990;97(1):4955. Intraocular pressure abnormalities associated with central
223. Weinreb RN. Blood rheology and glaucoma. J Glaucoma. and hemicentral retinal vein occlusion. Ophthalmology.
1993;2:1534. 2004;111(1):13341.
224. OBrien C, Butt Z, Ludlam C, Detkova P. Activation of the 244. Hayreh SS, Peer J, Zimmerman MB. Morphologic changes
coagulation cascade in untreated primary open-angle glau- in chronic high-pressure experimental glaucoma in rhesus
coma. Ophthalmology. 1997;104(4):7259; discussion 930. monkeys. J Glaucoma. 1999;8(1):5671.
225. Donati G, Pournaras CJ, Munoz JL, Poitry S, Poitry-Yamate 245. Hayreh SS. Inter-individual variation in blood supply of the
CL, Tsacopoulos M. Nitric oxide controls arteriolar tone in optic nerve head. Its importance in various ischemic disor-
the retina of the miniature pig. Invest Ophthalmol Vis Sci. ders of the optic nerve head, and glaucoma, low-tension
1995;36(11):222837. glaucoma and allied disorders. Doc Ophthalmol. 1985;
226. Kondo M, Wang L, Bill A. The role of nitric oxide in hyper- 59(3):21746.
aemic response to flicker in the retina and optic nerve in 246. Hayreh SS. In vivo choroidal circulation and its watershed
cats. Acta Ophthalmol Scand. 1997;75(3):2325. zones. Eye. 1990;4(Pt 2):27389.
References 109
247. Hayreh SS. Pathogenesis of visual field defects. Role of the 257. Sato Y, Tomita G, Onda E, Goto Y, Oguri A, Kitazawa Y.
ciliary circulation. Br J Ophthalmol. 1970;54(5):289311. Association between watershed zone and visual field defect
248. Hayreh SS. Optic disc changes in glaucoma. Br J in normal tension glaucoma. Jpn J Ophthalmol. 2000;44(1):
Ophthalmol. 1972;56(3):17585. 3945.
249. Hayreh SS. Colour and fluorescence of the optic disc. 258. Hayreh SS, Joos KM, Podhajsky PA, Long CR. Systemic
Ophthalmologica. 1972;165(2):1008. diseases associated with nonarteritic anterior ischemic
250. Hayreh SS. Anterior ischaemic optic neuropathy. II. Fundus optic neuropathy. Am J Ophthalmol. 1994;118(6):76680.
on ophthalmoscopy and fluorescein angiography. Br J 259. Anderson DR. Glaucoma: the damage caused by pressure.
Ophthalmol. 1974;58(12):96480. XLVI Edward Jackson memorial lecture. Am J Ophthalmol.
251. Hayreh SS. The choriocapillaris. Albrecht Von Graefes 1989;108(5):48595.
Arch Klin Exp Ophthalmol. 1974;192(3):16579. 260. Hayreh SS. Progress in the understanding of the vascular
252. Hayreh SS. Submacular choroidal vascular pattern. etiology of glaucoma. Curr Opin Ophthalmol. 1994;5(11):
Experimental fluorescein fundus angiographic studies. 2635.
Albrecht Von Graefes Arch Klin Exp Ophthalmol. 1974; 261. Hayreh SS. Blood supply of the optic nerve head in health
192(3):18196. and disease. In: Lambrou GN, Greve EL, editors. Ocular
253. Hayreh SS. The long posterior ciliary arteries. An experi- blood flow in glaucoma: means, methods and measure-
mental study. Albrecht Von Graefes Arch Klin Exp ments. Amstelveen: Kugler & Ghedini; 1989. p. 354.
Ophthalmol. 1974;192(3):197213.
254. Hayreh SS. Segmental nature of the choroidal vasculature.
Br J Ophthalmol. 1975;59(11):63148.
255. Hayreh SS. Anterior ischemic optic neuropathy. New York:
Springer; 1975. For a complete bibliography and detailed review of previously
256. Hayreh SS. In: Bernstein EF, editor. Amaurosis fugax. New published studies in the literature on the subject, please refer to
York: Springer; 1988. p. 93113. bibliography in previous publications listed here.

Measurement of the Optic Nerve Head
Blood Flow 6
Ischemic disorders of the optic nerve head (ONH) Laser Doppler velocimeter
constitute an important cause of acute visual loss. Magnetic resonance imaging
Anterior ischemic optic neuropathy is a well-known dis- Microsphere method
order belonging to this category. Also, during the past Morphological studies
four decades a large body of evidence has emerged sug- Nonradioactive colored microspheres
gesting that vascular insufficiency in the ONH may play Oxygen tension method
an important role in development of glaucomatous optic Pulsatile ocular blood flow
neuropathy [111]. Both types of optic neuropathy are Scanning laser Doppler flowmetry
common and cause serious visual loss. All this has natu- Scanning laser fluorescein angiography
rally resulted in the current major thrust of research to Temperature measurement
explore the various aspects of the ONH circulation in Transcranial Doppler ultrasound
health and disease. To understand the mechanism and
Some of these techniques have been used only in
management of ischemic disorders of the ONH, the first
humans, others only in experimental animals, and still
essential is to be able to evaluate reliably the state of the
others in both. The literature is full of studies on these
circulation in the ONH, clinically and in experimental
various methods. Before accepting the data derived
studies. To achieve that, one has to have a reliable method
from these techniques, the first essential is to assess
of measuring the blood flow in the ONH. With the
their validity and the limitations of each method.
increased awareness of the role of vascular insufficiency
Reproducibility of findings of a method of testing
in the ONH, particularly in glaucomatous optic neuropa-
ONH blood flow has often been considered a proof of
thy, there has recently been a great surge of interest in
its reliability; however, that is not necessarily always
measuring the blood flow in a eye and ONH and, conse-
valid. This is because, if a method is providing invalid
quently, a search for methods to evaluate reliably the
information in the first place, the reproducibility of the
state of circulation in the ONH, both clinically, in patients
same invalid information does not make it a valid
with anterior ischemic optic neuropathy and glaucoma-
method to evaluate ONH blood flow. Moreover, some
tous optic neuropathy, and experimentally, in related
of the methods used to evaluate circulation in other
basic studies. Many innovative and diverse techniques
organs are unthinkingly applied to the ONH when in
have been used. Following is a list of the main various
fact the patterns of circulation and the tissue structure
methods employed or advocated in the literature:
of those organs have nothing in common with the
C-2-deoxyglucose (2-DG) method
14
ONH. The information from such studies does not nec-
Color Doppler imaging essarily enhance our understanding of the ONH blood
Fluorescein fundus angiography flow in health and disease, or ultimately, ischemic dis-
Heidelberg retina flowmeter orders of the ONH. All this has introduced a good deal
Hydrogen clearance method of confusion and controversy on the ONH blood flow
Iodoantipyrine autoradiography and our understanding of ischemic disorders of the
Laser Doppler flowmetry ONH. In view of this, accurate measurement of ONH
Laser speckle flowgraphy blood flow is an extremely important subject.

112 6 Measurement of the Optic Nerve Head Blood Flow
All those who design, use or interpret the findings disc. This fact, although crucial in the evaluation of
of methods used to evaluate the ONH blood flow to the ONH ischemic disorders, has been totally
investigate ischemic disorders of the ONH, have to be ignored by the advocates of many of the methods,
aware first and foremost of the following crucial fact such as laser Doppler flowmetry, scanning laser
about the blood supply of the ONH: Doppler flowmetry, Heidelberg retina flowmeter,
The ONH is supplied by two totally different temperature measurement, oxygen tension method,
sources the surface nerve fiber layer (i.e. surface hydrogen clearance method and most probably laser
layer of the optic disc) is supplied primarily by the speckle flowgraphy. Any method that measures
retinal arterial circulation, and the deeper parts (i.e. only or mostly the blood flow in the superficial
prelaminar, lamina cribrosa and retrolaminar layer of the ONH and not the deeper layers pro-
regions) are supplied mainly by the posterior ciliary vides invalid information related to ONH ischemic
artery (PCA) circulation (see Chap. 3; Figs.6.1 and disorders.
6.2). We have ample evidence now that the primary Fluorescein fundus angiography has provided the
site of lesion in both anterior ischemic optic neuropa- most vital proof on this topic. It has shown that when
thy [11,1316] and glaucomatous optic neuropathy fluorescein fills the capillaries in the surface layer of
[24,7,8,10] is in the part of the ONH nourished by the ONH, they act like a thick cloud, completely
the PCA circulation and NOT by the retinal circula- masking the blood vessels in the deeper layers of the
tion. Therefore, our primary objective is to obtain ONH, and that makes it impossible to evaluate the
information about the blood flow in the deeper layers circulation in the deeper layers of the ONH. This is
of the ONH only. very well demonstrated by the angiograms in [17]
In view of this, obviously any method used to Fig.6.3. This eye of a rhesus monkey had a cilioreti-
evaluate the role of the ONH circulatory disorders in nal artery; this artery was supplied by the PCA cir-
anterior ischemic optic neuropathy and glaucoma- culation. Experimental occlusion of the PCA in this
tous optic neuropathy must provide reliable infor- eye resulted in absence of filling of the deeper parts
mation about the blood flow in the deeper part of the of the ONH and choroid in the region of the occluded
ONH tissue and NOT the surface layer of the optic PCA, clearly seen before filling of the surface layer
of the ONH and the retina (Fig. 6.3a). During the
later phase of angiography (Fig. 6.3b), when
the retinal circulation filled completely, the part of
NFL the area supplied by the cilioretinal artery on the
R surface of the optic disc and in the retina did not fill.
In this angiogram (Fig.6.3b), in the area where the
C RA
surface layer of the optic disc filled completely, it
PLR
has totally masked the underlying absence of filling
S of the deeper layers of the ONH and the choroid;
however, where the surface layer of the disc supplied
by the cilioretinal artery did not fill (arrow), one can
LC still see non-filling of the deeper layers of the ONH.
Thus, all methods which measure the blood flow
from the surface of the optic disc are primarily
measuring the blood flow in the surface layer of the
Branch, ONH (i.e. the retinal arterial circulation) rather
CRA
PCA P than the PCA circulation supplying the deeper lay-
ON
ers and are irrelevant to ischemic disorders of
Fig.6.1 Schematic representation of blood supply of the optic theONH. This serious flaw was clearly demonstrated
nerve head. C choroid, CRA Central retinal artery, LC lamina by the experimental study by Petrig etal. [18]. and
cribrosa, NFL surface nerve fiber layer of the disc, ON optic
Wang et al. [19]. on laser Doppler flowmetry and
nerve, P pia, PCA posterior ciliary artery, PLR prelaminar
region, R retina, RA retinal arteriole, S sclera (Reproduced from scanning laser Doppler flowmetry respectively, dis-
Hayreh [7]) cussed below.
6 Measurement of the Optic Nerve Head Blood Flow 113
CZ
C
Retinal vein
Retinal artery R S
PCA
OD
PR
LC
PCA
Fig.6.2 Schematic representation of blood supply of the optic artery, PR prelaminar region, R retina, S sclera (Modified from
nerve head. C choroid, CZ circle of Zinn and Haller, LC lamina Hayreh [12])
cribrosa, layer of the disc, OD optic disc, PCA posterior ciliary
a b
Fig. 6.3 Fluorescein fundus angiograms of a rhesus monkey pigment at disc margin. (b) During the retinal arteriovenous
eye after experimental occlusion of the lateral PCA. (a) During phase: The retinal capillary bed has filled completely except in
the early retinal arterial phase: There is no filling of the tem- the area of supply by a cilioretinal artery, which did not fill
poral half of the choroid and optic disc, but normal filling of the because it was a branch of the lateral PCA. Note that the retinal
medial PCA, central retinal artery and nasal part of prelaminar capillary filling has completely masked the underlying infor-
region; also the superficial retinal capillaries fill in the inferior mation about the optic disc and choroid, except in the area
temporal part of the optic disc. Note: vessels in the nasal part where the retinal capillaries have not filled (arrow); that shows
of the prelaminar region fill from the adjacent peripapillary absence of filling of the choroid and optic disc (Reproduced
choroid, and the dark nasal crescent corresponds to superficial from Hayreh [17])
ethods Used to Measure Blood Flow

M measuring mostly the blood flow of the deeper PCA
of the ONH circulation without the masking effect by the overlying
superficial retinal circulation.
2. After Occlusion of the PCAs
Following is a brief discussion of the commonly used This was done to eliminate the PCA source of blood
methods, the validity of the data obtained and their flow to the deeper parts of the ONH, so that LDF was
limitations. measuring the blood flow in the surface layer supplied
by the retinal circulation, without the underlying cili-
ary circulation.
3. After Occlusion of Both the Central Retinal
Laser Doppler Flowmetry (LDF) Artery and the PCAs
This was meant to eliminate both sources of blood
This is a commonly used method to evaluate ONH supply to the ONH.
blood flow. This technique was described by Riva
LDF measured flow during central retinal artery
etal. [20]. in 1992. According to them, in LDF a laser
occlusion alone was highly significantly decreased (p
beam illuminates a small volume of a tissue and some
<0.001) compared to the normal baseline, by an aver-
of the light scattered by the tissue and red blood cells
age of 3921% (mean95% confidence interval).
is detected by a photodetector. They [20, 21] felt that
Combined occlusion of the central retinal artery and
the LDF measures a depth of greater than 520m of
PCAs reduced LDF flow by 5727% (p<0.0005).
the ONH and thus actually measured the blood flow
After occlusion of the PCAs alone, however, LDF
in the deeper layers of the ONH (i.e. the ones sup-
measurements showed a non-significant decrease in
plied by the PCA circulation) and not simply of the
LDF flow of 1737%. Thus, on average, central reti-
superficial nerve fiber layer which is supplied by the
nal artery occlusion alone significantly reduced LDF
retinal circulation [2, 2224]. Using LDF in a model
flow measurements, but PCA occlusion alone did not.
eye, Koelle etal. [21] reported that, invitro, the tech-
With the occlusion of both the central retinal artery
nique was able to detect the flow of microspheres in a
and the PCAs, LDF flow measurement was reduced
glass capillary tube behind a layer of excised ONH
more than with central retinal artery occlusion alone,
tissue of up to 600mm thickness [21] and concluded
but the difference between the flow reductions was not
that LDF measurements of blood flow in the superfi-
significant (p>0.20). The findings are discussed at
cial and deeper layers of the ONH are possible. Since
length elsewhere [18].
then a large number of studies have reported measure-
The results of this study showed that, with the stan-
ments of blood flow parameters in the ONH with the
dard LDF technique, the volume of the LDF flow
fundus camera-based LDF method [20, 25, 26], with
parameter includes both superficial and deeper tissue
a scanning laser ophthalmoscope-based LDF method
layers, but predominantly the blood flow in the super-
[27, 28], or with a laser Doppler velocimetry-based
ficial retinal layers and less from the prelaminar and
method [29, 30].
deeper regions of the ONH. In this LDF technique, the
The whole concept that LDF was measuring the
weaker Doppler signal from the deep layers cannot be
blood flow in the deeper layers of the ONH up to about
separated from the dominant signal from the superfi-
600mm depth was based on the invitro study by Koelle
cial layers, to study exclusively the circulation in the
etal. [21] only. To test the validity of that assumption
deep layers, which is of primary interest in ONH isch-
invivo, Petrig etal. [18] did an experimental study in
emic disorders, including anterior ischemic optic neu-
rhesus monkeys. In that study, LDF measurements
ropathy and glaucoma. Thus, the original invitro study
were first made in normal eyes. After that, LDF mea-
finding by Koelle et al. [21] that LDF penetrates to
surements were made following three types of proce-
ONH tissue of up to 600 mm thickness [21] was not
dures to obtain scientifically valid information from
valid about the invivo situation.
the various layers of the ONH circulation:
In this experimental study, contribution to the LDF
1. After Occlusion of the Central Retinal Artery blood flow measurements by the various sources of
This was done to eliminate the retinal source of blood blood supply in the ONH was studied by logical step-
flow to the surface layer of the ONH, so that LDF was by-step procedures, determining the contribution by
Methods Used to Measure Blood Flow of the ONH 115
different vascular layers in the ONH. It clearly showed The experimental studies in primates by both Petrig
that LDF does not provide valid information about the etal. [18] and Wang etal. [19] demonstrated, without
deeper layers of the ONH. In spite of that, it is sad to any ambiguity, that the penetration power of the laser
find that this method is still being advocated and used. in both LDF and scanning LDF is limited, and the
instruments primarily measure the microcirculation on
the surface layer of the ONH supplied by the central
retinal artery rather than that by the PCA. In spite of
Scanning Laser Doppler Flowmetry the fact that these studies showed that the information
provided by LDF and scanning LDF is not valid for
our understanding of the blood flow in ONH ischemia,
Michelson and Schmauss [27] combined laser Doppler
it is unfortunate to find that these invalid methods are
flowmeter with scanning laser ophthalmoscopy and
still extensively advocated and used and their results
claimed theoretically that with confocal scanning laser
applied in the understanding of the ONH ischemic dis-
Doppler flowmetry they could measure the blood flow
orders. That adds misinformation on the subject and in
in deeper layers of the ONH, by providing simultane-
the long run does harm to the cause of science. No
ously qualitative and quantitative evaluation of capil-
wonder we have so much conflicting and confusing
lary blood flow of distinct areas of a capillary
information on the subject.
meshwork. Nicolela etal. [31] used that technique to
measure the blood flow in the peripapillary retina and
lamina cribrosa in 92 glaucoma patients and 46 control
subjects, but obtained good quality images from only
33 glaucoma patients and 29 control subjects, and of Laser Speckle Flowgraphy
those qualitative examination of the vascular bed could
be made in 30 and 21 respectively; in the rest the qual- More recently, laser speckle flowgraphy has been used
ity of images was too poor to provide the required to evaluate the blood flow in the ONH, primarily by
information. Japanese investigators [3238]. This method is based
To estimate the measuring depth of the blood flow on a laser speckle phenomenon [39]. According to
with scanning LDF and to establish the vascular con- Tamaki etal. [32], based on estimations in human eyes,
tributions to these measurements, Wang etal. [19] con- the laser tissue penetration is about 0.51mm.
ducted an experimental study in primates somewhat As discussed above, systematic experimental stud-
similar to that by Petrig et al. [18] discussed above. ies in primates in LDF and scanning LDF have shown
They measured the ONH blood flow in each eye of that the original claims made about the layer of the
monkeys using scanning LDF before and following ONH of which those techniques measured the ONH
surgical occlusion of the central retinal artery or PCAs. blood flow reliably have been shown not to be valid. In
The regional blood flow in both eyes was also mea- laser speckle flowgraphy the measurements are made
sured using a nonradioactive microsphere method. by using the laser in a fashion similar to that in LDF
After central retinal artery occlusion, scanning LDF and scanning LDF. This raises exactly the same issue,
showed a reduction of 51% compared to normal base- about what exactly this method is measuring. There
line findings, while the microsphere method showed a has been no systematic experimental study so far on
reduction by 83% in the combined surface nerve fiber laser speckle flowgraphy, similar to those done in LDF
layer and the prelaminar regions. After PCA occlusion [18] and scanning LDF [19], to determine whether
only, scanning LDF showed no significant reduction claims made by various investigators using this tech-
(12%); however, the microsphere method showed a nique about ONH blood flow in the deeper layers are
51% reduction in the laminar and retrolaminar regions. valid or not. Until and unless that is done, one has to
They concluded that scanning LDF measures blood interpret its findings cautiously and accept the claimed
flow principally in the surface nerve fiber layer of the result of this technique with good deal of skepticism.
ONH, which is primarily supplied by the central reti- There also seem to be other problems with laser
nal artery. Blood flow in the laminar and retrolaminar speckle flowgraphy. For example, Tamaki et al. [40]
regions makes a small contribution to the scanning found that the laser speckle method does not give an
LDF measurement. absolute value for the tissue blood velocity, but a
relative value, because the normalized blur is also 1. There is a fundamental flaw in the entire concept
influenced by the reflectivity of the laser beam in the that color Doppler imaging measures the blood
tissue. Therefore, it is difficult to compare the normal- flow in the ONH. The finding that the main source
ized blur interindividually. Moreover, some investiga- of blood supply to the ONH is from the PCA cir-
tors [37] have noted other limitations of this technique. culation has misled researchers in the field into
They found that the laser speckle flowgraphy does not believing firmly that simple evaluation of the
measure the blood flow directly but relies on a linear blood flow in the PCAs in the orbit automatically
calibration curve; the estimated blood flow at certain gives information about the ONH circulation. The
extreme experimental conditions, such as very low reality is that almost invariably it does not, for sev-
ocular perfusion pressure, may deviate from the true eral reasons briefly discussed below.
values. Also, they found the maximum temporal reso- As explained in the previous chapter dealing with
lution of laser speckle flowgraphy for the blood flow in the blood supply of the ONH (see Chap. 3), there
the ONH measurement is once every 20s, 4s for each may be 15 main PCAs supplying an eye. The
measurement, and that is not fast enough to provide main PCAs divide into multiple branches before
valid information. entering the eyeball; these branches fall into three
groups supplying three different areas:
(a) Long PCAs: There are two long PCAs one
on the medial and the other on the lateral side
Heidelberg Retina Flowmeter (Fig.6.4). They mainly supply a segment of
the iris, ciliary body and peripheral choroid on
This method is based on confocal laser scanning and the medial and lateral sides respectively, and
LDF. It measures temporal intensity variation at each play no role in the ONH blood supply [55].
point in the two-dimensional scan field by multiple (b) Short PCAs (SPCAs): There may be any
scanning. The Doppler frequency shift of the reflected number up to 20 or so of these (Fig. 6.4a).
light is computed from the acquired data and used to They are further subdivided into two sub-
quantify the local velocity of moving particles. This groups [23, 24, 56]:
technique allows resolving blood flow in individual (i) Distal SPCAs: The majority of the short
capillaries and determines hemodynamic variables at PCAs belong to this group, and enter the
discrete locations of the retina and the optic disc [41]. eyeball midway between the optic nerve
A study showed that clinically, there is an unsatisfac- and the site of entrance of the long PCAs
tory intraindividual reproducibility of flow values in (Fig.6.4). They mainly supply the choroid.
different studied layers [42]. Thus, this method has all (ii) Paraoptic SPCAs: These are a few small
the limitations for measuring the ONH blood flow of SPCAs which enter the eyeball closest to
LDF and scanning LDF methods, and does not provide the optic nerve(Fig.6.4). Available evi-
suitable information about the blood flow in the deeper dence suggests that the ONH is mostly
layers of the ONH our area of primary interest in supplied by the paraoptic short PCAs.
ischemic disorders of the ONH. Thus, paraoptic short PCAs, which supply the
ONH, constitute only a tiny fraction of the entire
blood supply by the PCAs. Our primary interest is
in determining the blood flow in the paraoptic
Color Doppler Imaging SPCAs supplying the ONH and NOT in the rest
of the PCAs. This has the following practical
There has been a good deal of interest in this method implications:
of evaluating the ONH blood flow and vascular changes (a) In view of the above fact, the blood flow infor-
in glaucomatous optic neuropathy [4353]. This mation from the main PCAs or other SPCAs
method measures the velocity and vascular resistance has no relevance to the ONH blood supply.
index in the intraorbital vessels. Unfortunately, there (b) It is simply impossible for Color Doppler imag-
are several flaws in this method of determining the ing to differentiate individual paraoptic SPCAs
blood flow in the ONH. from the distal SPCAs and long PCAs because
Fig.6.4 (a) Diagrammatic a

representation of actual
branching pattern of medial
and lateral PCAs in 6 eyes. Long PCA
(b) Diagrammatic representa-
tion of the site of entry of the
various long and short PCAs,
as seen on the back of the
eyeball (Both a and b
reproduced from Hayreh [54])
OPTIC NERVE
LPCA
MPCA
MPCA
CA
MP
LP
CA LPCA
LPCA
LPC
LP MPCA
CA
CA
MPCA MP
b
Paraoptic
Long PCA Long PCA
SPCAs
Optic
Distal SPCAs Nerve Distal SPCAs
all of them are lying jumbled and intertwined about the amount of ONH blood flow. In fact, local-
as a vascular bundle (letalone to measure blood ized narrowing of a vessel produces an increase in
flow in them). I have done orbitotomy in about local Doppler velocity [58] and decreases the vol-
400 monkeys to perform various types of ume. Color Doppler imaging does not provide
experimental ocular vascular occlusion, and information about the lumen of the measured artery.
have seen this jumbled and intertwined mesh of Thus, it seems that the claims that this method is
ciliary arteries around the optic nerve. There is measuring blood flow in the ONH are not valid.
no way anyone can distinguish on color Doppler 3. In color Doppler imaging studies on ONH blood
imaging which artery one is dealing with. Thus flow, the frequently mentioned findings on veloc-
any claim that they are measuring the PCA ity and vascular resistance index in the central reti-
supply to the ONH has no validity. nal artery are totally irrelevant to the blood flow in
(c) The flow velocity and resistance index mea- the ONH, because that artery plays no role in the
sured by color Doppler imaging in the PCAs blood supply of the ONH (see Chap. 3). Similarly,
would not be altered even if all its paraoptic abnormal velocity and resistance in the main oph-
SPCAs were stenosed or completely occluded, thalmic artery on color Doppler imaging do not
because they constitute only a tiny fraction of necessarily mean reduced blood flow to the ONH,
the total blood flow, velocity and resistance in or vice versa, because the blood flow in the paraop-
the main PCAs (Fig.6.4a). tic short PCAs does not necessarily correlate with
(d) We know from carotid artery disease that that in the ophthalmic artery. For example, one or
stenosis of up to about 70% in the carotid more of the PCAs may be markedly stenosed or
arteries usually produces no hemodynami- completely occluded and yet the blood flow in the
cally significant blood flow reduction. ophthalmic artery may be perfectly normal.
(e) Our serotonin studies in atherosclerotic mon- Therefore, to apply color Doppler imaging find-
keys [57] showed that not all arteries are ings from the ophthalmic artery or central retinal
equally involved by the atherosclerotic pro- artery represents an unwarranted leap of faith.
cess and that may be true in other disease pro- 4. There are still other confounding factors applying
cesses as well. The paraoptic SPCAs may be blood flow findings from PCAs to the ONH circu-
diseased or occluded while the other SPCAs lation on color Doppler imaging studies, including
or the main PCAs may be uninvolved; under the following:
those circumstances blood flow in the PCAs (a) The blood supply by the PCA circulation in
may be within normal limits. the ONH is segmental in nature invivo (see
(f) Moreover, in the color Doppler imaging stud- Chap. 3).
ies, the terms PCA and short PCA are used (b) There is marked interindividual variation in
interchangeably, which simply confuses mat- thearea supplied by the various PCAs in the
ters. Again, many authors have claimed that ONH, and that makes the pattern of blood
they measured the flow in the PCA when in supply by the PCAs in the ONH extremely
fact they were evaluating one of the many variable (see Chap. 3)
small short PCAs in the retrobulbar region (c) As discussed above, the location of the
and not the PCA. Thus, evaluation of blood watershed zones between the PCAs in rela-
flow in the PCAs or the bundle containing tion to the optic disc plays an important role
their multiple branches gives no information in determining the vulnerability of the ONH
on the state of blood flow in the paraoptic to ischemic disorders (see Chap. 3).
SPCAs, which supply the ONH. Color Doppler imaging gives no information whatso-
2. The other fundamental flaw with measuring blood ever on any of these important issues.
flow in the ONH by color Doppler imaging is that These important flaws in the color Doppler imaging
it measures the velocity and vascular resistance method invalidate claims that this method measures
index. Unless the diameter of the lumen of the blood flow in the ONH. In spite of all these fundamen-
arteries supplying the ONH is known, those two tal flaws with color Doppler, unfortunately it is still
parameters do not necessarily provide information commonly used.
Transcranial Doppler Ultrasound is based on extrapolations and several assumptions

[69]. The two major assumptions are: (1) The pressure
volume relationship (ocular rigidity) is assumed to be
Using this method, Rojanapongpun and Drance [59,
standard for all persons. (2) The outflow to the venous
60] measured ophthalmic artery blood flow velocity in
system is presumed to be constant and not pulsatile.
eyes with primary open angle glaucoma, normal ten-
Neither assumption is true in the majority of persons.
sion glaucoma, and those without glaucoma. On
Moreover, this technique can only detect the pulsatile
immersing the hand in cold water, there was no change
component of blood flow and cannot estimate nonpul-
in the ophthalmic artery blood flow velocities [60]. In
satile ocular blood flow, and it is not known what pro-
a similar study, they investigated in glaucoma patients
portions of the total ocular blood flow are pulsatile
and normal persons the effect of nicotine on ophthal-
and nonpulsatile. The major part of the pulsatile ocu-
mic artery flow velocity, and compared that with glau-
lar blood flow is from the choroidal blood flow. Much
coma patients on placebo. They found that in the
more importantly, this test gives no information at all
ophthalmic artery peak (systolic) flow velocity, mean-
about the ONH blood flow, in spite of claims to that
envelope flow velocity and systolic BP significantly
effect. I liken the relationship of the choroidal vascu-
increased, while finger blood flow significantly
lar bed to the ONH vascular bed to that between a
decreased in the nicotine-tested glaucoma and normal
huge lake and a small pond situated near the lake and
persons when compared with the placebo-tested group
connected to it. The size of the waves in the big lake
[59]. The authors speculated that the increase in oph-
gives us no information at all about the state in the
thalmic artery velocity may reflect a mild vasocon-
pond! If I understand correctly, the ocular pulse
striction effect on the artery by nicotine or an increase
simply gives information about the height of the
in systemic BP. However, as discussed above, one has
waves. I feel we need much more information on the
to keep in mind that a reduction or increase in ophthal-
subject before we can assess the relevance of its find-
mic artery blood velocity does not necessarily mean a
ings to ONH blood flow. Moreover, total blood flow in
change in ONH blood flow.
the eye and changes in it do not give a true picture of
the blood flow in the ONH.
Another variant of this technique is ocular pneu-
moplethysmography where the intraocular pressure is
Pulsatile Ocular Blood Flow elevated by a suction cup [70]; Williamson and Harris
[69] have discussed the various drawbacks and prob-
This is another method often advocated to evaluate lems with this technique.
ONH blood flow. Langham etal. [61] and some other
authors [6267] have described pulsatile ocular blood
flow findings in patients with glaucomatous optic Fluorescein Fundus Angiography
neuropathy. They reported finding a significantly
reduced pulsatile ocular blood flow in normal tension Fluorescein fundus angiography has truly revolution-
glaucoma as compared to normals [64], and in primary ized our knowledge and understanding of the invivo
open angle glaucoma as compared to normal [63] and blood supply of the ONH and choroid, and their role in
ocular hypertensives [66], but no difference between ONH ischemic disorders; it constitutes one of the most
the normal and ocular hypertensives [62]. useful techniques in the study of various ocular vascu-
The exact relevance of pulsatile ocular blood flow lar disorders. Since 1964 my fluorescein angiographic
to blood flow in the ONH is very puzzling. This is studies have provided useful information about the
because ocular pulse measurements simply reflect the ONH circulation in health and disease [24, 10, 13
difference between systolic and diastolic BP in the 17, 2224, 57, 71]. Since we [71] first reported a
choroidal vascular bed, with an insignificant contribu- reduced fluorescence of the optic disc in glaucomatous
tion by the ONH or retinal vascular bed. Krakau [68], optic neuropathy suggestive of interference with the
based on a number of assumptions, described theoreti- ONH blood supply, many papers have described
cally a way to calculate pulsatile ocular blood flow. angiographic abnormalities in glaucomatous optic
The claim of blood flow measurement by this method neuropathy. Following is a review of the earlier
studies. In 1977 [7] I reviewed various angiographic because of changes in the ocular media, inadequate
studies dealing with the ONH, peripapillary choroid dilation of the pupil, and poor ocular and/or systemic
and the rest of the choroid in glaucoma. Schwartz and circulation.
co-workers [7277] published a series of papers on the 3. Inadequate Resolution of Disc Microvasculature
subject showing that optic disc filling defects, seen on The optic disc is only 1.5 mm in diameter, with the
fluorescein angiography, represent areas of ischemia vessels in it arranged in multiple superimposed layers.
which are highly correlated with loss of visual field Fluorescein angiography is incapable of resolving the
[72], and that the filling defects are more marked in disc microvasculature of various layers simultaneously
glaucomatous discs than in ocular hypertensives and and to demonstrate fine subtle vascular defects satis-
least in normal discs [73]. Their studies further sug- factorily (Fig.6.3).
gested a decreased blood flow and a smaller vascular 4. Difficulty in Outlining the Deeper ONH
bed together with narrower vessels in hypofluorescent Capillaries
areas of the optic disc [75]. They further confirmed the We are essentially interested in the prelaminar and
relationship of fluorescein filling defects in the optic deeper capillaries of PCA origin in the ONH in these
disc and nerve fiber layer defects to glaucomatous eyes. As discussed above and shown by Fig.6.3, once
abnormalities, and thus the association between vas- the dense capillary network of retinal origin, lying in
cular damage to the ONH and axonal loss in glaucoma the surface nerve fiber layer of the optic disc, is filled,
[77]. Geijssen [8] found angiographic filling defects of it completely masks the underlying ciliary vessels so
disc in 92% of normal tension glaucoma cases. that no information at all can be obtained about the
While fluorescein angiography has provided very deeper capillaries.
useful information about the blood supply of the ONH 5. Optic Disc Fluorescence on Fluorescein
in health and disease, it is essential to realize that this Angiography
technique does have multiple limitations. If these limi- In fluorescein fundus angiography investigators have
tations are not kept in mind, serious misinterpretations used the degree of fluorescence of the optic disc as one
can occur, and from them misleading information, of the parameters of ONH vascularity, under the impres-
such as one finds not infrequently in the literature. sion that that is an index of quantity of the ONH circu-
lation. In view of this, it is pertinent to examine the
sources of normal optic disc fluorescence, as I have
done [17]. Usually, fluorescence of the disc before the
Limitations of Fluorescein Fundus Angiography
retinal arterial filling is due to filling of the ciliary ves-
sels in the prelaminar region and/or deeper down. When
In the study of ONH circulation, fluorescein angiogra-
the retinal capillaries in the surface nerve fiber layer of
phy is a very crude and not always a reliable method,
the disc fill, they completely mask the deeper ciliary
for the following reasons:
vessels, so that the fluorescence of the disc then is
1. Reliability of a Single Angiographic Examination essentially due to the retinal capillary filling of the
Like a single recording of blood pressure, information superficial layer only, with little or no contribution from
obtained from a single angiographic examination may the deeper ciliary vessels (see Fig.6.3) [2, 15, 17, 22].
be misleading. The excitement and worry of visiting During the late phase, it is due to staining of the prelam-
the hospital, and of angiography, invariably alter the inar and lamina cribrosa regions, especially the colla-
systemic blood pressure, and give totally misleading gen tissues of the latter, by fluorescein leakage from the
information about the ocular circulation. Thus, a nor- adjacent peripapillary choroid. In experimental studies
mal angiogram does not rule out the possibility of cir- on rhesus monkey eyes where I cut PCAs or the ocular
culatory disturbances in the ONH and/or choroid at circulation was completely obliterated by a very high
other times, particularly during sleep in patients with IOP, deeper layers of the ONH still showed fluores-
marked nocturnal hypotension [78, 79]. cence in some eyes and not in others; the origin of that
2. Unsatisfactory Quality of the Angiograms fluorescence is obscure. This was also recorded in some
In a fair proportion of patients with ONH ischemic dis- human eyes before the PCAs and central retinal artery
orders, the quality of angiograms is not good enough started to fill. Before the capillaries in the surface layer
to outline vasculature in the optic disc and choroid of the disc fill, the temporal part of the disc is usually
more fluorescent than the nasal side we still have no Scanning Laser Fluorescein Angiography
satisfactory explanation for this. Also puzzling are the
inconsistencies, in both normal and atrophic discs,
Cantor etal. [80] have claimed that with this technique
between the ophthalmoscopic appearance and the fluo-
video fluorescein angiograms of high spatial and tem-
rescein angiographic findings during the late phase: (a)
poral resolution are produced, and digital image analy-
Why is there such a variable degree of fluorescence,
sis of the angiograms is used to measure capillary flow
with no evident ophthalmoscopic or etiologic correla-
velocities. The authors measured the blood flow veloc-
tion? (b) Why do some discs show mild to marked
ities in the superficial optic disc capillaries supplied by
staining in one or another sector of the peripapillary
the retinal circulation. Unfortunately, as far as one can
choroid while others show none? (c) Why does peri-
tell, this method suffers from the same limitations as
papillary chorioretinal degeneration, seen most fre-
fluorescein fundus angiography in providing informa-
quently in cases of primary open angle glaucoma and
tion about the blood flow in the deeper layers of the
normal tension glaucoma, cause leakage of fluorescein
ONH supplied by the PCA circulation the primary
into the adjacent disc tissue in some and not in others?
area of interest in ischemic disorders of the ONH.
All these variations may be manifestations of the
marked interindividual variation in the ONH blood
supply, discussed at length above and elsewhere (see
Chap. 3) [23, 24]. Magnetic Resonance Imaging
6. Significance of Presence or Absence of a Filling
Defect on Fluorescein Fundus Angiography
While the presence of a filling defect on fluores- Prunte etal. [81] imaged the rat optic nerve using a
cein angiography indicates vascular insufficiency, T2-weighted magnetic resonance imaging technique
the absence of such a defect does not mean that and contrast agent diethylenetriaminepentaacetate-
there has never been vascular insufficiency. My gadolinium, and calculated blood flow by image
clinical and experimental studies on acute vascular analysis from the change in signal intensity. They
occlusion involving the PCAs, central retinal claimed that their findings suggest that magnetic res-
artery, optic disc vessels and iris vessels have onance imaging provides quantification of optic
clearly demonstrated that even in well documented nerve blood flow. Garcia etal. [82] in a small prelimi-
arterial occlusions, days or weeks later the circula- nary study claimed that qualitative perfusion imaging
tion usually returns to normal although permanent of the human optic nerve is feasible. However, the
ischemic damage is done to the involved tissues. major limiting factor with this method is the very
In conditions where there is only transient isch- small size of the ONH and limited resolution of the
emia or hypoxia due to a temporary fall in the per- technique to detect focal or mild circulatory insuffi-
fusion pressure (e.g., in nonarteritic anterior ciency in the ONH.
ischemic optic neuropathy and glaucomatous optic
neuropathy during nocturnal arterial hypotension
[78, 79]), fluorescein fundus angiography during
the waking hours may fail to demonstrate any fill- Temperature Measurement
ing defect.
From this discussion it is evident that fluorescein Bill [83], in cats and rabbits, using a heated thermo-
fundus angiography has limited usefulness in any couple method to the surface of the eyeball, found that
sophisticated study of the circulation in ONH ischemic an increase in intraocular pressure decreased the uveal
disorders, and these limitations must be borne in mind blood flow. Ernest and Potts [84] measured blood flow
while interpreting the results. The recent introduction in the ONH by inserting a thermocouple needle into
of fluorescein videoangiography and scanning laser the optic disc via the cornea and vitreous, and found
ophthalmoscopy may help to overcome some of those that elevation of intraocular pressure within the range
limitations. In spite of all these limitations, fluorescein of primary open angle glaucoma caused decrease in
fundus angiography still does provide useful informa- blood flow in the optic disc and a fall in intraocular
tion about the ONH circulation. pressure increased the blood flow, and concluded that
there was no autoregulation in the intraocular pressure. be shown that the findings by the laser speckle method
The fact that now it is established that ONH has auto- are reliable for measuring ONH blood flow.
regulation would indicate that this method did not pro-
vide correct information.
This method has other limitations as well for evalu- C-2-Deoxyglucose (2-DG) Method
14
ation of ONH blood flow. First, it is an indirect method
for blood flow measurement. Second, most impor-
tantly, since it measures the temperature from the sur- Sokoloff etal. [88] developed this method for evalua-
face of the optic disc which is supplied by the central tion of blood flow in experimental studies in animals.
retinal artery circulation, it suffers from the same prob- This is an indirect test to determine the glucose con-
lem as discussed above regarding LDF and scanning sumption by the tissues. The basis of the test is that
LDF, i.e., it does not give any information about the
14
C-deoxyglucose, with the same transport system as
deeper layers of the ONH circulation our primary glucose, passes through capillaries and into cells. In
interest. the cells it is phosphorylated but cannot be metabo-
lized any further. After phosphorylation, it cannot
leave the cells. Thus, its accumulation in the cells
reflects glucose consumption. Bill and his co-workers
Oxygen Tension Method [8992] used this method in a number of studies on the
blood flow of the ONH in monkeys. Sperber and Bill
[90] found no appreciable effect on the 2-DG accumu-
Ernest [85] measured the oxygen tension 10 mm in lation in the retina and prelaminar region of the ONH
front of the center of the optic disc in cats and showed at perfusion pressure of 40cm water but at perfusion
autoregulation of the optic disc oxygen. As discussed pressure of 20cm water 2-DG accumulation was mark-
above, the surface layer of the disc is supplied by reti- edly increased, indicating ischemia resulting in anaer-
nal circulation so that this method, again, does not give obic glycolysis. There was no change in the lamina
any information about the deeper layers of the ONH cribrosa and behind the lamina cribrosa. Using this and
supplied by the PCA circulation. It suffers from the microsphere methods, they [91] found that normally
same limitation discussed above, concerning LDF and the blood flow in the lamina cribrosa is about five times
scanning LDF. Moreover, it is well-established that the more than that 5 mm behind the lamina cribrosa. At
retinal circulation has autoregulation. perfusion pressure of 15mmHg, there was a marked
reduction in blood flow in the prelaminar region and
also within the anteriormost part of the lamina crib-
rosa. At 30mmHg perfusion pressure, the flow in the
Hydrogen Clearance Method retina and the prelaminar region tended to be moder-
ately reduced while that in the lamina cribrosa and
This method was first reported by Aukland etal. [86] behind was normal. They concluded that their studies
and has been used in the cerebral cortex, myocardium indicate that partial ischemia and anaerobic glycolysis
and other organs. Ernest [87] measured the blood flow was produced in a normal ONH only at perfusion pres-
at the lamina cribrosa in rhesus monkeys by inserting a sure of <30mmHg, and it was marked at 15mmHg.
microelectrode in the lamina cribrosa, and found that They suggested that glaucoma developing at intraocu-
the blood flow rapidly decreased when the perfusion lar pressure of <2530 mmHg has its main cause in
pressure reached 50 mmHg. Since then this method vascular insufficiency. At intraocular pressure of
has been used by some other workers. For example, 4050mmHg or more, partial ischemia will be another
Sugiyama etal. [33] compared the ONH blood flow in damaging factor. In the retina, laminar and prelaminar
rabbits by comparing the findings between the laser regions and the first few mm behind the lamina crib-
speckle method and hydrogen clearance method, and rosa, the glucose consumption is similar to that in the
stated that the findings of the two methods were com- gray matter of the brain while in the retrolaminar part
parable. As discussed above regarding the laser speckle it was more like the white matter. We [92] used the
method, it has some limitations and it still remains to 2-DG method in old, atherosclerotic rhesus monkeys,
Morphologic Studies 123
and found that there was defective autoregulation in Microsphere Method

the retina and ONH in these monkeys. It would seem
that for experimental studies, this technique is more
Unlabelled or radioactive labelled or nonradioactive
physiologic than many of the other experimental meth-
colored microspheres of varying sizes have been used to
ods; however, it is a laborious method and requires
study the ocular and ONH blood flow experimentally in
great expertise to obtain satisfactory results.
animals [37, 98101]. While this method has provided
useful information on the ocular blood flow in many
instances, it does have limitations in the ONH. For
example, spatial resolution of the microsphere method
Iodoantipyrine Autoradiography is limited even if a large number of small spheres is used
[89]. In small tissues, such as ONH, the specimens con-
This method has been used by a number of investiga- tain a small number of microspheres which can intro-
tors in experimental studies in animals to evaluate the duce an error and give relatively low blood flow and a
blood flow [9396]. Sossi and Anderson [93] recorded high variability [102]. Also the microspheres can par-
blood flow in the retina, choroid and ONH in cats by tially obstruct the vessels and change the flow pattern.
using Iodoantipyrine125 and found autoregulation of For example, Wang etal. [103] found that 8mm micro-
blood flow throughout the range of IOP until the perfu- spheres are the optimal size for both retinal and optic
sion pressure was reduced to 25 mmHg. At higher nerve blood flow estimation and the 10mm for the chor-
IOPs, the blood flow was reduced in the retina, choroid oid; but 15mm are inappropriate for ocular blood flow
and ONH but in the lamina cribrosa it was reduced measurements. Axial streaming can cause further prob-
only with extreme pressure elevation and not reduced lems. All these limitations restrict the usefulness of this
at all in the intraorbital optic nerve. They concluded technique in evaluation of the blood flow in the ONH.
that there is efficient autoregulation in the ONH and
lamina cribrosa, so that the IOP over a wide range does
not much influence blood flow normally. Weinstein
etal. [94] used 14C-iodoantipyrine in cats and found a Morphologic Studies
gradient of flow rates in different parts of the ONH,
with relatively high flow in the prelaminar, laminar There are a number of morphologic studies on the eye-
and 1mm retrolaminar regions, and significantly lower ball and ONH of glaucoma and anterior ischemic optic
flow in the 46 mm posterior to the laminar region. neuropathy patients. While these studies give useful
They found that ONH blood flow demonstrated auto- information about morphologic changes in the tissues
regulation in all areas. of the ONH, they are not of any help in giving informa-
This method gives us information more about the tion about disturbances in the invivo blood flow in the
diffusion than the blood flow in the ONH [97]. ONH. In fact, available evidence indicates that such
Caprioli and Miller [96] showed that blood flow mea- studies can be highly misleading. For example, Quigley
surements in the ONH with this technique may be etal. [104] from a histologic examination of the ONH
largely influenced by diffusion of the tracer from the in eyes with primary open angle glaucoma, reported
choroid into the ONH. They found that with a total that many capillaries are present in the optic nerve
diffusion time of 60s, real blood flow may differ as heads with early and moderate glaucomatous loss of
much as at 60% 50 mm from the choroid and about nerve fibers. This observation was interpreted by them
12% at 300mm from the choroid. The blood flow gra- to mean that no vascular disorder is responsible for the
dient reported from anterior to posterior optic nerve ONH changes in glaucoma. In this connection, one has
with this technique may simply be due to diffusion to bear in mind two well-established facts: (1) capillar-
from the surrounding choroid. Thus, blood flow mea- ies are far more resistant to ischemic damage than the
surements of the ONH obtained with this technique neural tissue, and (2) the presence of capillaries, as
must be cautiously interpreted, and evidently this is seen in histologic section, in no way proves that there is
not an ideal method to evaluate the blood flow in the normal or adequate blood flow or that ischemia was not
ONH in glaucoma [97]. present. Similarly, the presence of red blood cells in
capillaries on histologic examination [105], also in no 3. Hayreh SS. Pathogenesis of visual field defects. Role of the
way indicates a normal or adequate blood flow in the ciliary circulation. Br J Ophthalmol. 1970;54(5):289311.
4. Hayreh SS, Revie IH, Edwards J. Vasogenic origin of visual
ONH. Therefore, this is not a suitable method to obtain field defects and optic nerve changes in glaucoma. Br J
information about the invivo circulation in the ONH. Ophthalmol. 1970;54(7):46172.
5. Kaiser HJ, Flammer J, Hendrickson P, editors. Ocular blood
flow: new insights into the pathogenesis of ocular diseases.
New York: Karger; 1996.
6. Spaeth GL. The pathogenesis of nerve damage in glaucoma:
Conclusions contributions of fluorescein angiography. New York: Grune
& Stratton; 1977.
7. Hayreh SS. Pathogenesis of optic nerve damage and visual
As is evident from the above brief review, the bottom field defects. In: Heilmann K, Richardson KT, editors.
line is that we do not have a single method that gives us Glaucoma: conceptions of a disease. Stuttgart: Georg Theime
clinically reliable information on the blood flow in the Publishers; 1978. p. 10437.
ONH in man neither in health nor in disease. All the 8. Geijssen HC. Studies on normal pressure glaucoma.
Amstelveen: Kugler & Ghedini; 1991.
methods discussed above have various limitations. More 9. Drance SM. International symposium on glaucoma, ocular blood
importantly, many of the claims made about the various flow, and drug treatment. Baltimore: Williams & Wilkins; 1992.
methods of measuring the ONH blood flow, relevant to 10. Hayreh SS. Progress in the understanding of the vascular
anterior ischemic optic neuropathy or glaucomatous etiology of glaucoma. Curr Opin Ophthalmol. 1994;5(2):
2635.
optic neuropathy, are not valid for these diseases. This 11. Hayreh SS. Acute ischemic disorders of the optic nerve:
latter fact has introduced misinformation about the pathogenesis, clinical manifestations, and management.
pathogenesis and management of these disorders. Ophthalmol Clin North Am. 1996;9(3):40742.
Unless and until we have a satisfactory method of 12. Hayreh SS. Trans Am Acad Ophthalmol Otolaryngol. Anatomy
and physiology of the optic nerve 1974;78:OP24054.
measuring and manipulating the blood flow in the ONH, 13. Hayreh SS, Baines JA. Occlusion of the posterior ciliary
we are not going to solve the problem of logical manage- arteryIII. Effects on the optic nerve head. Br J Ophthalmol.
ment of ischemic disorders of the ONH and prevent 1972;56(10):75464.
patients from going blind [4]. Ophthalmologists need 14. Hayreh SS. Anterior ischemic optic neuropathy. Arch
Neurol. 1981;38(11):6758.
readily available, simple, non-invasive and reliable clini- 15. Hayreh SS, Chopdar A. Occlusion of the posterior ciliary
cal method(s) that can be performed easily in their daily artery. V. Protective influence of simultaneous vortex vein
practice on patients with glaucoma, anterior ischemic occlusion. Arch Ophthalmol. 1982;100(9):148191.
optic neuropathy and other ONH vascular disorders we 16. Hayreh SS. Ischemic optic neuropathy. Prog Retin Eye Res.
2009;28(1):3462.
have none so far. Moreover, because of the recently dis- 17. Hayreh SS. Colour and fluorescence of the optic disc.
covered importance of nocturnal arterial hypotension in Ophthalmologica. 1972;165(2):1008.
the development and progression of non-arteritic anterior 18. Petrig BL, Riva CE, Hayreh SS. Laser Doppler flowmetry
ischemic optic neuropathy [78, 79, 106] and glaucoma and optic nerve head blood flow. Am J Ophthalmol.
1999;127(4):41325.
[78, 79, 107111], we need a method of measuring the 19. Wang L, Cull G, Cioffi GA. Depth of penetration of scan-
ONH blood flow continuously or frequently over a ning laser Doppler flowmetry in the primate optic nerve.
24-hour period or longer (similar to the 24-hour ambula- Arch Ophthalmol. 2001;119(12):18104.
tory blood pressure monitor), particularly during sleep. 20. Riva CE, Harino S, Petrig BL, Shonat RD. Laser Doppler
flowmetry in the optic nerve. Exp Eye Res. 1992;55(3):
The lack of availability of these facilities makes manage- 499506.
ment of ONH ischemic disorders and prevention of blind- 21. Koelle JS, Riva CE, Petrig BL, Cranstoun SD. Depth of tis-
ness caused by them a hard task. sue sampling in the optic nerve head using laser Doppler
flowmetry. Lasers Med Sci. 1993;8(1):4954.
22. Hayreh SS, Perkins ES. Clinical and experimental studies on
the circulation at the optic nerve head. Proceedings of the Wm
References MacKenzie centenary symposium on the ocular circulation in
health and disease; 1968. Glasgow: Kimpton; 1969. p. 7186.
23. Hayreh SS. The 1994 Von Sallman lecture. The optic nerve
1. Hayreh SS. Blood supply and vascular disorders of the optic head circulation in health and disease. Exp Eye Res.
nerve. An Inst Barraquer. 1963;4:7109. 1995;61(3):25972.
2. Hayreh SS. Blood supply of the optic nerve head and its role 24. Hayreh SS. The Blood supply of the optic nerve head and
in optic atrophy, glaucoma, and oedema of the optic disc. Br the evaluation of it: myth and reality. Prog Retin Eye Res.
J Ophthalmol. 1969;53(11):72148. 2001;20(5):56393.
References 125
25. Harino S, Riva CE, Petrig BL. Intravenous nicardipine in 42. Jonescu-Cuypers CP, Harris A, Bartz-Schmidt KU,
cats increases optic nerve head but not retinal blood flow. Kagemann L, Boros AS, Heimann UE, etal. Reproducibility
Invest Ophthalmol Vis Sci. 1992;33(10):288590. of circadian retinal and optic nerve head blood flow mea-
26. Harris A, Anderson DR, Pillunat L, Joos K, Knighton RW, surements by Heidelberg retina flowmetry. Br J Ophthalmol.
Kagemann L, etal. Laser Doppler flowmetry measurement 2004;88(3):34853.
of changes in human optic nerve head blood flow in response 43. Galassi F, Nuzzaci G, Sodi A, Casi P, Vielmo A. Color
to blood gas perturbations. J Glaucoma. 1996;5(4):25865. Doppler imaging in evaluation of optic nerve blood supply
27. Michelson G, Schmauss B. Two dimensional mapping of the in normal and glaucomatous subjects. Int Ophthalmol.
perfusion of the retina and optic nerve head. Br J Ophthalmol. 1992;16(45):2736.
1995;79(12):112632. 44. Galassi F, Nuzzaci G, Sodi A, Casi P, Cappelli S, Vielmo A.
28. Groh MJ, Michelson G, Langhans MJ, Harazny J. Influence Possible correlations of ocular blood flow parameters with
of age on retinal and optic nerve head blood circulation. intraocular pressure and visual-field alterations in glaucoma:
Ophthalmology. 1996;103(3):52934. a study by means of color Doppler imaging. Ophthalmologica.
29. Hamard P, Hamard H, Dufaux J, Quesnot S. Optic nerve 1994;208(6):3048.
head blood flow using a laser Doppler velocimeter and hae- 45. Harris A, Sergott RC, Spaeth GL, Katz JL, Shoemaker JA,
morheology in primary open angle glaucoma and normal Martin BJ. Color Doppler analysis of ocular vessel blood
pressure glaucoma. Br J Ophthalmol. 1994;78(6):44953. velocity in normal-tension glaucoma. Am J Ophthalmol.
30. Netland PA, Feke GT, Konno S, Goger DG, Fujio N. Optic 1994;118(5):6429.
nerve head circulation after topical calcium channel blocker. 46. Harris A, Arend O, Kopecky K, Caldemeyer K, Wolf S,
J Glaucoma. 1996;5(3):2006. Sponsel W, et al. Physiological perturbation of ocular and
31. Nicolela MT, Hnik P, Drance SM. Scanning laser Doppler cerebral blood flow as measured by scanning laser ophthal-
flowmeter study of retinal and optic disk blood flow in glau- moscopy and color Doppler imaging. Surv Ophthalmol.
comatous patients. Am J Ophthalmol. 1996;122(6):77583. 1994;38(suppl):S816.
32. Tamaki Y, Araie M, Kawamoto E, Eguchi S, Fujii H. Non- 47. Trible JR, Sergott RC, Spaeth GL, Wilson RP, Katz LJ,
contact, two-dimensional measurement of tissue circulation Moster MR, etal. Trabeculectomy is associated with retrob-
in choroid and optic nerve head using laser speckle phenom- ulbar hemodynamic changes. A color Doppler analysis.
enon. Exp Eye Res. 1995;60(4):37383. Ophthalmology. 1994;101(2):34051.
33. Sugiyama T, Utsumi T, Azuma I, Fujii H. Measurement of 48. Rankin SJ, Walman BE, Buckley AR, Drance SM. Color
optic nerve head circulation: comparison of laser speckle Doppler imaging and spectral analysis of the optic nerve vascu-
and hydrogen clearance methods. Jpn J Ophthalmol. lature in glaucoma. Am J Ophthalmol. 1995;119(6):68593.
1996;40(3):33943. 49. Yamazaki Y, Hayamizu F. Comparison of flow velocity of oph-
34. Tamaki Y, Araie M, Tomita K, Nagahara M, Tomidokoro A, thalmic artery between primary open angle glaucoma and nor-
Fujii H. Real-time measurement of human optic nerve head mal tension glaucoma. Br J Ophthalmol. 1995;79(8):7324.
and choroid circulation, using the laser speckle phenome- 50. Butt Z, McKillop G, OBrien C, Allan P, Aspinall P. Measurement
non. Jpn J Ophthalmol. 1997;41(1):4954. of ocular blood flow velocity using colour Doppler imaging in
35. Yaoeda K, Shirakashi M, Funaki S, Funaki H, Nakatsue T, low tension glaucoma. Eye. 1995;9(Pt 1):2933.
Abe H. Measurement of microcirculation in the optic nerve 51. Netland PA, Grosskreutz CL, Feke GT, Hart LJ. Color Doppler
head by laser speckle flowgraphy and scanning laser Doppler ultrasound analysis of ocular circulation after topical calcium
flowmetry. Am J Ophthalmol. 2000;129(6):7349. channel blocker. Am J Ophthalmol. 1995;119(6):694700.
36. Yaoeda K, Shirakashi M, Funaki S, Funaki H, Nakatsue T, 52. Nicolela MT, Buckley AR, Walman BE, Drance SM. A
Fukushima A, et al. Measurement of microcirculation in comparative study of the effects of timolol and latanoprost
optic nerve head by laser speckle flowgraphy in normal vol- on blood flow velocity of the retrobulbar vessels. Am J
unteers. Am J Ophthalmol. 2000;130(5):60610. Ophthalmol. 1996;122(6):7849.
37. Liang Y, Downs JC, Fortune B, Cull G, Cioffi GA, Wang L. 53. Nicolela MT, Walman BE, Buckley AR, Drance SM. Ocular
Impact of systemic blood pressure on the relationship hypertension and primary open-angle glaucoma: a com
between intraocular pressure and blood flow in the optic parative study of their retrobulbar blood flow velocity.
nerve head of nonhuman primates. Invest Ophthalmol Vis J Glaucoma. 1996;5(5):30810.
Sci. 2009;50(5):215460. 54. Hayreh SS. Blood supply of the optic nerve head A reality
38. Sugiyama T, Araie M, Riva CE, Schmetterer L, Orgul S. Use check. In: Pillunat LE, Harris A, Anderson DR, Greve EL,
of laser speckle flowgraphy in ocular blood flow research. editors. Current concepts on ocular blood flow in glaucoma.
Acta Ophthalmol. 2010;88(7):7239. The Hague: Kugler; 1999. p. 331.
39. Fujii H, Nohira K, Yamamoto Y, Ikawa H, Ohura T. 55. Hayreh SS. The long posterior ciliary arteries: an experi-
Evaluation of blood flow by laser speckle image sensing. mental study. Albrecht Von Graefes Arch Klin Exp
Part 1. Appl Opt. 1987;26(24):53215. Ophthalmol. 1974;192(3):197213.
40. Tamaki Y, Araie M, Hasegawa T, Nagahara M. Optic nerve 56. Ducournau D. La systmatisation vasculaire de la chorode.
head circulation after intraocular pressure reduction achieved [Vascularization of the choroid]. Annee Ther Clin Ophtalmol.
by trabeculectomy. Ophthalmology. 2001;108(3):62732. 1979.
41. Bohdanecka Z, Orgul S, Prunte C, Flammer J. Influence of 57. Hayreh SS, Piegors DJ, Heistad DD. Serotonin-induced con-
acquisition parameters on hemodynamic measurements with striction of ocular arteries in atherosclerotic monkeys: impli-
the Heidelberg retina flowmeter at the optic disc. J Glaucoma. cations for ischemic disorders of the retina and optic nerve
1998;7(3):1517. head. Arch Ophthalmol. 1997;115(2):2208.
58. Hayreh SS, Beach KW. Optic nerve sheath decompression 78. Hayreh SS, Zimmerman MB, Podhajsky P, Alward WLM.
may improve blood flow in anterior ischemic optic neuropa- Nocturnal arterial hypotension and its role in optic nerve
thy: discussion. Ophthalmology. 1993;100(3):3035. head and ocular ischemic disorders. Am J Ophthalmol.
59. Rojanapongpun P, Drance SM. The effects of nicotine on 1994;117(5):60324.
the blood flow of the ophthalmic artery and the finger 79. Hayreh SS, Podhajsky P, Zimmerman MB. Role of noctur-
circulation. Graefes Arch Clin Exp Ophthalmol. 1993;231 nal arterial hypotension in optic nerve head ischemic disor-
(7):3714. ders. Ophthalmologica. 1999;213(2):7696.
60. Rojanapongpun P, Drance SM. The response of blood flow 80. Cantor LB, Harris A, Wolf S, Sponsel WE, Arend O.
velocity in the ophthalmic artery and blood flow of the finger Measurement of superficial optic nerve head capillary blood
to warm and cold stimuli in glaucomatous patients. Graefes velocities by scanning laser fluorescein angiography. J
Arch Clin Exp Ophthalmol. 1993;231(7):3757. Glaucoma. 1994;3 Suppl 1:S614.
61. Langham ME, Farrell RA, OBrien V, Silver DM, Schilder 81. Prnte C, Flammer J, Markstein R, Rudin M. Quantification
P. Blood flow in the human eye. Acta Ophthalmol Suppl. of optic nerve blood flow changes using magnetic resonance
1989;191:913. imaging. Invest Ophthalmol Vis Sci. 1995;36(1):24751.
62. Trew DR, Smith SE. Postural studies in pulsatile ocular 82. Garcia GH, Donahue KM, Ulmer JL, Harris GJ. Qualitative
blood flow: I Ocular hypertension and normotension. Br J perfusion imaging of the human optic nerve. Ophthal Plast
Ophthalmol. 1991;75(2):6670. Reconstr Surg. 2002;18(2):10713.
63. Trew DR, Smith SE. Postural studies in pulsatile ocular 83. Bill A. Intraocular pressure and blood flow through the uvea.
blood flow: II. Chronic open angle glaucoma. Br J Oph Arch Ophthalmol. 1962;67:33648.
thalmol. 1991;75(2):715. 84. Ernest JT, Potts AM. Pathophysiology of the distal portion
64. James CB, Smith SE. Pulsatile ocular blood flow in patients of the optic nerve. Am J Ophthalmol. 1971;72(2):43344.
with low tension glaucoma. Br J Ophthalmol. 1991;75(8): 85. Ernest JT. Autoregulation of optic-disk oxygen tension.
46670. Invest Ophthalmol. 1974;13(2):1016.
65. James CB, Smith SE. The effect of posture on the intraocular 86. Aukland K, Bower BF, Berliner RW. Measurement of local
pressure and pulsatile ocular blood flow in patients with blood flow with hydrogen gas. Circ Res. 1964;14:16487.
non-arteritic anterior ischaemic optic neuropathy. Eye. 87. Ernest JT. Optic disc blood flow. Trans Ophthalmol Soc UK.
1991;5(pt 3):30914. 1976;96(3):34851.
66. Trew DR. Blood flow in ocular hypertension and glaucoma. 88. Sokoloff L, Reivich M, Kennedy C, Des Rosiers MH, Patlak
In: Drance SM, editor. International symposium on glau- CS, Pettigrew KD, etal. The 14C-deoxyglucose method for
coma, ocular blood flow, and drug treatment. Baltimore: the measurement of local cerebral glucose utilization: the-
Williams & Wilkins; 1992. p. 11729. ory, procedure, and normal values in the conscious and anes-
67. Claridge KG. The effect of topical pilocarpine on pulsatile thetized albino rat. J Neurochem. 1977;28(5):897916.
ocular blood flow. Eye. 1993;7(pt 4):50710. 89. Bill A. Some aspects of the ocular circulation. Friedenwald
68. Krakau CE. Calculation of the pulsatile ocular blood flow. lecture. Invest Ophthalmol Vis Sci. 1985;26(4):41024.
Invest Ophthalmol Vis Sci. 1992;33(9):27546. 90. Sperber GO, Bill A. Blood flow and glucose consumption in
69. Williamson TH, Harris A. Ocular blood flow measurement. the optic nerve, retina and brain: effects of high intraocular
Br J Ophthalmol. 1994;78(12):93945. pressure. Exp Eye Res. 1985;41(5):63953.
70. Gee W, Oller DW, Homer LD, Bailey RC. Simultaneous 91. Bill A, Sperber GO. Blood flow and glucose consumption in
bilateral determination of the systolic pressure of the oph- the optic nerve: effects of high intraocular pressure. In:
thalmic arteries by ocular pneumoplethysmography. Invest Krieglstein GK, editor. Glaucoma update III. Heidelberg:
Ophthalmol Vis Sci. 1977;16(1):869. Springer; 1987. p. 517.
71. Hayreh SS, Walker WM. Fluorescent fundus photography in 92. Hayreh SS, Bill A, Sperber GO. Effects of high intraocular
glaucoma. Am J Ophthalmol. 1967;63(5):9829. pressure on the glucose metabolism in the retina and optic
72. Fishbein SL, Schwartz B. Optic disc in glaucoma: topogra- nerve in old atherosclerotic monkeys. Graefes Arch Clin
phy and extent of fluorescein filling defects. Arch Exp Ophthalmol. 1994;232(12):74552.
Ophthalmol. 1977;95(11):19759. 93. Sossi N, Anderson DR. Effect of elevated intraocular
73. Schwartz B, Rieser JC, Fishbein SL. Fluorescein angio- pressure on blood flow. Occurrence in cat optic nerve head
graphic defects of the optic disc in glaucoma. Arch studied with iodoantipyrine I 125. Arch Ophthalmol. 1983;
Ophthalmol. 1977;95(11):196174. 101(1):98101.
74. Talusan E, Schwartz B. Specificity of fluorescein angio- 94. Weinstein JM, Duckrow RB, Beard D, Brennan RW.
graphic defects of the optic disc in glaucoma. Arch Regional optic nerve blood flow and its autoregulation.
Ophthalmol. 1977;95(12):216675. Invest Ophthalmol Vis Sci. 1983;24(12):155965.
75. Sonty S, Schwartz B. Two-point fluorophotometry in the 95. Quigley HA, Hohman RM, Sanchez R, Addicks EM. Optic
evaluation of glaucomatous optic disc. Arch Ophthalmol. nerve head blood flow in chronic experimental glaucoma.
1980;98(8):14226. Arch Ophthalmol. 1985;103(7):95662.
76. Nagin P, Schwartz B, Reynolds G. Measurement of fluores- 96. Caprioli J, Miller JM. Measurement of optic nerve blood
cein angiograms of the optic disc and retina using computer- flow with iodoantipyrine: limitations caused by diffusion
ized image analysis. Ophthalmology. 1985;92(4):54752. from the choroid. Exp Eye Res. 1988;47(4):64152.
77. Nanba K, Schwartz B. Nerve fiber layer and optic disc fluo- 97. Hayreh SS. Factors determining the glaucomatous optic
rescein defects in glaucoma and ocular hypertension. nerve head damage. In: Krieglstein GK, editor. Glaucoma
Ophthalmology. 1988;95(9):122733. update III. Heidelberg: Springer; 1987. p. 406.
References 127
98. Alm A, Bill A. The oxygen supply to the retina. II Effects 105. Minckler DS, Bunt AH, Johanson GW. Orthograde and ret-
of high intraocular pressure and of increased arterial carbon rograde axoplasmic transport during acute ocular hyperten-
dioxide tension on uveal and retinal blood flow in cats. A sion in the monkey. Invest Ophthalmol Vis Sci.
study with radioactively labelled microspheres including 1977;16(5):42641.
flow determinations in brain and some other tissues. Acta 106. Hayreh SS, Podhajsky PA, Zimmerman B. Non-arteritic
Physiol Scand. 1972;84(3):30619. anterior ischemic optic neuropathy: time of onset of visual
99. Alm A, Bill A. Ocular and optic nerve blood flow at normal loss. Am J Ophthalmol. 1997;124(5):6417.
and increased intraocular pressures in monkeys (Macaca 107. Hayreh SS. The role of age and cardiovascular disease in
irus): a study with radioactively labelled microspheres glaucomatous optic neuropathy. Surv Ophthalmol. 1999;43
including flow determinations in brain and some other tis- Suppl 1:S2742.
sues. Exp Eye Res. 1973;15(1):1529. 108. Kaiser HJ, Flammer J, Graf T, Stmpfig D. Systemic blood
100. Geijer C, Bill A. Effects of raised intraocular pressure on pressure in glaucoma patients. Graefes Arch Clin Exp
retinal, prelaminar, laminar, and retrolaminar optic nerve Ophthalmol. 1993;231(12):67780.
blood flow in monkeys. Invest Ophthalmol Vis Sci. 109. Bchetoille A, Bresson-Dumont H. Diurnal and nocturnal
1979;18(10):103042. blood pressure drops in patients with focal ischemic glau-
101. Orgl S, Cioffi GA, Bacon DR, Bhandari A, Van Buskirk coma. Graefes Arch Clin Exp Ophthalmol. 1994;232(11):
EM. Measurement of optic nerve blood flow with nonra- 6759.
dioactive colored microspheres in rabbits. Microvasc Res. 110. Graham SL, Drance SM, Wijsman K, Douglas GR,
1996;51(2):17586. Mikelberg FS. Ambulatory blood pressure monitoring in
102. Buckberg GD, Luck JC, Payne DB, Hoffman JI, Archie JP, glaucoma. The nocturnal dip. Ophthalmology. 1995;102(1):
Fixler DE. Some sources of error in measuring regional 619.
blood flow with radioactive microspheres. J Appl Physiol. 111. Meyer JH, Brandi-Dohrn J, Funk J. Twenty four hour blood
1971;31(4):598604. pressure monitoring in normal tension glaucoma. Br J
103. Wang L, Fortune B, Cull G, McElwain KM, Cioffi GA. Ophthalmol. 1996;80(10):8647.
Microspheres method for ocular blood flow measurement
in rats: size and dose optimization. Exp Eye Res.
2007;84(1):10817.
104. Quigley HA, Addicks EM, Green WR, Maumenee AE.
Optic nerve damage in human glaucoma II. The site of For a complete bibliography and detailed review of previously
injury and susceptibility to damage. Arch Ophthalmol. published studies in the literature on the subject, please refer to
1981;99(4):63549. bibliography in previous publications listed here.

Part
II
Clinical Sciences

Optic Disc
7
Jost B. Jonas and Sohan Singh Hayreh
Definition of Optic Disc and Terminology (Fig.7.2), and (3) with a small optic disc (Fig.7.3). In a
normal Caucasian population, the optic disc area varies by
a factor of about 17 and the disc diameter by about 12.5
In the literature, the ophthalmoscopic term optic disc
[14]. The optic disc area is independent of age beyond an
has been applied interchangeably to either the whole or
age of about 310years. In regard to gender, body length
a part of the most anterior part of the optic nerve head
and refractive error, the majority of recent studies suggest
(i.e., the surface nerve fiber layer and the prelaminar
that the disc size is correlated with body height. Within a
region). Similarly the term papilla has been used as a
range of 5 to +5 diopters of refractive error, optic disc size
synonym for the optic disc or optic nerve head. The term
is mostly independent of ametropia [2, 4]. The optic disc is
papilla was coined by Briggs in 1676, based on an
significantly larger in highly myopic eyes, and signifi-
erroneous impression that the normal optic nerve head
cantly smaller in eyes with marked hyperopia (more than
was elevated like a papilla (see Chap. 2). Since the struc-
+5 diopters), than ineyes with a normal refractive error.
ture usually is not elevated above the level of the adja-
The size of the optic disc also depends on ethnic back-
cent retina but lies flat in the same plane as the retina and
ground [46]; Caucasians have relatively small optic discs,
has a central depression (i.e., the physiological cup), the
followed by Mexicans, Asians, and Afro-Americans.
term papilla is a misnomer. It must, however, be con-
Based on the Gaussian-like distribution curve of the optic
ceded that the appearance of the optic nerve head shows
disc area, very small discs or microdiscs and very large
considerable variations, all of which may be within
discs or macrodiscs can be defined morphometrically: a
physiological limits. In reviews of work by other
microdisc as being smaller than the mean minus twofold
authors, it is not clear how much of the optic nerve head
standard deviations, and a macrodisc as being larger than
has been included by them under the term optic disc.
the mean plus twofold standard deviations.
This interindividual variability in optic disc size is
Optic Disc Size morphologically important, because eyes with large
optic discs have a larger neuroretinal rim area than
The size of the optic disc varies among individuals. For eyes with small discs [14]; more optic nerve fibers [7,
example, there are normal eyes: (1) with a normal sized 8]; less nerve fiber crowding [8]; a higher count and a
and normal-shaped optic disc (Fig.7.1a, b), (2) with a large larger total area of lamina cribrosa pores [9]; a higher
optic disc and cup without any pathologic abnormality ratio of interpore connective tissue area to total lamina
cribrosa area [9]; a higher count of cilioretinal arteries
[10]; and a higher count of retinal photoreceptors [11]
J.B. Jonas
and retinal pigment epithelium cells [12] as well as a
Department of Ophthalmology and Eye Hospital, larger retinal surface area and longer horizontal and
The Medical Faculty Mannheim of the Ruprecht-Karls-University vertical diameters of the globe [13].
Heidelberg, 68167 Mannheim, Germany The optic disc size variability is also pathogeneti-
S.S. Hayreh () cally important because some optic nerve anomalies
Department of Ophthalmology and Visual Science,
University of Iowa Hospital and Clinics, College of Medicine,
and diseases are correlated with the optic disc size. For
Hawkins Dr. 200, Iowa City, IA 522421091, USA example, optic disc drusen [14, 15] (Fig. 7.4a),
e-mail: sohan-hayreh@uiowa.edu pseudopapilledema [16, 17] (Fig.7.4b) and nonarteritic
132 7 Optic Disc
a b
Fig.7.1 Photographs of two (a, b) examples of normal sized optic discs with a normal shape: the vertical disc axis is slightly longer
than the horizontal axis; Note: normal shape of the neuroretinal rim according to the ISNT rule
Fig.7.2 Photograph of a large physiologic cup in the large disc;

normal shape of the neuroretinal rim with its widest part in the
inferior disc region, followed by the superior disc region, the
nasal region, and finally the temporal disc region; unremarkable
parapapillary atrophy (physiologic alpha zone) Fig.7.3 Photograph of a small optic disc, with no cup
Optic Disc Shape 133
a b
Fig.7.4 Photographs of examples of (a) optic disc with disc drusen, (b) pseudopapilledema, and (c) fellow optic disc of a patient
with NA-AION
anterior ischemic optic neuropathy (NA-AION) [1824] than the horizontal one [2] (Fig. 7.1a, b). The ratio
(Fig.7.4c see Chap. 16) occur significantlymore often between minimal-to-maximal disc diameter ranges
in small optic discs/cups. Congenital pits of the optic between 0.64 and 0.98, corresponding to an interindi-
disc [25] (Fig. 7.5a, b) and the Morning-Glory- vidual variability of 1:1.53. The ratio between the hori-
Syndrome [26] (Fig.7.6a, b) are more common in large zontal to vertical disc diameters varies between 0.70
optic nerve heads. Hypoplastic optic nerves have small and 1.37 (variability of 1:1.96).
optic discs (Fig.7.7ad). The disc form is not correlated with age, gender,
right and left eye, or body weight and height [2]. An
abnormal optic disc shape is significantly correlated
Optic Disc Shape with increased corneal astigmatism and amblyopia [27],
and the amount of corneal astigmatism is significantly
The optic disc has a slightly vertically oval form with (p<0.001) correlated with an increasingly elongated
the vertical diameter being about 7% to 10% larger optic disc shape. Corneal astigmatism is significantly
134 7 Optic Disc
a1 a
a2
Fig. 7.6 Photographs of two (a, b) examples of eyes with

Morning-Glory-Syndrome
(p<0.01) highest in eyes with tilted discs (Fig.7.8ad).

Amblyopia is significantly (p<0.05) associated with an
elongated optic disc shape and high corneal astigma-
tism. The orientation of the longest disc diameter can
indicate the axis of corneal astigmatism. Furthermore,
Fig.7.5 Photographs of two eyes as examples of optic disc pit:

In one eyes, fundus photograph (a1) showing the pit (arrow) and
fluorescein angiogram (a2) showing late staining of the pit
(arrow). In the second eye, fundus photograph (b) showing the
pit (arrow) and associated serous retinal detachment of the mac-
ular region
Neuroretinal Rim Size 135
a b
c d
Fig.7.7 Photographs of four (AD) eyes as examples of hypoplastic optic nerves
eyes with a tilted optic disc can exhibit visual field interindividual variability. The rim area is correlated
defects mimicking a temporal hemianopsia. with the optic disc area [14]. The increase of rim area
The optic disc shape in eyes with NA-AION does with enlarging disc area is most marked for eyes with
not differ significantly from that of normal eyes. no disc cupping, medium pronounced for eyes with a
temporal flat sloping of the optic cup, and it is least
marked in eyes with a circular steep disc cupping [2].
The correlation between rim area and disc area corre-
Neuroretinal Rim Size sponds with the positive correlation between optic disc
size, optic nerve fiber count [7, 8], and number and
The neuroretinal rim contains the retinal nerve fibers total area of the lamina cribrosa pores [9]. It points
and it is, therefore, one of the main targets in the oph- towards a greater anatomic reserve capacity in eyes
thalmoscopic evaluation of the optic disc. Like the with large optic discs as compared to eyes with small
optic disc and cup, the neuroretinal rim shows a high optic discs.
136 7 Optic Disc
a b
c d
Fig.7.8 Photographs of four (AD) eyes as examples of tilted optic disc: Note: small optic disc, the longest disc axis is obliquely
orientated; inferior scleral crescent, except in Fig. D
Due to the association between NA-AION and (Figs.7.1a, b and 7.2). The assessment of the charac-
absent or small disc size, and the association between teristic shape of the rim is helpful in the diagnosis of
disc size and rim size, eyes with NA-AION have a early glaucomatous optic nerve damage. The physio-
small neuroretinal rim. logic shape of the neuroretinal rim is associated with:
1. The diameter of the retinal arterioles which are sig-
nificantly wider in the inferotemporal arcade than
Neuroretinal Rim Shape in the superotemporal arcade [28, 29];
2. The visibility of the retinal nerve fiber bundles; they
In normal eyes, the neuroretinal rim shows a charac- are significantly more often detected more easily in
teristic configuration. It is based on the vertically oval the inferotemporal region than in the superotempo-
shape of the optic disc and the horizontally oval shape ral region [29, 30];
of the optic cup. The neuroretinal rim usually follows 3. The location of the foveola 0.530.34mm inferior
the ISNT rule: broadest in the Inferior disc region, to the optic disc center [30];
followed by the Superior disc region, the Nasal 4. The morphology of the lamina cribrosa with the
disc area, and finally the Temporal disc region [2] largest pores and relatively the smallest amount of
Configuration and Depth of the Optic Disc Cup 137
interpore connective tissue in the inferior and supe- into account, however, that the distribution of the
rior regions, as compared to the temporal and nasal optic cup area is not Gaussian. The macrocups can be
sectors [9, 31, 32]; and subdivided into:
5. The distribution of the thin and thick nerve fibers in
1. Primary macrocups: These occur in primary mac-
the optic nerve just behind the globe with the thin
rodiscs [37]. They appear pseudoglaucomatous but
fibers in the temporal part of the nerve [8, 33].
are physiologic. They are constant in size after the
In those eyes with NA-AION which have a cup, the first years of life.
neuroretinal rim follows the ISNT rule (see above), 2. Acquired or secondary macrocups: These can
both before and after the event of NA-AION. The rea- further be sub classified into two types. (a) The
son is that, in contrast to glaucomatous optic neuropa- secondary, highly myopic macrocup is found in
thy, non-glaucomatous optic nerve damage (except of highly myopic eyes with secondary macrodiscs
giant cell arteritis induced AION) is not associated due to the myopic stretching of the optic nerve
with a loss of neuroretinal rim. Consequently, the shape head [38], (b) The other type of secondary macro-
of the neuroretinal rim does not markedly change. cups develops by the glaucomatous loss of neu-
roretinal rim.
In normal eyes, the areas of the optic disc and optic
Neuroretinal Rim Pallor cup are correlated with each other. In small optic discs,
cupping normally does not occur. Large optic discs
usually have a large optic cup. In the morphologic
Increasing pallor of the optic disc and especially of the diagnosis of glaucoma, this feature has to be taken into
neuroretinal rim is a typical sign of optic nerve damage account.
[3436]. The increase in pallor of the neuroretinal rim In contrast to glaucoma, the optic cup does not
is more marked in eyes with non-glaucomatous optic markedly enlarge in eyes with non-glaucomatous optic
neuropathy than in eyes with glaucoma. In other words, nerve damage such as NA-AION [39, 40]. Corre
if the neuroretinal rim looks rather pale, the disease is spondingly, the neuroretinal rim does not decrease
more probably non-glaucomatous optic neuropathy markedly in eyes with non-glaucomatous optic nerve
rather than glaucomatous optic neuropathy. Pallor of damage, except in arteritic AION where, on resolution
the neuroretinal rim is thus one among several vari- of optic disc edema in 68weeks, the optic disc devel-
ables which differentiate glaucomatous from non- ops optic atrophy with cupping indistinguishable from
glaucomatous optic neuropathy. This is very helpful to glaucomatous optic neuropathy, with a decrease in
differentiate optic disc cupping seen in glaucoma ver- neuroretinal rim [18, 4146].
sus that in arteritic AION (see Chap. 12). In glaucoma,
the overall pallor of the optic disc as the sum of the
optic cup and neuroretinal rim increases, mainly due to
the enlargement of the optic cup. onfiguration and Depth
C
In eyes with NA-AION, the neuroretinal rim of the Optic Disc Cup
becomes pale, usually in the affected optic disc sector
but not necessarily in that region only, due to the non- In normal eyes, the shape of the optic disc cup is hori-
glaucomatous type of optic nerve damage. zontally oval with the horizontal diameter being about
8% longer than the vertical diameter [2]. Combining
the horizontally oval shape of the cup and the verti-
cally oval shape of the optic disc explains the configu-
ptic Cup Size in Relation
O
ration of the normal neuroretinal rim, which has its
to the Optic Disc Size broadest parts in the inferior and superior disc regions
and its smallest parts in the temporal and nasal region
Like the optic disc and the neuroretinal rim, the optic of the optic disc [2].
cup also shows a high interindividual variability [1 The cup is ophthalmoscopically described by its
4] (Figs.7.17.4). Like macrodiscs, very large cups depth. In normal eyes, the cup depth depends on the
or macrocups may be defined as larger than the mean cup area, and indirectly on the disc size: the larger the
plus twofold standard deviations. One has to take cup, the deeper it is [47, 48].
138 7 Optic Disc
Eyes with NA-AION have no cup or a small cup Parapapillary Chorioretinal Atrophy
(Fig.7.4c, see Chap. 16). In those eyes with a cup, it usually
has a horizontally oval shape, so that the ISNT rule Ophthalmoscopically, parapapillary chorioretinal atro-
of the rim shape is fulfilled. Since the cup is small phy has been divided into a central beta zone and a
(if present) in eyes with NA-AION, it is usually shallow. peripheral alpha zone [49, 50] (Fig.7.9a, b). A periph-
eral zone (alpha zone) is characterized by an irregular
hypopigmentation and hyperpigmentation. On its
peripheral side it is adjacent to the normal appearing
Cup/Disc Ratio retina, and on its central side it is in touch with a zone
characterized by visible sclera and visible large chor-
Due to the vertically oval optic disc and the horizon- oidal vessels (beta zone), or with the peripapillary
tally oval optic cup, the cup/disc diameter ratio in scleral ring, respectively. Features of the inner zone
normal eyes is significantly larger horizontally than (beta zone) are marked atrophy of the retinal pigment
vertically [2]. In less than 7% of normal eyes the hori- epithelium and of the choriocapillaris, good visibility
zontal cup/disc ratio is smaller than the vertical one. of the large choroidal vessels and the sclera, thinning
That indicates that the quotient of the horizontal-to- of the chorioretinal tissues, with adjacent alpha zone
vertical-cup/disc ratios is usually higher than 1.0. This on its peripheral side and the peripapillary scleral ring
is important for the diagnosis of glaucoma, in which, on its central side. If both zones are present, the beta
in the early to medium advanced stages, the vertical zone is always closer to the optic disc than the alpha
cup/disc diameter ratio increases faster than the hori- zone. On indirect and direct clinical-histological com-
zontal one. It leads to an increase of the quotient of parisons, the beta zone correlates with a complete loss
horizontal-to-vertical-cup/disc ratios to values lower of retinal pigment epithelium cells and a markedly
than 1.0. diminished count of retinal photoreceptors [51]. The
Since the cup diameter depends on the disc diame- alpha zone is the equivalent of pigmentary irregulari-
ter, the cup/disc ratio also depends on the size of the ties in the retinal pigment epithelium. Correspondingly,
optic disc. The high interindividual variability of the the beta zone corresponds to an absolute scotoma, and
optic disc and cup diameters explains why the cup/disc the alpha zone to a relative scotoma [52]. In normal
ratios range in a normal population is between 0.0 and eyes, both alpha and beta zones are largest and most
almost 0.9 [2]. Due to the correlation between disc frequently located in the temporal horizontal sector,
area and cup area, the cup/disc ratios are low in small followed by the inferior temporal area and the superior
optic discs, and high in large ones. An unusually high temporal region. They are smallest and rarely found in
cup/disc ratio, therefore, can be physiologic in eyes the nasal parapapillary area. An alpha zone is present
with large optic nerve heads [37], while an average in almost all normal eyes and is thus more common
cup/disc ratio is uncommon in normal eyes with small than a beta zone (mean frequency in normal eyes:
optic discs. In the diagnosis of glaucomatous optic about 1520%). The two zones have to be differenti-
nerve damage, this interindividual variability of cup/ ated from the myopic scleral crescent in eyes with high
disc ratios and their dependence on the optic disc size myopia and from the inferior scleral crescent in eyes
has to be taken into account. with tilted optic discs (Fig.7.8). The myopic scleral
As both the cup diameter measure and the disc diam- crescent present in highly myopic eyes differs histo-
eter measure depend to an equal extent on the magnifi- logically from the glaucomatous beta zone in non-
cation by the optic media, the cup/disc ratios are highly myopic eyes. In the region of the myopic
independent of the magnification by the optical media crescent, only the inner limiting membrane and under-
of the examined eye and of the fundus camera or other lying retinal nerve fiber layer or its remnants cover the
instrument. Thus no method to correct for the ocular sclera while in the glaucomatous beta zone, Bruchs
and camera magnification need be applied. The quotient membrane and the choroid are interposed between the
of the horizontal-to-vertical-cup/disc ratios is addition- remnants of the retina and the sclera [53].
ally independent of the size of the disc, since both the Eyes with NA-AION, in contrast to glaucoma, do
horizontal cup/disc ratio and the vertical cup/disc ratio not show an enlargement of the beta zone nor of the
depend to an equal extent on the disc diameter. alpha zone of parapapillary atrophy [54, 55].
Retinal Nerve Fiber Layer 139
a b
Fig.7.9 Photographs of two examples (a, b) of glaucomatous optic nerve heads with alpha zone (arrow heads) and beta zone
(arrows) of parapapillary atrophy
Diameter of Retinal Arterioles
Diffuse narrowing of the retinal vessels has been

described in glaucomatous optic neuropathy and in
nonglaucomatous optic neuropathies such as descend-
ing optic nerve atrophy [48, 5456]. In glaucoma, the
vessel diameter decreases with decreasing area of the
neuroretinal rim, diminishing visibility of the retinal
nerve fiber layer and increasing visual field defects
[56]. Correspondingly, a recent study on healthy chil-
dren enrolled in the Singapore Cohort Study of the
Risk Factors for Myopia showed that a thinner neu-
roretinal rim and lower measurements of the retinal
nerve fiber layer were associated with narrower retinal
arterioles and venules [57].
Fig.7.10 Photograph of the retinal nerve fiber layer; Note: nerve
fiber layer best detectable in the temporal inferior sector, followed
by the temporal superior sector and the nasal superior sector
Retinal Nerve Fiber Layer
f ollowed by the temporal superior area, the nasal supe-
In normal eyes, visibility of the retinal nerve fiber layer rior region and finally the nasal inferior sector [48]. It
is regionally unevenly distributed (Fig.7.10). Dividing is least visible in the superior, inferior, temporal hori-
the fundus into eight regions, the nerve fiber bundles zontal and nasal horizontal regions. Correspondingly,
are most visible in the temporal inferior sector, the diameters of the retinal arterioles are significantly
140 7 Optic Disc
widest at the temporal inferior disc border, followed by

the temporal superior disc region, the nasal superior
area and finally the nasal inferior disc region. This is in
agreement with the location of the foveola below a
horizontal line drawn through the center of the optic
disc, and with a configuration of the neuroretinal rim
that is broadest at the temporal inferior disc border,
followed by the temporal superior disc region. The
sectors sequence concerning the best visibility of the
retinal nerve fiber layer correlates with the sectors
sequence in respect to rim configuration and retinal
artery caliber. Physiologically, it points towards an
anatomical and nutritional relationship. The visibility
of the retinal nerve fiber layer decreases with age. It
Fig.7.11 Photograph of an eye of a rhesus monkey with malig-
correlates with an age-related reduction of the optic
nant arterial hypertension and multiple cottonwool spots
nerve fiber count with an annual loss of about 4,000 (arrows); note extensive nerve fiber bundle loss (dark arcuate
5,000 fibers/year or about 0.3% out of an original pop- bands) in the regions of the cottonwool spots (Reproduced
ulation of presumably 1.4 million optic nerve fibers from Hayreh etal. [60])
[33], similar to the age-related loss of about 0.3% of
retinal photoreceptors and retinal pigment epithelium
cells [58]. These features of the normal retinal nerve detect than a localized defect. It is helpful to use the
fiber layer are important for diagnosis of retinal nerve variable sequence of fundus sectors concerning the
fiber layer changes secondary to optic nerve damage in best retinal nerve fiber layer visibility. If one detects
the diseased eye. that in an eye without fundus irregularities the retinal
Localized defects of the retinal nerve fiber layer nerve fiber layer is markedly better detectable in the
are defined as wedge-shaped and not spindle-like temporal superior fundus region than in the temporal
defects, running towards or touching the optic disc inferior sector, that points towards a loss of retinal
border [59]. If they are pronounced, they can have a nerve fiber layer located mainly in the temporal infe-
broad basis at the temporal raphe of the fundus. rior fundus region. This variable can be examined upon
Typically occurring in about 20% of all glaucoma ophthalmoscopy without applying sophisticated tech-
eyes, they can also be found in eyes with an atrophy niques. It is also helpful to evaluate whether the retinal
of the optic nerve due to other reasons such as optic vessels are clearly and sharply detectable. The retinal
disc drusen, toxoplasmic retinochoroidal scars, and vessels are normally embedded in the retinal nerve
ischemic retinopathies with cottonwool spots of the fiber layer. In eyes with a diffuse retinal nerve fiber
retina (Fig.7.11), after longstanding optic disc edema layer loss, the retinal vessels are covered only by the
or optic neuritis due to multiple sclerosis, to mention inner limiting membrane, resulting in a better visibility
some examples. Since the localized retinal nerve fiber and a sharper image of the large retinal vessels. This is
layer defects are not present in normal eyes, they an important variable in the diagnosis of optic nerve
almost always signify a pathological abnormality. damage.
The nerve fiber bundle defect can best be seen with The clinical introduction of the new generation of
red free ophthalmoscopy. optical coherence tomography is going to substan-
Besides localized retinal nerve fiber layer defects, a tially improve the imaging and quantification of the
diffuse loss of retinal nerve fibers occurs in eyes with retinal nerve layer in health and disease. It is of par-
damage to the optic nerve. It leads to decreased visibil- ticular importance for optic nerve diseases such as
ity of the retinal nerve fiber layer. After sectioning of glaucoma and NA-AION [6265]. Evaluation of the
the optic nerve in the orbit of monkeys, Quigley [61] retinal nerve fiber layer is very useful in eyes with
observed a disappearing of the visibility of the retinal nonglaucomatous optic nerve damage, such as eyes
nerve fiber layer, starting 1month after the operation, with NA-AION. The combination of (1) decreased
and being complete 4weeks later. Ophthalmoscopically, visibility of the retinal nerve fiber layer, (2) increased
diffuse retinal nerve fiber layer loss is more difficult to pallor of the neuroretinal rim, and (3) unremarkable
References 141
Fig.7.12 Photomicrograph
of the ONH and retrolaminar
optic nerve of a left eye with
18-day-old A-AION,
showing a sharply defined
area of necrosis in the
anterior part of the optic
nerve and a partially
disintegrated lamina cribrosa.
There is presence of
hyaluronidase-sensitive
material (blue) similar to that
found in cavernous
degeneration of the optic
nerve (Alcian blue stain)
(Reproduced from Hinzpeter
and Naumann [70])
size and shape of the optic disc, neuroretinal rim and the prelaminar, laminar and retrolaminar regions of the
parapapillary atrophy characterizes the nonglaucoma- optic nerve, along with a massive presence of acid
tous optic nerve atrophy in eyes with NA-AION, and mucopolysaccharides in the retrolaminar optic nerve
distinguishes it from the glaucomatous type of optic tissue (Fig. 7.12). This finding was interpreted as an
nerve damage. intrusion of vitreous material, resulting from breaks in
the internal limiting membrane and the pressure gradi-
ent from intraocular to extraocular tissues, an analogy
Histological Findings to Schnabel degeneration in acute glaucoma [70]. The
histology of the alpha and beta zones of parapapillary
atrophy has been described above.
Histological examinations have shown that, in agree-
ment with the ophthalmoscopic appearance, the retinal
nerve fiver layer is thickest at the inferior optic disc
pole, followed by the superior disc pole, the nasal disc Conclusion
margin and finally the temporal disc margin [66, 67]. It
agrees with the physiological shape of the neuroretinal This shows that a comprehensive understanding of
rim following the ISNT rule [2]. (Figs.7.1a, b, and various aspects of normal optic disc morphology,
7.2) In a similar manner, in agreement with intravital its physiological variations, and clinical evaluation
measurements, histomorphometry has shown that the on ophthalmoscopy and other tests provide a valu-
optic disc is larger in axially highly myopic eyes than able basis of knowledge of various types of optic
in normal sized eyes [53]. These studies additionally neuropathies.
revealed that the lamina cribrosa and the sclera are
thinner in highly myopic eyes, probably due to the
myopic stretching of the posterior fundus of the eye
References
[68, 69]. Interestingly, our recent study on the lamina
cribrosa in monkeys with optic nerve damage after an
1. Bengtsson B. The variation and covariation of cup and
experimental temporary central retinal artery occlu- disc diameters. Acta Ophthalmol (Copenh). 1976;54(6):
sion showed that the lamina cribrosa was thicker in 80418.
these eyes than in monkey eyes with advanced glau- 2. Jonas JB, Gusek GC, Naumann GO. Optic disc, cup and neu-
comatous optic nerve damage. There has been one roretinal rim size, configuration and correlations in normal
eyes. Invest Ophthalmol Vis Sci. 1988;29(7):11518.
report on the histology of an arteritic ischemic optic 3. Ramrattan RS, Wolfs RC, Jonas JB, Hofman A, de Jong PT.
neuropathy. The patient died 18days after the onset of Determinants of optic disc characteristics in a general population:
the disease. Histology revealed an ischemic necrosis of the Rotterdam study. Ophthalmology. 1999;106(8):158896.
142 7 Optic Disc
4. Varma R, Tielsch JM, Quigley HA, Hilton SC, Katz J, Spaeth 24. Jonas JB, Gusek GC, Naumann GO. Anterior ischemic optic
GL, etal. Race-, age-, gender-, and refractive error-related neuropathy: nonarteritic form in small and giant cell arteritis
differences in the normal optic disc. Arch Ophthalmol. in normal sized optic discs. Int Ophthalmol. 1988;12(2):
1994;112(8):106876. 11925.
5. Chi T, Ritch R, Stickler D, Pitman B, Tsai C, Hsieh FY. 25. Jonas JB, Naumann GO. Papillengruben in groen papillae
Racial differences in optic nerve head parameters. Arch nervi optici. Papillometrische Charakteristika in 15 Augen.
Ophthalmol. 1989;107(6):8369. [Pits of the optic papilla in large optic nerve papillae.
6. Tsai CS, Zangwill L, Gonzalez C, Irak I, Garden V, Hoffman R, Papillometric characteristics in 15 eyes]. Klin Monbl
et al. Ethnic differences in optic nerve head topography. Augenheilkd. 1987;191(4):28791.
J Glaucoma. 1995;4(4):24857. 26. Jonas JB, Koniszewski G, Naumann GO. Morning-Glory-
7. Quigley HA, Coleman AL, Dorman-Pease ME. Larger optic Syndrom bzw. Handmannsche Anomalie in kongeni-
nerve heads have more nerve fibers in normal monkey eyes. talen Makropapillen. Extremvariante konfluierender
Arch Ophthalmol. 1991;109(10):14413. Papillengruben? [Morning glory syndrome and
8. Jonas JB, Schmidt AM, Muller-Bergh JA, Schlotzer- Handmanns anomaly in congenital macropapilla.
Schrehardt UM, Naumann GO. Human optic nerve fiber Extreme variants of confluent optic pits?]. Klin Monbl
count and optic disc size. Invest Ophthalmol Vis Sci. Augenheilkd. 1989;195(6):3714.
1992;33(6):20128. 27. Jonas JB, Kling F, Grundler AE. Optic disc shape, corneal
9. Jonas JB, Mardin CY, Schlotzer-Schrehardt U, Naumann astigmatism, and amblyopia. Ophthalmology. 1997;
GO. Morphometry of the human lamina cribrosa surface. 104(11):19347.
Invest Ophthalmol Vis Sci. 1991;32(2):4015. 28. Jonas JB, Nguyen XN, Naumann GO. Parapapillary retinal
10. Jonas JB, Guggenmoos-Holzmann I, Naumann GO. vessel diameter in normal and glaucoma eyes. I. Mor
Cilioretinal arteries in large optic disks. Ophthalmic Res. phometric data. Invest Ophthalmol Vis Sci. 1989;30(7):
1988;20(5):26974. 1599603.
11. Panda-Jonas S, Jonas JB, Jakobczyk M, Schneider U. Retinal 29. Jonas JB, Schiro D. Visibility of the normal retinal nerve
photoreceptor count, retinal surface area, and optic disc size in fiber layer correlated with rim width and vessel caliber.
normal human eyes. Ophthalmology. 1994;101(3):51923. Graefes Arch Clin Exp Ophthalmol. 1993;231(4):20711.
12. Panda-Jonas S, Jonas JB, Jakobczyk-Zmija M. Retinal 30. Jonas JB, Nguyen NX, Naumann GO. The retinal nerve fiber
pigment epithelial cell count, distribution, and correlations in layer in normal eyes. Ophthalmology. 1989;96(5):62732.
normal human eyes. Am J Ophthalmol. 1996;121(2):1819. 31. Ogden TE, Duggan J, Danley K, Wilcox M, Minckler DS.
13. Papastathopoulos KI, Jonas JB, Panda-Jonas S. Large optic Morphometry of nerve fiber bundle pores in the optic nerve
discs in large eyes, small optic discs in small eyes. Exp Eye head of the human. Exp Eye Res. 1988;46(4):55968.
Res. 1995;60(4):45961. 32. Radius RL. Regional specificity in anatomy at the lamina
14. Mullie MA, Sanders MD. Scleral canal size and optic nerve cribrosa. Arch Ophthalmol. 1981;99(3):47880.
head drusen. Am J Ophthalmol. 1985;99(3):3569. 33. Jonas JB, Muller-Bergh JA, Schlotzer-Schrehardt UM,
15. Jonas JB, Gusek GC, Guggenmoos-Holzmann I, Naumann Naumann GO. Histomorphometry of the human optic nerve.
GO. Optic nerve head drusen associated with abnormally Invest Ophthalmol Vis Sci. 1990;31(4):73644.
small optic discs. Int Ophthalmol. 1987;11(2):7982. 34. Hayreh SS. Colour and fluorescence of the optic disc.
16. Rosenberg MA, Savino PJ, Glaser JS. A clinical analysis of Ophthalmologica. 1972;165(2):1008.
pseudopapilledema. I. Population, laterality, acuity, refrac- 35. Hitchings RA. The optic disc in glaucoma, III: diffuse optic
tive error, ophthalmoscopic characteristics, and coincident disc pallor with raised intraocular pressure. Br J Ophthalmol.
disease. Arch Ophthalmol. 1979;97(1):6570. 1978;62(10):6705.
17. Jonas JB, Gusek GC, Guggenmoos-Holzmann I, Naumann 36. Miller JM, Caprioli J. Videographic quantification of optic
GO. Pseudopapilledema associated with abnormally small disc pallor. Invest Ophthalmol Vis Sci. 1988;29(2):3203.
optic discs. Acta Ophthalmol (Copenh). 1988;66(2):1903. 37. Jonas JB, Zach FM, Gusek GC, Naumann GO. Pseudoglau
18. Hayreh SS. Pathogenesis of cupping of the optic disc. comatous physiologic large cups. Am J Ophthalmol.
Br J Ophthalmol. 1974;58(10):86376. 1989;107(2):13744.
19. Beck RW, Savino PJ, Repka MX, Schatz NJ, Sergott RC. 38. Jonas JB, Gusek GC, Naumann GO. Optic disk morphome-
Optic disc structure in anterior ischemic optic neuropathy. try in high myopia. Graefes Arch Clin Exp Ophthalmol.
Ophthalmology. 1984;91(11):13347. 1988;226(6):58790.
20. Feit RH, Tomsak RL, Ellenberger Jr C. Structural factors 39. Jonas JB, Nguyen NX, Naumann GO. Optic disc morphom-
in the pathogenesis of ischemic optic neuropathy. Am etry in simple optic nerve atrophy. Acta Ophthalmol
JOphthalmol. 1984;98(1):1058. (Copenh). 1989;67(2):199203.
21. Doro S, Lessell S. Cup-disc ratio and ischemic optic neu- 40. Jonas JB, Xu L. Parapapillary chorioretinal atrophy in
ropathy. Arch Ophthalmol. 1985;103(8):11434. normal-pressure glaucoma. Am J Ophthalmol. 1993;115
22. Beck RW, Servais GE, Hayreh SS. Anterior ischemic optic (4):5015.
neuropathy. IX. Cup-to-disc ratio and its role in pathogene- 41. Hayreh SS. Anterior ischaemic optic neuropathy. II. Fundus
sis. Ophthalmology. 1987;94(11):15038. on ophthalmoscopy and fluorescein angiography. Br J
23. Hayreh SS, Zimmerman MB. Nonarteritic anterior ischemic Ophthalmol. 1974;58(12):96480.
optic neuropathy: refractive error and its relationship to cup/ 42. Hayreh SS. Anterior ischemic optic neuropathy. New York:
disc ratio. Ophthalmology. 2008;115(12):227581. Springer; 1975.
References 143
43. Hayreh SS. Ischemic optic neuropathy. Int Ophthalmol. 59. Jonas JB, Schiro D. Localized retinal nerve fiber layer
1978;1:918. defects in nonglaucomatous optic nerve atrophy. Graefes
44. Hayreh SS, Podhajsky PA, Zimmerman B. Ocular manifes- Arch Clin Exp Ophthalmol. 1994;232(12):75960.
tations of giant cell arteritis. Am J Ophthalmol. 60. Hayreh SS, Servais GE, Virdi PS. Cotton-wool spots (inner
1998;125(4):50920. retinal ischemic spots) in malignant arterial hypertension.
45. Danesh-Meyer HV, Savino PJ, Sergott RC. The prevalence Ophthalmologica. 1989;198:197215.
of cupping in end-stage arteritic and nonarteritic anterior 61. Quigley HA, Davis EB, Anderson DR. Descending optic
ischemic optic neuropathy. Ophthalmology. 2001;108(3): nerve degeneration in primates. Invest Ophthalmol Vis Sci.
5938. 1977;16(9):8419.
46. Danesh-Meyer H, Savino PJ, Spaeth GL, Gamble GD. 62. Danesh-Meyer HV, Boland MV, Savino PJ, Miller NR,
Comparison of arteritis and nonarteritic anterior ischemic Subramanian PS, Girkin CA, etal. Optic disc morphology in
optic neuropathies with the Heidelberg Retina Tomograph. open-angle glaucoma compared with anterior ischemic optic
Ophthalmology. 2005;112(6):110412. neuropathies. Invest Ophthalmol Vis Sci. 2010;51(4):
47. Fernandez MC, Jonas JB, Naumann GO. Parapapillare cho- 200310.
rioretinale Atrophie in Augen mit flacher glaukomatser 63. Sung KR, Wollstein G, Bilonick RA, Townsend KA,
Papillenexkavation. [Para-papillary chorioretinal atrophy in Ishikawa H, Kagemann L, et al. Effects of age on optical
eyes with shallow glaucomatous optic disk cupping]. coherence tomography measurements of healthy retinal
Fortschr Ophthalmol. 1990;87(5):45760. nerve fiber layer, macula, and optic nerve head.
48. Jonas JB, Budde WM, Panda-Jonas S. Ophthalmoscopic Ophthalmology. 2009;116(6):111924.
evaluation of the optic nerve head. Surv Ophthalmol. 64. Leung CK, Lam S, Weinreb RN, Liu S, Ye C, Liu L, etal.
1999;43(4):293320. Retinal nerve fiber layer imaging with spectral-domain opti-
49. Jonas JB, Fernandez MC, Naumann GO. Glaucomatous cal coherence tomography analysis of the retinal nerve fiber
parapapillary atrophy. Occurrence and correlations. Arch layer map for glaucoma detection. Ophthalmology.
Ophthalmol. 1992;110(2):21422. 2010;117(9):168491.
50. Jonas JB, Naumann GO. Parapapillary chorioretinal atrophy 65. Scott CJ, Kardon RH, Lee AG, Frisn L, Wall M. Diagnosis
in normal and glaucoma eyes. II. Correlations. Invest and grading of papilledema in patients with raised intracra-
Ophthalmol Vis Sci. 1989;30(5):91926. nial pressure using optical coherence tomography vs clinical
51. Jonas JB, Konigsreuther KA, Naumann GO. Optic disc his- expert assessment using a clinical staging scale. Arch
tomorphometry in normal eyes and eyes with secondary Ophthalmol. 2010;128(6):70511.
angle-closure glaucoma. II. Parapapillary region. Graefes 66. Varma R, Skaf M, Barron E. Retinal nerve fiber layer thick-
Arch Clin Exp Ophthalmol. 1992;230(2):1349. ness in normal human eyes. Ophthalmology. 1996;
52. Rensch F, Jonas JB. Microperimetric changes after intravit- 103(12):21149.
real bevacizumab for exudative age-related macular degen- 67. Dichtl A, Jonas JB, Naumann GO. Retinal nerve fiber layer
eration. Acta Ophthalmol. 2008;86(6):6967. thickness in human eyes. Graefes Arch Clin Exp Ophthalmol.
53. Dichtl A, Jonas JB, Naumann GO. Histomorphometry of the 1999;237(6):4749.
optic disc in highly myopic eyes with absolute secondary 68. Jonas JB, Berenshtein E, Holbach L. Lamina cribrosa thick-
angle closure glaucoma. Br J Ophthalmol. 1998;82(3):2869. ness and spatial relationships between intraocular space and
54. Jonas JB, Fernandez MC, Naumann GO. Parapapillary cerebrospinal fluid space in highly myopic eyes. Invest
atrophy and retinal vessel diameter in nonglaucomatous Ophthalmol Vis Sci. 2004;45(8):26605.
optic nerve damage. Invest Ophthalmol Vis Sci. 1991;32 69. Ren R, Wang N, Li B, Li L, Gao F, Xu X, etal. Lamina cri-
(11):29427. brosa and peripapillary sclera histomorphometry in normal
55. Jonas JB, Xu L. Optic disc morphology in eyes after nonar- and advanced glaucomatous Chinese eyes with normal and
teritic anterior ischemic optic neuropathy. Invest Ophthalmol elongated axial length. Invest Ophthalmol Vis Sci.
Vis Sci. 1993;34(7):22605. 2009;50(5):217584.
56. Frisn L, Claesson M. Narrowing of the retinal arterioles in 70. Hinzpeter EN, Naumann G. Ischemic papilledema in giant-
descending optic atrophy. A quantitative clinical study. cell arteritis. Mucopolysaccharide deposition with normal
Ophthalmology. 1984;91(11):13426. intraocular pressure. Arch Ophthalmol. 1976;94(4):
57. Lim LS, Saw SM, Cheung N, Mitchell P, Wong TY. 6248.
Relationship of retinal vascular caliber with optic disc and
macular structure. Am J Ophthalmol. 2009;148(3):36875.
58. Panda-Jonas S, Jonas JB, Jakobczyk-Zmija M. Retinal pho- For a complete bibliography and detailed review of previously
toreceptor density decreases with age. Ophthalmology. published studies in the literature on the subject, please refer to

Clinical Evaluation of Patients
with Ischemic Optic Neuropathies 8
In my Ocular Vascular Clinic at the University of Iowa eventually to the correct diagnosis. Therefore, I spend
Hospitals & Clinics, when ischemic optic neuropathy a lot of my time to obtain a detailed history from the
patients are seen for the first time with a history of patient when he/she is first seen in my clinic. It is very
visual loss (usually sudden), their evaluation is done in time-consuming, but an investment of time which pays
the following two stages: dividends in arriving at the correct diagnosis.
(a) Initial evaluation in all patients with visual loss.
(b) Evaluation of outlier patients, who are judged, History of Present Illness
following the initial evaluation, not to have the
typical anterior ischemic optic neuropathy (AION) Since the University of Iowa Hospitals & Clinics is a
or posterior ischemic optic neuropathy (PION); tertiary referral center, the majority of the patients are
problem cases, in other words. referral patients. When I first ask the patient to tell me
what happened, it is not uncommon for them to start
by giving me the diagnosis of the previous ophthal-
mologist/neuro-ophthalmologist. I then tell them that
I nitial Evaluation in All Patients I am interested in what exactly his/her primary prob-
with Visual Loss lem is, and ask the patient to tell me in his/her own
words, in detail, all about what happened, right from
The following is a detailed account of my method of the beginning. I let the patient describe the process at
evaluating these patients. The data of my various clini- length and listen to what he/she has to say without
cal studies on ischemic optic neuropathies (summa- interrupting, although sometimes guiding them towards
rized in this book) are based on the findings from these the information which may be relevant. After that,
evaluations at the initial and follow-up visits. I am well I question the patient to obtain various specific pieces
aware that everyone has his/her own way of examining of information which the patient may not have given,
these patients, but this is how I have evaluated all my without hinting in any way what the right answer is.
patients in the various studies over the years. I particularly ask the patient for the following informa-
The entire work-up of all new patients is done by tion, if he/she has not already given it.
me personally. 1. Mode of onset of visual loss, i.e., whether sudden or
gradual.
2. How did the patient first discover the visual loss? That
is, suddenly noticing visual loss, or happening to cover
Detailed History one eye and discover the visual loss accidentally? Or
did he become aware of the visual loss for some other
The first, critical step in the diagnosis and management reason? How long ago was this discovered?
of these or any patients is to have a detailed history 3. The time of the day when he/she first noticed the
because this is the bed rock. What the patient has to visual loss. As far as I am concerned, this is critical
tell will provide critical information, and may lead information. Sometimes the patient may volunteer

146 8 Clinical Evaluation of Patients with Ischemic Optic Neuropathies
this information, while at other times one has to consult an ophthalmologist. Lack of awareness of this
extract it. To get the correct information, the right very basic fact can result in missing the correct diagno-
person has to ask the right question in the right sis. The eye is a part of the entire body and we know
manner. In my clinic, all new patients are person- that many systemic diseases can play important roles
ally seen by me and asked this question, again within the development of many ophthalmic diseases. This
out any suggestion of a right answer. is particularly so in the case of ischemic optic neuropa-
4. Was the visual loss accompanied by other ophthal- thies. Therefore, I quiz the patient repeatedly and thor-
mic or systemic symptoms? oughly, to obtain a detailed medical history. Since
5. What other eye problem had he/she already had arteritic AION is due to giant cell arteritis and that is
before the onset of this visual loss? This is impor- an ophthalmic emergency, a full questioning about the
tant; for example, a patient may have had glaucoma various symptoms and signs of that disease is critical
or ocular hypertension, which can play a role in the for immediate diagnosis and start of therapy to prevent
development of non-arteritic AION (NA-AION). severe and irreversible visual loss (see Chap. 12). We
6. I ask them to describe the location of their visual know that NA-AION is a multifactorial disease with
loss. many systemic risk factors playing roles (see Chap.
7. Since the onset, has the visual loss been stable, or 14). Our large prospective study [1], based on 406
progressively worsening, or improving? patients with NA-AION, showed association of many
8. Did he/she experience amaurosis fugax before the systemic diseases with NA-AION. Therefore, I ques-
visual loss? This is important because amaurosis tion the patient about all those, including diabetes mel-
fugax is seen in about one-third of patients with litus, hyperlipidemia, arterial hypertension, malignant
arteritic AION. Some patients with incipient arterial hypertension, ischemic heart and other cardio-
NA-AION also give such a history. vascular diseases, cardiac and valvular disease, carotid
9. I ask the patient if he/she has any other relevant artery disease, cerebrovascular disease, neurological
information to offer about the present illness. This disease, endocrinal disorders, hematologic disorders,
makes them think twice, and sometimes recall vital gastro-intestinal ulcers, hemodialysis, collagen vascu-
information at this point. lar diseases (including systemic lupus erythematosus,
polyarteritis nodosa and Wagners granulomatosis),
migraine and other vasospastic disorders, herpes
History of Previous Eye Diseases zoster, sleep apnoea, and renal, rheumatologic, infec-
tious, and other diseases. Any history of syphilis and
Detailed information about this is essential because HIV infection should also be obtained. Surgical PION
those may well be relevant, as for example, one eye develops following various types of systemic surgery
already had visual loss, glaucoma or ocular hyperten- (see Chap. 20); therefore, it is essential to find out
sion, cataract surgery, uveitis, etc. Development of about that. We know that distant repeated hemorrhages
NA-AION following intravitreal injection of anti-VEGF can result in the development of NA-AION (see Chap.
for age-related macular degeneration has recently been 15). Currently erectile dysfunction drugs are widely
reported (see Chap. 15). There is the possibility that that used, and those have been found to cause NA-AION
may also be occurring when intravitreal injection of (see Chap. 15). I have seen NA-AION develop follow-
anti-VEGF is given for other diseases, e.g., diabetic ing chemotherapy treatment of astrocytoma with
macular edema and macular edema due to other causes. BCNU (Carmustine). Finally, I ask every patient about
other major events any time in their life. Sometimes,
when asked about systemic diseases, the patient
responds at first that they never have had anything
Medical History wrong with them. In such a case, my next question
invariably is, have they ever seen a doctor, and why?
I always tell my Residents and Fellows that it is a That sometimes reveals new information. Thus, to get
patient with two eyes who is coming to consult us. It a full picture about systemic diseases is a time con-
is not one eye, an optic nerve, retina, cornea, lens or suming job but can provide important clues about
other part of the eye which walks into the clinic to ischemic optic neuropathies.
Initial Evaluation in All Patients with Visual Loss 147
Family History arterial hypotensive drugs to treat arterial hypertension

by physicians is not uncommon (some patients being
There is a rare entity of familial NA-AION where more treated aggressively for simple white-coat hyperten-
than one member of a family develops NA-AION (see sion) and that can have the same effect. I have seen
Chap. 15). Therefore, it is helpful to ask about that. development of NA-AION when patients were started
Some patients give a history that most of their relatives on Hytrin (Terazosin hydrochloride an alpha-1-selec-
have died of myocardial infarction or stroke at an early tive adrenoceptor blocking agent) or allied drugs (all
age, which is suggestive of familial hyperlipidemia. arterial hypotensive) for benign prostate hypertrophy,
We know that hyperlipidemia is a risk factor for devel- which is taken at bed time. Therefore, as a rule, it is
opment of NA-AION. In Lebers hereditary optic neu- most important to ask that question in the age group. If
ropathy there is a history of visual loss in the family. a patient is on any arterial hypotensive drug, it is
NA-AION is common among diabetics and a history important to ask how many of those drugs he/she is
of diabetes in the family may be relevant. I always ask taking and when. In my experience of referral cases,
if there is any history of blindness, poor vision or glau- I find that this important question about arterial
coma in the family and that information may be rele- hypotensive drugs is rarely asked by the ophthalmolo-
vant in some cases. gist or neuro-ophthalmologist, and an important clue
about the diagnosis and management of NA-AION is
thereby missed. It is important to ask men if they are
Social History using erectile dysfunction drugs specifically, because
those cause arterial hypotension and are usually taken
I ask each and every patient about smoking if he/she at bedtime they can cause NA-AION, and patients
has ever smoked, and if so when did he/she start, how are often unwilling to volunteer that information unless
much did he/she smoke, and if still smoking. I also ask asked directly.
them about drug abuse, and the use of alcohol (how
much and how often).
Detailed Ophthalmic Evaluation

Medications
It is crucial to evaluate the eye from A to Z, irrespec-
Information about the various drugs the patient is tak- tive of the suspected site of visual loss. It is not a good
ing, their dosage and frequency, is essential. This is practice to focus simply on the eye the patient is com-
because those may play a role in the development of plaining about or only the part of the eye, which one
NA-AION. Currently a large number of very potent suspects to be responsible for the visual problem.
drugs have emerged to treat arterial hypertension I have found that a very common problem in ophthal-
(a common disease in the middle-aged and elderly mology. For example, neuro-ophthalmologists are
when NA-AION is also common), and those are being concerned only with what is relevant to them in the
widely prescribed. My studies have shown that noctur- eye, retina specialists investigating only the retina, cor-
nal arterial hypotension plays an important role in the nea persons only examining the cornea and so on. For
development of NA-AION (see Chap. 14). My 24-h example, there was a cornea specialist who performed
ambulatory blood pressure recording study in more a corneal graft in a patient and put her on corticoster-
than 700 patients has shown that when arterial hypoten- oid eye drops on a long-term basis. He was interested
sive drug(s) is taken in the evening or at bedtime, it only in checking the corneal graft on regular follow-up
often results in abnormal nocturnal arterial hypoten- visits; he never measured the intraocular pressure.
sion. If a patient is taking these drugs in the evening, When I saw her, she had advanced glaucoma damage
I invariably ask him/her whether he/she is doing that with massive visual field loss and poor visual acuity
on the advice of his/her doctor or not. Not infrequently and intraocular pressure in the 40 s. That means her
the patients decide to take their medication in the eve- visual acuity and intraocular pressures were never
ning or at bedtime on their own, for no apparent reason. evaluated by the cornea specialist, whose sole interest
My studies have also shown that overmedication with was the state of the graft. Such a disaster can be avoided
if an eye is routinely examined fully. Over the years I the patients and the ophthalmologists visual acuity
have seen many similar examples. information is of major interest. Yet it is usually
Moreover, ophthalmologists sometimes evaluate assessed by persons who do not have full knowledge
only the eye about which the patient is complaining of the disease and the factors which influence the
and forget that the patient has two eyes and that the visual acuity testing. I have had patients referred to
second eye may also have something going on, which me (even from other clinics in my own department)
may provide crucial information about the cause. For with only hand motion visual acuity recorded by a
example, I discovered that asymptomatic optic disc technician, but less than half an hour later I can often
edema precedes the visual loss in NA-AION [2] and get as good as 20/200 or even better in them. This is
that incipient NA-AION [3] exists as a definite clinical because patients with a large central scotoma or with
entity by examining the fellow eye regularly at each a visual field defect involving central fixation area,
visit, irrespective of whether the patient had any com- have a natural tendency to fixate at the center of the
plaint related to the fellow eye or not. In a few cases visual acuity chart and say they cannot see anything.
I found early signs of arteritic AION in the fellow The technician then puts his/her fingers in the center
asymptomatic eye. The fact an eye is asymptomatic to find out if the eye can count fingers but since the
does not automatically rule out the possibility of any fingers are lying in the central scotoma, the patient
pathology in it. Similarly, patients with central retinal denies seeing that. The next step is that the technician
vein occlusion almost always have normal intraocular waves the hand, which goes into the seeing area and
pressure in the involved eye, but they may have a very a visual acuity of hand motion is recorded.
high intraocular pressure or glaucoma in the fellow eye Throughout my ophthalmic career, I have tested
(unknown to the patient), which was responsible for visual acuity in all my patients myself. For the reasons
the development of central retinal vein occlusion in the discussed above, it is well worth the time spent on that,
first place [4]. Missing that information puts the fellow to get valid information. As described above, there is a
eye at risk of developing glaucomatous visual loss or natural tendency for the patient to fixate at the center
even central retinal vein occlusion. These are just a of the chart, but when they have a central scotoma or
few examples of why a thorough evaluation of both when a visual field defect involves central fixation
eyes is crucial at the first and also at each follow-up area, they are not able to read the chart. Therefore,
visit. right at the start of testing, I tell them that by fixating
at the center of the chart they may not be able to see
properly or anything at all, but that they should look
esting Visual Function in Patients with Ischemic
T around to find out a spot where they can see best; that
Optic Neuropathies is, I encourage them to use eccentric fixation. Patients
often do not know what eccentric fixation means, so
I use the following four tests. I try to give them a simple example. I tell them that, for
example. If you go to a party and want to know what
someone is wearing, but you do not want that person to
Snellen Visual Acuity Testing see you are staring at him/her, what do you do? He/
she immediately says that you look ahead but look at
This is usually delegated to technicians, but I have them out of the corner of your eye. That gives them
found that testing of visual acuity is an art and a the idea what eccentric fixation is. I give a patient all
science. It is perhaps the most important test for all the time he/she needs to discover by various head or
ophthalmic patients for two reasons: (a) A patients eye movements where he/she can see best and how
main interest obviously is to find out how much he/ much, making sure that the other eye is completely
she can see and whether it is getting better or worse covered. Often patients are surprised how much better
that is primarily why they have come to consult an they can see with eccentric fixation. This also acts as
ophthalmologist. (b) Similarly, ophthalmologists are useful education for the patient to use eccentric fixa-
primarily interested to find out from published stud- tion. I have found that some patients are delighted to
ies whether a particular treatment had a beneficial find that they see better, and are able to watch televi-
effect or not on the visual acuity. Therefore, for both sion with eccentric fixation. On follow-up visits the
same procedure is repeated to get reliable visual acuity patient to look at the Amsler chart with the normal eye
data. Of course, valid visual acuity is the best corrected so that he/she gets the idea what it looks like normally.
visual acuity. Then I ask them to look at the chart with the involved
This limitation of simple visual acuity testing has eye only, covering the good eye. I ask them to imagine
important implications in the visual acuity results where the center of the chart is from the outline of the
reported by various studies dealing with various modes entire chart. While testing visual acuity, I was stressing
of treatment. When a patient has a central scotoma or a eccentric fixation all the time, but now I tell them that
visual field defect which involves the central fixation they have to fixate firmly at the center of the chart all
area, he/she tests poorly at initial testing, with poor the time, even if they do not see that, and not use eccen-
visual acuity for the reasons mentioned above. In due tric fixation at all. Then I ask them to outline the defect
course many of them may learn to fixate eccentrically with a pencil/pen and tell me what is wrong with that
and see much better and that apparent improvement area. To know the nature of the defect is important,
may erroneously be attributed to the beneficial effect because at subsequent visits not only the change in the
of a treatment, when in fact that really is not the case at size of the defect but also the nature of the defect is
all. This is not rare among the reported claims of ben- important. For example, the involved area may initially
eficial effects of some treatments in some published be totally dark but on subsequent visits the darkness
studies. may become less and less marked, showing progres-
sive improvement. I keep those charts in the patients
chart for subsequent comparison. I have found that that
Near Visual Acuity has provided very useful information in visual
evaluation.
This is also recorded in both eyes. Since most patients
with NA-AION use bifocal glasses, the near visual Visual Field Testing
acuity is tested with those glasses, holding the test
chart at the right distance from the eye. I have always All eyes with ischemic optic neuropathies have optic
used the following near vision test chart: Reading nerve related visual field defects. Therefore, perimetry
Test Types Approved by the Faculty of Ophthalmologists, is essential in each and every case to get the correct
London, England. Compared to the commonly used information. This provides important information
Jaeger chart (with words only), this near vision test which is not provided by the visual acuity or Amsler
chart has the distinct advantage of being in the form of grid testing. This is because visual acuity is a function
a 4-page booklet containing 11 paragraphs in ascend- primarily of the foveal region (maculopapillary nerve
ing print sizes; the experience is like reading a book. fibers), and it gives no information at all about what is
Patients prefer that to the Jaeger chart. going on in the rest of the retina or the optic nerve.
Currently visual field testing is done by two methods:
(1) manual kinetic perimetry, and (2) automated static
Amsler Grid Testing
threshold perimetry. Both types have their advantages
and disadvantages, and one should be aware of those
This is very useful to test the central visual field
when interpreting the findings.
defect. I have seen time and time again that when
there is a metamorphopsia in the central region which 1. Automated static threshold perimetry: Currently,
results in visual acuity as bad as 20/200, regular automated threshold perimetry [Humphrey 30-2 or
perimetry may show no defect but Amsler grid test- 24-2 SITA (Swedish Interactive Testing Algorithm)]
ing immediately outlines that defect. This is because is the most widely used form of perimetry. It rou-
with the Amsler grid we are testing resolution but not tinely gives information about the central 24 or
so with perimetry. 30. It is quantitative and can provide information
Since most patients with NA-AION use bifocal on the depth of threshold loss in that region. Patients
glasses, like the near visual acuity testing, Amsler grid with ischemic optic neuropathies and severe visual
testing is also done with those glasses, holding the test field loss have profound loss of threshold over the
chart at the right distance from the eye. I first ask the entire 24 or 30. In patients who can perform in
a reliable fashion (with only a small number of false highly useful manual kinetic perimetry is rapidly
positive and negative results and few fixation being replaced by automated perimetry.
losses), the probability plot in the computerized The entire subject of perimetry in NA-AION is dis-
report is useful because it provides a statistical basis cussed at length in Chap. 16.
for how abnormal each tested point is with regard to
the threshold, compared to age matched normal
eyes. However, automated threshold perimetry has Color Vision Testing
some severe limitations compared to manual kinetic
perimetry (recorded with a Goldmann perimeter). I have found that in eyes with poor visual acuity due to
2. Manual Kinetic Perimetry: The visual field informa- ischemic optic neuropathies, this is not a reliable test,
tion provided by manual kinetic perimetry performed and does not provide any useful information. I seldom
with a Goldmann perimeter may be very different use this.
from that by automated static threshold perimetry.
The latter does not provide any information beyond
the central 2430. Kinetic perimetry, by contrast, Complete Ophthalmic Evaluation
provides information all the way to about 8090
temporally, 70 inferiorly, 6070 nasally and 5060 This involves examining both eyes thoroughly from A
superiorly. This has important implications in char- to Z, without any exception. Both the anterior and pos-
acterizing the visual field defects in NA-AION and terior segments must be evaluated meticulously.
evaluating visual disability caused by NA-AION.
nterior Segment and Slitlamp Biomicroscopic
A
haracterization of Visual Field Defect Based
C Evaluation
on the Two Types of Perimetry in NA-AION
Each part of the eye must be examined meticulously.
Information provided by automated perimetry may
result in different characterization of pattern of the 1. External Evaluation: This may show abnormalities
visual field defects compared to that by kinetic perim- which may be important, e.g., proptosis. Evaluation
etry. For example, in automated perimetry a large sco- of the extraocular muscle function is essential in
toma or defect extending all the way to 2430 is any neurological assessment.
almost invariably interpreted as an altitudinal field 2. Penlight Evaluation: This includes evaluating eye-
defect (Fig.8.1a), when kinetic perimetry may show lids, conjunctiva, episclera, sclera and cornea. The
a normal peripheral field beyond that (Fig.8.1b); this most important test with this method is the relative
has important implications: (1) This may be one of the afferent pupillary defect and measuring it in log
reasons that an inferior altitudinal visual field defect is units with various filters by the swinging flashlight
almost invariably described as the classical field defect test. For example, if the left eye has ischemic optic
in NA-AION, which it is not in fact the case in most neuropathy, when the light is shining on the normal
NA-AION eyes (see Chap. 16) (Fig.8.1), and (2) Full right eye both pupils constrict, but when the light is
information about the peripheral visual fields (which switched to the left eye (with ischemic optic neu-
the automated does not provide) is critical to evaluate ropathy) both pupils dilate; once again, when the
the degree of visual disability, since the peripheral field light is swung back to the normal right eye, both
plays a critical role in our navigating and day-to-day pupils constrict again. The degree of relative affer-
living. In addition, I have found that most patients pre- ent pupillary defect is measured by placing various
fer manual kinetic perimetry over automated perime- filters in sequence in front of the normal right eye
try, and patients who are too old or too young to and the above test is repeated; when the appropriate
participate in automated perimetry do well on manual filter is in place, that neutralizes the afferent pupil-
kinetic perimetry. It is most unfortunate that the lary defect and both pupils constrict equally.
Fig. 8.1 This shows the difference in pattern of visual field Inferior paracentral scotoma seen in (b) plotted with Goldmann
defect revealed by visual fields plotted with an automated perim- perimeter was interpreted in automated perimetry field in (a) as
eter (a) and Goldmann perimeter (b) in an eye with NA-AION. inferior altitudinal defect
b flow in the optic nerve head depends upon the perfu-

sion pressure (mean blood pressure minus intraocular
pressure) in it. A rise of intraocular pressure may lower
the perfusion pressure below the critical level and
result in optic nerve head ischemia (see Chap. 5), as
seen in eyes with ocular ischemic syndrome and in
post-cataract extraction NA-AION. Testing of intraoc-
ular pressure is an art and science. This is because the
intraocular pressure is influenced by the following
factors;
1. Holding the breath during testing is common, a
natural tendency on the part of the patients. That
raises the CO2 blood level, which immediately
causes dilation of the uveal vascular bed. That in
turn, results in a rise of intraocular pressure I have
seen that to go as high 5mmHg or even higher in
some cases.
Fig.8.1 (continued) 2. Looking up, once again not an uncommon reaction
on the part of patients, results in a rise of intraocular
pressure by compression of the eyeball by the con-
tracted extraocular muscles. This plays an impor-
tant part in thyroid eye disease.
3. Slitlamp Biomicroscopic Evaluation: A thorough 3. Squeezing the eyelids again raises the intraocular
slitlamp examination of the entire anterior segment pressure.
from front to back is essential. The conjunctiva may 4. Lifting the eyelid by the examiner can also com-
show marked chemosis, sarcoid nodules or other press the eyeball in some case and result in a rise of
abnormalities. There may be marked congestion of intraocular pressure.
blood vessels in the episclera and sclera. I have seen
Thus there are several pitfalls in measuring the intraoc-
patients with scleritis which involved the optic
ular pressure accurately, which are often not realized.
nerve head and resulted in a diagnosis of NA-AION.
Since the level of intraocular pressure is important for
In the cornea, detection of some abnormalities, such
several reasons, not only in ischemic optic neuropa-
as amiodarone keratopathy, may be helpful.
thies but in also several other ocular vascular disor-
Evaluation of the anterior chamber for evidence of
ders, I evaluate the intraocular pressure with a
uveitis or bleeding or narrow angle or any other
Goldmann applanation tonometer myself, in all my
abnormality is important. I have seen a few cases
cases. I give the following instructions to the patients
where an eye had a transient bleed from an old cata-
before starting to measure the intraocular pressure: (1)
ract operation site, causing recurrent episodes of
keep breathing normally and do not stop breathing, (2)
amaurosis fugax, which resulted in extensive,
look straight ahead all the time, and (3) do not
expensive and unnecessary neurological evaluation;
squeeze.
this could have been totally avoided by a careful slit
Examination of the lens is done after dilating the
lamp evaluation of the anterior chamber for red
pupil because that provides the best information about
blood cells. Examination of the iris for evidence of
various lens changes, e.g., nuclear sclerosis, lens opac-
iritis, posterior synechiae, neovascularization or
ities, posterior subcapsular cataract and pseudoexfolia-
other abnormalities can be helpful. Evaluation of
tion. It is well-know that poor visual acuity may simply
the angle on gonioscopy may show evidence of
be due to lens changes.
angle closure or neovascularization in some cases.
A detailed examination of the vitreous is impor-
Evaluation of the intraocular pressure is absolutely tant in all cases. In ischemic optic neuropathies,
essential in eyes with NA-AION, because the blood there are usually age-related vitreous changes. The
presence of posterior vitreous detachment in the age 3. Contact Lens Biomicroscopic Examination:
group of ischemic optic neuropathies is not uncom- For this I always use a Hruby lens, which gives
mon; however, vitreous traction has rarely been pos- me a very clear three dimensional image. I use
tulated as responsible for the development of this only when I have any doubt about the optic
NA-AION. The most important thing to look for is disc and macular findings from direct
the presence of cells in the vitreous; this is critical ophthalmoscopy.
for the differential diagnosis of NA-AION from other
conditions (see Chap. 16).
Fundus Photography
I always take 30 and often also 60 stereoscopic fun-
Fundus Evaluation dus photographs of the involved eye. Stereoscopic
fundus photographs of the optic disc provide very use-
In neuro-ophthalmology the focus of examination of ful information to evaluate optic disc changes initially
the fundus in eyes with ischemic optic neuropathies and on a serial follow-up of patients with AION. If
usually tends to be mainly on the optic disc evalua- there is any suspicion of macular changes, stereoscopic
tion and the fundus in its immediate vicinity. As a photographs of the macular region are also taken
routine, I examine every eye using (1) indirect oph- separately.
thalmoscopy, (2) direct ophthalmoscopy, and (3) con-
tact lens in that order, after dilating the pupil. Each
of these methods of examination supplies different
Fluorescein Fundus Angiography
information. One could compare this to examining a
histological section under the low, medium and high In 1964 I was one of original pioneers of fluorescein
powers of a microscope. fundus angiography. I have used it extensively to eval-
uate various ocular circulatory beds in my experimen-
1. Indirect Ophthalmoscopy: With this method, I exam- tal and clinical studies ever since. The advent of
ine the entire fundus thoroughly, to find any abnormal- fluorescein angiography revolutionized our under-
ity anywhere in the fundus. I have found this very standing of the retinal, optic disc and choroidal circu-
useful in eyes with ischemic optic neuropathies, to lations in health and disease.
rule out any other factor which may be playing a role. I have always performed fluorescein angiography
2. Direct Ophthalmoscopy: This is invariably the (normally stereoscopic) regularly in all eyes with
method used by neuro-ophthalmologists. No doubt AION, some 1,300 eyes now. It provides important
this is a very useful technique, but it mainly evalu- and useful information when performed during the
ates the posterior fundus and totally misses the find- initial stages of the disease. The telltale impaired cir-
ings beyond that. A detailed evaluation of the optic culation and its location in AION are seen on angiog-
disc and adjacent retina is critical in AION. I have raphy only during early stages of the disease and
found that further evaluation with a red free filter in during the very early retinal arterial phase of dye
direct ophthalmoscopy often reveals findings not filling in the fundus. In my clinical studies on AION,
evident from ordinary direct ophthalmoscopy. For angiography has provided crucial information about
example, it enhances nerve fibers, abnormal blood its pathogenesis as well as the differential diagnosis of
vessels, hemorrhages, microaneurysms, lipid depos- arteritic from NA-AION. The fluorescein fundus
its (the so-called hard exudates) and microcystic angiograms given in Chap. 3 demonstrate the value of
macular edema; this can provide very useful addi- angiography in diagnosis of arteritic and NA-AION,
tional information. On direct ophthalmoscopy, as well as its use in revealing the optic nerve head
I have found that in eyes with optic disc drusen, if blood supply in AION. Its role in NA-AION is dis-
the drusen are retroilluminated by putting the oph- cussed in Chap. 16. It is most unfortunate that this
thalmoscope light at the margin of the optic disc, extremely useful and readily available test is not
the drusen light up. That makes it easy to find the used by neuro-ophthalmologists as often as it should
drusen even when they may not be clearly visible be in evaluation of AION. Many NA-AION patients
on direct ophthalmoscopy per se. who come to consult me, who have previously seen
neurologists/neuro-ophthalmologists come loaded I see no point in doing detailed thrombophilic testing

with optical coherence tomography, magnetic in patients with NA-AION since it is not a thrombotic
resonance imaging and/or magnetic resonance angiog- disorder but a hypotensive disorder (see Chap. 14).
raphy films ordered by the neurologists/neuro-
ophthalmologists, showing no abnormality at all. I
have also seen patients who lost vision in one eye due
to occult giant cell arteritis and had multiple neurora- Temporal Artery Biopsy
diological evaluations done by neurologists/neuro-
ophthalmologists, when doing fluorescein angiography
This is done when a diagnosis of arteritic AION is sus-
would have immediately provided a definite diagno-
pected from clinical findings (see Chap. 12), or giant
sis (see Chap. 12); delay in some of these cases
cell arteritis is suspected from the elevated ESR and
resulted in bilateral complete blindness which could
CRP, and other clinical finding (see Chap. 10). The
have been prevented completely. I find that fluores-
subject of temporal artery biopsy in giant cell arteritis
cein angiography provides much more useful infor-
is discussed at length in Chap. 11. In cases suspected
mation in evaluation of AION than neuroradiological
to have giant cell arteritis, there is no urgency to do a
evaluations, but it is rarely performed. Fluorescein
temporal artery biopsy immediately, but an immediate
angiography is also much cheaper than those neurora-
start of the treatment with high-dose corticosteroid
diological evaluations. My experience of dealing with
therapy is most critical in these cases to prevent any
AION during the past 4decades has been that neuro-
visual loss. Corticosteroid therapy does not alter the
radiological evaluations only rarely reveal any
outcome of temporal artery biopsy, as discussed in
abnormality.
Chap. 11. To wait for a temporal artery biopsy result to
start corticosteroid therapy may result in permanent
visual loss which can be prevented; therefore, my
Hematological Evaluation advice always has been not to wait for the temporal
artery biopsy result to start the corticosteroid therapy.
I do not recommend doing temporal artery biopsy on
I routinely do the following hematological evaluations both sides at the same time after all, temporal artery
in all patients first seen with visual loss. biopsy is not a benign procedure and I have seen scalp
1. Erythrocyte sedimentation rate (ESR) and necrosis following this practice. To get reliable infor-
C-reactive protein (CRP) testing: This is done reg- mation from a temporal artery biopsy, it is essential to
ularly on the initial visit in all AION, PION and keep the following two most important considerations
retinal arterial occlusion patients aged 50years or in mind: (1) there should be at least a one inch piece of
older. Testing for ESR and CRP is crucial for all temporal artery (Fig.8.2), and (2) the artery must be
these patients, without any exception, to rule out serially sectioned. For example, in one of my cases,
giant cell arteritis, since that is an ophthalmic emer- only 1 of about 300 sections showed definite evidence
gency and its early detection and immediate inten- of giant cell arteritis.
sive high-dose corticosteroid therapy can prevent
visual loss (see Chap. 13). In my clinic, I receive the
result of both tests in 11 h. My patients do not
leave the clinic till I have the results of both ESR Systemic Evaluation
and CRP available to make sure no patient with
giant cell arteritis is missed. If the diagnosis of AION or PION is definite, then the
2. Fasting lipid profile and glucose evaluation: This patient requires complete systemic evaluation because
is done on a subsequent follow-up visit, or as soon both NA-AION and PION are multifactorial in nature,
as possible. with systemic risk factors playing a role in their
Unilateral Optic Disc Edema with Visual Field Loss 155
urther Evaluation of Optic Disc Edema

F
in Outlier Cases
Randy Kardon
When a patient presents with unilateral or bilateral

optic disc edema associated with a history and exami-
nation that is supportive of a diagnosis of NA-AION,
most experienced clinicians would not pursue further
ancillary testing, such as neuro-radiologic imaging,
optical coherence tomography (OCT), extensive blood
tests, or carotid and heart ultrasound imaging. However,
some cases of optic disc edema may not show the typi-
cal history, examination features and epidemiological
profile expected with NA-AION. These patients often
Fig. 8.2 Temporal artery biopsy. Arrows show the superficial need further evaluation to look for an alternative
temporal artery ligated on two sites before at least 1in. piece of diagnosis.
it is cut for serial sectioning Patients with the following conditions often warrant
further evaluation:
development. This includes evaluation for various risk
factors, including diabetes mellitus, hyperlipidemia,
arterial hypertension, arterial hypotension, collagen
vascular diseases, sleep apnoea and other possible risk nilateral Optic Disc Edema with Visual
U
factors, depending upon a patients status of health Field Loss
(see Chap. 14). Information about these risk factors is
essential in the management of NA-AION (see Chap. Consider the following alternative causes of unilateral
17) and non-arteritic PION (see Chap. 22). I advise all optic disc edema besides NA-AION, particularly in
patients to have a complete physical checkup. young patients or patients with a history of vision loss
that does not follow the typical pattern of NA-AION.
(a) Acute neuro-retinitis before a macular

ollow-up Evaluation of Eyes
F
star develops
with AION and PION These patients usually have vitreous cells and a central
visual field scotoma on acute presentation in one eye,
The follow-up protocols for these eyes are discussed in but rarely both eyes may be involved. If the scotoma is
the chapters dealing with the management of AION due to a neurosensory detachment with subretinal
(Chaps. 13 and 17) and PION (Chap. 22). At each visit, fluid, then the visual field loss is often reversible.
a brief history is recorded about the change in visual However, some cases of neuroretinitis involve direct
status, any new ocular or systemic development, and if injury to the neural layers in the retina and may result
on corticosteroid therapy a detailed evaluation of any in permanent loss. In this setting, optical coherence
side-effects of the therapy and dosage. At each visit, tomography (OCT) may help to identify the extent
the detailed ophthalmic evaluation is exactly the same (area) of neurosensory detachment to determine if it
as described above. corresponds to the area of the visual field defect. OCT
of the macula may also show thickening in the outer which may be confused with hyperemia. These patients
plexiform layer, a location where exudates associated may present with the first eye involved or after the sec-
with a macular star may develop over time. Exudates ond eye becomes involved, at which time optic atrophy
are seen as highly reflective discrete bodies within the may already be present in the eye that was first affected.
outer plexiform layer with OCT. If there is still a cen- When there is clinical suspicion of Lebers Hereditary
tral scotoma when fluid has resorbed and is not present Optic Neuropathy, then genetic testing should be per-
on OCT, then multifocal electroretinography (mfERG) formed to look for one of the three known mitochon-
may reveal decreases in evoked potentials from the drial mutations.
inner retina. When neuroretinitis is suspected, a serum (d) Acute optic neuritis with disc edema
sample is submitted for analysis of antibodies to In younger patients (i.e., under the age of 45), acute
Bartonella henselae, a protobacterium, which provides optic neuritis can sometimes be a challenge to differ-
both IgG and IgM levels. However, not all cases of entiate from NA-AION occurring in a young person.
neuroretinitis will have antibodies to this organism, so There are also rare patients that can develop acute optic
a negative serum test does not rule out neuroretinitis. neuritis older than age 45. With time, most patients
In most cases the central visual field defect resolves with acute optic neuritis having visual field loss recover
over time, but this may take many months; some cases to a greater extent than has been reported with
show permanent loss of neurons in the inner retina NA-AION. However, some patients with severe optic
with permanent residual visual field loss. neuritis and poor vision during the acute stage may not
(b) Optic disc vasculitis have significant recovery of vision, and this is espe-
Patients with optic disc vasculitis, by definition have cially true of neuromyelitis optica, a rarer form of optic
vitreous cells, a disc that leaks fluorescein on fluores- neuritis not associated with multiple sclerosis.
cein angiography, and usually have chronic disc edema Clinically, there is usually only mild optic disc edema
that does not resolve spontaneously over time, as with acute optic neuritis, compared to more significant
would be the case with classic NA-AION. These optic disc edema seen with NA-AION. However, with
patients usually have very little visual field defect, the advent of OCT, the percentage of patients with reti-
except for an enlarged blindspot, awareness of which nal nerve fiber layer thickening is much greater than
often brings them for consultation. Usually they com- previously reported, often greater than 50% in some
plain more about visual symptoms than revealed by series. Although pain with eye movement is a common
visual testing. It is a self-limiting disease. There may feature of acute optic neuritis, this is not a differentiat-
be an underlying inflammatory disorder responsible ing feature, since some patients with acute NA-AION
for the vasculitis, so serum tests for autoimmune disor- also may have orbital pain. In cases of acute optic neu-
ders, sarcoid, and syphilis may be indicated. ritis, neuroimaging with magnetic resonance imagin-
(c) Lebers hereditary optic neuropathy ing (MRI) usually reveals enhancement of the inflamed
These patients are classically in the second or third area of the optic nerve with gadolinium contrast,
decade of life and the visual loss is frequently preceded whereas NA-AION typically does not result in
by binge alcohol drinking. However, there are a num- enhancement of the optic nerve with MRI imaging.
ber of case reports presenting with sequential visual The presence of other areas of demyelination in the
loss in one eye and then the other within 612months white matter of the brain is also a common finding on
that are older, even in as old as 80. Unlike NA-AION, brain MRI imaging in patients with optic neuritis.
the optic disc appearance is more hyperemic than Patients with NA-AION may have similar changes on
frankly edematous. It has also been reported that the MRI imaging due to small vessel ischemia, but these
retinal nerve fiber layer (RNFL) as measured by OCT are not restricted to just white matter, as in the case of
is thickened even in sectors without visual loss and demyelination of the brain with multiple sclerosis
also in the asymptomatic fellow eye. As atrophy devel- associated with acute optic neuritis.
ops, the RNFL becomes thin in the sectors correspond- (e) Orbital compressive optic neuropathy with disc
ing to the visual field loss, which is usually in the edema
ceco-central location. The major distinguishing feature Most patients with compressive optic neuropathy (e.g.
is that the visual field loss is out of proportion to the Gravess orbitopathy) and visual field loss do not show
disc appearance and that there is absence of disc edema optic disc edema, except when the compression occurs
Unilateral Optic Disc Edema Without Visual Field Loss 157
within the orbit. Enlarged extraocular muscles, orbital associated neuroimaging signs such as a concave sella
tumors such as meningiomas or hemangiomas may pro- turcica, flattening of the posterior globes and focal nar-
duce optic disc edema. The optic disc edema is usually rowing along the transverse/sagittal venous sinuses.
chronic, not segmental, and may also be due to venous With a high index of suspicion, a lumbar puncture
congestion. In cases of visual field loss associated with should be considered to evaluate opening pressure and
chronic optic disc edema, an orbital imaging study is fluid content.
indicated (orbital ultrasound, CT or MRI scan). (h) Elevated optic nerve due to buried optic disc
(f) Inflammatory optic neuropathy other than optic drusen or anomalous disc
neuritis These patients are usually asymptomatic and are dis-
These, for example, include toxoplasmosis or sarcoid covered on routine fundus evaluation. The nerve appears
disc granuloma, syphilitic optic neuropathy, and poste- elevated but rarely appears swollen as is the case with
rior scleritis. Inflammatory causes of visual field loss NA-AION and do not have associated superficial hem-
that also cause optic disc edema are usually located at orrhages or frank edema of the retinal nerve fiber layer.
the anterior portion of the optic nerve. Such cases may Sometimes there is an associated subretinal hemorrhage
be associated with prominent pain (posterior scleritis) in the peripapillary retina. These patients have normal
and may require MRI imaging to investigate for enhance- vision and most of the time have normal visual fields. In
ment of the optic nerve within the orbital portion. some cases, where the peripheral visual field is mea-
Inflammation of the optic nerve head may also show an sured with kinetic perimetry, there may be peripheral
abnormal topography of the optic nerve on examination visual field loss in an arcuate or a scalloped pattern, but
if a granuloma is present within the disc tissue. Usually the central visual field is almost always normal. If there
the visual field and vision loss is progressive and the are buried drusen that are calcified, but cannot be seen
optic disc edema is chronic. At the time of discovery, on fundus examination, then orbital ultrasound may be
these unusual causes of visual field loss associated with of benefit in detecting drusen. A noncontrast CT scan
disc edema may be difficult to differentiate from may also disclose the calcified drusen. Since axoplas-
NA-AION. Some cases of posterior uveitis have an mic flow stasis is common, an OCT scan of the retinal
associated, secondary optic disc edema, and are associ- nerve fiber layer may show mild thickening.
ated with other retinal signs of uveitis and vitritis.
(g) Asymmetric or unilateral optic disc edema due
to raised intracranial pressure
Most cases of raised intracranial pressure have bilateral
nilateral Optic Disc Edema Without
U
optic disc edema, and the associated visual field loss
usually consists of an enlarged blind spot. However,
Visual Field Loss
some cases may have neurosensory detachment associ-
ated ceco-central visual field loss or an arcuate bundle (a) Incipient AION
visual field defect associated optic nerve damage. The As described in another chapter (see Chap. 15), incipi-
latter may be a form of optic disc ischemia associated ent NA-AION presents with asymptomatic optic disc
with optic disc edema caused by raised intracranial pres- edema, without visual field loss. These patients usually
sure. Some cases of optic disc edema from raised intrac- have risk factors associated with NA-AION (see Chap.
ranial pressure are asymmetric, owing to asymmetric 14). If these factors are present, then interventions to
anatomical resistance to the transmission of fluid pres- reduce the risk factors should be undertaken, providing
sure within the arachnoidal trabeculae within the optic there are no other signs, such as vitreous cell or symp-
canal. Most of these cases have associated symptoms of toms of inflammatory optic neuropathy or raised
raised intracranial pressure, such as pulse synchronous intracranial pressure (see above).
tinnitus and headache. The disc edema, even if asym- (b) Vitreous traction syndrome
metric is usually chronic. Asymmetric optic disc edema In some patients the vitreous attachment to the optic
with some elevation of the nerve head in the lesser nerve head may cause traction, resulting in chronic optic
affected eye accompanied by symptoms of raised intrac- disc edema and leakage of fluorescein angiography.
ranial pressure may warrant neuroimaging to look for OCT of the optic nerve head is a very sensitive imaging
causes of raised intracranial pressure and other test to demonstrate vitreous traction with tenting up of
disc tissue. These cases usually do not show visual field ain Ancillary Tests Useful in the
M
defects except for an enlarged blind spot. Many are self Differential Diagnosis of the Above
limiting and resolve following a posterior vitreous
Outlier Cases
detachment. If associated with macular edema and
visual field loss, vitrectromy should be considered.
(c) Pseudopapilledema and buried optic disc Radiological Tests
drusen
See above (some cases have no associated visual field These include magnetic resonance imagining (MRI)
loss). of the orbit (fat suppression) and of the brain with and
without gadolinium (to look for compressive optic
neuropathy, enhancement of optic nerve from inflam-
matory optic neuropathies, and infiltrative tumors)
ilateral Optic Disc Edema and Visual
B
Field Loss
MRI Scan
(a) Raised intracranial pressure In this setting, the purpose of neuroradiologic investi-
See above for description under asymmetric optic disc gation is to provide evidence for causes of visual loss
edema. Raised intracranial pressure may be due to and optic disc edema other than NA-AION. The fol-
intracranial masses, venous sinus thrombosis, sleep lowing are the main outlier causes of optic neuropa-
apnea causing both increase in intracranial pressure thy not due to NA-AION that cause the following
and/or NA-AION, and idiopathic intracranial hyper- abnormalities on MRI scan:
tension with visual field loss due to NA-AION-like
1. Enhancement of the optic nerve with gadolinium
ischemia, especially in hypertensives. As described
and fat suppression
above, these patients usually have symptoms of raised
Enhancement of the anterior visual pathway usually
intracranial pressure and chronic optic disc edema,
indicates a breakdown of the blood-optic nerve barrier
unlike NA-AION. They usually require neuroimaging
at the capillary endothelial level, which can occur in
with MRI and MRV and lumbar puncture.
pathology such as granulomatous inflammation (sar-
(b) Malignant arterial hypertension
coidosis or Wegners granulomatosis), non-granu-
These patients may or may not show other retinal or
lomatous inflammation of the optic nerve (acute
choroidal signs of high blood pressure. This is why
demyelinating optic neuritis, autoimmune optic neuri-
blood pressure should be measured in patients with
tis, neuromyelitis optica), infiltration of the optic nerve
optic disc edema. An automated blood pressure cuff
by cancer (metastatic disease, malignant glioma of
measurement may sometimes be more reliable than a
adults, and in rare cases lymphomatous infiltration),
manual blood pressure reading when the ancillary staff
and in some cases of PION.
does not have enough experience.
2. Compression of the optic nerve by external
masses
These include meningioma, enlarged extraocular mus-
o Optic Disc Edema, but Decrease in
N cles (Graves compressive ophthalmopathy), retro-
orbital masses (i.e., cavernous hemangioma, orbital
Visual Function (Tentative Diagnosis
varix, carcinomas) and intracranial aneurysms com-
of Posterior Ischemic Optic Neuropathy) pressing the optic nerve.
3. Compression of the optic nerve by internal
Since posterior ischemic optic neuropathy (PION) is a masses
diagnosis of exclusion, in them one has to rule out This includes intrinsic tumors of the optic nerve such
compressive optic neuropathy, infiltration of optic as optic nerve glioma
nerve with tumor, sarcoid or Wegners optic neuropathy, 4. MRI evidence of arteritic AION
and ethambutal toxicity and retinopathy. See Chap. 21 Recently, there have been reports showing periarterial
on PION. enhancement of the temporal and occipital arteries
Main Ancillary Tests Useful in the Differential Diagnosis of the Above Outlier Cases 159
on fat suppressed MRI scans with gadolinium, which expertise and technical specifications of imaging cen-
help support a diagnosis of temporal arteritis. ters, it is still considered the gold standard upon
5. MRI evidence for bilateral postgeniculate which other emerging technology is compared.
pathology
This affects both hemifields of each eye due to stroke
or masses (e.g. bilateral occipital stroke), in which
case the optic nerves appear normal on exam, but the
Ultrasound of the Orbit
patient has visual loss in each eye.
Unlike other imaging modalities, such as CT and MRI,
orbital ultrasound requires an examiner with the
CT Scan knowledge and experience required to produce the
correct orientation of the ultrasound probe with respect
Orbital CT scan without contrast is often very sensitive to the orbital tissue to define a specific soundwave-tis-
for diagnosing Graves compressive optic neuropathy sue interaction that yields accurate diagnostic informa-
when the clinical examination raises this suspicion and tion. Many institutions do not have the resources to
in this setting may be substituted for an orbital MRI train and maintain echographers experienced in both
scan; in addition, if there is any suspicion of optic canal A and B scans, so ultrasound is being used to a lesser
narrowing (causing compressive optic neuropathy) by extent presently, especially in A scan mode. However,
hyperostosis of bone caused by meningioma, Crousons in the right hands, it can be an important diagnostic
disease in children, and tumors of osseous tissue (i.e., tool for defining orbital tumors, vascular abnormalities
fibrous dysplasia). Orbital CT with contrast can be and optic nerve pathology, including compressive optic
helpful in diagnosing retro-orbital vascular abnormali- neuropathy due to pathologically enlarged extraocular
ties such as hemangioma and varix and with newer CT muscles.
angiograms, aneurysms compressing the optic nerve.
Optical Coherence Tomography (OCT)

PET (Positron Emission Tomography) Scan
To evaluate the extent of optic nerve fiber loss, retinal
Since PET provides a relatively low resolution inven- thickness (in old branch or central retinal artery occlu-
tory of highly metabolic tissue, it may be useful in pro- sion), and subretinal fluid, OCT has seen a great
viding additional evidence of inflammatory causes of increase in use in ophthalmology and in neuro-oph-
optic neuropathy, such as sarcoidosis and Wegners thalmology in recent years, but the indication for its
granulomatosis and in some cases, metastatic tumors use in the setting of visual loss where the cause is still
in and around the optic nerve. Conversely, tissue with uncertain is still being refined, especially as the tech-
abnormally low metabolic consumption can be diag- nology is evolving. The following is a list of its present
nosed and may be useful in cases of Alzheimers vari- use as an ancillary test in patients in whom a diagnosis
ant of bilateral vision loss, in which case the occipital of NA-AION is equivocal:
cortex shows very low metabolic activation on PET
(a) Optic atrophy without a clear history of
scans.
preceding optic disc edema
In some cases of presumed NA-AION where the
patient comes to medical attention with a pale nerve
Conventional Cerebral Angiography and permanent visual field loss, it may be helpful to
rule out other causes, such as previous branch or cen-
With the increase in sensitivity and specificity of MRI tral retinal artery occlusion. Although both NA-AION
and CT angiography, conventional angiography is now and artery occlusions will show thinning of the retinal
used primarily to define vascular pathology during nerve fiber layer after a number of months, patients
interventional treatment, as in the case of aneurysms with artery occlusions have more thinning of the total
and vascular malformations. Depending upon the retinal thickness (i.e., macula thickness shows highly
statistically significant thinning) compared to patients settings where the diagnosis of ischemic optic neurop-
with NA-AION, owing to the damage to the nerve fiber athy may be in doubt:
layer, ganglion cell layer and a portion of the bipolar (a) Optic atrophy associated with visual field loss with-
cell layer. Therefore a helpful footprint of a previous out a clear history of preceding optic disc edema
retinal artery occlusion besides narrowing of arteries is As stated above, some of these patients may have suf-
that the total retinal thickness is less than one would fered from a central or branch retinal artery occlusion
expect purely from an optic neuropathy such as and not NA-AION. In these cases, the evoked poten-
NA-AION or non-arteritic PION, for the same degree tials from the inner retina, which includes the bipolar
of visual field loss and retinal nerve fiber layer cell layer is usually reduced; after a central retinal
thickness. artery occlusion, the full field ERG usually shows
(b) Very subtle optic disc edema in early, mild reduced b-wave amplitudes from the inner retina and
NA-AION or during resolution of edema in branch retinal artery occlusions the multifocal ERG
Some patients may have equivocal optic disc edema on may be more useful in detecting focal loss that corre-
clinical examination and in these cases the OCT may sponds to the location of the visual field deficit.
help to substantiate that the retinal nerve fiber layer of (b) Acute visual field loss with a non-edematous
the eye with suspected AION is thicker when com- optic nerve
pared to the other fellow eye. The OCT may also help In this setting when PION is suspected, other causes must
substantiate whether the optic disc edema is resolving be ruled out (compressive optic neuropathy or retinal
over time when studied at different visits. causes). Multifocal electroretinography may be useful for
(c) Optic disc edema due to vitreous traction on the diagnosing cancer-associated retinopathy, focal visual
optic nerve head field loss due to retinopathy, acute zonal occult outer retin-
In patients with chronic unilateral optic disc edema, opathy (AZOOR), choroidal hemangioma, and sorting
the OCT can help confirm that this is the underlying out how much visual field loss is due to optic nerve versus
cause rather than other pathologies. retinal damage in cases where both are present (i.e., age-
(d) Optic disc edema with fluid in the subretinal related macular degeneration with visual field loss that
space and visual field loss may also have compressive optic neuropathy).
In some cases of acute NA-AION the optic disc edema (c) Visual evoked potentials (VEP)
may be associated with fluid in the subretinal space VEP recorded from occipital scalp electrodes in
causing a neurosensory retinal detachment which can response to a pattern stimulus may be useful in helping
track from the disc to the macula. OCT may be helpful to differentiate NA-AION from acute optic neuritis in
in quantifying the presence and amount of fluid in this young patients; in optic neuritis, the VEP latency is
location and may help explain a portion of the visual usually delayed in addition to a decrease in amplitude,
field and vision loss, which may be partly reversible but in AION usually the amplitude is reduced, but the
when the fluid resolves. latency is relatively unaffected.
(e) Optic disc edema not due to NA-AION
In a small number of patients who may have neurore-
tinitis and not NA-AION, OCT may reveal fluid in the
outer plexiform layer with very small highly reflective Blood Tests
exudates, even before a clinically apparent macular
star is seen. This may be helpful in differentiating These tests are ordered only in cases where the diagno-
patients with early neuroretinitis from AION, espe- sis of NA-AION is not definite and there is a high pre-
cially in the younger age group. test clinical likelihood that these unusual cases might
be mimicking ischemic optic neuropathy, not routinely.
These include routine complete blood count and dif-
ferential count, and serum testing for anti-neutrophil
Electrophysiologic Testing cytoplasmic antibodies (for Wegners granulomatosis),
angiotensin converting enzyme for sarcoid, fluores-
Evoked retinal potentials (multifocal ERG and cent treponemal antibody absorbed test (for syphilis),
Ganzfeld full field ERG) may have a role in two Lyme titer, Bartonella henselae immunoglobulin titers
References 161
for suspected neuro-retinitis, and genetic testing for References

Lebers hereditary optic neuropathy.
1. Hayreh SS, Joos KM, Podhajsky PA, Long CR. Systemic
diseases associated with nonarteritic anterior ischemic optic
neuropathy. Am J Ophthalmol. 1994;118(6):76680.
Procedures 2. Hayreh SS. Anterior ischemic optic neuropathy V. Optic
disc edema an early sign. Arch Ophthalmol. 1981;99:
103040.
Skin biopsy of non-sun exposed skin for autoimmune 3. Hayreh SS, Zimmerman MB. Incipient non-arteritic anterior
optic neuritis, lumbar puncture for cases of suspected ischemic optic neuropathy. Ophthalmology. 2007;114:
raised intracranial pressure and neuro-sarcoid, 176372.
4. Hayreh SS, Zimmerman MB, Beri M, Podhajsky P.
conjunctival biopsy for suspected sarcoid. These pro-
Intraocular pressure abnormalities associated with central
cedures should only be done if there is a strong clinical and hemi-central retinal vein occlusion. Ophthalmology.
suspicion of these disorders and not routinely. 2004;111:13341.

General Characteristics of Giant
Cell Arteritis 9
Giant cell arteritis (GCA) is an important cause of the be due to misdiagnosis of GCA as PMR because of the
development of anterior and also posterior ischemic common rheumatological symptoms. A study has
optic neuropathy, usually with devastating visual loss. shown that 21% of temporal artery biopsy positive GCA
It is an ophthalmic emergency. However, loss of vision patients with visual loss are free of systemic symptoms
in GCA is preventable with early diagnosis and imme- (i.e., they have occult GCA) [9], indicating that some of
diate, intensive corticosteroid therapy (see Chap. 13). the patients with systemic symptoms who may be diag-
Therefore, a brief discussion of GCA is essential in nosed as having PMR actually have GCA.
any account of ischemic optic neuropathies.
GCA is a disease of persons aged 50years and older.
In 1890 Jonathan Hutchinson [1] first reported the case
of an English man who had pain on wearing a hat Demographic Characteristics
because of pain in his temples, and he called it throm-
botic arteritis. This disease was later described by Available evidence indicates that age, gender and race
Schmidt [2] in1930. Horton etal. [3] in 1932 described play some role in the development of GCA. It is almost
it under the term temporal arteritis. It was not till 1941 universally agreed that GCA is a disease of persons
that the term giant cell arteritis was used for this dis- aged 50years and older; it is more common in women
ease [4]. Cooke etal. [5] in 1946 established that it was than men; and in Caucasians than other races.
a generalized disease. Over the years, various terms have
been used to describe this disease, including throm-
botic arteritis, temporal arteritis, cranial arteritis,
arteritis of the aged and giant cell arteritis. Age Distribution
Since polymyalgia rheumatica (PMR) is often con-
fused with GCA, a brief discussion of that is also essen- The American College of Rheumatology defined GCA
tial here. PMR is a common nonfatal disease. The as a disease of persons50years old [10]. In our study
findings show that it is relatively common in middle- [11] the youngest patient was 56 years old (median
aged and older persons, generally runs a self-limited 75.8years, mean 757.9years). Several studies have
course and has no effect on survival [6]. There are shown that the incidence of GCA increases with age
authors who are of the opinion that GCA and PMR are [1214]. For example, Hauser et al. [15] found the
frequently overlapping diseases [7], while others feel prevalence of GCA being 10.7% in age 5059, and
that PMR and GCA frequently occur together but that 55.5% in over-80 years old. The incidence of GCA
no definitive conclusions can be drawn about the nature peaks in the 7079years age group [16, 17]. Gonzalez-
of this association [8]. This may be due to the fact that Gay etal. [17] reported the mean ( SD) age at the time
the vast majority of GCA patients have rheumatological of diagnosis as 75.06.9years. Although it is so widely
symptoms similar to those seen in PMR. One of the agreed that GCA is a disease of persons aged 50years
arguments put forward for PMR and GCA frequently and older, however, there are numerous reports in the
occurring together is that some patients diagnosed as literature of infants, juveniles and adults under 50
PMR have a positive temporal artery biopsy; this may claimed to have had GCA [4, 1828]. Wagenvoort

164 9 General Characteristics of Giant Cell Arteritis
etal. [23] called their cases infantile giant-cell arteri- GCA. For example, in occult GCA, as mentioned
tis. A critical review of these under-50-years patients above, there are no systemic symptoms of GCA [9].
reveals that they do not meetall the criteria of the clas-
sical GCA syndrome. Confusion is also caused by the
entity called juvenile temporal arteritis which affects Seasonal Variation
the branches of the external carotid artery, principally
the superficial temporal artery, and is a localized and
benign condition [28]. Because of such reports, some Some studies have reported seasonal variations or a
of our referring physicians often insist that we do tem- cyclic pattern [14, 37]. On the other hand Gonzalez-
poral artery biopsy in persons younger than 50years. Gay etal. [17] in Spain observed no seasonal pattern
Of the 4 young (<56 years) patients in our study of for the diagnosis of the disease. Smeeth etal. [38] in
about 400 patients who had the temporal artery biopsy the United Kingdom found both GCA and PMR more
done, none had a positive biopsy. Features of a number commonly diagnosed in the summer months. Gonzalez-
of systemic vasculitides can mimic and overlap GCA Gay etal. [39] concluded that the presence of peaks
[29] and result in confusion and misdiagnosis. It is in the incidence and a cyclic pattern observed in some
well-known that GCA is a masquerader of many sys- studies suggests that infectious agents may play a role
temic diseases and vice versa. Not infrequently in the in the pathogenesis of both diseases. They could be the
literature, the diagnosis of GCA has been made purely triggers for the development of these conditions in
on clinical features and not confirmed by temporal genetically predisposed individuals.
artery biopsy. This can result in inclusion of conditions In our series of 116 patients with GCA in the
other than GCA in this clinical entity. Midwest of the United States, Fig.9.1 shows the distri-
bution for the various months of the year (1=January
and so on). Following was the seasonal distribution of
development of GCA:
Gender Distribution
Winter (December to February) in 27 patients
Spring (March to May) in 34 patients
This has been described differently by various studies; Summer (June to August) in 26 patients
but it is generally agreed that GCA is much more com- Fall (September to November) in 29 patients
mon in women than in men. For example, Ostberg [30]
in his postmortem study in Sweden found it signifi- A statistical analysis, using the chi-square goodness of
cantly more common in females than in males. Hauser fit test, showed no significant difference in either the
etal. [15] found that women were affected four times months (p=0.165I) or the seasons (p=0.747) for the
more than men. In Scandinavian countries, a 3:1 ratio onset of GCA.
of women to men has been reported [12, 13, 3133]. In
our study [11] the ratio between women and men was
2.65 to 1. A lower ratio of women to men was observed Epidemiology
in Israel and Southern Europe [14, 34]. Bengtsson and
Malmvall [35, 36] in their study found the PMR syn- There are several epidemiological studies on the inci-
drome more common among women with GCA (79%) dence of GCA and PMR, reported from different coun-
than among men (56%). Their group of patients tries with varying results. PMR was reported to be two
without muscular symptoms contained an equal num- to three time more common than GCA [12, 16, 38, 40].
ber of men and women. Eye complications were seen
in 15 patients (12%). In six, the ocular symptoms were
transient, while nine suffered permanent loss of vision.
In three of these patients, temporal artery biopsy Epidemiology of GCA
revealed no evidence of arteritis, and five had no clini-
cal signs of localized temporal arteritis. This may be Three of the studies on GCA are based on a survey in
due to problems with the reliability of temporal artery Olmsted County, Minnesota in the USA. Hauser etal.
biopsy and systemic symptoms for the diagnosis of [15], in the study published in 1971, found an average
Epidemiology 165
Fig.9.1 Distribution of Seasonal distribution

onset of GCA for the various 16
months of the year in my
series of 116 patients 14
(1=January and so on)
12
10
Number
8
0
1 2 3 4 5 6 7 8 9 10 11 12
Months
incidence of 2.9/100,000. The prevalence increased Reports from Scandinavian countries give a different
with age, being 10.7 in aged 5059, and 55.5 in over- incidence. Ostberg [30] in his postmortem study in
80 year olds. In the study published in 1978 [41], it Sweden found it in 2/100,000 higher than clinically
was reported that during a 25-year period an average diagnosed. A study from Finland [12] reported an
annual incidence/100,000 population aged 50 and annual incidence of GCA in a 44 month period of
older rose from 5.1 in 19501959 to 17.4 in 1970 30.4/100,000 in the whole population, and 94.4/100,000
1974. The prevalence of patients with GCA on 1 in residents aged 50 years or older. From Sweden,
January 1975 was 133/100,000 population aged 50 Bengtsson and Malmvall [35, 36], based on histologi-
and older. According to that study, GCA had no sig- cal evidence of GCA, found an annual incidence of
nificant effect on survival. In the next study, published GCA 5.5/100,000 of the total population and in the age
in 1995 by that group [37], the age- and sex-adjusted group over 50 16.8/100,000.
incidence/100,000 persons 50 years of age or older There are several studies from the Mediterranean
was 17.8, significantly higher in women than men (24 region. Gonzlez-Gay etal. [34], in a study from Spain,
versus 8). They found that age-specific incidence rates found the average annual incidence for the population
increased with age (P<0.0001). This study showed a aged 50 and older was 10.24/100,000 (men
regular cyclic pattern in incidence rates over time 11.00/100,000, women 9.57/100,000), and there was
period, and concluded that that supported the hypoth- a progressive increase in the incidence in men an
esis of an infectious cause for GCA. In the study pub- annual increase of 8% and in women 11%. This group,
lished in 2004 [42], they found that in persons 50years in a subsequent study [17] of biopsy-proven GCA,
of age or older there was a significant (P=0.017) profound an age- and sex-adjusted annual incidence rate
gressive increase in GCA from 1950 to 1979, with an of 10.13/100,000 populations aged 50years and older.
overall age-and sex-adjusted incidence/100,000 per- The annual incidence rate for women was slightly
sons of 18.8 the incidence was higher in women than higher than for men (10.23 versus 9.92). The annual
men (24.4 versus 10.3); however, there was no change incidence rate increased with advancing age up to a
over a period of 20years after that, and the overall sur- maximum of 23. Sixteen in the 7079year age-group.
vival in GCA remained similar. In this study, they They also observed a progressive increase in the inci-
again found the cyclic pattern of annual incidence rates dence from 1981 through 2000 (p=0.001). Salvarani
over a 50-year period. etal. [43] found in their Italian population a average
annual incidence rate of GCA of 8.8/100,000 popula- women and 83.2 in men. Hamrins data [45] from
tion aged 50years or older. In a subsequent study [16] Sweden suggested an annual incidence rate of about
this group found the average annual incidence of GCA 12.0/100,000 population over a period of 7.5 years.
was 6.9/100,000 in a population aged 50 years or Like GCA, the reported incidence varies widely from
older. country to country.
A British study [38] found an age adjusted inci-
dence rate of GCA 2.2/10,000 person-years, with no
increase observed. It was more common in the south
than in the north of the country, which may be partly pidemiology of Combined Population
E
attributable to a risk factor which is more prevalent in of GCA and PMR
the south and east of the United Kingdom.
Thus the reported incidence is highly variable, not
There are some studies which have reported the inci-
only from different countries but also from the same
dence of a combined population of GCA and PMR,
region at different times. This may be due to different
instead of reporting separately, as above. A study from
criteria used for diagnosis of GCA. This makes it hard
Norway [44] reported an annual incidence of PMR/
to reach definite conclusions about the epidemiology
GCA of 141.7/100,000 populations aged 50 years or
of GCA.
more more in women than men (177.6 versus 99.5).
A British study [46], from private practice, found that
the annual incidence of PMR/GCA in those aged 65
Epidemiology of PMR and over was about 4/1,000.
The incidence of PMR, like GCA, has been reported

from various countries. A survey study [6], in Olmsted
County, Minnesota, USA, found that the average Racial Differences
annual incidence of PMR increased during a 10-year
period (19701979) from 19.8/100,000 in persons There is a marked racial predilection for development
5059years of age, to a maximum of 112.2/100,000 of GCA. It is far more common among Caucasians
in persons 7079 years of age. A subsequent study than other races. Even among Caucasians, it differs
[40] found that among cases of PMR seen between between different ethnic groups, as is evident from the
1970 and 1991 the average annual age-and sex- above epidemiological review, and so does the inci-
adjusted incidence of PMR/100,000 population in the dence of GCA and PMR. On the whole, it is higher in
age group of 50years was 52.5, with a significantly persons of Scandinavian origin (the population of
higher incidence in females than in males (61.7 versus Olmsted County, Minnesota has a high proportion of
39.9). They concluded that the incidence and clinical persons of Scandinavian origin) than in the southern
manifestations have varied but remained relatively European population, which was also evident in the
stable over the 2 decades. An Italian study [43] found British study. It is very uncommon among non-Cauca-
average annual incidence rates of PMR of 12.8/100,000 sians. For example, only a rare case of GCA has been
populations aged 50 years or older. In a subsequent reported among persons from China [47, 48], India
study [16] this group found the average annual inci- [4951] and Thailand [52]. A nationwide survey in
dence of PMR 12.7/100,000 in a population aged Japan for GCA disclosed that its prevalence was
50years or older. A British study [38] found the age extremely low compared with other countries; more-
adjusted incidence rate of PMR was 8.4/10,000 per- over, permanent visual loss, jaw claudication, and
son-years. Again it was more common in the south PMR were infrequent among Japanese individuals
than in the north of the country, and that may be partly with GCA [53]. In Israel GCA was found three times
attributable to risk factors from the south and east to more often in Ashkenazi (originally from Western and
the north of the United Kingdom. In a study from Central Europe, mainly from Germany) than in
Norway [44] annual incidence of pure PMR/100,000 Sephardic Jews (from Spain and Portugal) and only
residents aged 50years or more was 112.6, 137.7 in rarely in Arabs [54]. A lower incidence of GCA was
Pathophysiology 167
reported in black persons from Tennessee in the USA mediated disease. Immunological studies indicate that
[55]. In a study from Los Angeles, California, positive GCA and PMR have a polygenic basis. The current
temporal artery biopsy was seen in 19 of 66 whites, concept of the pathophysiology of GCA is based
1 of 9 Asians, none of 40 Hispanics (Mexican) and mainly on the extensive immunological studies by
none of 6 African Americans [56]. In that study [56], Weyand and colleagues [6874]. The characteristic
the incidence of biopsy-proven GCA in a retrospective feature in this disease is the formation of granuloma in
study of 121 consecutive patients who underwent tem- the wall of the medium and large arteries. In GCA,
poral artery biopsy was as follows: among 66 white according to Weyand etal. [73], in the affected arteries
patients 29% had positive biopsy, whereas only 11% the adventitia is the primary site of immunological
of the 9 Asian patients, none of the 40 Hispanic and injury. They stated that The typical lesions, granulo-
none of the 6 African American patients had a positive mas in the vessel wall, are formed by IFN-gamma-
biopsy. Johnson and Arnold [57], among 12 cases of producing CD4+ T cells and macrophages. CD4+ T
arteritic anterior ischemic optic neuropathy due to cells undergo in situ activation in the adventitia, where
GCA had 11 white, one Hispanic and no African they interact with indigenous dendritic cells. Tissue
American or Asians. Tomsak [58] found positive tem- injury is mediated by several distinct sets of mac-
poral artery biopsy for GCA in 17% of Caucasians, 7% rophages that are committed to diverse effector func-
of African Americans but none in Hispanics. Mader tions. The dominant tissue injury in the media results
et al. [59] reported a very low incidence of biopsy- from oxidative stress and leads to smooth muscle cell
proven GCA among Alaska Natives. In our study [11] apoptosis and nitration of endothelial cells.
of 106 patients with positive temporal artery biopsy, Macrophage-derived growth factors are instrumental
we had only one African American patient but that in driving the response-to-injury program of the artery
may partly be due to the predominantly white popula- that causes intimal hyperplasia and vessel occlusion.
tion in our part of the U.S.A. In North America, it is That may be the reason that steroid therapy may be
more common among persons of Scandinavian origin effective in GCA. According to Martinez-Taboada
[42, 60]. etal. [75], PMR and GCA are characterized by a pro-
The incidence of PMR has also been found to be duction of more IL-6, and corticosteroid therapy is
higher in individuals of Scandinavian background [31, followed by significant decrease of IL-6 levels.
40, 61, 62]. Salvarani et al. [16] determined frequencies of HLA
These racial differences in the development of GCA antigens. When they compared that with HLA findings
suggest a genetic predisposition. Many studies have in healthy controls, DR4 was not found to be signifi-
shown the implication of genetic variants in key com- cantly associated with PMR (24% in PMR patients
ponents of immune and inflammatory pathways in versus 14% in controls). Patients with GCA also
GCA and PMR susceptibility [63, 64], because GCA showed an increased frequency of DR4 compared with
and PMR are polygenic diseases. An increased risk of controls (36% versus 14%), but this difference was
visual ischemic complications in biopsy-proven GCA also not statistically significant. They discussed immu-
patients and an increased frequency of relapses in nogenetic features of PMR and GCA and the relation-
patients with PMR were found to be associated with ship between the immunogenetic and epidemiologic
the carriage of HLADRB1*04 alleles [65, 66]. patterns in different populations.
Moreover, a functional variant of the vascular endothe- In PMR there is mild synovitis of the shoulder.
lial growth factor gene was associated with severe According to Meliconi etal. [76] Synovitis was pres-
ischemic complications in patients with GCA [67]. ent in 10 of 12 (83%) untreated patients and in only 2
of 7 (29%) treated patients. The synovitis was charac-
terized by vascular proliferation and leukocyte infiltra-
tion. Infiltrating cells consisted predominantly of
Pathophysiology macrophages and T Lymphocytes. Almost all T lym-
phocytes were CD45RO positive. A few neutrophils,
GCA is a systemic vasculitis preferentially involving but no B cells, natural killer cells, or gamma/delta T
medium and large-size arteries. The pathophysiology cells were found. Intense expression of HLA class II
of GCA is not yet fully understood. It is an immune antigens was found in the lining layer cells as well as
in macrophages and lymphocytes. DR, but not DP or effects. It is frequently stated that systemic symptoms
DQ, was expressed by the endothelium of a few ves- rather than erythrocyte sedimentation rate (ESR) and
sels. Class II antigen expression correlated with the C-reactive protein (CRP), should be the guide for
number of macrophages and lymphocytes. They con- tapering down or stopping steroid therapy; this can be
cluded that their findings supported the hypothesis that a dangerous approach, because these patients may lose
synovitis is a major cause of the musculoskeletal vision without any warning systemic or ocular symp-
symptoms of PMR. toms [89, 90] (see Chap. 13). Moreover, 21.2% of the
patients with visual loss have the occult GCA [9], i.e.,
no systemic symptoms at all (see below). Thus, from
the point of view of prevention of visual loss, one
Is GCA a Self-limited Disease? should not consider GCA a self-limited disease.
From the management point of view and for the pre-

vention of visual loss, this is an important issue. There
is controversy about the duration of GCA. The self- Recurrence/Relapses of GCA
limiting nature of GCA is stressed in text-books and by
many physicians (particularly rheumatologists) who In the literature there are many reports of recurrence
feel strongly that there is no need for these patients to or relapses of GCA. My studies of GCA patients
take prolonged steroid therapy, because of its side- have shown that those represent inadequate corticos-
effects. For example, Huston et al. [41], stated that teroid treatment or stopping the therapy too soon,
GCA lasts for less than a year. Similarly, other rheu- under the commonly held belief among rheumatolo-
matologists believe that stopping steroid therapy after gists that GCA is (1) a self-limited disease and stop-
2 years is reasonable [7782]. However, others have ping corticosteroid therapy after 2years is reasonable
shown that that is not true, because evidence of active [77] and (2) that systemic symptoms are the guide to
disease has been seen even after many years [8385]. duration of the therapy. In my study of about 300 GCA
For example, repeat temporal artery biopsy has shown patients since 1969, with and without visual loss, none
evidence of active disease even after 9years of steroid suffered a recurrence, because the guiding principle in
therapy [8488]. Cullen [83, 84] in his serial biopsies my study has always been to control corticosteroid
in GCA patients showed evidence of active disease in therapy by the levels of ESR and CRP- and NOT by
the temporal arteries 3, 6, 12, 21 and 26months and systemic symptoms [89, 90].
7years later. Cullen [84] and Cohen [85] reported that
ocular complications can still occur as long as 7years
after onset. This contradicts the claims by others that
GCA is a self-limiting disease after a variable time Occult GCA
period. Clinically, as discussed in a Chap. 13, many
reports have confirmed that GCA patients require ste- It is a widely held belief among physicians and oph-
roid therapy for much longer than 2 years, and even thalmologists that GCA patients always present with
may need it for the rest of their life as in many other systemic symptoms and signs, and that the absence of
systemic rheumatologic diseases. such symptoms and signs rules out GCA. This impres-
Our study [89] has shown that almost all GCA sion has sometimes, sadly, caused misdiagnosis and
patients require life long steroid therapy to prevent consequent blindness. Development of visual symp-
visual loss. I have had patients who went blind after toms as the first or the only sign of GCA was first
their local physician had stopped steroid therapy after reported in 1952 [91, 92]. Since then a number of com-
a year or two, influenced by the rheumatology litera- munications, mostly anecdotal, have reported the same
ture [41, 7782]. In this connection, it is pertinent to [35, 49, 93110].
point out that fear of visual loss is exceeded only by In 1962 Simmons and Cogan [98] called this type
the fear of death; and patients will always opt for a of GCA occult temporal arteritis. Occult (or silent
mode of therapy that prevents them from going blind, or atypical) GCA is now a well-established entity;
even when they are fully informed of the possible side- however, a certain amount of confusion about it still
References 169
persists for a number of reasons, including the 300 GCA patients with and without ocular involve-
following: ment, I have found that patients will retain their existing
vision if they are treated immediately and adequately
1. Several studies [103, 106, 108, 110, 111] reported
with high-dose systemic corticosteroid therapy, pro-
in the literature have defined occult GCA in dif-
vided they do not suffer any further visual loss within
ferent ways. Many reports include patients with a
the first week after starting the corticosteroid therapy
number of systemic symptoms of GCA.
[90, 113]. Thus, blindness due to GCA is preventable
2. The reported incidence of occult GCA varies from
in the vast majority of cases, if diagnosed early and
5% to 38% [35, 78, 106, 108, 109, 112]. This wide
treated properly. Therein lies the importance of recog-
variation may be due to the fact that most of the
nizing the occult GCA cases early.
studies include some cases where the diagnosis of
Therefore, any patient over the age of 55 who
GCA was based purely on the clinical judgment of
presents with a history of amaurosis fugax, diplopia
the physicians and never confirmed on temporal
or visual loss, and clinical findings of arteritic AION,
artery biopsy.
central retinal artery occlusion, cilioretinal artery
3. Until recently, there was no study in the literature
occlusion or posterior ischemic optic neuropathy
where a large series of patients have been studied
should be suspected of having GCA and should have
prospectively using the strict definition of occult
ESR and CRP done on an emergency basis, irrespec-
GCA, i.e., where ocular disorders were the sole pre-
tive of whether he/she shows systemic symptoms of
senting symptoms and signs of GCA in otherwise
GCA. If, based on this information, there is a high
perfectly healthy individuals, and the diagnosis was
index of suspicion, the patient should be started
confirmed on temporal artery biopsy in all.
immediately on high doses of systemic corticosteroid,
All these limitations make it impossible to evaluate and a temporal artery biopsy done to confirm or rule
from the previous literature the true incidence of out GCA. It is well-established now that early diagno-
occult GCA among patients who first present with sis of GCA and early and aggressive treatment are
the visual symptoms and signs of GCA. We [9] inves- the keys to preventing blindness.
tigated the incidence of occult GCA in a prospective
study of 85 patients who fulfilled the following inclu-
sion criteria: (a) They all had GCA confirmed on tem-
poral artery biopsy, (b) all had ocular involvement and References
(c) they had no symptoms and signs of GCA whatso-
ever. In that study, 21.2% fulfilled all the criteria for
1. Hutchinson J. Diseases of the arteries. I on a peculiar form of
occult GCA, and in the majority of them elevated ESR thrombotic arteritis of the aged which is sometimes
and CRP and/or the presence of arteritic AION were productive of gangrene. Arch Surg (Lond). 18891890;1:
the main signs of GCA. Because they have no systemic 3239.
2. Schmidt M. Intracranial aneurysms. Brain. 1930;53:
symptoms, such patients do not go to rheumatologists 489540.
and would not be included in a rheumatologic study. 3. Horton BJ, Magath TB, Brown GE. An unusual form of arteri-
Thus, if systemic manifestations of GCA are the pri- tis of the temporal vessels. Mayo Clin Proc. 1932;7:7001.
mary and sole criterion for diagnosing GCA, 21.2% of 4. Gilmour JR. Giant cell chronic arteritis. J Pathol Bacteriol.
1941;53(2):26377.
GCA patients with visual loss are going to be missed. 5. Cooke WT, Cloake PCP, Govan ADT, Colbeck JC. Temporal
Since occult GCA is a potential cause of blindness, arteritis: a generalized vascular disease. Q J Med.
that is an unacceptably high number to be missed. 1946;15(1):4775.
Therefore, the absence of systemic symptoms and 6. Chuang TY, Hunder GG, Ilstrup DM, Kurland LT.
Polymyalgia rheumatica: a 10-year epidemiologic and clini-
signs does not rule out GCA an extremely important cal study. Ann Intern Med. 1982;97(5):67280.
point to be borne in mind. If the physician considers 7. Gonzalez-Gay MA. Giant cell arteritis and polymyalgia
systemic symptoms and signs of GCA an essential part rheumatica: two different but often overlapping conditions.
of the syndrome, there is a high risk that he/she will Semin Arthritis Rheum. 2004;33(5):28993.
8. Cantini F, Niccoli L, Storri L, Nannini C, Olivieri I, Padula
miss the disease in those GCA patients whose unilat- A, etal. Are polymyalgia rheumatica and giant cell arteritis
eral or bilateral ocular involvement is the sole reason the same disease? Semin Arthritis Rheum. 2004;33(5):
for initial consultation. In my experience with about 294301.
9. Hayreh SS, Podhajsky PA, Zimmerman B. Occult giant cell 27. Fernandez-Herlihy L. Temporal arteritis: clinical aids to
arteritis: ocular manifestations. Am J Ophthalmol. 1998; diagnosis. J Rheumatol. 1988;15(12):1797801.
125(4):5216. 28. Tomlinson FH, Lie JT, Nienhuis BJ, Konzen KM, Groover
10. Hunder GG, Bloch DA, Michel BA, Stevens MB, Arend RV. Juvenile temporal arteritis revisited. Mayo Clin Proc.
WP, Calabrese LH, et al. The American College of 1994;69(5):4457.
Rheumatology 1990 criteria for the classification of giant 29. Vidal E, Liozon F, Rogues AM, Cransac M, Berdha JF,
cell arteritis. Arthritis Rheum. 1990;33(8):11228. Liozon E. Concurrent temporal arteritis and Churg-Strauss
11. Hayreh SS, Podhajsky PA, Raman R, Zimmerman B. Giant syndrome. J Rheumatol. 1992;19(8):13124.
cell arteritis: validity and reliability of various diagnostic 30. Ostberg G. Temporal arteritis in a large necropsy series. Ann
criteria. Am J Ophthalmol. 1997;123(3):28596. Rheum Dis. 1971;30(3):22435.
12. Franzn P, Sutinen S, von Knorring J. Giant cell arteritis and 31. Boesen P, Sorensen SF. Giant cell arteritis, temporal arteri-
polymyalgia rheumatica in a region of Finland: an epidemio- tis, and polymyalgia rheumatica in a Danish county. A pro-
logic, clinical and pathologic study, 19841988. J Rheumatol. spective investigation, 19821985. Arthritis Rheum.
1992;19(2):2736. 1987;30(3):2949.
13. Petursdottir V, Johansson H, Nordborg E, Nordborg C. The 32. Baldursson O, Steinsson K, Bjornsson J, Lie JT. Giant cell
epidemiology of biopsy-positive giant cell arteritis: special arteritis in Iceland. An epidemiologic and histopathologic
reference to cyclic fluctuations. Rheumatology (Oxford). analysis. Arthritis Rheum. 1994;37(7):100712.
1999;38(12):120812. 33. Haugeberg G, Paulsen PQ, Bie RB. Temporal arteritis in
14. Bas-Lando M, Breuer GS, Berkun Y, Mates M, Sonnenblick Vest Agder County in southern Norway: incidence and clini-
M, Nesher G. The incidence of giant cell arteritis in cal findings. J Rheumatol. 2000;27(11):26247.
Jerusalem over a 25-year period: annual and seasonal 34. Gonzalez-Gay MA, Garcia-Porrua C, Rivas MJ, Rodriguez-
fluctuations. Clin Exp Rheumatol. 2007;25(1 suppl 44): Ledo P, Llorca J. Epidemiology of biopsy proven giant cell
S157. arteritis in northwestern Spain: trend over an 18year period.
15. Hauser WA, Ferguson RH, Holley KE, Kurland LT. Temporal Ann Rheum Dis. 2001;60(4):36771.
arteritis in Rochester, Minnesota, 1951 to 1967. Mayo Clin 35. Bengtsson BA, Malmvall BE. The epidemiology of giant
Proc. 1971;46(9):597602. cell arteritis including temporal arteritis and polymyalgia
16. Salvarani C, Macchioni P, Zizzi F, Mantovani W, Rossi F, rheumatica. Incidences of different clinical presentations
Castri C, etal. Epidemiologic and immunogenetic aspects of and eye complications. Arthritis Rheum. 1981;24(7):
polymyalgia rheumatica and giant cell arteritis in northern 899904.
Italy. Arthritis Rheum. 1991;34(3):3516. 36. Bengtsson BA, Malmvall BE. Giant cell arteritis. Acta Med
17. Gonzalez-Gay MA, Miranda-Filloy JA, Lopez-Diaz MJ, Scand Suppl. 1982;658:1102.
Perez-Alvarez R, Gonzalez-Juanatey C, Sanchez-Andrade 37. Salvarani C, Gabriel SE, OFallon WM, Hunder GG. The
A, etal. Giant cell arteritis in northwestern Spain: a 25-year incidence of giant cell arteritis in Olmsted County,
epidemiologic study. Medicine (Baltimore). 2007;86(2): Minnesota: apparent fluctuations in a cyclic pattern. Ann
618. Intern Med. 1995;123(3):1924.
18. Roberts AM, Askey JM. Temporal arteritis; relief of head- 38. Smeeth L, Cook C, Hall AJ. Incidence of diagnosed poly-
ache by injection of procaine hydrochloride. J Am Med myalgia rheumatica and temporal arteritis in the United
Assoc. 1948;137(8):697. Kingdom, 19902001. Ann Rheum Dis. 2006;65(8):
19. Meyers L, Lord Jr JW. Cranial arteritis; report of its occur- 10938.
rence in a young woman. J Am Med Assoc. 1948;136(3): 39. Gonzalez-Gay MA, Vazquez-Rodriguez TR, Lopez-Diaz
16971. MJ, Miranda-Filloy JA, Gonzalez-Juanatey C, Martin J,
20. Gelfand M. Giant cell arteritis with aneurysmal formation in et al. Epidemiology of giant cell arteritis and polymyalgia
an infant. Br Heart J. 1955;17(2):2646. rheumatica. Arthritis Rheum. 2009;61(10):145461.
21. Hollenhorst RW, Brown JR, Wagener HP, Shick RM. 40. Salvarani C, Gabriel SE, OFallon WM, Hunder GG.
Neurologic aspects of temporal arteritis. Neurology. Epidemiology of polymyalgia rheumatica in Olmsted
1960;10:4908. County, Minnesota, 19701991. Arthritis Rheum.
22. Bugg Jr EI, Coonrad RW, Grim KB. Giant-cell arteritis: an 1995;38(3):36973.
acute hand syndrome. J Bone Joint Surg Am. 41. Huston KA, Hunder GG, Lie JT, Kennedy RH, Elveback
1963;45:126972. LR. Temporal arteritis: a 25-year epidemiologic, clinical,
23. Wagenvoort CA, Harris LE, Brown Jr AL, Veeneklaas GM. and pathologic study. Ann Intern Med. 1978;88(2):
Giant-cell arteritis with aneurysm formation in children. 1627.
Pediatrics. 1963;32:8617. 42. Salvarani C, Crowson CS, OFallon WM, Hunder GG,
24. Bethlenfalvay NC, Nusynowitz ML. Temporal arteritis: a Gabriel SE. Reappraisal of the epidemiology of giant cell
rarity in the young adult. Arch Intern Med. 1964;114: arteritis in Olmsted County, Minnesota, over a fifty-year
4879. period. Arthritis Rheum. 2004;51(2):2648.
25. Kent DC, Arnold H. Aneurysm of the aorta due to giant cell 43. Salvarani C, Macchioni PL, Tartoni PL, Rossi F, Baricchi R,
aortitis. Successful surgical correction. J Thorac Cardiovasc Castri C, etal. Polymyalgia rheumatica and giant cell arteri-
Surg. 1967;53(4):5727. tis: a 5-year epidemiologic and clinical study in Reggio
26. Vilaseca J, Gonzalez A, Cid MC, Lopez-Vivancos J, Ortega Emilia, Italy. Clin Exp Rheumatol. 1987;5(3):20515.
A. Clinical usefulness of temporal artery biopsy. Ann Rheum 44. Gran JT, Myklebust G. The incidence of polymyalgia rheu-
Dis. 1987;46(4):2825. matica and temporal arteritis in the county of Aust Agder,
References 171
south Norway: a prospective study 198794. J Rheumatol. 64. Gonzlez-Gay MA, Amoli MM, Garcia-Porrua C, Ollier
1997;24(9):173943. WE. Genetic markers of disease susceptibility and severity
45. Hamrin B. Polymyalgia arteritica. Acta Med Scand Suppl. in giant cell arteritis and polymyalgia rheumatica. Semin
1972;533:1131. Arthritis Rheum. 2003;33(1):3848.
46. Kyle V, Silverman B, Silman A, King H, Oswald N, Reiss B, 65. Gonzlez-Gay MA, Garcia-Porra C, Vzquez-Caruncho
et al. Polymyalgia rheumatica/giant cell arteritis in a M, Dababneh A, Hajeer A, Ollier WE. The spectrum of
Cambridge general practice. Br Med J (Clin Res Ed). polymyalgia rheumatica in northwestern Spain: incidence
1985;291(6492):3857. and analysis of variables associated with relapse in a 10year
47. Kwok AK, Lam DS, Liew CT. Bilateral arteritic central reti- study. J Rheumatol. 1999;26(6):132632.
nal artery occlusion in a Chinese patient. Aust N Z J 66. Gonzlez-Gay MA, Garcia-Porra C, Llorca J, Hajeer AH,
Ophthalmol. 1998;26(2):1756. Braas F, Dababneh A, etal. Visual manifestations of giant
48. Hu Z, Yang Q, Zeng S, Li J, Wu X, Cao L, etal. Giant cell cell arteritis. Trends and clinical spectrum in 161 patients.
arteritis in China: a prospective investigation. Angiology. Medicine (Baltimore). 2000;79(5):28392.
2002;53(4):45763. 67. Rueda B, Lopez-Nevot MA, Lopez-Diaz MJ, Garcia-Porrua
49. Mohan K, Gupta A, Jain IS, Banerjee CK. Bilateral central C, Martin J, Gonzalez-Gay MA. A functional variant of vas-
retinal artery occlusion in occult temporal arteritis. J Clin cular endothelial growth factor is associated with severe
Neuroophthalmol. 1989;9(4):2702. ischemic complications in giant cell arteritis. J Rheumatol.
50. Desai MC, Vas CJ. Temporal arteritis. The Indian scene. 2005;32(9):173741.
J Assoc Physicians India. 1989;37(9):60911. 68. Weyand CM, Hicok KC, Hunder GG, Goronzy JJ. Tissue
51. Sood R, Zulfi H, Ray R, Handa R, Wali JP. Giant cell arteri- cytokine patterns in patients with polymyalgia rheumatica
tis a rare cause of fever of unknown origin in India. J Assoc and giant cell arteritis. Ann Intern Med. 1994;121(7):
Physicians India. 2002;50:8468. 48491.
52. Phanthumchinda K. Giant cell arteritis. J Med Assoc Thai. 69. Weyand CM, Goronzy JJ. Arterial wall injury in giant cell
1990;73(6):3537. arteritis. Arthritis Rheum. 1999;42(5):84453.
53. Kobayashi S, Yano T, Matsumoto Y, Numano F, Nakajima 70. Weyand CM, Goronzy JJ. Pathogenic mechanisms in giant
N, Yasuda K, etal. Clinical and epidemiologic analysis of cell arteritis. Cleve Clin J Med. 2002;69 suppl 2:S112832.
giant cell (temporal) arteritis from a nationwide survey in 71. Weyand CM, Goronzy JJ. Medium-and large-vessel vasculi-
1998 in Japan: the first government-supported nationwide tis. N Engl J Med. 2003;349(2):1609.
survey. Arthritis Rheum. 2003;49(4):5948. 72. Weyand CM, Goronzy JJ. Giant-cell arteritis and polymyal-
54. Fainaru M, Friedman G, Friedman B. Temporal arteritis in gia rheumatica. Ann Intern Med. 2003;139(6):50515.
Israel. A review of 47 patients. J Rheumatol. 1979;6((3):3305. 73. Weyand CM, Ma-Krupa W, Goronzy JJ. Immunopathways
55. Smith CA, Fidler WJ, Pinals RS. The epidemiology of giant in giant cell arteritis and polymyalgia rheumatica.
cell arteritis. Report of a ten-year study in Shelby County, Autoimmun Rev. 2004;3(1):4653.
Tennessee. Arthritis Rheum. 1983;26(10):12149. 74. Weyand CM, Ma-Krupa W, Pryshchep O, Groschel S,
56. Liu NH, LaBree LD, Feldon SE, Rao NA. The epidemiology Bernardino R, Goronzy JJ. Vascular dendritic cells in giant
of giant cell arteritis: a 12-year retrospective study. cell arteritis. Ann NY Acad Sci. 2005;1062:195208.
Ophthalmology. 2001;108(6):11459. 75. Martinez-Taboada VM, Alvarez L, RuizSoto M, Marin-
57. Johnson LN, Arnold AC. Incidence of nonarteritic and arter- Vidalled MJ, Lopez-Hoyos M. Giant cell arteritis and poly-
itic anterior ischemic optic neuropathy. Population-based myalgia rheumatica: role of cytokines in the pathogenesis
study in the state of Missouri and Los Angeles County, and implications for treatment. Cytokine. 2008;44(2):
California. J Neuroophthalmol. 1994;14(1)):3844. 20720.
58. Tomsak RL. Giant cell arteritis. Ophthalmology. 76. Meliconi R, Pulsatelli L, Uguccioni M, Salvarani C,
2002;109(2):21920. Macchioni P, Melchiorri C, et al. Leukocyte infiltration in
59. Mader TH, Werner RP, Chamberlain DG, Doornbos D. synovial tissue from the shoulder of patients with polymyal-
Giant cell arteritis in Alaska Natives. Can J Ophthalmol. gia rheumatica. Quantitative analysis and influence of corti-
2009;44(1):536. costeroid treatment. Arthritis Rheum. 1996;39(7):
60. Machado EB, Michet CJ, Ballard DJ, Hunder GG, Beard 1199207.
CM, Chu CP, etal. Trends in incidence and clinical presenta- 77. Kyle V. Treatment of polymyalgia rheumatica/giant cell
tion of temporal arteritis in Olmsted County, Minnesota, arteritis. Baillires Clin Rheumatol. 1991;5(3):48591.
19501985. Arthritis Rheum. 1988;31(6):7459. 78. Fauchald P, Rygvold O, Oystese B. Temporal arteritis and
61. Schaufelberger C, Bengtsson BA, Andersson R. polymyalgia rheumatica. Clinical and biopsy findings. Ann
Epidemiology and mortality in 220 patients with polymyalgia Intern Med. 1972;77(6):84552.
rheumatica. Br J Rheumatol. 1995;34(3):2614. 79. Kyle V, Hazleman BL. Stopping steroids in polymyalgia
62. Elling P, Olsson AT, Elling H. Synchronous variations of the rheumatica and giant cell arteritis. Br Med J.
incidence of temporal arteritis and polymyalgia rheumatica 1990;300:3445.
in different regions of Denmark; association with epidemics 80. Myles AB. Prognosis of polymyalgia rheumatica and giant
of Mycoplasma pneumoniae infection. J Rheumatol. cell arteritis. Ballieres Clin Rheumatol. 1991;5:
1996;23(1):1129. 493503.
63. Gonzalez-Gay MA. Genetic epidemiology. Giant cell arteri- 81. Salvarani C, Cantini F, Bolardi L, Hunder GG. Polymyalgia
tis and polymyalgia rheumatica. Arthritis Res. rheumatica and giant-cell arteritis. N Engl J Med.
2001;3(3):1547. 2002;347:26171.
82. Salvarani C, Cantini F, Hunder GG. Polymyalgia rheumatica 99. Cullen JF. Occult temporal arteritis. Trans Ophthalmol Soc
and giant-cell arteritis. Lancet. 2008;372:23445. UK. 1963;83:72536.
83. Cullen JF. Ischaemic opticneuropathy temporal arteritis. 100. Moro F. Pseudo-papillite vascolare in arterite temporale
Mackenzie symposium on the ocular circulation in health occulta. Atti Soc Oftal Lomb. 1963;18:7480.
and disease. London: 1968. p. 1426. 101. Russell RW, Earl CJ. Loss of vision from cranial arteritis
84. Cullen JF. Temporal arteritis: occurrence of ocular compli- without other symptoms. Br J Ophthalmol. 1964;48:
cations 7 years after diagnosis. Br J Ophthalmol. 61921.
1972;56:5848. 102. Friedman JJ. Occult temporal arteritis. Am J Ophthalmol.
85. Cohn DN. Temporal arteritis improvement in visual 1965;60:3335.
prognosis and management with repeated biopsies. Trans 103. Cullen JF. Occult temporal arteritis. A common cause of
Am Acad Ophthalmol Otolaryngol. 1973;77:7485. blindness in old age. Br J Ophthalmol. 1967;51(8):
86. Blumberg S, Giansiracusa DF, Docken WP, Kantrowitz FG. 51325.
Recurrence of temporal arteritis. Clinical recurrence nine 104. Gutrecht JA. Occult temporal arteritis. JAMA. 1970;213(7):
years after initial illness. JAMA. 1980;244((15):1713. 11889.
87. Graham E, Holland A, Avery A, Russell RW. Prognosis in 105. Hayreh SS. Anterior ischemic optic neuropathy. New York:
giant-cell arteritis. Br Med J Clin Res Ed. 1981; Springer; 1975. p. 115.
282(6260):26971. 106. Healey LA, Wilske KR. Presentation of occult giant cell
88. Gouet D, Marchaud R, Le Berre D, Alcalay M, Becq- arteritis. Arthritis Rheum. 1980;23(6):6413.
Giraudon B, Boissonnot L, et al. Prognosis of treated 107. Radda TM, Binder S. Larvierte arteriitis temporalis. [Occult
temporal arteritis. Retrospective study of 87 cases. Presse temporal arteritis a case report]. Wien Klin Wochenschr.
Md. 1986;15(13):6036. 1980;92(8):2935.
89. Hayreh SS, Zimmerman B. Management of giant cell arteri- 108. Desmet GD, Knockaert DC, Bobbaers HJ. Temporal arteri-
tis. Our 27-year clinical study: new light on old controver- tis: the silent presentation and delay in diagnosis. J Intern
sies. Ophthalmologica. 2003;217(4):23959. Med. 1990;227(4):23740.
90. Hayreh SS. Steroid therapy for visual loss in patients with 109. Varin J, Guillotde Suduiraut C, Muffat-Joly M, Christoforov
giant-cell arteritis. Lancet. 2000;355:15723. B, Portos JL, Morin Y. Aspects cliniques et epidemi-
91. Parsons-Smith G. Ophthalmic manifestations of temporal ologiques de la maladie de Horton selon le milieu de
arteritis. Br J Ophthalmol. 1952;36(11):61525. recrutement, ophtalmologique ou de medecine interne.
92. Siegert P. Akute Ischamie der Papille bei Arteriitis cranialis [Clinical and epidemiological aspects of Horton disease
(temporalis). [Acute ischemia of the papilla caused by based on survey methods in ophthalmology or internal
arteritis temporalis]. Klin Monbl Augenheilkd Augenarztl medicine]. Ann Med Interne (Paris). 1994;145:
Fortbild. 1952;120(3):25473. 398404.
93. Kreibig W. Optikomalazie, die Folge eines Gefassverschlusses 110. Healey LA, Wilske KR. Anemia as a presenting manifesta-
im retrobulbren Abschnitt des Sehnerven. [Opticomalacia tion of giant cell arteritis. Arthritis Rheum.
caused by vascular occlusion in the retrobulbar portion of 1971;14(1):2731.
optic nerves.]. Klin Monbl Augenheilkd Augenarztl Fortbild. 111. Calamia KT, Hunder GG. Giant cell arteritis (temporal
1953;122(6):71931. arteritis) presenting as fever of undetermined origin.
94. Birkhead NC, Wagener HP, Shick RM. Treatment of tempo- Arthritis Rheum. 1981;24(11):14148.
ral arteritis with adrenal corticosteroids; results in fifty-five 112. Malmvall BE, Bengtsson BA. Giant cell arteritis. Clinical
cases in which lesion was proved at biopsy. J Am Med features and involvement of different organs. Scand
Assoc. 1957;163(10):8217. J Rheumatol. 1978;7(3):1548.
95. Wagener HP, Hollenhorst RW. The ocular lesions of tempo- 113. Hayreh SS, Zimmerman B. Visual deterioration in giant
ral arteritis. Am J Ophthalmol. 1958;45(5):61730. cell arteritis patients while on high doses of corticosteroid
96. Palm E. The ocular crisis of the temporal arteritis syn- therapy. Ophthalmology. 2003;110(6):120415.
drome (Horton). Acta Ophthalmol (Copenh). 1958;36(2):
20843.
97. Parsons-Smith G. Sudden blindness in cranial arteritis. Br J
Ophthalmol. 1959;43(4):20416. For a complete bibliography and detailed review of previously
98. Simmons RJ, Cogan DG. Occult temporal arteritis. Arch published studies in the literature on the subject, please refer to
Ophthalmol. 1962;68:818. bibliography in previous publications listed here.
Clinical Features of Giant Cell Arteritis
10
Giant cell arteritis (GCA) shares many clinical features synovitis, distal extremity pain, normal erythrocyte
with polymyalgia rheumatica (PMR). Therefore, for a sedimentation rate, and mild weakness. They pointed
correct diagnosis of GCA, it is essential is to know out that while concurrent asymptomatic GCA is com-
clinical features of both PMR and GCA. mon, there are no data to support obtaining a temporal
artery biopsy in patients with pure PMR symptoms.
Bird etal. [6] put forward the seven most valuable cri-
teria to standardize a diagnosis of PMR: bilateral shoul-
ystemic Clinical Features
S
der pain or stiffness, onset of illness of less than
of Polymyalgia Rheumatica 2weeks duration, initial ESR greater than 40mm/h,
duration of morning stiffness exceeding 1 h, age
PMR was first described as senile rheumatic gout and 65 years or more, depression and/or weight loss, and
named PMR in 1957 [1]. Compared with GCA, PMR is bilateral tenderness in the upper arms. They suggested
much more common [2, 3]. PMR is characterized by that a patient might be regarded as probably having
the presence of pain and stiffness in the neck, shoulders PMR if any three or more of these criteria are fulfilled.
and hips, weight loss, low-grade fever, fatigue and ele- Pitting edema of the distal extremity has also been
vated erythrocyte sedimentation rate (ESR), with bursi- described in PMR. In the Olmsted County population-
tis, synovitis and tenosynovitis of the proximal shoulder based cohort of 245 patients with PMR seen over a
and hip girdles being characteristic features [4]. 22-year period, 19 patients (13 female, 6 male) had one
Stiffness is greatest in the morning and lasts up to about or more episodes of painful distal extremity swelling
an hour after the patient gets up in the morning. Muscle with pitting edema [7]. Jones and Hazelman [8] con-
weakness may be a prominent feature, so much so that cluded that PMR is not a benign disease.
they may have difficulty getting up from a sitting posi- The symptoms of PMR in GCA have been reported
tion or getting out of bed. The diagnosis is primarily in about 4060% of patients and those have been the
based on clinical findings but those can mimic other initial symptom complex in 2040% [913]. This has
diseases; therefore one has to exclude other causes for resulted in confusion about the clinical features of the
those symptoms. Arthralgias and myalgias may develop two diseases and their differentiation.
abruptly or evolve insidiously over several weeks or The diagnosis of PMR is clinical and depends on its
months, and malaise, fatigue, and depression, along classical symptoms. There seems to be a general opin-
with aching and stiffness, may be present for months ion that there is no need to do a temporal artery biopsy
before the diagnosis is made [5]. Pain at night is com- for PMR patients who have no clinical evidence of
mon, and movement during sleep may awaken the GCA. Where musculoskeletal symptoms are mild and
patient. According to Brooks and McGee[1], PMR is a of recent onset, or where they are stable and of long
clinical syndrome of proximal muscle pain in older duration (a year or more) without current or previous
patients that often presents a diagnostic challenge evidence of GCA, it may be reasonable to follow the
because of the large differential diagnosis, lack of patient closely without biopsy [5]. However, in our
definitive diagnostic criteria, and relatively frequent study [14] we found that 21% of the patients who
atypical clinical findings, such as peripheral developed visual loss had no systemic symptoms of

174 10 Clinical Features of Giant Cell Arteritis
GCA (i.e. had occult GCA). Therefore simple reliance In 1990 the American College of Rheumatology
on systemic symptoms of GCA can be misleading and [20]defined criteria to differentiate GCA from other
have tragic results. types of vasculitis, based on a comparison of 214 cases
diagnosed as GCA and 593 with other types of vascu-
litis. Of the 214 GCA patients, 18 had negative tempo-
ral artery biopsy and the diagnosis was based entirely
ystemic Clinical Features of Giant
S
on other criteria. They selected five criteria to diagnose
Cell Arteritis GCA: age 50 years at disease onset, new onset of
localized headache, temporal artery tenderness or
Paulley and Hughes [15] in 1960 stated that when decreased pulse, elevated ESR (Westergren) 50mm/h,
elderly people begin to fail mentally and physically, and temporal artery biopsy showing necrotizing
this disorder should be one of the first to be consid- arteritis, characterized by a predominance of mononu-
ered. GCA patients may complain of one or more of clear cell infiltrates or a granulomatous process with
the following symptoms: headache, scalp tenderness, multinucleated giant cells. In that study, the presence
neck pain, malaise, myalgia, anorexia, weight loss, of three or more of these criteria was associated with a
anemia, jaw claudication, PMR symptoms, abnormal sensitivity of 93.5% and a specificity of 91.2%. They
temporal arteries, flu-like symptoms, fever of used new headache and scalp tenderness or nodules as
unknown etiology and other vague systemic symp- a surrogate for temporal artery biopsy in those without
toms. In the rheumatologic literature headache has positive temporal artery biopsy.
been stressed as the most characteristic symptomatic Chmelewski et al. [21] retrospectively compared
expression of this vasculitic process, occurring in the presenting features of 30 patients with positive
>80% of patients [1618]. temporal artery biopsy and 68 with negative temporal
But there is a large amount of contradictory data artery biopsy. In their study, the presenting features
available on the signs, symptoms and diagnostic tests of temporal artery biopsy positive patients were head-
used to diagnose GCA. For example, in a review of ache in 93%, polymyalgia rheumatica in 57%, jaw-
the literature, Keltner [19] pointed out that the claudication in 50%, fever in 37%, scalp abnormality
reported incidence of headache in these patients varin 47% and neurologic symptoms in 17%. Patients
ied from 4% to 100%, tender temporal artery 2891%, with positive temporal artery biopsy had significantly
myalgia and arthralgia 2886%, fever 30100%, (p = 0.003) increased headache (93% vs. 62%) and jaw
weight loss 1676%, jaw claudication 467%, claudication (50% vs. 18%) when compared with neg-
anorexia 1469%, malaise 1297% and leg claudica- ative temporal artery biopsy. PMR symptoms were
tion in 243%. Variable and sometimes conflicting present in similar frequency in the two groups. To dif-
information in different studies may partly be due to: ferentiate the temporal artery biopsy positive from
(a) selection bias in patients, because of the different negative cases, headache had low specificity (40%),
specialties of the authors, e.g., internal medicine, jaw claudication more specificity (56%) and PMR had
rheumatology, neurology or ophthalmology, (b) low sensitivity (23%) but higher specificity (78%).
different diagnostic criteria used to define GCA, for Vilaseca et al. [22] suggested that the simultaneous
example, diagnosis based strictly on presence of a presence of jaw claudication, abnormal temporal arter-
positive temporal artery biopsy, or only on clinical ies on examination and headache of recent onset had
manifestations, and (c) variations in demographics a specificity of 94.8% for positive temporal artery
and sample size of the patient population in the biopsy. They concluded that concurrence of these three
studies. clinical features points to a definite diagnosis of GCA.
The essential requirement to prevent the dreaded Fernandez-Herlihy [23], on comparison of 29 tempo-
complication of blindness in patients with GCA is ral artery biopsy positive with 68 biopsy negative
early and correct diagnosis, and the start of intensive patients, found significantly more frequent jaw claudi-
high-dose corticosteroid therapy immediately. cation (p<0.0005), anorexia (p<0.001), anemia
Therefore, the question arises: which of all these symp- (p<0.01) and ESR of 45148 (average 102) mm/h
toms and signs are more helpful than others, and what (p<0.001) in the biopsy positive cases. They reported
importance should be attached to each of them? that a combination of headache, jaw claudication and
Systemic Clinical Features of Giant Cell Arteritis 175
abnormal temporal artery had a high specificity (100%) temporal artery biopsy and did not use other criteria.
but low sensitivity (17%) for predicting a positive tem- The symptoms and signs that showed significant asso-
poral artery biopsy for GCA. ciation were jaw claudication (p<0.0001), neck pain
Kearns [24] stressed that GCA ranks as the prime (p=0.0003), anorexia/weight loss (p=0.0005), fever
medical emergency in ophthalmology, there being no (p=0.0400) and scalp tenderness (p=0.058). However,
other disease in which the prevention of blindness when we considered a multivariate logistic regression
depends so much on prompt recognition and early model which included ESR and age, the only sys-
treatment. Since GCA is a well-known masquerader, temic symptoms that were significant were jaw clau-
and its early diagnosis is the key to correct manage- dication and neck pain (mostly in the occipital and
ment of patients and prevention of its dreaded compli- back parts of the neck). This implies that jaw claudi-
cations, it is essential to ascertain the validity, reliability, cation and neck pain were indicators of a positive
sensitivity and specificity of the various signs, symp- biopsy independent of ESR and age and that they
toms and diagnostic tests of GCA. With that as our were more highly correlated to the outcome of a posi-
primary objective, we conducted a prospective study tive biopsy than anorexia/weight loss, fever and scalp
specifically to investigate that particular aspect [25]. It tenderness. Other variables may also be helpful in the
was therefore imperative that the patients from whom diagnosis of GCA but not to the same extent as these
data were drawn must meet the definitive criterion for parameters. The conclusions of previous publications
diagnosis of GCA (i.e. positive temporal artery biopsy). on the subject were mainly based on univariate analy-
Using that criterion strictly, in our study we investi- sis. Our study [25] showed that the odds of having a
gated the incidence of the various systemic symptoms positive temporal artery biopsy were 9 times greater
and signs in patients with positive and negative tempo- with jaw claudication, 3.3 times with neck pain, 2.1
ral artery biopsy. Our findings are given in Table10.1. times greater with ESR 47107mm/h and 2.7 times
Figure 10.1 shows prominent and nodular temporal greater with ESR >107 mm/h relative to ESR
artery in one of my patients with biopsy positive GCA <47 mm/h, and 2.0 times greater when the patients
and bilateral A-AION. In our study [14] we found that were aged75 years compared to age <75 years
21% of the patients who developed visual loss had no (Table 10.2). The odds were 3.2 times greater with
systemic symptoms of GCA (i.e. had occult GCA). C-reactive protein (CRP) above 2.45mg/dl compared
Unlike the American College of Rheumatology to CRP2.45mg/dl.
and several other studies, we included only those Other systemic symptoms have been described
patients in the GCA group who had a positive in the literature, including vertigo, tinnitus and
Table10.1 Incidence of various systemic symptoms and signs in patients with positive and negative temporal artery biopsy for
GCA[25]
Signs and symptoms Temporal artery biopsy
Positive (n=106) Negative (n=257) P-value
Jaw claudication 51(48.1%) 22(8.6%) <0.0001
Anorexia/weight loss 55(51.9%) 84(32.7%) 0.0005
Neck pain 17(16.0%) 11(4.3%) 0.0003
Fever 28(26.4%) 42(16.3%) 0.040
Scalp tenderness 19(17.9%) 27(10.5%) 0.058
Headache 59(55.7%) 117(45.5%) 0.084
Abnormal temporal artery 21(19.8%) 33(12.8%) 0.105
Malaise 40(37.7%) 78(30.4%) 0.177
Myalgia 31(29.2%) 68(26.5%) 0.606
Anemia 14(13.2%) 31(12.1%) 0.730
Reproduced from Hayreh etal. [25]
he/she may still say that he/she can see you and
everything around. I have seen this in an occasional
GCA patient. This euphoria may be due to coinci-
dental cerebral ischemia, but visual symptoms of
cerebral origin are rare [28]. Psychiatric disorders
have also been mentioned. There are several reports
of scalp necrosis and I have seen that. Skin or
mucous membrane manifestations are like purpura,
necrotic ulcers, or necrosis of the anterior part of
tongue [31]
Systemic symptoms indicate a more generalized
arteritis and this has been proven by several necropsy
studies since 1941 [29, 3241]. These have shown the
involvement of various arteries: aorta, pulmonary, cor-
onary, innominate, carotids, temporal, subclavian,
Fig.10.1 This shows prominent, nodular and tender superficial
radial, mesenteric, coeliac, renal, iliac, vertebral, fem-
temporal artery (arrow) in a patient with giant cell arteritis
oral, popliteal, dorsalis pedis and, in the orbit, the oph-
thalmic, central retinal, and posterior ciliary arteries
hearing loss [26], hepatosplenomegaly and arthritis (especially the posterior ciliary arteries). Thus, GCA is
[27], cerebral or brain stem infarction [28], and a systemic disease. This contradicts the commonly
myocardial infarction [28, 29]. A frequent and used terms temporal arteritis and cranial arteritis
striking feature in these patients is the degree of for this disease. I have found that the term temporal
euphoria they have [28, 30], so that even bilateral arteritis has resulted in a misconception among
complete blindness is accepted philosophically as patients that visual loss is due to the involvement of the
if nothing significant had gone wrong; this may temporal arteries, when in fact the temporal arteries
trick a beginner in estimating their visual loss, play no role in the blood supply of the eye and optic
because even if the patient cannot perceive light, nerve.
Table10.2 Clinical variables which showed an association with increased odds of a positive biopsy for giant cell GCA[25]
Variable Odds ratio 95% Confidence interval P-value
for odds
Using sample with ESR dataa
Jaw claudication 9.1 5.0, 16.7 <0.0001
Neck pain 3.4 1.4, 8.3 0.0085
Age75years 2.0 1.2, 3.4 0.0105
ESR 47107mm/hb 2.0 1.0, 4.1 0.0454
ESR > 107mm/hb 2.7 1.2, 5.9 0.0106
Using sample with both ESR and CRP datac
Jaw claudication 7.5 3.1, 18.0 <0.0001
Neck pain 4.5 1.2, 17.2 0.0263
Age75years 2.3 1.0, 5.3 0.0412
ESR 47107mm/hb 1.4 0.5, 4.5 0.5326
ESR >107mm/hb 2.3 0.6, 8.4 0.2085
CRP >2.45mg/dld 3.2 1.2, 8.6 0.0208
Reproduced from Hayreh etal. [25]
a
n=363: 106 positive biopsy and 257 negative biopsy
b
Compared to patients with ESR < 47mm/h
c
n=223: 43 positive biopsy and 180 negative biopsy
d
Compared to patients with CRP < 2.45mg/dl
References 177
Conclusion 17. Gonzlez-Gay MA, Barros S, Lopez-Diaz MJ, Garcia-

Porrua C, Sanchez-Andrade A, Llorca J. Giant cell arteritis:
disease patterns of clinical presentation in a series of 240
Our study [25] showed that symptoms of jaw claudica- patients. Medicine (Baltimore). 2005;84(5):26976.
18. Gonzalez-Gay MA, Garcia-Porrua C, Miranda-Filloy JA,
tion, anorexia, weight loss and neck pain, particu-
Martin J. Giant cell arteritis and polymyalgia rheumatica:
larly jaw claudication, and elevated CRP are highly pathophysiology and management. Drugs Aging.
suggestive of GCA, and normal ESR or absence of 2006;23(8):62749.
any systemic symptoms does not rule out GCA. 19. Keltner JL. Giant-cell arteritis. Signs and symptoms.
Ophthalmology. 1982;89(10):110110.
20. Hunder GG, Bloch DA, Michel BA, Stevens MB, Arend
WP, Calabrese LH, et al. The American College of
References Rheumatology 1990 criteria for the classification of giant
cell arteritis. Arthritis Rheum. 1990;33(8):11228.
21. Chmelewski WL, McKnight KM, Agudelo CA, Wise CM.
1. Brooks RC, McGee SR. Diagnostic dilemmas in polymyal- Presenting features and outcomes in patients undergoing
gia rheumatica. Arch Intern Med. 1997;157(2):1628. temporal artery biopsy. A review of 98 patients. Arch Intern
2. Chuang TY, Hunder GG, Ilstrup DM, Kurland LT. Med. 1992;152(8)):16905.
Polymyalgia rheumatica: a 10-year epidemiologic and clini- 22. Vilaseca J, Gonzalez A, Cid MC, Lopez-Vivancos J, Ortega
cal study. Ann Intern Med. 1982;97(5):67280. A. Clinical usefulness of temporal artery biopsy. Ann Rheum
3. Salvarani C, Gabriel SE, OFallon WM, Hunder GG. Dis. 1987;46(4):2825.
Epidemiology of polymyalgia rheumatica in Olmsted 23. Fernandez-Herlihy L. Temporal arteritis: clinical aids to
County, Minnesota, 19701991. Arthritis Rheum. 1995; diagnosis. J Rheumatol. 1988;15(12):1797801.
38(3):36973. 24. Kearns TP. Collagen and rheumatic diseases: ophthalmic
4. Miller D, Allen SH, Walker SE. A primary care physicians aspects. In: Mausolf FA, editor. The eye and systemic dis-
guide to polymyalgia rheumatica. Primary Care Rep. ease. St. Louis: Mosby; 1975. p. 10518.
1998;4(10):91985. 25. Hayreh SS, Podhajsky PA, Raman R, Zimmerman B. Giant
5. Hunder GG. Giant cell arteritis and polymyalgia rheumatica. cell arteritis: validity and reliability of various diagnostic
Med Clin North Am. 1997;81(1):195219. criteria. Am J Ophthalmol. 1997;123(3):28596.
6. Bird HA, Esselinckx W, Dixon AS, Mowat AG, Wood PH. 26. Font JH. Otorhinolaryngological considerations on the tem-
An evaluation of criteria for polymyalgia rheumatica. Ann poral arteritis syndrome. JAMA. 1960;174:8536.
Rheum Dis. 1979;38(5):4349. 27. Lavy S, Liban E. Temporal arteritis; report of an unusual
7. Salvarani C, Gabriel SE, Hunder G. Distal extremity swell- case. Harefuah. 1958;54(1):610.
ing with pitting edema in polymyalgia rheumatica. Arthritis 28. Meadows SP. Temporal or giant cell arteritis: ophthalmic
Rheum. 1996;39(1):7380. aspects. In: Smith JL, editor. Neuro-ophthalmology: sympo-
8. Jones JG, Hazleman BL. Prognosis and management of sium of the university of Miami and the Bascom palmer eye
polymyalgia rheumatica. Ann Rheum Dis. 1981;40(1):15. institute. St. Louis: Mosby; 1968. p. 14857.
9. Horton BJ, Magath TB, Brown GE. An unusual form of 29. Spencer WH, Hoyt WF. A fatal case of giant cell arteritis
arteritis of the temporal vessels. Mayo Clin Proc. (temporal or cranial arteritis) with ocular involvement. Arch
1932;7:7001. Ophthalmol. 1960;64:8627.
10. Fauchald P, Rygvold O, Oystese B. Temporal arteritis and 30. Meadows SP. Temporal arteritis and the loss of vision. Trans
polymyalgia rheumatica. Clinical and biopsy findings. Ann Ophthalmol Soc UK. 1954;74:1324.
Intern Med. 1972;77(6):84552. 31. Kinmont PD, McCallum DI. Skin manifestations of giant-
11. Healey LA, Wilske KR. The systemic manifestations of tem- cell arteritis. Br J Dermatol. 1964;76:299308.
poral arteritis. New York: Grune & Stratton; 1978. 32. Gilmour JR. Giant cell chronic arteritis. J Pathol Bacteriol.
12. Calamia KT, Hunder GG. Giant cell arteritis (temporal 1941;53(2):26377.
arteritis) presenting as fever of undetermined origin. Arthritis 33. Cooke WT, Cloake PCP, Govan ADT, Colbeck JC. Temporal
Rheum. 1981;24(11):14148. arteritis: a generalized vascular disease. Q J Med.
13. Nordborg E. Bengtsson BA Epidemiology of biopsy-proven 1946;15(1):4775.
giant cell arteritis (GCA). J Intern Med. 1990;227(4):2336. 34. Sproul EE. A case of temporal arteritis. NY State J Med.
14. Hayreh SS, Podhajsky PA, Zimmerman B. Occult giant cell 1942;42:3456.
arteritis: ocular manifestations. Am J Ophthalmol. 35. Chasnoff J, Vorzimer JJ. Temporal arteritis: a local manifestation
1998;125(4):5216. of a systemic disease. Arch Intern Med. 1944;20(2):32733.
15. Paulley JW, Hughes JP. Giant-cell arteritis, or arteritis of the 36. Curtis HC. Cranial arteritis: a review of the literature with report
aged. Br Med J. 1960;2(5212):15627. of a case with total blindness. Am J Med. 1946;1(4):43746.
16. Huston KA, Hunder GG, Lie JT, Kennedy RH, Elveback 37. Cardell BS, Hanley T. A fatal case of giant-cell or temporal
LR. Temporal arteritis: a 25-year epidemiologic, clinical, arteritis. J Pathol Bacteriol. 1951;63(4):58797.
and pathologic study. Ann Intern Med. 1978;88(2): 38. Kreibig W. Optikomalazie, die Folge eines Gefassverschlusses
1627. im retrobulbren Abschnitt des Sehnerven. [Opticomalacia
caused by vascular occlusion in the retrobulbar portion of 41. Crompton MR. The visual changes in temporal (giant-cell)
optic nerves.]. Klin Monatsbl Augenheilkd Augenrztl arteritis. Report of a case with autopsy findings. Brain.
Fortbild. 1953;122((6):71931. 1959;82:37790.
39. Meneely Jr JK, Bigelow NH. Temporal arteritis; a critical
evaluation of this disorder and a report of three cases. Am
J Med. 1953;14(1):4651. For a complete bibliography and detailed review of previously
40. Heptinstall RH, Porter KA, Barkley H. Giant-cell (temporal) published studies in the literature on the subject, please refer to
arteritis. J Pathol Bacteriol. 1954;67(2):50719. bibliography in previous publications listed here.
Diagnostic Criteria for Giant Cell Arteritis
11
Giant cell arteritis (GCA) is the most important medical easons for Controversy on
R
emergency in ophthalmology, as was rightly stressed Management of GCA to Prevent
by Kearns [1] when he stated that GCA ranks as the
Visual Loss
prime medical emergency in ophthalmology, there
being no other disease in which the prevention of blind-
ness depends so much on prompt recognition and early This important topic requires a detailed discussion
treatment. A study [2] further emphasized that fact, before actually discussing the management of GCA.
when it found that a high proportion of patients with I have investigated various aspects of GCA, particu-
permanent visual loss have a delayed diagnosis and larly the visual loss due to GCA, in my clinical studies
treatment; in that study 35% had systemic symptoms since 1970 [421], and have found the following
for an average of 10.8 months before visual loss and reasons for the controversy on management of GCA to
65% had premonitory visual symptoms for an average prevent visual loss.
of 8.5days. Thus, blindness in GCA is preventable if
these patients are diagnosed early and treated imme-
diately and aggressively (see Chap. 13). This obviously
iffering Perspectives on GCA Between
D
raises two critical issues about management of GCA:
Rheumatologists and Ophthalmologists
1. How to establish early a definite diagnosis of
GCA?
This is the most important reason for the controversy.
2. What is the proper treatment to prevent blindness?
Rheumatologists deal with patients essentially with
There is a voluminous literature on both these topics; rheumatologic manifestations, while ophthalmologists
however, there is still a good deal of controversy on see GCA patients with visual loss or patients with
both issues [3]. This has been partly responsible for occult GCA who lose vision without having any rheu-
missing a correct diagnosis of GCA, inadequate man- matologic or other systemic symptoms [15]. Therefore,
agement and unnecessary blindness. for the ophthalmologists GCA is blinding disease with
It is essential to have an in-depth insight into the tragic consequences, whereas for rheumatologists it is
management of GCA to prevent visual loss. I will address a disease mainly with rheumatologic complaints and
the following 3 questions relevant to this topic: not particularly serious. This philosophy is very evident
from the most recent review on the subject by a group
1. What are the reasons for disagreement on the man-
of rheumatologists [22, 23]. The reviews give the topic
agement of GCA to prevent visual loss?
of visual loss in GCA short shrift, despite the fact that
2. How can a definite diagnosis of GCA be quickly
the most dreaded complication of GCA is visual loss.
established?
Moreover, the authors recommend a management
3. What is the proper treatment of GCA to prevent
which may be appropriate for polymyalgia rheumatica
visual loss?
(PMR) (with no risk of blindness), but falls well short
In this chapter I discuss issues number 1 and 2; the third of the kind of therapy required in GCA patients to pre-
issue is in Chap. 13 dealing with the treatment of GCA. vent blindness (see Chap. 13).

180 11 Diagnostic Criteria for Giant Cell Arteritis
onsidering GCA and PMR as One

C that study advocated that in such cases new headache
Disease Entity and scalp tenderness or nodules be used as a
surrogate.
We conducted a prospective study from 1973 to
For management, most rheumatologists believe this to 1995 in 363 consecutive patients [13]. As discussed
be the case [24], although the most dreaded complica- above, the essential requirement to prevent the dreaded
tion of GCA is visual loss which does not occur in complication of blindness in patients with GCA is early
PMR. Healey [25] described PMR and GCA as sepa- and correct diagnosis. The purpose of our study was to
rate syndromes rather than two manifestations of the establish the relative degree of validity, reliability,
same disease; however, Wilke and Hoffman [26] sensitivity and specificity of the various signs, symp-
argued that they represent a spectrum of the same dis- toms and diagnostic tests for GCA. It was therefore
ease. The clinical behavior of PMR and GCA supports imperative that the patients from whom data were
Healys views [25]; it is generally agreed that the dos- drawn must meet the definitive criterion for diagnosis
age of steroid therapy required to manage PMR is of GCA (i.e., positive temporal artery biopsy). All
much smaller than that for GCA, and the overall prog- these patients were suspected of having GCA by physi-
nosis of PMR is much better than that of GCA [3, 22]. cians from the various medical specialties in the multi-
GCA is a often a widespread chronic vasculitis and specialty University of Iowa Hospitals and Clinics and
visual loss is its major complication but not of PMR. referred for temporal artery biopsy to our department,
Therefore, it is illogical to treat the two diseases as one which does the biopsies for the entire medical center;
entity. That seems to be the root cause of most of the this gave us an unbiased group of patients. Of the 363
confusion on the management of GCA. A review of patients, 106 were found to have positive temporal
the rheumatologic literature reveals that, more often artery biopsy and 257 negative for GCA. A compari-
than not, the two types of cases are grouped into one son of the systemic and hematologic findings in patients
category, without any clear distinction between the with positive versus negative temporal artery biopsy
two, when describing diagnosis and management of [13] showed the following significant differences from
GCA. Consequently the heterogeneity of cases in the criteria advocated by the ACR study [27].
various studies has contributed to controversy on the
management of GCA. Likewise, when discussing
cases with rheumatologists over the years, I have found Occult GCA
that they often describe and manage temporal artery
biopsy confirmed GCA patients no differently from A vital fact that emerged from our study was that 21%
those with PMR or even diagnose them to have PMR. of patients with visual loss and positive temporal artery
biopsy for GCA had no systemic symptoms of any
kind whatsoever at any time and visual loss was the
sole complaint, i.e., they had occult GCA [15]. Because
Clinical Criteria Used to Diagnose GCA they have no systemic symptoms, such patients do not
go to rheumatologists and would not be included in a
The following five criteria advocated by the American rheumatologic study. Thus, when systemic manifesta-
College of Rheumatologists (ACR) [27] are generally tions of GCA are the primary and sole criterion for
accepted as the gold standard to classify GCA: (1) diagnosing GCA, 21% of GCA patients with visual
age 50 years at onset, (2) new onset of localized loss are going to be missed. That is an unacceptably
headache, (3) temporal artery tenderness or decreased high number to be missed to prevent blindness.
temporal artery pulse, (4) elevated erythrocyte sedi-
mentation rate (ESR Westergren) 50mm/h, and (5)
positive temporal artery biopsy for GCA. They stated I mportant Clinical Criteria Missing from the
that: A patient shall be classified as having GCA if at Five Criteria of ACR
least three of these five criteria are met. But in the ACR
study, 18 of 214 (8.4%) patients temporal artery biopsy In our study [13], logistic regression analysis showed
was either negative for GCA (15) or not done (3), and that the odds of having temporal artery biopsy positive
Reasons for Controversy on Management of GCA to Prevent Visual Loss 181
GCA were 9.0 times greater with jaw claudication, 3.4 that none of 167 patients fulfilling the ACR criteria for
times with neck pain, and 3.2 times with C-reactive GCA and low ESR (i.e., <40mm/h) developed blind-
protein (CRP) above 2.45 mg/dl than those without ness. In our studies [14, 15] of 85 consecutive GCA
them. The ACR studys five criteria do not include any patients with positive temporal artery biopsy for GCA
of these important criteria, which are very helpful in and with visual loss, the lowest level of ESR was
diagnosis of GCA. 5 mm/h, and in another study in 96 temporal artery
biopsy positive GCA patients with visual loss, it was
4mm/h; these findings refute the claim by Salvarani
Headache and Hunder [29].
The ACR study advocates using new onset of local-

ized headache as an important criterion. Headache of New Headache and Scalp Tenderness
recent onset was often a major reason why internists,
rheumatologists, neurologists and other physicians In patients with negative temporal artery biopsy for
referred patients to us for temporal artery biopsy. In our GCA, the ACR study advocated using new headache
study [13], however, 56% of patients with positive tem- and scalp tenderness or nodules as a surrogate. In our
poral artery biopsy complained of headache, but so did study [13], 56% with positive temporal artery biopsy
45.5% of those with negative biopsy (P=0.084). Thus, and 45.5% with negative biopsy (P=0.084) com-
although headache is frequently stressed as an impor- plained of recent onset of headache (see above), and
tant sign of GCA in the rheumatologic literature, in our scalp tenderness was complained of by 18% with posi-
study we did not find that always reliable for the diag- tive temporal artery biopsy and 10.5% with negative
nosis of GCA because headache is not an uncommon biopsy. Thus both these criteria are unreliable for the
complaint and it can be due to many other diseases. diagnosis of GCA.
Clinically Abnormal Temporal Artery Temporal Artery Biopsy
The ACR study advocated using temporal artery ten- A positive temporal artery biopsy for GCA is definite
derness or decreased temporal artery pulse as a crite- proof of GCA. In the ACR study 8% of the patients
rion. In our study [13], abnormal temporal artery on had negative temporal artery biopsy; in our study [13]
physical examination was seen in 20% with positive 100% had positive temporal artery biopsy. That is an
temporal artery biopsy and 13% with negative biopsy important difference.
(P=0.105). Moreover, patients with positive temporal The reason for the differences between the two
artery biopsy not infrequently had perfectly normal studies is the difference in their patient populations;
temporal arteries clinically. since the ACR study was conducted by rheumatolo-
gists, it would seem their primary criterion of inclusion
was the presence of rheumatologic systemic symptoms,
Elevated ESR while our study had an unbiased group of patients with
rheumatologic as well as visual symptoms, since the
The ACR study advocated elevated ESR50mm/h as patients were referred to us by physicians from all the
one of the criteria. In contrast, in our study [13], ESR medical specialties, including rheumatologists. Since
at initial visit in those with positive temporal artery the most serious complication of GCA is visual loss,
biopsy varied between 4 and 140 mm (mean 8533 the criteria advocated by the ACR study may be ade-
SD, median 87.5) and in those with negative biopsy quate for diagnosing GCA patients with rheumatologic
between 2 and 155mm (mean 7239 SD, median 68). symptoms, but they are not good enough to prevent all
Most importantly, our study showed that normal ESR the GCA patients from going blind, particularly
does not rule out GCA. Many reports in the literature patients with occult GCA [15] who never develop any
have also reported normal or low ESR in 530% of systemic symptoms of GCA and will probably not be
GCA patients [28]. Salvarani and Hunder [29] stated seen by rheumatologists [22].
Thus, in conclusion, it is evident that the ACR study example, Sparrow etal. [30] determined ESR in 1,480
criteria are likely to result in some false-negative or healthy men and reported a significant (p<0.001)
false-positive diagnoses of GCA, risking visual loss. correlation between increasing age and higher ESR
values. Based on ESR in 4,341 healthy normal persons
(2,140 men and 2,201 women), Gilbertsen [31], con-
cluded that the range of normal ESR which include
stablishing a Definite Diagnosis
E 90% of persons aged 4579years was 032mm/h in
of GCA Without Delay men and 148 mm/h in women. Bottlinger and
Svedberg [32] in 2,500 subject aged >50years found
The key first step in management of GCA, to forestall upper limits of normal 20mm in men and 30mm in
the possibility of loss of vision, is its early and accu- women. Miller etal. [33], based on a study of 27,912
rate diagnosis. A patient with GCA presents to his/her individuals, aged 2065 years, derived an empirical
physician with either systemic or ophthalmic manifes- formula for ESR that included 98% of healthy persons;
tations of GCA or both. But GCA is a well-known it was: age in years divided by two for men, and age in
masquerader. Let us briefly review the merits of various years plus 10 divided by two for women. However,
systemic and ophthalmic manifestations of GCA and Hanger et al. [34], based on a survey of 194 elderly
of various diagnostic tests that can help us to achieve a subjects, advocated that in Miller etal.s [33] formula,
definite, early diagnosis. three be used instead of two as the dividing factor.
There are numerous other conflicting reports based on
smaller series.
There are, however, numerous anecdotal case
Systemic Manifestations of GCA reports of low ESR in patients with temporal artery
biopsy positive GCA - between 5 and 20 mm/h [11,
These are discussed at length in Chap. 10. 3542] or between 22 and 30mm/h [38, 4143].
ESR is affected by age, gender, and many other
hematologic and other parameters unrelated to GCA
and it is a non-specific test. Diabetics tend to have
Various Diagnostic Tests for GCA higher ESR. Intercurrent infections also cause a rise of
ESR. Factors affecting ESR level include the degree of
rouleaux formation by the blood cells, fibrinogen,
Various Hematologic Tests immunoglobulins, other plasma proteins, lipids, and
even drugs. It can also be affected by the time lapse
These include the following: between the venipuncture and testing of ESR in the
laboratory. (In our study it was always done within
a few minutes.) It is increased in anemia, inflammatory
Erythrocyte Sedimentation Rate (Westergren) disease, malignancy, infection, connective tissue
diseases, pregnancy, hypercholesterolemia and a host
An elevated ESR is universally emphasized as a sine of other systemic conditions. Very low ESR is seen in
qua non for the diagnosis of GCA and that its estima- a number of hematologic disorders (including poly-
tion is an important test in diagnosis. There is, however, cythemia, hypochromic microcystic anemia, hereditary
a common misconception among physicians about spherocytosis, hemoglobinopathy and hypofibrinogen-
normal and abnormal levels of ESR for diagnosis emia), congestive heart failure and treatment with anti-
and management of GCA patients. Therefore, the most inflammatory drugs. These factors make ESR a
important question in this context is: What is the nor- non-specific indicator of disease. In our GCA group,
mal and abnormal ESR in persons 50years and older? there was no patient whose elevated ESR could be
There are many reported normal values in the litera- attributed to any disease other than GCA.
ture and several proposed formulae for calculating We investigated ESR in 749 individuals aged
normal values. All reports agree that ESR increases 16100 years old without any systemic disease, and
with age and women have higher ESR than men. For found it 159 mm/h; it showed a sensitivity and
Various Diagnostic Tests for GCA 183
95 % UPPER CONFIDENCE BOUND

MALES FEMALES
Positive biopsy for GCA Positive biopsy for GCA

160 160
No GCA No GCA
140 140
120 120
100 100
ESR
ESR
80 80
60 60
40 40
20 20
0 0
10 20 30 40 50 60 70 80 90 100 10 20 30 40 50 60 70 80 90 100
AGE AGE
Fig.11.1 Graphs showing regression analysis of ESR (in mm/h) the control group. At places empty circles for various patients
with age (in years) and gender in our patients with positive tem- overlap each other giving an erroneous impression of open
poral artery biopsy for GCA (empty circles) and control group circles, corresponding to patients with GCA, being too low
(dots). Dotted line represents the 95% upper confidence limit for (Reproduced from Hayreh etal. [13])
specificity of 92% with the cutoff criterion of C-Reactive Protein (CRP)

33 mm/h for men and 35 mm/h for women
(Fig.11.1) [13]. This meant that 8% of patients with This is an acute-phase plasma protein of hepatic origin
GCA had ESR within normal levels, the lowest being and has been shown to be elevated in GCA [4548].
4mm/h. CRP rises with tissue necrosis, infectious diseases,
In our 106 patients with positive temporal artery inflammatory diseases, surgical tissue injury, myocar-
biopsy the ESR varied between 4 and 140mm/h, with dial infarction, rheumatic disease, transplantation, and
a median 87.5 mm/h and mean 8533 mm/h a host of other conditions. Concentration of CRP can
(Westergren); while in our 749 normal individuals it reach abnormal levels within 46h after the insult and
was between 1 and 59 mm/h [13]. Thus there is can increase as much as a 1000-fold. Similarly, after
a marked overlap at lower levels of ESR between the the insult is removed, its concentration can drop
normal individuals and GCA patients with positive rapidly. Unlike ESR, CRP is not influenced by various
temporal artery biopsy. In conclusion, the rule for hematologic factors or age; making CRP a more sensi-
ESR should therefore be: if it is markedly elevated, it tive diagnostic test than ESR. According to Deodhar
is suggestive of GCA but normal or low ESR does not [49], CRP has many advantages over ESR for diagnosis
rule out GCA an extremely important point to be of GCA, because: (1) it is highly sensitive, reproduc-
borne in mind because in many patients the diagnosis ible, quantitative and easy and rapid to perform (we get
of GCA is dismissed outright if the ESR is not elevated. the results of CRP in less than 1 h after drawing the
Those with no visual loss had a significantly greater blood), (2) it responds to disease activity much more
initial ESR (median of 102mm/h; interquartile range rapidly than ESR, and (3) it is not influenced by age,
of 70124mm/h) than those with visual loss (median gender or hematologic factors. However, CRP, like
of 80 mm/h; interquartile range of 52108 mm/h) ESR, is a non-specific test. CRP testing is superior to
(p=0.004) [44]. ESR on clinical, scientific, and practical grounds.
two patients (1.7%) had a normal CRP despite an elevated

140 CONTROL
GCA ESR. The study concluded that Although most patients
with GCA have both an elevated ESR and CRP, there
120
can be nonconcordance of the two blood tests. Although
100 such nonconcordance is most often a normal ESR but
an elevated CRP, the finding of an elevated ESR and
80 a normal CRP also is consistent with GCA. The use of
ESR
both tests provides a slightly greater sensitivity for the

60 diagnosis of GCA than the use of either test alone.
However, it depends upon what is considered as nor-
40 mal ESR and CRP. For example, in our hematology
laboratory at the University of Iowa Hospital & Clinics,
20
during the period of our study [13], CRP was measured
from 1mg onwards in increments of 1mg; however,
0
<0.05 0.1 1 10 100
lately it has been changed to <0.5mg as the normal. In
CRP our study [13], we had GCA patients who initially had
45mg CRP, which decreased with steroid therapy to
Fig.11.2 This is a scatter plot showing relationship of ESR (in 1mg, so that in such a patient a value of <0.5mg CRP
mm/h) and CRP (in mg/dl) in our patients with positive temporal
as normal is deceptive. This factor may be responsible
artery biopsy for GCA (empty circles) and control group (dots)
(Reproduced from Hayreh etal. [13]) for the presence of normal CRP in one case in the
above study [50].
Some have argued that ESR and CRP should not be
We have also found it good for monitoring disease used as independent variables since they are measuring
activity [44]. the same phenomenon (acute phase response), and that
In our study [13], CRP levels in GCA patients var- it may be more appropriate to assay interleukin-6, the
ied between 0.5 and 34.7 (median 4.35 and mean cytokine that directly stimulates CRP synthesis and
6.66.7) mg/dl, and Fig.11.2 shows the relationship other protein production leading to elevated ESR. For
of ESR and CRP in patients with GCA and control sub- example, Weyand etal. [51], to determine the value of
jects. In GCA patients a high CRP is associated with a the ESR and plasma interleukin-6 as biologic markers
high ESR. Control subjects with high ESR values do for monitoring disease activity in GCA, investigated 25
not necessarily have high CRP. Sensitivity and speci- biopsy positive GCA patients in a prospective study.
ficity of CRP in detection of GCA in men were 100% They concluded that plasma interleukin-6 is more sen-
and 83%, and in women 100% and 79% respectively. sitive than ESR for indicating disease activity in
On logistic regression analysis of ESR and CRP with untreated and treated GCA patients. We [13] conducted
positive temporal artery biopsy, only CRP showed a a small pilot study in 23 biopsy positive GCA patients
significant association with positive biopsy. Thus, we on treatment to find out the relationship of interleukin-6
have found CRP a highly useful test in diagnosing and with ESR and CRP. We found no significant linear rela-
monitoring the activity of GCA. Unlike ESR, there tionship between ESR and interleukin-6 (r=0.004,
was no significant difference in initial CRP between p=0.982). There was a linear relationship between
those without visual loss (median of 9.1; interquartile CRP and interleukin-6 (r=0.533, p=0.028). This would
range of 417.1) and those with visual loss (median of suggest that interleukin-6 may not provide any better
6.8; interquartile range of 3.111.7) (p=0.207) [44]. information than CRP. Moreover, CRP, unlike interleu-
A retrospective study [50] evaluated ESR and CRP kin-6, is a quick, cheap and readily available test.
in 119 temporal artery biopsy positive GCA patients.
In this study, the ESR had a sensitivity of 7686%,
whereas an elevated CRP had a sensitivity of 97.5%. Thrombocytosis
The sensitivity of the ESR and CRP together was 99%.
Only 1 of the 119 patients (0.8%) presented with a In 1963 Olhagen [52] first reported thrombocytosis in 7
normal ESR and normal CRP (double false negative); of his 12 GCA patients. Since then several reports of
thrombocytosis in GCA have appeared in the literature complications has been debated in the literature [56,
[5370]. These studies have suggested a role for esti- 59, 60, 63, 67, 68, 7173]. For example, De Keyser
mation of platelet count in diagnosis and management etal. [60]. reported a correlation between thrombocy-
of GCA because a decrease in platelet count with ste- tosis and ischemic lesions because their group with
roid therapy has also been reported. Some studies have ischemic complications (18 patients with visual loss,
linked thrombocytosis to a higher risk of ischemic stroke and transient ischemic attack) had a signifi-
complications in GCA [59, 60, 64, 68, 71, 72] while cantly (P<0.01) higher prevalence of thrombocytosis
others did not find this relationship [57, 67, 73]. In the and a higher median platelet count than the group
literature there were marked disparities among various without those ischemic complications (38 patients).
studies about the prevalence rate of thrombocytosis, the Similarly Liozon etal. [68] concluded that there was
cut-off values used to define thrombocytosis, as well as a correlation between increased risk of vision loss and
the criteria used to diagnose GCA (Table11.1). higher platelet counts. However, a careful review of
We [70] investigated thrombocytosis in a study of the data and statistical methods in the latter study
121 patients with temporal artery biopsy positive GCA; revealed that in the analysis used to identify the inde-
57.0% had thrombocytosis, defined by the hematology pendent risk factors for permanent vision loss, there
laboratory of our University Hospitals and Clinics, as were 17 variables in the logistic regression model of
a serum platelet count >400 103/ml. Overall, the 23 patients with permanent vision loss and 151 with no
platelet count was 421 (313504) 103/ml, and this was vision loss. In such an analysis, a general rule is to
a significantly (P<0.0001) higher value for platelet include no more than one independent variable per ten
count than in the control group, with sensitivity of subjects in the smaller of the two outcome groups to
57.0% and specificity of 96.5%; comparable values for derive a model that is reliable [74]. Therefore, Liozon
ESR were 94.2% and 80.5% respectively and for CRP et al.s [68] analysis incorporated too many indepen-
98.6% and 75.7% respectively. dent variables in the logistic regression model for the
The finding of an elevated platelet count in GCA small sample size, and that resulted in a model that is
has been described as reactive thrombocytosis and over-fitted. Thus, the conclusions derived from their
attributed to the systemic inflammation present with multiple logistic regression model may not be valid.
this disorder [54, 64]. In our study as well as in most Those who believe in a possible relationship
other studies [53, 54, 57, 6072] platelet count was between thrombocytosis and ischemic lesions in GCA,
evaluated among GCA patients at the time of initial have proposed various mechanisms to account for it.
presentation, before the start of steroid therapy, as an Bengtsson [59] speculated that the arteritic process
acute phase reactant. Some authors have also studied may adversely affect the prostacyclin synthesis in the
the change in platelet count over time or in response arterial wall, thereby disturbing the balance between
to corticosteroid therapy among patients with GCA prostacyclin and thromboxane A2 release from the
[54, 59, 62, 64, 66, 67]. Lincoff etal. [64] studied the activated platelets, a substance known to be a potent
platelet count in the year prior to diagnosis among 19 vasoconstrictor as well as a proaggregatory agent [75].
patients with biopsy-proven GCA, and described an Gibb et al. [71], in one case of fatal basilar artery
escalating thrombocytosis among 13 patients; nota- thrombosis and GCA, found a high platelet count
bly, all 13 patients showed a drop in their platelet count (566 103 m/l), and concluded that ischemia could
after treatment with corticosteroids, which has also result from embolization of platelet clumps from
been reported by most of the other studies in the inflamed intimal surfaces or by release of thrombox-
literature. ane A2 by activated platelets producing platelet aggre-
There are reports in the literature claiming that gation and vasoconstriction.
GCA patients with systemic symptoms have higher In contrast to that, other authors have found no
platelet counts than those without systemic symptoms such association between thrombocytosis and an
[57, 66]. This difference was not observed in our study increased risk of ischemic complications in GCA [67,
[70] when we compared GCA patients with and 73]. For example, Cid et al. [73], in a study of 200
without systemic symptoms. consecutive biopsy-proved GCA patients, on compar-
The role of elevated platelet count and the possi ing patients with ischemic lesions (32 patients) versus
ble association of thrombocytosis with ischemic those without ischemic lesions (168 patients), found
186
Table11.1 Summary of reports of thrombocytosis in GCA, based on large series (Reproduced from Costello etal. [70])
Author(s) Patients Mean age Visual loss in ESR initial (Mean) Percent of Initial platelet count in Temporal
(number) (years) (patients) (mm/h) patients with 103/ml (Mean) artery biopsy
thrombocyto- positive for
sis (103/ml) GCA in
Malmvall etal. [53] 68 69 8 (reversible) 88 35% (>350) 338 42
a a
Bengtsson [59] 80 70 NA NA 54% (350) 349 50a
De Keyser etal. [60] 56 76 13 96 in patients 37% (>400) 475 in patients NA
with ischemia; 91 with ischemia; 338
in without ischemia in without
Price etal. [61] 10 NA NA 72 NA 526 10
Gonzalez-Gay 108 76 Permanent in 17 GCA=97 NA GCA=413 108
etal. [63]
GCA=63 74 Transient in 13 GCA+PMR=91 GCA+PMR=429
GCA+PMR
= 45
Lincoff etal. [64] 19 75men (12) Permanent in 10 74 68% (>400) 477 19
73women (7) Transient in 2
Vrij etal. [66] 19 72 (median) NA 82 NA 507 19
Gonzalez Alegre 34 72 14 patients 98 44% (>400) 379 18
etal. [67]
Liozon etal. [68] 174 75 48 90 50% (>400) 419 147
Foroozan etal. [69] 47 78 NA 82 57% (>400) 433 47
a
From a total of 117 patients diagnosed as having GCA, platelet study performed in 80 patients. Author gives no demographic information on the latter group. Reported 62% of 117
had positive biopsy; hence extrapolating positive biopsy in 62% of 80 would be 50
FUO Fever of unknown origin, NA Not available, PMR Polymyalgia rheumatica
11 Diagnostic Criteria for Giant Cell Arteritis
no significant difference in platelet count between the ocular ischemic complications compared to those
two. Wu [76] found none of the 11 patients with reac- without in GCA, treatment with aspirin or other
tive thrombocytosis had thrombotic or hemorrhagic platelet anti-aggregating agents to prevent visual loss
complications. In our study [70] of 121 GCA patients, in GCA may not be justified. It seems the faulty ratio-
we also found no significant difference in the mean nale for the use of aspirin in GCA to prevent ischemic
platelet counts between GCA patients with vision loss complications stems from confusion between two
(406 103/ml) and those without (425 103/ml). It is very different types of thrombocytosis: reactive throm-
relevant to point out that permanent visual loss, as a bocytosis (seen in GCA) and essential thrombocytosis
complication of biopsy-proved GCA, was much (a chronic progressive myeloproliferative disorder of
higher in our study than has been reported in other insidious onset). With essential thrombocytosis there
GCA studies dealing with thrombocytosis [53, 56, 57, is a high risk of thrombotic involvement of major ves-
60, 6365, 6769, 73]. Therefore, our study provides sels and the microcirculation [78], and aspirin is usu-
much more reliable information on this issue than pre- ally effective at relieving vasomotor and microvascular
vious studies. occlusive symptoms, although there is only limited
The study by Kaiser etal. [72]. showed that it is the evidence that its use reduces the risk of larger vessel
degree of intimal proliferation that plays a role in the thrombosis even in those cases [10]. Also there is no
development of ischemic lesions and not an embolic evidence that anticoagulants help to prevent blindness
phenomenon. The lumen of arteries involved by GCA in GCA [56, 79].
is usually reduced by severe thickening of the intima,
and thrombus is often formed at the site of active
inflammation. Our fluorescein fundus angiographic Other Hematologic Tests
studies of GCA patients with A-AION (due to occlu-
sion of the posterior ciliary artery), central retinal To determine whether other routinely perfumed hema-
artery occlusion and cilioretinal artery occlusion indi- tologic tests can help in diagnosis of GCA, we evalu-
cate that visual loss is primarily due to occlusion of the ated various routine hematologic variables in a study
involved artery by arteritis and associated thrombosis of 121 GCA patients with positive temporal artery
and not due to embolism [5, 10, 44]. However, the pos- biopsy. Examination of the relationship between the
sibility of detached microscopic clumps of platelets or hematocrit and ESR showed a significant negative cor-
other material from the surface of a freshly forming relation (P<0.0001) (Fig.11.3). There was also a neg-
thrombus producing retinal cotton-wool spots [14], ative correlation between ESR and hemoglobin
before the thrombus completely occludes the common (P<0.0001). A lower degree of correlation was seen
trunk of the posterior ciliary artery and central retinal
artery [77], cannot be ruled out; on the other hand it is
160
important to bear in mind that such transient micro-
scopic emboli are almost invariably too small to cause 140
occlusion of main ocular arteries and produce the dev- 120
astating permanent visual loss of GCA. Thus, there is
ESR (MM/hr)
100
no convincing evidence that ischemic manifestations
80
occur as a direct consequence of reactive thrombocy-
tosis in GCA, particularly with the rather moderate 60
increases in platelets seen in GCA patients. 40 r = 0.50

As regards the role of aspirin or other platelet anti- 95% CI: (0.34, 0.66)
20
aggregating agents to prevent visual loss in GCA, to 0
date there are no studies in support of their efficacy to 20.0 24.0 28.0 32.0 36.0 40.0 44.0 48.0
do that. Gonzales-Alegre etal. [67]. also found no evi- Hematocrit (%)
dence to support the idea of starting platelet aggrega-
Fig.11.3 This scatter plot shows the relationship between the
tion inhibitors to prevent ischemic complications of
hematocrit and ESR in GCA. There was a significant negative
GCA. In our study [70], because there was no signifi- correlation between the two (r = 0.50; 95% CI: 0.35, 0.63;
cant difference in the platelet counts of patients with P<0.0001) (Reproduced from Costello etal. [70])
between platelet count and ESR (P=0.0003) (p=0.757), hematocrit (p=0.094), hemoglobin
(Fig.11.4), with a non-significant correlation between (p=0.269), and white blood cells (p=0.913).
ESR and white blood cells (P=0.104). Comparison of ESR, CRP, and hematologic vari-
In the 71 GCA patients where we tested CRP, CRP ables of GCA patients, and of A-AION with the
did not show a significant correlation with platelets NA-AION group, showed significantly (P< 0.0001)
higher median levels of ESR, CRP, platelet count, and
160 white blood cell count and lower levels of hemoglobin
140 and hematocrit in the GCA patients and A-AION than
120
in NA-AION.
The Receiver Operating Characteristic (ROC)
ESR (MM/hr)
100
curves for predicting GCA using ESR, platelet count,
80 hemoglobin, hematocrit, WBC count, and the combi-
60 nation of ESR and platelet count are shown in
40 Fig. 11.5a. ESR (AUC=0.946) was a better predictor
r = 0.22
20 95% CI: (0.02, 0.40) of GCA compared to platelet count (AUC=0.834),
hemoglobin (AUC=0.854) and hematocrit (AUC=0.841).
0
0 200 400 600 800 1000 1200 The combination of ESR and platelet count resulted in a
Platelet Count slight improvement in predictive ability (AUC=0.953)
over ESR alone. WBC was least predictive of GCA
Fig. 11.4 This scatter plot shows the relationship between
platelet count (103/ml)) and ESR in GCA. There was a weak cor-
(AUC=0.666).
relation between the two (r=0.33; 95% CI: 0.16, 0.48; P=0.0003) To be able to compare the predictive ability of CRP
(Reproduced from Costello etal. [70]) with ESR and platelet, the ROC analysis was
a 1.0 b 1.0
0.9 0.9
0.8 0.8
0.7 0.7
0.6 0.6
Sensitivity
Sensitivity
0.5 0.5
CRP (AUC=0.978)
0.4 0.4
Platelet&CRP (AUC=0.976)
ESR (AUC=0.946)
ESR (AUC=0.909)
0.3 Platelet (AUC=0.834) 0.3
Platelet&ESR (AUC=0.920)
Platelet&ESR (AUC=0.953)
0.2 Platelet (AUC=0.816)
Hemoglobin (AUC=0.854) 0.2
Hemoglobin (AUC=0.852)
Hematocrit (AUC=0.841)
0.1 WBC (AUC=0.666) 0.1
0.0 0.0
0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0
1Specificity 1Specificity
Fig.11.5 Receiver Operating Characteristic (ROC) Curves for from Costello etal. [70]). (b) ROC curves based on a subgroup
prediction of GCA from logistic regression analysis: (a) ROC of patients with CRP measured (71 GCA and 103 NA-AION),
curves based on all patients in study (121 GCA and 287 with CRP, ESR, platelet count, hemoglobin, combination of
NA-AION), with ESR, platelet count, hemoglobin, hematocrit, ESR and platelet count, or combination of CRP and platelet
WBC, or combination of ESR and platelet count as independent count as independent variables. Comparing area under the curve
variable. Comparing area under the curve (AUC), ESR is a bet- (AUC), CRP is a better predictor of GCA compared to all other
ter predictor of GCA compared to the other hematologic hematologic variables. There is no increase in AUC using
variables. There is a small increase in AUC using platelet count platelet count in combination with CRP compared to CRP alone
in combination with ESR compared to ESR alone (Reproduced (Reproduced from Costello etal. [70])
performed on a subgroup of patients that had CRP between our GCA and NA-AION patients. The former
measured (71 GCA and 103 N-AION) (Fig. 11.5b). had a significantly (P<0.0001) higher level of white
CRP(AUC=0.978) was a better predictor of GCA than blood count and lower levels of hemoglobin and hema-
ESR (AUC=0.909) and platelet count (AUC=0.816), tocrit than the latter. Thus, our study [70] suggests that
as well as the combination of ESR and platelet count in the rare cases where the ESR and CRP may not be
(AUC=0.920). The combination of platelet count with conclusive, thrombocytosis and other hematologic
CRP (AUC=0.976) did not improve predictive ability abnormalities (e.g., low hemoglobin, hematocrit) may
compared to CRP alone. be helpful in suspecting and diagnosing GCA and
The combinations of sensitivity and specificity that A-AION, although they may not, individually, add
were used to construct the ROC curve for the ESR- very much to the value of ESR and/or CRP.
platelet count combination were examined to select Other hematological tests, including serum amy-
a cut-off with desirable levels of false negative and false loid A apolipoprotein, anticardiolipin antibodies,
positive rates. Since missing a diagnosis of GCA could Willebrand factor antigen, plasma viscosity, anti-
result in visual loss in one or both eyes, we would want neutrophil cytoplasm antibodies and HLA-DR, -DQ
to have the lowest possible false negative rate while AND -DP, have occasionally been advocated in diag-
keeping the false positive rate at an acceptable level. nosis of GCA. However, none of them has so far been
Figure11.6 shows three cut-off lines, A, B, and C, that shown to be as readily available and reliable as ESR
correspond to the lowest false negative rate (4.9%, 5.6% and CRP.
and 8.9% respectively) with the false positive rate no
higher than 24.5%, 14.6%, and 9.6%, respectively. The
points that fall in the region to the right or above the line Temporal Artery Biopsy
are predicted to have GCA and the points in the region
to the left or below the line are predicted not to have Although this is considered the definitive criterion for
GCA. For example, for line B in Fig.11.4, sensitivity is diagnosis of GCA, false-negative biopsies have been
86.8%, specificity is 93.7%, with a false negative rate of reported and attributed to skip areas of arteritis in
5.6% and false positive rate of 14.6%. superficial temporal arteries. Hedges etal. [80] found
It is well-established that GCA patients have ane- 5% false-negative temporal artery biopsies for GCA;
mia [53, 57, 69]. Therefore, in our study [70] we com- however, they found that age and ESR alone or in com-
pared median levels of other hematologic variables bination with other signs or symptoms of GCA, lacked
ESR and Platelet Count in GCA and Non-Arteritic AION

1000
GCA
Non arteritic AION
900
A B C
Sensitivity 93.8% 90.7% 82.5%
800
Fig.11.6 Three selected Specificity 85.8% 92.9% 96.2%
False Positive 24.8% 14.6% 9.1%
cut-off lines for prediction of 700 False Negative 3.2% 4.4% 7.7%
GCA status based on
combination of ESR and
Platelets
600
platelet count. The cut-off C
lines, A, B, and C, corre-
500 B
spond to the three lowest
false negative rates with the A
400
false positive rates. The
points that fall in the region
to the right or above the line 300
are predicted to have GCA
and the points in the region to 200
the left or below the line are
predicted to not have GCA 100
(Reproduced from Costello 0 15 30 45 60 75 90 105 120 135 150
etal. [70]) ESR (mm/hr)
the sensitivity and specificity of biopsy results. An with GCA. Patients with negative temporal artery
important reason for false negative temporal artery biopsy but with high ESR/CRP were referred to their
biopsy is the practice of examining only a few sections internists for a thorough systemic evaluation to find
on frozen sectioning of a biopsy. Breuer etal. [81] in out the cause(s) of the elevated ESR/CRP and to
their study found that when GCA patients with positive institute appropriate management.
biopsy were compared with negative biopsy, positive Granulomatous inflammation in the temporal artery
biopsy was significantly longer than in biopsy-negative may also occur due to other systemic vasculitides and
cases (p=0.008). The rate of positive biopsies was only may result in a false-positive, misleading biopsy for
19% with biopsy length of 5mm or less, but increased GCA. The latter may be the reason that some young
to 89% when temporal artery biopsy length was longer patients are misdiagnosed as having GCA. We have
than 20mm. In our study [13] we took at least a one occasionally seen patients who have evidence of arteri-
inch piece of the superficial temporal artery and all tis (e.g., polyarteritis nodosa) in the temporal artery
biopsies were serially sectioned throughout the entire but who do not suffer from GCA.
length; we had positive temporal artery biopsy in 100% We recommend that temporal artery biopsy should
of 106 patients and a close follow-up of the remaining be done in every patient suspected of having GCA,
257 patients with negative temporal artery biopsy even if the diagnosis is evident from other findings dis-
showed no stigmata of GCA develop in any of them, cussed above. This is for two reasons: (a) GCA is
indicating that we did not miss any patients with GCA. a well-known masquerader; and (b) the only definitely
Our study indicated that failing to do meticulous serial proved treatment of GCA is prolonged steroid therapy
sectioning of the entire length of the biopsy is the most and there is a high risk of development of steroid
common cause of false negative results. For example, related systemic complications. If a patient develops
in one of the biopsies in our study, only one of about serious complications of steroid therapy, resulting in
300 serial sections of the biopsy showed definite evi- a medico-legal case, morphologic evidence of GCA in
dence of GCA. Similarly, we have had several patients temporal artery biopsy provides definite proof that the
whose local pathologist had diagnosed biopsies from treatment was appropriate.
both sides to be negative for GCA; when we got the There is frequently debate on whether temporal artery
biopsy blocks and did serial sectioning ourselves, we biopsy should be done on one or both sides, and, if on
found evidence of florid GCA in both negative biop- both sides, should it be done simultaneously, or on sec-
sies. Typical histopathological changes diagnostic of ond side only if the first is found to be negative. Boyev
GCA are shown in figure 13.5 in chapter 13. etal. [83] reported that performing a bilateral simultane-
Immunohistochemistry study of temporal artery ous or sequential temporal artery biopsy improves the
biopsy in ten patients with GCA and ten control patients diagnostic yield in at least 3% of cases while Pless etal.
showed angiotensin I receptor staining primarily in the [84] found that in 5%. Danesh-Meyer etal. [85], on the
medial layer of the temporal arteries and higher in the other hand, concluded that there is a low yield of infor-
patients with GCA than in the control patients. The mation from a second temporal artery biopsy and did not
authors postulated that angiotensin type I receptors seem to favor doing a biopsy on the other side. In this
may play a role in the development of GCA [82]. context, it is important to bear in mind that, contrary to
In our study [13], when the temporal artery biopsy the prevalent impression, temporal artery biopsy is not
results on one side were equivocal or negative for GCA an entirely benign procedure; although it is usually safe,
but there was a very strong index of suspicion for GCA a number of complications can develop from it, includ-
based on clinical findings, the biopsy was usually ing severe hemorrhage, scalp necrosis and infection.
repeated on the other side (in 76 patients second Because of the risk of complications following temporal
biopsy was positive in seven of them). It could be artery biopsy and the infrequent need to do temporal
argued that by not doing a second biopsy in the rest artery biopsy on the other side to establish diagnosis, we
with a negative biopsy we might have missed some of find no justification for doing temporal artery biopsy on
the patients with GCA. A close follow-up of all patients both sides at the same time.
with negative temporal artery biopsy in our study Another commonly debated issue is whether steroid
showed no stigmata of GCA develop in any of them, therapy alters the temporal artery biopsy results. Most
thereby indicating that we did not miss any patients of the available evidence indicates that it does not. For
example, Achkar etal. [86] in a consecutive cohort of pronounced inflammatory cell infiltration of the vessel
535 patients found that previous corticosteroid treat- wall. However, patients with no ultrasonographic
ment did not influence the positive rate of temporal changes (33%) presented histological signs of initial
artery biopsy. Ray-Chaudhuri et al. [87] found that inflammation such as isolated inflammatory cells
temporal artery biopsy after 4weeks or more of steroid around the vasa vasorum and/or in the adventitial layer.
therapy was still positive. In fact repeat temporal artery This shows that ultrasonography would miss one third
biopsy even after 9years of steroid therapy has shown of the GCA patients, which is unacceptable in view of
evidence of active disease [8892]. We have found that the fact that GCA is blinding disease. Pfadenhauer
steroid therapy does not influence the temporal artery etal. [97] found that color coded duplex sonography
biopsy results. We feel that in patients with a strong cannot differentiate between inflammatory and degen-
index of suspicion of GCA, it is dangerous to withhold erative artery disease and has spatial resolution limita-
steroid therapy till temporal artery biopsy results are tions, and that it is only a complementary method in
available, because there is every possibility that the evaluation of patients suspected to suffer from GCA
patients might have suffered irreversible visual loss in [100]. Schmid etal. [96] found the sensitivity of color
one or both eyes by the time temporal artery biopsy duplex sonography compared to temporal artery biopsy
results are available. as 50%. Prez Lpez et al. [101] stated that color-
duplex ultrasonography may be helpful in the diagno-
sis of GCA, but its evaluation of the ophthalmic artery
is less useful for diagnosis of GCA and that it had no
Doppler Tests in Diagnosis of GCA relationship with blindness. Vianna et al. [99] found
the halo sign and/or the intra-arterial filling in 55.5% of
their negative temporal artery biopsy cases. Salvarani
The following Doppler tests have been advocated by
etal. [102] found that the halo around temporal arteries
some for diagnosis of GCA:
did not improve the diagnostic accuracy of a careful
physical examination. Bley etal. [103] on a review of
the literature on the subject concluded that color-coded
olor Doppler Sonography of Superficial
C Duplex sonography is a strongly observer-dependent
Temporal Arteries imaging modality, and it may be difficult to distinguish
inflammatory from atherosclerotic mural changes.
Schmidt etal. [93] stated that a dark halo around the In our studies on GCA patients, we have found that
lumen of the temporal arteries in GCA patients is a perfectly normal and pulsating temporal arteries in
specific sign and claimed that it may be possible to patients with positive temporal artery biopsy are not
make a diagnosis GCA without performing temporal rare. Thus, the chances of false negative results with
artery biopsy. According to them, this hypoechoic halo color Doppler sonography of superficial temporal
is due to an edema of the artery wall and stenosis. They arteries are high and hence it has little role in establish-
found that in GCA this hypoechoic halo is also seen in ing a definite diagnosis of GCA to prevent blindness.
large vessels at the axillary and brachial arteries and in
temporal arteries in polyarteritis nodosa. Since then
several workers have investigated the role of color Color Doppler Imaging
Doppler sonography of superficial temporal arteries in
the diagnosis of GCA [94101]. Stammler etal. [95] Since visual loss is the most frequent and important
concluded that it remains to be proven whether sonog- complication of GCA, color Doppler imaging has been
raphy without biopsy is reliable enough for the diagno- advocated by some as a diagnostic test to evaluate the
sis and treatment of asymptomatic GCA. Schmidt etal. blood flow in the ophthalmic, central retinal and pos-
[98], based on a study of ultrasonography and biopsy terior ciliary arteries [104107]. Lauwerys etal. [104]
of superficial temporal artery in 36 patients with GCA, claimed that decreased blood flow velocity in the
found 67% of the patients with a distinct dark halo superficial temporal artery is very common in GCA
around the lumen of the temporal arteries, demon- patients and rare in PMR patients. Therefore, they
strated by ultrasonography, also showed corresponding concluded that color Doppler imaging examination
may contribute to the diagnosis of GCA. Ghanchi etal. magnetic resonance imaging, magnetic resonance
[106] stated that color Doppler imaging allows the angiography, cerebral angiography and positron emis-
detection of blood flow in a posterior ciliary artery and sion tomography [108121]. These techniques have
monitoring of alterations in orbital blood flow, and been used primarily for evaluation of vasculitides of
that it may be used to aid management decisions. large arteries. Bley etal. [119] investigated contrast-
To place the role of color Doppler imaging in GCA enhanced, high-resolution magnetic resonance imag-
in proper perspective, one has to understand some of ing in 32 patients with positive temporal artery biopsy
the very basic issues involved in visual loss in GCA. It for GCA. Mural thickness, lumen diameter, and
is well-established now that visual loss is the most a mural contrast enhancement score were assessed
dreaded complication of GCA and that arteritic AION with T1-weighted spin-echo images with submillime-
is the most common cause of visual loss in GCA (see ter in-plane spatial resolution. Evaluation of the mural
Chap. 12). It is also well-established that posterior inflammatory magnetic resolution imaging signs for
ciliary arteries are the main source of blood supply to diagnosing vasculitis resulted in a sensitivity of 80.6%
the optic nerve (see Chap. 3). It is the occlusion of and a specificity of 97.0%. The mean wall thickness
posterior ciliary arteries in GCA that produce arteritic increased significantly (P <.001), and the lumen
AION (see Chap. 12). As discussed at length in Chap. diameter decreased significantly (P <.05) for patients
6, color Doppler imaging does not provide reliable with GCA. They concluded that this test allows non-
information on posterior ciliary artery blood flow. invasive assessment of mural inflammation in vessels
Having dealt with patients with visual loss in GCA in GCA.
since 1970, I find that the readily available test of Magnetic resonance angiography has been advo-
fluorescein fundus angiography, unlike color Doppler cated by some in GCA [108, 109, 111] Harada etal.
imaging, provides precise and most reliable informa- [109] in one case showed stenosis of both superficial
tion about the circulation in the posterior ciliary arteries temporal arteries. Stanson etal. [111] reported steno-
as well as the central retinal artery [4, 5, 8, 10, 12, 14] sis of arteries of only the upper and lower extremities;
(see angiograms in Chap. 3). Fluorescein fundus however, inflammation of the arterial wall could not
angiography is the gold standard test for informa- be detected by this technique. They went on to state
tion on the circulation in the posterior ciliary artery, that standard computed tomography imaging with
central retinal artery, retina and optic nerve in GCA. contrast enhancement, and certain magnetic reso-
By sharp contrast, color Doppler imaging is not a nance imaging sequences as well as ultrasound permit
readily available technique, requires extra technical identification of the edema and inflammation of the
help and expertise, and, above all, it has multiple limi- vessel wall.
tations (discussed in Chap. 6) to evaluate posterior The use of positron emission tomography has also
ciliary artery and optic nerve head circulation and that been investigated in GCA. Fletcher etal. [114] in one
makes it an unreliable test for this purpose. patient with GCA found that positron emission
While evaluating the merits of any diagnostic test, tomography revealed diffuse abnormal F-18 fluoro-
one must ask a basic question: Is it ethically justified to deoxyglucose (FDG) arterial uptake consistent with
use a test of questionable sensitivity and specificity to GCA. After treatment with steroids, a repeat positron
diagnose GCA, the consequence (blindness) of missing emission tomography scan showed near-complete
which can be disastrous for the patient? In my opinion, resolution of the abnormal FDG uptake. de Leeuw
the answer is unequivocally NO. Thus, I find no place et al. [113] in 5 GCA patients found that positron
for Doppler tests for routine diagnosis of GCA. emission tomography visualized inflamed arteries in
all five patients, but there was a discrepancy between
positron emission tomography and angiography or
magnetic resonance angiography. In eight arteries of
Neuro-Radiological Tests for GCA four patients only positron emission tomography
showed disease involvement, while in 5 arteries of 2
More recently neuro-radiological tests for diagnosis patients only angiography or magnetic resonance
of GCA have been investigated, including computed angiography showed involvement. They concluded
tomography, computed tomography angiography, that positron emission tomography is a promising
Conclusions 193
new technique for the diagnosis of large vessel crucial information [5, 7, 1012, 14, 44, 122, 123] (see
vasculitides. Furthermore, positron emission tomog- angiograms in Chap. 3). Having done fluorescein fun-
raphy appears to be a valuable tool for the assess- dus angiography in more than 300 patients with GCA
ment of disease activity during follow-up and of the and visual loss, I find that this provides the most criti-
response to therapy. Blockmans et al. [117] evalu- cal information about the evidence of occlusion of
ated FDG uptake in the large vessels of 35 GCA those arteries or their narrowing before they are com-
patients with positron emission tomography. At diag- pletely closed. None of the Doppler or neuro-
nosis, vascular FDG uptake was noted in 83%, espe- radiological tests provide that crucial information.
cially in the subclavian arteries (74%), but also in the Most importantly, fluorescein fundus angiography is
aorta (>50%) and up to the femoral arteries (37%). not only the most reliable and sensitive test to pro-
They did not evaluate this in temporal arteries. Bley vide information about the blood flow in the ocular
etal. [118] stated that compared to positron emission arteries and their occlusion but is also much cheaper
tomography with 18F-fluoro-2-deoxy-d-glucose, and more readily available than the Doppler or
which is a very sensitive whole-body screening tool neuro-radiological tests. It is most unfortunate that it
for detecting extracranial involvement of large vessel is not used as frequently as it should be in the diagnosis
vasculitis, magnetic resonance imaging allows visu- of GCA.
alization and assessment of both the superficial cra-
nial arteries in high detail and the extracranial large
artery involvement in the same investigation. Both
et al. [120], based on their studies concluded that Conclusions
magnetic resonance imaging and positron emission
tomography are unreliable for assessing large-vessel Visual loss in GCA is the most important, devastating
inflammation in patients with GCA and pre-existing complication. The key to preventing that from
immunosuppressive therapy. Magnetic resonance happening is to establish its diagnosis early, so that
imaging is valuable for its ability to detect morpho- intensive systemic corticosteroid therapy can be insti-
logical vessel lesions, such as aneurysms and tuted immediately. Combined information from sys-
stenosis. temic symptoms, ESR and CRP provide the most useful
Optic nerve sheath enhancement with magnetic information. It is important to remember that normal
resonance imaging of the brain and orbits with gado- erythrocyte sedimentation rate does not rule out GCA
linium has been reported in isolated cases of arteritic [13]. Our study [15] on GCA has shown that 21% of
AION [110, 112] GCA patients have no systemic symptoms of GCA
Zborowska et al. [116], based on only one case, whatsoever, so that their absence does not rule out
advocated using cerebral angiography to determine the GCA. Our study [13] also showed that among the vari-
most appropriate artery to biopsy to confirm the diag- ous systemic symptoms, jaw claudication (most com-
nosis of GCA. mon), neck pain, anorexia, fever and weight loss are the
most significant symptoms. In the diagnosis of GCA,
elevated CRP was found to be the most reliable test.
Temporal artery biopsy is the gold standard but it is
Fluorescein Fundus Angiography not advisable to wait for its results to initiate steroid
therapy if other tests are strongly suggestive of GCA,
Since visual loss is the most important and dreaded because by the time biopsy results are available, the
complication of GCA due almost invariably to occlu- patient may have suffered more visual loss, which is
sion of the posterior ciliary arteries, and in some due to irreversible [21, 44]. My studies have shown that fluo-
associated occlusion of the central retinal artery or rescein fundus angiography provides the most critical
cilioretinal artery, the first essential is to find out the information about the evidence of occlusion of arteries
blood flow in those arteries in patients suspected to in GCA. By contrast, all the neuro-radiological and
have GCA from systemic symptoms and/or hemato- Doppler evaluations discussed above are highly expen-
logic tests. It is well-established that fluorescein fundus sive, not as widely available and much less reliable than
angiography is the most reliable test to provide that the routine clinical and hematological evaluations and
fluorescein fundus angiography for diagnosis of GCA. 18. Hayreh SS. Steroid therapy for visual loss in patients with
Therefore, neuroradiological and Doppler tests are not giant-cell arteritis. Lancet. 2000;355(9215):15723.
19. Hayreh SS, Jonas JB. Optic Disc morphology after Arteritic
cost-effective when compared with the other readily Anterior Ischemic Optic Neuropathy. Ophthalmology.
available, cheap and highly reliable tests for diagnosis 2001;108(9):158694.
of GCA. They are not indicated in routine clinical 20. Hayreh SS, Zimmerman B, Kardon RH. Visual improve-
practice except for a very few special cases, for research ment with corticosteroid therapy in giant cell arteritis report
of a large study and review of literature. Acta Ophthalmol
investigations or for evaluation of occlusion of large Scand. 2002;80(4):35567.
arteries in the body. 21. Hayreh SS, Zimmerman B. Visual deterioration in giant cell
arteritis patients while on high doses of corticosteroid ther-
apy. Ophthalmology. 2003;110(6):120415.
22. Salvarani C, Cantini F, Boiardi L, Hunder GG. Polymyalgia
References rheumatica and giant-cell arteritis. N Engl J Med.
2002;347(4):26171.
23. Salvarani C, Cantini F, Hunder GG. Polymyalgia rheu-
1. Kearns TP. Collagen and rheumatic diseases: ophthalmic matica and giant-cell arteritis. Lancet. 2008;372(9634):
aspects. In: Mausolf FA, editor. The eye and systemic dis- 23445.
ease. St. Louis: Mosby; 1975. p. 10518. 24. Cohen MD, Ginsburg WW. Polymyalgia rheumatica. Rheum
2. Font C, Cid MC, Coll-Vinent B, Lopez-Soto A, Grau JM. Dis Clin North Am. 1990;16(2):32539.
Clinical features in patients with permanent visual loss due 25. Healey LA. Relation of giant cell arteritis to polymyal-
to biopsy-proven giant cell arteritis. Br J Rheumatol. gia rheumatica. Baillires Clin Rheumatol. 1991;5(3):
1997;36(2):2514. 3718.
3. Hazelman BL, Bengtsson BA. Giant cell arteritis and polymy- 26. Wilke WS, Hoffman GS. Treatment of corticosteroid-
algia rheumatica. Baillires Clin Rheumatol. 1991;5:371507. resistant giant cell arteritis. Rheum Dis Clin North Am.
4. Hayreh SS. Anterior ischaemic optic neuropathy I. 1995;21(1):5971.
Terminology and pathogenesis. Br J Ophthalmol. 1974;58(12): 27. Hunder GG, Bloch DA, Michel BA, Stevens MB, Arend
95563. WP, Calabrese LH, etal. The American college of rheuma-
5. Hayreh SS. Anterior ischaemic optic neuropathy II. Fundus tology 1990 criteria for the classification of giant cell arteri-
on ophthalmoscopy and fluorescein angiography. Br J tis. Arthritis Rheum. 1990;33(8):11228.
Ophthalmol. 1974;58(12):96480. 28. Zweegman S, Makkink B, Stehouwer CD. Giant-cell arteri-
6. Hayreh SS. Pathogenesis of cupping of the optic disc. Br J tis with normal erythrocyte sedimentation rate: case report
Ophthalmol. 1974;58(10):86376. and review of the literature. Neth J Med. 1993;42(34):
7. Hayreh SS. Anterior ischemic optic neuropathy. New York: 12831.
Springer-Verlag; 1975. 29. Salvarani C, Hunder GG. Giant cell arteritis with low eryth-
8. Hayreh SS. Anterior ischemic optic neuropathy. Arch rocyte sedimentation rate: frequency of occurence in a
Neurol. 1981;38(11):6758. population-based study. Arthritis Rheum. 2001;45(2):
9. Hayreh SS. Posterior ischemic optic neuropathy. 1405.
Ophthalmologica. 1981;182(1):2941. 30. Sparrow D, Rowe JW, Silbert JE. Cross-sectional and longi-
10. Hayreh SS. Anterior ischaemic optic neuropathy differentia- tudinal changes in the erythrocyte sedimentation rate in men.
tion of arteritic from non-arteritic type and its management. J Gerontol. 1981;36(2):1804.
Eye. 1990;4(1):2541. 31. Gilbertsen VA. Erythrocyte sedimentation rates in older
11. Hayreh SS. Ophthalmic features of giant cell arteritis. patients. A study of 4,341 cases. Postgrad Med. 1965;38(5):
Baillires Clin Rheumatol. 1991;5(3):43159. A4452.
12. Hayreh SS. Acute ischemic disorders of the optic nerve 32. Bottiger LE, Svedberg CA. Normal erythrocyte sedimenta-
pathogenesis, clinical manifestations, and management. tion rate and age. Br Med J. 1967;2(5544):857.
Ophthalmol Clin North Am. 1996;9(4):40742. 33. Miller A, Green M, Robinson D. Simple rule for calculating
13. Hayreh SS, Podhajsky PA, Raman R, Zimmerman B. Giant normal erythrocyte sedimentation rate. Br Med J Clin Res
cell arteritis: validity and reliability of various diagnostic Ed. 1983;286(6361):266.
criteria. Am J Ophthalmol. 1997;123(3):28596. 34. Hanger HC, Sainsbury R, Gilchrist NL, Beard ME.
14. Hayreh SS, Podhajsky PA, Zimmerman B. Ocular manifes- Erythrocyte sedimentation rates in the elderly: a community
tations of giant cell arteritis. Am J Ophthalmol. study. N Z Med J. 1991;104(909):1346.
1998;125(4):50920. 35. Cullen JF. Occult temporal arteritis. Trans Ophthalmol Soc
15. Hayreh SS, Podhajsky PA, Zimmerman B. Occult giant cell UK. 1963;83:72536.
arteritis: ocular manifestations. Am J Ophthalmol. 1998; 36. Bethlenfalvay NC, Nusynowitz ML. Temporal arteritis: a
125((4):5216. 893. rarity in the young adult. Arch Intern Med. 1964;114:
16. Hayreh SS. Masticatory muscle pain: an important indicator 4879.
of giant cell arteritis. Spec Care Dentist. 1998;18(2):605. 37. Eagling EM, Sanders MD, Miller SJ. Ischaemic papillopa-
17. Hayreh SS. Ocular manifestations of giant cell arteritis. Am thy. Clinical and fluorescein aniographic review of forty
J Ophthalmol. 1998;126:7434. cases. Br J Ophthalmol. 1974;58(12):9901008.
References 195
38. Kansu T, Corbett JJ, Savino P, Schatz NJ. Giant cell arteritis 57. Calamia KT, Hunder GG. Giant cell arteritis (temporal
with normal sedimentation rate. Arch Neurol. 1977;34(10): arteritis) presenting as fever of undetermined origin. Arthritis
6245. Rheum. 1981;24(11):14148.
39. Wong RL, Korn JH. Temporal arteritis without an elevated 58. Riddle JM, Bluhm GB, Pitchford WC, McElroy H, Jimenea
erythrocyte sedimentation rate. Case report and review of C, Leisen J, etal. A comparative study of platelet reactivity
the literature. Am J Med. 1986;80(5):95964. in arthritis. Ann NY Acad Sci. 1981;370:229.
40. Grodum E, Petersen HA. Temporal arteritis with normal 59. Bengtsson BA. Hematological observations in giant cell
erythrocyte sedimentation rate. J Intern Med. 1990; arteritis. Acta Med Scand. 1982;658(Suppl):447.
227(4):27980. 60. De Keyser J, De Klippel N, Ebinger G. Thrombocytosis and
41. Jundt JW, Mock D. Temporal arteritis with normal erythro- ischaemic complications in giant cell arteritis. BMJ.
cyte sedimentation rates presenting as occipital neuralgia. 1991;303(6806):825.
Arthritis Rheum. 1991;34(2):2179. 61. Price N, Clearkin LG. Thrombocytosis and giant cell arteri-
42. Wise CM, Agudelo CA, Chmelewski WL, McKnight KM. tis. Lancet. 1994;343(8898):672.
Temporal arteritis with low erythrocyte sedimentation rate: a 62. Krishna R, Kosmorsky GS. Implications of thrombocytosis
review of five cases. Arthritis Rheum. 1991;34(12):15714. in giant cell arteritis. Am J Ophthalmol. 1997;124(1):
43. Neish PR, Sergent JS. Giant cell arteritis. A case with 103.
unusual neurologic manifestations and a normal sedimenta- 63. Gonzlez-Gay MA, Garca-Porrua C, Vzquez-Caruncho
tion rate. Arch Intern Med. 1991;151(2):37880. M. Polymyalgia rheumatica in biopsy proven giant cell
44. Hayreh SS, Zimmerman B. Management of giant cell arteri- arteritis does not constitute a different subset but differs
tis. Our 27-year clinical study: new light on old controver- from isolated polymyalgia rheumatica. J Rheumatol.
sies. Ophthalmologica. 2003;217(4):23959. 1998;25(9):17505.
45. Small JM, Ga Vrilescu K. The serum protein changes in giant 64. Lincoff NS, Erlich PD, Brass LS. Thrombocytosis in tempo-
cell arteritis. J Neurol Neurosurg Psychiatry. 1963;26:25761. ral arteritis rising platelet counts: a red flag for giant cell
46. Eshaghian J, Goeken JA. C-reactive protein in giant cell arteritis. J Neuroophthalmol. 2000;20(2):6772.
(cranial, temporal) arteritis. Ophthalmology. 1980;87(11): 65. Hu Z, Yang Q, Zheng S, Tang J, Lu W, Xu N, etal. Temporal
11606. arteritis and fever: report of a case and a clinical reanalysis
47. Espinoza LR, Jara LJ, Silveira LH, Martinez-Osuna P, of 360 cases. Angiology. 2000;51(11):9538.
Zwolinska JB, Kneer C, etal. Anticardiolipin antibodies in 66. Vrij AA, Rijken J, Van Wersch JW, Stockbrugger RW.
polymyalgia rheumatica-giant cell arteritis: association with Platelet factor 4 and -thromboglobulin in inflammatory
severe vascular complications. Am J Med. 1991;90(4): bowel disease and giant cell arteritis. Eur J Clin Invest.
4748. 2000;30(3):18894.
48. Berthelot JM, Lamour A, Mottier D, Pennec YL, Dueymes 67. Gonzalez-Alegre P, Ruiz-Lopez AD, Abarca-Costalago M,
M, Jouquan J, etal. Anti-neutrophil cytoplasm antibodies in Gonzalez-Santos P. Increment of the platelet count in tempo-
patients with giant cell arteritis and/or polymyalgia rheu- ral arteritis: response to therapy and ischemic complications.
matica. Ann Med Interne (Paris). 1992;143(7):42932. Eur Neurol. 2001;45(1):435.
49. Deodhar SD. C-reactive protein: the best laboratory indica- 68. Liozon E, Herrmann F, Ly K, Robert PY, Loustaud V, Soria
tor available for monitoring disease activity. Cleve Clin J P, etal. Risk factors for visual loss in giant cell (temporal)
Med. 1989;56(2):12630. arteritis: a prospective study of 174 patients. Am J Med.
50. Parikh M, Miller NR, Lee AG, Savino PJ, Vacarezza MN, 2001;111(3):2117.
Cornblath W, etal. Prevalence of a normal C-reactive pro- 69. Foroozan R, Danesh-Meyer H, Savino PJ, Gamble G,
tein with an elevated erythrocyte sedimentation rate in Mekari-Sabbagh ON, Sergott RC. Thrombocytosis in patients
biopsy-proven giant cell arteritis. Ophthalmology. with biopsy-proven giant cell arteritis. Ophthalmology. 2002;
2006;113(10):18425. 109(7):126771.
51. Weyand CM, Fulbright JW, Hunder GG, etal. Treatment of 70. Costello F, Zimmerman MB, Podhajsky PA, Hayreh SS.
giant cell arteritis: Interleukin-6 as a biologic marker of dis- Role of thrombocytosis in diagnosis of giant cell arteritis
ease activity. Arthritis Rheum. 2000;43:10418. and differentiation of arteritic from non-arteritic anterior
52. Olhagen B. Polymyalgia rheumatica. A form of senile arteri- ischemic optic neuropathy. Eur J Ophthalmol.
tis? Acta Rheumatol Scand. 1963;9:15764. 2004;14(3):24557.
53. Malmvall BE, Bengtsson BA. Giant cell arteritis. Clinical 71. Gibb WR, Urry PA, Lees AJ. Giant cell arteritis with spinal
features and involvement of different organs. Scand J cord infarction and basilar artery thrombosis. J Neurol
Rheumatol. 1978;7(3):1548. Neurosurg Psychiatry. 1985;48(9):9458.
54. Bergstrm AL, Bengtsson BA, Olsson LB, Malmvall BE, 72. Kaiser M, Weyand CM, Bjornsson J, Goronzy JJ. Platelet-
Kutti J. Thrombokinetics in giant cell arteritis, with special derived growth factor, intimal hyperplasia, and ischemic
reference to corticosteroid therapy. Ann Rheum Dis. complications in giant cell arteritis. Arthritis Rheum.
1979;38(3):2447. 1998;41(4):62333.
55. Giordano M. Thrombocytosis in giant cell arteritis. Ann 73. Cid MC, Font C, Oristrell J, de la Sierra A, Coll-Vinent B,
Rheum Dis. 1980;39(3):298. Lopez-Soto A, et al. Association between strong
56. Calamia KT, Hunder GG. Clinical manifestations of giant- inflammatory response and low risk of developing visual
cell (temporal) arteritis. Clin Exp Rheumatol. 1980;6(2): loss and other cranial ischemic complications in giant cell
389403. (temporal) arteritis. Arthritis Rheum. 1998;41(1):2632.
74. Harrell FE. The LOGIST Procedure. SUGI Supplemental 93. Schmidt WA, Kraft HE, Vorpahl K, Volker L, Gromnica-
Library Users Guide. NC: SAS Institute Inc; 1986. Ihle EJ. Color duplex ultrasonography in the diagnosis of
p. 26993. Chapter 23, Version5. Cary. temporal arteritis. N Engl J Med. 1997;337(19):133642.
75. Moncada S, Vanse JR. Arachidonic acid metabolites and the 94. Schmidt WA. Doppler ultrasonography in the diagnosis of
interactions between platelets and blood-vessel walls. giant cell arteritis. Clin Exp Rheumatol. 2000;18(4 Suppl
N Engl J Med. 1979;300:11427. 20):S4042.
76. Wu KK. Platelet hyperaggregability and thrombosis in 95. Stammler F, Ysermann M, Mohr W, Kuhn C, Goethe S.
patients with thrombocythemia. Ann Intern Med. Stellenwert der farbkodierten Duplexsonographie bei
1978;88(1):711. Patienten mit Polymyalgia rheumatica ohne klinische
77. Singh (Hayreh) S, Dass R. The central artery of the retina. I. Zeichen einer Arteriitis temporalis. [Value of color-coded
Origin and course. Br J Ophthalmol. 1960;44:193212. duplex ultrasound in patients with polymyalgia rheumatica
78. Pearson TC. The risk of thrombosis in essential thrombo- without signs of temporal arteritis]. Dtsch Med Wochenschr.
cythemia and polycythemia vera. Semin Oncol. 2002; 2000;125(42):12506.
29(3 Suppl 10):1621. 96. Schmid R, Hermann M, Yannar A, Baumgartner RW.
79. Russell RWR. Giant cell arteritis. Q J Med. 1959; Farbduplexsonographie der temporalarterien: biopsieersatz
28:47189. bei arteriitis temporalis? [Color duplex ultrasound of the
80. Hedges 3rd TR, Gieger GL, Albert DM. The clinical value temporal artery: replacement for biopsy in temporal arteri-
of negative temporal artery biopsy specimens. Arch tis]. Ophthalmologica. 2002;216(1):1621.
Ophthalmol. 1983;101(8):12514. 97. Pfadenhauer K, Weber H. Duplex sonography of the tem-
81. Breuer GS, Nesher R, Nesher G. Effect of biopsy length on poral and occipital artery in the diagnosis of temporal
the rate of positive temporal artery biopsies. Clin Exp arteritis A prospective study. J Rheumatol. 2003;30(10):
Rheumatol. 2009;27(1 Suppl 52):S1013. 217781.
82. Dimitrijevic I, Malmsjo M, Andersson C, Rissler P, 98. Schmidt D, Hetzel A, Reinhard M, Auw-Haedrich C.
Edvinsson L. Increased angiotensin II type 1 receptor expres- Comparison between color duplex ultrasonography and
sion in temporal arteries from patients with giant cell arteri- histology of the temporal artery in cranial arteritis (giant
tis. Ophthalmology. 2009;116(5):9906. cell arteritis). Eur J Med Res. 2003;8(1):17.
83. Boyev LR, Miller NR, Green WR. Efficacy of unilateral ver- 99. Vianna RN, Mansour M, Ozdal PC, Souza Filho JP,
sus bilateral temporal artery biopsies for the diagnosis of Bakalian S, Saraiva VS, etal. The role of ultrasound biomi-
giant cell arteritis. Am J Ophthalmol. 1999;128(2):2115. croscopy in predicting the result of temporal artery biopsy
84. Pless M, Rizzo 3rd JF, Lamkin JC, Lessell S. Concordance in temporal arteritis patients: a preliminary study. Eur J
of bilateral temporal artery biopsy in giant cell arteritis. Ophthalmol. 2005;15(6):6559.
J Neuroophthalmol. 2000;20(3):2168. 100. Pfadenhauer K, Weber H. Ultrasonography of the temporal,
85. Danesh-Meyer HV, Savino PJ, Eagle Jr RC, Kubis KC, periorbital and carotid arteries in the diagnosis of giant cell
Sergott RC. Low diagnostic yield with second biopsies in arteritis and its neuroophthalmological complications.
suspected giant cell arteritis. J Neuroophthalmol. Ultraschall Med. 2006;27(4):32935.
2000;20(3):2135. 101. Prez Lpez J, Solans Laqu R, Bosch Gil JA, Molina
86. Achkar AA, Lie JT, Hunder GG, OFallon WM, Gabriel SE. Cateriano C, Huguet Redecilla P, Vilardell Tarrs M.
How does previous corticosteroid treatment affect the biopsy Colour-duplex ultrasonography of the temporal and oph-
findings in giant cell (temporal) arteritis? Ann Intern Med. thalmic arteries in the diagnosis and follow-up of giant cell
1994;120(12):98792. arteritis. Clin Exp Rheumatol. 2009;27(1 Suppl
87. Ray-Chaudhuri N, Kine DA, Tijani SO, Parums DV, 52):S7782.
Cartlidge N, Strong NP, etal. Effect of prior steroid treat- 102. Salvarani C, Silingardi M, Ghirarduzzi A, Lo Scocco G,
ment on temporal artery biopsy findings in giant cell arteri- Macchioni P, Bajocchi G, etal. Is duplex ultrasonography
tis. Br J Ophthalmol. 2002;86(5):5302. useful for the diagnosis of giant-cell arteritis? Ann Intern
88. Cullen JF. Temporal arteritis. Occurrence of ocular compli- Med. 2002;137(4):2328.
cations 7 years after diagnosis. Br J Ophthalmol. 103. Bley TA, Brink I, Reinhard M. Bildgebende Verfahren bei
1972;56(8):5848. Riesenzellarteriitis (M. Horton). [Imaging procedures for
89. Cohen DN. Temporal arteritis: improvement in visual prog- giant cell arteritis (Hortons disease)]. Ophthalmologe.
nosis and management with repeat biopsies. Trans Am Acad 2006;103(4):30816.
Ophthalmol Otolaryngol. 1973;77(1):OP7485. 104. Lauwerys BR, Puttemans T, Houssiau FA, Devogelaer JP.
90. Blumberg S, Giansiracusa DF, Docken WP, Kantrowitz FG. Color Doppler sonography of the temporal arteries in giant
Recurrence of temporal arteritis. Clinical recurrence nine cell arteritis and polymyalgia rheumatica. J Rheumatol.
years after initial illness. JAMA. 1980;244(15):1713. 1997;24(8):15704.
91. Graham E, Holland A, Avery A, Russell RW. Prognosis in 105. Ho AC, Sergott RC, Regillo CD, Savino PJ, Lieb WE,
giant-cell arteritis. Br Med J Clin Res Ed. Flaharty PM, etal. Color doppler hemodynamics of giant
1981;282(6260):26971. cell arteritis. Arch Ophthalmol. 1994;112(7):93845.
92. Gouet D, Marechaud R, Le Berre D, Alcalay M, Becq- 106. Ghanchi FD, Williamson TH, Lim CS, Butt Z, Baxter GM,
Giraudon B, Boissonnot L, etal. Pronostic de larterite tem- McKillop G, et al. Colour doppler imaging in giant cell
porale traitee. Etude retrospective de 87 observations. (temporal) arteritis: serial examination and comparison
[Prognosis of treated temporal arteritis. Retrospective study with non-arteritic anterior ischaemic optic neuropathy. Eye
of 87 cases]. Presse Md. 1986;15(13):6036. (Lond). 1996;10(4):45964.
References 197
107. Ghanchi FD, Dutton GN. Current concepts in giant cell 117. Blockmans D, de Ceuninck L, Vanderschueren S, Knockaert
(temporal) arteritis. Surv Ophthalmol. 1997;42(2):99123. D, Mortelmans L, Bobbaers H. Repetitive 18 F-
108. Mitomo T, Funyu T, Takahashi Y, Murakami K, Koyama K, fluorodeoxyglucose positron emission tomography in giant
Kamio K. Giant cell arteritis and magnetic resonance cell arteritis: a prospective study of 35 patients. Arthritis
angiography. Arthritis Rheum. 1998;41(9):1702. Rheum. 2006;55(1):1317.
109. Harada S, Mitsunobu F, Kodama F, Hosaki Y, Mifune T, 118. Bley TA, Uhl M, Markl M, Frydrychowicz A, Langer M.
Tsugeno H, etal. Giant cell arteritis associated with rheu- Magnetresonanztomografie der arteriitis temporalis horton.
matoid arthritis monitored by magnetic resonance angiog- [MRI in giant cell (temporal) arteritis]. Rofo. 2007;
raphy. Intern Med. 1999;38(8):6758. 179(7):70311.
110. Lee AG, Eggenberger ER, Kaufman DI, Manrique C. Optic 119. Bley TA, Uhl M, Carew J, Markl M, Schmidt D, Peter HH,
nerve enhancement on magnetic resonance imaging in etal. Diagnostic value of high-resolution MR imaging in
arteritic ischemic optic neuropathy. J Neuroophthalmol. giant cell arteritis. AJNR Am J Neuroradiol. 2007;28(9):
1999;19(4):2357. 17227.
111. Stanson AW. Imaging findings in extracranial (giant cell) 120. Both M, Ahmadi-Simab K, Reuter M, Dourvos O, Fritzer
temporal arteritis. Clin Exp Rheumatol. 2000;18(4 Suppl E, Ullrich S, etal. MRI and FDG-PET in the assessment of
20):S4348. inflammatory aortic arch syndrome in complicated courses
112. Morgenstern KE, Ellis BD, Schochet SS, Linberg JV. of giant cell arteritis. Ann Rheum Dis. 2008;67(7):
Bilateral optic nerve sheath enhancement from giant cell 10303.
arteritis. J Rheumatol. 2003;30(3):6257. 121. Reinhard M, Schmidt WA, Hetzel A, Bley TA. Bildgebung
113. de Leeuw K, Bijl M, Jager PL. Additional value of positron der riesenzellarteriitis. Sonographie und MRT. [Imaging
emission tomography in diagnosis and follow-up of patients techniques for giant cell arteritis. Ultrasound and MRI]. Z
with large vessel vasculitides. Clin Exp Rheumatol. Rheumatol. 2009;68(2):10816.
2004;22(6 Suppl 36):S2126. 122. Hayreh SS. Inter-individual variation in blood supply of the
114. Fletcher TM, Espinola D. Positron emission tomography in optic nerve head. Its importance in various ischemic disor-
the diagnosis of giant cell arteritis. Clin Nucl Med. ders of the optic nerve head, and glaucoma, low-tension
2004;29(10):6179. glaucoma and allied disorders. Doc Ophthalmol.
115. Schmidt WA. Use of imaging studies in the diagnosis of 1985;59(3):21746.
vasculitis. Curr Rheumatol Rep. 2004;6(3):20311. 123. Hayreh SS. Ischemic optic neuropathy. Prog Retin Eye
116. Zborowska B, Ell J, McGhee-Collett M, Scolyer R, Res. 2009;28(1):3462.
McCluskey PJ. Progressive visual loss in a patient with pre-
sumed temporal arteritis despite treatment: how to make For a complete bibliography and detailed review of previously
the diagnosis. Clin Experiment Ophthalmol. 2004; published studies in the literature on the subject, please refer to
32(3):3356. bibliography in previous publications listed here.

Ophthalmic Manifestations of Giant
Cell Arteritis 12
The most important complication of giant cell arteritis neuropathy (AION) - most of them, unfortunately,
(GCA) is the visual loss and from that point of view based on lack of knowledge of the blood supply of the
GCA is the prime ophthalmic emergency, as discussed optic nerve and ONH till 1969 [2].
in previous chapters. Most authors before 1969, while describing a circu-
latory lesion in the optic nerve, made nonspecific, vague
statements as to the exact lesion. It was considered to be
a circulatory disturbance in the optic nerve due to patho-
Site of Lesion for Visual Loss
logic changes [3]. Others thought the change mainly
in Giant Cell Arteritis affected the central retinal artery. Smith and Greene
mentioned the possibility that arteritis of vessels of the
This has been a controversial topic for a long time and optic nerve and retina is responsible for this [4]. Some
requires a detailed historical discussion because it has authors stated that these small vessel changes were
important implications. Originally visual loss due to located in the anterior retrobulbar part of the optic
GCA was thought to be due to central retinal artery nerve [5, 6], while according to others, these were in the
occlusion (CRAO). Meadows in 1954 [1], however, vessels of the optic disc [1, 79]. Meadows [10] men-
found that in his 12 patients with visual loss, only one tioned that there is ischemic damage to the optic disc
had CRAO and in the rest there was ischemia of the and anterior portion of the optic nerve.
optic nerve head (ONH). It is interesting to review the Francois etal. [1116] and others [1719] postu-
old literature and find how disparate the views were on lated that it was due to occlusion of a central artery
the exact location and cause of anterior ischemic optic of the optic nerve (Fig. 12.1), an artery whose
Fig.12.1 Central artery of

the optic nerve, described by
Franois and Neetens [20].
optic nerve, CRA Central CRA CAON
retinal artery (Modified from
Fig.12.9 by Franois and
Neetens [20])

200 12 Ophthalmic Manifestations of Giant Cell Arteritis
existence they assumed. According to them, this with GCA; it is not clear about the nature of AION in
artery arises directly from the ophthalmic artery, the report by Knox etal. [49], which seems to contain
proximal to the origin of the central retinal artery a mixture of both arteritic and non-arteritic-AION
[20], enters the optic nerve behind the central retinal (NA-AION). In the reports dealing with GCA [1, 6, 7,
artery, and divides into anterior and posterior branches 3544, 46, 47, 49], involvement of the central retinal
on reaching the center of the optic nerve; the anterior artery [6, 7, 35, 37, 38, 40, 4246], the PCA [6, 7, 37,
and posterior branches run to the lamina cribrosa and 38, 40, 42, 4446, 49], and the ophthalmic artery [7,
optic foramen respectively in the center of the optic 3840, 42, 45, 49] have been mentioned. There is only
nerve. The anterior branch anastomoses with the cir- one case in which none of these arteries was involved
cle of Haller and Zinn. They stated that it is impos- [47], and this finding was explained as due either to
sible to explain the symptoms without the presence the long-term use of corticosteroids before the patient
of an individualized axial arterial system in the optic died, or to the self-limiting nature of the disease.
nerve and concluded that a central optic nerve Rodenhauser [6] is the only investigator to report
vascular system always exists [16]. My anatomic arteritis in the choroidal arteries and retinal arterioles.
studies on the blood supply of the optic nerve and Knox and Duke [48] reported focal necrosis of the
branches of the ophthalmic artery in about 100 human temporal half of the ONH and retrolaminar optic
specimens [2126] and also those of Beauvieux and nerve, with some distension of the lamina cribrosa in
Ristitch [27] and Steele and Blunt [28] failed to that region, in a patient with occlusion of the left com-
reveal a central artery of the optic nerve corre- mon carotid artery and generalized arteriosclerosis,
sponding to the description by Franois et al. [20, 2 weeks after the onset of visual disturbance. Cogan
2931]. Hence, occlusion of the central artery of the [45] also reported a case of AION with well-defined
optic nerve cannot be a factor in the etiology of infarction of the retrolaminar optic nerve in the entire
anterior ischemic optic neuropathy (AION). All this temporal half.
was before I found in my studies that the ONH was It is interesting to note that Ellenberger and
primarily supplied by the posterior ciliary arteries Netsky [50]. examined 40 optic nerves of people
(PCAs) [2] (see Chap. 3). older than 45 years for atherosclerosis and arterio-
Interference with the blood supply to the retina as sclerosis and found that changes in the optic nerve
the cause of visual loss has been considered by some vessels were similar to those seen elsewhere in the
workers [4, 32, 33] because central retinal artery pres- body, suggesting that the optic nerve vessels were
sure fell, as judged by ophthalmodynamometry. constantly involved in systemic atherosclerosis and
Uhthoff [34] and Smith and Green [4] thought it was arteriosclerosis. In AION, vascular changes in the
due to the involvement of vessels in both the optic optic nerve correlated slightly or not at all with those
nerve and retina. in the retina [51, 52].
There are many histopathologic studies on GCA Various mechanisms for the production of isch-
patients with visual loss, which have provided most emic necrosis in the anterior part of the optic nerve
important information about the location of the lesion have been postulated. Kreibig [38] considered it to
in the optic nerve [1, 6, 7, 3549]. These studies have be due to combined involvement of the axial and
shown that the area of greatest involvement in the peripheral (from short PCAs) vascular system due to
optic nerve is the area supplied by the PCAs, i.e., ret- occlusion of the ophthalmic artery and short PCAs.
rolaminar and ONH (mentioned as lamina cribrosa in Spencer and Hoyt [40] agreed with Kreibigs [38]
reports pertinent to the lamina cribrosa and prelami- hypothesis. Meadows [1], Wagener and Hollenhorst
nar regions) [6, 7, 38, 40, 4549]. The lesion begins [53], and Macfaul [46] thought that ischemic
with infarction and evolves through liquefaction necrosis is due to involvement of the short PCAs as
necrosis (in 4weeks) [49], a reactive increase in astro- they form the circle of Zinn and Haller. According
cytes and lymphocytes in 8weeks [7], and finally to to Birkhead et al. [33] anoxic damage of the optic
retrolaminar fibrosis in 4months [47]. The involved nerve is probably due to extreme narrowing of the
part of the optic nerve usually is well-defined and cir- lumen of the ophthalmic artery or central retinal
cumscribed. All but two of these reports are of patients artery.
Site of Lesion for Visual Loss in Giant Cell Arteritis 201
Experimental and Clinical Studies lesion in A-AION is the occlusion of the PCAs, which
Related to Arteritic Anterior Ischemic produces infarction not only of the ONH but also of the
adjacent retrolaminar part of the optic nerve, by virtue
Optic Neuropathy
of their distribution to the prelaminar, lamina cribrosa,
and retrolaminar regions.
GCA produces arteritic AION (A-AION). My studies Fluorescein angiographic studies in patients with
of the blood supply of the ONH [2], as well as our A-AION have provided the most critical and confirma-
experimental studies [5457] and clinical studies in tory evidence of the presence of occlusion of the PCAs
eyes with A-AION [5871] have provided the crucial in these eyes [59, 61, 62, 6466, 6871]. During the
information. As discussed in Chap. 3, the ONH is sup- acute phase, at the onset of visual loss, they showed
plied by the PCAs. Experimental studies [5457] were occlusion of the PCA, resulting in complete absence of
conducted in rhesus monkeys to investigate the patho- filling of the choroid and also of the optic disc in the
genesis of AION, in which the various PCAs were cut region supplied by the occluded PCA (see Figs.3.25,
in different combinations in 85 rhesus monkey eyes. 3.26, 3.283.34, 3.36, 3.37, 3.43, 3.49, 3.50). In
These eyes developed AION that ophthalmoscopically patients with cilioretinal artery, the retinal area sup-
showed swelling of the optic disc when the eyes were plied by this artery is also infarcted (Figs. 12.9 and
examined the day after occlusion of the PCAs (Fig.12.2) 12.10), and that was also seen in my experimental stud-
and, within 56weeks, optic atrophy. Histologic exam- ies on occlusion of the PCA in eyes with cilioretinal
ination of these eyes showed that the ONH and retro- artery (see Fig.3.4) [72]. Since the central retinal artery
laminar region (supplied by the PCA) were involved does not supply the ONH, there is no involvement of
(Fig.12.3), similar to the findings in histopathological the central retinal artery per se, unless it arises by a
studies of A-AION (Figs.12.412.7) and embolic non- common trunk with the PCA from the ophthalmic
arteritic AION (Fig.12.8). These showed that the basic artery (see Figs.3.21 and 3.22), and when the common
trunk is occluded by giant cell arteritis such eyes then
present with occlusion of both the PCA and the central
retinal artery (Figs.12.11 and 12.12). Foulds [73] also
attributed AION to interference with PCA circulation.
Thus, fluorescein angiographic studies in patients
with A-AION and experimental studies in rhesus mon-
keys established the role of occlusion of the PCAs in
A-AION. Histopathologic studies [1, 6, 7, 3549], men-
tioned above, further confirmed this, because the infarct
involved the area supplied by the PCA, i.e., ONH and
immediate retrobulbar region (Figs. 12.412.7), in all
eyes. The PCAs were frequently involved in these eyes
[6, 7, 37, 38, 40, 42, 4446, 49] - l0 of 16 case reports
mention PCA involvement, while five make no men-
tion of the absence of PCA involvement; presumably
changes in the PCAs were not recorded. Only one
case [47] showed no changes in the PCAs and the
explanation for the absence of these changes is given
above. The possible role of PCA occlusion in the eti-
ology of A-AION was also speculated by some work-
ers who found obstructed PCA histopathologically
[1, 38, 46, 53]. We now know that the site of the
lesion for visual loss in GCA almost invariably lies
Fig. 12.2 Fundus photograph of a rhesus monkey, showing
chalky white optic disc edema 2days after occlusion of the lat-
in the ONH and the adjacent retrolaminar region
eral PCA (Reproduced from Hayreh and Baines [55]) and is the result of occlusion of the PCA, although
a b
Fig.12.3 Photomicrographs of the ONH and retrolaminar optic p roducing an appearance resembling cavernous degeneration.
nerve in rhesus monkeys. (a) 36days after occlusion of all the (Massons trichrome stain) (Reproduced from Hayreh [61]) (c)
PCAs (Massons Trichrome stain) showing atrophy and degen- 42days after occlusion of the lateral PCA (Massons Trichrome
erative changes in retrolaminar region. (Reproduced from Hayreh stain) showing atrophy and degenerative changes involving the
and Baines [55]) (b) Higher magnification of a part of (a) showing temporal half of the optic nerve, with nasal half normal. (Massons
marked degeneration of neural tissue in retrolaminar region, trichrome stain) (Reproduced from Hayreh and Baines [55])
occasionally it may be due to involvement of the ret- similar to that seen in A-AION and occlusion of the
ina (as mentioned above). In the old literature, it seems PCA. In A-AION when fluorescein angiography is per-
arteritic and NA-AION were lumped together as one formed many days and weeks after its onset, similar
disease and their pathogeneses were considered to be absence of a filling of the PCA circulation is not a true
similar. guide to an earlier occurrence of a filling defect, which
In NA-AION, by contrast, there is no occlusion of is invariably seen at the time of onset of A-AION. This
the PCAs, but only delayed filling of the peripapillary is also true in many other arterial occlusive disorders,
choroid and adjacent part of the optic disc and/or the including the retinal vascular occlusions [76].
watershed zone passing through the optic disc [6466,
71, 74] (see Chap. 14). We [75] demonstrated experi-
mentally that circulation in the optic disc, peripapillary
Incidence of Ocular Involvement
choroid, and choroid (all supplied by the PCA circula-
tion) depends upon the difference between intraocular in Giant Cell Arteritis
and mean blood pressure in those vessels. This is dis-
cussed at length in the chapter dealing with pathogene- The most common ocular complication in GCA is loss
sis of NA-AION (see Chap. 14). Therefore, in NA-AION of vision in one or both eyes, which was first reported
fluorescein angiography does not show filling defect in 1937 by Horton and Magath [77] and in 1938 by
Incidence of Ocular Involvement in Giant Cell Arteritis 203
optic nerve of right eye with
4-week-old A-AION showing
a well-defined area of
infarction of the ONH and
retrolaminar region
(Verhoeffs modified elastic
MacMichael and Cullen [49])
optic nerve of a left eye with
18-day-old A-AION, showing
a sharply defined area of
necrosis in the anterior part of
the optic nerve and a partially
disintegrated lamina cribrosa.
There is presence of
hyaluronidase-sensitive
material (blue) similar to that
found in cavernous degenera-
tion of the optic nerve (Alcian
blue stain) (Reproduced from
Hinzpeter and Naumann [107])
Jennings [78]. Since then, an enormous amount of lit- [36], 33.3% [85], and 42% [83]; amaurosis fugax has
erature has accumulated on the subject. Visual loss is been reported in 20%, visual impairment in 44% and
now well established as the most dreaded and irrevers- bilateral blindness in 12% [53]. Liu et al. [86] in 45
ible complication of GCA, and that makes GCA an patients with positive temporal artery biopsy found
ophthalmic emergency. visual loss in 41 individuals (63 eyes), and it was uni-
The incidence of ocular involvement given in the lateral in 46% of the patients and bilateral in 54%
literature varies widely 38% [36, 79], 40% [1, 80], (sequential in 37%, and simultaneous in 17%); arter-
44% [53], 45% [81], 50% [4], 3357% [82], more than itic AION developed in 88% of eyes. A study by
half [83] and 70% [84]. The incidence of visual loss in a rheumatology group [87] in 161 biopsy-proven GCA
the earlier literature was 20% [39], 25% [1], 26.3% patients reported the following incidence of visual
optic nerve of left eye with
4-month-old A-AION and no
light perception vision. It
shows atrophy of prelaminar
tissue and fibrosis and gliosis
of laminar and retrolaminar
tissue. (Verhoeff elastic stain)
(Reproduced from Henkind
etal. [47])
Fig.12.7 Photomicrograph of the ONH and retrolaminar optic

nerve of an eye with A-AION showing atrophy of prelaminar
tissue and fibrosis of the laminar and retrolaminar tissue
(Reproduced by courtesy of late Paul Henkind)
Fig.12.8 Photomicrograph of the ONH and retrolaminar optic

nerve of an eye with non-arteritic AION showing infarction of
manifestations: visual ischemic complications in temporal half of the ONH and retrolaminar optic nerve
26%, and irreversible blindness, mainly due to AION (Reproduced from Cogan [45])
and frequently preceded by amaurosis fugax, in 15%.
They reported that patients with visual ischemic com-
plications had lower clinical and laboratory biologic According to them, the best predictors of any visual
markers of inflammation. HLA-DRB1*04-positive complication were HLA-DRB1*04 phenotype and the
patients had visual manifestations more commonly. absence of anemia at the time of admission. The best
a b
Fig. 12.9 Right eye with A-AION and superior cilioretinal with watershed zone above (arrow); (c) shows normal filling of
artery occlusion. (a) Fundus photograph shows opacity of the the retinal circulation in the lower half with box-carring (cattle-
upper half of the retina, marked pallor in the upper half of the trucking) in the retinal vessels above (seen typically in retinal
optic disc but no edema, and chalky white edema of the lower arterial occlusion) and uniform filling of the choroid. (d) Fundus
half of the optic disc. (b, c) Fluorescein angiograms (b) shows photograph 8 months after onset of A-AION shows optic
very slow filling of the cilioretinal artery above, normal filling of atrophy and disc cupping, with peripapillary chorioretinal
the central retinal artery below, and slow filling of the choroid degeneration
predictors of irreversible blindness (permanent visual response may be a protective factor against the devel-
loss) were amaurosis fugax and cerebrovascular acci- opment of cranial ischemic events. A later study by the
dents. The presence of other ischemic complications same group [88] reported visual ischemic manifesta-
constitutes an alarm for the development of irrevers- tions and irreversible visual loss in 22.4% and 12.5%
ible blindness. In contrast, a higher inflammatory of the 255 patients, respectively. Salvarani etal. [89] in
a b
Fig.12.10 Left eye with A-AION and cilioretinal artery occlu- plied by the lateral PCA, but no filling of the choroid and optic
sion. (a) Fundus photograph showing chalky-white optic disc disc supplied by the medial PCA as well as of the cilioretinal
edema and a patch of retinal opacity in the distribution of the artery (arrow). Note the supply by the medial PCA extends all
cilioretinal artery occlusion. (b) Fluorescein angiogram showing the way up to the fovea involving the entire optic disc
normal filling of the central retinal artery and of the choroid sup- (Reproduced from Hayreh [64])
a b
Fig.12.11 Right eye with A-AION and central retinal artery (b) Fluorescein angiogram during the late phase showing box-
occlusion. (a) Fundus photograph showing pallor with no disc carring (cattle-trucking) in the retinal vessels (seen typically in
edema and box-carring (cattle-trucking) in the retinal vessels. retinal arterial occlusion), with disc staining in the lower half
a study of 136 Italians with biopsy-proven GCA found This wide variation reported in the literature is
that the proportion of Italian patients with GCA that because the degree of loss detected depends upon
developed visual loss was similar to that reported from a number of factors, including the following: (1) The
other countries. The patients with low inflammatory most important factor is how early the diagnosis of
response had a higher risk of visual loss. GCA is established and the treatment started - the
a sooner GCA is diagnosed and managed satisfactorily,

the lower the incidence of visual loss. (2) GCA is a
well known masquerader of many systemic diseases
and vice versa. In the literature, not infrequently, the
diagnosis of GCA has been made purely on clinical
features and not confirmed by temporal artery biopsy.
This can result in inclusion of conditions other than
GCA. (3) The type of patients seen by the author(s)
(e.g., rheumatology, internal medicine, ophthalmology
or neurology) and referral patterns are other factors
responsible for the wide variation in the reported inci-
dence of blindness.
Because of all that, we conducted a study to inves-
tigate the ocular manifestations of GCA in a system-
b atic, prospective study, using the strict criterion of
positive temporal artery for GCA diagnosis [68]. There
were 170 patients with GCA proven by a temporal
artery biopsy in our study. In our series, the incidence
of ocular involvement was 50% and the incidence of
visual loss was 49%. We feel that this may reflect
a more realistic incidence, because every one of our
patients had the strict criterion of positive temporal
artery biopsy used for the diagnosis of GCA and
patients were referred from many kinds of physicians
(rheumatologists, internists, neurologists, ophthalmol-
ogists and other physicians). In our series, the patients
without ocular involvement were significantly younger
c (p=0.026), had higher ESR (p=0.0163), and were
more likely to have headache (p=0.0018), myalgia
(p<0.0001), fever (p=0.0062) and at least one of the
other systemic symptoms of GCA (p=0.0007) than
those with ocular involvement. This difference in sys-
temic symptoms between patients with and without
visual symptoms may possibly have been because the
latter group is more likely to be referred for temporal
artery biopsy because of high ESR and systemic symp-
toms than the former group. In the 85 patients with
ocular involvement, the ocular symptoms were visual
loss of varying severity in 83 (98%), amaurosis fugax
in 31%, diplopia in 6% and eye pain in 8%; ocular
ischemic lesions consisted of A-AION in 81%, CRAO
in 14%, PION in 7%, and cilioretinal artery occlusion
in 22%. In eyes with A-AION, the optic disc showed
chalky white edema in 69% during the acute phase
Fig. 12.12 Right eye with A-AION and central retinal artery
(Figs. 12.10a, 12.12a, 12.13, 12.14, 12.15b, 12.16,
occlusion. (a) Fundus photograph showing chalky-white optic
disc edema and some mild box-carring (cattle-trucking) in the reti- 12.17, 12.18a, 12.19a). The ocular symptoms and isch-
nal vessels and cherry red spot with perifoveal retinal opacity seen emic lesions were seen in a variety of combinations.
in central retinal artery occlusion. (b) Fluorescein angiogram Our study does not agree with the very low incidence
during the late phase showing box-carring (cattle-trucking) in the
of visual manifestations reports by a rheumatology
retinal vessels (seen typically in retinal arterial occlusion), with
almost no disc staining. (c) Fundus photograph showing optic group [87, 88] which is due to selection bias caused by
atrophy and optic disc cupping, with peripapillary degeneration two factors: (a) patients with visual loss invariably
consult ophthalmologists rather than rheumatologists, Second Eye Involvement in Giant

and (b) more importantly, 21% of patients with occult Cell Arteritis
GCA [69] will never consult a rheumatologist because
their sole symptom is visual loss without any systemic
symptoms. Bilateral visual loss in GCA has been reported in the
literature but that varies widely among different
reports. It also varies between the early reports com-
pared to the more recent reports. For example, Rucker
Fig. 12.13 Fundus photograph of right eye with A-AION

showing chalky-white optic disc edema with two small superfi-
cial retinal hemorrhages in upper part b
Fig. 12.15 Fundus photographs of right eye (ac)of a patient

with GCA who developed A-AION first in the left eye with no
light perception, and later, while in the hospital on high-dose
steroid therapy for its management, developed A-AION in the
right eye with no light perception (see case report in Chap. 13).
(a) Normal right eye one day after the left eye developed A-AION.
Fig.12.14 Fundus photograph of right eye with A-AION show- (b) Right eye one day after development of A-AION [(5 days
ing chalky-white optic disc edema after (a)]. (c) Right eye 4months after developing A-AION
Second Eye Involvement in Giant Cell Arteritis 209
Fig.12.17 Fundus photograph of right eye with A-AION show-

ing chalky-white optic disc edema, a small area of adjacent retinal
infarction (arrow) due to associated occlusion of tiny cilioretinal
artery and small peripapillary hemorrhage in inferior nasal region
involvement of the two eyes was mentioned as 1 day to

1 week - the longest interval being 4 weeks. More
recently, Liu etal. [86] in 1994 reported a retrospective
review of records of 41 GCA patients with visual loss;
three developed visual loss in the fellow eye while on
high-dose corticosteroid therapy within days of initial
visual loss. In our series [68], reported in 1998, bilat-
eral visual loss at the initial visit was seen in 27 (32%)
of 85 GCA patients. Of the 27 patients, 24 could give
definite information of the time interval in onset of
visual loss in the two eyes; 17% of them became aware
of the visual loss in both eyes at the same time, whereas
in the rest of the patients, the time interval between the
visual loss in the two eyes was 17 days in 46%,
814 days in 8%, 1530 days in 8%, 12 months in
Fig.12.16 Fundus photograph of right eye with A-AION showing 12.5%, and more than 6months (7 and 9months) in
chalky-white optic disc edema (Reproduced from Hayreh [61]) 8%. In all our patients who complained of simultane-
ous loss of vision in both eyes, fundus examination
revealed that in fact one eye had older changes than the
[90] in 1956 stated that when the vision is lost in one, newly involved eye. This indicates that these patients
chances that it will be lost in the other eye within the were not aware of visual loss in the first eye until the
next few days or weeks are even. Mosher [91] in 1959 second eye was involved, giving an erroneous impres-
reported bilateral moderate visual loss in 6%. Wagener sion of simultaneous visual loss in both eyes. I have
and Hollenhorst [53] in 1958 stated that if the second seen patients where premature reduction of corticos-
eye is not affected for 2months, the prognosis is good. teroids below the critical dosage or its stoppage by the
In a Mayo Clinic series [53, 92] in 1960, bilateral local internists resulted in visual loss in the previously
visual loss was seen in 29% of 73 GCA patients. uninvolved second eye from within a few days to about
This was described as simultaneous bilateral loss in 2 months. This raises an important issue in the
six patients, and the usual interval between the management of corticosteroid therapy in GCA patients
a a
Fig. 12.18 Fundus photograph (a) and fluorescein fundus

angiogram (b) of right eye with A-AION and occlusion of a
small cilioretinal artery during the initial stages. (a) Fundus pho-
tograph shows chalky white optic disc edema with retinal infarct Fig.12.19 Fundus photographs of left eye with A-AION. (a) It
in the distribution of occluded cilioretinal artery (arrow) and shows chalky white optic disc edema and a cotton-wool spot
small inferior temporal peripapillary hemorrhage. (b) Fluorescein (arrow) during the acute phase. (b) It shows later on optic disc
fundus angiogram shows evidence of occlusion of the medial atrophy and cupping
PCA and no filling in the region of the cilioretinal artery (arrow)
to prevent visual loss. There is an almost universal experience with some 300 GCA patients with and
practice among rheumatologists to guide steroid ther- without ocular involvement, we have found that patients
apy purely based on systemic symptoms [93, 94] rather will retain their existing vision and will not get the sec-
than ESR and CRP levels; however, our study [95, 96] ond eye involved, if they do not suffer any further
showed that is a dangerous strategy with a risk of the visual loss within the first week after starting adequate
second eye losing vision (see Chap. 13). systemic corticosteroid therapy, when satisfactory
Thus, the incidence of initial ocular involvement as therapy is maintained [96, 97]. Thus, blindness due to
well as involvement of the second eye in GCA, very GCA is preventable in the vast majority of cases, if
much depends upon when the diagnosis is made, how diagnosed early and treated properly. Our systemic
early the patient is seen, and how aggressively corticosteroid therapeutic regimen in patients with
systemic corticosteroid therapy is used. In our GCA is discussed elsewhere [96] and in Chap. 13.
Optic Nerve Ischemic Lesions 211
Various Ophthalmic Manifestations complication of GCA. It has been reported by a large

of Giant Cell Arteritis number of publications; however, the frequency of
amaurosis fugax has been reported variably. For exam-
ple, Hollenhorst etal. [92] reported amaurosis fugax in
The ocular lesions seen in GCA are ischemic in nature 47% of their 73 patients and 50% of them later devel-
due to thrombosis of one or more of the PCAs supplying oped permanent visual loss, Gonzlez-Gay etal. [98].
the eye and optic nerve by granulomatous inflammation. in 14.6% of 52 patients, Cid etal. [99] in 32.3% of 32
GCA is a disease of the medium-sized and larger arteries. patients, Gonzlez-Gay et al. [87] in 14.9% of 161
As discussed above, histopathologic studies have reported patients, Nesher etal. [100] in one of 31 consecutive
involvement of the PCAs, ophthalmic artery, nutrient patients (but they found it later in 4 of the 5 patients
arteries of the optic nerve, or central retinal artery by with permanent visual loss), and Schmidt etal. [101]
GCA. Among the orbital arteries, GCA has a special pre- in 46%. In our study [68] of 170 GCA patients, in
dilection for the PCAs, and that has also been shown by most cases, it preceded the visual loss. It was reported
our [59, 68] fluorescein fundus angiographic studies (see by 31% of the 85 patients with ocular involvement or
Figs.3.25, 3.26, 3.283.34, 3.36, 3.37, 3.43, 3.49, 3.50). 27% of the involved eyes; in 64% of these eyes, amau-
The PCAs always supply the choroid and ONH, and in rosis fugax preceded the development of permanent
a number of eyes also give out the cilioretinal artery. visual loss - in 39% of the A-AION, 15% of the CRAO,
3% of the posterior ischemic optic neuropathy and 6%
of those with A-AION combined with cilioretinal
Classification of Ophthalmic Ischemic artery occlusion. Our study showed that the com-
monest cause of amaurosis fugax in GCA is tran-
Lesions of GCA
sient ischemia of the ONH and, contrary to the
prevailing impression, not of the retina, because the
These may be classified according to various anatomi- incidence of anterior or posterior ischemic optic
cal parts of the eye involved, e.g., the optic nerve, retina, neuropathy is much higher than that of CRAO in
choroid, anterior segment and extraocular muscles. these cases (88% versus 13%). The most common
As mentioned above, visual loss is the most feared factor precipitating amaurosis fugax in our study was a
and irreversible complication of GCA. Therefore, oph- change in posture, e.g., suddenly standing up from
thalmologists are likely to be the first physicians con- sitting position and associated with orthostatic hypoten-
sulted by GCA patients with visual loss, especially sion, or stooping down with associated increase in
those with occult GCA [69] who have no associated intraocular pressure in the eyes; both these factors
systemic symptoms at all. Apart from the visual loss, reduce the perfusion pressure in the ONH vessels and
patients with GCA can also develop other ophthalmic further compromise the already poor circulation in the
complaints which may be early manifestations of ONH [102].
GCA. In view of all this, it is essential that ophthal- Raymond etal. [103] reported one patient with GCA
mologists must be fully aware of the various ophthal- complaining of fleeting visual blurring with heat or exer-
mic manifestations of GCA, to enable them to suspect cise (Uhthoffs symptom). Finelli [104] reported the
and diagnose GCA early and initiate corticosteroid case of a 77-year-old man who had numerous episodes
therapy immediately, to prevent further visual loss. of transient alternating loss of vision for several days.
The following is a brief account of the various The findings indicate that amaurosis fugax is an
ophthalmic manifestations of GCA. important early visual symptom of GCA and an omi-
nous sign of impending blindness.
Amaurosis Fugax
Optic Nerve Ischemic Lesions
Usually there is sudden, dramatic visual loss due to
GCA; however, in some cases it is preceded by amau- These consist of arteritic anterior and posterior isch-
rosis fugax. Amaurosis fugax is a well recognized emic optic neuropathies.
Arteritic Anterior Ischemic Optic lateral PCA correspondingly supplied the other part
Neuropathy (A-AION) of the ONH or the entire ONH (see Fig. 3.24) in
about 40%, and that agrees with the marked interin-
dividual variation reported in the blood supply of the
This is by far the most common and devastating ocular ONH by the PCAs [65, 105, 106]. Since A-AION is
complication of GCA, resulting in sudden, permanent, the major cause (in 81%) of visual loss in GCA,
partial or complete visual loss, involving one or both occlusion of the medial PCA by GCA seems to be
eyes. The reported incidence of A-AION in GCA in mainly responsible for that. When only the medial or
the literature varies between 18% and 88% [61]. In our lateral PCA is occluded, that usually results in seg-
series [68], 81% of 85 patients or 76% of 123 eyes mental involvement of the ONH (see Figs. 3.10,
developed A-AION. 3.27, 3.30, 3.323.36, 3.38, 3.39), but when both
PCAs are occluded or the occluded PCA supplies
the entire ONH (see Figs. 3.26, 3.25, 3.28, 3.29,
Pathogenesis of A-AION 3.31, 3.37, 3.49, 3.50), that results in involvement of
the entire ONH. Since A-AION is a thrombotic dis-
A-AION, as discussed above, is due to ischemia of order, it results in much more severe ischemic dam-
the ONH [55, 58]. PCAs are the main source of blood age than in NA-AION, which is NOT a thrombotic
supply to the ONH (see Chap. 3) [2, 65, 105, 106]. disorder but primarily a hypotensive disorder (see
For some unknown reason, GCA has a special predi- Chap. 14), resulting in massive visual loss in one or
lection to involve the PCA, resulting in its throm- both eyes in A-AION.
botic occlusion. Therefore, occlusion of the PCA The most important fact in this context is that
results in infarction of a segment or the entire ONH, AION is a common disease in the middle-aged and
depending upon the area of the ONH supplied by the elderly population and etiologically is of two types:
occluded PCA, which shows marked interindividual (i) A-AION due to GCA, and (ii) NA-AION due to
variation (see Figs.3.24, 3.25, 3.283.34, 3.36, 3.37, other causes. NA-AION is the more common of the
3.43, 3.49, 3.50). That results in development of two types and is one of the most prevalent visually
A-AION. In fact, in our studies, in all patients with crippling diseases in the middle-aged and elderly.
visual loss due to GCA fluorescein fundus angiogra- Although the two types of AION have many similar
phy showed evidence of PCA occlusion (see clinical features and presentation, their management
Figs. 3.25, 3.26, 3.283.34, 3.36, 3.37, 3.43, 3.49, is entirely different. A-AION is an ophthalmic emer-
3.50) [58, 59, 64, 66, 68, 102]. That has also been gency and these patients require immediate and
demonstrated by histopathologic studies of eyes gone aggressive treatment with high-dose systemic corti-
blind due to GCA [40, 4547, 49, 107], and has been costeroids to prevent further visual loss; visual loss
confirmed by experimental occlusion of the PCA in GCA is preventable. In sharp contrast to that,
(Fig.12.2) [54, 55]. NA-AION is not an ophthalmic emergency. Therefore,
In our study [68] of fluorescein fundus angiogra- when a patient is diagnosed as having AION, the
phy of eyes during the early stage of A-AION, there first crucial step is to identify immediately whether it
was occlusion of the medial PCA in 36% (see is arteritic or non-arteritic. That naturally raises the
Figs.3.283.32), lateral PCA in 8% (see Figs.3.24, question of how to differentiate between the two types
3.333.37), and both medial and lateral PCAs in of AION.
56% of the eyes (see Figs.3.26 and 3.25) thus the
medial PCA is the artery most commonly involved
by GCA. In some eyes choroidal watershed zone fill- Differential Diagnosis of A-AION and NA-AION
ing defects were seen (see Figs. 3.38, 3.40, 3.55)
[74]. In that study, fluorescein fundus angiography The differential diagnosis of the two types of AION
almost always revealed involvement of the PCAs by is discussed at length elsewhere [66, 67, 71, 102].
the GCA; it also revealed that the medial PCA sup- Following is a brief account. I have found that the
plied the entire (see Figs. 3.28, 3.29, 3.31, 3.37, collective information provided by the following
3.49, 3.50) or a part (see Figs. 3.10, 3.27, 3.30, criteria helps to differentiate the two types of AION
3.323.36, 3.38, 3.39) of the ONH in 93% and the reliably.
1. Systemic Symptoms of GCA However, the presence of perfectly normal visual

These are discussed in detail in Chap. 10. However, acuity does not rule out A-AION.
21.2% with occult GCA have no systemic symptoms 5. Chalky White Optic Disc Edema
of any kind, ever, and visual loss is the sole complaint This is seen commonly in A-AION (Figs. 12.10a,
[69]. Therefore, absence of systemic symptoms does 12.12a, 12.13, 12.14, 12.15b, 12.16, 12.17, 12.18a,
not rule out GCA this is an extremely important con- 12.19) [59, 61, 62, 66, 68, 71, 102]. It was seen in 69%
sideration, because in some patients GCA is missed for of the A-AION eyes in our study [68]. The presence
that reason, with catastrophic visual loss. Patients with ofchalky white optic disc edema is almost diagnostic
NA-AION have no systemic symptoms of GCA. of A-AION. In NA-AION, chalky white optic disc
2. Visual Symptoms edema occurs only very rarely with embolic occlusion
As discussed above, amaurosis fugax in a patient with of the PCA.
AION is highly suggestive of A-AION [68]. Amaurosis 6. AION Associated With Cilioretinal Artery Occlusion
fugax is extremely rare in NA-AION. This also is almost diagnostic of A-AION (Figs.12.9,
3. Hematologic Abnormalities 12.10, 12.17, 12.18) [59, 61, 66, 68, 71, 102]. This is
These are discussed in detail in Chap. 11. Immediate because both the ONH and cilioretinal artery derive
evaluation of erythrocyte sedimentation rate (ESR) their blood supply from the PCA, and occlusion of the
and C-reactive protein (CRP) is vital in all AION PCA naturally results in both lesions.
patients aged 50 and over THIS IS A MUST. 7. Evidence of Occlusion of the PCA on Fluorescein
Elevated ESR and CRP, particularly CRP, is helpful in Fundus Angiography
the diagnosis of GCA. Patients with NA-AION do not If angiography is performed during the first few days
show any of these abnormalities, except when a patient after the onset of A-AION, the choroid supplied by
has some other intercurrent systemic disease. There is one or more of the PCAs does not fill [59, 61, 65, 66,
almost a universal impression that the presence of nor- 68, 71, 102] (see Figs. 3.24, 3.25, 3.283.34, 3.36,
mal ESR rules out GCA. But I have had temporal artery 3.37, 3.43, 3.49, 3.50) This is almost diagnostic of
biopsy confirmed elderly GCA patients with ESR as A-AION as well. However, later on, with the establish-
low as 45mm/h. Therefore, the presence of normal ment of collateral circulation, this information may be
ESR does not rule out GCA this is an extremely lost. In NA-AION, in my studies, I found that such a
important fact, because in some patients GCA is missed non-filling of a PCA on angiography is not seen, except
for that reason, with catastrophic visual loss. in an extremely rare case when there is embolic occlu-
4. Early, Massive Visual Loss sion of the PCA (see Figs.3.27 and 3.35).
Table12.1 gives the initial visual acuity in NA-AION 8. Temporal Artery Biopsy
[108] and A-AION [68] eyes in our studies. This shows This finally establishes the diagnosis and its role is dis-
that early, massive visual loss is extremely suggestive cussed in Chap. 11.
of A-AION; however, the presence of perfectly normal Although none of these findings is present in 100% of
visual acuity does not rule out A-AION. the GCA patients, combined information from all of them
put together should establish the diagnosis of A-AION.
Table12.1 Initial visual acuity in NA-AION and A-AION in Thus, clinically the following 3 signs in particular,
our series when present, are virtually diagnostic of A-AION
NA-AION (n=386 eyes) A-AION (n=123 eyes) and are not seen in NA-AION:
Visual acuity Percentage Visual acuity Percentage 1. Chalky white optic disc edema.
20/1520/30 46 20/40 or better 21 2. AION associated with cilioretinal artery
20/4020/60 18 20/5020/60 8 occlusion.
3. Evidence of complete occlusion of the PCA on
20/7020/100 11 20/7020/100 9
fluorescein fundus angiography.
20/20020/400 11 20/20020/400 8
Count fingers 14 Count fingers 54
or worse or worsea
Clinical Features of A-AION
a
Of the 66 eyes in this group, 47% had count fingers to hand
motion visual acuity, 24% had only light perception and 29% There is sudden dramatic visual loss. It is not unusual
no-light perception to find that the visual loss and/or progression in visual
loss is discovered on waking up from sleep in the ischemia due to PCA occlusion. As discussed in Chap.
morning or from a nap. This suggests that nocturnal 3, there is marked interindividual variation in the area
arterial hypotension [109, 110] helps to complete a of supply in the ONH by the various PCAs. For exam-
thrombotic occlusion by stasis or a drop in perfusion ple, occlusion of the medial PCA in GCA may: (a) not
pressure below the critical level in the partially involve the ONH at all (see Fig.3.41) and consequently
occluded PCA; that compromises the circulation for a cause no visual loss, or (b) it may involve the whole of
long enough period to produce infarction of the ONH the ONH and cause complete infarction of the ONH
and A-AION. (see Figs.3.28, 3.29, 3.31, 3.50) resulting in complete
loss of vision, or (c) it may involve only any one part of
the ONH (see Figs. 3.27, 3.30, 3.32, 3.43), in which
Age, Gender and Race case the extent of the visual loss depends upon which
part of the ONH is involved. Similarly, occlusion of the
Since GCA invariably is cause of A-AION, these fac- lateral PCA may: (a) not involve the ONH at all (see
tors (discussed in chapter 9) play important role in Figs.3.37 and 3.49), or (b) may involve the whole of it
A-AION as well. Thus, A-AION is a disease of late (see Fig.3.24), or (c) only one part of it (see Figs.3.33
middle-aged and elderly persons, more common in 3.36). Therefore, unfortunately, the common concept
women and mostly seen in Caucasians. In our study that occlusion of the PCA in GCA causes a similar type
[68] of 85 GCA patients with A-AION, meanSD age of visual loss in all eyes is not valid. For example, in
was 76.27.0 (range 5793years), 71% were women the eye in Fig. 3.32, occlusion of the medial PCA
and 29% men, and all but one Caucasians. involved all the ONH except a small part of the tempo-
ral region (supplied by the normal lateral PCA), and
the eye had 20/20 visual acuity and tiny central visual
Ophthalmic Symptoms field defect. On the other hand, occlusion of the medial
PCA in Figs.3.28, 3.29, 3.31, 3.50 involved the entire
As discussed above, amaurosis fugax is an important ONH and resulted in no light perception.
visual symptom and an ominous sign of impending
visual loss in GCA [68]. Most patients with GCA
develop visual loss suddenly without any warning. Visual Acuity
Simultaneous bilateral visual loss has been reported;
however, as discussed above, our study [68] indicated, Table12.1 gives the initial visual acuity in our series of
that it generally represented cases where the patient 123 eyes with visual loss due to GCA [68]. Thus,
was unaware of vision loss in one eye until the second although usually there is a marked deterioration of
eye is also involved. Also, the incidence of bilateral visual acuity in GCA, almost normal visual acuity does
involvement depends upon how early the patient is not rule it out. Visual loss in 76% of these eyes was due
seen, when the diagnosis is made, and how aggres- to A-AION. Liu etal. [86] in 55 eyes with A-AION
sively systemic corticosteroid therapy is used the reported the visual acuity 20/200 or worse in 70% and
longer the time interval from the onset of visual symp- 21% no light perception. In our series [68] of 123 eyes
toms in one eye without adequate corticosteroid ther- with visual loss, visual acuity was 20/200 or worse in
apy, the higher the risk of second eye involvement. 62% and no light perception in 15%.
I have seen a rare patient with GCA suffering from
euphoria and even denying any visual loss.
Visual Fields
Visual Loss in GCA At mentioned above, the type and extent of the visual
field defect depends upon the extent and location of the
In the vast majority of cases, the visual loss in GCA is part of the ONH involved by the occluded PCA, which
due to A-AION. The extent and severity of visual loss varies widely. Compared to NA-AION, the visual defects
(i.e., field defects and visual acuity) depend upon the are much more extensive and severe in A-AION because
severity and location of ONH damage caused by of more extensive and severer ONH ischemic damage.
Optic Disc Changes contrast, in NA-AION such cupping of the optic disc
is very rare. Danesh-Meyer etal. [111] in a study of
1. During the acute phase: Initially the optic disc 92 A-AION and 102 NA-AION found optic disc
develops edema in all eyes. Optic disc edema, com- cupping in 92% in A-AION and in only 2% of
pared to NA-AION, usually has a diagnostic appear- NA-AION. These authors [112] also compared cup
ance in A-AION, i.e., chalky white color (Figs.12.9a, size in involved versus the uninvolved eyes in patients
12.10a, 12.12a, 12.13, 12.14, 12.15b, 12.16, 12.17, with unilateral A-AION and unilateral NA-AION; in
12.18a, 12.19) [59, 61, 62, 66, 68, 71, 102]. In our eyes affected with A-AION there was a significant
series [68], this was seen in 69% of the eyes with excavation and enlargement of the optic cup when
A-AION during the acute phase, with normal reti- compared with contralateral uninvolved eyes but
nal circulation, except when there was associated not so in NA-AION.
cilioretinal artery (Figs.12.9, 12.10, 12.17, 12.18)
or CRAO (Figs. 12.11 and 12.12). As discussed
below, when there is combined CRAO and A-AION, Fluorescein Fundus Angiography
there may not be any optic disc edema (Fig.12.11a).
Similarly Fig.12.9a shows chalky white optic disc As a routine I do fluorescein fundus angiography in
edema only in the lower part of the optic disc but all eyes with fresh AION because it provides extremely
not in the upper part where cilioretinal artery occlu- useful information. My studies [59, 61, 62, 6568, 71,
sion resulted in infarction of the entire upper half of 102] on A-AION, showed evidence of occlusion of
the retina. As discussed below, infarction of the PCA. Our latest study [68] showed that lateral and/or
retina in these conditions results in death of all gan- medial PCA circulation was abnormal in all eyes. In a
glion cells, resulting in absent axoplasmic flow and few eyes when an additional very small superior
consequently no optic disc edema. I have had the (Fig.12.20) or inferior or both superior and inferior
opportunity to see rare cases soon after the onset of PCAs were present, those were not involved. When
visual loss due to A-AION when the disc was only angiography was performed soon after the onset of
hyperemic with mild edema, later on changing to visual loss, the involved eye invariably showed
chalky white edema; in one case immediately after
onset, at first examination in the morning the disc
was hyperemic and at 4PM it was chalky white.
2. A
fter resolution of optic disc edema: With resolution
of optic disc edema, optic atrophy usually develops
within 68weeks; the optic disc in the vast majority
shows cupping indistinguishable from that seen in
glaucomatous optic neuropathy (Figs.12.9d, 12.12c,
12.15c, 12.19b) [5962, 68, 71, 111, 112] except that
the disc rim is pale whereas it is of normal color in
glaucomatous optic neuropathy. Figure12.15 of the
right eye is an excellent example of development of
cupping of the disc in A-AION. This patient was first
seen with A-AION in the left eye. Five days later he
developed A-AION in the right eye. Figure12.15a is
fundus photograph of the right eye when the eye was
normal; Fig. 12.15b is 3 days after the A-AION
developed in that eye and shows optic disc edema;
and Fig.12.15c is 4months later when the optic disc
developed cupping (see detailed report of this patient
in Chap. 13). If eyes with A-AION are seen late,
Fig. 12.20 Fluorescein fundus angiogram of an eye with
cupping of the optic disc may result in a misdiagno- A-AION, showing the sites of entry of two superior short PCAs
sis of old, burned out glaucoma and visual loss. By and their course in the choroid but not of any other PCA
complete occlusion of the PCA (see Figs.3.24, 3.25, Retinal Ischemic Lesions
3.283.34, 3.36, 3.37, 3.43, 3.49, 3.50); however,
when angiography was not performed within the first
These include central or cilioretinal artery occlusion or
few days of onset of visual loss, there was only mod-
cotton-wool spots.
erate to marked delay in filling of the choroidal vascu-
lar bed in the distribution of the involved PCA,
depending upon the time gap between the onset of
A-AION and angiography. The circulatory abnormal- Central Retinal Artery Occlusion (CRAO)
ity involved the medial PCA only in 36%, lateral PCA
only in 8%, and both lateral and medial PCAs in 56%. The incidence of CRAO in GCA has been given in the
There was marked interocular variation in the supply literature between 2% and 18% [53, 68, 92]. In our
by the medial and lateral PCAs in the choroid and the study [68], this was seen in 14% (12 of 85) of the
optic disc (see Figs.3.24, 3.25, 3.283.34, 3.363.38, patients or 12% (15 of 123) of the eyes. CRAO is almost
3.40, 3.43, 3.49, 3.50) [65, 74, 102, 106]. These find- invariably combined with PCA occlusion the latter
ings indicate that the major source of blood supply to detected only on fluorescein fundus angiography. In our
the ONH is usually the medial PCA (but not always) studies [59, 68], while ophthalmoscopy revealed a clas-
[68]. A few eyes showed only a choroidal watershed sical picture of CRAO with or without optic disc edema
zone [74] filling defect between the two PCAs, like a in these cases, fluorescein fundus angiography in fact
non-fluorescent band (see Figs. 3.38, 3.40, 3.41). revealed a combined occlusion of the central retinal
Thus, fluorescein angiography almost always revealed artery and one of the PCAs (Figs.12.11b and 12.12b).
involvement of the PCAs by the GCA. This provides This is because the central retinal artery arises from the
very useful information and angiography must be ophthalmic artery by a common trunk with one or more
performed in all eyes with acute AION without of the PCAs (Figs.12.21 and 12.22, see asterisks for the
exception. It helps not only to determine the circula- common trunk) in 60% of human cases [22, 24]. In such
tory defect, but also most importantly to differentiate cases, if arteritis involves the common trunk and causes
arteritic from NA-AION. Fluorescein fundus angiog- its thrombosis and occlusion, that results in occlusion of
raphy during the early stages constitutes a critical both the PCA (manifesting as A-AION) and the CRA.
diagnostic test for A-AION. It is unfortunate that The absence of optic disc edema in eyes with combined
most neuro-ophthalmologists do not do angiography CRAO and A-AION (i.e., PCA occlusion) (Fig.12.11a)
and miss important information. That is responsible is due to stoppage of axoplasm formation by the isch-
for a good deal of confusion and controversy on the emic retinal ganglion cells, and we know that optic disc
subject of AION. In all our studies of A-AION cases edema in AION is due to axoplasmic flow stasis at the
[59, 61, 62, 6568, 71, 102], as discussed above, ONH [118]. Similarly Fig. 12.9a shows chalky white
absent filling of the PCA on angiography has been optic disc edema only in the lower part of the optic disc
well documented. but not in the upper part where cilioretinal artery occlu-
sion resulted in infarction of the entire upper half of the
retina. In some eyes with both CRAO and PCA occlu-
sion, the possibility that the two arteries may be inde-
Arteritic Posterior Ischemic Optic
pendently involved by the arteritic process cannot be
Neuropathy (A-PION) ruled out. Also, the possibility of an eye developing
CRAO only due to GCA cannot be ruled out. Therefore,
The posterior part of the optic nerve may be involved as a rule, when persons 50years or older present with
by acute ischemic lesions caused by GCA and produce CRAO, fluorescein fundus angiography must be
A-PION [113115]. This is due to occlusion of orbital performed to find out if there is underlying PCA occlu-
arteries which supply small nutrient arteries to this part sion as well, because its presence is almost diagnostic
of the nerve (see Figs.3.3, 3.153.18). In GCA, devel- of CRAO due to GCA. Such patients require immediate
opment of A-PION been described in the past [33, 53, and aggressive corticosteroid therapy to prevent cata-
84, 92, 116, 117], under different eponyms. It is dis- strophic visual loss - which is preventable with adequate
cussed at length in Chap. 21. corticosteroid therapy (see Chap. 13).
Retinal Ischemic Lesions 217
Fig. 12.21 Example of a common trunk of origin of central

retinal artery and PCA from the ophthalmic artery as seen from
below central retinal artery and lateral PCA arising by a com-
mon trunk. CRA Central retinal artery, LPCA Lateral PCA, OA
Ophthalmic artery, PPS Point of penetration into the sheath by Fig. 12.22 Example of a common trunk of origin of central
CRA, * Common trunk of origin of CRA and PCA retinal artery and PCA from the ophthalmic artery as seen from
below central retinal artery and medial PCA arising by a com-
mon trunk. CRA Central retinal artery, LPCA Lateral PCA, MPCA
Cilioretinal Artery Occlusion Medical PCA, OA Ophthalmic artery, ON Optic nerve, * Common
trunk of origin of CRA and PCA (Reproduced from Hayreh
[71])
The cilioretinal artery arises directly or indirectly from
the PCA. Since GCA has a special predilection for picture of GCA, i.e., a combination of chalky white
involving the PCAs, it is not surprising to see simulta- optic disc edema, retinal infarct in the region of the
neous development of both A-AION and cilioretinal occluded cilioretinal artery and PCA occlusion on
artery occlusion, when an eye has a cilioretinal artery fluorescein angiography. Unfortunately, cilioretinal
(Figs.12.9, 12.10, 12.17, 12.18) [59, 62, 66, 68, 71]. artery occlusion has often been misdiagnosed as
In our series [68], this was seen in 22% of the patients a branch retinal artery occlusion with GCA [119],
or 21% of the eyes. In one of our patients cilioretinal but the so-called branch retinal arteries are in fact
artery occlusion was the sole presentation of GCA, arterioles, and GCA is a disease of the medium-sized
without any associated A-AION, because the occluded and large arteries and not of the arterioles; hence, it is
lateral PCA (the source of the cilioretinal artery) in illogical to have branch retinal artery occlusion due to
that eye did not supply the ONH. The area of the retina GCA [68, 96]. Because of that reason, I have never
supplied by a cilioretinal artery can vary widely from seen a true branch retinal artery occlusion with GCA.
large (Fig.12.9) to only tiny (Figs.12.17 and 12.18) I have seen patients with cilioretinal artery occlusion
These eyes present with a classical, diagnostic clinical misdiagnosed by ophthalmologists as ordinary branch
retinal artery occlusion and left untreated, resulting in Other Retinal Lesions
catastrophic visual loss in both eyes, which could have
been prevented, if the possibility of GCA as one of its
In addition to these well-known retinal lesions, in the
causes had been borne in mind. Thus, as a rule, in all
literature there is an occasional mention of the pres-
patients 50years and older, it is essential to rule out
ence of phlebitis of one of the retinal veins [77, 121] or
GCA in all patients with cilioretinal artery occlusion
of central retinal vein occlusion [53, 84]. I have never
(or so-called branch retinal artery occlusion) to
seen this and there is no logical reason to expect it,
prevent catastrophic visual loss.
since GCA is not a disease of the veins. I think that it
is purely a chance occurrence.
Cotton-Wool Spots
One third of the eyes with visual loss in our series [68] Choroidal Ischemic Lesions
had retinal cotton-wool spots at the posterior pole dur-
ing early stages of the disease (Figs. 12.19a, 12.23, In GCA, occlusion of the PCAs may, in addition to
12.24). They represent focal inner retinal ischemic A-AION, produce patches of choroidal infarcts which
lesions. In GCA, most probably these are due to plate- 23 weeks later appear as peripheral chorioretinal
let microembolization from the partially thrombosed degenerative lesions. They are usually located in the
regional arteries, i.e., ophthalmic artery, central retinal mid-peripheral region of the fundus and frequently
artery or the common trunk of central retinal artery and are triangular in shape with their base towards the
PCA (Figs.12.21 and 12.22). Since GCA is a disease equator and apex towards the posterior pole
of medium-sized and large arteries, it cannot involve (Fig. 12.25) [59, 61, 122]. In our series [68], these
terminal retinal arterioles to produce cotton-wool were seen in only 10 of the 85 eyes with visual loss
spots. Isolated cotton-wool spots can be an early oph- due to GCA.
thalmoscopic finding in GCA and precede severe irre-
versible visual loss [120].
Fig.12.23 Fundus photograph of left eye of a patient with GCA Fig.12.24 Fundus photograph of right eye with A-AION and
with two cotton-wool spots (arrows) a cotton-wool spot (arrow)
Miscellaneous Ocular Ischemic Lesions 219
a Hwang et al. [125] showed evidence of ophthalmic

artery occlusion in their case. This shows an extremely
rare occurrence of this condition in GCA. It is possible
that in the literature patients with A-AION, arteritic
PION, CRAO and/or choroidal ischemia have been
categorized as having ocular ischemia, rather than
the classical ocular ischemic syndrome. The usual
explanation given has been that it was caused by
thrombosis and occlusion of the ophthalmic artery by
GCA; however, that has mostly been a presumptive
b explanation in these reports.
Anterior Segment Ischemic Lesions
There are several anecdotal case reports of this in GCA

in the literature [61, 67, 68, 124138] Ocular hypotony
as a manifestation of anterior segment ischemia in
GCA has been reported [61, 128133, 136]. Other
anterior segment manifestations of GCA include
uveitis [61, 139143] (this may actually represent
protein and cells leaking from ischemic iris/ciliary
body vessels rather than genuine uveitis) or associated
with corneal changes [61, 133, 144], marginal corneal
ulceration [134], reduced corneal sensitivity [145], fall
Fig.12.25 Fundus photograph of left eye of a patient with GCA of corneal temperature [146], necrotizing anterior
and A-AION several months after visual loss. (a) A composite
fundus photograph showing peripheral chorioretinal lesions
scleritis [147] scleritis and episcleritis [148], or neovas-
with optic atrophy. (b) A magnified view of the area indicated by cular glaucoma [44]. In our recent series of 170 GCA
arrow in (a) patients [68] there was only one patient with anterior
segment ischemia and associated neovascular
glaucoma.
Miscellaneous Ocular Ischemic Lesions
Ocular Ischemic Syndrome

Pupillary Abnormalities
There are many vague mentions of the presence of
ocular ischemia in patients with GCA, without any In GCA, the commonest abnormality is a relative affer-
specific information about its clinical features, fre- ent pupillary defect because of visual loss in one eye
quency or cause. In our study [123] of 32 consecutive due usually to A-AION and in some cases to PION or
patients with ocular ischemic syndrome, GCA was CRAO. Anterior segment ischemia, by causing iris
responsible in none of them for its development. ischemia, can also cause pupillary abnormality.
Similarly in our study [68] of ocular manifestations of Development of Horners syndrome associated with
GCA in 170 patients, none had what is classically GCA has been reported [149151]. Some have reported
described as ocular ischemic syndrome. I have been development of tonic pupils from GCA [152154] and
able to find only 4 case reports of development of it has been postulated that may be due to ischemia of
ocular ischemic syndrome with GCA [124127]. the ciliary body [153].
Extra-Ocular Motility Disorders lesions which are not seen in these cases. The myo-
genic view seems more plausible; according to it,
There are many anecdotal case reports of development diplopia is due to ischemic myopathy of one or more
of diplopia in GCA due to involvement of the extraocular of the extraocular muscles, as a result of arteritic occlu-
muscle, with or without visual disturbance. The sion of one or more of the arteries supplying the
reported incidence in larger series has been 1015%, extraocular muscles [24, 68]. Several authors have
e.g., Whitfield etal. [82] reported that in 12 out of 72 suggested that ophthalmoplegia in GCA is due to isch-
GCA patients, Caselli etal. [155] in only 2%, Jonasson emia of the extraocular muscles [162, 165, 166].
etal. [156] in 5%, and Glutz von Blotzheim and Borruat Sibony and Lessell [167] found transient oculomotor
[157] in 4 of 47 patients. In our series [68] of 170 GCA synkinesis or aberrant regeneration in a patient with
patients, the incidence was 6% and all of them gave a GCA and ophthalmoplegia, and they argued that this
history of only transient diplopia and had no diplopia indicates that ophthalmoplegia was neurogenic rather
on examination in our clinic; however, it is possible than myogenic in origin.
that this may be an under estimate because if one eye is In an elderly patient with diplopia and an ill-defined
blind or has severe visual loss, no diplopia may be pattern of ophthalmoplegia, one should always rule out
experienced in spite of extraocular muscle palsy. In one GCA [168]. Complete or partial recovery with steroid
woman in our series, however, a history of transient therapy within 24 h [161], up to some months, is gen-
diplopia was the only presenting symptom of GCA. erally the rule [10, 82, 154, 161163].
Meadows [10] stated that diplopia usually appears I have heard from physicians who missed GCA,
during the active phase of arteritis, some weeks after arguing that the patient did not have diplopia, which
the headache, although occasionally headache and seems to have been stressed in some textbooks as an
diplopia may appear together as the first signs, or important sign of GCA. This is unfortunate and
extraocular muscle palsy may precede visual loss. It inaccurate.
has been suggested that unexplained diplopia in elderly
persons should raise a suspicion of GCA and lead to
a search for other manifestations of this disease [158].
According to some authors, the vertical eye muscles Orbital Complications
are involved first and most often [92], while according
to others, there is usually sixth nerve palsy [159], or These have been described in patients with GCA. They
third nerve [160], and still others have reported involve- mostly represent what has been called orbital pseudo-
ment of any one or all of the oculomotor nerves, with tumor (or orbital inflammatory syndrome) [67, 135,
sparing of the pupil [161] or, most frequently, limita- 169176], associated with orbital pain, conjunctival
tion of up-gaze [162]. Cranial nerve palsy is often injection, chemosis, proptosis, orbital muscle enlarge-
reported as incomplete and may be unilateral or bilat- ment, diplopia, and in some cases optic nerve sheath
eral. Crompton et al. [163] reported a patient with involvement. Presence of proptosis with pseudotumor
bilateral internuclear ophthalmoplegia (with horizontal or alone has been reported by several authors [159,
diplopia, weakness of adduction and abducting nystag- 175, 177179]. In addition to these, orbital infarction
mus) as the initial presenting sign of GCA. Rarely with ischemia of all orbital structures [157, 180],
there may be ptosis [162, 164]. orbital apex syndrome [181, 182] or periorbital ecchy-
There are two theories on the cause of diplopia in mosis [183] have all been reported. In our recent series
GCA: neurogenic and myogenic. According to the of 170 GCA patients [68] we did not see any patient
neurogenic theory, diplopia is generally thought to be with orbital complications.
due to ischemia of one or more of the three oculomotor
nerves, or possible to brain-stem ischemia [10]. That
seems unlikely because the various oculomotor nerves
are supplied by fine nutrient vessels which cannot be Miscellaneous Ophthalmic Manifestations
involved by GCA, which is basically a disease of the
medium-sized and large arteries; if there was involve- There are rare reports of formed visual hallucinations
ment of big arteries in those regions, then one would in GCA patients [184]. Nesher et al. [100] studied
see far more extensive brain stem damage or other 31 consecutive patients with GCA, and reported that 4
References 221
of the 5 patients who suffered permanent visual loss Conclusions

experienced intermittent visual hallucinations over
a period of 110 days before the visual loss the
Since GCA is a potentially blinding disease, its early
patients were aware of the unreal nature of the visions.
diagnosis and treatment is the key to preventing
Hallucinations disappeared within 2weeks, but in one
blindness - visual loss due to GCA is preventable. If
patient, recurred 6months later in association with fur-
a patient is older than 50years, presents with a his-
ther visual deterioration. Hallucinations are thought to
tory of amaurosis fugax, diplopia or sudden visual
be due to temporal lobe involvement [185].
loss in one or both eyes, and has AION, PION, CRAO,
cilioretinal artery occlusion or other acute ocular
ischemic lesions, then the physician must first rule
Cerebral Ischemic Lesions Producing out GCA by doing ESR and CRP evaluation on an
Visual Loss emergency basis THIS IS ESSENTIAL. If there is
a high index of suspicion for GCA from these and
systemic findings, the patient should be started imme-
Involvement of intracranial arteries is well known in
diately on high doses of systemic corticosteroid
GCA, since this disease has often been called cranial
therapy. Temporal artery biopsy should be done as
arteritis. Therefore, cerebral ischemic lesions can
soon as is convenient to confirm the diagnosis, but the
produce visual loss; however, such an occurrence is
start of the treatment must not wait for result of the
extremely rare. I have had only one such case among
temporal artery biopsy, because by the time that result
about 300 GCA patients seen in my clinic. Graham
is available, the patient may suffer further irreversible
[161] reported cortical blindness in only one of 90
visual loss.
patients with GCA. Wiszniewska etal. [186] in a series
of 4,086 patients found stroke in 6 patients (0.15%);
the CT and MRI scans showed lacunar infarction in 3
patients, territorial infarction in 2 and both were nor-
References
mal in 1; and stroke latency ranged from 0 to 2months.
Gonzlez-Gay etal. [187] in a series of 287 consecu-
1. Meadows SP. Temporal arteritis and the loss of vision. Trans
tive biopsy proven GCA patients found vertebrobasilar Ophthalmol Soc UK. 1954;74:1324.
stroke in 7 (2.4%). Stroke in GCA may be due to 2. Hayreh SS. Blood supply of the optic nerve head and its role
involvement of the posterior cerebral or calcarine or in optic atrophy, glaucoma, and oedema of the optic disc. Br
carotid artery by arteritis [10]. This may result in hom- J Ophthalmol. 1969;53(11):72148.
3. Loewenstein A. Cavernous degeneration, necrosis and other
onymous hemianopia. Bilateral cortical blindness regressive processes in the optic nerve with vascular disease
rarely may develop due to infarction of both occipital of the eye. Arch Ophthalmol. 1945;34:2206.
lobes [39, 188190]; in this case there is a sudden 4. Smith C, Greene PB. Ocular complications of temporal
onset of bilateral blindness, with normal pupils and arteritis. Am J Ophthalmol. 1949;32(5):68590.
5. Igersheimer J. Serious visual troubles caused by unusual cir-
fundi. In these cases, because of sparing of the area of culatory disturbances in the region of the optic nerve. Am
fixation, the central vision is sometimes preserved, J Ophthalmol. 1959;47(6):76772.
with tubular fields of a few degrees only. There may be 6. Rdenhauser JH. ber pathologisch-anatomische augenver-
associated cerebral symptoms to point towards the aenderungen bei generalisierter riesenzellarteriitis. Klin
Monatsbl Augenheilkd. 1964;145:41429.
cerebral etiology of the visual loss. Liozon etal. [191] 7. Crompton MR. The visual changes in temporal (giant-cell)
have discussed the risk factors for development of irre- arteritis. Report of a case with autopsy findings. Brain.
versible cerebral ischemia due to GCA. Cid etal. [99] 1959;82:37790.
reported that the presence of a strong acute-phase 8. Foulds WS. Ischaemic optic neuropathy. In: Cant JS, editor.
Proceedings, William Mackenzie Centenary Symposium on
response defines a subgroup of patients at very low the Ocular Circulation in Health and Disease. London: Royal
risk of developing cranial ischemic complications, and College of Physicians and Surgeons; 1968. p. 13641.
conversely, a low inflammatory response and the pres- 9. Sanders MD. Ischaemic papillopathy. Trans Ophthalmol Soc
ence of transient cranial ischemic events provide UK. 1971;91:36986.
10. Meadows SP. Temporal or giant cell arteritis: ophthalmic
a high risk of developing irreversible ischemic compli- aspects. In: Smith JL, editor. Neuro-ophthalmology:
cations and require a prompt therapeutic intervention. Symposium of the University of Miami and the Bascom
There is, however, controversy on this topic. Palmer Eye Institute. St. Louis: Mosby; 1968, v. 4. p. 14857.
11. Franois J. Pseudo-papillites vasculaires [vascular pseudo- 33. Birkhead NC, Wagener HP, Shick RM. Treatment of tempo-
papillitis]. Bull Soc Ophtalmol Fr. 1956;69:3657. ral arteritis with adrenal corticosteroids; results in fifty-five
12. Francois J, Verriest G, De Rouck A. Lelectro-oculographie cases in which lesion was proved at biopsy. J Am Med
en tant quexamen fonctionnel de la retine. [electrooculogra- Assoc. 1957;163(10):8217.
phy in the functional examination of the retina]. Bibl 34. Uhthoff W. Zu den entzndlichen Sehnerven-affektionen bei
Ophthalmol. 1957;13(49):167. Arteriosklerose (Atherosklerose). Ber Zusammenkunft
13. Franois J, Verriest G, Baron A. Pseudopapillites vasculai- Dtsch Ophthalmol Ges. 1924;44:196208.
res. Acta Ophthalmol (Copenh). 1957;35:3252. 35. Cooke WT, Cloake PCP, Govan ADT, Colbeck JC. Temporal
14. Franois J, Verriest G, Neetens A, De Rouck A. arteritis. A generalized vascular disease. Q J Med.
Pseudopapillites vasculaires [vascular pseudo-papillitis]. 1946;15:4775.
Bull Soc Belge Ophtalmol. 1961;129:41327. 36. Cardell BS, Hanley T. A fatal case of giant-cell or temporal
15. Franois J, Verriest G, Neetens A, de Rouck A, Hanssens M. arteritis. J Pathol Bacteriol. 1951;63(4):58797.
Pseudonvrites optiques oedmateuses dorigine vasculaire 37. Greenfield JG. A case of giant-cell arteritis. Proc R Soc
[vascular pseudopapillitis]. Ann Ocul (Paris). 1962;195: Med. 1951;44:8557.
83085. 38. Kreibig W. Optikomalazie, die Folge eines Gefassverschlusses
16. Franois J, Neetens A. Physio-anatomy of the axial vascu- im retrobulbren Abschnitt des Sehnerven. [Opticomalacia
larization of the optic nerve. Doc Ophthalmol. 1969;26: caused by vascular occlusion in the retrobulbar portion of
3849. optic nerves]. Klin Monatsbl Augenheilkd Augenrztl
17. Piper HF, Unger L. Hemianopsia horizontalis inferior bei Fortbild. 1953;122(6):71973.
akuten Durchblutungsstorungen des Sehnerven. Ophthalmo 39. Heptinstall RH, Porter KA, Barkley H. Giant-cell (temporal)
logica. 1957;134(3):16980. arteritis. J Pathol Bacteriol. 1954;67(2):50719.
18. Ponte F. La pseudopapillite vascolare degli arteriosclerot- 40. Spencer WH, Hoyt WF. A fatal case of giant-cell arteritis
ici. Aspetti clinici e consideraziioni patogenetiche. G Ital (temporal or cranial arteritis) with ocular involvement. Arch
Oftalmol. 1964;17(3):17019. Ophthalmol. 1960;64(6):8627.
19. Saraux H, Murat JP. Les pseudo-papillites dorigine vascu- 41. Seitz R. ber die Pathogenese des Verschlusses der
laire. [pseudopapillitis of vascular origin]. Ann Ocul (Paris). Netzhautarterien. Ber Zusammenkunft Dtsch Ophthalmol
1967;200(1):119. Ges. 1961;64:3217.
20. Franois J, Neetens A. Vascularization of the optic pathway. 42. Cullen JF. Occult temporal arteritis. Trans Ophthalmol Soc
I. Lamina cribrosa and optic nerve. Br J Ophthalmol. UK. 1963;83:72536.
1954;38(8):47288. 43. Manschot WA. A fatal case of temporal arteritis with ocular
21. Hayreh SS. A study of the central artery of the retina in symptoms. Ophthalmologica. 1965;149:12130.
human beings. Panjab University, India: Thesis for Master of 44. Wolter JR, Phillips RL. Secondary glaucoma in cranial
Surgery; 1958. arteritis. Am J Ophthalmol. 1965;59:62534.
22. Singh (Hayreh) S, Dass R. The central artery of the retina. I. 45. Cogan DG. Neurology of the visual system. Springfield, Ill:
Origin and course. Br J Ophthalmol. 1960;44:193212. C.C. Thomas; 1966. p. 137, 173, 1858.
23. Singh (Hayreh) S, Dass R. The central artery of the retina. II. 46. MacFaul PA. Ciliary artery involvement in giant cell arteritis.
A study of its distribution and anastomoses. Br J Ophthalmol. Br J Ophthalmol. 1967;51(8):50512.
1960;44:28099. 47. Henkind P, Charles NC, Pearson J. Histopathology of isch-
24. Hayreh SS. The Ophthalmic Artery. III. Branches. Br J emic optic neuropathy. Am J Ophthalmol. 1970;69(1):
Ophthalmol. 1962;46(4):21247. 7890.
25. Hayreh SS. The central artery of the retina. Its role in the 48. Knox DL, Duke JR. Slowly progressive ischemic optic neu-
blood supply of the optic nerve. Br J Ophthalmol. ropathy. A clinicopathologic case report. Trans Am Acad
1963;47:65163. Ophthalmol Otolaryngol. 1971;75(5):10658.
26. Hayreh SS. Blood supply and vascular disorders of the optic 49. MacMichael IM, Cullen JF. Pathology of ischaemic optic
nerve. An Inst Barraquer. 1963;4:7109. neuropathy. In: Cant JS, editor. Proceedings, 2nd William
27. Beauvieux J, Ristitch K. Les vaisseaux centraux du nerf Mackenzie Symposium on the optic nerve. London:
optique. Etude anatomique. Arch Ophtalmol (Paris). Kimpton; 1972. p. 10816.
1924;41:35269. 50. Ellenberger Jr C, Netsky MG. Infarction in the optic nerve.
28. Steele EJ, Blunt MJ. The blood supply of the optic nerve and J Neurol Neurosurg Psychiatry. 1968;31(6):60661.
chiasma in man. J Anat (Lond). 1956;90:48693. 51. Igersheimer J. Sehnervenerkrankung bei maligner Sklerose.
29. Franois J, Neetens A, Collette JM. Vascular supply of the Zeitschrift fr Augenheilkunde. 1929;69:4759.
optic pathway. II. Further studies by micro-arteriography of 52. Peters W. ber die Neuritis optici arteriosklerotischer
the optic nerve. Br J Ophthalmol. 1955;39(4):22032. Genese. Klin Monatsbl Augenheilkd Augenrztl Fortbild.
30. Franois J, Neetens A. Vascularization of the optic pathway. 1958;132(3):36377.
III. Study of intra-orbital and intracranial optic nerve by 53. Wagener HP, Hollenhorst RW. The ocular lesions of tempo-
serial sections. Br J Ophthalmol. 1956;40(1):4552. ral arteritis. Am J Ophthalmol. 1958;45(5):61730.
31. Franois J, Neetens A. Central retinal artery and central optic 54. Hayreh SS, Baines JAB. Occlusion of the posterior ciliary
nerve artery. Br J Ophthalmol. 1963;47(1):2130. artery I. Effects on choroidal circulation. Br J Ophthalmol.
32. Paillas J-E, Guillot P, Duplay J. Six observations de neuro- 1972;56(10):71935.
papillites aiges. Exploration cu chiasma optique; constata- 55. Hayreh SS, Baines JA. Occlusion of the posterior ciliary
tions opratoires; intrt du traitement neuro-chirurgical. artery. III. Effects on the optic nerve head. Br J Ophthalmol.
Bull Soc Ophtalmol Fr. 1949;7:20210. 1972;56(10):75464.
References 223
56. Hayreh SS, Chopdar A. Occlusion of the posterior ciliary 79. Bonnet P. Larterite temporale et ses manifestations oculaires;
artery V. Protective influence of simultaneous vortex vein revue critique. [temporal arteritis and its ocular manifesta-
occlusion. Arch Ophthalmol. 1982;100(9):148191. tions]. Arch Ophtalmol Rev Gn Ophtalmol. 1954;14(1):
57. Loeffler KU, Hayreh SS, Tso MOM. The effects of simulta- 2447.
neous occlusion of the posterior ciliary artery and vortex 80. Bruce GM. Temporal arteritis as a cause of blindness; review
veins. Arch Ophthalmol. 1994;112(5):67482. of literature and report of a case. Trans Am Ophthalmol Soc.
58. Hayreh SS. Anterior ischaemic optic neuropathy. I. 1949;47:30016.
Terminology and pathogenesis. Br J Ophthalmol. 81. Bessiere E, Julien RG. Les manifestations papillaires dans le
1974;58(12):95563. syndrome dit arterite temporale, frequence de loedeme
59. Hayreh SS. Anterior ischaemic optic neuropathy. II. Fundus papillaire ischemique. [papillary manifestations in the syn-
on ophthalmoscopy and fluorescein angiography. Br J drome called temporal arteritis; frequency of ischemic papil-
Ophthalmol. 1974;58(12):96480. lary edema]. Arch Ophtalmol Rev Gn Ophtalmol. 1950;
60. Hayreh SS. Pathogenesis of cupping of the optic disc. Br J 10(6):70114.
Ophthalmol. 1974;58(10):86376. 82. Whitfield AG, Bateman M, Cooke WT. Temporal arteritis.
61. Hayreh SS. Anterior ischemic optic neuropathy. New York: Br J Ophthalmol. 1963;47:55566.
Springer; 1975. 83. Benedict WL, Wagener HP, Horton BT. Temporal arteritis.
62. Hayreh SS. Ischemic optic neuropathy. Int Ophthalmol. Ophthalmol Ibero Am. 1957;19:1604.
1978;1:9l8. 84. Parsons-Smith G. Sudden blindness in cranial arteritis. Br
63. Hayreh SS, Podhajsky P. Visual field defects in anterior isch- J Ophthalmol. 1959;43(4):20416.
emic optic neuropathy. Doc Ophthalmol Proc Ser. 1979; 85. Crosby RC, Wadsworth RC. Temporal arteritis; review of
19:5371. the literature and report of five additional cases. Arch Med
64. Hayreh SS. Anterior ischemic optic neuropathy. Arch Interna. 1948;81(4):43164.
Neurol. 1981;38(11):6758. 86. Liu GT, Glaser JS, Schatz NJ, Smith JL. Visual morbidity in
65. Hayreh SS. Inter-individual variation in blood supply of the giant cell arteritis. Clinical characteristics and prognosis for
optic nerve head. Its importance in various ischemic disorders vision. Ophthalmology. 1994;101(11):177985.
of the optic nerve head, and glaucoma, low-tension glaucoma 87. Gonzlez-Gay MA, Garca-Porra C, Llorca J, Hajeer AH,
and allied disorders. Doc Ophthalmol. 1985;59(3):21746. Braas F, Dababneh A, etal. Visual manifestations of giant
66. Hayreh SS. Anterior ischaemic optic neuropathy. Differen cell arteritis. Trends and clinical spectrum in 161 patients.
tiation of arteritic from non-arteritic type and its manage- Medicine (Baltimore). 2000;79(5):28392.
ment. Eye. 1990;4(1):2541. 88. Gonzlez-Gay MA, Miranda-Filloy JA, Lopez-Diaz MJ,
67. Hayreh SS. Ophthalmic features of giant cell arteritis. Perez-Alvarez R, Gonzlez-Juanatey C, Sanchez-Andrade A,
Baillires Clin Rheumatol. 1991;5(3):43159. etal. Giant cell arteritis in northwestern Spain. A 25-year epi-
68. Hayreh SS, Podhajsky PA, Zimmerman B. Ocular manifes- demiologic study. Medicine (Baltimore). 2007;86(2):
tations of giant cell arteritis. Am J Ophthalmol. 1998; 618.
125(4):50920. 89. Salvarani C, Cimino L, Macchioni P, Consonni D, Cantini F,
69. Hayreh SS, Podhajsky PA, Zimmerman B. Occult giant cell Bajocchi G, etal. Risk factors for visual loss in an Italian
arteritis. ocular manifestations. Am J Ophthalmol. 1998; population-based cohort of patients with giant cell arteritis.
125(4):5216. 893. Arthritis Rheum. 2005;53(2):2937.
70. Costello F, Zimmerman MB, Podhajsky PA, Hayreh SS. 90. Rucker CW. Circulatory disorders. Trans Am Acad
Role of thrombocytosis in diagnosis of giant cell arteritis Ophthalmol Otolaryngol. 1956;60(1):602.
and differentiation of arteritic from non-arteritic anterior 91. Mosher HA. The prognosis in temporal arteritis. Arch
ischemic optic neuropathy. Eur J Ophthalmol. 2004; Ophthalmol. 1959;62(4):6414.
14(3):24557. 92. Hollenhorst RW, Brown JR, Wagener HP, Shick RM.
71. Hayreh SS. Ischemic optic neuropathy. Prog Retin Eye Res. Neurologic aspects of temporal arteritis. Neurology. 1960;
2009;28(1):3462. 10:4908.
72. Hayreh SS. Colour and fluorescence of the optic disc. 93. Salvarani C, Cantini F, Boiardi L, Hunder GG. Polymyalgia
Ophthalmologica. 1972;165:1008. rheumatica and giant-cell arteritis. N Eng J Med.
73. Foulds WS. Visual disturbances in systemic disorders. Optic 2002;347:26171.
neuropathy and systemic disease. Trans Ophthalmol Soc 94. Salvarani C, Cantini F, Hunder GG. Polymyalgia rheumatica
UK. 1969;89:12546. and giant-cell arteritis. Lancet. 2008;372(9634):23445.
74. Hayreh SS. In vivo choroidal circulation and its watershed 95. Hayreh SS. Steroid therapy for visual loss in patients with
zones. Eye. 1990;4(2):27389. giant-cell arteritis. Lancet. 2000;355(9215):15723.
75. Hayreh SS, Revie IH, Edwards J. Vasogenic origin of visual 96. Hayreh SS, Zimmerman B. Management of giant cell arteri-
field defects and optic nerve changes in glaucoma. Br J tis. Our 27-year clinical study. new light on old controver-
Ophthalmol. 1970;54(7):46172. sies. Ophthalmologica. 2003;217(4):23959.
76. Hayreh SS. Pathogenesis of occlusion of the central retinal 97. Hayreh SS, Zimmerman B. Visual deterioration in giant cell
vessels. Am J Ophthalmol. 1971;72(5):998101. arteritis patients while on high doses of corticosteroid ther-
77. Horton BT, Magath TB. Arteritis of the temporal vessels: apy. Ophthalmology. 2003;110(6):120415.
report of seven cases. Proc Staff Meet Mayo Clin. 98. Gonzlez-Gay MA, Blanco R, Snchez-Andrade A,
1937;12:54853. Vzquez-Caruncho M. Giant cell arteritis in Lugo, Spain: a
78. Jennings GH. Arteritis of the temporal vessels. Lancet. more frequent disease with fewer classic features.
1938;1:4248. J Rheumatol. 1997;24(11):216670.
99. Cid MC, Font C, Oristrell J, de la Sierra A, Coll-Vinent B, 118. McLeod D, Oji EO, Kohner EM, Marshall J. Fundus signs
Lpez-Soto A, etal. Association between strong inflamma- in temporal arteritis. Br J Ophthalmol. 1978;62(9):5914.
tory response and low risk of developing visual loss and 119. Fineman MS, Savino PJ, Federman JL, Eagle Jr RC. Branch
other cranial ischemic complications in giant cell (tempo- retinal artery occlusion as the initial sign of giant cell arteri-
ral) arteritis. Arthritis Rheum. 1998;41(1):2632. tis. Am J Ophthalmol. 1996;122(3):42830.
100. Nesher G, Nesher R, Rozenman Y, Sonnenblick M. Visual 120. Daudin JB, Bluwol E, Chaine G, Rohart C. Arterite a cel-
hallucinations in giant cell arteritis: association with visual lules geantes revelee par des nodules cotonneux. [Cotton-
loss. J Rheumatol. 2001;28(9):20468. wool spots as first ocular manifestation of giant cell
101. Schmidt WA, Moll A, Seifert A, Schicke B, Gromnica-Ihle arteritis]. J Fr Ophtalmol. 2006;29(10):e28.
E, Krause A. Prognosis of large-vessel giant cell arteritis. 121. Shannon EW, Solomon J. Bilateral temporal arteritis with
Rheumatology (Oxford). 2008;47(9):14068. complete loss of vision. J Am Med Assoc. 1945;
102. Hayreh SS. Acute ischemic disorders of the optic nerve - 127(11):6479.
Pathogenesis, clinical manifestations, and management. 122. Amalric P. Acute choroidal ischaemia. Trans Ophthalmol
Ophthalmol Clin North Am. 1996;9(4):40742. Soc UK. 1971;91:30522.
103. Raymond LA, Sacks JG, Choromokos E, Khodadad G. 123. Mizener JB, Podhajsky P, Hayreh SS. Ocular ischemic
Short posterior ciliary artery insufficiency with hyper- syndrome. Ophthalmology. 1997;104(5):85964.
thermia (Uhthoffs symptom). Am J Ophthalmol. 124. Hamed LM, Guy JR, Moster ML, Bosley T. Giant cell
1980;90(5):61923. arteritis in the ocular ischemic syndrome. Am J Ophthalmol.
104. Finelli PF. Alternating amaurosis fugax and temporal 1992;113(6):7025.
arteritis. Am J Ophthalmol. 1997;123(6):8501. 125. Hwang JM, Girkin CA, Perry JD, Lai JC, Miller NR,
105. Hayreh SS. The 1994 Von Sallman Lecture. The optic Hellmann DB. Bilateral ocular ischemic syndrome second-
nerve head circulation in health and disease. Exp Eye Res. ary to giant cell arteritis progressing despite corticosteroid
1995;61(3):25972. treatment. Am J Ophthalmol. 1999;127(1):1024.
106. Hayreh SS. The blood supply of the optic nerve head and 126. Casson RJ, Fleming FK, Shaikh A, James B. Bilateral
the evaluation of it - myth and reality. Prog Retin Eye Res. ocular ischemic syndrome secondary to giant cell arteritis.
2001;20(5):56393. Arch Ophthalmol. 2001;119(2):3067.
107. Hinzpeter EN, Naumann G. Ischemic papilledema in giant- 127. Fathilah J, Jamaliah R. Giant cell arteritis with panocular
cell arteritis. Mucopolysaccharide deposition with normal involvement in an Indian male. Med J Malaysia.
intraocular pressure. Arch Ophthalmol. 1976;94(4):6248. 2003;58(1):1114.
108. Hayreh SS, Zimmerman MB. Nonarteritic anterior isch- 128. Haimbock K. The effect of smoking on the retinal arterial
emic optic neuropathy: natural history of visual outcome. pressure of normal persons. Wien Klin Wochenschr.
Ophthalmology. 2008;115(2):298305. 1961;73:52931.
109. Hayreh SS, Zimmerman MB, Podhajsky P, Alward WL. 129. Bettelheim H. Die akute Ophthalmomalacie-eine
Nocturnal arterial hypotension and its role in optic nerve Folgeerscheinung der Arteriitis cranialis. [Acute ophthal-
head and ocular ischemic disorders. Am J Ophthalmol. momalacia-as caused by cranial arteritis]. Albrecht Von
1994;117(5):60324. Graefes Arch Klin Exp Ophthalmol. 1968;174(4):35966.
110. Hayreh SS, Podhajsky PA, Zimmerman B. Nonarteritic 130. Goder G. Durchhbluungsstrungen des auges und biopsy
anterior ischemic optic neuropathy. Time of onset of visual der arteria temporalis. Leipzig: G. Thieme; 1968.
loss. Am J Ophthalmol. 1997;124(5):6417. 131. Nagy F, Juhasc J. Ophthalmologic alterations of giant cll
111. Danesh-Meyer HV, Savino PJ, Sergott RC. The prevalence of arteritis. Szemescr. 1970;107:1629.
cupping in end-stage arteritic and nonarteritic anterior isch- 132. Daicker B, Keller HH. Giant cell arteritis with endocular
emic optic neuropathy. Ophthalmology. 2001;108(3):5938. spreading and hypotonia bulbi dolorosa. A clinico-patholog-
112. Danesh-Meyer H, Savino PJ, Spaeth GL, Gamble GD. ical report. Klin Monbl Augenheilkd. 1971;158(3):35872.
Comparison of arteritis and nonarteritic anterior ischemic 133. Verdich M, Nielsen NV. Acute transient ophthalmomalacia
optic neuropathies with the Heidelberg Retina Tomograph. in giant-cell arteritis. Report of a case. Acta Ophthalmol
Ophthalmology. 2005;112(6):110412. (Copenh). 1975;53(6):8758.
113. Hayreh SS. Posterior ischemic optic neuropathy. 134. Gerstle CC, Friedman AH. Marginal corneal ulceration
Ophthalmologica. 1981;182(1):294. (limbal guttering) as a presenting sign of temporal arteritis.
114. Hayreh SS. Posterior ischaemic optic neuropathy. Clinical Ophthalmology. 1980;87(11):11736.
features, pathogenesis, and management. Eye. 2004;18(11): 135. Ghanchi FD, Dutton GN. Current concepts in giant cell
1188206. (temporal) arteritis. Surv Ophthalmol. 1997;42(2):
115. Sadda SR, Nee M, Miller NR, Biousse V, Newman NJ, 99123.
Kouzis A. Clinical spectrum of posterior ischemic optic 136. Huna-Baron R, Mizrachi IB, Glovinsky Y. Intraocular pres-
neuropathy. Am J Ophthalmol. 2001;132(5):74350. sure is low in eyes with giant cell arteritis. J Neuroophthalmol.
116. Saraux H, Laroche L, Foels A, Martin-Gousset D. Ischemie 2006;26(4):2735.
aigue posterieure du nerf optique. [Posterior acute ischemic 137. McKillop E, Tejwani D, Weir C, Jay J. Anterior segment
optic neuropathy]. J Fr Ophtalmol. 1982;5(3):16717. ischaemia with giant cell arteritis. Can J Ophthalmol.
117. Koenig F, Antoine C, Michel D, Rousset H, Maugery J. 2006;41(2):2013.
Manifestations ophtalmologiques de la maladie de Horton. 138. Zion VM, Goodside V. Anterior segment ischemia with
[Eye manifestations of Hortons disease]. Bull Soc ischemic optic neuropathy. Surv Ophthalmol. 1974;19:
Ophtalmol Fr. 1989;89(5):6937. 1930.
References 225
139. Landblad Y. Temporal arteritis complicated by iritis. Nord 163. Crompton JL, Burrow DJ, Iyer PV. Bilateral internuclear
Med. 1956;56:1794. ophthalmoplegiaan unusual initial presenting sign of giant
140. Dasgupta B, Pitzalis C, Panayi GS. Inflammation of the cell arteritis. Aust N Z J Ophthalmol. 1989;17(1):
uveral tract as a presenting feature of temporal arteritis. 714.
Ann Rheum Dis. 1989;48(11):9645. 164. Duverne J, Mounier R, Brochier C, Margolliet MF. Arterite
141. Rajesh CV, Cole M. Panuveitis as a presenting feature of temporale bilaterale avec manifestations oculaires.
giant cell arteritis. Br J Ophthalmol. 2000;84(3):340. [Bilateral temporal arteritis with ocular manifestations].
142. Vdrine L, Algayres JP, Coutant G. Giant-cell arteritis. N Bull Soc Fr Dermatol Syphiligr. 1969;76(5):6902.
Eng J Med. 2003;348:14978. 165. Martin EA. Double vision and temporal arteritis. J Ir Med
143. Bandini F, Benedetti L, Ceppa P, Corallo G. Uveitis as a Assoc. 1970;63(395):1913.
presenting sign of giant cell arteritis. J Neuroophthalmol. 166. Barricks ME, Traviesa DB, Glaser JS, Levy IS.
2005;25(3):2478. Ophthalmoplegia in cranial arteritis. Brain. 1977;100(2):
144. Radda TM, Bardach H, Riss B. Acute ocular hypotony. A 20922.
rare complication of temporal arteritis. Ophthalmologica. 167. Sibony PA, Lessell S. Transient oculomotor synkinesis in
1981;182(3):14852. temporal arteritis. Arch Neurol. 1984;41(1):878.
145. Galves MJ, Molina NP. Ocular manifestations in arteritis. 168. Hayreh SS. Ocular manifestations of giant cell arteritis.
Archivos Soc Oftalmol Hisp Am. 1966;26:1915. Am J Ophthalmol. 1998;126:7434.
146. Horven I. Corneal temperature in normal subjects and arte- 169. Nassani S, Cocito L, Arcuri T, Favale E. Orbital pseudotu-
rial occlusive disease. Acta Ophthalmol (Copenh). mor as a presenting sign of temporal arteritis. Clin Exp
1975;53(6):86374. Rheumatol. 1995;13(3):3679.
147. Watson PG, Hazleman BL. The sclera and systemic disor- 170. de Heide LJ, Talsma MA. Giant-cell arteritis presenting
ders. Philadelphia: W.B. Saunders; 1976. p. 274. as an orbital pseudotumor. Neth J Med. 1999;55(4):
148. Long RG, Friedmann AI, James DG. Scleritis and temporal 1968.
arteritis. Postgrad Med J. 1976;52(613):68992. 171. Lee AG, Tang RA, Feldon SE, Pless M, Schiffman JS,
149. Bell TA, Gibson RA, Tullo AB. A case of giant-cell arteri- Rubin RM, etal. Orbital presentations of giant cell arteritis.
tis and Horners syndrome. Scott Med J. 1980;25(4): Graefes Arch Clin Exp Ophthalmol. 2001;239(7):
302. 50913.
150. Arunagiri G, Santhi S, Harrington T. Horner syndrome and 172. Islam N, Asaria R, Plant GT, Hykin PC. Giant cell arteritis
ipsilateral abduction deficit attributed to giant cell arteritis. mimicking idiopathic orbital inflammatory disease. Eur J
J Neuroophthalmol. 2006;26(3):2312. Ophthalmol. 2003;13(4):3924.
151. Shah AV, Paul-Oddoye AB, Madill SA, Jeffrey MN, Tappin 173. Gibbs P. Temporal arteritis. onset with pseudotumour orbit
AD. Horners syndrome associated with giant cell arteritis. or red eye. Practitioner. 1974;213(1274):20521.
Eye (Lond). 2007;21(1):1301. 174. Laidlaw DA, Smith PE, Hudgson P. Orbital pseudotumour
152. Davis RH, Daroff RB, Hoyt WF. Tonic pupil after temporal secondary to giant cell arteritis: an unreported condition.
arteritis. Lancet. 1968;1(7546):822. BMJ. 1990;300(6727):784.
153. Currie J, Lessell S. Tonic pupil with giant cell arteritis. Br J 175. Nielsen NV, Eriksen JS. Unusual ocular manifestation in
Ophthalmol. 1984;68(2):1358. temporal arteritis. A case report. Acta Ophthalmol
154. Foroozan R, Buono LM, Savino PJ, Sergott RC. Tonic (Copenh). 1979;57(3):3628.
pupils from giant cell arteritis. Br J Ophthalmol. 176. Chertok P, Leroux JL, Le Marchand M, Aboukrat P, Alary
2003;87(4):5102. JC, Blotman F. Orbital pseudo-tumor in temporal arteritis
155. Caselli RJ, Hunder GG, Whisnant JP. Neurologic disease in revealed by computerized tomography. Clin Exp Rheumatol.
biopsy-proven giant cell (temporal) arteritis. Neurology. 1990;8(6):5879.
1988;38(3):3529. 177. Liou LM, Khor GT, Lan SH, Lai CL. Giant cell arteritis
156. Jonasson F, Cullen JF, Elton RA. Temporal arteritis. with multiple cranial nerve palsy and reversible proptosis:
A 14-year epidemiological, clinical and prognostic study. a case report. Headache. 2007;47(10):14513.
Scott Med J. 1979;24(2):1117. 178. Cockerham KP, Cockerham GC, Brown HG, Hidayat AA.
157. Glutz von Blotzheim S, Borruat FX. Neuro-ophthalmic Radiosensitive orbital inflammation associated with tem-
complications of biopsy-proven giant cell arteritis. Eur J poral arteritis. J Neuroophthalmol. 2003;23(2):11721.
Ophthalmol. 1997;7(4):37582. 179. Looney BD. Unilateral proptosis resulting from giant-cell
158. Lockshin MD. Diplopia as early sign of temporal arteritis. arteritis. J Am Optom Assoc. 1999;70(7):4439.
Report of two cases. Arthritis Rheum. 1970;13(4):41942. 180. Borruat FX, Bogousslavsky J, Uffer S, Klainguti G, Schatz
159. Simmons RJ, Cogan DG. Occult temporal arteritis. Arch NJ. Orbital infarction syndrome. Ophthalmology.
Ophthalmol. 1962;68:818. 1993;100(4):5628.
160. Asensio-Snchez VM, Morales-Gmez I, Rodrguez-Vaca 181. Stein R, Regenbogen L, Romano A, Kam J, Tadmor R,
I. Third nerve palsy as the only manifestation of occult tem- Godel V. Orbital apex syndrome due to cranial arteritis.
poral arteritis. Arch Soc Esp Oftalmol. 2009;84(8):3958. Ann Ophthalmol. 1980;12(6):70813.
161. Graham E. Survival in temporal arteritis. Trans Ophthalmol 182. Puerto B, Noval S, Veiga C, Contreras I, Mateos E. Apical
Soc UK. 1980;100(Pt 1):10810. orbital inflammation in giant cell arteritis. Arch Soc Esp
162. Dimant J, Grob D, Brunner NG. Ophthalmoplegia, ptosis, Oftalmol. 2007;82(1):479.
and miosis in temporal arteritis. Neurology. 1980;30(10): 183. Fernandez S, Adhiyaman V. Giant cell arteritis presenting
10548. as peri-orbital ecchymosis. Age Ageing. 2005;34(1):856.
184. Hart CT. Formed visual hallucinations. A symptom of cra- 189. Wilkinson IM, Russell RW. Arteries of the head and neck in
nial arteritis. Br Med J. 1967;3(5566):6434. giant cell arteritis. A pathological study to show the pattern
185. Walsh FB. Clinical neuro-ophthalmology. 2nd ed. of arterial involvement. Arch Neurol. 1972;27(5):37891.
Baltimore: Williams & Wilkins; 1957. p. 71. 190. Chisholm IH. Cortical blindness in cranial arteritis. Br J
186. Wiszniewska M, Devuyst G, Bogousslavsky J. Giant cell Ophthalmol. 1975;59(6):3323.
arteritis as a cause of first-ever stroke. Cerebrovasc Dis. 191. Liozon E, Herrmann FR, Ly K, Jauberteau MO, Loustaud
2007;24(23):22630. V, Soria P, etal. Risk factors for irreversible cerebral isch-
187. Gonzlez-Gay MA, Vazquez-Rodriguez TR, Gomez- emia complications from Hortons disease: prospective
Acebo I, Pego-Reigosa R, Lopez-Diaz MJ, Vazquez- study of 178 patients. Rev Md Interne. 2001;22(1):
Trinanes MC, etal. Strokes at time of disease diagnosis in 3041.
a series of 287 patients with biopsy-proven giant cell arteri-
tis. Medicine (Baltimore). 2009;88(4):22735.
For a complete bibliography and detailed review of previously
188. Symonds C, Mackenzie I. Bilateral loss of vision from
published studies in the literature on the subject, please refer to
cerebral infarction. Brain. 1957;80(4):41555.
bibliography in previous publications listed here.
Management of Giant Cell Arteritis
to Prevent Visual Loss 13
As has been emphasized in earlier chapters, the most ophthalmologists (discussed in Chap. 11). This
dreaded complication of giant cell arteritis (GCA) is difference in outlook is evident from most recent
permanent visual loss in one or both eyes and fear of reviews on the subject by rheumatologists [22, 23].
visual loss is exceeded only by fear of death. If GCA is Most GCA patients are initially seen by rheuma-
diagnosed early and treated immediately, aggres- tologists, except those for whom the first manifestation
sively and adequately with systemic corticosteroids, is loss of vision. Most rheumatologists essentially deal
blindness is entirely preventable; that makes GCA the with patients with rheumatological manifestations, while
prime ophthalmic emergency. A study [1] further ophthalmologists see GCA patients with tragic visual
emphasized that fact, when it found that a high propor- loss in one or both eyes or with occult GCA (losing
tion of patients with permanent visual loss had had vision without having any rheumatological or other sys-
a delayed diagnosis and treatment; in that study 35% temic symptoms [14]). So for ophthalmologists GCA is
had experienced systemic symptoms for an average of a blinding disease, whereas for rheumatologists it is a
10.8months before visual loss and 65% had had pre- less serious disease with mainly rheumatological com-
monitory visual symptoms for an average of 8.5days. plaints. Moreover, rheumatologists usually recom-
This obviously raises two critical issues about the mend a management which may be appropriate for
management of GCA: (1) how to establish an early and polymyalgia rheumatica (PMR) (with no risk of blind-
a definite diagnosis of GCA? and (2) what is the proper ness), but falls well short of the kind of therapy required
treatment to prevent blindness? There is a voluminous to prevent blindness in GCA patients.
literature on both these topics; however, there is still
a good deal of controversy on both issues [2]. This has
been partly responsible for missing the correct diagno-
Corticosteroid Therapy in GCA
sis of GCA, inadequate management and unnecessary
blindness. In Chap. 11, I discussed how to establish an to Prevent Visual Loss
early and a definite diagnosis of GCA.
It is well established now that the only proven and effec-
tive treatment for GCA is systemic corticosteroids. This
Reasons for Disagreement is because in GCA an immune insult in the vascular wall
initiates a reaction in the artery that leads to structural
on Management of GCA to Prevent
changes, intimal hyperplasia, and luminal occlusion
Visual Loss [24]. The mechanisms triggering the immune stimula-
tion are unknown; however, the process is strictly
This is an important issue. I have investigated vari- dependent on T cells [24]. It is also universally agreed
ous aspects of GCA, particularly the visual loss due that if there is a reasonable index of suspicion of GCA,
to GCA, in our clinical studies since 1970 [321]. high doses of systemic corticosteroid therapy should
I have found that the most important reason for the be started IMMEDIATELY, as an EMERGENCY
controversy has been the differing perspectives on MEASURE. There is, however, tremendous controversy
GCA as a disease, between rheumatologists and about the correct corticosteroid therapy regimen for

228 13 Management of Giant Cell Arteritis to Prevent Visual Loss
patients with GCA, primarily because, as discussed also on the initial (i) mode of corticosteroid therapy,
above (and discussed at length in Chap. 11), rheumatol- i.e., oral or intravenous, and (ii) the dosage.
ogists and ophthalmologists have very different perspec-
tives on GCA. In the rheumatologic literature dealing
with treatment, it is not uncommon to find patients with
PMR and GCA mixed together. This has resulted in the Intravenous Corticosteroid Therapy
treatment of a collection of heterogeneous cases by
rheumatologists and not purely of GCA. We know that Intravenous methylprednisolone as the initial therapy
the risk of visual loss, of prime importance in GCA, has been investigated by several workers. Cornblath
does not exist in PMR. If a patient diagnosed as having etal. [25] found that results of high-dose intravenous
PMR loses vision, the original diagnosis was not cor- methylprednisolone treatment of patients with visual
rect. It is also well-established that patients with PMR loss from GCA were similar to the results of treatment
usually require much lower doses of corticosteroid ther- with oral corticosteroids. Chan et al. [26] in a retro-
apy to control their disease than do those with GCA. spective study of 73 biopsy positive GCA patients (43
Therefore, the criteria for successful management with with worse initial visual acuity treated with intrave-
corticosteroid therapy among rheumatologists and oph- nous methylprednisolone 100 mg for 25 days; the
thalmologists are entirely different for the two groups of remaining 30 treated with oral prednisolone 50100mg
patients for rheumatologists the aim is essentially a daily) found that the visual acuity improved in 40% of
control of rheumatologic symptoms (with a proportion those treated with intravenous therapy, compared to
of their cases having PMR) but for ophthalmologists it is 13% with oral corticosteroid. They concluded that
primarily the prevention of visual loss due to GCA. intravenous corticosteroid therapy may offer a greater
prospect of improvement than oral corticosteroid ther-
apy. Chevalet et al. [27] evaluated the effect of an
initial intravenous pulse 240 mg methylprednisolone
Controversy on Mode of Administration
followed by oral prednisone (0.7 mg/kg/day) versus
of Corticosteroid Therapy in GCA oral Prednisone therapy (0.7 mg/kg/day); they found
to Prevent Visual Loss that the intravenous therapy had no significant long
term benefit in simple GCA and should be limited to
While corticosteroid therapy is well-established as the complicated forms. Salvarani etal. [22] recommended
treatment of choice in GCA to prevent visual loss, dis- that initial high-dose intravenous methylprednisolone
agreement is still widespread about the initial dosage, (1,000mg every day for 3days) can be tried in patients
mode (oral or intravenous), tapering regimen, duration with recent visual loss, although they found no docu-
of therapy, method of monitoring corticosteroid ther- mentation suggesting that it helps more than oral pred-
apy, relapse on reduction of therapy, and side-effects nisone. Mazlumzadeh et al. [28] conducted a
and their management. In view of these controversies, randomised controlled trial of 27 GCA patients where
we conducted a 27-year planned, longitudinal clinical all patients were on 40mg/day prednisone; in addition,
study of corticosteroid therapy in temporal artery intravenous methylprednisolone (15 mg/kg of body
biopsy confirmed GCA patients to find answers to weight/day) was given for 3days to 14, and not to the
these controversies [21]. In the following discussion, I remaining 13 patients. They concluded that initial
summarize the literature on various aspects of corti- treatment of GCA with intravenous methylpredniso-
costeroid therapy in GCA, as well as the findings of lone pulses, in addition to the regular oral dose of pred-
our study [21] on the subject. nisone, allowed for more rapid tapering of oral
prednisone and had long-term benefits, with a higher
frequency of patients experiencing sustained remission
of their disease after discontinuation of treatment. This
Initial Dosage of Corticosteroid Therapy demonstrates the level of disagreement about the role
of intravenous corticosteroid therapy and the dosage.
A review of the literature shows that the reported ini- In our study [21] of 145 temporal artery biopsy
tial dose of corticosteroid to control GCA has varied confirmed GCA patients, there were 96 with and 49
widely among different studies. There is disagreement without visual loss. In this study, the factors that
Initial Dosage of Corticosteroid Therapy 229
influenced the initial dosage, frequency, duration and that, my current recommendation for intravenous
mode (intravenous or oral) of corticosteroid therapy corticosteroid therapy is to give initially one intrave-
regimen included the following: (a) at the initial visit, a nous mega dose (equivalent to 1,000mg of Prednisone)
history of amaurosis fugax before visual loss occurred; followed by high-dose (80120mg) oral Prednisone to
(b) complete or marked loss of vision in one eye; (c) patients who present with: (1) history of amaurosis
early signs of involvement of both eyes; (d) severity of fugax, (2) complete or marked loss of vision in one eye
GCA disease process, as judged from levels of sys- or (3) early signs of involvement of the second eye.
temic symptoms, erythrocyte sedimentation rate (ESR) This is because these patients have a high risk of fur-
and C-reactive protein (CRP); (e) progressive visual ther visual loss, and one must try to achieve a high
loss with or without corticosteroid therapy; (f) response blood concentration of corticosteroids immediately
of ESR, CRP and symptoms to the corticosteroid ther- which is not possible with oral therapy.
apy; and (g) other confounding factors such as other
systemic diseases. In view of these factors and also
because of the marked inter-individual variation in
response to corticosteroid therapy, we found that nei- Oral Corticosteroid Therapy
ther the mg/kg strategy for Prednisone nor any one
standard regimen advocated by some workers was a This is the most common mode of corticosteroid ther-
valid approach to the management of GCA. My overall apy in GCA. Almost all of the information is based on
strategy for the corticosteroid therapy regimen was: the recommendations of the rheumatologists. There is
the greater the visual loss or potential for visual loss, marked controversy about the starting dose of
the more aggressive the corticosteroid therapy, to pre- Prednisone. For example, various authors have recom-
vent any further visual loss. In the light of these factors, mended an initial dose of Prednisone as low as 20mg/
initially intravenous therapy was given to 46% with day or as high as 60mg/day, lower doses being advo-
visual loss and 8% without visual loss. Of the patients cated for patients without ocular or cerebrovascular
who had initial intravenous therapy, a median of 3 symptoms of GCA [29, 30]. Myles etal. [31] stated that
doses (equivalent to 1,000mg of Prednisone in 75%, patients with a starting dose of 20mg of prednisolone or
800850mg in 13%, 500mg in 6% and 300330mg in less daily fared at least as well as those starting on
6%) were given every 8h over a median of 24h. The higher doses. It has been suggested that mean initial
initial intravenous dose in patients without visual loss dose of Prednisone of 0.7mg/kg/day [27, 32] and the
was not significantly different from those with visual mean maintenance dose of 10mg/day [32] are adequate.
loss. After the initial intravenous corticosteroid therapy Wilke and Hoffman [30] commented that generally
or for those without intravenous corticosteroid therapy, patients who require prolonged corticosteroid therapy
all patients were started on high-dose oral Prednisone. of more than 1520mg Prednisone daily may constitute
The initial starting oral Prednisone dose was 5560mg a minority of all GCA patients. Kyle [29] advised giving
in 22%, 80mg in 55% and 100160mg in 23%; this 40 mg/day prednisolone although, according to her,
was not significantly different in patients without and some may need 6080 mg; she recommended a slow
with visual loss. At the end of the study, data analysis reduction in corticosteroid dose, and a maintenance
showed that there was no evidence that intravenous dose of 7.5mg prednisolone after 69months. Salvarani
megadose corticosteroid therapy was more effective et al. [22] recommended giving an initial dose of
than oral therapy in improving vision [19] or prevent- 4060mg/day of Prednisone, and if the patient does not
ing visual deterioration due to GCA [20]. respond promptly, the dose should be increased.
This marked variation in the starting dose of
Prednisone recommended by rheumatologists may be
partly due, as has been said, to their dealing with a
My Recommendation for Intravenous
mixed population of GCA and PMR. It is well-
Corticosteroid Therapy established that patients with PMR require a much
smaller dose than GCA patients because they have
Available evidence, therefore, indicates that there is no only rheumatic symptoms with no risk of visual loss.
difference in visual outcome in GCA between the By comparison, in GCA patients the primary objective
intravenous and oral corticosteroid therapy. In view of of corticosteroid therapy is prevention of visual loss.
In our study [21] of oral corticosteroid therapy, the patients within 1year. This prevented serious adverse
median starting dose of Prednisone was 80 mg/day. side effects related to GCA, indicating that less toxic
The starting dose was 80mg/day in 44% of the patients, therapeutic measures are needed. In this study, relapses
100 mg/day in 40%, and 5560 mg/day (only 1 or recurrences occurred in 47.5%, which means that
patient had an initial dose of 55 mg) in 16%. The almost half of the patients received inadequate corticos-
starting dose of Prednisone was significantly teroid therapy and were at risk of losing vision with this
(p<0.0001) greater for those with visual loss (55 treatment regimen, i.e., this was not adequate therapy to
60mg/day in 8.3%, 80mg/day in 36.5%, and 100mg/ prevent blindness in GCA. Most recently a
day in 55%) than those without visual loss (60mg/day rheumatologist group [23] advocated the following
in 31%, 80mg/day in 59%, and 100mg/day in 10%). regimen of tapering of corticosteroid therapy. The ini-
The various factors that influenced the dosage are dis- tial dose of glucocorticosteroids is usually given for
cussed above. Doses of less than 80 mg Prednisone 24weeks until all reversible signs and symptoms have
were all started by the referring ophthalmologists and resolved and acute phase reactants are back to normal.
the patients were usually left on those. Then, the dose can be gradually reduced each week or
every 2weeks by a maximum of 10% of the total daily
dose. The necessary duration of glucocorticosteroids
therapy is variable, but in most cases it can be discon-
Tapering Down of Corticosteroid
tinued within 12years.. Rheumatologists in general
Therapy to Prevent Visual Loss recommend systemic symptoms as their guide in taper-
ing down of corticosteroid therapy and its duration.
This is an even more controversial issue, and almost all In sharp contrast to this view by the rheumatologists,
of the information is based on studies of rheumatolo- our 27-year long systematic, longitudinal study [21] of
gists who deal with both PMR and GCA; as mentioned 145 temporal artery biopsy confirmed GCA patients to
above, they not infrequently mix the two diseases in prevent visual loss has provided totally different infor-
their management with corticosteroid therapy. We mation. This study showed that there is a marked inter-
know that PMR is not associated with visual loss individual variation in the required tapering regimen,
whereas irreversible visual loss is the dreaded compli- maintenance dose required and the time it takes to reach
cation of GCA. Therefore, from the ophthalmic point that goal among various GCA patients. Therefore, no
of view, it is extremely important to prevent visual loss generalization at all is possible regarding tapering
from GCA during the tapering phase and duration of down the Prednisone; it has to be individualized; like-
corticosteroid therapy. wise, there is no set formula or any other way to pre-
Myles etal. [31] recommended tapering Prednisone dict the maintenance dose required by a particular
by 5mg monthly to 10mg and then by 1mg monthly. patient. The cook-book formulae about tapering pred-
Proven etal. [33] retrospectively evaluated the course nisone therapy in GCA offered by rheumatologists are
of glucocorticoid therapy in residents of Olmsted inadequate to prevent visual loss.
County, Minnesota, USA, by obtaining follow-up infor- Table13.1 shows the prednisone tapering doses and
mation of 120 GCA patients treated by different physi- the time it took to reach various levels, and Table13.2
cians. In this study, all responded rapidly to corticosteroid shows the lowest maintenance dose of Prednisone
therapy (median initial dosage 60mg prednisone/day). achieved and the time it took to reach that level, as
The dosage was later reduced according to the treating seen in our study [21]. The median time to reach the
physicians judgment. The median duration required to lowest maintenance dose of Prednisone at which the
reach 7.5 mg/day was 6.5 months and the median ESR and CRP stayed stable or went lower was
duration required to reach 5 mg/day was 7.5 months. 48.7months (95% CI: 34.6, 71.4months). Comparing
Relapses or recurrences occurred in 57 (47.5%) patients. patients with and without ocular involvement showed
For the 87 patients followed to discontinuation of no significant difference for (1) tapering and the time it
therapy and permanent remission of GCA (median of took to reach various levels (p-value between 0.33 and
22 months), the total median dose of prednisone was 0.69) and (2) to achieve the lowest maintenance dose
6.47g. They concluded that the prednisone dosage was of prednisone achieved and the time it took to reach
reduced to physiologic levels in three-fourths of the that level (log-rank test p=0.359).
Tapering Down of Corticosteroid Therapy to Prevent Visual Loss 231
Table 13.1 Kaplan-Meier curve estimates of the median, 25th and 75th percentile of the length of time (months) to reach
a Prednisone dose of 40, 30, 20, 10, and 5mg/day in 145 temporal artery biopsy confirmed GCA patients [21]
Prednisone dose Median time (months) 25th percentile 75th percentile
(mg/day) to reach dose (95% CI) (95% CI) (months) (95% CI) (months)
40 2.17 (1.81, 3.12) 1.12 (0.95, 1.48) 5.42 (3.91, 6.38)
30 5.19 (3.72, 6.38) 2.43 (2.07, 2.70) 10.32 (7.99, 13.68)
20 9.27 (7.36, 11.84) 4.41 (3.88, 5.95) 21.70 (13.18, 29.95)
10 26.66 (18.35, 38.01) 11.28 (9.67, 13.91) 50.33 (39.45, 73.12)
5 58.59 (49.25, 78.87) 30.87 (23.41, 43.99) 95.84 (72.69, a)
Estimate could not be calculated from the data
a
Table13.2 The Kaplan-Meier product-limit estimates of the cumulative proportion of GCA patients that attained the lowest oral
Prednisone dose at selected time points in 145 temporal artery biopsy confirmed GCA patients [21]
Duration of Number of patients still Total number that had Percentage that had attained
corticosteroid use being followed that had attained lowest dose lowest dose (95% CI)
(months) not reached lowest dose
6 119 7 5.11% (1.41%, 8.81%)
9 107 11 8.37% (3.61%, 13.13%)
12 98 15 11.88% (6.20%, 17.56%)
18 82 21 17.64% (10.70%, 24.58%)
24 65 30 27.35% (18.80%, 35.90%)
36 44 41 41.07% (30.98%, 51.16%)
48 34 47 49.91% (39.13%, 60.69%)
60 26 51 55.85% (44.89%, 66.81%)
72 19 55 63.03% (51.80%, 74.26%)
84 12 59 72.07% (60.60%, 83.54%)
I have seen some patients whose corticosteroid for at least 3months. Table13.3 gives details of these
therapy could be tapered fairly rapidly and to a very ten patients.
low maintenance dosage (as low as 1 mg Prednisone The Kaplan-Meier curve showing the distribution
daily) but others who took several years before the of continued use of corticosteroid therapy for patients
optimum lowest maintenance dosage was achieved with and without visual loss in our study is shown in
(based on stable ESR and CRP levels), which is usu- Fig. 13.1. Comparison of these two distributions
ally much less than 10 mg of Prednisone; but any showed that those with visual loss were more likely
attempt to go below that dosage caused the ESR and to stay on corticosteroid therapy longer than those
CRP to rise immediately. I have not found any side- without (log-rank test p=0.05). At 2years after start
effects among those patients on the low maintenance of corticosteroid therapy, 98.36% (95% CI: 95.17%,
dose (5 mg Prednisone) for years or even decades. 100%) of those with visual loss were still on corti-
We have found that, unfortunately, most physicians costeroid therapy compared to 92.64% (95% CI:
are not aware of the complexity of corticosteroid ther- 84.56%, 100%) of those without visual loss. After
apy management in GCA. Of the 145 patients in the 5 years, 92.47% (95% CI: 83.94%, 100%) of those
study and followed in my clinic, only 10 (7 without with visual loss and 86.11% (95% CI: 74.59%,
visual loss, 3 with visual loss) were able to stop corti- 97.63%) of those without visual loss were still on
costeroid use completely and maintained this status corticosteroid therapy.
Table13.3 Duration of corticosteroid therapy before stopping loss. The primary criterion invariably advocated by
that and the length of time been off corticosteroid therapy [21] rheumatologists is the use of systemic symptoms, and
Ocular Duration Length of time
only secondarily the levels of ESR and CRP. As dis-
involvement of corticosteroid have been off
use before corticosteroid cussed above, there are markedly conflicting recom-
stopping (months) mendations about the tapering of corticosteroid
(months) therapy, based almost invariably by rheumatologists
No 8.5 95.6 on a set cook-book formula for all individuals with
No 9.7 12 GCA. However, our study [21] showed that this one-
size-fits-all regimen is totally inadequate to prevent
No 21.7 21.6
visual loss from GCA, because of the marked inter-
No 29.5 17.5 individual variation in the tapering regimen, the
No 34.6 16.0 maintenance dose required and the time the patient
takes to reach the goal of the lowest possible mainte-
No 68.6 22.5
nance dose.
No 142.0 48.4 My experience of dealing with GCA patients over
Yes 13.4 129.1 the past 40years has shown that the guiding principal
Yes 31.8 61.2 in monitoring and tapering down the corticosteroid
therapy and finding the maintenance dose is to achieve
Yes 37.8 43.9
the lowest levels of ESR and CRP (acute phase reac-
tants) with the lowest possible dose of Prednisone. It is
frequently stated by rheumatologists that systemic
Criteria Used for Monitoring
symptoms, rather than ESR and CRP, should be the
Corticosteroid Therapy guide for tapering down or stopping corticosteroid
therapy; this can be a dangerous approach, because
This is a key topic because it determines the tapering these patients may lose vision without any warning
regimen of corticosteroid therapy initially and during systemic or ocular symptoms, and moreover, 21.2% of
follow-up, and the duration of therapy to prevent visual the patients with visual loss have the occult GCA [14].
1.00
Distribution of Patients Still on Steroid Therapy
0.80 Log-rank test p=0.051
0.60
0.40
Fig.13.1 The Kaplan-Meier

curve showing the distribu- 0.20
tion of continued use of With visual loss
corticosteroid therapy for
GCA patients with and Without visual loss
without visual loss over
0.00
a period of 20years
0 24 48 72 96 120 144 168 192 216 240
and Zimmerman [21]) Months
Criteria Used for Monitoring Corticosteroid Therapy 233
I have found a titration of the corticosteroid dosage monitor corticosteroid therapy to prevent visual loss.
with the levels of ESR and CRP is the only safe and In my studies I have found that normal levels of ESR
reliable method for tapering down and follow-up of and CRP, particularly of the ESR, vary widely (see
corticosteroid therapy in order to prevent any further Chap. 11). Diabetics usually tend to have higher lev-
visual loss. I have found no other parameter as sensitive els of both ESR and CRP than non-diabetics. The
and reliable a guide as ESR and CRP to monitor treat- tapering down of corticosteroid therapy should not
ment in these patients [17, 21]. be started until both the ESR and CRP have reached
their lowest and stable levels.
1. Initially, Soon after the Start of Corticosteroid In our study [21], the median ESR at first visit was
Therapy 87 (interquartile range of 56113) mm/h with minimum
Based on our study [21] I have found the only reli- of 4mm/h and maximum of 150mm/h. Those with no
able method is by doing ESR and CRP. After the start visual loss had a significantly greater initial ESR
of high-dose corticosteroid therapy, both ESR and (median of 102 mm/h; interquartile range of
CRP are repeated while the patient is on the high 70124mm/h) than those with visual loss (median of
dose, every 23 days. As shown in Figs. 13.2 and 80 mm/h; interquartile range of 52108 mm/h)
13.6 both ESR and CRP progressively come down (p=0.004). The median lowest ESR achieved after cor-
till they reach a stable level. ESR takes longer than ticosteroid therapy was 9 (interquartile range of 515)
the CRP to reach a stable level. The lowest levels of mm/h, with a minimum of 1 mm/h and a maximum
ESR and CRP achieved by this method actually rep- 41mm/h. Figure13.3 gives the initial and lowest ESR
resent the baseline ESR and CRP for that individual. in patients with and without visual loss. The median
After that, that level acts as the benchmark to main- time to reach the lowest level of ESR with high dose
tain while subsequently tapering the therapy. I have corticosteroid therapy was 24days (interquartile range
found that to be the only satisfactory means to of 15days to 38days). The median CRP at first visit
a 25
20
C-Reactive Protein (mg/dl)
b 100
15 1 2 3 4 5 6
90
80
70
ESR in MM/Hr
10 60
50
40
5 30
20
10
0 0
0 5 10 15 0 2 4 6 8 10 12 17 24 27 31 39 45 50 67 75 90
Days since start of steroid therapy Days since start of steroid therapy
Fig.13.2 Graphs of C-reactive protein levels (left) and erythrocyte sedimentation rates (ESR right) of six patients with GCA,
showing their initial responses to high dose corticosteroid therapy (Reproduced from Hayreh and Zimmerman [21])
160 is to maintain, as far as possible, the previously

achieved low levels of ESR and CRP with the lowest
possible dose of Prednisone. Because of the tremen-
140
dous variation in response of different patients, no
generalization is possible. I have found that one can
120 taper down Prednisone fairly fast in some, while it
takes a much longer time for others. For example, I
100
have found that some patients take many months or
even over a year to go down from a high dose of
ESR mm/Hr
80mg Prednisone to about 40mg Prednisone. If at

80 any stage of lowering Prednisone dose, the ESR and
CRP levels started to rise sharply (without any other
60
evident cause), the Prednisone dose is immediately
increased to the previous level and that lowers the
ESR and CRP to stable levels; thus, the Prednisone
40 dosage must be individually tailored, to maintain
stable ESR and CRP levels for that individual. The
20 patients should be advised to consult their physicians
for a regular physical check-up for any side-effects
of corticosteroid therapy. Initially, patients are fol-
0 lowed at least every 24 weeks till the Prednisone
Initial Lowest Initial Lowest
Without Visual Loss With Visual Loss dose goes down to 40 mg daily, and thereafter at
slightly longer intervals, but never longer than every
Fig.13.3 Scattergram showing initial (before the start of corti- 46weeks, till they reach a low, stable maintenance
costeroid therapy) and lowest (after high dose corticosteroid
therapy) erythrocyte sedimentation rates (ESR) of 145 GCA
dose, after which they are followed at least once
patients without (49) and with (96) visual loss (Reproduced every 3months. Finally, when they have been stable
from Hayreh and Zimmerman [21]) for a long time on a small maintenance dose of
Prednisone, they can be followed every 6months. To
prevent visual loss, as stated above, the follow-up
was 8.1 (interquartile range of 413.2) mg/dl, with min- protocol has to be individualized, depending upon
imum of 0.5 and maximum of 60.5mg/dl. There was no the patients response to tapering of the corticoster-
significant difference in initial CRP between those oid therapy. In my study, where the patients have
without visual loss (median of 9.1; interquartile range been followed for many years or even decades, not a
of 417.1) and those with visual loss (median of 6.8; single patient suffered any further visual loss after
interquartile range of 3.111.7) (p=0.207). The median the first 5days from the start of high dose corticos-
lowest CRP achieved after corticosteroid therapy was teroid therapy.
0.2 (interquartile range of 0.10.4) mg/dl, with a mini- 3. Does a Reduction of 1Mg of Prednisone Make a
mum of 0.1 mg/dl and a maximum 6.6 mg/dl. The Difference?
median time to reach the lowest CRP with high-dose The answer is definitely yes, from my studies deal-
corticosteroid therapy was 11days (interquartile range ing with GCA and other rheumatologic and other con-
of 7days to 21days). ditions requiring systemic Prednisone. I have seen that
2. For Tapering Down Corticosteroid Therapy and frequently. I have known several GCA patients who
to Determine the Maintenance Dose of Therapy were on a 1 mg of Prednisone maintenance dose for
After ESR and CRP have reached their lowest, stable 23years with stable ESR and CRP, but when I stopped
levels, Prednisone dosage is slowly decreased, guided their Prednisone, there was an immediate rise of both
entirely by the levels of ESR and CRP. As a rule, ESR and CRP to high levels, so that they had to resume
during follow-up, I repeat ESR and CRP for every Prednisone therapy immediately to prevent visual loss.
patient at every visit, and before deciding about I tried that experiment with them several times, with
lowering the dose of Prednisone. The basic principle exactly the same response. Thus, for them 1 mg
Duration of Corticosteroid Therapy 235
Prednisone was enough to control the disease. I have reasonable [22, 23, 29, 31, 3638] because they believe
also seen that happen to a patient with orbital myositis that GCA is a self-limited disease, lasting less than a
and scleritis; lowering their Prednisone dose by 1mg year [37]. But repeat temporal artery biopsy has shown
resulted in a flare up of the disease. Like most physi- evidence of active disease even after 9years of corti-
cians, I did not believe initially that that could happen, costeroid therapy [32, 3942]. Some have stated that
but repeated experience has made me a firm believer some patients have a chronic relapsing course and
that in some cases 1mg of Prednisone does make all might need low doses of glucocorticosteroids for sev-
the difference. Therefore, while tapering down the eral years [33, 43, 44]. Multiple reports in the literature
Prednisone dose in GCA, I have found that if going have concurred that GCA patients require corticoster-
down in steps of a higher dose results in a rise of ESR oid therapy for much longer than 2years. For example,
and CRP, tapering down by a smaller dose can prevent Fernandez-Herlihy [45] reported a mean duration of
that from happening. Tapering down of Prednisone in corticosteroid therapy of 6years. Andersson etal. [43],
GCA is certainly laborious and time consuming in on a follow-up of 90 GCA patients for 916 years
some cases, but one has to individualize the tapering (median 11.3years), found the mean duration of corti-
regimen in each case. costeroid therapy 5.8years (range 012.8years), with
43% of them on therapy for 5 years, and 25% still
being treated with 1.2510mg of prednisolone daily
(median dose 5mg) after 9years. Gouet etal. [32], in
Alternate Day Corticosteroid Therapy 87 temporal artery biopsy positive GCA patients,
reported that although corticosteroid could be discon-
Treatment with alternate-day corticosteroid therapy tinued in 21 cases after 36 months on average, the
has been proposed, to reduce the risk of adverse reac- other patients could not be weaned off, even though
tions related to this therapy. Hunder etal. [34], based four of them had been on corticosteroid for more than
on a prospective study of 60 GCA patients, con- 10years. There are other reports of corticosteroid ther-
cluded that alternate-day therapy is associated with a apy being required in some patients for an indefinite
higher rate of treatment failure than is daily adminis- period [42, 43, 46, 47]. Graham [48], on a follow up of
tration (70% vs 20%) and that it does not satisfacto- GCA patients from 6months to 12years reported that
rily control symptoms in most patients and cannot be there appear to be two groups of patients who suffer
recommended for those with severe disease. with the disease: those who respond well to treatment
Bengtsson and Malmvall [35] initially treated GCA and those who follow a chronic course despite treat-
patients daily with corticosteroid, followed by a ment. Arashvand etal. [49] recently reported, in a ret-
gradual transition to alternate-day treatment, and rospective study of 30 temporal artery biopsy proven
concluded that it was possible to maintain 67% of 27 GCA patients, the duration of treatment in patients
patients on an alternate-day treatment regimen, but with and without eye involvement. They found that the
the regimen was abandoned in 33% because of clini- mean duration in patients with eye involvement was
cal symptoms on the day off. My experience of significantly (p=0.0018) longer than in those without
dealing with about 300 GCA patients supports the eye involvement.
view that alternate day corticosteroid therapy is not The findings of our study [21] showed that GCA
indicated at any stage. patients require corticosteroid therapy for much
longer than the 12years suggested by some rheuma-
tologists [23]. This is evident from the data in
Tables13.113.3 above. The median time to reach the
Duration of Corticosteroid Therapy lowest maintenance dose of Prednisone at which the
ESR and CRP stayed stable or went lower was
Because of the frequent systemic side-effects of 48.7 months (95% CI: 34.6, 71.4 months). For the
chronic corticosteroid therapy, the total duration of patients who attained the lowest corticosteroid dose,
corticosteroid therapy in GCA has become highly con- the median lowest dose was 7 mg/day (interquartile
troversial. The view among rheumatologists is that range of 116mg/day). Only 10 of the 145 patients
stopping corticosteroid therapy after 2 years is followed in our study were able to stop corticosteroid
therapy without a rise of ESR and CRP; Table 13.3 during corticosteroid tapering, with a mean 1.57
gives the duration of their corticosteroid therapy relapses per patient. In 56 patients free of treatment
before stopping. Thus, our data show that most GCA (mean treatment duration: 40months), a relapse of the
patients require corticosteroid therapy virtually indef- disease was observed in 48% of patients 125months
initely. As discussed above, in our study [21] patients after the end of the treatment. Proven etal. [33] reported
with visual loss were more likely to stay on corticos- relapses in 47.5%. Hernandez-Rodriguez et al. [53]
teroid therapy longer than those without (p=0.05) reported that GCA patients with a strong initial sys-
(Fig.13.1). temic inflammatory reaction have higher and more
The invariable reason for giving GCA patients only prolonged corticosteroid requirements (i.e., time taken
small doses of corticosteroids, for only a short time, is to achieve a maintenance dose of<10 mg/day
the risk of systemic side-effects of prolonged corticos- Prednisone), and experience more disease flares during
teroid therapy. But it is important to bear in mind corticosteroid therapy than patients with a weak sys-
a very important basic fact from the patients point of temic acute phase response. Wilke and Hoffman [30]
view: fear of going blind is next to the fear of death. If advocated that there should be a follow-up of at least
the balance of risks and benefits of corticosteroid ther- 1 year after discontinuation of treatment to capture
apy is clearly explained to GCA patients, I have found relapses.
them willing to risk the side-effects rather than risk
going blind, especially if they have already suffered
some visual loss. Guiding Rules for Reduction
In GCA, as in most other rheumatological dis-
of Corticosteroid Therapy in Our
eases, the corticosteroid therapy is purely suppressive
and not curative; the vast majority of patients require Study [21] to Prevent Relapses
a highly variable, carefully adjusted maintenance
dose for many years and perhaps the rest of their life The dose of corticosteroid therapy for GCA was
to prevent visual loss. guided solely by the levels of ESR and CRP, NOT by
systemic symptoms. As discussed above, tapering
down of Prednisone from the initial high dose was not
started till the levels of both ESR and CRP had stabi-
Relapses of GCA on Reduction
lized at a low level (Fig. 13.2). After that, tapering
of Corticosteroid down of the dosage of Prednisone was done very grad-
ually, once again guided only by the ESR and CRP
Relapses or flare-up of symptoms on reduction or stop- levels. Any rise of ESR and CRP was a sign that the
page of corticosteroid therapy in GCA patients has dose of Prednisone was not adequate or the tapering
been commonly reported [32, 33, 43, 4652]. was done too soon unless this rise was due to some
Andersson etal. [43] recorded a relapse rate of about unrelated cause or intercurrent disease. ESR and CRP
50%, regardless of the time after diagnosis, when an rise quickly any time the corticosteroid dosage goes
attempt was made to withdraw the treatment. Forty-six below the optimal level in GCA patients, and that level
per cent of the relapses occurred within 1month and varies widely not only among various patients but also
96% within 1year of the end of treatment. Most of the in the same patient at different dosages (Table13.1).
flare-ups occurred during the first year of therapy and In the event of a rise of the levels of ESR and CRP, it
in 55% of the patients on a prednisolone dosage of is essential to go up to the next higher level of
5mg or less. Weyand etal. [51], in a prospective treat- Prednisone dose or even higher. I found ESR and CRP
ment study with 60mg/day Prednisone, found 60% of the most sensitive and reliable tests in monitoring the
the patients developed symptoms of recurrent disease, activity of GCA [17, 21] and for tapering corticoster-
while the Prednisone dosage was being reduced. They oid therapy. A close follow-up of these patients is
concluded that smoldering disease activity may expose essential to monitor the dosage and side-effects of cor-
GCA patients to the risk of progressive vascular dis- ticosteroid therapy. I have discussed my follow-up regi-
ease. Hachulla et al. [52], in a study of 176 GCA men above. On the regimen of treatment described
patients, reported a relapse in 62.4% of the patients above, none of about 300 GCA patients, whom I have
Is GCA a Self-Limited Disease? 237
treated over the past 40years had a relapse or lost any develop systemic symptoms of GCA at any stage, i.e.,
vision except for the following one unique patient they have occult GCA [14]. I have seen GCA patients
who did suffer a relapse. who suffered relapses and lost vision, often in both
I first saw this 78-year old woman in February 1998 eyes, while their corticosteroid therapy was managed
with anorexia, weight loss and pains in shoulder and by other physicians; it was invariably caused by a rapid
hip, without any visual loss. Her temporal artery biopsy tapering down or even stopping of corticosteroid ther-
was positive for GCA. She was treated with 80 mg apy, usually guided by physical symptoms rather than
Prednisone daily to begin with and that was gradually by ESR and/or CRP. In my study, when a rise in ESR
tapered down, guided by ESR and CRP levels, and she and CRP occurred soon after the corticosteroid dosage
went down to a 4 mg Prednisone daily maintenance was lowered, there was NEVER a simultaneous cor-
dose in September 2001 with stable ESR (4mm/h) and responding change in the systemic symptoms of GCA.
CRP (<0.5mg). When seen in July 2003 for follow-up, So any reliance on systemic symptoms of GCA as a
she complained of weight loss, anorexia and jaw and guide to corticosteroid therapy exposes the patient to the
ear pain for 3 months. At that time her ESR was risk of blindness. Our study [21] showed that, as dis-
>100 mm/h and CRP 13.4 mg. She was started on cussed above, ESR and CRP are the only tests to pre-
80 mg Prednisone once again. At next visit 3 weeks vent relapse and loss of vision. No other cookie-cutter
later the ESR was 1mm/h and CRP <0.5mg. Prednisone formula is effective to prevent visual loss.
was gradually tapered down, guided by ESR and CRP It is essential to follow these patients very closely
levels, and when last seen in March 2007 she was on even after corticosteroid therapy is stopped, guided by
6mg Prednisone with normal ESR and CRP. The rea- ESR and CRP levels. As previously discussed, I have
son why she had a flare up, after having been stable for had a few patients who were on 1mg of Prednisone for
years on a small maintenance dose of Prednisone and 23years with stable ESR and CRP, and then stopped
stable ESR and CRP levels, remained a mystery. taking 1mg of Prednisone; they immediately experi-
Thus, in my study, only one out of about 300 enced a rise of both ESR and CRP to high levels, so
patients had a relapse for no apparent reason when she that they had to go back immediately on Prednisone
was doing very well and was stable on a low mainte- therapy. I tried that several times in those patients, with
nance dose of Prednisone for 22 months. This is in exactly similar responses. Thus, in their case 1 mg
sharp contrast to the high rate of relapses (from about Prednisone was just enough to control the disease
50% to two thirds of patients) reported in the rheuma- something most physicians would find hard to
tologic literature [33, 43, 51, 52]. The implication is believe.
that those patients relapsed because they either received
an inadequate dose of corticosteroid therapy, or the
therapy was stopped prematurely. During a relapse
a GCA patient is at potential risk of losing vision. That Is GCA a Self-Limited Disease?
is not acceptable from the ophthalmic and ethical point
of view, when the primary objective of corticosteroid The self-limited nature of GCA is often stressed in
therapy in GCA is to prevent visual loss. text-books and by many physicians (particularly rheu-
matologists) who feel strongly that there is no need to
keep these patients on prolonged corticosteroid ther-
apy, because of its side-effects. For example, Huston
Role of Systemic Symptoms in Regulation
et al. [37] stated that GCA lasts for less than a year.
of Corticosteroid Therapy Because of that belief, other rheumatologists believe
that stopping corticosteroid therapy after 2 years is
Common advice in the rheumatologic literature is that reasonable [22, 23, 29, 31, 3638]. But, as discussed
alterations in treatment should be based on the clinical above, repeat temporal artery biopsy has shown
picture, rather than on laboratory tests [29] This can evidence of active disease even after 9years of corti-
be dangerous patients may lose vision irrevocably costeroid therapy [32, 3942], and there are multiple
without any warning systemic symptoms of GCA; reports which confirm that GCA patients require corti-
moreover, 21% of the patients with visual loss never costeroid therapy for much longer than 2years.
Our study [21] has shown that most GCA patients resistant GCA is timidity; an initial dose of Prednisone
require lifelong corticosteroid therapy to prevent visual as low as 20 mg/day is totally inadequate to control
loss. I have had patients who went blind after their active GCA and prevent visual loss, although it may be
local physician had stopped corticosteroid therapy sufficient for PMR. I feel that the basis for this confu-
after a year or two under the impression that GCA is sion may be, as discussed above, the mixing of PMR
a self-limited disease, and also under the belief that and GCA patients in rheumatologic studies, with more
prolonged corticosteroid therapy is deleterious. In this cases of PMR than of GCA.
connection, it is pertinent to remind the reader that the
patients fear of visual loss is almost as great as the
fear of death, and they will always accept a mode of
Visual Improvement with
therapy that prevents them from going blind, even at
the risk of a certain amount of side-effects. In GCA, as Corticosteroid Therapy in GCA
in most other rheumatological diseases, the corticos-
teroid therapy is purely suppressive and NOT curative; A review of the literature reveals highly conflicting
the vast majority of patients require a highly variable, information about visual improvement with corticos-
carefully adjusted maintenance dose, probably for the teroid therapy in eyes with visual loss due to GCA,
rest of their life. varying from claims of complete recovery or dramatic
improvement of visual acuity, recovery in almost half
of cases, or only a mild recovery of visual function.
The only way to pinpoint the reasons for these highly
Corticosteroid Resistant GCA variable claims of visual improvement is by critically
reviewing the details of the major reports of visual
Rheumatologists often mention corticosteroid resis- improvement and the methods used to evaluate it.
tant GCA patients. For example, Wilke and Hoffman It appears that Whitfield etal. [55], in 1953, were
[30] defined corticosteroid resistant GCA patients as the first to ask the question: Has cortisone or ACTH
those who after a few months of treatment, may any power to restore sight when visual loss has
require daily Prednisone doses of 15mg or more or the occurred? They reported that out of 10 patients with
very uncommon patient who does not respond to initial GCA related visual loss, 5 with partial visual loss for
therapy using 2060mg Prednisone per day. Ghanchi 10days showed some visual acuity and field improve-
and Dutton [54] stated that failure to control arteritis ment after treatment; but they warned that the degree
even with high doses of corticosteroid is not uncom- of change was relatively small and in no case was
mon and relapses while on corticosteroid treatment normal sight restored. They gave no ophthalmoscopic
could indicate corticosteroid resistance. Chevalet etal. details, and in one case alleged only that visual acuity
[27] stated that corticosteroid resistance is a real risk and fields showed considerable improvement.
factor for GCA complications. Bennett [56] reported visual acuity improvement in 2
My experience of treating about 300 GCA patients of 9 eyes (in 6 GCA patients) from NLP to HM and
during the past 40years does not support the concept 20/200.
of corticosteroid resistant GCA at all. I have had The paper by Parsons-Smith [57] is invariably cited
many patients referred to me by outside physicians as evidence for improvement in visual acuity with corti-
with that diagnosis but when I treated them with ade- costeroid therapy; hence it deserves careful evaluation.
quate doses of corticosteroid, every single one of them He concluded that: ACTH given on the same day as
immediately responded to corticosteroid therapy. The blindness has restored sight in every eye so treated
problem is that, as stated above, the corticosteroid (my italics). ACTH may prove effective after a longer
dose, length of corticosteroid therapy and the mainte- period of blindness in cases of retrobulbar neuritis (my
nance dose required all vary tremendously among interpretation, arteritic-anterior ischemic optic neurop-
GCA patients, as shown by the data in Tables 13.1 athy (AION) ). Unfortunately, this retrospective study
13.3, and that no one set corticosteroid therapy regi- is hampered by drawbacks that confound the interpreta-
men is successful for all GCA patients. The basic tion of the results and conclusions; for instance: (a) None
reason for this misleading concept of corticosteroid of the patients had temporal artery biopsy to confirm the
Visual Improvement with Corticosteroid Therapy in GCA 239
diagnosis of GCA. (b) Improved visual acuity was acuity of 20/200 and an altitudinal field defect; in all 3
claimed in 13 of 25 patients with visual loss, whose cases treatment was started within 24h of the visual
diagnoses were: central retinal artery occlusion (6), loss and improvement occurred within 48h. In one of
branch retinal artery occlusion (1), retinal hemor- the eyes, they reported visual acuity improvement
rhages and exudates (4), papilloedema (7), retrobul- within 1 h of treatment from 20/200 to 20/30 which
bar neuritis (3), optic atrophy (2), central retinal seems unrealistic; in this eye and in others with altitu-
vein branch thrombosis (2). But a careful evaluation of dinal field defect, the possibility of simple eccentric
the fundus description in each eye reveals that most of fixation causing apparently improved vision later on
the eyes possibly had arteritic AION, then an unknown cannot be ruled out.
clinical entity. The possibility of nonarteritic AION Aiello et al. [59], in a retrospective study of 34
cases (with better visual outlook spontaneous visual acu- patients with visual loss due to GCA (25 of them with
ity improvement in non-arteritic AION in about 40% see a positive temporal artery biopsy), reported visual acu-
Chap. 17) having been included in this study cannot be ity improvement of more than 2 lines in 5 eyes. From
ruled out. We know that GCA is not a disease of veins, their Table13.3, it seems that none of these five patients
so that central/branch retinal vein thrombosis could not had temporal artery biopsy done. Only one eye had
have been due to GCA. (c) No visual fields were plotted. visual fields recorded. Again, the possibility of eccen-
(d) The reliability of the recorded visual acuity in some tric fixation and a less firm diagnosis as the reason for
patients is subject to question. In the light of all these the improvement cannot be ruled out. Moreover, spon-
limitations, the conclusions of this often cited study have taneous visual acuity improvement in non-arteritic
to be accepted with great reservations. Unfortunately, AION has been reported in about 40% (see Chap. 17).
the problems discussed here occur in many of the stud- Liu etal. [60], in a retrospective study of 41 GCA
ies reporting visual improvement, obviously resulting in patients (63 eyes) with visual loss, reported visual
controversy on the subject. improvement in at least one of the two eyes of 14
Cohen [40] reported 14 temporal artery biopsy con- patients (13 had AION and 1 posterior ischemic optic
firmed GCA patients with arteritic AION and stated neuropathy (PION)), in 8 from worse than 20/200 to
that visual acuity improved by more than 2 lines in 7 between 20/25 and 20/80. The time interval between
eyes, concluding: corticosteroids in high doses . the initial and the final improved visual acuity in AION
seem to be able to improve the prognosis for retention patients varied from 2 months to 3 years
of useful vision and recouping some of the lost vision.; (13.39.9 months; median 10 months) and 1 day in
but on a review of his data, we found actual improve- the PION patient.
ment in only 4 eyes. No other details were given, Gonzales-Gay et al. [61] in a retrospective study
except for one eye with an inferior visual field defect reported partial improvement of visual acuity in 12%
in which visual acuity improved from 20/400 to 20/40 of 69 biopsy positive GCA patients with visual involve-
in about 2 weeks, where the possibility of patient- ment. They concluded that after adjustment for the
learned eccentric fixation causing improved vision treatment regimen (intravenous pulse methylpredniso-
cannot be ruled out. Kearns (1973), discussing this lone versus oral prednisone), early treatment (within
paper, questioned the claims of visual improvement in the first day of visual loss) was the only predictor of
this study and stated that one should not expect real improvement.
improvement in vision in these cases. Kupersmith etal. [62], in a multicenter uncontrolled
Russell [58] reported treating 11 GCA patients (15 study of 7 patients (9 eyes) with visual loss, concluded
eyes) with visual loss with corticosteroid and antico- that patients with GCA related visual loss can improve
agulants and stated that visual acuity improved in one with treatment. The authors claimed that 4 of the 9
patient (one eye). Graham etal. [42] retrospectively eyes showed visual acuity improvement within 1month
reviewed 43 GCA patients with visual loss due to uni- of starting corticosteroids; however, they found no cor-
lateral AION (33), bilateral AION (5), central retinal responding central visual field improvement. Without
artery occlusion (4) and branch retinal artery occlusion concomitant visual field improvement, visual acuity
(1), treated with 250 mg intravenous hydrocortisone testing artifacts could not be ruled out; moreover the
followed by 80 mg oral prednisone. They reported reported visual improvement was of such a low magni-
some visual recovery in 3 AION eyes with a visual tude as to be of little practical importance [17].
Foroozan etal. [63] in a retrospective study of 32 history and not the effect of the megadose intravenous
consecutive patients with biopsy-proven GCA with methylprednisolone.
visual loss from arteritic-AION or central retinal artery Can claims of marked visual improvement with
occlusion reported improvement in visual acuity in corticosteroid therapy be reconciled with the very rare
13% with visual loss from biopsy-proven GCA, and all visual improvement seen by others? For this reason,
had arteritic-AION. Despite the improvement of visual we conducted a study [19] to investigate the incidence
acuity in these patients, perimetry revealed marked and extent of visual improvement with systemic corti-
constriction of the visual field in each affected eye. costeroids in patients with GCA in a large comprehen-
There is a problem with several reports claiming sive study. Our study [19] was based on 84 consecutive
visual recovery or visual improvement in the literature, patients (114 eyes) with visual loss, all of whom had
particularly after many months. For severely infarcted GCA confirmed by temporal artery biopsy and were
nerve fibers in the optic nerve head to survive for treated by us with high-dose systemic corticosteroid
months and recover much visual function later on therapy. Visual loss was due to arteritic AION (91%),
seems unrealistic, if the mechanism is to be explained central retinal artery occlusion (10.5%), cilioretinal
as survival of optic nerve fibers. In some of these artery occlusion (10%), and/or PION (4%) alone or
reports there are claims of visual acuity improvement in different combinations. Improvement in visual
without any visual field improvement. For a genuine acuity (of 2 lines) was present in 10.5% of 114 eyes,
visual acuity improvement, there has to be improve- but only 4% showed a corresponding central visual
ment in the central visual field, otherwise the claimed field improvement. Thus, improvement in both visual
visual acuity improvement simply represents the acuity and central visual field was found in only 5 (4%)
patient learning to use eccentric fixation to see better. eyes and of the 5 patients, 3 were treated with intra-
There are several more reports claiming visual venous and 2 with oral corticosteroid therapy. There
improvement with corticosteroid therapy, based on was no significant (p=0.672) difference in visual out-
only one or two cases, many of them with problems come among patients treated initially with intravenous
similar to those mentioned above. Where adequate corticosteroid versus those treated with only oral corti-
information was provided in the case reports, the costeroid. Comparison of patients with visual improve-
improved visual acuity without concomitant improvement in both visual acuity and fields versus those with
ment in visual field most probably represented learn- none suggested a shorter (p=0.065) interval between
ing to fixate eccentrically [64, 65]. onset of visual loss and start of corticosteroid therapy
The report by Matzkin etal. [66] illustrates another in the improved patients. Disparity between improve-
problem in the literature on visual improvement with ment in visual acuity and visual fields may be due to:
corticosteroid therapy in GCA. These authors reported (a) deceptively poor visual acuity at the first visit due
visual improvement with megadose intravenous meth- to emotional upheaval from sudden visual loss, and (b)
ylprednisolone in 2 eyes diagnosed with central reti- the recorded visual acuity improvement may repre-
nal artery occlusion. A detailed review of the cases sent simply a patients learning by experience to read
offers a different interpretation of their results. One of the test chart better by eccentric fixation. Similarly,
the two eyes in fact had a cilioretinal artery occlusion Danesh-Meyer etal. [69] in a recent multicenter pro-
and not central retinal artery occlusion, because the spective case series of 34 consecutive biopsy positive
ischemia involved only the inferotemporal retina and GCA patients, found that visual recovery is uncom-
there was a corresponding superior nasal quadrantic mon (5%) in patients who lose vision from GCA.
field defect a pattern inconsistent with central retinal Recovery in visual acuity was not associated with
artery occlusion. In our natural history study [67] of visual field or color vision improvement in this series.
cilioretinal artery occlusion, the final visual acuity According to the evidence-based classification
was 20/40 or better in 100% without treatment. [70, 71], in the literature on visual improvement with
Similarly in our natural history study of 260 eyes with corticosteroid therapy in GCA related visual loss, there
central retinal artery occlusion, there was spontaneous is neither a Class I study (evidence provided by a ran-
visual acuity improvement in 22% [68]. Thus, it is domized controlled clinical trial) nor a Class II study
possible that the improvement in visual acuity reported (evidence provided by prospective matched group
by Matzkin et al. [66] may simply represent natural cohort study), because from present knowledge it
Visual Improvement with Corticosteroid Therapy in GCA 241
would be unethical to conduct such studies, since common trunk results in both arteritic-AION and
corticosteroid therapy has proven to be the definitive central retinal artery occlusion, simultaneously [4,
treatment to prevent blindness in GCA. Similarly, in 13]. Since the posterior ciliary arteries also supply
our study [19], although ideally we would have liked the cilioretinal artery, occlusion of the former natu-
to give an identical regimen of corticosteroid therapy rally results in occlusion of the latter (see Figs.12.9,
to all patients, this was often not possible, because the 12.10, 12.17 and 12.18 in Chap. 12) [9, 13].
corticosteroid dose was influenced by a number of d. In a small number of eyes, arteries supplying the
clinical factors, as discussed above. Most of the posterior part of the optic nerve may become
reported studies are Class III type (based on a case involved, producing PION [13].
series subjected to a given intervention) or Class IV e. The most crucial issue influencing the possibility of
type (based on an interventional case report). Class IV recovery is the length of time the retina and optic nerve
studies, though possibly of interest, are mainly anec- can survive an acute ischemic insult, before they are
dotal in nature. irreversibly damaged. This depends upon two factors.
(i) The Severity of Ischemia
In arteritic-AION there is almost invariably com-
plete occlusion of the posterior ciliary arteries and
Why Is There Little Chance no blood flow in the optic nerve head (see fluores-
of Improvement in Visual Loss cein angiograms in Chap. 3), resulting in infarction
Due to GCA in Spite of High-Dose and severe damage (see Figs.12.312.8 in Chap.
Corticosteroid Therapy? 12). However, on fluorescein fundus angiography,
I have documented two rare exceptions to this:
(a) I have occasionally documented some residual
The answer to this question lies in understanding the
circulation still present in the retinal and/or
mechanism causing visual loss in GCA and the factors
posterior ciliary artery circulatory beds at the
influencing the optic nerve head circulation. The fol-
patients first visit, within a day or so of the
lowing pieces of information help to explain:
visual loss. The mechanisms responsible for
a. Visual loss in the 84 patients (114 eyes) in our this residual circulation in the optic nerve head
study [19] was due to arteritic AION (91%), cen- in complete occlusion of the posterior ciliary
tral retinal artery occlusion (10.5%), cilioretinal arteries are discussed elsewhere [8082].
artery occlusion (10%), and/or PION (4%) alone (b) In arteritic-AION eyes there is occasionally a
or in different combinations. markedly delayed choroidal circulation, indi-
b. GCA almost invariably causes occlusion of the cating that the posterior ciliary artery is
posterior ciliary artery, occasionally of the central severely narrowed but not completely
retinal artery, and rarely of the ophthalmic artery occluded. Such a severe narrowing would
[3, 6, 13, 7276]. cause a marked fall of perfusion pressure in
c. As discussed in Chap. 3, the posterior ciliary arteries the optic nerve head circulation. A further fall
are the main source of blood supply to the optic of perfusion pressure during nocturnal arterial
nerve head, so that their occlusion causes ischemic hypotension [8385] may result in only tran-
damage to the optic nerve head, resulting in arteritic- sient hypoperfusion or nonperfusion of the
AION. Histological studies [7276] in arteritic- optic nerve head and the restoration of blood
AION have demonstrated infarction of the optic pressure to normal levels upon awakening
nerve head, involving only a part or the whole of it could restore some circulation in the optic
(see Figs.12.312.8 in Chap. 12), depending upon nerve head. These two conditions, with some
the number of posterior ciliary arteries supplying the residual circulation, may help to explain true
optic nerve head and the area supplied by each [77 neuronal recovery, resulting in visual improve-
79]. In some eyes, the central retinal artery and the ment in a few eyes.
posterior ciliary artery arise by a common trunk (ii) The Duration of Acute Ischemia
from the ophthalmic artery (see Figs. 12.21 and In our experimental acute central retinal artery
12.22 in Chap. 12) [77], so that occlusion of their occlusion studies in elderly rhesus monkeys, central
retinal artery occlusion of less than 100 min pro- and safer [60, 89, 90]. Liu etal. [60] treated 9 patients
duced no apparent morphometric evidence of reti- with intravenous methylprednisolone and 5 with oral
nal or optic nerve damage, but central retinal artery prednisone, and noted a slight trend toward visual
occlusion of 105 up to 240min produced progres- improvement associated with intravenous corticoster-
sive damage, and central retinal artery occlusion of oid in patients with visual loss.
240min or more produced total optic nerve atrophy As mentioned above, our study [19] showed no sig-
and retinal damage [18, 86]. Thus the retina and nificant difference in visual outcome in patients treated
optic nerve can only survive severe acute ischemia initially with intravenous corticosteroid versus those
for a relatively short time. As mentioned above, his- treated with only oral corticosteroid.
tological studies in arteritic-AION have demon- The above critical review of the literature and dis-
strated infarction of the optic nerve head (see cussion, combined with my experience of multiple
Figs.12.312.8 in Chap. 12). Hence, after a critical clinical studies during the past 40years, seems to sug-
period, acute ischemia results in permanent neu- gest that the major limitations of most studies about
ronal damage, with no chance of visual recovery. the claimed visual improvement in GCA are;
By the time the vast majority of patients with visual
loss due to GCA are seen by an ophthalmologist, (a) The recording of apparent improvement in
they have already suffered permanent, irreversible visual acuity which may be simply the result of
damage from infarction. Moreover, if the arterial limitations inherent in testing visual acuity.
supply to the optic nerve head or retina is com- (b) The use of visual acuity improvement without
pletely occluded, neither corticosteroid therapy nor concomitant central visual field improvement as
any other treatment can re-establish the circulation the sole criterion for improved visual function.
and reverse the ischemic damage. Thus, in such (c) The omission of a consideration of the natural his-
eyes there seems to be little scientific rationale for tory of the visual outcome in a disease can result
any possibility of visual improvement with corti- in invalid claims. For example, Richard etal. [91]
costeroid therapy. Indeed, most of the cases reported reported that local fibrinolysis treatment in central
in the literature show no visual improvement with retinal artery occlusion resulted in visual acuity
corticosteroid therapy, and in our study 96% of the improvement in 66% of 46 treated eyes. However,
eyes belonged to this type. This was also stressed my study [92] showed no real difference in final
by Kearns [87], Bengtsson and Malmvall [50] and visual acuity between Richard etal.s [91] treated
Thystrup etal. [88]. Therefore, it is vital to empha- central retinal artery occlusion and my [92]
size that the goal of corticosteroid therapy in GCA untreated central retinal artery occlusion patients.
is to help prevent visual loss, not to reverse it.
Unfortunately, claims of marked or dramatic visual
improvement with corticosteroid therapy reported in
Is Intravenous Corticosteroid Therapy the literature can mislead physicians and, even worse,
More Effective than Oral Therapy in desperate GCA patients, into expecting significant
improvement in visual function with corticosteroid
Causing Visual Improvement in GCA
therapy. This can have medicolegal implications.
Patients? Since there is little or no chance of recovery or
a cure for visual loss in the vast majority of GCA patients,
A review of the old literature reveals widespread dis- prevention of visual loss is crucial. Kearns [87] has cor-
agreement on the role of intravenous versus oral corti- rectly stated that GCA ranks as the prime medical emer-
costeroid therapy in visual recovery in GCA. The gency in ophthalmology, there being no other disease in
majority of the reported patients with visual loss due to which the prevention of blindness depends so much on
GCA in the various studies were initially started on prompt recognition and early treatment. Therefore, it is
intravenous corticosteroids, followed by oral therapy. essential to know what criteria are most helpful in early
Some authors have recommended intravenous therapy diagnosis, so that the patients are treated promptly. I have
[6466] while others feel that oral therapy is as good discussed those diagnostic criteria in Chap. 11.
Visual Deterioration in GCA Patients While on High Doses of Corticosteroid 243
Visual Deterioration in GCA Patients taking high doses of systemic corticosteroids during the
While on High Doses of Corticosteroid early stages of treatment, (2) the various factors which
may influence that, and (3) whether intravenous mega-
dose corticosteroid therapy was more effective than oral
A review of the English literature reveals many reports therapy. In our study [20], after the start of the high-dose
of further visual loss after patients had started systemic corticosteroid therapy: (a) of the 53 who were first seen
corticosteroid therapy [25, 31, 37, 40, 5961, 66, 69, in my clinic without any visual loss, none suffered any
93114] visual deterioration on a long-term follow-up; and (b) of
Liu et al. [60], in a retrospective review of the the 91 patients who came with visual loss already in one
records of 41 GCA patients with visual loss, found or both eyes, 9 patients (11 eyes) developed further
worsening vision in 17%. Intravenous methylpredniso- visual deterioration in spite of high doses of systemic
lone was given to 25 patients, and the rest received oral corticosteroids (intravenous megadose corticosteroid in
prednisone alone. Danesh-Meyer etal. [69], in a multi- seven patients 8 eyes, and oral in the rest). When visual
center prospective case series of 34 consecutive biopsy- deterioration occurred, it almost invariably started
proven GCA patients, found visual deterioration in within 5days after the start of high-dose systemic corti-
27% of the eyes within the first 6 days despite high- costeroid therapy, and was usually progressive over a
dose IV corticosteroids. Loddenkemper etal. [114] ret- period of 13 days, after which the vision stabilized.
rospectively reviewed 90 consecutive patients with Among the eyes that had normal vision when high-dose
biopsy positive GCA and found progressive visual loss corticosteroid therapy was started, only one eye in our
in 23% despite corticosteroid therapy. They found that series developed new visual loss. There was no signifi-
risk factors for progressive visual loss included older cant difference between those treated with intravenous
age, elevated CRP and disc swelling. corticosteroid therapy versus oral corticosteroids. Thus,
In the reported studies, the total number of patients, in this study, not one of the patients who did not lose any
number of eyes involved, initial corticosteroid therapy further vision during the first 5days of the start of high-
dose, time between start of therapy and visual deterio- dose corticosteroid therapy, lost vision thereafter, with a
ration, and dosage of corticosteroid therapy at the time range of follow-up of 3.4weeks to 20.2years (Median
of visual deterioration are summarized in the Table13.4 2.1years). Therefore, during the first week of high-dose
[20]. Most of these patients suffered further visual loss corticosteroid therapy, one should give a guarded prog-
during the initial stages while already on high doses of nosis for any further visual loss, after which the outlook
corticosteroids, although a few experienced it only is invariably good provided the maintenance corticos-
when the corticosteroid dose was reduced later on, to teroid dose is adequate. (I have seen many other patients
what may have been an inadequate level. In the litera- belonging to this category who were not included in our
ture, almost all of the reports on large numbers of study [20] because they did not fulfill the inclusion and
patients were based on retrospective retrieval of infor- exclusion criteria [20]).
mation from charts of patients seen over several years
by different physicians; the others are from a few
selected patients only. The inherent limitations of stud-
ies based on retrospective retrieval of information from Case Report 1
charts of a large number of patients, with data recorded
and collected over several years by different persons The following unfortunate patient in my series is an
with varying professional expertise, are well known. instructive example of relentless visual loss in both eyes
To get more reliable information, we [20] conducted during the first 5days in spite of aggressive high-dose
a study from 1974 to 1999 in 144 consecutive temporal corticosteroid therapy. A 78-year-old white man noticed
artery positive GCA patients (271 eyes); of these blurred vision all over in the left eye when he got up one
patients, 91 were seen initially with visual loss, while 53 morning; it progressively deteriorated during the day
patients were initially seen without any visual loss. The (Day 1). He was seen that day at 1p.m. by a local oph-
objective of the study was to investigate: (1) the inci- thalmologist, who recorded a visual acuity of hand
dence and extent of visual deterioration in GCA patients motions in the left eye and normal in the right eye.
Table13.4 Review of Literature Dealing with Giant Cell arteritis-Related Visual Deterioration While on Steroid Therapy (Reproduced from Hayreh and Zimmerman [20])
244
Author(s) and Year Total GCA Number Initial corticosteroid Time between Corticosteroid therapy at the time
reference number patients in of eyes therapy start of therapy of visual deterioration
study involved Corticosteroid used Dose (mg) and visual Corticosteroid used Dose (mg)
deterioration
Meadows [93] 1954 12 1 Cortisone NA 11d Cortisone NA
Birkhead etal. [94] 1957 55 2 Cortisone 150300 7d, 10d Cortisone 150300
Mosher [95] 1959 11 4 NA NA NA NA NA
Whitfield etal. [96] 1963 72 1 Prednisone 3040 NA Prednisone 3040
Meadows [97] 1968 >80 5 Cortisone NA Within 11d Cortisone NA
Smith [98] 1968 3 3 Prednisone 80100 Within 12w Prednisone 80100
Cohen [40] 1973 14 5 Prednisone 100 Within first w Prednisone 100
Boghen and Glaser [99] 1975 13 1 Prednisone 80 3d Prednisone 80
Healey and Wilske [100] 1977 50 1 Prednisone 40 2m Prednisone 10
Srensen and Lorenzen [101] 1977 63 2 Prednisone 3060 NA Prednisone 2.5
Huston etal. [37] 1978 42 1 Cortisone 300 7d Cortisone 75
Hugod and Sacheibel [102] 1979 1 1 Prednisone 60 5d Prednisone 60
Calamia and Hunder [103] 1980 100 3 (2 pts) Prednisone 120 1d Prednisone 120
Prednisone 50 1d Prednisone 50100
Prednisone 50 NA Prednisone 60
Herrick [104] 1980 19 2 Prednisone 60 1d, 4d Prednisone 60
Jones and 1981 85 2 Prednisone NA 2m Prednisone 8
Hazleman [105] Prednisone NA 7m Prednisone 10
Bengtsson [106] 1982 126 2 Prednisolone 16 2d, 5d Prednisolone 16
Delecoeuillerie etal. [107] 1988 78 2 Prednisone 60 NA Prednisone 60
Slavin and Margolis [108] 1988 2 2 Prednisone 60 5d Prednisone 80
NA NA NA NA NA
Kyle and Hazleman [109] 1989 35 1 Prednisone 40 1w Prednisone 40
Faavang and Thyssen [110] 1989 1 1 Prednisone 80 63d Prednisone 10
Myles etal. [31] 1992 96 1 Prednisone 60 4w Prednisone 15
13 Management of Giant Cell Arteritis to Prevent Visual Loss
Matzkin etal. [66] 1992 3 1 IVMP 2,000 5d IVMP 2,000
Aiello etal. [59] 1993 245 7(4 pts) Prednisone 60 4d Prednisone 60
Prednisone 120 for 10d 24d Prednisone 120 alternate
days
Prednisone 60 1m Prednisone 60
Weinberg etal. [111] 1994 3 2 (1pt) IVMP 1,000 NA IVMP 1,000
1 IVMP 240500 13d Prednisone 40
Liu etal. [60] 1994 45 7 Prednisone 60 <12d Prednisone 60
IVMPOral NA 8m Prednisone NA
IVMP NA 3d IVMP NA
Prednisone 60 14m Prednisone 5
Cornblath and 1997 5 5 IVMP 5004 2d IVMP 5004
Eggenberger [25] Prednisone 60 Several d Prednisone 60
IVMP 2504 2d IVMP 2504
Hwang etal. [112] 1999 1 1 IVMP 2504 4d IVMP 2504
Visual Deterioration in GCA Patients While on High Doses of Corticosteroid
Schmidt etal. [113] 2000 5 3 IVMP 2504 5d IVMP 2504

d day(s), IVMP intravenous methylprednisolone, m month(s), NA information not available, pt patient, w week(s)
245
On the second day (Day 2) he was seen in my clinic, dose of corticosteroid therapy or discontinuing it com-
with a visual acuity of hand motions in a temporal island pletely because of the absence of any systemic symp-
in the left eye and 20/20 in the right eye. Fundus exami- toms on tapering off the therapy. In this context, the
nation in the left eye revealed chalky white optic disc following recommendation of corticosteroid therapy in
swelling (typical of arteritic-AION), and a normal GCA by rheumatologists [23] is relevant: The initial
fundus in the right eye (see Fig.12.15a in Chap. 12). dose of glucocorticosteroids is usually given for
He was admitted to the hospital, immediately started on 24weeks until all reversible signs and symptoms have
intravenous Dexamethasone sodium phosphate 150mg resolved and acute phase reactants are back to normal.
every 8h and given 3 doses; he also had temporal artery Then, the dose can be gradually reduced each week or
biopsy done which was positive for GCA. On Day 3, in every 2weeks by a maximum of 10% of the total daily
the morning he was started on 120mg Prednisone oral dose. The necessary duration of glucocorticosteroids
daily. The vision in both eyes was stable till Day 6 (while therapy is variable, but in most cases it can be discontin-
he was still in hospital), when he woke up in the morn- ued within 12years. Our study [21] of corticosteroid
ing and found that he could not read; at that time his therapy in GCA to prevent visual loss has shown the
visual acuity was counting fingers at 2 feet in the right cook-book regimen of therapy advocated by this group
eye and light perception in the left; he had suffered is not at all valid and may be dangerous because of
visual deterioration in both eyes during the night. The marked inter-individual variability in response, dosage
right optic disc had chalky white optic disc swelling (see required and duration by various GCA patients. My
Fig.12.15b in Chap. 12). Intravenous Dexamethasone experience of dealing with these patients for about
150mg was again started immediately and repeated 6 40years has indicated that GCA patients, like other suf-
hourly 4 times that day. On Day 7, in the morning his ferers from rheumatologic diseases, generally require
visual acuity was hand motions in the right eye and no adequate maintenance corticosteroid therapy for years,
light perception in the left. I felt that he had had suffi- most of the time for the rest of their lives, to keep the
cient intravenous corticosteroids, and so he was switched disease under control.
to 120mg Prednisone orally. On Day 8 in the morning
the visual acuity in the right eye was bare light percep-
tion and in the left no light perception. On Day 9 in the Why Do a Few Eyes Develop Visual
morning it was no light perception in either eye, and this Deterioration in Spite of High-Dose
did not change subsequently in spite of continued high
corticosteroid dosage orally. He was followed for over
Corticosteroid Therapy?
4years in my clinic (till his death), on a small mainte-
nance dose of Prednisone (to prevent development of This is most probably due to two factors. To under-
other serious systemic ischemic disorders caused by stand them, it is essential first to review briefly the
GCA, e.g., myocardial infarction or cerebrovascular mechanism of visual loss in GCA.
accident), without experiencing any change in vision.
Over the years I have had the experience of seeing
patients who had suffered further visual loss on either
rapid tapering of corticosteroid therapy (because of Mechanism of Visual Loss in GCA
their physicians apprehension of side-effects of corti-
costeroid therapy) or completely stopping it under the It is well-established now that the visual loss in
impression that GCA is a self-limiting disease and GCA is due to arteritic-AION in the vast majority of
there is no need to keep them on corticosteroid therapy cases [4, 13]. The mechanism is discussed at length
after about 12 years [23]. In discussion with other in Chap. 12.
ophthalmologists, I have heard of a number of similar Our study [13], as well as many others in the
cases. literature [6, 115118], has shown that many patients
Rheumatologists who recommend tapering of corti- give a history of having episodes of amaurosis fugax
costeroid therapy by systemic symptoms rather than before the development of visual loss in GCA. On doing
ESR and CRP, as discussed above, may be an uninten- fluorescein fundus angiography in GCA patients with
tional cause of visual deterioration, due to an inadequate amaurosis fugax, I have found evidence of markedly
sluggish and impaired choroidal circulation in the indicating that the posterior ciliary arteries (the main
distribution of one or the other posterior ciliary artery. source of blood supply to the ONH) were markedly nar-
In some GCA patients who had recently developed rowed by the arteritic process, but not completely
visual loss in one eye, I also found early evidence of occluded yet. All this was almost diagnostic of GCA,
sluggish and impaired choroidal circulation in the fel- which was later confirmed on temporal artery biopsy.
low eye with normal visual acuity; these angiographic The patient was immediately admitted to the hospital
findings are due to low perfusion pressure in the poste- and started on intravenous Solu-Medrol 150mg three
rior ciliary artery(s), caused by partial occlusion of times daily for 3days to prevent him from losing vision
these arteries by an arteritic process in GCA. Thus, permanently in that eye or even both eyes, and then
fluorescein fundus angiography provides very useful switched to 120 mg Prednisone daily orally. He had
information about the mechanism of visual loss in GCA another episode of visual loss that evening at 10PM.
and the factors that influence its progress (see below). He had couple of episodes of short transient visual
The following case report is an instructive example of blurring on January 28, 1999 and one minor episode on
an initial misdiagnosis of amaurosis fugax, and imme- January 30, 1999, but none after that. His episodes of
diate high-dose corticosteroid therapy preventing visual transient visual blurring occurred whenever the perfu-
loss in eyes presenting with amaurosis fugax. sion pressure in the ONH (in the posterior ciliary
arteries as well) fell below the critical levels for rea-
sons discussed below. He has since been followed in
my clinic for about 12years and has been doing fine
Case Report 2 on a maintenance dose of 1mg Prednisone, with ESR
1mm/h and CRP <0.5mg. The instructive part of this
On January 19,1999, a 74-year-old man noticed flash- case is that his initial amaurosis fugax was misdiag-
ing little butterflies in his right eye at noon, lasting nosed as due to embolization and he was sent home
510 min. On January 24, 1999, at 2 PM a curtain for later evaluation for embolization, when in fact he
moved over the temporal side and he could see only was on the verge of going blind and needed immedi-
grey for 10min. When first seen on January 25, 1999 in ate, aggressive high-dose systemic corticosteroid ther-
the general clinic of our department, his visual acuity apy to prevent visual loss. He has maintained 20/20
was 20/20 in both eyes and the ophthalmic evaluation visual acuity in both eyes at all times, with no other
was normal. Differential diagnosis by the ophthalmolo- evidence of visual loss.
gist was migraine, transient ischemic attack and poste-
rior vitreous detachment. The ophthalmologist told the
patient that she was going to arrange for him to have Factors Influencing Visual Loss
carotid Doppler and echocardiographic studies, as soon Progression
as possible, to find the source of the embolism in case
it was transient ischemic attack. On January 27, 1999,
The following two factors, individually or in combina-
the ophthalmologist met me and casually mentioned
tion, may be responsible for the progression of visual
having seen this patient. When I heard about the history
loss in spite of high-dose corticosteroid therapy.
of visual symptoms, I asked her to get the patient back
immediately to do ESR and CRP urgently, because a 1. The Latent Period between the Start
74-year-old man with those symptoms needs those tests of Corticosteroid Therapy and Control
to rule out GCA straight away, to prevent any further of the Arteritic Process in GCA
visual loss. When seen that day, he related that he had As is well-known, there is always a latent period
lost vision in that eye at 8AM for about 10min that between the start of a therapy and its effectively con-
day, and on further questioning, gave a history of hav- trolling the disease process. Histopathological studies
ing had jaw claudication all along and temporal head- of the ophthalmic and posterior ciliary arteries in
aches. His ESR was 45mm and CRP 7.4mg. Although GCA patients with visual loss have shown a
his fundus was normal on ophthalmoscopy (Fig.13.4a), granulomatous inflammatory reaction with giant cells
fluorescein fundus angiography revealed a markedly in the walls, intimal thickening and occlusion of the
delayed and impaired choroid circulation (Fig.13.4b, c), lumen by organizing thrombi [72, 73, 119, 120].
a b
Fig.13.4 (a). Normal fundus and optic disc of the right eye. (b and no filling of the choroidal and optic disc circulation at this stage.
c) Fluorescein fundus angiograms of the right eye. (a) During the (c) During the late retinal arterio-venous phase shows patchy filling
retinal arterial phase shows normal filling of the retinal arteries but of the choroid
Corticosteroid therapy, however aggressively given, p ressure (i.e., nocturnal arterial hypotension) [83,
takes several days to stop the arteritic process in the 85]. When the blood pressure in the ONH vessels
wall of the posterior ciliary arteries findings in our is already low, due to marked narrowing of the
study suggest that that may be up to 5days. One could posterior ciliary arteries (as discussed above), even
compare the situation to an emergency stop in a trav- a normal fall of blood pressure during sleep is
eling vehicle; even standing on the brakes does not enough to compromise the ONH blood supply,
stop the vehicle immediately but only after it has gone resulting in ischemia, and development of arteritic-
for a variable distance, depending upon the speed it AION. If these patients are on arterial hypotensive
was moving before. Therefore, it is apparent that if therapy, particularly taking it at bedtime or in the
the posterior ciliary arteries are already markedly nar- evening, that would aggravate the nocturnal
rowed by the arteritic process in their wall, (although hypotension [6, 13] and make them more vulnera-
not enough to cause visual symptoms or visual loss), ble to visual loss. Hence, it may be relevant to
starting high-dose aggressive corticosteroid therapy at check their hypotensive medication regimen. I
that stage may not stop the relentless progression of have also seen GCA patients developing amauro-
the inflammatory process soon enough to avoid occlu- sis fugax when they suddenly stood up, because of
sion of the artery and consequent development of development of transient orthostatic arterial
arteritic-AION. Obviously, the earlier the treatment hypotension.
is started, the better are the chances of preventing
blindness.
2. Role of fall of Perfusion Pressure in the Optic
Nerve Head (ONH) Is Intravenous Megadose Corticosteroid
Blood flow in the ONH depends upon the perfusion Therapy More Effective than Oral Therapy
pressure in its vascular bed. Perfusion pressure is equal in Preventing Further Visual Loss?
to mean blood pressure minus the intraocular pressure.
Therefore, a fall of blood pressure, a rise of intraocular In our study [20], 6 of the 48 patients (13%) who were
pressure or a combination of both can compromise the on intravenous megadose corticosteroid therapy devel-
ONH blood supply, resulting in visual loss. Marked oped visual deterioration compared to 3 of 97 of
narrowing of the posterior ciliary arteries or of the patients (3%) on oral therapy only. Of the nine patients
ophthalmic artery by arteritic process in GCA [72, 73, with progressive visual loss while on high-dose corti-
119, 120] would result in a fall of blood pressure in the costeroid therapy, the fellow eye either did not suffer
vascular bed of the posterior ciliary arteries and the any further visual loss or remained normal with intra-
ONH (as demonstrated on angiography in Fig.13.4b, c). venous megadose therapy in 4 and with oral corticos-
In such a situation, the blood supply of the ONH may be teroid therapy in 3; however, in one patient the fellow
readily compromised by the following two factors, normal eye went blind in spite of intravenous mega-
resulting in ischemia of the ONH and consequent dose therapy (as discussed in one of the case reports
development of either amaurosis fugax or arteritic- above). Therefore, it seems that there was no evidence
AION: from our study that intravenous megadose corticoster-
(a) Rise in intraocular pressure: Even a small rise oid therapy was more effective than oral therapy in
(e.g., on stooping down) may be enough to tem- preventing further visual loss; however, in our study
porarily compromise the ONH blood flow and the patients who were started on intravenous therapy
precipitate amaurosis fugax, which is often were those who had worse visual acuity to start with
reported by GCA patients before the loss of vision than did those on oral therapy, and that is a confounding
[6, 13]. factor. Other authors have also reported progressive
(b) Fall of blood pressure: On enquiring the exact time visual loss in spite of intravenous megadose therapy
when these patients first discovered permanent [25, 60, 66, 69, 108, 111, 113].
visual loss, I have often found that to be upon In spite of the lack of any definite evidence, in our
awaking from sleep (for example, in the case report study or in the reported literature, that intravenous
with complete visual loss described above); we megadose corticosteroid therapy is more effective than
know that during sleep there is a fall of blood oral therapy in preventing further visual loss in the
already involved eye or in the fellow uninvolved eye, reported cases were most probably purely coincidental
one cannot rule out completely the possibility that in a occurrence, unrelated to GCA or corticosteroid therapy.
rare case it may help to prevent visual loss. Thus, to Moreover, tens of thousands of GCA patients are treated
give the benefit of the doubt to GCA patients, I still with corticosteroid therapy. If there were a true associa-
recommend using it in those few patients who have: (i) tion, it would be seen far more commonly.
a history of amaurosis fugax but no visual loss at initial Osteoporosis is a well-known major risk factor of
visit (as in the case reported above), (ii) complete or long term corticosteroid therapy. Lespessailles et al.
marked loss of vision in one eye, and (iii) early signs [125] based on a study of 121 women with a mean age
of involvement of both eyes, just in case it may of 60.4 +/14.3 years on long-term glucocorticoid
prevent further, irreversible visual loss. therapy concluded that there is compelling evidence
that bone mineral density is a major determinant of the
fracture risk in patients with glucocorticoid-induced
osteoporosis. Cacoub etal. [126] compared bone mass
Side-Effects of Corticosteroid Therapy
loss due to deflazacort versus prednisone in long-term
in GCA and Their Prevention treatment of 74 patients with GCA in a randomized
double blind comparative trial over a minimum of
As discussed above, the invariable reason for giving 12months period. Their study showed that with older
GCA patients too little, for too short a time corticos- patients who are at risk of osteoporosis taking long-
teroid therapy is the common fear among physicians term glucocorticoids, deflazacort did not result in less
about the risks of systemic side-effects of prolonged bone loss than prednisone. Chevalet etal. [127] com-
corticosteroid therapy, particularly in the elderly. Kyle pared the clinical and therapeutic courses (corticoster-
and Hazleman [121] found that side effects were sig- oid response, corticosteroid amount, complications) in
nificantly related to an initial prednisolone dose of people with GCA under and over 75years, during the
more than 30mg and to the cumulative prednisolone first year of treatment, in a series of 164 patients. The
dose. Patients taking a mean daily dose of 5mg pred- data were retrospectively analyzed (mean age:
nisolone or less were significantly less likely to develop 73.3years) for the two subgroups. They concluded that
side effects. Nesher etal. [122] in a study of 43 con- even if corticosteroid response was identical in the
secutive prednisone treated GCA patients, on a mean therapeutic course of GCA, rheumatic side effects
follow-up of 3years found that 58% developed major appeared more frequent in the elderly group (over
corticosteroid related complications. The most com- 75years).
mon were fractures and severe infections. But these Since osteoporosis is the major risk factor in long
complications were age related, occurring twice as term corticosteroid therapy, several suggestions have
often in patients older than 75 years as in younger been made to prevent or treat that. For example, Saag
patients. Corticosteroid related side effects were also et al. [128]. carried out two 48-week, randomized,
dose related, occurring more commonly in patients placebo-controlled studies of two doses of Alendronate
starting with doses of >40mg/day Prednisone and in (Fosamax) in 477 men and women, 1783 years of
patients taking high maintenance dosage. age, who were receiving glucocorticoid therapy. They
There has been a rare anecdotal case report of devel- concluded that Alendronate increases bone density in
opment of iatrogenic central serous chorioretinopathy patients receiving glucocorticoid therapy. The
in GCA patients treated with corticosteroid therapy American College of Rheumatology [129] recom-
[123, 124], but in more than 40years of treating about mended that glucocorticoid-induced bone loss should
300 GCA patients with corticosteroid therapy I have be prevented. Supplementation with calcium and vita-
never seen this complication. In addition to that, during min D at a dosage of 800IU/day, or an activated form
this period, I have also treated several thousand patients of vitamin D (e.g., alfacalcidiol at 1 mg/day or calcit-
with high-dose corticosteroid therapy for various ocu- riol at 0.5 mg/day), should be offered to all patients
lar rheumatologic and other diseases and have yet to receiving glucocorticoids, to restore normal calcium
see a case developing central serous chorioretinopathy. balance.
Central serous chorioretinopathy is not a rare disease In view of the fact that osteoporosis is the major
and develops without any such association. The risk factor in long term corticosteroid therapy, I advise
Corticosteroid Sparing Therapy in GCA 251
all my patients to take calcium and vitamin D, and also clearly shown that alternate therapy does not control
to consult their physician about taking Alendronate GCA satisfactorily. (iii) The relapse rate was >60%
sodium (Fosamax) or Ibandronate (Boniva) to prevent and they excluded isolated elevation of ESR as a crite-
osteoporosis. Both Fosamax and Boniva are specific rion for relapse, which means the relapse rate was most
inhibitors of osteoclast-mediated bone resorption. probably higher than was claimed. All relapses were in
fact accompanied by elevations in ESR, which means
the disease was not under control. Most relapses
Corticosteroid Sparing Therapy in GCA occurred after an alternate-day prednisone schedule was
initiated (51%) or after prednisone was entirely discon-
tinued (34%). (iv) There was a high incidence of vision
High-dose corticosteroids are the mainstay of treatment
loss, which occurred after treatment in both groups (5 of
for systemic vasculitides, including GCA. However,
47 in the corticosteroid plus placebo group, 4 of 51 in
since corticosteroid therapy is associated with systemic
the corticosteroid plus methotrexate group) and was
side-effects and complications, there has been a constant
isolated to the first 6months of follow-up (16months).
search to find alternative therapies for the treatment of
Among the eight patients with delayed vision loss
various systemic vasculitides, including GCA. The
(>1month from initiation of therapy), prednisone had
most common agent which has been studied is metho-
been tapered and entirely withdrawn in 4. As men-
trexate. Others include Imuran (Azathioprine), Cyclos
tioned before, in our study [20] not a single patient lost
porine A, Infliximab (Remicade), Enbrel (Etanercept),
any more vision after the start of high dose corticoster-
Adalimumab (Humira), Cyclophosphamide, Dapsone,
oid therapy and tapering guided by ESR and CRP. This
and Chlorambucil.
high incidence of relapses, associated with elevated
ESR and high incidence of visual loss in both wings of
their study clearly shows that the disease was not under
Methotrexate satisfactory control after all the primary purpose of a
treatment in GCA is to prevent visual loss and that was
The use of methotrexate as a corticosteroid sparing not satisfactorily achieved in this study. In the 1980s in
agent for treatment of GCA is often advocated. van der collaboration with our rheumatology department at the
Veen et al. [130] found no difference between their University of Iowa Hospitals & Clinics, I did a pre-
methotrexate group and the placebo group concerning liminary study of methotrexate with prednisone in
time to achieve remission, duration of remission, num- GCA patients without any visual loss but the results
ber of relapses, or cumulative Prednisone doses, and were inconclusive.
no corticosteroid sparing effect of methotrexate. Spiera On the other hand, Jover etal. [133] in a random-
et al. [131] in a randomized, controlled, double- ized, double-blind, placebo-controlled trial, investi-
masked trial comparing methotrexate (12 patients) gated the safety and efficacy of combined therapy with
versus placebo (9 patients) in addition to corticosteroid corticosteroids and methotrexate in GCA in 42 patients
therapy in patients with newly diagnosed GCA, also with new-onset biopsy positive GCA. They concluded
found no corticosteroid-sparing benefit. Similarly, that treatment with methotrexate plus corticosteroid is
Hoffman et al. [132] in a multicenter, randomized, a safe alternative to corticosteroid therapy alone in
double-masked, placebo-controlled trial of adjuvant patients with GCA and is more effective in controlling
methotrexate treatment for GCA, found that their disease. Mahr et al. [134], based on an individual
results did not support the adjunctive use of methotrex- patient data meta-analysis of three randomized pla-
ate to control disease activity or to decrease the cumu- cebo-controlled trials in patients with newly diagnosed
lative dose and toxicity of corticosteroids in patients GCA, concluded that in GCA, adjunctive treatment
with GCA. There are, however, several problems with with methotrexate lowers the risk of relapse and
this study, including the following. (i) In this study, reduces exposure to corticosteroids. While meta-
17% did not have positive temporal artery biopsy. analyses often provide useful information, that is not
(ii) Corticosteroids were initially administered daily, always the case, because their conclusions depend
and after 4weeks were tapered according to an alter- upon the credibility of the studies included in the
nate-day schedule. The study by Hunder etal. [34] has meta-analysis.
Thus, the evidence is equivocal but tends to indicate patients with newly diagnosed GCA that was in gluco-
that there is no real benefit in using methotrexate as a corticosteroid-induced remission. In that study, inflix-
corticosteroid sparing agent or to control GCA. imab therapy did not increase the proportion of patients
without relapse at week 22 compared with placebo
(P=0.65), nor did it increase the proportion of patients
whose glucocorticosteroid dosages were tapered to
Azathioprine (Imuran) 10 mg/d without relapse (P=0.31). The incidence of
infection was 71% with infliximab and 56% with
This is an immunosuppressive antimetabolite. De Silva placebo. Their conclusion was that using infliximab as
and Hazleman [135] used azathioprine to reduce the maintenance therapy in patients in glucocorticoid-
maintenance prednisolone requirement of 31 patients induced remission of newly diagnosed GCA is of no
with PMR or GCA, and noted a statistically significant benefit and may be harmful.
difference (p<0.05) in mean prednisolone dose
between the two groups at the end of 52weeks, there
being a fall in corticosteroid requirement in the aza-
thioprine treated group. Enbrel (Etanercept)
This is a tumor necrosis factor antagonist. Martinez-

Taboada etal. [138] conducted a randomised, double-
Cyclosporine A blind, placebo controlled trial with etanercept in 17
patients with biopsy-proven GCA with side effects sec-
This is a potent immunosuppressant, primarily through ondary to corticosteroids. They were given Etanercept
its direct suppression of T-cells. Schaufelberger etal. (n=8) or placebo (n=9) over 1year together with cor-
[136] conducted a 12-month, multicentre, prospective, ticosteroids. They concluded that the limited number
controlled, randomized, open study that was designed of patients included in this study did not allow them to
to evaluate the corticosteroid-sparing potency of draw definitive conclusions.
Cyclosporin A. The patients were randomly assigned
to one of two treatment groups (Group A: corticoster-
oids alone, Group B: corticosteroids+Cyclosporin A).
There was a high rate of premature termination and Adalimumab (Humira)
treatment modification due to adverse events that
appeared in the Cyclosporin A treatment group. Only This is a fully human recombinant IgG1, anti- tumor
3 of 29 patients were treated without adverse events. necrosis factor-alpha monoclonal antibody. There is
They concluded that their study, as well as a previous one anecdotal case report of a patient who was treated
study, clearly showed that Cyclosporin A has no corti- with adalimumab and success claimed [139].
costeroid sparing potency in GCA patients, primarily I could not find any report specifically dealing with
because of extremely poor tolerability. Therefore, they the use of cyclophosphamide, dapsone, and chloram-
did not recommend Cyclosporin A either in early or in bucil in GCA alone.
longstanding GCA. Nesher and Sonnenblick [140], in a review of
corticosteroid-sparing treatments, concluded that no
definite recommendation can be drawn regarding the
use of corticosteroid-sparing agents in GCA. After a
Infliximab (Remicade) review of the published reports dealing with the
addition of corticosteroid-sparing agents, Wilke and
This is a chimeric IgG1k monoclonal antibody and it Hoffman [30] stated that: the small number of patients,
neutralizes the biological activity of tumor necrosis lack of controls, and brevity of these reports do
factor-alpha. Since this is present in the arteries in not allow even preliminary conclusions about the
GCA, Hoffman et al. [137] conducted a multicenter advantages of addition of a second agent or about
clinical trial to evaluate the efficacy of infliximab in 44 whatmight be the alternative treatment of choice for
Corticosteroid Sparing Therapy in GCA 253
corticosteroid-resistant GCA. They concluded that Liozon et al.s [150] analysis incorporated too many
studies in which a second agent has been added to independent variables in the logistic regression model
corticosteroid therapy have demonstrated that typical for the small sample size, and that resulted in a model
GCA patients usually do very well with corticosteroids that is over-fitted. Thus, the conclusions derived from
alone . Cantini et al. [141] found that the data sup- their multiple logistic regression model may not be
porting the usefulness of corticosteroid-sparing effect valid. Those who believe in a possible relationship
of methotrexate and anti-tumor necrosis factor-alpha between thrombocytosis and ischemic lesions in GCA
drugs are limited and non-conclusive. have proposed various mechanisms to account for it.
Thus, all the available evidence so far indicates that Bengtsson and Malmvall [143] speculated that the
the best treatment option for GCA is corticosteroid arteritic process may adversely affect the prostacyclin
therapy. I have no personal experience of using any of synthesis in the arterial wall, thereby disturbing the
the above corticosteroid-sparing therapies, except in a balance between prostacyclin and thromboxane A2
small preliminary trial of methotrexate mentioned release from the activated platelets, a substance
above. known to be a potent vasoconstrictor as well as a pro-
aggregatory agent [152]. Gibb etal. [144] in one case
of fatal basilar artery thrombosis and GCA, found a
high platelet count (56610 [3]m/l), and concluded
Role of Aspirin in Management of GCA that ischemia could result from embolization of plate-
let clumps from inflamed intimal surfaces or by release
Several studies have shown the presence of reactive of thromboxane A2 by activated platelets producing
thrombocytosis in GCA, as discussed by us [142]. That platelet aggregation and vasoconstriction.
has led some to assume that aspirin or other anti-plate- In contrast to that, other authors have found no such
let aggregating agents might have a role in the manage- association between thrombocytosis and an increased
ment of GCA, the prevention of visual loss and other risk of ischemic complications in GCA [148, 149]. For
ischemic lesions. Therefore, it is essential to discuss example, Cid etal. [147], in a study of 200 consecutive
the association between thrombocytosis in GCA and biopsy-proved GCA patients, on comparing patients
ischemic complications. The role of an elevated plate- with ischemic lesions (32 patients) versus those with-
let count and the possible association of thrombocyto- out ischemic lesions (168 patients), found no signifi-
sis with ischemic complications has been debated in cant difference in platelet count between the two. Wu
the literature [103, 143150]. For example, De Keyser [153] found none of the 11 patients with reactive
etal. [145] reported a correlation between thrombocy- thrombocytosis had thrombotic or hemorrhagic com-
tosis and ischemic lesions because their group with plications. In our study [142] of 121 temporal artery
ischemic complications (18 patients with visual loss, biopsy confirmed GCA patients, we also found no sig-
stroke and transient ischemic attack) had a signifi- nificant difference in the mean platelet counts between
cantly (P<0.01) higher prevalence of thrombocytosis GCA patients with vision loss (406103ml) and those
and a higher median platelet count than the group without (425103ml). It is relevant to point out that
without those ischemic complications (38 patients). permanent visual loss, as a complication of biopsy-
Similarly Liozon etal. [150] concluded that there was proved GCA, was much higher in our study than has
a correlation between increased risk of vision loss and been reported in other GCA studies dealing with
higher platelet counts. However, a careful review of thrombocytosis [103, 145, 147150, 154158].
the data and statistical methods in the latter study Therefore, our study provides much more reliable
reveals that in the analysis used to identify the inde- information on this issue than previous ones.
pendent risk factors for permanent vision loss, there There are several studies dealing with treating GCA
were 17 variables in the logistic regression model of patients with aspirin, with conflicting results. For
23 patients with permanent vision loss and 151 with no example, Nasher et al. [159] reviewed charts of 175
vision loss. In such an analysis, the general rule is to consecutive GCA patients. At the time of the diagnosis
include no more than 1 independent variable per of GCA, 21% were already taking low-dose aspirin
10 subjects in the smaller of the 2 outcome groups (100mg/day). They evaluated cranial ischemic com-
to derive a model that is reliable [151]. Therefore, plications of GCA in those with and without aspirin
therapy. Acute visual loss was present in 17%, stroke in is no convincing evidence that ophthalmic ischemic
6% and both in 1%. They found that 8% of the aspirin- manifestations occur as a direct consequence of reac-
treated patients developed cranial ischemic complica- tive thrombocytosis in GCA, particularly with the
tions, compared with 29% of the non-aspirin-treated rather moderate increases in platelets seen in GCA
patients (P=0.01). Despite the use of corticosteroid patients.
therapy, cranial ischemic complications developed in It seems the whole controversy on the association of
9% of the 166 patients followed up for 3 months or thrombocytosis in GCA and ischemic lesions has
longer, and cranial ischemic complications devel- emerged from confusing essential thrombocytosis and
oped in only 3% of the aspirin-treated patients, com- polycythemia vera with reactive thrombocytosis (seen
pared with 13% of the patients treated with prednisone in GCA). It is well known that patients with essential
only (P=0.02). They concluded that their data suggest thrombocytosis and polycythemia vera have a high risk
that low-dose aspirin decreases the rate of visual loss of thrombotic involvement of major vessels and the
and stroke in patients with GCA. Lee et al. [160] microcirculation [162]. However, reactive thrombocy-
investigated in 104 biopsy-proven GCA patients in a tosis associated with GCA is not the same thing as
retrospective chart review whether antiplatelet or anti- essential thrombocytosis and polycythemia vera the
coagulant therapy reduces ischemic complications in latter two are chronic, progressive, myeloproliferative
patients with GCA followed for a mean period of disorders of insidious onset, with a much higher platelet
4 years. Long-term antiplatelet or anticoagulant ther- count. As discussed above, there is no convincing evi-
apy was given to 60% and none to 40%. They con- dence that ischemic manifestations occur as a direct
cluded that antiplatelet or anticoagulant therapy may consequence of reactive thrombocytosis in GCA, par-
reduce the risk of ischemic events in patients with ticularly with the rather moderate increases in platelets
GCA. Cantini etal. [141], based on a review, concluded [142]. So there is little justification for giving aspirin or
that the data supporting the usefulness of antiplatelet other platelet anti-aggregating agents to prevent visual
agents and anticoagulants combined with corticoster- loss in GCA. Gonzales-Alegre etal. [149] also found no
oids to prevent ischemic complications in GCA are evidence to support the idea of starting platelet aggrega-
limited and non-conclusive. tion inhibitors to prevent ischemic complications of
A study by Kaiser etal. [146] showed that it is the GCA. There is also no specific evidence that anticoagu-
degree of intimal proliferation that plays a role in the lants help to prevent blindness in GCA [57, 103].
development of ischemic lesions, and not an embolic
phenomenon. The lumen of arteries involved by GCA
is usually reduced by severe thickening of the intima,
Morbidity and Mortality of GCA Patients
and thrombus is often formed at the site of active
inflammation. Our fluorescein fundus angiographic on Corticosteroid Therapy
studies of GCA patients with arteritic-AION (due to
occlusion of the posterior ciliary artery), central retinal All GCA patients are treated with systemic corticoster-
artery occlusion and cilioretinal artery occlusion indi- oid therapy to prevent visual loss for variable length of
cate that visual loss is primarily due to occlusion of the time. There is a common impression among physicians
involved artery by arteritis and associated thrombosis that GCA patients on prolonged corticosteroid therapy
and not due to embolism [4, 9, 13]. However, the pos- are at greater risk of morbidity and mortality than the
sibility of detached microscopic clumps of platelets or general population. Therefore, one often hears that
other material from the surface of a freshly forming GCA patients should be on systemic corticosteroid
thrombus, producing retinal cotton-wool spots [13], therapy for a minimum period. A review of the litera-
before the thrombus completely occludes the common ture reveals conflicting information about that topic.
trunk of the posterior ciliary artery and central retinal For example, Nesher etal. [122] in a study of 43 con-
artery [161], cannot be ruled out. On the other hand, it secutive prednisone treated GCA patients, on a mean
is important to bear in mind that such transient micro- follow-up of 3 years concluded that corticosteroid
scopic emboli are almost invariably too small to cause treatment for GCA may cause major morbidity and
occlusion of the main ocular arteries and produce the increased mortality, and that the therapeutic regimen in
devastating permanent visual loss of GCA. Thus, there GCA should be individualized, and the patients age,
Morbidity and Mortality of GCA Patients on Corticosteroid Therapy 255
severity of GCA, and coexistent medical conditions activity is not sufficiently controlled by the treatment.
should be considered. Crow etal. [163], in 44 cases, As discussed below, it is well-established that GCA
concluded that GCA patients were more likely than can result in myocardial infarction or strokes due to
age- and gender-matched controls to die within the involvement of the corresponding arteries. Gonzalez-
first 5years following diagnosis. Graham [164], in 90 Gay et al. [169] concluded that although ischemic
biopsy proven GCA patients on a follow-up of 6months complications, in particular irreversible visual loss or
to 12 years, reported that those who required a high aortic aneurysmal disease, may occur in patients with
maintenance level of corticosteroids (greater than GCA, most population-based studies have shown no
10 mg daily) experienced a significantly increased excess mortality in GCA and PMR. That was also
mortality rate. Uddhammar et al. [165] in 136 GCA shown by our study [21] of 145 temporal artery biopsy
patients, found that death due to cardiovascular disease proved GCA patients. Among patients with visual loss,
was increased, and increased mortality was related to 17 were already diabetic (5 of them on insulin) when
either the corticosteroid therapy itself or insufficient they developed GCA and one patient developed it dur-
control of inflammation. ing follow-up, while among those without visual loss 3
In sharp contrast to those reports, Anderson and were already diabetic (1 on insulin) and one developed
Malmwall [43], on a follow-up of 916years (median diabetes during follow-up. Giving long term corticos-
11.3 years) of 90 GCA patients, neither found any teroid therapy to these diabetics did not pose a problem
increase in morbidity in their patients compared to the while their physicians regulated their diabetes as
general population, nor did they see any significant before.
complications which they could attribute to the corti- To place this controversy in proper perspective, one
costeroid treatment. Gouet et al. [32] evaluated 87 has to go to the basic aspects of GCA. In this disease
temporal artery biopsy proven GCA patients in 21 of systemic vasculitis causes vascular occlusion which is
them corticosteroid therapy could be discontinued the primary cause of major complications, e.g., visual
after 36months on average but the other patients could loss, myocardial infarction, cerebrovascular disorders
not be weaned, even though 4 of them had been on and other serious arterial occlusive diseases. Systemic
steroids for more than 10years. They found that in the corticosteroid therapy is given to control vasculitis and
long run the mortality rate of GCA patients was exactly thereby prevent arterial occlusion. If vasculitis is not
the same as in the general population. Gran etal. [166] controlled adequately because of insufficient corticos-
compared deaths among the 64 GCA patients and their teroid therapy or stopping it altogether prematurely for
controls, and found no difference in mortality (RR=1.2, fear of corticosteroid therapy side-effects (a common
95% CI 0.552.74). The mean initial dose of predniso- phenomenon in my experience), that can result in death
lone, the mean maintenance dose of prednisolone, and due to arterial occlusion causing myocardial infarc-
the mean initial ESR and CRP did not differ between tion, cerebrovascular disorders and other serious arte-
survivors and those dying during the observation rial occlusive diseases. I have emphasized that patients
period. Nuenninghoff et al. [167] in a study of 168 do suffer visual loss when they are not treated with
residents of Olmsted County, Minnesota, USA, found adequate amounts of corticosteroid therapy either ini-
that, overall, mortality in the whole group of patients tially or later on, or if corticosteroid therapy is stopped
with GCA with large-artery complication was similar altogether. Under such circumstances, it is the inade-
to that in patients with GCA without large-artery com- quate control of vasculitis and not the corticosteroid
plication. Pipitone et al. [168] found that mortality therapy per se which is responsible for higher morbid-
rates of GCA patients are comparable to those of the ity and mortality in GCA. Pipitone etal. [168] rightly
general population, but there is evidence for an pointed out that there is evidence for an increased fre-
increased frequency of potentially life-threatening quency of potentially life-threatening ischemic events,
ischemic events, such as myocardial infarction and especially early on in the disease course, and that risk
cerebrovascular accidents, especially early on in the decreases with time, presumably as a consequence of
disease course. The risk conferred by the disease corticosteroid therapy controlling the vasculitis,
appears to decrease with time, presumably as a conse- whereas it can remain significantly elevated in patients
quence of glucocorticoid treatment, whereas it can whose disease activity is not sufficiently controlled by
remain significantly elevated in patients whose disease the treatment. My studies in about 300 GCA patients
fully support this view. A good control of vasculitis by older age, elevated CRP and disc swelling. Gonzales-
corticosteroid therapy is the most important consider- Gay et al. [61], in a retrospective study, investigated
ation not only to prevent visual loss but also to reduce this in 239 biopsy positive GCA patients. Visual
morbidity and mortality from vasculitis involving other involvement was observed in 69 patients. The predic-
large vessels in GCA. The only reliable guides to the tors of permanent visual loss were amaurosis fugax,
activity of the GCA are the levels of ESR and CRP jaw claudication, normal levels of liver enzymes, and
(particularly the CRP level) [21], and not the univer- absence of constitutional syndrome. After adjustment
sally advocated systemic symptoms; that may be the for the treatment regimen (intravenous pulse methyl-
culprit in inadequate control of vasculitis, causing life prednisolone versus oral prednisone), early treatment
threatening complications, which are erroneously (within the first day of visual loss) was the only predic-
attributed to corticosteroid therapy. tor of improvement. In my experience, the most
I must comment on a very important topic, the side- important factors for permanent visual loss in GCA
effects of corticosteroid therapy. Many physicians are are: (i) late diagnosis of GCA, (ii) consequently late
uncomfortable treating patients, particularly the elderly, start of high-dose corticosteroid therapy, (iii) inade-
with high-dose steroid therapy and on a long-term ther- quate corticosteroid therapy, (iv) therapy guided only
apy. Over 45 years, I have treated not only 300 GCA by systemic symptoms and not ESR and CRP levels,
patients but also several thousand patients for a variety and (v) corticosteroid therapy stopped early alto-
of serious acute ophthalmic diseases, including scleritis, gether. I have seen that time and again.
orbital myositis, uveitis, retinal vasculitis and other acute
vision threatening eye diseases. Based on that clinical
practice, I find that corticosteroid therapy can be used
safely, provided patients are followed closely and metic- Patient Education About Systemic
ulously. I find that there seems to be a good deal of mis- Corticosteroid Therapy in GCA
information, almost amounting to a phobia, about this
therapy in the medical community. No doubt corticoster-
This is most critical in the management of GCA
oid therapy is associated with systemic side-effects,
patients. As a rule, before starting a patient on high-
which differ markedly from patient to patient; however,
dose corticosteroid therapy, I make sure to have full
in my dealings with these patients, I have found that the
discussion about the following aspects of the therapy
side-effects can be adequately managed by close fol-
with them, to educate the patient about the disease and
low-up and proper medical care. There is also a preva-
treatment.
lent impression among physicians that no dose is safe.
In my prolonged careful follow-up of these patients, I 1. The first, most important point is that the whole
have found that a small maintenance dose of about 5mg objective of treatment is to prevent any further visual
Prednisone or so, taken daily to keep the disease under loss [20, 21] and that there is practically no chance
control, for as long as 10 or 15years, never produced any of recovering the existing visual loss [19, 21]. This
side-effect attributed to corticosteroid therapy. That is a is critical for the success of corticosteroid therapy as
perfectly safe dose for prolonged maintenance treatment. well as the patients compliance with the long-term
My studies, following these patients closely for the past corticosteroid therapy. There is almost invariably an
45years, do not support the prevalent impressions about impression among patients that intensive corticos-
the side-effects of very low-dose corticosteroid therapy. teroid therapy (particularly the intravenous therapy)
is going to result in visual improvement. Since that
is not going to happen in the vast majority, the
patient who believes that will be disappointed and
Predictors for Permanent Visual
may stop corticosteroid therapy, which is going to
Loss in GCA put him/her at risk of further visual loss which is
preventable in GCA with the continued adequate
Loddenkemper et al. [114] retrospectively reviewed corticosteroid therapy. They may then go on having
90 consecutive patients with biopsy positive GCA and further consultations with other ophthalmologists to
found risk factors for progressive visual loss included try to get visual improvement and waste a lot of
Unique Experience of Giant Cell Arteritis 257
money without achieving anything. Thus, it is essen- as systemic complications. Therefore, a careful
tial to make it absolutely clear at the beginning that regular follow-up with the physician is essential.
the whole purpose of treatment is to prevent any 6. I tell them that I am always available in case there is
further visual loss and not to regain any lost any question or problem.
vision.
Over the years I have had tremendous co-operation
2. I also stress that during the first 56days of starting
and compliance from my GCA patients, some of whom
the high-dose corticosteroid therapy, there is a risk
I have followed for 20years or even longer.
of further visual loss in spite of the treatment. If
they do not know that and it happens, then that
could result in them thinking that they were not
treated properly (which could result in medico-legal Misconceptions About Preventing
problems). I tell them that if they do not lose vision
Visual Loss in Arteritic AION and GCA
during that the first week of the therapy and they
continue to take corticosteroid therapy in adequate
dose during follow-up, then I guarantee that they 1. That to diagnose GCA, the patient must have sys-
will not lose any more vision. My guarantee has temic symptoms and signs of GCA. My study
never failed with any of the several hundred GCA showed that 21% of GCA patients have no systemic
patients under my care over 40 years, as long as symptoms and signs whatsoever and the only pre-
they follow the regimen of corticosteroid therapy senting sign is visual loss, i.e., occult GCA [14]
discussed above. 2. That to diagnose GCA, the patient must have
3. I stress that most probably this may be a life-long elevated ESR. As discussed above (also in Chap. 11),
treatment, to prevent any visual loss, but at a much it has been shown that normal ESR does not rule out
lower and safer dose finally. This is also important GCA.
because most patients are under the impression that 3. That corticosteroid therapy can be tapered accord-
they will need treatment only during the acute ing to a set regimen [23]. As discussed above, it has
phase, as with most illnesses been shown that there is marked inter-individual
4. I also explain fully the various side-effects of short variation in the tapering regimen of steroid therapy
and long term corticosteroid therapy, and how to and NO one size fits all.
cope with those. This is also crucial, because if 4. That steroid therapy can be regulated by using clini-
side-effects do develop and the patient is not pre- cal symptoms and signs of GCA. That is not valid
pared for them, there is a high likelihood that the to prevent visual loss. The only reliable method is
patient will stop therapy, may not return for follow- to use ESR and CRP as the guideline.
up, and may suffer further visual loss. A patient 5. That steroid therapy can be stopped after 12years
who is prepared for side-effects and knows that the because the disease burns itself out. That is not
alternative to treatment is the risk of blindness is a true at all. A vast majority of patients require life-
co-operative, successful patient. In GCA, as in long steroid therapy to prevent visual loss.
most other rheumatological diseases, the corticos-
teroid therapy is purely suppressive and not cura-
tive; the vast majority of patients require a highly
variable, carefully adjusted maintenance dose for Unique Experience of Giant Cell Arteritis
perhaps the rest of their life. I make them fully
aware of this fact. I have never had a patient so far I have investigated and published on various aspects of
who opted to stop corticosteroid therapy because giant cell arteritis during the past 40 years. In April
of side-effects. 2011, ironically, I myself was diagnosed to have occult
5. I point out that they must not suddenly stop or cut giant cell arteritis. The disease was originally sus-
down corticosteroid therapy on their own because pected when testing revealed that I had an unexplained
that can result in serious complications. With the elevated erythrocyte sedimentation rate and C-reactive
development of side-effects, they may do that, with protein which were very high for me. This was con-
hazardous consequences both for visual loss as well firmed on temporal artery biopsy, which showed
c lassical pathological findings, i.e. presence of multi- Figure13.6 shows my erythrocyte sedimentation rate
nucleated giant cell (Fig. 13.5a), absence of internal and C-reactive protein response to 80mg Prednisone
elastic lamina in the area of scar (Fig.13.5b), and pres- for 3weeks this response is an exact replica of what
ence of epithelioid histiocytes in intima (Fig. 13.5c). I found in my studies (see Fig.13.2).
Fig.13.5 (a) It shows a high

magnification of a multi-
nucleated giant cell inside the
internal elastic lamina.
(b) Elastic stain shows
absence of internal elastic
lamina (dark black wavy line)
in area of scar. (c) This shows
high magnification of
scattered epithelioid
histiocytes in intima
Conclusions 259
c
8 120
CRP (mg) ESR (mm/hr)
7 100
6
Fig.13.6 Graphs of 80
C-reactive protein (CRP) 5
levels (Left figure) and of 4 60
erythrocyte sedimentation
rate (ESR) levels (Right 3
40
figure) showing their initial
2
responses to Prednisone 80
20
mg daily for 3 weeks and 1
then 70 mg daily. Y axis in
the left figure shows CRP in 0 0
mg and in the right ESR in 0 5 10 15 20 25 0 5 10 15 20 25
mm/hr Westergren Day Day
Conclusions My 27-year clinical study [21] dealing with corti-

costeroid therapy in GCA has shown that if there is a
GCA is the most important medical emergency in reasonable index of suspicion of GCA, high doses of
ophthalmology, because of its high risk of visual loss, systemic corticosteroid therapy must be started
which is preventable if these patients are diagnosed IMMEDIATELY, as an EMERGENCY MEASURE.
early and treated immediately and aggressively. This It is not advisable to wait for the result of the tempo-
obviously raises two critical issues about manage- ral artery biopsy, because by the time it is available,
ment of GCA: (1) how can a definite diagnosis of the patient may have lost further vision irreversibly in
GCA be quickly established? and (2) what is the one or both eyes. The temporal artery biopsy should be
proper treatment to prevent blindness? I have dis- done as soon as possible but not necessarily the same
cussed both these issues in detail above, and the con- day; starting the corticosteroid therapy does not inter-
troversies surrounding them, in the light of findings of fere with the biopsy results. I have briefly described
my studies. the regimen of corticosteroid therapy in GCA patients
which, in my study, I have found to be safe and effec- 14. Hayreh SS, Podhajsky PA, Zimmerman B. Occult giant cell
tive in preventing visual loss. ESR and CRP are the arteritis: ocular manifestations. Am J Ophthalmol.
1998;125(4):5216. 893.
two most reliable and sensitive parameters to monitor 15. Hayreh SS. Masticatory muscle pain: an important indicator
GCA activity and to regulate corticosteroid therapy, of giant cell arteritis. Spec Care Dentist. 1998;18(2):605.
and not the clinical picture or systemic symptoms of 16. Hayreh SS. Ocular manifestations of giant cell arteritis.
GCA. Patients with GCA show marked inter-individ- Correspondence. Am J Ophthalmol. 1998;126(5):7434.
17. Hayreh SS. Steroid therapy for visual loss in patients with
ual variation in the dose and duration of corticoster- giant-cell arteritis [erratum 2001;356(9227):434]. Lancet.
oid therapy required, so that therapy must always be 2000;355(9215):15723.
individualized and no generalization is possible. My 18. Hayreh SS, Jonas JB. Optic disk and retinal nerve fiber layer
data show that most GCA patients require corticoster- damage after transient central retinal artery occlusion: an
experimental study in rhesus monkeys. Am J Ophthalmol.
oid therapy almost indefinitely. The overall manage- 2000;129(6):78695.
ment of GCA is taxing and laborious, for both the 19. Hayreh SS, Zimmerman B, Kardon RH. Visual improve-
patient and the physician, and it requires the trust and ment with corticosteroid therapy in giant cell arteritis. Report
cooperation of all, including the patients local physi- of a large study and review of literature. Acta Ophthalmol
Scand. 2002;80(4):35567.
cian, because of the systemic side-effects of the pro- 20. Hayreh SS, Zimmerman B. Visual deterioration in giant cell
longed corticosteroid therapy. arteritis patients while on high doses of corticosteroid ther-
apy. Ophthalmology. 2003;110(6):120415.
21. Hayreh SS, Zimmerman B. Management of giant cell arteri-
tis. Our 27-year clinical study: new light on old controver-
References sies. Ophthalmologica. 2003;217(4):23959.
22. Salvarani C, Cantini F, Boiardi L, Hunder GG. Polymyalgia
rheumatica and giant-cell arteritis. N Engl J Med.
1. Font C, Cid MC, Coll-Vinent B, Lopez-Soto A, Grau JM. 2002;347(4):26171.
Clinical features in patients with permanent visual loss due 23. Salvarani C, Cantini F, Hunder GG. Polymyalgia rheumatica
to biopsy-proven giant cell arteritis. Br J Rheumatol. and giant-cell arteritis. Lancet. 2008;372(9634):23445.
1997;36(2):2514. 24. Weyand CM, Goronzy JJ. Pathogenic principles in giant cell
2. Hazelman BL, Bengtsson B-A. Giant cell arteritis and polyarteritis. Int J Cardiol. 2000;75 Suppl 1:S915. S15S19.
myalgia rheumatica. Baillires Clin Rheumatol. 1991;5:371 25. Cornblath WT, Eggenberger ER. Progressive visual loss
507 (see below). from giant cell arteritis despite high-dose intravenous meth-
3. Hayreh SS. Anterior ischaemic optic neuropathy. I. ylprednisolone. Ophthalmology. 1997;104(5):8548.
Terminology and pathogenesis. Br J Ophthalmol. 26. Chan CC, Paine M, ODay J. Steroid management in giant
1974;58(12):95563. cell arteritis. Br J Ophthalmol. 2001;85(9):10614.
4. Hayreh SS. Anterior ischaemic optic neuropathy. II. Fundus 27. Chevalet P, Barrier JH, Pottier P, Magadur-Joly G, Pottier
on ophthalmoscopy and fluorescein angiography. Br J MA, Hamidou M, et al. A randomized, multicenter, con-
Ophthalmol. 1974;58(12):96480. trolled trial using intravenous pulses of methylprednisolone
5. Hayreh SS. Pathogenesis of cupping of the optic disc. Br J in the initial treatment of simple forms of giant cell arteritis:
Ophthalmol. 1974;58(10):86376. a one year followup study of 164 patients. J Rheumatol.
6. Hayreh SS. Anterior ischemic optic neuropathy. New York: 2000;27(6):148491.
Springer; 1975. 28. Mazlumzadeh M, Hunder GG, Easley KA, Calamia KT,
7. Hayreh SS. Anterior ischemic optic neuropathy. Arch Matteson EL, Griffing WL, et al. Treatment of giant cell
Neurol. 1981;38(11):6758. arteritis using induction therapy with high-dose glucocorti-
8. Hayreh SS. Posterior ischemic optic neuropathy. Ophthal coids: a double-blind, placebo-controlled, randomized pro-
mologica. 1981;182(1):2941. spective clinical trial. Arthritis Rheum. 2006;54(10):
9. Hayreh SS. Anterior ischaemic optic neuropathy. 33108.
Differentiation of arteritic from non-arteritic type and its 29. Kyle V. Treatment of polymyalgia rheumatica/giant cell
management. Eye. 1990;4(1):2541. arteritis. Baillires Clin Rheumatol. 1991;5(3):48591.
10. Hayreh SS. Ophthalmic features of giant cell arteritis. 30. Wilke WS, Hoffman GS. Treatment of corticosteroid-resis-
Baillires Clin Rheumatol. 1991;5(3):43159. tant giant cell arteritis. Rheum Dis Clin North Am.
11. Hayreh SS. Acute ischemic disorders of the optic nerve: 1995;21(1):5971.
pathogenesis, clinical manifestations, and management. 31. Myles AB, Perera T, Ridley MG. Prevention of blindness in
Ophthalmol Clin North Am. 1996;9(3):40742. giant cell arteritis by corticosteroid treatment. Br J
12. Hayreh SS, Podhajsky PA, Raman R, Zimmerman B. Giant Rheumatol. 1992;31(2):1035.
cell arteritis: validity and reliability of various diagnostic 32. Gouet D, Marechaud R, Le Berre D, Alcalay M, Becq-
criteria. Am J Ophthalmol. 1997;123(3):28596. Giraudon B, Boissonnot L, etal. Pronostic de larterite tem-
13. Hayreh SS, Podhajsky PA, Zimmerman B. Ocular manifes- porale traitee. Etude retrospective de 87 observations.
tations of giant cell arteritis. Am J Ophthalmol. 1998; [Prognosis of treated temporal arteritis. Retrospective study
125(4):50920. of 87 cases]. Presse Md. 1986;15(13):6036.
References 261
33. Proven A, Gabriel SE, Orces C, OFallon WM, Hunder GG. 53. Hernandez-Rodriguez J, Garcia-Martinez A, Casademont J,
Glucocorticoid therapy in giant cell arteritis: duration and Filella X, Esteban MJ, Lopez-Soto A, etal. A strong initial
adverse outcomes. Arthritis Rheum. 2003;49(5):7038. systemic inflammatory response is associated with higher
34. Hunder GG, Sheps SG, Allen GL, Joyce JW. Daily and alter- corticosteroid requirements and longer duration of therapy
nate-day corticosteroid regimens in treatment of giant cell in patients with giant-cell arteritis. Arthritis Rheum.
arteritis: comparison in a prospective study. Ann Intern Med. 2002;47(1):2935.
1975;82(5):6138. 54. Ghanchi FD, Dutton GN. Current concepts in giant cell
35. Bengtsson BA, Malmvall BE. An alternate-day corticoster- (temporal) arteritis. Surv Ophthalmol. 1997;42(2):99123.
oid regimen in maintenance therapy of giant cell arteritis. 55. Whitfield AG, Cooke WT, Jameson-Evans P, Rudd C.
Acta Med Scand. 1981;209(5):34750. Temporal arteritis and its treatment with cortisone and
36. Fauchald P, Rygvold O, Oystese B. Temporal arteritis and A.C.T.H. Lancet. 1953;1(6757):40812.
polymyalgia rheumatica. Clinical and biopsy findings. Ann 56. Bennett G. Cortisone therapy of visual loss in temporal
Intern Med. 1972;77(6):84552. arteritis. Br J Ophthalmol. 1956;40(7):4303.
37. Huston KA, Hunder GG, Lie JT, Kennedy RH, Elveback LR. 57. Russell RWR. Giant cell arteritis. Q J Med. 1959;28(112):
Temporal arteritis: a 25-year epidemiologic, clinical, and 47189.
pathologic study. Ann Intern Med. 1978;88(2):1627. 58. Parsons-Smith G. Sudden blindness in cranial arteritis. Br J
38. Kyle V, Hazelman BL. Stopping steroids in polymyalgia rheu- Ophthalmol. 1959;43(4):20416.
matica and giant cell arteritis. BMJ. 1990;300(6721):3445. 59. Aiello PD, Trautmann JC, McPhee TJ, Kunselman AR,
39. Cullen JF. Temporal arteritis. Occurrence of ocular compli- Hunder GG. Visual prognosis in giant cell arteritis.
cations 7 years after diagnosis. Br J Ophthalmol. 1972; Ophthalmology. 1993;100(4):5505.
56(8):5848. 60. Liu GT, Glaser JS, Schatz NJ, Smith JL. Visual morbidity in
40. Cohen DN. Temporal arteritis: improvement in visual prog- giant cell arteritis. Clinical characteristics and prognosis for
nosis and management with repeat biopsies. Trans Am Acad vision. Ophthalmology. 1994;101(11):177985.
Ophthalmol Otolaryngol. 1973;77(1):OP7485. 61. Gonzlez-Gay MA, Blanco R, Rodriguez-Valverde V,
41. Blumberg S, Giansiracusa DF, Docken WP, Kantrowitz FG. Martinez-Taboada VM, Delgado-Rodriguez M, Figueroa M,
Recurrence of temporal arteritis. Clinical recurrence nine etal. Permanent visual loss and cerebrovascular accidents in
years after initial illness. JAMA. 1980;244(15):17134. giant cell arteritis: predictors and response to treatment.
42. Graham E, Holland A, Avery A, Russell RW. Prognosis in Arthritis Rheum. 1998;41(8):1497504.
giant-cell arteritis. Br Med J (Clin Res Ed). 1981;282(6260): 62. Kupersmith MJ, Langer R, Mitnick H, Spiera R, Spiera H,
26971. Richmond M, etal. Visual performance in giant cell arteritis
43. Andersson R, Malmvall BE, Bengtsson BA. Long-term cor- (temporal arteritis) after 1year of therapy. Br J Ophthalmol.
ticosteroid treatment in giant cell arteritis. Acta Med Scand. 1999;83(7):796801.
1986;220(5):4659. 63. Foroozan R, Deramo VA, Buono LM, Jayamanne DG,
44. Salvarani C, Macchioni PL, Tartoni PL, Rossi F, Baricchi R, Sergott RC, Danesh-Meyer H, etal. Recovery of visual func-
Castri C, etal. Polymyalgia rheumatica and giant cell arteri- tion in patients with biopsy-proven giant cell arteritis.
tis: a 5-year epidemiologic and clinical study in Reggio Ophthalmology. 2003;110(3):53942.
Emilia, Italy. Clin Exp Rheumatol. 1987;5(3):20515. 64. Rosenfeld SI, Kosmorsky GS, Klingele TG, Burde RM,
45. Fernandez-Herlihy L. Duration of corticosteroid therapy in Cohn EM. Treatment of temporal arteritis with ocular
giant cell arteritis. J Rheumatol. 1980;7(3):3614. involvement. Am J Med. 1986;80(1):1435.
46. Beevers DG, Harpur JE, Turk KA. Giant cell arteritis: the 65. Postel EA, Pollock SC. Recovery of vision in a 47-year-old
need for prolonged treatment. J Chronic Dis. 1973;26(9): man with fulminant giant cell arteritis. J Clin Neuroophthal
57184. mol. 1993;13(4):26270.
47. Bengtsson BA, Malmvall BE. Prognosis of giant cell arteri- 66. Matzkin DC, Slamovits TL, Sachs R, Burde RM. Visual
tis including temporal arteritis and polymyalgia rheumatica. recovery in two patients after intravenous methylpredniso-
A follow-up study on ninety patients treated with corticos- lone treatment of central retinal artery occlusion secondary
teroids. Acta Med Scand. 1981;209(5):33745. to giant-cell arteritis. Ophthalmology. 1992;99(1):6871.
48. Graham E. Survival in temporal arteritis. Trans Ophthalmol 67. Hayreh SS, Podhajsky PA, Zimmerman MB. Branch retinal
Soc UK. 1980;100(Pt 1):10810. artery occlusion: Natural history of visual outcome.
49. Arashvand K, Alexander WL, Dasgupta B. Duration of Ophthalmology. 2009;116(6):118894.
treatment after eye involvement in giant cell arteritis. 68. Hayreh SS, Zimmerman MB. Central retinal artery occlusion:
J Rheumatol. 2007;34(9):1945. visual outcome. Am J Ophthalmol. 2005;140(3):37691.
50. Bengtsson BA, Malmvall BE. Giant cell arteritis. Acta Med 69. Danesh-Meyer H, Savino PJ, Gamble GG. Poor prognosis of
Scand Suppl. 1982;658:6572. visual outcome after visual loss from giant cell arteritis.
51. Weyand CM, Fulbright JW, Hunder GG, Evans JM, Goronzy Ophthalmology. 2005;112(6):1098103.
JJ. Treatment of giant cell arteritis: interleukin-6 as a bio- 70. Grogan PM, Gronseth GS. Practice parameter: Steroids,
logic marker of disease activity. Arthritis Rheum. acyclovir, and surgery for Bells palsy (an evidence-based
2000;43(5):10418. review): report of the Quality Standards Subcommittee of
52. Hachulla E, Boivin V, Pasturel-Michon U, Fauchais AL, the American Academy of Neurology. Neurology.
Bouroz-Joly J, Perez-Cousin M, etal. Prognostic factors and 2001;56(7):8306.
long-term evolution in a cohort of 133 patients with giant 71. Sharma S. Levels of evidence. Evid Based Eye Care.
cell arteritis. Clin Exp Rheumatol. 2001;19(2):1716. 2001;2(3):1312.
72. Spencer WH, Hoyt WF. A fatal case of giant cell arteritis 92. Hayreh SS. Retinal arterial occlusion with LIF using rTPA
(temporal or cranial arteritis) with ocular involvement. AMA (Correspondence). Ophthalmology. 1999;106(7):12369.
Arch Ophthalmol. 1960;64:8627. 2044.
73. MacFaul PA. Ciliary artery involvement in giant cell arteri- 93. Meadows SP. Temporal arteritis and the loss of vision.
tis. Br J Ophthalmol. 1967;51(8):50512. Trans Ophthalmol Soc UK. 1954;74:1324.
74. Henkind P, Charles NC, Pearson J. Histopathology of 94. Birkhead NC, Wagener HP, Shick RM. Treatment of tem-
ischemic optic neuropathy. Am J Ophthalmol. 1970;69(1): poral arteritis with adrenal corticosteroids; results in fifty-
7890. five cases in which lesion was proved at biopsy. J Am Med
75. MacMichael IM, Cullen JF. Pathology of ischaemic optic Assoc. 1957;163(10):8217.
neuropathy. In: Cant JS, editor. The optic nerve: proceedings 95. Mosher HA. The prognosis in temporal arteritis. AMA
of the second William Mackenzie Memorial Symposium Arch Ophthalmol. 1959;62:6414.
held in Glasgow, 710 September 1971. London: Henry 96. Whitfield AG, Bateman M, Cooke WT. Temporal arteritis.
Kimpton Publishers; 1972. p. 10816. Br J Ophthalmol. 1963;47:55566.
76. Hinzpeter EN, Naumann G. Ischemic papilledema in giant- 97. Meadows SP. Temporal or giant cell arteritis: ophthalmic
cell arteritis. Mucopolysaccharide deposition with normal aspects. In: Smith JL, editor. Neuro-ophthalmology: sym-
intraocular pressure. Arch Ophthalmol. 1976;94(4):6248. posium of the University of Miami and the Bascom Palmer
77. Hayreh SS. The ophthalmic artery: III. Branches. Br J Eye Institute, vol. 4. St. Louis: Mosby; 1968. p. 14857.
Ophthalmol. 1962;46(4):21247. 98. Smith JL. Editors note regarding Meadows SP. Temporal or
78. Hayreh SS. The 1994 Von Sallman lecture: The optic nerve giant cell arteritis: ophthalmic aspects. In: Smith JL, editor.
head circulation in health and disease. Exp Eye Res. Neuro-ophthalmology: Symposium of the University of
1995;61(3):25972. Miami and the Bascom Palmer Eye Institute. St. Louis:
79. Hayreh SS. The blood supply of the optic nerve head and the Mosby; 1968: 4. p. 156.
evaluation of it: myth and reality. Prog Retin Eye Res. 99. Boghen DR, Glaser JS. Ischaemic optic neuropathy. The
2001;20(5):56393. clinical profile and history. Brain. 1975;98(4):689708.
80. Hayreh SS, Baines JA. Occlusion of the posterior ciliary 100. Healey LA, Wilske KR. Manifestations of giant cell arteri-
artery. I. Effects on choroidal circulation. Br J Ophthalmol. tis. Med Clin North Am. 1977;61(2):26170.
1972;56(10):71935. 101. Srensen S, Lorenzen I. Giant-cell arteritis, temporal arteri-
81. Hayreh SS, Chopdar A. Occlusion of the posterior ciliary tis and polymyalgia rheumatica. A retrospective study of
artery. V. Protective influence of simultaneous vortex vein 63 patients. Acta Med Scand. 1977;201(3):20713.
occlusion. Arch Ophthalmol. 1982;100(9):148191. 102. Hugod C, Scheibel M. Temporal arteritisprogressive
82. Petrig BL, Riva CE, Hayreh SS. Laser Doppler flowmetry affection of vision during high-level corticosteroid therapy.
and optic nerve head blood flow. Am J Ophthalmol. A case report. Acta Med Scand. 1979;205(5):4456.
1999;127(4):41325. 103. Calamia KT, Hunder GG. Clinical manifestations of
83. Hayreh SS, Zimmerman MB, Podhajsky P, Alward WL. giant-cell (temporal) arteritis. Clin Exp Rheumatol.
Nocturnal arterial hypotension and its role in optic nerve 1980;6(2):389403.
head and ocular ischemic disorders. Am J Ophthalmol. 104. Herrick PC. Cranial arteritis. A preventable cause of blind-
1994;117(5):60324. ness. Trans Ophthalmol Soc N Z. 1980;32:1069.
84. Hayreh SS, Podhajsky PA, Zimmerman B. Nonarteritic 105. Jones JG, Hazleman BL. Prognosis and management of
anterior ischemic optic neuropathy: time of onset of visual polymyalgia rheumatica. Ann Rheum Dis. 1981;40(1):15.
loss. Am J Ophthalmol. 1997;124(5):6417. 106. Bengtsson BA, Malmvall BE. Eye complications in giant
85. Hayreh SS, Podhajsky P, Zimmerman MB. Role of noctur- cell arteritis. Acta Med Scand Suppl. 1982;658:3843.
nal arterial hypotension in optic nerve head ischemic disor- 107. Delecoeuillerie G, Joly P, Cohen de Lara A, Paolaggi JB.
ders. Ophthalmologica. 1999;213(2):7696. Polymyalgia rheumatica and temporal arteritis: a
86. Hayreh SS, Zimmerman MB, Kimura A, Sanon A. Central retrospective analysis of prognostic features and different
retinal artery occlusion. Retinal survival time. Exp Eye Res. corticosteroid regimens (11 year survey of 210 patients).
2004;78(3):72336. Ann Rheum Dis. 1988;47(9):7339.
87. Kearns TP. Discussion: temporal arteritis: Improvement in 108. Slavin ML, Margolis AJ. Progressive anterior ischemic
visual prognosis and management with repeat biopsies. Trans optic neuropathy due to giant cell arteritis despite high-
Am Acad Ophthalmol Otolaryngol. 1973;77(1):OP845. dose intravenous corticosteroids. Arch Ophthalmol.
88. Thystrup J, Knudsen GM, Mogensen AM, Fledelius HC. 1988;106(9):1167.
Atypical visual loss in giant cell arteritis. Acta Ophthalmol 109. Kyle V, Hazleman BL. Treatment of polymyalgia rheumat-
(Copenh). 1994;72(6):75964. ica and giant cell arteritis. I. Steroid regimens in the first
89. Clearkin L, Caballero J. Recovery of visual function in ante- two months. Ann Rheum Dis. 1989;48(8):65861.
rior ischemic optic neuropathy due to giant cell arteritis. Am 110. Faarvang KL, Pontoppidan Thyssen E. Giant cell arteritis:
J Med. 1992;92(6):7034. loss of vision during corticosteroid therapy. J Intern Med.
90. Clearkin LG. IV steroids for central retinal artery occlusion 1989;225(3):2156.
in giant-cell arteritis. Ophthalmology. 1992;99(10):14824. 111. Weinberg DA, Savino PJ, Sergott RC, Bosley TM. Giant
91. Richard G, Lerche RC, Knospe V, Zeumer H. Treatment of cell arteritis. Corticosteroids, temporal artery biopsy, and
retinal arterial occlusion with local fibrinolysis using recom- blindness. Arch Fam Med. 1994;3(7):6237.
binant tissue plasminogen activator. Ophthalmology. 112. Hwang JM, Girkin CA, Perry JD, Lai JC, Miller NR,
1999;106(4):76873. Hellmann DB. Bilateral ocular ischemic syndrome
References 263
s econdary to giant cell arteritis progressing despite Towards a reduced initial dose]. Rev Md Interne. 2001;
corticosteroid treatment. Am J Ophthalmol. 1999;127(1): 22(7):62430.
1024. 128. Saag KG, Emkey R, Schnitzer TJ, Brown JP, Hawkins F,
113. Schmidt D, Vaith P, Hetzel A. Prevention of serious oph- Goemaere S, et al. Alendronate for the prevention and
thalmic and cerebral complications in temporal arteritis? treatment of glucocorticoid-induced osteoporosis.
Clin Exp Rheumatol. 2000;18(4 Suppl 20):S613. Glucocorticoid-Induced Osteoporosis Intervention Study
114. Loddenkemper T, Sharma P, Katzan I, Plant GT. Risk Group. N Engl J Med. 1998;339(5):2929.
factors for early visual deterioration in temporal arteri- 129. American College of Rheumatology Ad Hoc Committee
tis. J Neurol Neurosurg Psychiatry. 2007;78(11): on Glucocorticoid-Induced Osteoporosis. Recommendations
12559. for the prevention and treatment of glucocorticoid-induced
115. Wagener HP, Hollenhorst RW. The ocular lesions of tem- osteoporosis: 2001 update. Arthritis Rheum. 2001;
poral arteritis. Am J Ophthalmol. 1958;45(5):61730. 44(7):1496503.
116. Hollenhorst RW, Brown JR, Wagener HP, Shick RM. 130. van der Veen MJ, Dinant HJ, van Booma-Frankfort C, van
Neurologic aspects of temporal arteritis. Neurology. Albada-Kuipers GA, Bijlsma JW. Can methotrexate be
1960;10:4908. used as a steroid sparing agent in the treatment of polymy-
117. Raymond LA, Sacks JG, Choromokos E, Khodadad G. algia rheumatica and giant cell arteritis? Ann Rheum Dis.
Short posterior ciliary artery insufficiency with hyper- 1996;55(4):21823.
thermia (Uhthoffs symptom). Am J Ophthalmol. 1980; 131. Spiera RF, Mitnick HJ, Kupersmith M, Richmond M,
90(5):61923. Spiera H, Peterson MG, etal. A prospective, double-blind,
118. Gonzlez-Gay MA, Garcia-Porra C, Llorca J, Hajeer AH, randomized, placebo controlled trial of methotrexate in the
Braas F, Dababneh A, etal. Visual manifestations of giant treatment of giant cell arteritis (GCA). Clin Exp Rheumatol.
cell arteritis Trends and clinical spectrum in 161 patients. 2001;19(5):495501.
Medicine Baltimore. 2000;79(5):28392. 132. Hoffman GS, Cid MC, Hellmann DB, Guillevin L, Stone
119. Cooke WT, Cloake PCP, Govan ADT, Colbeck JC. JH, Schousboe J, etal. A multicenter, randomized, double-
Temporal arteritis: a generalized vascular disease. Q J Med. blind, placebo-controlled trial of adjuvant methotrexate
1946;15(1):4775. treatment for giant cell arteritis. Arthritis Rheum.
120. Cardell BS, Hanley T. A fatal case of giant-cell or temporal 2002;46(5):130918.
arteritis. J Pathol Bacteriol. 1951;63(4):58797. 133. Jover JA, Hernandez-Garcia C, Morado IC, Vargas E,
121. Kyle V, Hazleman BL. Treatment of polymyalgia rheumat- Banares A, Fernandez-Gutierrez B. Combined treatment of
ica and giant cell arteritis. II. Relation between steroid dose giant-cell arteritis with methotrexate and prednisone. a ran-
and steroid associated side effects. Ann Rheum Dis. domized, double-blind, placebo-controlled trial. Ann Intern
1989;48(8):6626. Med. 2001;134(2):10614.
122. Nesher G, Sonnenblick M, Friedlander Y. Analysis of ste- 134. Mahr AD, Jover JA, Spiera RF, Hernandez-Garcia C,
roid related complications and mortality in temporal arteri- Fernandez-Gutierrez B, Lavalley MP, et al. Adjunctive
tis: a 15-year survey of 43 patients. J Rheumatol. 1994; methotrexate for treatment of giant cell arteritis: an indi-
21(7):12836. vidual patient data meta-analysis. Arthritis Rheum. 2007;
123. Bevis T, Ratnakaram R, Smith MF, Bhatti MT. Visual loss 56(8):278997.
due to central serous chorioretinopathy during corticoster- 135. De Silva M, Hazleman BL. Azathioprine in giant cell
oid treatment for giant cell arteritis. Clin Experiment arteritis/polymyalgia rheumatica: a double-blind study.
Ophthalmol. 2005;33(4):4379. Ann Rheum Dis. 1986;45(2):1368.
124. Steichen O, Chauveheid MP, Lidove O, Doan S, Papo T. 136. Schaufelberger C, Mollby H, Uddhammar A, Bratt J,
Baisse de lacuite visuelle chez une patiente traitee par Nordborg E. No additional steroid-sparing effect of
corticoides pour maladie de Horton: chorioretinopathie cyclosporine A in giant cell arteritis. Scand J Rheumatol.
sereuse centrale iatrogene. [Iatrogenic central serous cho- 2006;35(4):3279.
rioretinopathy during glucocorticoid therapy for temporal 137. Hoffman GS, Cid MC, Rendt-Zagar KE, Merkel PA,
arteritis]. Rev Md Interne. 2006;27(9):7025. Weyand CM, Stone JH, etal. Infliximab for maintenance of
125. Lespessailles E, Poupon S, Adriambelosoa N, Pothuaud L, glucocorticosteroid-induced remission of giant cell arteri-
Siroux V, Bouillon S, etal. Glucocorticoid-induced osteo- tis: a randomized trial. Ann Intern Med. 2007;146(9):
porosis: is the bone density decrease the only explanation? 62130.
Joint Bone Spine. 2000;67(2):11926. 138. Martinez-Taboada VM, Rodriguez-Valverde V, Carreno L,
126. Cacoub P, Chemlal K, Khalifa P, Wechsler B, De Gennes Lopez-Longo J, Figueroa M, Belzunegui J, etal. A double-
C, Belmatoug N, et al. Deflazacort versus prednisone in blind placebo controlled trial of etanercept in patients with
patients with giant cell arteritis: effects on bone mass loss. giant cell arteritis and corticosteroid side effects. Ann
J Rheumatol. 2001;28(11):24749. Rheum Dis. 2008;67(5):62530.
127. Chevalet P, Barrier JH, Glmarec J, el Kouri D, Hamidou 139. Ahmed MM, Mubashir E, Hayat S, Fowler M, Berney SM.
M, de Wazires B, etal. Maladie de Horton du sujet age de Treatment of refractory temporal arteritis with adalimumab.
plus de 75 ans: particularites evolutives, complications de Clin Rheumatol. 2007;26(8):13535.
la corticotherapie. Etude comparative sur une serie de 164 140. Nesher G, Sonnenblick M. Steroid-sparing medications
patients. Vers une dose dattaque reduite. [Hortons disease in temporal arteritis: report of three cases and review of
in elderly patients aged over 75: clinical course, complica- 174 reported patients. Clin Rheumatol. 1994;13(2):
tions of corticotherapy. Comparative study of 164 patients. 28992.
141. Cantini F, Niccoli L, Nannini C, Bertoni M, Salvarani C. 155. Calamia KT, Hunder GG. Giant cell arteritis (temporal
Diagnosis and treatment of giant cell arteritis. Drugs Aging. arteritis) presenting as fever of undetermined origin.
2008;25(4):28197. Arthritis Rheum. 1981;24(11):14148.
142. Costello F, Zimmerman MB, Podhajsky PA, Hayreh SS. 156. Lincoff NS, Erlich PD, Brass LS. Thrombocytosis in tem-
Role of thrombocytosis in diagnosis of giant cell arteritis poral arteritis rising platelet counts: a red flag for giant cell
and differentiation of arteritic from non-arteritic anterior arteritis. J Neuroophthalmol. 2000;20(2):6772.
ischemic optic neuropathy. Eur J Ophthalmol. 2004; 157. Hu Z, Yang Q, Zheng S, Tang J, Lu W, Xu N, etal. Temporal
14(3):24557. arteritis and fever: report of a case and a clinical reanalysis
143. Bengtsson BA, Malmvall BE. Hematological observations of 360 cases. Angiology. 2000;51(11):9538.
in giant cell arteritis. Acta Med Scand Suppl. 1982; 158. Foroozan R, Danesh-Meyer H, Savino PJ, Gamble G,
658:447. Mekari-Sabbagh ON, Sergott RC. Thrombocytosis in
144. Gibb WR, Urry PA, Lees AJ. Giant cell arteritis with spinal patients with biopsy-proven giant cell arteritis.
cord infarction and basilar artery thrombosis. J Neurol Ophthalmology. 2002;109(7):126771.
Neurosurg Psychiatry. 1985;48(9):9458. 159. Nesher G, Berkun Y, Mates M, Baras M, Rubinow A,
145. De Keyser J, De Klippel N, Ebinger G. Thrombocytosis Sonnenblick M. Low-dose aspirin and prevention of cranial
and ischaemic complications in giant cell arteritis. BMJ. ischemic complications in giant cell arteritis. Arthritis
1991;303(6806):825. Rheum. 2004;50(4):13327.
146. Kaiser M, Weyand CM, Bjornsson J, Goronzy JJ. Platelet- 160. Lee MS, Smith SD, Galor A, Hoffman GS. Antiplatelet and
derived growth factor, intimal hyperplasia, and ischemic anticoagulant therapy in patients with giant cell arteritis.
complications in giant cell arteritis. Arthritis Rheum. Arthritis Rheum. 2006;54(10):33069.
1998;41(4):62333. 161. Singh S, Dass R. The central artery of the retina. I. Origin
147. Cid MC, Font C, Oristrell J, de la Sierra A, Coll-Vinent B, and course. Br J Ophthalmol. 1960;44:193212.
Lopez-Soto A, etal. Association between strong inflamma- 162. Pearson TC. The risk of thrombosis in essential thrombo-
tory response and low risk of developing visual loss and cythemia and polycythemia vera. Semin Oncol. 2002;29(3
other cranial ischemic complications in giant cell (tempo- Suppl 10):1621.
ral) arteritis. Arthritis Rheum. 1998;41(1):2632. 163. Crow RW, Katz BJ, Warner JE, Alder SC, Zhang K,
148. Gonzlez-Gay MA, Garca-Porrua C, Vzquez-Caruncho Schulman S, et al. Giant cell arteritis and mortality. J
M. Polymyalgia rheumatica in biopsy proven giant cell Gerontol A Biol Sci Med Sci. 2009;64(3):3659.
arteritis does not constitute a different subset but differs 164. Graham E. Survival in temporal arteritis. Trans Ophthalmol
from isolated polymyalgia rheumatica. J Rheumatol. Soc UK. 1980;100(Pt 1):10810.
1998;25(9):17505. 165. Uddhammar A, Eriksson AL, Nystrom L, Stenling R,
149. Gonzlez-Alegre P, Ruiz-Lopez AD, Abarca-Costalago M, Rantap-Dahlqvist S. Increased mortality due to cardio-
Gonzalez-Santos P. Increment of the platelet count in tem- vascular disease in patients with giant cell arteritis in north-
poral arteritis: response to therapy and ischemic complica- ern Sweden. J Rheumatol. 2002;29(4):73742.
tions. Eur Neurol. 2001;45(1):435. 166. Gran JT, Myklebust G, Wilsgaard T, Jacobsen BK. Survival
150. Liozon E, Herrmann F, Ly K, Robert PY, Loustaud V, Soria in polymyalgia rheumatica and temporal arteritis: a study
P, etal. Risk factors for visual loss in giant cell (temporal) of 398 cases and matched population controls.
arteritis: a prospective study of 174 patients. Am J Med. Rheumatology Oxford. 2001;40(11):123842.
2001;111(3):2117. 167. Nuenninghoff DM, Hunder GG, Christianson TJ,
151. Harrell FE. The LOGIST procedure. SUGI supplemental McClelland RL, Matteson EL. Mortality of large-artery
library users guide, version 5. Cary: SAS Institute, Inc; complication (aortic aneurysm, aortic dissection, and/or
1986. p. 26993. large-artery stenosis) in patients with giant cell arteritis: a
152. Moncada S, Vane JR. Arachidonic acid metabolites and population-based study over 50 years. Arthritis Rheum.
the interactions between platelets and blood-vessel walls. 2003;48(12):35327.
N Engl J Med. 1979;300(20):11427. 168. Pipitone N, Boiardi L, Bajocchi G, Salvarani C. Long-term
153. Wu KK. Platelet hyperaggregability and thrombosis in outcome of giant cell arteritis. Clin Exp Rheumatol.
patients with thrombocythemia. Ann Intern Med. 2006;24(2 Suppl 41):S6570.
1978;88(1):711. 169. Gonzlez-Gay MA, Vazquez-Rodriguez TR, Lopez-Diaz
154. Malmvall BE, Bengtsson BA. Giant cell arteritis. Clinical MJ, Miranda-Filloy JA, Gonzalez-Juanatey C, Martin J,
features and involvement of different organs. Scand J etal. Epidemiology of giant cell arteritis and polymyalgia
Rheumatol. 1978;7(3):1548. rheumatica. Arthritis Rheum. 2009;61(10):145461.
Pathogenesis of Classical Non-arteritic
Anterior Ischemic Optic Neuropathy 14
Non-arteritic anterior ischemic optic neuropathy Knox and Duke [3] reported a 55-year old man with
(NA-AION) is the most common type of ischemic visual loss for 2weeks who committed suicide while
optic neuropathy. Although findings of sudden onset in the hospital. The patient had occlusion of the left
of segmental visual field defect, initially with optic common carotid artery and generalized arteriosclerosis.
disc edema with a few peripapillary hemorrhages and Histopathology revealed edema of the temporal half of
later on optic atrophy were reported in the very early the optic nerve head. The optic nerve just posterior to
literature, the pathogenesis and management of the lamina cribrosa showed, in the temporal part, a
NA-AION had been the focus of continuing contro- focal necrosis and swollen and fragmented nerve
versy. To explain its pathogenesis, multiple theories fibers, with distension of the lamina cribrosa.
have been postulated, varying from some degree of Lieberman etal. [4] described a 68-year old man who
logic to some quite preposterous. NA-AION was had had no ophthalmologic examination and died of
referred to in the past as the arteriosclerotic variety, septicemia; only postmortem information about his
to distinguish it from arteritic AION associated with eyes and optic nerve was available. The short posterior
giant cell arteritis (GCA). It seems Uhthoff in 1924 ciliary artery (PCA) entering the choroid near the
first described three cases of the arteriosclerotic type center of the macula was occluded by fibrin-
[1]. The next major description seems to be by Francois containing thrombus. There was no outer retinal or
et al. in 1956 [2]; they described this clinical entity retinal pigment epithelial change. The optic nerve
under the title of vascular pseudopapilitis, and showed focal infarction 3mm behind the lamina crib-
stressed the etiological importance of arteriosclerotic rosa in the region of a thromboembolic occlusion of
small vessel disease. three discrete pial and pia-derived arterioles. There
To find a rational management for a disease, the was loss of astrocytes, nerve fibers and myelin, and a
first essential is to understand its pathogenesis. mild infiltration of swollen histiocytes (Gitter cells) in
Therefore, to begin with, a detailed discussion of the the infarcted region. In view of the lack of any prior
pathogenesis of NA-AION is crucial. ophthalmic information, it is hard to determine the full
implications of their findings. They did not find muco-
polysaccharide deposition in the area of infarction.
Guyer etal. [5] reported histopathological changes in
Histopathology of the Optic Nerve
a patient with non-Hodgkins mixed-cell lymphoma
in NA-AION with bilateral ischemic optic neuropathy and retinal
vascular occlusions. Bilateral NA-AION developed
Pathological studies provide important information in the left eye 2 weeks and right eye 1 week before
about the pathogenesis of a disease. Unlike arteritic death. Histopathology revealed necrosis of the right
AION due to GCA (see Chap. 12), there are only a few optic nerve head and the left optic nerve head had isch-
histopathology reports of NA-AION, because, in con- emic infarction. Both nerves showed fibrovascular pial
trast to GCA, it is not a systemic disease with a risk of septal infiltration by tumor from the eyeball to the dis-
death. We have the following information available on tal 23mm of the intracanalicular portion. Tumor cells
histopathology in NA-AION. were crowded in the extravascular area. The nerves

266 14 Pathogenesis of Classical Non-arteritic Anterior Ischemic Optic Neuropathy
were slightly swollen with ischemic necrosis and gen- intact nerve from ischemic edema and cavernous
eral loss of cellularity and myelin staining of all nerve swelling, or a second ischemic lesion.
fiber bundles from the eyeball to the mid area of the Ellenberger and Netsky [21] studied 40 optic nerves
intracanalicular part; the infarcted area tapered to a in persons over the age of 45years. They found diffuse
0.5mm central area. The central retinal artery (CRA) thickening of the PCAs and arterioles within the nerve.
and central retinal vein of the right eye and CRA of the They considered these changes due to atherosclerosis
left eye were occluded by thrombi with bacteria. Thus, and arteriosclerosis.
this patient had both anterior (AION) and posterior
(PION) ischemic optic neuropathies and central reti-
nal artery occlusion in both eyes, as well as central
retinal vein occlusion in the right eye. Knox etal. [6] Histopathological Reports of a Case
described histopathological changes without differen- by Levin and Colleagues [22, 23]
tiating them into arteritic and non-arteritic ischemic
optic neuropathy. The time interval between the onset
I find this case very interesting and instructive, requir-
of ischemic optic neuropathy and histopathological
ing a detailed discussion. They reported histopatho-
study in their cases varied from 1 day to as long as
logical findings in NA-AION in a 70-year old man.
3years. The histopathologic criteria for a diagnosis of
The patient had peripheral vascular disease (including
ischemic optic neuropathy used by them were the
femoral-popliteal grafting), coronary artery disease,
presence of localized ischemic edema, cavernous
congestive heart failure, atrial fibrillation, hypertension
degeneration, or an area of atrophy located superior or
and renal failure. He had thrombolectomy for femoral-
inferior in the optic nerve. They combined that infor-
popliteal thromboembolism. The patient experienced
mation with the cases reported in the literature till
sudden, painless loss of vision in the left eye the morn-
1993, with a history of abrupt loss of vision [720].
ing after thrombolectomy of an aortofemoral graft.
They reported histopathological changes in 193 eyes
On examination his visual acuity was 9/200 and there
with ischemic optic neuropathy (a combination of
was an inferior altitudinal field defect. He was diag-
their own and those reported in the literature), but,
nosed to have NA-AION [22]. He died 20days after
unfortunately, their study had a mixture of both arter-
the onset of NA-AION from complications of renal
itic and non-arteritic AION which makes it hard to
failure, pancreatitis and hypercalcemia. The eye was
determine the reliability of its findings for NA-AION,
removed 12h and 25min after death, fixed in formalin
although in both types of ischemic optic neuropathy
and serially sectioned [23]. This eye formed the basis
the site of lesions is in the optic nerve head. They
of the following two reports by that group.
found: Cavernous degeneration was present in 69
nerves (36%). Mucopolysaccharide (MPS) was pres- 1. Levin and Louhab [22] reported a loss of axons in
ent in 37 cavernous lesions 1 month or longer after the superior part of the optic nerve and a moderate
loss of vision. Cavernous lesions were seen in three amount of gliosis in this area. The cavernous spaces
eyes in which peripapillary retinal nerve fiber layer contained hyaluronidase-sensitive Alcian blue-pos-
hemorrhage had been observed prior to death. Atrophic itive material, consistent with Schnabels cavernous
lesions, the most common pattern, were observed in atrophy. Pial septa were intact. On the margin of the
133 optic nerves (66.8%). More than one ischemic infarct, areas of hypercellularity with relatively pre-
lesion was seen in 38 optic nerves (19.7%). Bilateral served axonal elements were present. No emboli
ischemic lesions were seen in 50 (35.2%) of 142 paired were seen in pial, PCA or CRA. They identified
eyes. They concluded that ischemic optic nerve retinal ganglion cells undergoing apoptosis, and
lesions are initially acellular and later show mac- stated that the most likely explanation for the apop-
rophage infiltration. Cavernous lesions with MPS are tosis is the injury to the retinal ganglion cell axon.
present 4weeks or longer after vision loss. The loca- 2. Tesser etal. [23], in the same optic nerve, on serial
tion of MPS posteriorly and along the internal margin sections, studied the three-dimensional anatomic con-
suggests that MPS is produced at the edges of lesions. figuration of the infarct in NA-AION. They identified
They postulated that progressive vision loss in ischemic and reconstructed the area of axonal loss within each
optic neuropathy may be secondary to compression of section of the optic nerve. The loss was in the superior
Histopathology of the Optic Nerve in NA-AION 267
part of the nerve, encircling the CRA at its greatest blood supply of the optic nerve. The CRA gives out
extent. Remaining areas of the nerve appeared healthy, the following two sets of branches to supply the
and, notably, the periphery of the uninvolved inferior anterior part of the optic nerve:
portion of the nerve was normal. Three-dimensional (a) Pial Branches: Before entering the optic nerve,
analysis revealed two distinct areas of infarct at the the CRA during its intravaginal course (i.e. on
posterior extent of the lesion which coalesced toward the pial surface of the optic nerve before entering
the center of the lesion and finally tapered as the infarct the optic nerve) gives pial branches in 100% of
reached the optic nerve head. Sagittal reconstructions cases [24, 25]. They contribute to the peripheral
gave the appearance of a two-pronged fork posteriorly centripetal vascular supply of the anterior part of
connecting to a single handle anteriorly. The total the optic nerve invariably the inferior part and
length of the nerve involved by the infarct was approx- less frequently the medial and lateral parts of the
imately 1.5 mm. Based on their findings they con- optic nerve (Fig.14.1).
cluded that: The morphology of this NA-AION (b) Intraneural Branches: During its intraneural
infarct is not consistent with disease of large or small course within the optic nerve, the CRA gives
vessels and, more likely, represents a form of compart- one to eight branches in 75% of the cases [24,
ment syndrome that causes tissue ischemia. 25]. They constitute the centrifugal vascular
supply system of the axial part of the optic
Unfortunately, while their findings are valuable, their
nerve when those branches are present. There is
interpretation and conclusion are not correct. A full dis-
wide interindividual variation in the area sup-
cussion of this important case is required to explain why
plied by these intraneural branches (Figs.14.1
their conclusion is not valid and what new information
and 14.2).
it provides us about pathogenesis of NA-AION.
2. I have always stressed to my residents and fellows
To understand the mechanism of the development
that it is a patient who is coming to see an ophthal-
of infarct in the axial part of the optic nerve in this
mologist and not an eyeball, optic nerve, retina,
case, it is crucial to consider the following facts about
cornea or other parts of the eye. What matters is not
the blood supply of the optic nerve.
only what is going on in the eye or the optic nerve,
1. The location and distribution of the infarct was but in the patient as whole. Unfortunately, that fact
exactly in the region of distribution of fine intraneu- is often completely ignored, resulting in misleading
ral branches of the CRA supplying the superior information and wrong conclusions. This is of fun-
axial part of the optic nerve. In this connection, it is damental importance in this patient for the follow-
essential to consider the role of the CRA in the ing reasons:
C
R S Col. Br.
PCA
D
A
Pia
Fig.14.1 Schematic
representation of blood supply
of the optic nerve (Modified
from Hayreh [26]). A
arachnoid, C choroid, CRA ON
central retinal artery, Col. Br.
Collateral branches, CRV PR
central retinal vein, D dura, LC
LC lamina cribrosa, ON optic
SAS
R retina, S sclera, SAS CRV
PCA CRA
subarachnoid space
Fig.14.2 Course of the

a b
central artery of the retina
(CAR) as seen from below,
before (*) and after its site of
8
penetration into the optic
nerve (ON). Its intraneural 7 6
(I/N) course is seen on
splitting open the optic nerve
anterior to the site of 5
penetration. It shows branches
of the central artery of the
retina two (1, 2) before its ON
penetration (PPN) into the split open
optic nerve, and six (38) I/N
during its intraneural course.
(b) is Camera Lucida drawing 4
of (a) for better clarity
3 1
2
CAR
ON
unsplit
(a) This patient [22, 23] had peripheral vascular dis- gave no information about the medications or the
ease, coronary artery disease, congestive heart time those were taken. This information is
failure, atrial fibrillation, hypertension and renal extremely pertinent in all NA-AION cases
failure, and underwent thrombolectomy for fem- because multiple arterial hypotensive drugs, par-
oral-popliteal thromboembolism. Following that, ticularly those administered in the evening or
his hematocrit dropped to a very low level and he before going to bed, are a potent source of noc-
had four units of blood transfusion. The patient turnal arterial hypotension (as was very well
developed visual loss in the left eye the morning demonstrated in my 24-h ambulatory blood pres-
after thrombolectomy. That means he had severe sure monitoring in more than 700 patients [27
vasculopathic changes in all his arteries, includ- 29]). In our study [30] on the time of onset of
ing the CRA and its fine intraneural branches, NA-AION, we found that 73% discovered the
and several other vascular problems. He died visual loss on waking up from sleep or the first
from complications of renal failure, pancreatitis opportunity to use the vision critically thereafter,
and hypercalcemia. The statement by the authors while the remainder were not sure of the time of
that: The morphology of this NA-AION infarct onset, suggesting that more than 73% actually
is not consistent with disease of large or small developed NA-AION during sleep. It is well
vessels is not at all consistent with the presence established that during sleep there is nocturnal
of widespread marked systemic arterial disease. arterial hypotension, which acts as the precipitat-
(b) The patient experienced sudden, painless loss of ing factor for the development of NA-AION in
vision in the left eye the morning after throm- persons with predisposing risk factors. This
bolectomy of an aortofemoral graft. For his patient had all the possible predisposing risk fac-
various cardiovascular problems, he must have tors one can think of.
been prescribed various drugs which are potent 3. There is a universal misconception that unless one
arterial hypotensives. Unfortunately the authors can demonstrate on histopathology the presence of
Histopathology of the Optic Nerve in NA-AION 269
actual occlusion, thrombus or embolus, optic nerve a

damage cannot be attributed to the disease of large
or small vessels. This is not always true. I have
stressed time and again that NA-AION in the vast
majority is caused by transient hypoperfusion or
nonperfusion of the optic nerve head vascular bed
during sleep due to nocturnal arterial hypotension
in persons with various predisposing risk factors
[3032]. This has been very well documented in my
fluorescein angiographic studies in NA-AION. (I
perform fluorescein angiography in all patients with
NA-AION). In arteritic AION there is occlusion
of the PCAs (see angiograms in Figs. 3.24, 3.25,
3.283.34, 3.36, 3.37, 3.43, 3.49 and 3.50; by sharp b
contrast, in the vast majority of NA-AION eyes
there is NO actual occlusion of the vessels in the
optic nerve head but only transient hypoperfusion
(see Fig. 14.3).
In such a complex scenario, let us put all these fac-
tors together to understand what caused the devel-
opment of infarct in the axial part of the optic nerve
and NA-AION in this case.
(a) The patient had peripheral vascular disease
(including femoral-popliteal grafting), coronary
artery disease, congestive heart failure, atrial
fibrillation, hypertension and renal failure. He
had femoral-popliteal thromboembolism; thus,
he had severe generalized vasculopathy. There
must have been marked vascular changes in the Fig.14.3 Two fluorescein fundus angiograms at 6-s interval in
a patient with NA-AION. (a) During the retinal arterial phase of
CRA as a part of that, and in its fine intraneural
the dye circulation (15 s after the injection of fluorescein), shows
branches supplying the axial region of the optic no filling of the temporal, superior and inferior peripapillary
nerve. Thus, the statement by the authors that choroid and superior choroidal watershed zone (arrow), and the
The morphology of this NA-AION infarct is optic disc, with normal filling of both the medial and lateral pos-
terior ciliary arteries. (b) During the retinal arteriovenous phase
not consistent with disease of large or small
of the dye circulation (21 s after the injection of fluorescein),
vessels is not accurate. there is complete filling of the peripapillary choroid (main
(b) The patient must have been taking multiple source of blood supply to the ONH) and the watershed zone.
drugs for his coronary artery disease, conges- These angiograms illustrate the most important point that in
NA-AION there is only delayed filling of the peripapillary chor-
tive heart failure, atrial fibrillation, hyperten-
oid and the watershed zone and no occlusion of the PCA
sion and renal failure. It is well-known that
most of those are powerful hypotensive drugs.
pressure in the optic nerve head vessels below a
Their administration at bedtime causes a marked
critical level without any actual occlusion of its
degree of nocturnal arterial hypotension, as
vessels [3133]. Therefore, histopathology
demonstrated by our [2729] 24-h ambulatory
would not show any occlusion of the fine ves-
blood pressure recording. This is further sup-
sels in the optic nerve head.
ported by the fact that he noticed his visual loss
the morning after thrombolectomy Thus, when all these facts are put together, a clear
(c) In the vast majority of patients, NA-AION is picture emerges. In a highly vasculopathic person, mor-
produced by transient hypoperfusion or nonper- phological changes in the CRA and its fine intraneural
fusion of circulation in the optic nerve head branches predisposed the optic nerve to ischemic dam-
during sleep, caused by the fall of perfusion age. Under such circumstances, the sequence of events
for development of NA-AION was most probably as 2. However, as discussed above, the case reported by
follows: nocturnal arterial hypotensionfall of perfu- Tesser et al. [23] shows for the first time that, in
sion pressure below the critical level in the intraneural some patients, NA-AION can be caused by isch-
branches of the CRAdevelopment of axial infarct in emia of the axial part of the optic nerve supplied by
the optic nerveNA-AION. There was no demonstra- the intraneural branches of the CRA.
ble histopathological vascular occlusion because there
was only transient hypoperfusion or nonperfusion dur-
ing sleep and no actual occlusion caused by thrombosis
or embolism. In classical NA-AION, which is primarily Historical Review of the Various
due to involvement of the PCA circulation to the optic Hypotheses on Pathogenesis of NA-AION
nerve, fluorescein fundus angiography has almost invari-
ably shown hypoperfusion or nonperfusion in the peri-
Since I have been actively and closely involved with
papillary choroid or PCA watershed zones, but no actual
investigation of the pathogenesis and other features
occlusion (see Fig.14.3). Unfortunately, lack of knowl-
of NA-AION since 1969, I will discuss in detail the
edge of this complex scenario misled the authors.
progression of our knowledge about it. The follow-
I have entered into this lengthy discussion about the
ing brief historical review of the literature is essential
mechanism of development of NA-AION in this
to explain the evolution of our concept about the
patient, because the histopathological report by Tesser
pathogenesis of NA-AION over the decades. It
etal. [23]. shows for the first time that in some patients
reveals multiple contradictory and thoroughly con-
NA-AION can be caused by ischemia of the axial part
fusing views. There is voluminous literature. The
of the anterior optic nerve supplied by the intraneural
discovery in 1969 that the optic nerve head was pri-
branches of the CRA, although in classical NA-AION
marily supplied by the PCA circulation [35] acted
it is due to circulatory impairment in PCA circulation
as a watershed between the old concepts and the
supplying the optic nerve head, which can almost
newones.
invariably be demonstrated on fluorescein fundus
Historically, we can divide the literature about vari-
angiography during the very early stages. By contrast,
ous theories postulated to explain the pathogenesis of
unfortunately, it is not possible to get that information
NA-AION into the following phases:
clinically when it is due to impaired circulation in the
axial circulation supplied by the central retinal artery. 1. Old literature on the pathogenesis of AION before
1969
2. Literature on pathogenesis of NA-AION proposed
since 1969
(a) Pathogenesis of NA-AION originally postu-
Site of Lesion in AION as a Whole
lated by me in 1969
Information about this is best provided by a large num- (b) Pathogenesis of NA-AION proposed by Ar-
ber of histopathological studies dealing with arteritic nold in 2003
AION, discussed in detail in Chap. 12. Those studies (c) Pathogenesis of NA-AION proposed by Levin
showed that the site of the lesion in AION is in the optic and Danesh-Meyer in 2008
nerve head, in the region supplied by the PCA circula- 3. Multifactorial pathogenesis of non-arteritic AION
tion (see Chap. 3), as was also shown by the develop- Following is a brief discussion of each of these
ment of AION after experimental occlusion of the PCA topics.
in rhesus monkeys [34]. Unfortunately, in NA-AION
there are only a few histopathological studies, which
are discussed above; however, they do show that the Review of Old Literature on the
area involved is the same as in arteritic AION. Based on
Pathogenesis of AION Before 1969
this information, one can conclude the following:
1. Like arteritic AION, the site of lesion in the classi- The site of the lesion in AION was thought to lie in the
cal NA-AION is in the area supplied by the PCA optic nerve, although some authors considered it to be
circulation in the optic nerve head [3133]. due to involvement of the retina. It was almost
Historical Review of the Various Hypotheses on Pathogenesis of NA-AION 271
universally agreed that it was due to interference with branches of the ophthalmic artery in 100 human speci-
the circulation of one of these structures. Skillern and mens [24, 25, 7780] and also other studies [8184]
Lockhart [36], however, thought that AION in diabetes showed that there was no mythical central artery of
was due to the toxic effect of prolonged hyperglycemia the optic nerve corresponding to the description men-
and failure of glucose utilization, and did not consider tioned by Francois etal. [70, 7376]. Hence, occlusion
arterial circulatory disorders a significant factor. of the central artery of the optic nerve as the reason
Most authors, while describing a circulatory lesion for the development of AION is an exploded concept.
in the optic nerve, made vague, nonspecific statements Miller and Smith [85] in 1966 described 11 cases
as to the exact lesion. It was considered to be a circula- with NA-AION and called it ischemic optic neuropa-
tory disturbance in the optic nerve [37, 38] due to thy. They did not discuss its pathogenesis except for a
pathologic changes [3943] or occlusive disease brief review of the literature on the subject. They felt
[4447] of the small vessels of the nerve. Occlusion of that in diabetics it was due to disease of the small
the small nutrient vessels of the optic nerve was gener- vessels in the optic nerve. Boghen and Glaser [86],
ally considered to be due to arteriosclerosis [1, 3841, based on a retrospective study of 29 patients with what
43, 4855]; others thought the change mainly affected they called idiopathic ischemic optic neuropathy
the CRA [5659] although Igersheimer [41] and concluded: At the present time neither the pathophys-
Rintelen [43] were of the opinion that arteriosclerosis iology nor anatomical background of primary, non-
of the CRA had little effect on the optic nerve. The arteritic ischaemic optic neuropathy is completely
possibility that arteritis of vessels of the optic nerve understood. There is no conclusive evidence either
and retina was responsible was also mentioned [60]. from previous reports or from the present study to
Some authors stated that these small vessel changes implicate the extracranial carotid system, and the roles
were located in the anterior retrobulbar part of the of arteriosclerotic vascular disease and hypertension
optic nerve [12, 50, 61, 62]; while according to others are uncertain.
these were in the vessels of the optic disc [10, 6365]. Interference with the blood supply to the retina was
Meadows [66] mentioned that there was ischemic considered the cause of AION by some workers [60,
damage to the optic disc and anterior portion of the 87, 88] because CRA pressure fell, as judged by
optic nerve. ophthalmodynamometry. Uhthoff [1] and Smith and
Francois and co-workers [53, 6770] and others Greene [60] thought it was due to the involvement of
[47, 71, 72] postulated that AION was due to occlusion vessels in both the optic nerve and retina.
of the central artery of the optic nerve (Fig. 14.4) It is interesting to note that Ellenberger etal. [21]
described by Francois etal. [70, 7376]. My anatomic examined 40 optic nerves of people older than 45years
studies on the blood supply of the optic nerve and for atherosclerosis and arteriosclerosis and found that
Fig.14.4 Central artery of

the optic nerve, described by
Francois and Neetens [1].
optic nerve; CRA Central CRA CAON
retinal artery (Modified from
Fig.9 by Francois and
Neetens [1])
changes in the optic nerve vessels were similar to those involve more and more of the peripapillary choroid.
seen elsewhere in the body, suggesting that the optic The literature contains many histopathological reports
nerve vessels were constantly involved in systemic showing the involvement of the posterior ciliary arter-
atherosclerosis and arteriosclerosis. In AION, vascular ies by temporal arteritis and infarction in the optic
changes in the optic nerve correlated slightly or not at nerve. The pale oedema involving only the optic disc
all with those in the retina [38, 40]. can be produced only by occlusion of the ciliary arte-
This brief historical review of the old literature on rial supply to the disc and not by an occlusion of the
the subject reflects the extent of confusion and contro- central retinal artery; the latter would produce a gener-
versy about the nature of the vascular basis of the alized oedema of the retina.
AION. It seems the major factor responsible for contro- Based on my experience of dealing with a large vol-
versy was the lack of understanding and contradictory ume of both arteritic and non-arteritic AION patients
views about the blood supply of the optic nerve head. since 1969, the only modification I would make to the
above description is about the stress I laid on pale
optic disc edema at that time, which was mainly
because many of the cases that I had seen at that time
Review of Literature on Pathogenesis
were of arteritic AION; however, my subsequent stud-
of NA-AION Since 1969 ies showed that that does not hold true for NA-AION,
as discussed regarding the clinical features of NA-AION
My Postulated Pathogenesis of NA-AION in Chap. 16.
in 1969
Based on my studies, in 1969 I described the optic Pathogenesis of NA-AION Proposed

nerve head as primarily supplied by the PCA circula- by Arnold in 2003
tion [35] I made the following observation at that time,
describing for the first time the pathogenesis of AION: Arnold [90], in 2003, proposed a hypothesis about the
lschaemic optic neuropathy with a pale oedema of the pathogenesis of NA-AION, based on a review of the
optic disc is a well documented condition. It is seen literature dealing with: (1) histopathology, (2) electron
mostly in elderly persons. The retinal vessels are microscopic cast, (3) optic nerve head blood flow, (4)
normal for the age. There is a sudden complete loss of various risk factors, (5) vasospasm and impaired regu-
vision or development of a variety of visual field lation, and (6) cellular mechanism. He [90] concluded
defects, e.g., central, altitudinal, or sectoral defects or as follows. (1) structurally crowded optic discs are
peripheral constriction. This is caused in most cases by predisposed; (2) laminar and retrolaminar regions are
temporal arteritis but may be due to arteriosclerotic the most common locations for infarction; (3) there is
occlusive disorder (Cullen, 1967 [89]). The pathogen- a flow impairment in the prelaminar optic disc during
esis of the pale oedema of the optic disc and the visual the acute phase; (4) lack of consistent choroid flow
loss in this condition has been debated by various impairment and retrolaminar location of infarcts sug-
authors. The basic lesion is the occlusion of the poste- gest vasculopathy within or distal to the paraoptic
rior ciliary arteries, which produces infarction not only branches of the posterior choroid arteries; (5) diabetes
of the disc but also of the retrobulbar part of the optic is the most consistently identified vasculopathic risk
nerve by virtue of their distribution to the prelaminar factor; (6) impaired autoregulation of the disc circula-
and lamina cribrosa regions of the optic disc and the tion by atherosclerosis, with possible contribution from
adjacent optic nerve. In patients with cilio-retinal arter- serotonin and endothelial-mediated vasospasm may
ies, the retinal area supplied by these vessels is also play a role; and (7) progression may be caused by sec-
seen to be involved. The infarction of the optic disc ondary cell death after the initial ischemic insult or
leads to the pale oedema. The area of the disc involved compression from cavernous degeneration and mecha
depends upon the extent of PCA involvement, i.e. the nical axonal distortion.
occlusion of the main PCA produces extensive infarc- On first glance, these conclusions seem quite valid.
tion of the disc, but involvement of one small branch Let us critically review the various parameters on
leads to a sectoral lesion. The lesion may progressively which the conclusions by Arnold [90] are based. To
base conclusions simply on a review of the literature is do not give us full information about invivo conditions
likely to involve some serious flaws, because of short- and the pathogenesis of typical NA-AION.
comings in the originally published studies. 2. Electron Microscopic Cast Studies Of the Optic
1. Histopathological Studies Nerve Head Vessels
These studies are in the vast majority of arteritic AION No doubt these casts are impressive to look at, but one
cases (see Chap. 12), and only a rare one of NA-AION must be extremely careful in applying findings from
of embolic nature, but almost none are of the most the postmortem casts to the in vivo circulation. We
common, typical NA-AION (see above). Infarction of have an excellent example of that flaw in the choroidal
the retrolaminar optic nerve head is described almost vascular bed. Since Federick Ruysch [92] in 1737 first
invariably in AION with occlusion of the PCA (due to described findings based on evaluation of casts of the
GCA or embolism). Cavernous degeneration is seen choroidal vascular bed, a large number of studies have
when there is infarction, as documented in eyes with evaluated corrosion casts of the choroidal and optic
arteritic AION or following experimental occlusion of disc vascular bed (examined both by light and electron
the PCAs [34]. As discussed above, no doubt histo- microscopy); without any exception, all of them have
pathological studies in arteritic AION have given valu- concluded that the entire choroidal and optic disc vas-
able information about the location of the ischemic cular bed, including the choriocapillaris, forms a freely
lesion in the optic nerve head when there is severe communicating and uninterrupted vascular bed, with
ischemia caused by complete occlusion of the PCA; interarterial and arteriovenous anastomoses. That has
however, the severity of ischemia is different between been the classical teaching for almost two and a half
the two types of AION severe in arteritic AION and centuries, because of those universal findings. However,
mild to moderate in NA-AION. Occlusion of the PCA my invivo experimental [9396] and clinical fluores-
in arteritic AION (see Figs. 3.24, 3.25, 3.283.34, cein angiographic studies in the early 1970s [31, 32,
3.36, 3.37, 3.43, 3.49 and 3.50), and also in an occa- 91, 97, 98], dealing with the PCAs and the choroidal
sional eye with emboli going to the PCA (see Figs.3.27 vascular bed, showed the old view to be totally invalid
and 3.35), is well documented on fluorescein fundus invivo, because angiographic studies showed clearly
angiography [91] and histopathological studies (see that the PCAs all the way up to the choriocapillaris
Chap. 12). By sharp contrast, fluorescein angiographic have a strictly segmental distribution, with no anasto-
studies in NA-AION during waking hours have clearly moses whatsoever between the adjacent segments at
shown NO similar occlusion of the PCA or its branches any level, i.e. the PCAs are end-arteries [99]. The
supplying the optic nerve head; there is simply a slow finding of that pattern, after almost two and a half cen-
and delayed filling of the vascular bed in the peripapil- turies, explained for the first time why inflammatory,
lary choroid and/or the choroidal watershed zone, and metastatic and degenerative diseases of the choroid are
of the optic disc [3133, 91] (Fig.14.3 above; also see localized in nature. That is an excellent example of the
Figs.3.39, 3.41, 3.45, 3.52, 3.54, 3.57, 3.58 and 3.59). utterly misleading nature of the cast findings as being
This is due to low perfusion pressure in them and that applied to the invivo situation. This shows that one has
would be aggravated by nocturnal arterial hypotension to be extremely cautious of applying the findings from
during sleep (the most common time when NA-AION the corrosion casts of the optic nerve head to invivo
develops [30]) in persons with predisposing risk fac- blood flow pattern.
tors. This results in mild to moderate ischemia of the 3. Optic Nerve Head Blood Flow Studies
optic nerve head in NA-AION. This differentiation is As discussed in detail in Chap. 6, dealing with mea-
extremely critical, because the severity of the ischemia surements of the optic nerve head blood flow, all the
determines the severity of the ischemic damage, and methods used so far to measure the optic nerve head
that in turn determines the extent of visual loss and blood flow have serious flaws, so that we have currently
visual recovery. This is very evident from the fact that NO method to measure the blood flow in the optic
in arteritic AION there is practically no visual improve- nerve head reliably [100]. So all the methods upon
ment (see Chap. 12) while in NA-AION about 40% which Arnold [90] based his conclusions are flawed to
show visual improvement (see Chap. 17). Thus, histo- provide reliable information. His argument that lack
pathological studies, apart from giving us the informa- of consistent choroid flow impairment and retrolami-
tion about the location of the ischemic lesions, really nar location of infarcts suggests vasculopathy within or
distal to the paraoptic branches of the posterior choroid e.g., development of NA-AION in migraine or in
arteries is flawed, because, as discussed above, in persons with atherosclerotic carotid artery disease
NA-AION, fluorescein angiography done during the associated with release of serotonin [102]. Athero
day usually shows delayed filling only in the peripapil- sclerotic carotid artery disease is not present in all
lary choroid and/or the choroidal watershed zone and patients with NA-AION.
the adjacent optic disc, but no evidence of complete 6. Cellular Mechanism
occlusion of the PCA [3133, 91]. We have no means Arnold [90] mentions a study where experimental
to determine the blood flow in the paraoptic branches optic nerve ischemia was associated with elevated lev-
of the PCA, which play an important role in the blood els of glutamate in the overlying vitreous [103]. Some
supply of the optic nerve head (see Chap. 3) nor to time ago, the presence of elevated glutamate was
determine the blood flow in the optic nerve head during claimed in experimental glaucoma eyes but that study
sleep, the time when NA-AION usually develops. was found to be flawed [104]. Kwon et al. [105], in
There is another serious flaw in measuring optic their study on the evaluation of vitreous glutamate
nerve head blood flow during waking hours. It has following experimental CRA occlusion in primates,
been demonstrated by 24-h ambulatory blood pressure found no elevation of glutamate at all in the vitreous,
monitoring that the extent of fall of blood pressure is in spite of massive ischemic cell damage caused by
directly proportional to the depth of sleep, so that blood CRA occlusion. That would totally contradict the find-
pressure during waking hours has little relevance to ing that experimental optic nerve ischemia causes ele-
that occurring during deep sleep [27, 28]; thus angio- vated levels of glutamate in the overlying vitreous.
graphic findings or any other mode of evaluation of the Arnold [90] states that progression may be caused
optic nerve head blood flow during waking hours gives by secondary cell death after the initial ischemic insult
us no reliable information at all about the state of cir- or compression from cavernous degeneration and
culation during sleep. mechanical axonal distortion. Unfortunately, we do
4. Various Risk Factors for Development not have any scientific evidence in support of this state-
of NA-AION ment. The mechanism of the progression of initial
NA-AION is a multifactorial disease. As discussed visual loss in NA-AION is discussed at length else-
below, one can divide these risk factors into two cate- where [106] (see Chap. 16); it does not support the
gories: (a) predisposing risk factors and (b) precipitat- view of Arnold [90].
ing risk factor(s). There is an invariable tendency to
blame one risk factor as the villain, which is totally In conclusion, one has to evaluate the theory by Arnold
mistaken in a multifactorial scenario. A person with [90] on the pathogenesis of NA-AION in the light of
only predisposing risk factors may go through life the above limitations and multiple problems.
without ever developing NA-AION. Moreover, we do
not have the technology and means yet to detect each
and every risk factor, as discussed below. One of the Pathogenesis of NA-AION Proposed
arguments often put forward by those who believe the by Levin and Danesh-Meyer in 2008
pathogenesis of NA-AION is not yet known is that
some patients with NA-AION are apparently perfectly Levin and Danesh-Meyer [107] put forward a new
healthy and have no known cardiovascular risk factors hypothesis about the pathogenesis of NA-AION. They
and yet develop NA-AION the fallacy of this con- hypothesized that obstruction of venous outflow,
cept is discussed below. rather than arterial hypoperfusion, is the causative
5. Vasospasm and Impaired Autoregulation mechanism of NAION. Vasodilating arteries second-
As discussed in Chap. 5, impaired autoregulation arily block venous outflow in a tight compartment and
seems to play an important role but as yet we have no thereby cause venous engorgement. In the anterior
means to evaluate that in the human. We did assess it optic nerve head, the central retinal artery and vein
in rhesus monkeys [101] but that is not possible in the share a common adventitial sheath. In a disc at risk,
human. Whether, and (if so) how much vasospasm prolonged arterial dilation within the fibrous sheath
plays a role in development of NA-AION is not clear, would decrease optic nerve drainage to the central
except that it is theoretically possible in some cases, retinal vein via its tributaries, resulting in optic nerve
head engorgement, hyperemia, and, ultimately, NAION. To explain their hypothesis, they [107] apply the
According to them [107]: (1) NAION (non-arteritic concept of cytotoxic and vasogenic edema from
AION) often does not have clinical characteristics of the brain to the optic nerve. This is completely
an arterial disease. (2) NAION does not have patho invalid for several reasons. The morphology, vascu-
physiological characteristics of an arterial disease. (3) lar pattern and many other basic aspects of the optic
Risk factors for NAION are not specific to arterial nerve are totally different from those of the brain.
disease. They hypothesized that in NA-AION in For example, unlike the brain, in the optic nerve
many cases the ischemia results from venous conges- behind the prelaminar region, the blood vessels lie
tion. They have put forward several speculations in in the fibrous septa of the optic nerve, separated
support of their hypothesis, discussed below. from the nerve fiber bundles. There is no grey mat-
I [108] find that the hypothesis by Levin and ter in the optic nerve. The so-called congestion of
Danesh-Meyer [107] is based more on arm-chair the venous tributaries in the septa is incapable of
philosophy than well-established scientific facts causing edema or ischemia of the axons in the nerve
dealing with NA-AION and central retinal vein fiber bundles and optic disc edema. Thus, to apply
occlusion (CRVO). My views, on the contrary, are automatically findings from the brain to the optic
based on my extensive, multifaceted studies on the nerve in this case is to ignore the known facts.
blood supply of the optic nerve conducted since 2. The authors [107] stated that there is occlusion of
1955 (see Chap. 3), on the anatomy of the optic CRV tributaries. They [107] further state that the
nerve (see Chap. 2), and experimental and prospec- clinical picture of CRV occlusion probably reflects
tive clinical studies on various aspects of NA-AION occlusion of the CRV anteriorly and resultant hem-
(on more than a 1,000 patients) and CRVO (more orrhage and leakage of capillaries within the retina
than a 1,000 patients). that drain into retinal veins. My prospective stud-
Levin and Danesh-Meyer [107] stated that ies on central retinal vein occlusion absolutely con-
NA-AION is not an arterial disease, in that (they said) tradict this statement. Almost all central retinal vein
it does not have the clinical and pathophysiological occlusion eyes develop optic disc edema, but in my
characteristics of an arterial disease, nor risk factors study of more than a 1,000 eyes with CRVO none
specific to arterial disease. But they have overlooked developed NA-AION attributable to CRVO. If the
some well-established, fundamental facts about the hypothesis of the authors were valid, the vast major-
clinical, pathophysiological and risk factors involved ity should have suffered associated NA-AION.
in the pathogenesis of NA-AION, confirmed time and 3. They [107] stated that Increased venous pressure
again by a whole host of studies in the past (discussed from sleep apnea or recumbency during sleep may
at length elsewhere [32, 33].). It seems the authors contribute to a cascade of events leading to
[107] have a poor understanding of the basic facts NA-AION. It seems the authors are not aware that
about the anatomy of the optic nerve, vascular and in obstructive sleep apnea there is actually an
venous system of the optic nerve, and pathogenesis increase in sympathetic activity, vasoconstriction
and clinical features of CRVO and NA-AION, which and increase in blood pressure during sleep [109].
have misled them to postulate a hypothesis with no 4. According to them [107], Vasodilation of the CRA
scientific validity at all. can compress the CRV and its tributaries within
Very briefly, there is no scientifically valid evidence their shared sheath. This mechanism would explain
in support of any of the following main statements by NAION in the setting of prolonged hypotension,
the authors: nocturnal hypotension, shock, or phosphodi-
esterase-5 inhibitor erectile dysfunction drugs, all
1. They [107] stated: The degree of visual impairment of which are characterized by arterial dilation.
of NAION shows a similar pattern to venous occlu- There is no evidence whatsoever that there is dila-
sive disease of the central nervous system. Based tion of the CRA in these conditions. It seems the
on my basic, experimental and clinical (in about authors are not aware of the various factors that
1,500 cases) studies dealing with retinal venous influence the blood flow in the optic nerve head
occlusive disease and associated changes in the optic (see Chap. 5) and the role arterial hypotension plays
nerve, I find that statement without foundation. in that [110].
5. They [107] contend that Decreased venous out- a significant increase in retinal hemorrhages in
flow within the anterior optic nerve would not be persons on these therapies, and that usually causes
expected to cause the hemorrhagic retinal findings destruction of the retinal tissue and consequently
of CRVO or delayed retinal venous filling because poorer visual outcome [115]. Thus, use of these
there are significant collateral venous drainage agents for treatment of retinal vein occlusion per
pathways from the retina. All of my central retinal se is contraindicated [115118].
vein occlusion studies contradict this statement. (b) There is a common perception among ophthalmol-
6. They [107] state that Cilioretinal artery occlusion, ogists and neurologists that NA-AION and cere-
which often occurs in conjunction with CRVO, may bral stroke are similar in nature pathogenetically
be the result of a congested CRV causing secondary and in management. This has resulted in major
cilioretinal artery compression. This is an invalid controversy on the pathogenesis and management
statement. In our study [111] of 38 such eyes, we of NA-AION. The following evidence, however,
discussed fully the mechanism of cilioretinal artery indicates that NA-AION, when compared to cere-
occlusion in central retinal vein occlusion. bral stroke, is pathogenetically a distinct clinical
7. Levin and Danesh-Meyer [107] reasoned that entity [33, 119].
Although diabetes predisposes to both arterial and (i) Difference in Association with Smoking:
venous disease, it also is associated with capillary There is a huge volume of literature showing
endothelial leakage and would be expected to produce a significant association between smoking
more vasogenic edema for a given amount of venous and cerebrovascular accident [120] (a throm-
pressure elevation. A study dealing with 931 eyes boembolic disorder). Two large prospective
with NA-AION (306 in diabetics, and 625 in non- studies have shown no association between
diabetics) lends no support to this speculation [112]. smoking and NA-AION [121, 122].
8. Their [107] explanation for development of incipi- (ii) Difference In Response To Aspirin: While
ent NA-AION is not valid. The pathogenesis of this the beneficial effect of aspirin in cerebrovas-
condition is discussed at length elsewhere [113] and cular accident (usually a thromboembolic
in Chap. 15, which contradicts their statement. disorder) is well-established, NA-AION stud-
ies have shown that aspirin has no beneficial
It seems the authors [107] arguments are primarily
effects in NA-AION [119, 121, 123, 124]
based on their concept of what happens in the brain.
(being a hypotensive disorder).
In Medicine, there has been a frequent tendency to
(iii) Difference in Association with Thrombophilic
carry over findings from disease of one organ to
Risk Factors: While an association has been
explain pathogenesis and management in another
reported between thrombophilic risk factors
organ, irrespective of the fact that there are morpho-
and cerebrovascular accident, no such asso-
logical, physiological, pathogenetic and other major
ciation has been found between NA-AION
differences between the two; this is true of the optic
and thrombophilic risk factors, for the same
nerve and brain as well. We have several examples
reason [125128].
where applying findings from other organs has resulted
(iv) A Hypotensive Disorder: As discussed below,
in misconceptions and mismanagement of diseases.
our findings show that NA-AION is primarily
Following are just two examples in ophthalmology:
a hypotensive disorder and not a thromboem-
(a) In the past it was taken for granted that the patho- bolic disorder, in the vast majority of cases.
genesis of central retinal vein occlusion was simi- (v) Seasonal Differences in Onset of NA-AION
lar to that of deep vein thrombosis in the leg, and and Stroke: As discussed below, there is
therefore anticoagulants or antiplatelet therapy a distinct difference in the seasonal varia-
was the treatment of choice in retinal venous tion in the onset of NA-AION and vascular
diseases. This ignored the basic fact that the mor- accidents, such as myocardial infarction and
phology of retinal veins, retinal tissue and the cerebrovascular accident. Our study [30]
pathogenesis of retinal venous occlusion [114] are showed that NA-AION onset was signifi-
totally different from those of deep vein thrombo- cantly more frequent in the Iowa (USA)
sis and the calf tissues an eye is not a leg! In my Summer (or the hot months, with temperature
studies on eyes with retinal vein occlusion, I found in the 60s to low 90 Fahrenheit) than in the
Winter (or the cold months, when the maxi- central retinal vein results in a marked rise of
mum temperature is usually much below intraluminal pressure in the entire retinal capillary
freezing) (p=.0030). This is in contrast to the bed; when that intraluminal pressure rises above
reported incidence of systemic vascular acci- that in the cilioretinal artery, the result is a hemody-
dents, such as myocardial infarction and cere- namic block in the cilioretinal artery. [111]. This is
brovascular accidents which are higher during like comparing apples and oranges because cilioret-
winter than other seasons. Thus, there is a inal artery occlusion with central retinal vein occlu-
distinct difference in the seasonal variation in sion is a totally different clinical entity from what
onset of NA-AION and systemic vascular they discussed in their hypothesis.
accidents. Moreover, in our study [129] there 2. Levin [131] rhetorically asked, Does he (Dr.
was no evidence that NA-AION patients are Hayreh) seriously doubt that diabetes is a risk factor
at increased risk of developing subsequent for NAION at a young age, given the wide evidence
cerebrovascular disease, myocardial infarc- to the contrary? Hardly, in fact I [132] was the first
tion or mortality. to describe clinical findings of NA-AION in young
diabetics, and more recently published a large series
This shows that it is serious mistake to equate
[112] showing that. He [131] has taken my com-
NA-AION with stroke and other systemic vascular
ments totally out of context.
accidents.
3. Levin [131] stated that attempts to produce NAION
This dangerous practice of explaining pathogenesis
by posterior ciliary occlusion have failed to estab-
and recommending management by extrapolating find-
lish arterial hypoperfusion as the underlying
ings from one organ to another has the potential to
pathophysiology. The published photograph of the
introduce serious errors, as is the hypothesis postulated
optic nerve head from Dr Hayrehs 1972 experi-
by Levin and Danesh-Meyer [107].
mental occlusion of the PCAs demonstrates pallid
In conclusion, venous occlusion plays no role
swelling of the optic nerve head more similar to
whatsoever in development of NA-AION. Their [107]
arteritic ischemic optic neuropathy than NAION.
comparison of arteritic and nonarteritic AION in sup-
This result supports our reasoning that from a clini-
port of their hypothesis is misleading on several issues.
cal point of view, NAION has more similarities
NA-AION is a multifactorial disease caused by arte-
with venous infarction than arterial infarction. This
rial insufficiency to the optic nerve head due to a
statement is seriously inaccurate and demonstrates
combination of various systemic and local risk factors
a profound lack of basic knowledge of the (admit-
[32, 33]. The hypothesis by Levin and Danesh-Meyer
tedly complex) subject. This is because my [34]
[107] introduces an invalid concept on the subject of
experimental occlusion of PCA in rhesus monkeys
pathogenesis of NA-AION. Without solid facts, an
(cited by Levin) produced ischemic optic neuropa-
hypothesis is simply a speculation. Thomas Henry
thy by cutting off the blood supply to the optic nerve
Huxley said: The great tragedy of Science the slay-
head i.e. that produced ischemic damage. For the
ing of a beautiful hypothesis by an ugly fact [130].
first time, that study showed that the optic nerve
The hypothesis must be backed up by all the facts, and
head was indeed supplied by the PCA, confirming
in the case of the hypothesis by Levin and Danesh-
my previous anatomical studies, and that its occlu-
Meyer [107], it is not.
sion causes its ischemia. There is ischemia in both
In responding to my critique of their hypothesis
arteritic and non-arteritic AION it is simply the
[108], Levin [131] cited my previous publications to
severity of ischemia which is different between the
refute my criticism of their hypothesis. Unfortunately,
two types of ischemic optic neuropathy (as discussed
he made irrelevant and misleading statements about
above). In many of my publications, based on fluo-
the subject under discussion and cited my paper out of
rescein angiographic studies of NA-AION and
context. Following are some examples of that.
A-AION patients, 24-h ambulatory blood pressure
1. He [131] stated that I [108] doubt our suggestion monitoring studies and clinical studies of NA-AION
that congestion from CRVO could cause secondary and A-AION, I [32, 33] have stressed repeatedly
ciliary retinal artery occlusion. Yet his theory is that in NA-AION there is a transient non-perfusion
similar, only differing in the location of the block or hypoperfusion of the optic nerve head circulation
due to the congested vein: Sudden occlusion of the due to nocturnal arterial hypotension since visual loss
is first discovered on waking in the morning or first of blood supply is from the PCA circulation, as
opportunity to use central vision [30], and not discussed in Chap. 3 (Fig.14.1). Therefore, NA-AION
occlusion of the PCA (Fig. 14.3 above; also see represents an ischemic disorder of PCA circulation in
Figs. 3.39, 3.41, 3.45, 3.52, 3.54, 3.57, 3.58 and the optic nerve head. As discussed in Chap. 3, the optic
3.59), while in A-AION there is occlusion of the nerve head has marked inter-individual variation in its
PCA (see Figs. 3.24, 3.25, 3.283.34, 3.36, 3.37, blood supply and blood flow patterns, and the optic
3.43, 3.49 and 3.50). nerve head blood flow is influenced by a whole host of
4. Levin [131] stated that Dr Hayreh asserts that factors (see Chap. 5); all those exercise a profound
because the optic nerve is different from the brain, influence on the pathogenesis and clinical features of
having pial septa but not gray matter, it cannot have NA-NAION.
cytotoxic edema, vasogenic edema, or venous con- Three misconceptions concerning NA-AION are
gestion. We agree that the optic nerve is different being perpetuated:
from the brain and does not contain neuronal cell
bodies. But neither does the liver. Again to cite the 1. That NA-AION is presumably due to ischemia
liver when considering the optic nerve is like [136]. With the currently available information, based
equating apples and orange because the anatomy, on clinical and experimental findings, we can now
physiology, blood supply, circulation and everything say with confidence that all the available evidence
else of the liver, brain and optic nerve have nothing shows that it is definitely an ischemic disorder.
in common at all. It makes no sense whatsoever to 2. That NA-AION, like arteritic AION is due to com-
state that the two would behave in similar fashion. plete occlusion of the PCA. As discussed above and
below, there is no evidence at all of complete occlu-
This clearly shows that their hypothesis has no scien- sion of the PCA in NA-AION in the vast majority
tific basis, but is based on armchair philosophy and of cases [32, 33, 91]. There is only delayed filling of
inadequate knowledge of a highly complex subject. It the peripapillary choroid and/or the choroidal
is unfortunate that such unscientific hypotheses result watershed zone and the optic disc on fluorescein
in unnecessary confusion and controversy and do not fundus angiography (Fig.14.3; also see Figs.3.39,
serve the cause of Science. 3.41, 3.45, 3.52, 3.54, 3.57, 3.58 and 3.59).
3. That AION represents acute ischemia of only the
prelaminar region [136]. Histopathological studies on
the optic nerves in arteritic [14, 15, 137] and non-
Multifactorial Pathogenesis arteritic [138] AION, as well as in experimentally pro-
of Non-arteritic AION duced AION in monkeys [34, 135, 139], show definite
involvement of a variable length of the retrolaminar
I am going to discuss this at length. In 1974, I [133] first region (depending upon the extent of supply by the
discussed the multifactorial pathogenesis of NA-AION, PCA in this region), in addition to the prelaminar and
based on my studies on the blood supply of the optic laminar regions of the optic nerve (Fig.14.1).
nerve head, and experimental and clinical studies at that Following is a detailed discussion of the various
time. Since then, a tremendous amount of additional aspects of the highly complex pathogenesis of
information has emerged. The concept of the pathogen- NA-AION, based on the currently available evidence.
esis of NA-AION presented below is based on the cumu-
lative information provided by various studies reported
in the literature on NA-AION, as well as on various
studies conducted by me on the following subjects:
Mechanism of Vascular Insufficiency
(a) arterial blood supply of the optic nerve head, (b)
experimental studies in rhesus monkeys, and (c) multi- in the Optic Nerve Head in NA-AION
faceted clinical studies in a large cohort of NA-AION
patients seen in my clinic over a period of 40years. First of all, it is essential to understand the basic mech-
NA-AION is due to acute ischemia of the optic anism of vascular insufficiency in the optic nerve head
nerve head [31, 34, 91, 133135], whose main source in NA-AION. As discussed in Chap. 3, the optic nerve
Mechanism of Vascular Insufficiency in the Optic Nerve Head in NA-AION 279
head is primarily supplied by the PCA circulation. In

NA-AION, unlike arteritic AION (see Chap. 12), it is
important to understand that it is not essential for the
PCAs to be completely occluded to produce NA-AlON.
In fact, in the majority of eyes there is never a com-
plete occlusion of the PCAs (Fig. 14.3; also see
Figs.3.39, 3.41, 3.45, 3.52, 3.54 and 3.573.59) [32,
33, 91, 99, 140].
To understand the mechanism of vascular insuffi-
ciency in the optic nerve head in NA-AION, one has to
consider the factors which influence the blood flow in
the optic nerve head, discussed at length elsewhere
[110, 141] and in Chap. 5. Briefly, circulation in the
optic nerve head depends upon the perfusion pressure
and vascular resistance in its vessels. The perfusion
pressure depends upon the difference between the
intraocular pressure and mean blood pressure in the
PCAs [142]. Thus, an imbalance in the perfusion pres-
sure can be produced by either lowering the blood
pressure or raising the intraocular pressure or a combi-
nation of both. On lowering the perfusion pressure in Fig.14.5 Fluorescein fundus angiogram of a cynomolgus mon-
the intraocular blood vessels, the susceptibility of the key (after experimental central retinal artery occlusion to unmask
PCA supply in the optic disc and choroid) at 70mmHg intraocu-
various intraocular blood vessels to obliteration varies lar pressure. It shows selective complete obliteration of the peri-
considerably: papillary choroidal and prelaminar vessels (Reproduced from
Hayreh etal. [142])
(a) The vessels in the prelaminar part of the optic
nerve head: These are most susceptible to oblitera-
tion and are usually the first to be affected. than the perfusion pressure in the CRA. Thus, the
(b) The peripapillary choroid: This is either equally retinal circulation is the last to be obliterated.
or slightly less susceptible to obliteration In the light of these observations, it is evident that once
(Fig.14.5). The obliteration of the peripapillary the perfusion pressure in the PCAs falls, the optic
choroid itself would in turn involve the blood nerve head circulation is the first to be involved, then
supply to the prelaminar region of the disc and the peripapillary choroid and the watershed zone, and
retrolaminar optic nerve as well (Fig.14.1), par- the rest of the choroid last of all. In these cases the
ticularly in arteriosclerotic individuals. The chor- PCAs may still be patent and showing normal filling of
oidal watershed zones [98] (between the territories the choroid in spite of poor circulation in the peripapil-
of distribution of the medial and lateral PCAs) lary choroid, watershed zone and optic nerve head
behave like the peripapillary choroid in the event (Fig.14.3; also see Figs.3.10, 3.383.41, 3.45, 3.52,
of a fall of perfusion pressure in the PCA circula- 3.543.59); that has misled people into believing that
tion. When the watershed zone is related to the in NA-AION, because of normal filling of the choroi-
optic disc [98], that results in poor or absent fill- dal vascular bed, there is no reason why vascular insuf-
ing of the adjacent optic disc also (see Figs.3.10, ficiency in the PCA circulation has any role to play in
3.383.41, 3.44, 3.45, 3.51b and 3.523.59). the pathogenesis of NA-AION, unless the PCA is com-
(c) The rest of the choroidal circulation: This is also pletely occluded. As mentioned above, the peripapil-
susceptible to obliteration but much less so than (a) lary choroid and choroidal watershed zones [98] are
and (b). most susceptible to obliteration when the perfusion
(d) The retinal circulation: Because of the presence pressure falls in the PCA circulation, and the peripapil-
of autoregulation in it, it does not show any oblit- lary choroid plays an important role in the blood sup-
eration unless the intraocular pressure is higher ply of the optic nerve head (see Chap. 3). Thus, the
determining factor for the production of NA-AION is nocturnal arterial hypotension) was revealed by our
the level of the perfusion pressure in the PCAs and 24-h ambulatory blood pressure monitoring studies
not the absence or presence of complete occlusion of [27, 28, 30, 32, 33]. It is further confirmed by the fact
the PCA. This is demonstrated very well by fluores- that the vast majority of NA-AION patients first notice
cein fundus angiography which provides most useful visual loss on waking up in the morning [30]; so that
information about the circulation in NA-AION. In it nocturnal arterial hypotension is an important precipi-
fluorescein fundus angiography usually reveals absent tating factor. Drusen in the optic nerve head could pro-
or delayed filling of the optic disc, watershed zones duce NA-AION by direct pressure on the vessels.
and peripapillary choroid, with almost normal choroi- NA-AION may start as a sectoral change (due to
dal circulation [32, 33, 91, 99, 140] (Fig.14.3; also see segmental optic nerve head ischemia), but may soon
Figs.3.10, 3.383.41, 3.45, 3.52, 3.543.59). progress to involve more of the optic nerve head; this
The other very important factors to consider, when is because of the following two factors: (a) In eyes
interpreting filling of the PCA circulation on fluores- with NA-AION, there is usually no optic disc cup,
cein angiography in these cases, are: which results in limited space in the prelaminar region
and crowding of the axons. (b) In NA-AION there is
1. The time-lapse between the onset of NA-AION
axonal swelling, which produces segmental optic disc
and fluorescein angiographic examination: This
edema. A combination of these two factors must com-
is because, even in eyes with complete absence of
press the small blood vessels in the parts adjoining the
filling of the PCA circulation at the onset of the dis-
edematous part in a crowded situation, resulting in
ease, a slow restoration of circulation starts to
ischemia of those parts of the optic nerve head.
occurs fairly soon, so that a significant filling defect
Thus, NA-AION is a manifestation of ischemia of
is no longer seen after a few days or weeks [93, 134,
the optic nerve head caused by interference with its
140]. Thus, the fact that no filling defect appears in
circulation, as a result of multiple factors influencing
the PCA circulation on fluorescein angiography
the circulation (see Chap. 5). It is the complexity of
does not prove that there has been no earlier filling
this mechanism which is responsible for the confusion
defect.
and controversy about the role of the PCA circulation
2. Fluorescein angiography performed during wak-
in the development of NA-AION.
ing hours: The other confounding factor, as dis-
cussed above, is that fluorescein angiography is
performed during waking hours when the blood
pressure is normal, but NA-AION usually develops
during sleep [30] when there is nocturnal arterial Multifactorial Pathogenesis of NA-AION
hypotension which results in a fall of perfusion
pressure. Thus, angiography during waking hours Etiologically and pathogenetically, NA-AION cases
does not give correct information about the state of can be broadly classified into two groups [31]:
circulation during deep sleep.
1. Due to Embolic Lesions of the Arteries/Arterioles
These are serious limitations in the usefulness of fluo- Feeding the Optic Nerve Head
rescein angiography for evaluating the optic nerve Embolic occlusion of the PCAs (see Figs. 3.27 and
head circulation; unfortunately, the vast majority of 3.35) or of the optic nerve head arterioles seems to
ophthalmologists are not aware of them. It is still, occur only occasionally, but this impression may be
sadly, almost invariably stated in the literature that erroneous because of our inability to see the emboli in
NA-AION is due to occlusion of the PCA. This is these vessels on ophthalmoscopy, as compared with the
simply inaccurate. ease with which we can see emboli in the retinal arteri-
This imbalance between the blood pressure in the oles. Only fluorescein angiography can help in that
PCA circulation and intraocular pressure was also determination [91]. As a routine I perform fluorescein
pointed out by Foulds [63, 143]. The fall in perfusion angiography in all new NA-AION and arteritic AION
pressure in the optic nerve capillaries may be due to patients, so that I have done that in more than a 1,000
several causes, local, systemic or ocular. The role cases during the past 40years; while in arteritic AION
played by systemic arterial hypotension (particularly I have invariably found evidence of occlusion the PCA
Multifactorial Pathogenesis of NA-AION 281
(see Figs.3.24, 3.25, 3.283.34, 3.36, 3.37, 3.43, 3.49 embolic type, by contrast, fluorescein angiography,
and 3.50), and I have found that only rarely in NA-AION during the first few days after the onset of NA-AION,
(see Figs.3.27 and 3.35). Multiple emboli in the ves- shows gross filling defects in the deep vessels of the
sels of the anterior part of the optic nerve have been optic disc and adjacent choroid, depending on the dis-
demonstrated histopathologically in NA-AION [4]. tribution pattern of the occluded artery [31, 32, 91, 98]
2. Due to Transient Nonperfusion or Hypoperfusion (Figs.3.27 and 3.35).
of the Nutrient Vessels in the Optic Nerve Head
As discussed above (see Chap. 5), the blood flow in the
optic nerve head depends upon the perfusion pressure
in its vessels. A transient nonperfusion or hypoperfu- Risk Factors for Development of NA-AION
sion of the optic nerve head can occur due to a tran-
sient fall of perfusion pressure in its vessels, which in All the available evidence indicates that NA-AION is
turn, in susceptible persons, produces NA-AION. It is multifactorial in nature and its many risk factors can
extremely important to remember that in this mode of be divided into two main categories:
development of NA-AION there is NO actual occlu-
sion of the PCA (Fig.14.3; also see Figs.3.39, 3.41, 1. Predisposing Factors: These may be systemic or
3.45, 3.52, 3.54, 3.57, 3.58 and 3.59). Almost all local in the eye and/or optic nerve head (see
NA-AION cases belong to this group. In view of this, below), and they make the optic nerve head suscep-
the common assertion in the literature that PCA occlu- tible to ischemic disorders.
sion causes NA-AION is simply not correct. 2. Precipitating Factor(s): These act as the final
insult (last straw), resulting in ischemia of the
A fall of the perfusion pressure below the critical level optic nerve head and NA-AION. Nocturnal arterial
in the capillaries of the optic nerve head may be caused hypotension is an important factor in this category
either by a marked fall in mean blood pressure, e.g., in [27, 28, 30].
shock, nocturnal hypotension [27, 28] during sleep at
night or a nap during the day, severe internal carotid
artery and/or ophthalmic artery stenosis or occlusion Serious Difficulties in Determining Risk Factors
[144, 145], and other causes, or by a marked rise in the for NA-AION
intraocular pressure [146] (as in acute angle closure
glaucoma); or by a combination of the two (e.g., in Controversy on the pathogenesis of NA-AION has
neovascular glaucoma associated with ocular ischemia arisen largely because of the difficulty in determining
due to internal carotid artery [144, 145] and/or oph- these risk factors (discussed elsewhere [147]). A com-
thalmic artery stenosis or occlusion). The optic nerve mon argument put forward by those who believe that
head damage, which may be mild to marked, depends the pathogenesis of NA-AION is not known, or who
upon the severity and the duration of the transient isch- even question that it is ischemic in nature, is that some
emia, but is usually less extensive and less severe than patients with NA-AION are apparently perfectly
in thrombotic/embolic cases. Available clinical evi- healthy, have no complaint of any kind, and routine
dence indicates that certain segments of the optic nerve physical examination (including electrocardiography
head (e.g., the upper part) are more vulnerable to isch- done in the clinic) show no abnormality, and have no
emia than others. Since the ischemia caused by hypop- known risk factor. Therefore, it is argued how can one
erfusion is mostly due to transient non-perfusion, explain the development of NA-AION in such per-
usually during sleep, without an organic block in the sons? I have heard this argument repeatedly during
vessels, fluorescein fundus angiography done during discussions on NA-AION. In view of that, they argue
the day may reveal minimal or no filling defect or that surely there must be some other mechanism rather
delay in filling in the optic disc and/or peripapillary than vascular disturbance responsible for the develop-
choroid or choroidal watershed zone, either through- ment of NA-AION? This has led to controversy and in
out its vertical length (Fig. 14.3; also see Figs. 3.39, postulating new and unsupported theories about the
3.41, 3.52, 3.54, 3.57 and 3.58) or only in one-half of pathogenesis of NA-AION (as discussed above). But
it (see Figs. 3.45, 3.52, 3.54, 3.573.59). In the unfortunately, it is not always possible to detect all
risk factors in patients with NA-AION, for several interest, the cardiologist felt I should wear a Holter
reasons; following are some examples: monitor for a few days to find more about that. Since
I was totally asymptomatic, feeling fine and per-
1. A totally asymptomatic person, who looks perfectly fectly normal, and working about 80h a week on my
healthy and feels fine and whose routine physical research and clinical work, I was reluctant to go
evaluation does not reveal any systemic, ocular or through that hassle of Holter monitoring; however
optic nerve head abnormality is not necessarily with- finally I went along with the advice of the cardiolo-
out risk factors. A recent experience of my own is gist. It was a great surprise when next day I was
instructive here [147]. I made an appointment totally asked to go to the cardiology clinic urgently because
casually to see a cardiologist just to find, out of curi- the electrocardiogram on the Holter monitor that
osity, about my entirely asymptomatic fluctuation of night showed that I was having, every few minutes,
heart rate, some days in 70s and other days as high recurrent and prolonged periods of complete asys-
as 100/min; i happened to have noticed that for some tole during sleep, some lasting as long as about 10s
time on occasional feeling of my pulse but I was (with a flat electrocardiogram Fig. 14.6); occa-
totally asymptomatic. Most persons would consider sionally the asystoles were also recorded during the
that of no consequence since pulse rate fluctuates day. That required urgent pacemaker implantation
from time to time in all persons. Out of academic to prevent me from going into cardiac arrest. Had I
Fig.14.6 Three segments of electrocardiogram recorded by Holter monitor (at 12.50AM, 2.07AM and 2.10AM) showing variable
periods of cardiac asystole. In the segment at 2.07AM, cardiac asystole lasted for 9.8s with flat electrocardiogram
Ambulatory Blood Pressure Date

280
260
240
220
200
Heart Rate (BPM) Blood Pressure (mmHg)
180
160
140
120
100
80
60
40
20
200
180
160
140
120
100
80
60
40
20
10 12 14 16 18 20 22 0 2 4 6 8 10
Time (hour)
Fig. 14.7 Ambulatory blood pressure and heart rate monitoring sure is ideal; however, the systolic blood pressure dropped from
records (based on individual readings) over a 24-h period, starting about 140mmHg during the day to 90mmHg on going to sleep,
from about 11AM, in a 58-year old woman with bilateral NA-AION, and diastolic blood pressure dropped from about 80mmHg to low
and on no medication. This shows that during the day blood pres- 40s during sleep. (Reproduced from Hayreh etal. [28])
not made an appointment to see a cardiologist out of cases it dropped markedly during sleep to very low
simple curiosity, and had he not ordered the Holter levels, for no apparent reason, even in persons who
monitor for asymptomatic heart rate fluctuation, are not on any medication [27, 28]. For example,
which could normally have been considered of no the blood pressure recording in Fig.14.7 is from a
significance, it is quite possible that I would have (1) patient who had NA-AION first in one eye and then
died during sleep from cardiac arrest, (2) developed in the second eye and was not on any medication.
stroke, or (3). NA-AION. Everyone would have This shows that during the day the blood pressure
wondered how it could have happened to a perfectly was ideal; however, the systolic blood pressure
healthy and active person like me. This is also shown dropped from about 140mmHg during the day to
by the fact that from time to time athletes suddenly 90 mmHg on going to sleep, and diastolic blood
die of cardiac arrest while playing. That was an pressure dropped from about 80mmHg to low 40s
extremely important lesson about unknown risk fac- during sleep. Thus, daytime blood pressure is no
tors. It shows that being to all appearances perfectly guide at all to the nighttime blood pressure. That
healthy does not preclude the possibility of having has misled everyone into believing that NA-AION
unknown serious risk factors. patients may have normal or high blood pressure
2. In my 24-h ambulatory BP recording in about 700 but never any arterial hypotension. That is why
patients, I found time and again that though there many persons believe nocturnal arterial hypoten-
was no hypotension during waking hours, in some sion plays no role in the development of NA-AION
(see below, the role of nocturnal arterial hypoten- our findings of systemic diseases associated with
sion in NA-AION). NA-AION in a large prospective study, based on 406
3. It has been suspected that derangement of autoregu- patients with NA-AION, systematically investigated
lation of the optic nerve head blood flow plays an for that. It gave a detailed prevalence of arterial hyper-
important role in various types of circulatory disor- tension, diabetes mellitus, and various types of
ders of the optic nerve head (See Chap. 5). Our cardiovascular, endocrinal, cerebrovascular, renal,
study [101] in atherosclerotic and hypertensive rhe- rheumatologic, infectious, and other diseases and
sus monkeys showed marked derangement of auto- hematologic abnormalities, according to three age
regulation in the optic nerve head, but that required groups (young <45 years, middle-aged 4564 years,
enucleating the eyes and optic nerves, and doing elderly65years, and total number). Compared with
elaborate autoradiographic studies done by experts the prevalence reported in the general population,
in the field in Sweden. That is obviously not possi- young (<45 years), middle-aged (4564 years) and
ble in the human. Thus, we do not have the technol- elderly (65years) patients with NA-AION showed a
ogy currently to detect this and many other risk significantly higher prevalence of arterial hypertension
factors; they may be suspected, but not confirmed. (p<0.02), diabetes mellitus (p<0.01), and gastro-
4. Evaluation of the optic nerve head blood flow is intestinal ulcer (p<0.02). Also, middle-aged and
crucial in understanding the pathogenesis and man- elderly patients showed a significantly higher preva-
agement of various ischemic disorders of the optic lence of ischemic heart disease (p<0.01) and thyroid
nerve head, including NA-AION and glaucomatous disease (p<0.01). Middle-aged patients exhibited sig-
optic neuropathy. This is similar to the evaluation of nificantly higher rates of chronic obstructive pulmo-
intraocular pressure in the management of glau- nary disease and cerebrovascular disease (p<0.01). In
coma. As discussed in Chap. 6, in spite of enthusi- our study, NA-AION patients with both arterial hyper-
astic claims made by some advocates of various tension and diabetes mellitus had a significantly higher
methods to measure the optic nerve head blood incidence of cerebrovascular disease (p<0.01). Of the
flow, the hard fact is that we have no method as yet 406 patients in this study, 10% reported no previous or
that provides us with reliable information about current systemic disease at first visit to our clinic; 32%
that very important factor which plays a vital role of the patients had no major systemic disease, includ-
in the development of NA-AION [100]. ing arterial hypertension, diabetes mellitus, cardiovas-
cular disease, cerebrovascular disease, or migraine. Of
These are some examples of serious pitfalls in deter-
the individuals in the young, middle-aged and elderly
mining risk factors. Thus, the axiom should be that if
groups, 77%, 88% and 95% respectively reported one
we do not find any evident risk factor, that does not
or more previous or current systemic disease.
necessarily mean that patient has no risk factor hid-
A review of the various studies dealing with the
ing in his body.
prevalence of systemic diseases in NA-AION shows
that the differences in prevalence of various systemic
diseases among different studies depend upon the
sample size, referral pattern and study design. For
Known Risk Factors example, Repka etal. [150] (Table14.1) excluded all
young patients, presumably under the impression that
A tremendous amount of literature has accumulated on NA-AION did not develop in young persons (which
the subject. Following is a brief summary. we now know is not correct [32, 33, 112, 122, 129]).
The systemic diseases which showed a significant
association in our study [129] fall into the following
three classes:
Systemic Diseases Associated with NA-AION
and Risk Factors 1. The diseases which produce vascular changes in
the optic nerve head vessels must have a role to
Table 14.1 summarizes the data on various systemic play in the pathogenesis of non-arteritic NA-AION.
diseases in patients with NA-AION reported in the lit- These include arterial hypertension, diabetes mel-
erature from 1969 to 1990. In 1994, we [129] reported litus, hyperlipidemia, ischemic heart and other
Table14.1 Systemic diseases in patients with non-arteritic anterior ischemic optic neuropathy (Reproduced from Hayreh etal. [129])
Foulds etal. Ellenberger Eagling Boghen Repka etal. Guyer etal. Moro etal. [152] Guiffre
[143] etal. [148] etal. [149] etal. [86] [150] [151]a [153]
Year published 1969 1973 1974 1975 1983 1985 1989 1990
No. of patients 24 45 33 37 169 200 81 44
Study period 19611972 19621972 19691982 19751985 19751986 19791988
Multifactorial Pathogenesis of NA-AION
Age (years) 41 35 >40 45 1790 3387 >50

No. (%) with 5 (21) 7 (16) 6 (18) 15 (44) 65 (38) 34% 40% 39%
hypertension
No. (%) with diabetes 2 (8) 13 (29) 3 (9) 13 (37) 26 (15) 10% 12% 20%
No. (%) with hyperten- 5 (11) 7% 9%
sion and diabetes
No. (%) with CVA 2 (8) 4 (9) 3 (9) 3 (8) 5 (3)
No. (%) with ischemic 2 (8) 4 (9) 3 (8) 26 (15) 14%b
heart disease
No. (%) with migraine 2 (6) 3 (8) 2% 10%
c
No. (%) with misc. 9 (38) 11 (33) 10 (27) 1.3%
systemic diseases
No.(%) idiopathic 4 (17) 6 (18) 27% 11%
CVA cerebrovascular accidents
a
Data are from Table2 of Guyer and associates
b
Combined CVA+heart disease
c
Hypercholesterolemia, hypertriglyceridemia, and hyperglycemia in 45%, 16%, and 39%, respectively; statistically (P<.05) increased compared with control subjects
285
c ardiovascular diseases, and thyroid disease (prob- cardiac and valvular disease, carotid artery disease,
ably because of hypothyroidisms known associa- and hematologic disorders, can produce NA-AION by
tion with hyperlipidemia). microembolization to the optic nerve head.
2. In contrast to that, other diseases showing signifi-
cant association may in fact have nothing to do
with the development of non-arteritic NA-AION Systemic Diseases of Fortuitous Association
itself but simply represent another manifesta- with NA-AION
tion or cause of diseases common to both. For
example, cerebrovascular disease may fall in this It is worth adding a note of caution; some of the sys-
category. temic diseases recorded in patients with NA-AION, in
3. In still others, e.g., gastro-intestinal ulcers and the various studies, may be no more than purely coin-
chronic obstructive pulmonary disease, in spite of cidental findings, with no direct bearing at all on the
a significant association, we do not know exactly etiology of NA-AION.
how they relate to NA-AION. One can only specu- Guyer etal. [151] reported information on the influ-
late. It is possible that gastro-intestinal ulcers, by ence of systemic diseases on mortality and morbidity in
recurrent bleeding, may play a role in development NA-AION patients. It appears from their discussion that
of AION [158]. Also it is known that gastro-intes- any patient with NA-AION not found to have hyperten-
tinal ulcers are more common in persons with Type sion, presumed atherosclerosis (i.e. with cerebrovas-
A behavior pattern and in smokers [159163] cular or cardiovascular event), diabetes mellitus or
both of these factors would make a person more migraine was classified as idiopathic (28.5% of the
prone to cardiovascular disorders (see below), 200 cases). Using the same criteria, in our study [129]
which may be a common denominator. In chronic 32% were classified as idiopathic. On collating the prev-
obstructive pulmonary disease, smoking and its alence of major systemic vascular diseases seen in our
vascular sequelae combined with poor oxygen- study and that reported by Guyer etal. [151] according
ation may be related to the development of to the three age groups, ischemic heart disease was twice
NA-AION. Furthermore, I have seen, periodically, as common in the middle-aged and elderly and three
NA-AION patients with a Type A behavior pat- times as common in the young in the series of Guyer
tern, who have commented that episodes of stress etal. [151] as it was in ours [129]. Guyer etal. [151] had
or agitation produce either temporary or perma- about twice as many young patients with arterial hyper-
nent visual deterioration. A few patients even tension and diabetes mellitus as our [129] study did. It
attributed the onset of NA-AION to such an epi- would seem that the referral pattern of patients to Guyer
sode. The possible mechanism by which a Type A etal.s group [151] had a much higher incidence of car-
behavior pattern can produce NA-AION is dis- diovascular disease than in the unselected population
cussed below. seen by us [129].This could explain the differences in the
morbidity results of the two studies discussed below.
In that study [129], we had examined a number of There is no evidence in our study [129] that patients
patients undergoing hemodialysis who developed who experience an attack of NA-AION are at an
NA-AION, and the most likely cause seems to be arte- increased risk for subsequent cerebrovascular disease
rial hypotension associated with hemodialysis. or cardiac disease (myocardial infarction). These find-
Systemic lupus erythematosus and other collagen vas- ings are in sharp contrast to the report of Guyer etal.
cular diseases are well-known causes of ischemic vas- [151] who found a significantly higher incidence of
cular accidents, and the high incidence of increased subsequent cerebrovascular (p=0.006) and cardiovas-
erythrocyte sedimentation rate and high antinuclear cular disease (particularly myocardial infarction)
antibody titer in the our series may suggest such dis- (p<.001) than in their controls. This would suggest that
eases. There is evidence indicating that vasospasm in this may be partly due to the type of patient population
migraine may be responsible for the development of seen in that study. Our study [129] also found no
NA-AION in some individuals, particularly young significant increase in incidence of cerebrovascular
people who are otherwise perfectly healthy. Similarly, disease or myocardial infarction in NA-AION patients
systemic diseases associated with embolism, such as with arterial hypertension. However, our study [129]
did concur with Guyer etal. [151] on the finding that Vasculitis
patients with NA-AION in the presence of hyperten-
This may be due to systemic collagen vascular diseases
sion do not have an increased mortality rate over that
(e.g., systemic lupus erythematosus [167] and polyar-
which would be expected from an age, sex and race
teritis nodosa), herpes zoster [168], and other causes.
matched group.
AION has been seen in association with these disor-
We [129] found no significantly increased incidence
ders. Some would like to call AION in this group arter-
of cerebrovascular disease or myocardial infarction or
itic AION, because it is caused by vasculitis.
significantly increased mortality in NA-AION patients
Conventionally, arteritic AION is invariably consid-
with diabetes mellitus. Guyer etal. [151] did not have
ered to be due to GCA. To avoid that confusion, in my
enough NA-AION patients with diabetes mellitus or
description I have grouped this in the NA-AION
with both diabetes mellitus and arterial hypertension to
category.
examine these two subgroups. In our study [129],
NA-AION patients with both hypertension and diabe-
tes mellitus did not have statistically significantly Malignant Arterial Hypertension
increased mortality or statistically significantly
Patients with renovascular malignant arterial hyperten-
increased incidence of myocardial infarction; however,
sion, toxemia of pregnancy and renal disease can
they did have a statistically significantly increased
develop NA-AION [169].
incidence of cerebrovascular disease (p<0.01). Given
these results, we did not find strong evidence that
NA-AION patients in any of these four subpopulations
are at increased risk of developing myocardial infarc- Pathogenesis of NA-AION in Malignant
tion. Evidence of an increased risk of subsequently Arterial Hypertension
developing cerebrovascular disease was confined to
the NA-AION patients with both hypertension and There are strong indications from the available evi-
diabetes. dence that the renin-angiotensin-aldosterone system
Since the publication of our study [129], three more plays an important role in the development and mainte-
studies have been published on the subject. Of the 420 nance of malignant arterial hypertension. Angiotensin
patients who were found to be eligible for the Ischemic II is a powerful vasopressor substance. Since chorio-
Optic Neuropathy Decompression Trial [164], 47% capillaris are very leaky, the plasma, along with the
had arterial hypertension and 24% diabetes mellitus. angiotensin and other vasopressor agents in it, leaks
In the 406 patients in our prospective study [129] the freely into the choroidal interstitial fluid; there it causes
prevalence of arterial hypertension was 43% (26% in vasoconstriction and/or occlusion of the choroidal ves-
young, 41% in middle-aged and 49% in the elderly) sels, seen clearly on fluorescein fundus angiography
and of diabetes mellitus 23% (21% in young, 21% in [170] and in pathological studies [171] in all these eyes.
middle-aged and 26% in the elderly) these are com- Moreover, ischemic choroidal lesions (e.g., Elschnigs
parable. Jacobson etal. [165] on multivariate analysis spots, retinal pigment epithelial lesions, and serous
in a case control study of 51 NA-AION patients older retinal detachment) are almost a universal finding in
than 45 years found only diabetes (odds ratio 5.0, these eyes [170]. The leakage of the angiotensin and
95% confidence interval 1.417.3, P = .01) a signifi- other vasopressor agents into the choroidal interstitial
cant risk factor. Preechawat etal. [166] in a retrospec- fluid and the vasoconstriction and occlusion of the
tive study of NA-AION patients younger than 50years choroidal vascular bed caused by them would produce
found that anemia and type I diabetes were associated ischemia of the optic nerve head by two mechanisms:
significantly with fellow eye involvement. 1. As a part of the choroidal vascular involvement, the
In addition to the diseases mentioned in the above peripapillary choroid is also involved. As discussed
discussion, NA-AION patients may also have a in Chap. 3, the peripapillary choroid is the main
number of other systemic diseases which may directly source of blood supply to the optic nerve head
or indirectly act as risk factors for the development (Fig. 14.1). Vasoconstriction and occlusion of the
of NA-AION [129, 134], and these include the peripapillary choroid would secondarily cause
following: ischemia of the optic nerve head.
2. In addition, the angiotensin from the choroidal Blood Flow

interstitial fluid would diffuse into the optic nerve Autoregulation Range
head through the Border Tissue of Elschnig, which
is freely permeable. The angiotensin in the tissues Hyperten Absent Present Absent
of the optic nerve would produce vasoconstriction
and occlusion by direct action on the capillaries and
other vessels in the optic nerve head; this was also Normal Absent Present Absent
suggested by the experimental studies of Sossi and

Anderson [172]. Pathologic studies in these eyes Perfusion Pressure
showed swollen endothelial cells and degenerated
pericytes with obliterated lumen of vessels in the Fig. 14.8 A diagrammatic representation of blood flow auto-
optic nerve [173]. regulation range at different perfusion pressures in normal per-
sons and in hypertensives (Hyperten). Absent and present
Ischemia of the axons results in axoplasmic flow sta- denote absence or presence of the autoregulation (Reproduced
from Hayreh [184])
sis, which accumulates proximal to the ischemic site
and results in the axonal swelling at the optic disc [139,
173, 174]. Pathological studies in these eyes revealed autoregulation of blood flow in the respective organs.
that optic disc swelling was essentially due to ischemic It has been shown in the cerebral blood flow in patients
hydropic swelling of the axons [173]. Thus, studies with malignant hypertension of some duration that, as
strongly indicate that the optic disc edema (swelling) an adaptive phenomenon, the range of the autoregula-
in accelerated arterial hypertension is due to ischemia tion shifts to a higher level than in normal persons
of axons in the optic nerve head. This represents (Fig. 14.8) [185193]. As discussed in Chap. 5, this
NA-AION. This has been suggested by other authors adaptation makes the individual less tolerant to low
[175177] although their hypotheses about its mecha- blood pressure, while at the same time it improves his
nism were very different. Ischemic damage to the tolerance of high blood pressure. As previously dis-
axons in the optic nerve head finally results in optic cussed, the optic nerve head in patients with malignant
atrophy. hypertension and very high blood pressure is already
ischemic, due to various microvascular changes; any
sudden reduction of blood pressure under such condi-
Dangers of Precipitous Reduction of Blood Pressure tions inflicts further ischemic insult and makes the
in Malignant Arterial Hypertension optic nerve head especially vulnerable to severe acute
ischemic damage. This results in bilateral infarction of
In any discussion of NA-AION in malignant arterial the optic nerve head. In view of this, the most impor-
hypertension, from the clinical management point of tant precaution to be taken in patients with malignant
view, it is important to stress the dangers to the optic hypertension is that blood pressure should be lowered
nerve head of a rapid therapeutic reduction of blood slowly over many hours or even days, giving time for
pressure in such patients. There are a number of reports the autoregulation of the blood flow to adapt itself to
[178182] describing patients with malignant hyper- the falling blood pressure.
tension who had their blood pressure lowered precipi-
tously to prevent severe neurological and cardiovascular
Systemic Arterial Hypotension
complications, and suffered immediate and permanent
blindness or severe visual loss. In all these cases the Most important in this category is nocturnal arterial
optic nerve head was infarcted (i.e. NA-AION) and hypotension [27, 28] and its role is discussed below. In
later developed marked optic atrophy. The findings addition, acute arterial hypotension from other causes
were of severe bilateral NA-AION in all cases. In addi- (e.g., during surgical procedures, renal hemodialysis,
tion, these patients may develop acute ischemic neuro- massive hemorrhages, etc. [129].) is also responsible
logic lesions [180183]. It has been postulated that for the development of NA-AION in a number of
under such circumstances, acute ischemia of the brain, patients. Development of unilateral or bilateral
spinal cord, and/or optic nerve head is due to failure of NA-AION has been reported in patients undergoing
prolonged cardiovascular or spine and neck procedures examination. Thus, in view of these limitations, the
in orthopedic surgery [194204]. In the literature time of discovery of visual loss reported by a patient
NA-AION has been mixed with PION or not differen- has to be placed in its true perspective.
tiated into the two types. It has been stated that In the NA-AION Decompression Trial Study [164],
NA-AION is more prevalent among cardiac surgery it was reported that 42% (174 of 418) of patients
patients and PION among spine and neck procedures recalled that the onset of visual loss occurred within
[205]; however, development of arterial hypotension in 2h of awakening; that much lower than in our study
patients undergoing head and neck surgery has been [30]. However, it is well-known that to get the correct
reported. PION predominates especially among those information, the right person has to ask the right ques-
who have spine and neck orthopedic procedures on tion in the right manner. It is quite possible that the
their backs, lying face-down; in them arterial hypoten- discrepancy between our study [30] and the multi-
sion combined with pressure on the eyeball (associated center study [164] arose because their data were col-
with elevated IOP) results in development of NA-AION, lected by many individuals, not always well-informed
and they discover visual loss on recovering from anes- neuro-ophthalmologists, whereas in our study, if the
thesia [206]. The mechanism of development of isch- patient did not offer that spontaneously, I regularly
emic optic neuropathy in these cases is discussed in elicited it by careful non-suggestive questioning. In
Chap. 20. addition, our [27, 28] 24-h ambulatory blood pressure
monitoring studies independently showed a significant
correlation between progressive visual loss in NA-AION
Role of Nocturnal Arterial Hypotension and nocturnal arterial hypotension.
in Development of NA-AION The time of first discovery of visual loss in NA-AION
is a very critical piece of information because it points
I have discussed this at length elsewhere [27, 28, 30]. to the time of development of NA-AION and also
Typically, patients with non-arteritic and often those indirectly to its pathogenesis. As mentioned above, our
with arteritic AION complain of discovering visual study [30] indicated that NA-AION most often develops
loss on waking in the morning [30]. In our series [30], during sleep, and there must be something happening
among 871 NA-AION eyes (925 episodes), where the during sleep which precipitates it. Our 24-h ambula-
patients could give some information on the time of tory blood pressure monitoring studies [27, 28] in
onset of visual loss, in at least 73% of episodes, the patients with NA-AION revealed a marked interindi-
visual loss was discovered upon awakening (from vidual variation in the amount of fall of blood pressure
sleep during the night or a nap during the day) or soon during sleep. We have discussed elsewhere the role of
after, when the patient had the first opportunity to use nocturnal arterial hypotension in the pathogenesis of
vision critically. In the remaining 27% episodes of NA-AION [27, 28, 30, 32, 129]. The graph shown in
NA-AION, although the patients became aware of the Fig.14.9, based on our study, shows the mean hourly
visual loss later on during the day, they could not say systolic and diastolic blood pressure in hypertensives
definitely when it occurred, or rule out the possibility and normotensives over a 24 h period, starting at
that it had been there since awakening. The time when 10a.m. and continuing till 10a.m. next day. Figure14.7
a patient becomes aware of visual loss is extremely shows an actual blood pressure recording from a patient
variable, depending upon the location, severity and with bilateral NA-AION. There is a steep drop in blood
type of visual field defect, the occupation of the patient, pressure on falling asleep at night and recovery to nor-
and, above all, perceptiveness of the patient in detect- mal on waking in the morning.
ing a visual problem. For example, in my study, a Our study [27] showed that arterial hypertensives tak-
patient with superior visual field loss first noticed the ing oral hypotensive therapy have a significant (p=0.004)
visual field defect while hunting, when he tried to association between progressive visual field deterioration
shoot a flying duck, but was sure that his vision had in NA-AION and nocturnal hypotension. We [28] also
been normal the previous day; another patient, a physi- found that among the hypertensives on oral hypotensive
cian, was not aware of any visual loss, in spite of fairly drugs, the mean minimum night-time systolic blood pres-
marked visual loss in the inferior nasal sector of one sure was significantly lower in the patients with visual
eye, till it was discovered during an ophthalmic field deterioration than in those without visual field
Fig.14.9 Mean hourly MEAN HOURLY SYSTOLIC BLOOD PRESSURE

systolic and diastolic blood 150
pressure over a 24-h period
in patients with normal
blood pressure (normoten-
sive) and hypertension 140
(hypertensive) (Reproduced
Systolic Pressure (mmHg)

from Hayreh etal. [27])
130
120
110
Normotensive
Hypertensive
100
10am 2pm 6pm 10pm 1am 4am 7am 9am
HOUR
MEAN HOURLY DIASTOLIC BLOOD PRESSURE

90
80
Diastolic Pressure (mmHg)
70
60
Normotensive
Hypertensive
50
10am 2pm 6pm 10pm 1am 4am 7am 9am
HOUR
deterioration (p=.040, 106.03.5 vs. 119.53.9mmHg). difference between those with and without progressive
They also showed a significantly larger mean percentage visual field deterioration for these variables.
drop in systolic, diastolic and mean BP compared with The fall of blood pressure during sleep is a physi-
similar patients with no visual field deterioration (sys- ological phenomenon but it is much influenced by
tolic: p=0.026, 39.01.6% vs. 32.41.6%; diastolic: many factors, including drugs such as beta-blockers,
p=.023, 45.61.7% vs. 38.61.9%; mean: p<0.0001, calcium-channel blockers (Fig. 14.10), angiotensin
47.51.8% vs. 36.42.0%). Normotensives and hyper- converting enzyme inhibitors, Hytrin (Terazosin
tensives without medication showed only a small hydrochloride) used in enlarged prostate to reduce
280 a
260
240
220
Blood Pressure (mmHg)
200
180
160
140
120
100
80
60
40
20
280
b
260
240
220
200
180
160
140
120
100
80
60
40
20
Fig. 14.10 Ambulatory blood pressure monitoring records shows when she was taking Verapamil 160 mg morning and
(based on individual readings) over a 24-h period, starting from bedtime and 80mg at noon, there was a marked degree of noc-
about 10 AM in a 63 year-old woman who developed normal turnal arterial hypotension. (b) This record shows a marked
tension glaucoma while on Verapamil hydrochloride for improved in nocturnal arterial hypotension on stopping the bed-
migraine (Reproduced from Hayreh [184]). (a) This record time dose
nocturia, phosphodiesterase type 5 and other similar Landau etal. [207], based on a study of only 24
compounds particularly the number and amount of patients (13 from Switzerland and 11 from New
drugs taken and the time of day they are taken. For York) with NA-AION, claimed that a patient with
example, when these drugs were taken at bedtime, AION typically notices blurred vision in the
they produced a far more marked degree of nocturnal affected eye within 1 or 2h of arising in the morn-
hypotension than when taken in the morning because ing, (i.e. similar to the onset of many of the myo-
they aggravate the naturally occurring fall of blood cardial infarctions) and not first thing in the
pressure during sleep (Fig. 14.10). There are, how- morning, and that nocturnal arterial hypotension
ever, some patients who develop marked nocturnal does not play any role in the development NA-AION.
hypotension even without any medication (this is pre- The authors stressed that the pathogeneses of myo-
sumably due to defective cardiovascular autoregula- cardial infarction and NA-AION are similar, and
tion), as can be seen in Fig. 14.7, a blood pressure added that that helps to explain why patients with
tracing of a patient with bilateral NA-AION and pro- AION have a high incidence of myocardial infarc-
gressive visual loss. tion. [207] We pointed out elsewhere [208] multiple
flaws in that study. These include: (a) flawed data condition. In view of all this evidence, it seems
collection, analysis and interpretation of the ambula- NA-AION may be occurring as an iatrogenic disease
tory blood pressure monitoring data, (b) no informa- in some persons. There may be a paradoxical phenom-
tion on the short-term fluctuations of blood pressure enon: arterial hypertensives on excessive medications
which are critical in NA-AION, and (c) that our study to lower their blood pressure developing arterial
[129] in 406 NA-AION patients showed no evidence hypotension; a combination of arterial hypertension
of an increased incidence of myocardial infarction in and hypotension can play an important role in either
NA-AION patients on follow-up as compared to the the development or the progression of NA-AION [29].
age-sex matched general population (see above). We This brief discussion, then, gives some idea of the
believe that these and other faults in the study invali- great complexity of the mechanisms of development of
date the conclusions of Landau etal. [207]. NA-AION and the role of nocturnal hypotension in it.
To place the role of nocturnal arterial hypotension A whole host of systemic and local factors acting in
in proper perspective in the pathogenesis of NA-AION, different combinations and to different extents may
very briefly: as discussed above, our studies show that derange the optic nerve head circulation, with some
NA-AION is a multifactorial disease, with many risk making the optic nerve head susceptible to ischemia
factors acting in various combinations, some acting as and others acting as the final insult. Nocturnal hypoten-
predisposing factors, making an optic nerve head sus- sion seems to be an important precipitating factor in
ceptible to ischemia, and others as precipitating fac- the susceptible patient. Thus, nocturnal hypotension is
tors inflicting the final insult and resulting in one of the variables to produce NA-AION, and its role
development of NA-AION. In such a multifactorial must be placed in the overall context of the disease
scenario, any particular factor or combination of fac- process and not taken in isolation.
tors may be present in one case and not in another, or a Some patients develop NA-AION during the day-
factor may play a major role in one case and only a time, not associated with a nap. This may be due a
subsidiary role in another. The predisposing factors variety of causes, including the following:
may be systemic or local in the optic nerve head [32,
129]. When an optic nerve head has been rendered vul- (a) Embolism into the PCA circulation (the main source
nerable to ischemia by these conditions (in disparate of blood supply to the optic nerve head [91, 133])
combinations in different individuals), nocturnal can produce NA-AION which can develop any time
hypotension may act as the precipitating factor to proof the day. The eyes with the usual type of NA-AION
duce ischemia and NA-AION [27, 28, 30] the straw generally have no optic disc cup [106, 157], but
that breaks the camels back. In a normal, healthy those of an embolic nature generally have the nor-
optic nerve head with normal autoregulation, a similar mal cup size. On fluorescein fundus angiography, in
fall of blood pressure during the night may have no these cases, we have seen complete occlusion of a
deleterious effect at all. The very potent drugs, e.g., PCA (with negative temporal artery biopsy) [91]
beta-blockers, calcium channel blockers, angiotensin (see Figs.3.27 and 3.35). The source of the embolus
converting enzyme inhibitors and other drugs, which may be the carotid arteries, aorta or the heart.
are now available to treat hypertension and prescribed (b) Our experimental study [102] in atherosclerotic
commonly, have begun to emerge as an important risk monkeys has shown that serotonin in atherosclero-
factor for arterial hypotension (particularly nocturnal sis can produce vasospasm of the PCA, resulting in
hypotension see Fig. 14.10), especially when used its occlusion and development of NA-AION. We
aggressively and/or given at bedtime. Historically, it is speculate that the vasospasm is produced by the
interesting to note that before 1960 almost the only release of serotonin by platelet aggregation on ath-
treatment for arterial hypertension was diuretics and erosclerotic plaques.
NA-AION was a rare clinical entity at that time.
Thefirst beta-blocker drug was marketed in the early
Atherosclerosis and Arteriosclerosis
1960s, and since then a plethora of new potent arterial
hypotensive drugs have emerged to treat arterial hyper- As a part of generalized atherosclerosis and arterio-
tension; also since then, the incidence of NA-AION sclerosis, the internal carotid artery, ophthalmic artery,
has progressively increased, so that it is now a common PCAs and other nutrient arteries to the optic nerve may
be involved and contribute to the development of can result in development of NA-AION [144].
NA-AION by several mechanisms. Development of NA-AION has also been reported fol-
lowing carotid dissection [217221]. Fry et al. [222]
1. It can cause marked stenosis or occlusion of the evaluated 15 consecutive patients with NA-AION for
ocular or internal carotid arteries, resulting in cervical carotid artery stenosis and compared them with
hypoperfusion which makes a person vulnerable to 30 age- and sex-matched asymptomatic patients and
NA-AION. also with 11 age- and sex-matched patients experienc-
2. Embolism from an atherosclerotic plaque going to the ing transient monocular blindness. They found no dif-
PCA can result in the development of NA-AION [91]. ference in the mean stenosis of the internal carotid
3. Our experimental study [102] in atherosclerotic mon- artery between patients with NA-AION and asymptom-
keys showed that release of serotonin from platelets on atic patients, whereas patients with transient monocular
an atherosclerotic plaque can produce transient spasm blindness had significantly more stenosis in the cervical
and that can result in the development of NA-AION carotid arteries than control subjects (p<0.0001), and
[209]. As mentioned above, serotonin liberation from also than patients with NA-AION (p<0.0001). There
platelet aggregation on the atherosclerotic plaques was also no difference in the percentage of patients with
may also play a role in some cases [102]. stenosis30% between NA-AION and asymptomatic
patients, whereas 10 of 11 patients with transient mon-
ocular blindness had stenosis30%, significantly more
Migraine and Other Vasospastic Disorders
than patients with NA-AION (p<0.0001) and asymp-
There are anecdotal case reports of NA-AION associ- tomatic patients (p<0.0001). They concluded that
ated with migraine [129, 210214]. In our study [129], NA-AION is not a marker for atherosclerotic carotid
17 (4%) of 406 NA-AION patients gave a history of artery stenosis. I find that while their conclusion may be
migraine; according to age group, migraine was valid for carotid artery stenosis of 30%, which is con-
reported by 7% of the young, 6% of the middle-aged sidered minimal stenosis and does not cause any hemo-
and 2% of the elderly. It seems reasonable to assume dynamic problem, that is not true for high grade stenosis
that there is an association between NA-AION and (>70%) which can lower the perfusion pressure in the
vasospastic disorders. One case in particular seems eye and optic nerve head, and thus becomes a risk factor
interesting from this point of view: A woman with a for development of NA-AION, as seen by us in ocular
strong history of migraine since the age of 26years, ischemic syndrome [144, 145]. Thus, the degree of
first developed NA-AION in her right eye at age 27 stenosis of the internal carotid artery is the crucial fac-
and in the left eye at 31; she had one recurrence of tor in determining its role in NA-AION.
NA-AION in the right eye (10 years after the initial
episode) and two recurrences (first 7years and the sec-
Cardiac Valvular Diseases
ond 20 years after the initial episode) in the left eye
over a period of 25years follow-up. Each episode was These include mitral valve disease, mitral prolapse,
associated with an attack of migraine. This was aortic valve disease, artificial heart valve, rheumatic
strongly suggestive of NA-AION due to migraine. heart disease, and other valvular diseases [129]. I have
seen some NA-AION patients having mitral valve pro-
lapse, mostly in young patients; it is impossible to state
Internal Carotid Artery Disease
whether there is a cause-and-effect relationship
This can contribute to development of NA-AION either between it and NA-AION. The most likely mechanism
by embolism or by lowering the perfusion pressure of development of NA-AION in these patients is
because of stenosis or occlusion. The most common microembolism to the optic nerve head.
cause of carotid artery stenosis or occlusion and conse-
quent development of NA-AION is atherosclerotic
Defective Cardiovascular Autoregulation
change in the artery. Rarely it may be due to other con-
ditions, e.g., pulseless disease [215] or internal carotid Patients who develop orthostatic arterial hypotension
artery hypoplasia [216]. Ocular ischemic syndrome is would seem to have this disorder. It is known that
due to carotid artery occlusion or marked stenosis and diabetics have defective autoregulation. This can be
a factor, by producing arterial hypotension. That may NA-AION while being treated for sleep apnea syn-
play a role in development of NA-AION in some drome with CPAP. Of the 108 patients in the series,
cases. one had bilateral sequential NA-AION and two had
unilateral NA-AION despite treatment with CPAP for
sleep apnea syndrome. All the patients were being
Type A Behavior Pattern
treated with CPAP for a period ranging from 4months
As mentioned above, I have seen patients in whom the to 6years before the onset of NAION. They concluded
type A behavior pattern seems to play some role in the that CPAP did not prevent the development of
development of NA-AION [32, 129]. The following NA-AION in sleep apnea syndrome patients.
mechanism may be responsible for that. Increased lev- In my experience of dealing with more than a 1,000
els of plasma catecholamines have been well estab- NA-AION patients, I have found only a few who gave
lished in persons with a Type A behavior pattern under a history of sleep apnea on questioning, which was
stress [223, 224]. Leakage of catecholamines into the confirmed to be genuine by a sleep study.
optic nerve head (as in malignant arterial hypertension,
discussed above) can cause vasoconstriction [169],
Hematologic Disorders and Thrombophilia
resulting in optic nerve head ischemia.
There are many reports, mostly anecdotal, claiming
an association of various hematologic abnormalities
Sleep Apnea Syndrome
with NA-AION, e.g., various types of anemia, sickle
I was the first to point out in 1996 that sleep apnea cell disease, sticky platelet syndrome, polycythemia
may be a risk factor in the development of NA-AION vera, Henoch-Schnlein purpura, factor V Leiden,
[32]. Since it is associated with cardiovascular disor- activated protein C, homozygosity for the C677T
der, there is every reason to suspect that it may act as methylenetetrahydrofolate reductase mutation, anti
a risk factor in NA-AION. Since then an association of phospholipid syndrome and other hematologic abnor-
sleep apnea with NA-AION has been reported by malities [232246].
several authors [225231]. In sleep apnea syndrome Several studies have investigated the role of throm-
there are recurrent episodes of upper airway flow stop- bophilia in NA-AION. There is conflicting and con
page during sleep; these episodes may be partial or tradictory information in those studies, and other
complete. Cardiovascular disorders associated with problems. A number of studies over the years have
sleep apnea are well documented. For example, claimed that thrombophilia does play a role in the
Sommers etal. [109] in their study of obstructive sleep development of NA-AION. For example, Acheson and
apnea, concluded that patients with obstructive sleep Sanders [233] described seven cases in which there
apnea have high sympathetic activity when awake, was an associated thrombophilic (prothrombotic) state;
with further increases in blood pressure and sympa- four had deficiencies of the physiological anticoagu-
thetic activity during sleep. These increases are atten- lants proteins C and S and antithrombin III, and two
uated by treatment with continuous positive airway had anti-phospholipid antibody (lupus anticoagulant)
pressure (CPAP). syndromes. One other patient had reduced levels of the
The prevalence of sleep apnea syndrome in physiological fibrinolytic agent tissue plasminogen
NA-AION is in not known as yet because there has activator (t-PA). In five patients other risk factors for
been no large study with proper evaluation of sleep small vessel occlusive disease were also present, and
apnea in consecutive NA-AION cases. The extreme four had recurrent episodes of ischemic optic neuropa-
example is that of Palombi etal. [227], who, in a study thy in the same eye. They stated that the visual progno-
of 27 consecutive newly diagnosed NA-AION patients, sis of these patients may be improved by anticoagulation
found 89% exhibited a sleep apnea syndrome. They with warfarin. Glueck et al. [237] evaluated 12
concluded that sleep apnea is the most frequent disor- NA-AION patients and 36 healthy, normal race-, sex-,
der associated with NA-AION. That high prevalence is and age-matched controls by polymerase chain reac-
not supported by other studies and may be due to tion-complementary DNA assays of gene mutations
referral bias. For example, Behbehani etal. [226], in a associated with coagulation disorders and serologic
prospective review, identified patients who developed coagulation measurements. They claimed that they
found among the 12 NA-AION patients, 7 (58%) had 1. Thrombophilia is an inherited or acquired predispo-
at least 1 gene mutation in the C677T MTHFR, sition to thrombosis [249]. Therefore, thrombo-
G1691A V Leiden, or G20210A prothrombin gene, philic factors cause circulatory disorders by causing
compared with 5 of 36 controls (14%) (P=.002). Of thrombosis. The primary reason for considering
the eight women with NA-AION, five (63%) first NA-AION as due to thrombophilia (i.e. due to
experienced the condition while taking hormone thrombosis) is by equating it to cerebral strokes.
replacement therapy (n=4) or during pregnancy (n=1), As discussed above, NA-AION is a hypotensive
with superposition of estrogen-induced thrombophilia disorder and NOT a thrombotic disorder, while
on heritable thrombophilia and hypofibrinolysis. cerebral stroke is a thromboembolic disorder. Thus,
Kuhli-Hatternbach et al. [245] stated that they found NA-AION and stroke pathogenetically are two dif-
increased levels of factor VIII and lipoprotein (a) in ferent clinical entities.
NA-AION patients. Giambene et al. [246] reported 2. The principal clinical manifestation of thrombo-
plasma levels of homocysteine and lipoprotein (a), as philia is venous thromboembolism [249]. NA-AION
well as low vitamin B6 levels, significantly higher in is not a venous thrombotic disorder but an arterial
patients with NA-AION compared with controls, with ischemic disorder.
no significant difference in MTHFR C677T polymor- 3. Aspirin, which is effective in thromboembolic dis-
phism between the two groups. orders, has been shown to have no protective effect
However, there are studies which have shown no in NA-AION [121, 123, 124].
association between NA-AION and thrombophilic 4. As mentioned above, Salomon et al. [125] and
risk factors. For example, Salomon etal. [125], in 61 Biousse etal. [248] found no association between
patients with NA-AION found no association with NA-AION and thrombophilic risk factors.
Protein C, protein S, antithrombin III, lupus antico-
agulant, and three prothrombotic polymorphisms If testing of NA-AION patients reveals the presence of
(i.e., factor V G1691A, factor II G20210A, and meth- one or more abnormal thrombophilic factors, that does
ylenetetrahydrofolate reductase [MTHFR] C677T). not automatically imply a cause-and-effect relation-
Also that group [247] found no association between ship; it may be purely a coincidental finding. That may
NA-AION and angiotensin converting enzyme and be the explanation for some of the reported positive
angiotensin II type 1 receptor. Similarly, Biousse findings.
et al. [248] in 14 patients with NA-AION found I conclude that, in the light of the above facts, there
homocysteine within normal limits in all of them, and is no rational scientific basis for believing that throm-
mutation positive and mutation negative patients for bophilia has any role to play in NA-AION in the vast
MTHFR C677T did not differ with respect to clinical majority of cases, since NA-AION generally is not a
data concerning risk factors for NA-AION. Both thrombotic disorder but a hypotensive disorder. Of
studies concluded that there was no association course, the possibility that thrombophilia may be
between NA-AION and thrombophilic risk factors. involved in a rare case cannot be ruled out in Medicine
Often there are other problems with the studies there is no such thing as never. That raises the impor-
claiming an association between NA-AION and tant question whether all patients with NA-AION
thrombophilia, For example, Pianka etal. [234] and should be tested for thrombophilia, as advocated by
Stanger etal. [240] combined NA-AION with retinal some [246]. To quote Heit [249] Currently, there is no
vascular occlusions in their study on thrombophilia, single laboratory assay or simple set of assays that will
which is unwarranted because retinal vascular occlu- identify all thrombophilias. Consequently, a battery of
sions (thromboembolic disorders) and NA-AION complex and potentially expensive assays is usually
(hypotensive disorder) are not only totally different required. Many of these laboratory analyses are
pathogenetically but also as clinical entities. It is like affected by other conditions (e.g., warfarin reduces
mixing apples and oranges. protein C and S levels) such that the correct interpreta-
I [128] recently discussed the role of thrombophilia tion of the results can be complicated and always
in NA-AION. I pointed out that to evaluate whether requires clinical correlation. Leiden mutation of the
thrombophilia has any role in development of factor V gene is associated with a procoagulant state,
NA-AION, one has to consider the following facts. especially in the venous bed, and its association with
arterial thrombotic disease remains unclear. Thus, accidents. For example, Bull etal. [255] noted a positive
testing patients with NA-AION for thrombophilia is correlation between temperature and factor VII, anti-
unwarranted, unless there is a medical or family his- thrombin III and cholesterol and negatively with fibrin-
tory of thrombosis. The latter is well demonstrated by olytic activity. Keatinge et al. [256] reported that
Glueck and Wang [244] in a 12-member, 3-generation exposure to cold increases packed cell volume, platelet
family. They investigated them by using polymerase count, mean platelet volume, both whole blood and
chain reaction measures for thrombophilia (G1691A plasma viscosity, arterial BP, and cholesterol. They
factor V Leiden, G20210A prothrombin gene, C677T- speculated that the larger than normal platelets seen in
A1298C methylenetetrahydrofolate reductase, platelet these patients may be either due to cold induced activity
glycoprotein and hypofibrinolysis [plasminogen acti- of sympathetic nerves releasing sequestered large plate-
vator inhibitor-1 4G4G]). This extended family was lets or by production of large new platelets. Large plate-
associated with variegated thrombotic events, includ- lets aggregate and adhere more readily to blood vessels
ing ocular ischemic disorders (amaurosis fugax, than small platelets [257, 258]. Several other changes
NA-AION), transient ischemic attack, ischemic stroke, produced by cold exposure will tend to induce thrombo-
deep vein thrombosis, pulmonary embolus, and mes- sis, e.g., an increase in the number of red blood cells
enteric artery thrombosis. This is a rare occurrence. promotes platelet adhesion [259261], and increased
blood viscosity facilitates clotting of blood after the for-
mation of a platelet thrombus [256]. Stout and Crawford
Role of Seasonal Variations in the Development
[262] in subjects aged 75 and over, found significant
of NA-AION
seasonal effects for fibrinogen, plasma viscosity and
Apart from the relationship between the time of day and HDL cholesterol. Plasma fibrinogen concentrations
the onset of NA-AION, we investigated the role of sea- were 23% higher in the coldest 6months compared with
sonal variations in the development of NA-AION [30]. summer months. Fibrinogen significantly and negatively
This is because seasonal variation has been reported in related to core body temperature and all measures of
the incidence of vascular accidents, such as myocardial environmental temperature; they felt that this variation
infarction and cerebrovascular accident. For example, was large enough to increase the risk of myocardial
temperature has been reported to be closely correlated infarction and cerebrovascular accident during winter.
with both myocardial infarction and cerebrovascular Buchanan and Riglar [263] reported high titers of anti-
accident, with both of the latter more common in winter centromere antibodies in Raynauds phenomenon and
than in other seasons [250252]. Bull and Morton [251] patients with a high titer of anti-centromere antibodies
found a nearly linear fall in death rate as the minimum had a higher risk of developing thrombotic vascular
temperature rose between 10oC and +20oC, but above disease. Reilly etal. [264] reported a significant circan-
20oC deaths rose steeply with the rise in temperature nual variation in a large number of hematologic param-
and below 10oC deaths again rose steeply as tempera- eters in the elderly. Exton-Smith [265] reported that
ture fell, especially in the elderly; this was also reported exposure to low environmental temperature more read-
by Rogot and Padgett [253]. Bull and Morton [251] ily produced orthostatic hypotension or hypothermia in
found a stronger association of death rates with tem- old than in younger persons, as a consequence of
perature in the old than the young, probably because of impaired physiological thermoregulatory mechanisms
their failure to adapt physiologically to temperature in the elderly. Brennan etal. [266] and Keatinge etal.
change. Fersini et al. [254], in patients with transient [256] reported an increase in blood pressure with cold.
ischemic attacks, showed a seasonal pattern of disease We [30] investigated seasonal variation of the onset
onset in early spring, with a peak in MarchApril, and of NA-AION for 839 episodes. The data showed that
isolated peaks in January and October. NA-AION onset was significantly (p=.0030) more
In order to explain the seasonal variations of myocar- frequent in the Iowa (USA) Summer (the hot months,
dial infarction, cerebrovascular accidents and possibly with temperature in 60s to low 90s degrees Fahrenheit)
other vascular accidents, a number of studies have been than in the Winter (the cold months, when the maxi-
conducted to find the effects of temperature variations mum temperature is usually much below freezing),
on hematologic and other vascular changes in the body withSpring/Fall or the mild months intermediate. That
which could be contributory to development of vascular study suggests that factors other than hematologic
abnormalities may be playing an important role in the observation since then has been confirmed by many
development of NA-AION. Moreover, in that study, large studies [106, 157, 268271]. It is unfortunate
there was no evidence that NA-AION patients are at to find that tremendous stress has been placed upon
increased risk of developing subsequent cerebrovascu- the importance of a small or absent cup (the catch-
lar disease, myocardial infarction or mortality [129]. phrase is disc at risk [272]) in NA-AION [272
We have no explanation for this seasonal variation in 274], leading to almost a universal impression in the
development of NA-AION. ophthalmic community that a small or absent cup is
The reason for all these discrepancies is that myocar- actually the primary factor in the development of the
dial infarction and cerebral strokes are thromboembolic disease. I have met persons without any visual com-
disorders while NA-AION is a hypotensive disorder and plaint who have been alarmed by ophthalmologists
NOT a thromboembolic disorder in the vast majority of telling them that they were at risk of developing
cases. It is this highly prevalent misconception of equat- NA-AION simply because they have no cup or a
ing NA-AION with cerebral stroke (see above) that is small cup. In view of that, it is important to place
the primary reason for the confusion and controversy on this subject in proper perspective; to do that one has
the pathogenesis and management of NA-AION. to look at the overall pattern of the cup/disc (C/D)
ratio in the normal general population versus that in
NA-AION patients.
Ocular Conditions Associated with NA-AION
and Risk Factors
C/D Ratio in Normal General Population
The association of NA-AION with a number of ocular Information about the C/D ratio in persons without
and optic nerve head conditions has been reported by NA-AION is provided by four general population
various studies. These conditions include: based studies [154157, 275] (Table14.2).
The Framingham Eye Study [154, 275], in a survey
of a general population without any visual complaint,
Absent or Small Cup in the Optic Disc evaluated the C/D ratio in 5,054 eyes of persons of all
ages on ophthalmoscopy. Mean horizontal C/D ratio
In 1974, based on my [267] study in AION, I reported was 0.250 (SD 0.159) and mean vertical C/D ratio
that seven of ten eyes with NA-AION had no optic 0.255 (SD 0.166). There was no evidence that the C/D
disc cup and the remainder only a small cup. That ratio changes significantly with age.
Table14.2 Comparison of C/D ratio in patients with NA-AION versus the normal control group
Study group Normal control Number of eyes Method used to Cup/disc ratio
or with examined evaluate C/D ratio
NA-AION
Framingham eye study Control 5,054 Ophthalmoscopy Mean horizontal C/D ratio
[154] 0.2500.159; mean vertical
C/D ratio 0.2550.166
Rotterdam study [155] Control 5,114 Stereoscopic image Mean 0.490.14
analyzer
Rotterdam study [156] Control 5,143 Ophthalmoscopy Mean 0.30 (standard error
0.0021)
Beck etal. [157] Control 122 Ophthalmoscopy <0.15 in 20%; <0.25 in 39%
(mean 0.310.19)
Beck etal. [157] NA-AION 126 Ophthalmoscopy <0.15 in 48%; <0.25 in 71%
(mean 0.160.15)
Hayreh and Zimmerman NA-AION 65 Morphometric <0.15 in 37% and <0.25 in
[106] 75% (mean 0.1990.093)
C/D cup-disc ratio, NA-AION non-arteritic anterior ischemic optic neuropathy, standard deviation
The Rotterdam Study [155], based on a general estimate the C/D ratio in various studies. As mentioned
population study of 5,114 subjects, aged 55years and above, the semiautomated methods for measurement
older, using a stereoscopic image analyzer, showed a of the C/D ratio (used in the Rotterdam Study [155]),
mean vertical C/D ratio of 0.49 (SD, 0.14), and mean and our study [106] give a larger ratio than the ophthal-
horizontal cup-to-disc ratio of 0.40 (SD, 0.14). It moscopically measured ratio [156].
showed no association of cup area, vertical C/D ratio,
or horizontal C/D ratio with age, refractive error or
The Role of an Absent or Small Cup in the Pathogenesis
height [155]. In another investigation [156], this
of Development of NA-AION
Rotterdam Study group evaluated vertical C/D ratio in
5,143 subjects, aged 55years and older, by two meth- As discussed above, since 1974 [267], it has been
ods ophthalmoscopic and semiautomated. That study shown that eyes with NA-AION have a significantly
showed a mean ophthalmoscopic vertical C/D ratio higher prevalence of absent or small cup than the gen-
0.30 (standard error 0.0021; range 0.00, 1.00) com- eral population [106, 157, 268271] giving almost a
pared with a semiautomated measured vertical C/D universal impression in the ophthalmic community
ratio of 0.49 (standard error 0.0019; range 0.04, 0.86). that a small or absent cup is actually the primary factor
This showed that semiautomated measurement of the in the development of the disease or that those eyes are
vertical C/D ratio is, obviously, larger than the oph- at risk of developing NA-AION. This is not a valid
thalmoscopically measured ratio. According to the concept. The Framingham Eye [154, 275] and Beck
authors, the reason for this discrepancy is the use of etal.s [157] studies showed that not all eyes with no
different criteria for defining the cup and possibly the cup or small cup develop NA-AION and, conversely,
disc as well. The authors concluded that currently some eyes with a larger cup do develop NA-AION (see
there is no gold standard for assessment of the VCDR the case report below). The role of an absent or small
(vertical C/D ratio) [156]. cup in the pathogenesis of the development of
Beck etal. [157], in a study of 122 eyes of persons NA-AION is discussed in detail elsewhere [106, 157].
without NA-AION, showed that 12% had no cup, 8% Briefly, in the multifactorial scenario of the pathogen-
had a C/D ratio 0.050.149, and 19% had a C/D ratio esis of NA-AION, one has to consider the role of the
0.150.249. following two factors relevant to C/D ratio.
(a) The size of the physiological cup is determined
Comparison of Cup-Disc Ratio in Patients primarily by the size of the scleral canal embryo-
with NA-AION Versus Normal Control Population logically [276]. With a small scleral canal, and
associated small opening in the Bruchs membrane
A study of 248 eyes (126 with NA-AION and 122 in the region of the optic disc, there is a small or no
normal), by Beck et al. [157] showed that in normal cup. The C/D ratio in persons with NA-AION is
subjects 19% had no cup, 11% had a cup-disc ratio usually smaller than that seen in general popula-
0.050.149, and 16% had a cup-disc ratio 0.150.249. tion. This means the optic discs of NA-AION
That means that 46% of normal individuals had either patients tend to have smaller openings in their
no cup or a small cup, compared to 77% in NA-AION Bruchs membrane and scleral canal than the gen-
eyes. Our morphometric study [106] of the optic disc eral population, resulting in crowding of the optic
in the 65 fellow normal eyes of patients with unilateral nerve fibers as they pass through the restricted
NA-AION found a vertical C/D ratio of <0.15 in 37% space in the optic disc and lamina cribrosa.
and <0.25 in 75% and horizontal C/D ratio of <0.15 in (b) Ischemia or hypoxia of the axons in the optic nerve
31% and <0.25 in 82%. This shows that the C/D ratio head causes axoplasmic flow stasis [139, 174]
in persons with NA-AION is smaller than that of the Axoplasmic flow stasis causes swelling of the
normal control population. The methods used to evalu- axons, and that is responsible for optic disc edema
ate C/D ratio have been different in different studies, in NA-AION.
resulting in different C/D ratios. However, the magni-
tude of the difference between the control eyes and Thus, the sequence of events in the development of clas-
those with NA-AION suggests that this difference can- sical NA-AION is this: Initially, optic nerve head isch-
not be attributed solely to the different methods used to emia/hypoxia causes axoplasmic flow stasis, which
a b
c d
Fig.14.11 Fundus photographs of the right eye showing optic of optic disc edema. (c) Optic disc on recurrence of NA-AION.
disc changes during follow-up. (a) Optic disc at initial visit dur- (d) Optic disc at final visit
ing the first episode of NA-AION. (b) Optic disc on resolution
results in axonal swelling. If the optic disc has a cup, the optic disc produce secondary vascular changes [277]. It
swollen axons can expand into that space without com- has been shown that asymptomatic optic disc edema is
pressing any other tissues in the optic disc (see case the earliest sign of NA-AION [113, 278]. It has also
report below and Figs.14.11 and 14.12). But when there been demonstrated that nocturnal arterial hypotension
is no cup or only a small one, the swollen axons are precipitates the development of NA-AION [30]. Thus,
crowded in a restricted space in the optic disc. In that the available evidence indicates that the sequence of
case, the swollen axons can expand only by compress- events in the development of NA-AION is as follows:
ing the surrounding tissues. The tissues that are most Subclinical ischemia (hypoxia) of the optic nerve
vulnerable to compression in this case are the capillaries headaxoplasmic flow stasis in the optic nerve
and other fine vessels lying among the nerve fibers. fibersaxonal swellingasymptomatic optic disc
Thus, swollen axons in a restricted space within the edema [113, 278](incipient NA-AION)compression
a b
c d
Fig. 14.12 Fundus photographs of the left eye showing optic during the first, second and third episodes of NA-AION respec-
disc changes during follow-up. (a) Normal optic disc at initial tively. (e) Optic disc at final visit
visit before it developed NA-AION. (b, c) and (d) Optic disc
e (Fig. 14.11a) and in the left eye at 31 (Fig. 14.12b).

Over 35years of follow-up, she had one recurrence of
NA-AION in the right eye 10 years after the initial
episode (Fig.14.11c) and two in the left eye, the first
7 years (Fig. 14.12c) and the second 20 years
(Fig. 14.12d) after the initial episode. Each episode
was associated with an attack of migraine. During all
the episodes the visual acuity was 20/20 to 20/15 in
both eyes. Manual kinetic perimetry with a Goldmann
perimeter showed minimal visual field defects: in the
right eye, during the first episode there was a relative
superior centrocecal scotoma with I-2e isopter only,
touching fixation, and during the second episode only
an enlarged blind spot. In the left eye, during the first
episode there was a small superior centrocecal scotoma
with I-2e and I-4e isopters 5 from fixation, during the
second episode a small superior centrocecal scotoma
Fig.14.12 (continued) with I-2e isopter only extending up to 7 from fixation,
and during the third episode mild enlargement of the
of the intervening capillaries by swollen axons in a blind spot. In both eyes, during all these episodes, the
crowded discsetting up a vicious cycle, i.e. the greater peripheral fields were normal. At final evaluation
the compression of capillaries, the greater the blood 35years after her first episode, her visual acuity was
flow compromise, the greater the axoplasmic flow stasis 20/20 in the right eye and 20/25 in the left, and visual
and the more axonal swelling. Since compression of the fields showed only mild peripheral constriction in the
optic disc capillaries increases the vascular resistance in inferior nasal region in both eyes and a small centroce-
them and compromises their blood flow, a fall of blood cal scotoma in the right eye extending up to 2 from
pressure must further derange their blood flow. the fixation point (Fig.14.13). The optic discs in both
Therefore, in this situation, a fall of perfusion pressure eyes showed marked enlargement of the blind spots
in the optic disc capillaries due to nocturnal arterial (Figs.14.11d and 14.12e). She showed no evidence of
hypotension results in marked ischemia and that pre- demyelinating disease or optic neuritis with normal
cipitates visual loss (symptomatic NA-AION), which is brain and orbit imaging.
usually discovered on waking up in the morning [30]. Changes in the optic disc cup size during the various
From this sequence of events, it is evident that in episodes were highly instructive. The left eye had a
the multifactorial scenario of the pathogenesis of large optic disc cup before it developed NA-AION
NA-AION, contrary to the prevalent impression, an (Fig.14.12a); however, when the disc was edematous
absent or small cup is simply a secondary contribut- due to NA-AION, it was either reduced to a very small
ing factor, ONCE the process of NA-AION has size (Fig.14.12b) or completely obliterated (Fig.14.12c,
started, and NOT a primary factor [113, 157, 279]. d) due to swollen axons expanding into the cup.
What happens when an eye with a large optic Similarly, in the right eye, on resolution of optic disc
disc cup develops NA-AION? edema after the first episode, the optic disc had a large
When an eye has a large cup, the latter acts as a cup (Fig.14.11b) but when the disc was edematous, the
protective mechanism, providing space for the swollen cup was reduced to a small size (Fig.14.11a) or com-
axons to expand into, resulting in much less interfer- pletely obliterated by the swollen axons expanding into
ence in the optic nerve head circulation than wouldhave the cup (Fig.14.11c). Figures14.11d and 14.12e show
happened had there been no cup. This was very well marked enlargement of the cups in both eyes after
demonstrated by one of our patients (personal commu- recurrent episodes of NA-AION in them over 25years
nication from ES Birkholz, SS Hayreh, JJ Corbett). (compare these with the initial normal cup size of the
A27year old Caucasian woman, with a strong history right eye in Fig.14.11b and the left eye in Fig.14.12a).
of classical migraine since the age of 26 years, she In spite of repeated episodes of NA-AION in both eyes
first developed NA-AION in her right eye at age 27 (two in the right and three in the left), the visual acuity
of normal visual acuity and none or minimal demon-

strable visual field defect is also seen in eyes with glau-
comatous optic neuropathy. Thus, a diffuse loss of nerve
fibers is compatible with normal visual acuity and little
or no visual field loss in both NA-AION and glaucoma-
tous optic neuropathy. This case shows that, while a
large optic disc cup protects the eye from suffering
severe visual loss, it does not completely protects it
from having a diffuse nerve fiber loss.
Refractive Error
It is interesting to consider whether refractive error has

any role to play in NA-AION. This can be tested by the
following two studies:
(a) Relationship between Refractive Error

and C/D Ratio
Tomlinson and Phillips [280] stated that the more
hypermetropic the eye, the smaller the C/D ratio, and
conversely the more myopic an eye, the larger the C/D
ratio. The Rotterdam Study [155], by contrast, showed
no association of C/D ratio with refractive error in the
general population. That was also true in another study
[157]. Our morphometric study [106] in fellow normal
eyes of patients with unilateral NA-AION showed that
for the horizontal C/D ratio, a higher degree of myopia
and higher degree of hyperopia were both significantly
associated with a larger C/D ratio (p=0.021 for myopia
and p=0.004 for hyperopia). A similar pattern was
Fig.14.13 Visual fields plotted with a Goldmann perimeter, of seen for the vertical C/D ratio in hyperopia (p=0.057)
the right (A) and left (B) eyes at the last clinic visit
but not in myopia (p=0.428). Thus, none of the studies
based on much larger populations supports the conclu-
was normal in both eyes and there was finally only a sion of Tomlinson and Phillips [280]. This indicates
mild degree of visual field defect in both eyes that refractive error does not contribute to NA-AION.
(Fig. 14.13). This is because the large cup acted as a (b) Refractive Error in Eyes with NA-AION:
protective mechanism, providing space for the swollen In our study of 608 consecutive NA-AION patients
axons to expand into, resulting in much less interference [106], there was no significant (p=0.289) difference in
in the optic nerve head circulation than would have hap- spherical equivalent refraction in eyes with NA-AION
pened had there been no cup. The other interesting fea- compared to that of an age matched general population
ture of this case is that the cup size increased markedly in the Framingham Eye Study cohort [275]. There was a
in both eyes (see Figs.14.11d and 14.12e), which clearly significant (p<0.0001) association between age and
indicates that there was a marked diffuse ischemic nerve refractive error but no difference between males and
fiber loss in both optic nerves, due to recurrent episodes females (p=0.976). Our study [106] contradicted the
of NA-AION in both eyes, rather than localized nerve findings of three previous studies based on 126 [157], 50
fiber loss (usually seen in classical NA-AION with no [281] and 37 [282] eyes respectively, which suggested
cup). I have sometimes seen similar enlargement of the that patients with NA-AION are more hypermetropic
cup in other NA-AION eyes with cups. In spite of than the control group. Of the three previous studies, the
marked enlargement of the optic disc cup, the presence study by Beck et al. [157] had the largest number of
eyes, i.e. 126 eyes with NA-AION. In that study, the a

mean age of the NA-AION group was 62years while
that of the control group was 37years; that marked age
difference accounts for the finding of more hypermetro-
pia in the NA-AION group than in the control group.
This is further supported by the fact that in that study
there was no difference in refraction between NA-AION
and arteritic AION eyes because those groups were
almost similar in age. The other two studies [281, 282]
had age-matched control groups but the sample size was
small; it is possible that with a much larger sample size
their findings might have been different.
b
Thus, one can conclude that refraction does not play
any role in NA-AION.
Optic Disc Drusen
There are several reports of NA-AION in eyes with

optic disc drusen [283292]. Most of them are anec-
dotal cases reports rather than a large series. NA-AION
in such cases can occur in young persons. In one report
it was see in a young boy of only 13 years [287]. There
are reports of NA-AION with drusen seen in persons
in their 20s [283285]. However, it can occur at any
age. I have seen some cases of NA-AION in eyes with Fig.14.14 Optic disc drusen in a 28-year-old woman in her left
optic disc drusen. I have found that visual field defect eye: (a) fundus photograph and (b) showing auto-fluorescence
in such cases may be progressive in nature. These of the drusen
patients may report episodes of transient visual blur-
ring or loss. As regards management of these cases,
unfortunately there is no definite treatment available.
Raised Intraocular Pressure
I have treated some of these patients with intraocular
pressure lowering therapy when they had a progressive
The blood flow in the optic nerve head capillaries
visual loss, which resulted in stable vision; however, it
depends upon the perfusion pressure in them (perfu-
is hard to be sure whether that was effective or not.
sion pressure=mean blood pressure minus intraocu-
Following are the various ways of diagnosing optic
lar pressure). A rise in intraocular pressure reduces
disc drusen.
the perfusion pressure and is an important risk factor
1. Ophthalmoscopy: The best way to see optic disc in the production of NA-AION in susceptible indi-
drusen on ophthalmoscopy is to indirectly illumi- viduals, e.g., in acute angle closure glaucoma, dur-
nate the optic disc by shining the ophthalmoscope ing the immediate post-operative period after cataract
light just at the edge of the rim, and the drusen then extraction [146], in neovascular glaucoma with ocu-
usually light up. lar ischemia due to internal carotid artery and/or
2. Fundus photography: The drusen are autofluores- ophthalmic artery stenosis or occlusion [144, 145],
cent. Therefore, they can be seen easily with a fluo- or in eyes with already low perfusion pressure in the
rescein fundus angiography camera without the vessels of the optic nerve head due to other causes.
injection of fluorescein (Fig.14.14). During surgical procedures under general anesthe-
3. Ultrasound: this is helpful in eyes with buried drusen. sia, inadvertent pressure on the eyeballs, with or
4. Computed tomographic scanning is also helpful in without systemic arterial hypotension, can produce
eyes with buried drusen. NA-AION.
Post-cataract Extraction-NA-AION (PCE-NA-AION) In contrast to the PCE-NA-AION discussed above, a

number of cases of ordinary NA-AION in aphakic eyes
This is a type that develops during the period immedi- have been described in the literature [294]. In all the
ately after a cataract extraction [146]. As discussed eyes, the so-called optic neuritis which developed
above, blood flow in the optic nerve head capillaries weeks, months, or even years after the cataract extraction
depends on the perfusion pressure, which is equal to the cannot be attributed to the latter. Reese and Carroll [295]
mean blood pressure minus the intraocular pressure. commented that, for 12 months after the extraction,
Thus, a fall of the perfusion pressure (caused by either a visual acuity was good (usually normal), and the same
fall of mean blood pressure or a rise of intraocular pres- seemed to be implied in the majority of other reports.
sure) below a critical level results in ischemia of the optic This is in sharp contrast to eyes with PCE-NA-AlON,
nerve head. A rise of intraocular pressure during the where NA-AION develops during the immediate post-
immediate postoperative period has been reported [146]. operative period, with none of the patients ever having a
A fall of blood pressure does occur during deep sedation, normal visual function after the cataract extraction and
sleep and prolonged general anesthesia. Thus, in suscep- before the onset of NA-AION. Thus, the so-called optic
tible persons, with serious cardiovascular disorders, and neuritis in aphakic eyes in the previous reports did not
in eyes with poor circulation in the optic nerve head, any represent PCE- NA-AION but, in fact, represents normal
transient rise of intraocular pressure combined with a fall NA-AION occurring independently in persons past
of blood pressure can easily compromise the circulation middle age; aphakic eyes are as vulnerable to NA-AION
in the optic nerve head and produce NA-AION. It could as phakic eyes. There is a high incidence of involvement
be argued that retrobulbar injection for cataract extrac- of the second eye by NA-AION in phakic eyes; similarly
tion could directly press on the PCAs and produce isch- Carroll [294] emphasized the frequent involvement of
emia. This seems unlikely because, first, the rise in orbital the second eye in aphakic NA-AION cases but errone-
pressure is extremely short-lived after the injection; sec- ously attributed this to cataract extraction.
ond, opening the eye for cataract extraction reduces the In conclusion, patients with unilateral NA-AION are
intraocular pressure to zero and would increase the per- at a significantly higher risk of developing NA-AION in
fusion pressure; and third, there is no evidence of choroi- the fellow eye after cataract extraction; this is because
dal ischemic lesions in these eyes. If retrobulbar most of the risk factors which made a person susceptible
hemorrhage develops in the orbit secondary to the injec- to NA-AION following cataract extraction in the first
tion, that can cause the intraocular pressure to rise to high eye still exist. This can be prevented if it is anticipated
levels, which in persons with poor circulation in the optic and adequate prophylactic measures are taken against
nerve head can result in optic nerve ischemia. the rise of intraocular pressure and fall of systemic blood
Patients with PCE-NA-AION in the first eye run a pressure. The rarity of this condition in clinical practice
high risk of development of PCE-NA-AION in the sec- may be more apparent than real because of lack of
ond eye [146, 293]. In the study by Lam etal. [293], awareness. Eyes with ocular ischemic syndrome are at
they compared the incidence of development of risk of developing NA-AION and their intraocular pres-
NA-AION in the fellow eye of patients who developed sure should be kept as low as possible; caution: in these
PCE-NA-AION in the first eye versus those who devel- patients, avoid giving beta-blocker eye drops to lower
oped NA-AION in the first eye without cataract extrac- intraocular pressure, because our study [296] showed
tion. They found that risk of NA-AION occurrence in that that can cause significant nocturnal arterial hypoten-
the fellow eye was 3.6-fold in the former compared to sion and may result in development of NA-AION in
the latter. The occurrence of PCE-NA-AION in the persons with predisposing risk factors. To avoid devel-
first eye should be a definite warning to the surgeon opment of NA-AION during surgical procedures under
that the second eye is equally vulnerable to NA-AION. general anesthesia, avoid any pressure on the eyeballs.
With patients who are going to have cataract extraction
on the second eye and who have had PCE-NA-AION
in the first eye, one must institute all possible prophy- Marked Optic Disc Edema
lactic measures to prevent post-extraction ocular hyper-
tension. That also applies to those who undergo other When there is marked optic disc edema due to raised
surgical procedures associated with prolonged general intracranial pressure [277], transient obscuration of
anesthesia and/or systemic arterial hypotension. vision may be the presenting or a prominent feature.
The attacks may be brought about by a change of pos- symptoms resembled those seen with orthostatic
ture, sudden movement of the eye, stooping or rubbing hypotension. This change of posture produces a tran-
the eyes; these are due to a transient fall in perfusion sient fall of blood pressure and lowers the mean blood
pressure caused by either a rise in intraocular pressure pressure in the prelaminar capillaries, resulting in a
or fall of systemic blood pressure. The transient obscu- transient ischemia. Similarly, any factor causing a
ration may progress to development of NA-AION. riseof intraocular pressure, e.g., rubbing of the eyes,
The mechanism of development of transient obscu- stooping, a strong contraction of the lids, or strong
ration and the development of NA-AION in such cases contraction of the extraocular muscles, can produce
is as follows [297]: The normal prelaminar region of transient ischemia of the prelaminar region in a previ-
the optic nerve head lies in a restricted space, sur- ously existing state of circulatory stasis. A prolonged
rounded by an unyielding Bruchs membrane and fall of perfusion pressure in the optic nerve head capil-
contains nerve fibers, the intervening glial septa and laries (as for example due to nocturnal arterial hypoten-
the fine vessels of this region lying in the septa (see sion during sleep) finally results in development of
Chap. 2). Swelling of the optic disc in optic disc edema NA-AION. Thus, marked optic disc edema predis-
is primarily due to the swelling of the nerve fibers as a poses the eye to NA-AION.
result of axoplasmic flow stasis in raised intracranial It could be argued that axoplasmic flow stasis may
pressure [298, 299]. As discussed above, this swelling have some role in the development of visual distur-
of the axons in a restricted prelaminar region com- bances and optic atrophy in optic disc edema.
presses the fine vessels in this region; those fine ves- Axoplasmic flow itself is not concerned with transmis-
sels are likely to be the first victims, particularly the sion of a visual impulse [300] the two are indepen-
fine venules and the capillaries [277]. Compression of dent processes. This is very well demonstrated by optic
the blood vessels in the optic nerve head may be fur- disc edema in raised intracranial pressure, which,
ther aggravated by extracellular fluid accumulation though due to axoplasmic flow stasis [299], is not
due to secondary vascular changes caused by marked accompanied by any detectable axonal dysfunction per
optic disc edema. That this does happen is suggested se. There are other similar examples of optic disc
by our fluorescein fundus angiographic studies in the edema due to axoplasmic flow stasis without any visual
eyes with optic disc edema. These revealed circulatory dysfunction. Therefore, as discussed elsewhere [277],
stasis in the prelaminar region in eyes with moderate axoplasmic flow stasis per se cannot produce visual
and marked optic disc edema. In the normal optic disc, disturbances. There is not much evidence to suggest
the prelaminar capillaries start to fill either before the the possibility that long-term axoplasmic flow stasis
retinal arteries or during the very early part of the reti- may produce degeneration of the axons and thus optic
nal arterial phase, but in eyes with optic disc edema atrophy. The overwhelming evidence favors ischemia
they showed no filling until either the later stage of the as the major factor in the development of transient
retinal arterial phase or even later; this was particularly obscuration, which progresses to NA-AION and later
noticeable in eyes with moderate and marked optic on optic atrophy in some cases.
disc edema. The circulation in the prelaminar part of
the optic nerve head, as in other intraocular vascula-
ture, depends on the perfusion pressure in the capillar- Location of the Watershed Zone of the PCAs
ies. In the case of a swollen optic disc, the tissue in Relation to the Optic Disc
pressure surrounding the prelaminar capillaries is high.
In light of this evidence, it is logical that when the cir- This is discussed at length in Chap. 3. This almost uni-
culatory stasis develops in the prelaminar capillaries, a versally ignored factor plays a key role in the develop-
critical balance may exist between the blood pressure ment of optic nerve head ischemia. It is well-established
in the capillaries and the surrounding tissue pressure. that in the event of fall of perfusion pressure, a water-
Any factor that causes an increase in the intraocular shed zone, being an area of comparatively poor vascu-
pressure or optic nerve head tissue pressure, or a fall of larity, is most vulnerable to ischemia. For example, the
the blood pressure in the capillaries, would induce development of cerebral watershed zone infarcts fol-
transient obscuration. In most of these patients, attacks lowing a fall of perfusion pressure in the cerebral
of transient obscuration are brought on by suddenly arteries is well known. Similarly, in the event of a fall
standing up from a sitting position, and the visual in perfusion pressure in the choroidal vascular bed
Fig.14.15 Diagrammatic representation of some of the locations of the watershed zones (shaded area) between the medial and
lateral PCAs in human eyes (Reproduced from Hayreh [300])
supplied by the PCAs, the part of the optic disc located Vascular Disorders of the Nutrient Vessels
in the watershed zone is more susceptible to ischemia of the Optic Nerve Head
than the rest [98] (Fig. 14.15, see Figs. 3.39, 3.41,
3.45, 3.52, 3.53, 3.54, 3.563.59). As discussed in As detailed in Chap. 5, defective autoregulation, vasos-
Chap. 3, the location of the watershed zone in relation pasm, arteriosclerosis, atherosclerosis, and other vas-
to the optic nerve head shows marked inter-individual cular disorders of the ophthalmic artery and/or PCAs
variation. If the optic disc is situated away from the may make the optic nerve head susceptible to ischemic
watershed zone (Fig.14.15a), in the event of a fall in disorders [110, 141].
perfusion pressure in the choroidal vascular bed, it is
relatively safe; however, if the entire disc lies on the
watershed zone (Fig.14.15c), it is most vulnerable to Deranged Autoregulation of Optic Nerve Head
ischemic damage. Thus, in the event of a fall in perfu- Blood Flow
sion pressure in the PCA (choroidal vascular bed) and
development of a watershed zone filling defect, the The important role played by autoregulation of
vulnerability of the optic nerve head to ischemia blood flow in the optic nerve head is fully discussed
depends upon its relationship to the watershed zone in Chap. 5. Its derangement makes an optic nerve
filling defect. That evidently must determine the extent head vulnerable to development of ischemic disor-
and site of visual loss. ders. Unfortunately, we have no means to evaluate
Neural Ischemia Not an All or None Phenomenon 307
that, which has resulted in controversy about the none phenomenon. There is a whole spectrum of
pathogenesis of NA-AION [100]. ischemia, varying from very mild to severe: the former
producing only subtle physiological derangement of
some of the neural components, while the latter results
Absence of Blood-Optic Nerve Head Barrier in immediate, complete infarction of all neural tissues.
in the Prelaminar Region Thus, the concept that ischemia always produces
infarction is simply not correct. Moreover, there is
As discussed in Chap. 4, there is evidence that the a good deal of evidence available that different compo-
absence of a blood-optic nerve head barrier in the nents of neural tissues have a different susceptibility to
prelaminar region plays a role in the development of ischemic damage. The subject was discussed at length
hypertensive optic neuropathy [173] and methanol elsewhere [278]. Briefly, in AION, one end of the
toxic optic neuropathy [302] by the mechanism dis- spectrum is very mild ischemia, which in the optic
cussed there. It is quite possible that in some cases of nerve head produces only axoplasmic flow stasis, and
NA-AION, the absence of a blood-optic nerve barrier consequently optic disc swelling and incipient
in the prelaminar region may also be playing a role. NA-AION [113, 278], without interfering with the
For example, from time to time I have come across transmission of a visual impulse along the axon (hence
patients (particularly persons with type A personality no visual loss); while at the other extreme is the severe
[32]) who complained of developing of blurring of ischemia that causes infarction of all the neural tissues
vision and even NA-AION when exposed to a stressful in the optic nerve head, with swollen optic disc and
situation. There is a huge volume of literature showing complete loss of vision, as seen in arteritic AION with
the association between stress and the release of complete thrombotic occlusion of the PCA. There is a
catecholamines (norepinephrine and epinephrine). whole wide range between the two extremes. Thus,
Norepinephrine and epinephrine are vasoconstrictors, different grades of ischemia in the optic nerve head
which can leak from the choriocapillaris into the chor- have different effects and manifestations, and the clini-
oidal interstitial tissue, and through the Border tissue cal entity anterior ischemic optic neuropathy encom-
of Elschnig into the prelaminar region, and result in passes the entire range. In malignant hypertension, the
optic nerve head ischemia, by the mechanism dis- ischemic damage to the optic nerve head may vary
cussed above in hypertensive optic neuropathy. widely with different sequelae so that hypertensive
optic neuropathy, which in fact is NA-AION [169],
would manifest in different ways in different eyes.
Posterior Vitreous Detachment This wide spectrum has importance from the point
of view of the management of AION. A mild isch-
Some seem to imply that the presence or absence of emia results in impaired axonal function (visual loss)
posterior vitreous detachment plays a role in the devel- without the death of the axon, and given a chance it
opment of NA-AION. In our study [303], evaluation of can recover its function, as shown by the spontaneous
posterior vitreous detachment in patients with visual improvement, without any treatment, in at least
NA-AION showed that in patients with uniocular 40% of NA-AION eyes recorded by two large pro-
NA-AION, there was no significant difference in pos- spective studies [304, 305]. It is like a starving person
terior vitreous detachment between the affected eye who is unable to work due to starvation but is not
and the fellow normal eye. dead, who, given food, can recover to his normal level
of functioning. By contrast, severe ischemia produces
immediate death of neural tissue, as in arteritic AION
where there is no recovery of visual function [306,
Neural Ischemia Not an All or None 307]. Knowledge of the initial loss of visual function
Phenomenon does not give us any information about the transient or
permanent nature of that loss of function. The clinical
In any consideration of the effect of ischemia on the picture of AION thus can vary widely, reflecting this
neural tissue, it is of fundamental importance to appre- wide spectrum in severity of ischemia. Lack of appre-
ciate the fact that neural ischemia is not an all or ciation of this simple fact has caused marked
confusion on the subject of neural ischemia and 5. Guyer DR, Green WR, Schachat AP, Bastacky S, Miller NR.
AION, and posed a problem in its management. Thus, Bilateral ischemic optic neuropathy and retinal vascular occlu-
sions associated with lymphoma and sepsis. Clinicopathologic
when there is hypoxia of axons, as happens in most correlation. Ophthalmology. 1990;97(7):8828.
NA-AION cases, the use of the term infarction in 6. Knox DL, Kerrison JB, Green WR. Histopathologic studies
not justified. of ischemic optic neuropathy. Trans Am Ophthalmol Soc.
2000;98:20320; discussion 221222.
7. Goerlitz M. Histologische Untersuchung eines Falles von
Erblindung nach schwerem Blutverlust [Histologic Exami
nation of Blindness after Intense Loss of Blood]. Klin
Conclusion Monatsbl Augenheilkd. 1920;64:76382.
8. Greenfield JG. A case of giant-cell arteritis. Proc R Soc
Med. 1951;44:8557.
From this brief discussion about the pathogenesis of 9. Kreibig W. Optikomalazie, die Folge eines Gefassverschlusses
NA-AION and about the various factors that influ- im retrobulbren Abschnitt des Sehnerven [Opticomalacia
ence the blood flow in the optic nerve head (see caused by vascular occlusion in the retrobulbar portion of
Chap. 5), the great complexity of mechanisms of optic nerves]. Klin Monatsbl Augenheilkd. 1953;122(6):
71931.
development of NA-AION becomes clear. A whole 10. Crompton MR. The visual changes in temporal (giant-cell)
host of systemic and local factors, acting in differ- arteritis. Report of a case with autopsy findings. Brain.
ent combinations and to different extents, may 1959;82:37790.
derange the optic nerve head circulation, with some 11. Spencer WH, Hoyt WF. A fatal case of giant-cell arteritis
(temporal or cranial arteritis) with ocular involvement. Arch
making the optic nerve head susceptible to isch- Ophthalmol. 1960;64:8627.
emia and others acting as the final insult. Nocturnal 12. Rdenhauser JH. ber pathologisch-anatomische augenver-
hypotension seems to be an important precipitating aenderungen bei generalisierter riesenzellarteriitis. Klin
factor in the susceptible patient. The quest for a Monatsbl Augenheilkd. 1964;145:41429.
13. Manschot WA. A fatal case of temporal arteritis with ocular
single villain that causes NA-AION is a mistaken symptoms. Ophthalmologica. 1965;149:12130.
and misleading endeavor, which has resulted in 14. Henkind P, Charles NC, Pearson J. Histopathology of isch-
controversy and confusion. Contrary to the com- emic optic neuropathy. Am J Ophthalmol. 1970;69(1):
mon misconception, NA-AION is not a thromboem- 7890.
15. Hinzpeter EN, Naumann G. Ischemic papilledema in giant-
bolic disorder in the vast majority of cases (see cell arteritis. Mucopolysaccharide deposition with normal
above) but a hypotensive disorder. Thus, the patho- intraocular pressure. Arch Ophthalmol. 1976;94(4):6248.
genesis of NA-AION is complex but not, as often 16. Rootman J, Butler D. Ischaemic optic neuropathy: a com-
stated, unknown. bined mechanism. Br J Ophthalmol. 1980;64(11):82631.
17. Burde RM, Smith ME, Black JT. Retinal artery occlusion in
In the vast majority, NA-AION is due to ischemic the absence of a cherry red spot. Surv Ophthalmol.
insult in the optic nerve head supplied by the PCA 1982;27(3):1816.
circulation; however, rarely it can be caused by acute 18. Johnson MW, Kincaid MC, Trobe JD. Bilateral retrobulbar
ischemia of the axial part of the anterior optic nerve optic nerve infarctions after blood loss and hypotension.
A clinicopathologic case study. Ophthalmology. 1987;94(12):
supplied by the intraneural branches of the central 157784.
retinal artery (see above). 19. Stefani FH, Asiyo MN, Mehraein P, Eder M. Histopathologie
von Netzhaut, Fasciculus opticus und Corpus geniculatum
laterale bei langer bestehender, symmetrisch bilateraler
ischamischer Schnabelscher kavernoser Optikusatrophie.
[Histopathology of the retina, optic fascicle and lateral
References geniculate body in chronic, bilateral symmetric ischemic
Schnabels cavernous optic atrophy]. Klin Monatsbl
1. Uhthoff W. Zu den entzndlichen Sehnerven-affektionen bei Augenheilkd. 1990;197(2):1625.
Arteriosklerose (Atherosklerose). Ber Zusammenkunft 20. Kubota T, von Below H, Holbach LM, Naumann GO.
Dtsch Ophthalmol Ges. 1924;44:196208. Bilateral cavernous degeneration of the optic nerve associ-
2. Franois J, Verriest G, Baron A. Pseudopapillites vasculai- ated with multiple arteriosclerotic ischemic infarctions.
res. Acta Ophthalmol (Copenh). 1957;35:3252. Graefes Arch Clin Exp Ophthalmol. 1993;231(1):525.
3. Knox DL, Duke JR. Slowly progressive ischemic optic neu- 21. Ellenberger Jr C, Netsky MG. Infarction in the optic nerve. J
ropathy. A clinicopathologic case report. Trans Am Acad Neurol Neurosurg Psychiatry. 1968;31(6):60611.
Ophthalmol Otolaryngol. 1971;75(5):10658. 22. Levin LA, Louhab A. Apoptosis of retinal ganglion cells in
4. Lieberman MF, Shahi A, Green WR. Embolic ischemic anterior ischemic optic neuropathy. Arch Ophthalmol.
optic neuropathy. Am J Ophthalmol. 1978;86(2):20610. 1996;114(4):48891.
References 309
23. Tesser RA, Niendorf ER, Levin LA. The morphology of an lary edema]. Arch Ophtalmol Rev Gn Ophtalmol.
infarct in nonarteritic anterior ischemic optic neuropathy. 1950;10(6):70114.
Ophthalmology. 2003;110(10):20315. 46. Carroll FD. Optic neuritis; a 15year study. Am J Ophthalmol.
24. Hayreh SS. A study of the central artery of the retina in 1952;35(1):7582.
human beings. Thesis for the Master of Surgery. India: 47. Sanna M. Pseudopapilliti vascolari. Riv Oto-neuro-oftalmol.
Panjab University; 1958. 1958;33:44970.
25. Singh Hayreh S, Dass R. The central artery of the retina. II. 48. Stock W. ber kavernose Sehnervenatrophie bei Myopie.
A study of its distribution and anastomoses. Br J Ophthalmol. Klin Monatsbl Augenheilkd. 1908;46(1):34258.
1960;44:28099. 49. Henschen SE. ber circumscripte arteriosklerotische
26. Hayreh SS. Anatomy and physiology of the optic nerve head. Nekrosen (Erweichungen) in dem Sehnerven, im Chiasma
Trans Am Acad Ophthalmol Otolaryngol. 1974;78:OP24054. und in dem Tractus. Albrecht Von Graefes Arch Ophthalmol.
27. Hayreh SS, Zimmerman MB, Podhajsky P, Alward WL. 1911;78:21223.
Nocturnal arterial hypotension and its role in optic nerve 50. Beselin O. Subakute Funktionsstrung des Sehnerven bezie-
head and ocular ischemic disorders. Am J Ophthalmol. hungsweise der Netzhaut durch Arteriosklerose. Klin
1994;117(5):60324. Monatsbl Augenheilkd. 1925;75:3638.
28. Hayreh SS, Podhajsky P, Zimmerman MB. Role of noctur- 51. Loewenstein A. Cavernous degeneration, necrosis and other
nal arterial hypotension in optic nerve head ischemic disor- regressive processes in the optic nerve with vascular disease
ders. Ophthalmologica. 1999;213(2):7696. of the eye. Arch Ophthalmol. 1945;34:2206.
29. Hayreh SS. Role of nocturnal arterial hypotension in the 52. Dawson BH. On the blood vessels of the human optic chi-
development of ocular manifestations of systemic arterial asma, hypophysis and hypothalamus. M.D. Thesis,
hypertension. Curr Opin Ophthalmol. 1999;10(6):47482. Manchester; 1948.
30. Hayreh SS, Podhajsky PA, Zimmerman B. Nonarteritic 53. Franois JGV, DeRouck A. Llectro-oeulographie en tant
anterior ischemic optic neuropathy: time of onset of visual quexamen functionnel de la rtine [Electro-oculography in
loss. Am J Ophthalmol. 1997;124(5):6417. functional examination of the retina]. Adv Ophthalmol.
31. Hayreh SS. Anterior ischemic optic neuropathy. Arch 1957;7:164.
Neurol. 1981;38(11):6758. 54. Leopold M. Les signes ophtalmologiques de latherosclerose
32. Hayreh SS. Acute ischemic disorders of the optic nerve: [Ophthalmological signs of atherosclerosis]. J Md Lyon.
Pathogenesis, clinical manifestations and management. 1959;40:68391.
Ophthalmol Clin North Am. 1996;9(3):40742. 55. Calmettes L, Deodati F, Bechac G. Pseudopapillite vascu-
33. Hayreh SS. Ischemic optic neuropathy. Prog Retin Eye Res. laire. Rev Otoneuroophtalmol. 1963;35:645.
2009;28(1):3462. 56. Haab O. In: Norris WF, Oliver C, editors. System of the diseases
34. Hayreh SS, Baines JA. Occlusion of the posterior ciliary of the eye, vol. IV. London: Lippincott; 1900. p. 481537.
artery. III. Effects on the optic nerve head. Br J Ophthalmol. 57. Abelsdorff G. Sehnervenatrophie durch athero-sklerotischen
1972;56(10):75464. Verschluss der Zentralarterie. Ophthalmologica. 1924;
35. Hayreh SS. Blood supply of the optic nerve head and its role 52:2736.
in optic atrophy, glaucoma, and oedema of the optic disc. 58. Rucker CW. Circulatory disorders. Trans Am Acad
Br J Ophthalmol. 1969;53(11):72148. Ophthalmol Otolaryngol. 1956;60(1):602.
36. Skillern PG, Lockhart G. Optic neuritis and uncontrolled diabe- 59. Betteto G. Modificazioni strutturali dell arteria centrale
tes mellitus in 14 patients. Ann Intern Med. 1959;51:46875. della retina e delle arteriole del nervo ottico in rapporto
37. Kadlecov V. Oedema papillae ischaemicum [Ischemic pap- allarterio ed arteriosclerosi renale. Annali di Ottalmologia.
illary edema]. Cesk Oftalmol. 1951;7(6):38894. 1958;84:6179.
38. Peters W. ber die Neuritis optici arteriosklerotischer 60. Smith C, Greene PB. Ocular complications of temporal
Genese. Klin Monatsbl Augenheilkd Augenrztl Fortbild. arteritis. Am J Ophthalmol. 1949;32(5):68590.
1958;132(3):36377. 61. Bregeat P. Loedma papillaire. Bull Soc Fran Ophtal.
39. Fuchs E. ber senile Vernderungen des Sehnerven. 1956:441472.
Albrecht Von Graefes Arch Ophthalmol. 1920;103:30430. 62. Igersheimer J. Serious visual troubles caused by unusual cir-
40. Igersheimer J. Sehnervenerkrankung bei maligner Sklerose. culatory disturbances in the region of the optic nerve. Am J
Zeitschrift fr Augenheilkunde. 1929;69:4759. Ophthalmol. 1959;47(6):76772.
41. Igersheimer J. General hypertension and peripheral optic 63. Foulds WS. Ischaemic optic neuropathy. In: Cant JS, editor.
pathways. Am J Ophthalmol. 1940;23:124354. Proceedings, William Mackenzie Centenary Symposium on
42. von Stief A. Zur Histopathologie des Sehnerven bei the ocular circulation in health and disease (1968: Royal
Arteriosklerose und bei der senilen Demenz. College of Physicians and Surgeons, Glasgow). London:
Ophthalmologica. 1930;70:4152. Kimpton; 1968. p. 13641.
43. Rintelen F. ber arteriosklerotische Opticusatrophie. 64. Meadows SP. Temporal arteritis and the loss of vision. Trans
Ophthalmologica. 1946;111:28590. Ophthalmol Soc UK. 1954;74:1324.
44. Kurz O. ber Papillitis arteriosclerotica. Ophthalmologica. 65. Sanders MD. Ischaemic papillopathy. Trans Ophthalmol Soc
1948;116(45):2815. UK. 1971;91:36986.
45. Bessiere E, Julien RG. Les manifestations papillaires dans le 66. Meadows SP. Temporal or giant cell arteritis: ophthalmic
syndrome dit arterite temporale, frequence de loedeme aspects. In: Smith JL, editor. Neuro-ophthalmology: sympo-
papillaire ischemique [Papillary manifestations in the syn- sium of the University of Miami and the Bascom Palmer Eye
drome called temporal arteritis; frequency of ischemic papil- Institute, vol. 4. St. Louis: Mosby; 1968. p. 14857.
67. Franois J. Les pseudo-papillites vasculaires. Rapport Soc 89. Cullen JF. Ischaemic optic neuropathy. Trans Ophthalmol
Fran Ophthal. 1956. Soc UK. 1967;87:75974.
68. Franois J, Verriest G, Neetens A, De Rouck A. 90. Arnold AC. Pathogenesis of nonarteritic anterior ischemic
Pseudopapillites vasculaires [Vascular pseudo-papillitis.]. optic neuropathy. J Neuroophthalmol. 2003;23(2):15763.
Bull Soc Belge Ophtalmol. 1961;129:41327. 91. Hayreh SS. Inter-individual variation in blood supply of the
69. Franois J, Verriest G, Neetens A, de Rouck A, Hanssens M. optic nerve head. Its importance in various ischemic disor-
Pseudonvrites optiques oedmateuses dorigine vasculaire ders of the optic nerve head, and glaucoma, low-tension
[Vascular pseudopapillitis]. Ann Ocul (Paris). 1962; glaucoma and allied disorders. Doc Ophthalmol.
195:83085. 1985;59(3):21746.
70. Franois J, Neetens A. Physio-anatomy of the axial vascu- 92. Wedel C, Ruysch F. Epistola anatomica, problematica ad
larization of the optic nerve. Doc Ophthalmol. 1969;26: Fredericum Ruyschium. De oculorum tunicis.
3849. Amstelaedami: Janssonio-Waesbergios; 1737.
71. Piper HF, Unger L. Hemianopsia horizontalis inferior bei 93. Hayreh SS, Baines JA. Occlusion of the posterior ciliary
akuten Durchblutungsstorungen des Sehnerven. artery. I. Effects on choroidal circulation. Br J Ophthalmol.
Ophthalmologica. 1957;134(3):16980. 1972;56(10):71935.
72. Ponte F. La pseudopapillite vascolare degli arteriosclerot- 94. Hayreh SS. The choriocapillaris. Albrecht Von Graefes
ici. Aspetti clinici e consideraziioni patogenetiche. G Ital Arch Klin Exp Ophthalmol. 1974;192(3):16579.
Oftalmol. 1964;17(3):17091. 95. Hayreh SS. Submacular choroidal vascular pattern.
73. Franois J, Neetens A. Vascularization of the optic pathway. Experimental fluorescein fundus angiographic studies.
I. Lamina cribrosa and optic nerve. Br J Ophthalmol. Albrecht Von Graefes Arch Klin Exp Ophthalmol.
1954;38(8):47288. 1974;192(3):18196.
74. Franois J, Neetens A, Collette JM. Vascular supply of the 96. Hayreh SS. The long posterior ciliary arteries. An experi-
optic pathway. II. Further studies by micro-arteriography mental study. Albrecht Von Graefes Arch Klin Exp
of the optic nerve. Br J Ophthalmol. 1955;39(4): Ophthalmol. 1974;192(3):197213.
22032. 97. Hayreh SS. Physiological anatomy of the choroidal vascu-
75. Franois J, Neetens A. Vascularization of the optic pathway. lar bed. Int Ophthalmol. 1983;6:8593.
III. Study of intra-orbital and intracranial optic nerve by 98. Hayreh SS. In vivo choroidal circulation and its watershed
serial sections. Br J Ophthalmol. 1956;40(1):4552. zones. Eye. 1990;4(Pt 2):27389.
76. Franois J, Neetens A. Central retinal artery and central 99. Hayreh SS. Posterior ciliary artery circulation in health and
optic nerve artery. Br J Ophthalmol. 1963;47(1):2130. disease: the Weisenfeld lecture. Invest Ophthalmol Vis Sci.
77. Hayreh SS. The ophthalmic artery: III. Branches. Br J 2004;45(3):74957; 748.
Ophthalmol. 1962;46(4):21247. 100. Hayreh SS. Evaluation of optic nerve head circulation: review
78. Hayreh SS. Blood supply and vascular disorders of the optic of the methods used. J Glaucoma. 1997;6(5):31930.
nerve. An Inst Barraquer. 1963;4:7109. 101. Hayreh SS, Bill A, Sperber GO. Effects of high intraocular
79. Hayreh SS. The central artery of the retina. Its role in the pressure on the glucose metabolism in the retina and optic
blood supply of the optic nerve. Br J Ophthalmol. nerve in old atherosclerotic monkeys. Graefes Arch Clin
1963;47:65163. Exp Ophthalmol. 1994;232(12):74552.
80. Singh Hayreh S, Dass R. The central artery of the retina. I. 102. Hayreh SS, Piegors DJ, Heistad DD. Serotonin-induced
Origin and course. Br J Ophthalmol. 1960;44:193212. constriction of ocular arteries in atherosclerotic monkeys.
81. Beauvieux J, Ristitch K. Les vaisseaux centraux du nerf Implications for ischemic disorders of the retina and optic
optique. Etude anatomique. Arch Ophthalmol (Paris). nerve head. Arch Ophthalmol. 1997;115(2):2208.
1924;41:35269. 103. Kim TW, Kang KB, Choung HK, Park KH, Kim DM. Elevated
82. Blunt MJ, Steele EJ. The blood supply of the optic nerve and glutamate levels in the vitreous body of an invivo model of
chiasm in man. J Anat. 1956;90(4):48693. optic nerve ischemia. Arch Ophthalmol. 2000;118(4):5336.
83. Franois J. Vascularization of the optic nerve. Arch 104. Lotery AJ. Glutamate excitotoxicity in glaucoma: truth or
Ophthalmol. 1977;95(3):520. fiction? Eye. 2005;19(4):36970.
84. Franois J, Fryczkowski A. Microcirculation of the anterior 105. Kwon YH, Rickman DW, Baruah S, Zimmerman MB, Kim
part of the optic nerve. Ophthalmologica. 1977;175(4): CS, Boldt HC, etal. Vitreous and retinal amino acid con-
2229. centrations in experimental central retinal artery occlusion
85. Miller GR, Smith JL. Ischemic optic neuropathy. Am J in the primate. Eye. 2005;19(4):45563.
Ophthalmol. 1966;62(1):10315. 106. Hayreh SS, Zimmerman MB. Nonarteritic anterior isch-
86. Boghen DR, Glaser JS. Ischaemic optic neuropathy. The emic optic neuropathy: refractive error and its relationship
clinical profile and history. Brain. 1975;98(4):689708. to cup/disc ratio. Ophthalmology. 2008;115(12):227581.
87. Paillas J-E, Guillot P, Duplay J. Six observations de neuro- 107. Levin LA, Danesh-Meyer HV. Hypothesis: a venous etiol-
papillites aiges. Exploration cu chiasma optique; constata- ogy for nonarteritic anterior ischemic optic neuropathy.
tions opratoires; intrt du traitement neuro-chirurgical. Arch Ophthalmol. 2008;126(11):15825.
Bull Soc Ophtalmol Fr. 1949;7:20210. 108. Hayreh SS. Pathogenesis of nonarteritic anterior ischemic
88. Birkhead NC, Wagener HP, Shick RM. Treatment of tempo- optic neuropathy. Arch Ophthalmol. 2009;127(8):10823.
ral arteritis with adrenal corticosteroids; results in fifty-five 109. Somers VK, Dyken ME, Clary MP, Abboud FM.
cases in which lesion was proved at biopsy. J Am Med Sympathetic neural mechanisms in obstructive sleep apnea.
Assoc. 1957;163(10):8217. J Clin Invest. 1995;96(4):1897904.
References 311
110. Hayreh SS. Factors influencing blood flow in the optic 129. Hayreh SS, Joos KM, Podhajsky PA, Long CR. Systemic
nerve head. J Glaucoma. 1997;6(6):41225; 1998;7(1):71. diseases associated with nonarteritic anterior ischemic
111. Hayreh SS, Fraterrigo L, Jonas J. Central retinal vein occlu- optic neuropathy. Am J Ophthalmol. 1994;118(6):76680.
sion associated with cilioretinal artery occlusion. Retina. 130. Hayreh SS. Scientific literature and gospel truth. Indian J
2008;28(4):58194. Ophthalmol. 2000;48(2):939.
112. Hayreh SS, Zimmerman MB. Nonarteritic anterior isch- 131. Levin LA. Pathogenesis of nonarteritic anterior ischemic
emic optic neuropathy: clinical characteristics in diabetic optic neuropathy, authors reply. Arch Ophthalmol.
patients versus nondiabetic patients. Ophthalmology. 2009;127(8):10834.
2008;115(10):181825. 132. Hayreh SS, Zahoruk RM. Anterior ischemic optic neuropa-
113. Hayreh SS, Zimmerman MB. Incipient nonarteritic ante- thy. VI. In juvenile diabetics. Ophthalmologica. 1981;
rior ischemic optic neuropathy. Ophthalmology. 2007;114 182(1):1328.
(9):176372. 133. Hayreh SS. Anterior ischaemic optic neuropathy. I.
114. Hayreh SS. Central retinal vein occlusion. Ophthalmol Clin Terminology and pathogenesis. Br J Ophthalmol. 1974;58
North Am. 1998;11:55990. (12):95563.
115. Hayreh SS, Podhajsky P, Zimmerman MB. Central and 134. Hayreh SS. Anterior ischemic optic neuropathy. New York:
hemi-central retinal vein occlusion: role of anti-platelet Springer; 1975.
aggregation agents and anticoagulants. Ophthalmology 135. Hayreh SS, Chopdar A. Occlusion of the posterior ciliary
2011;118(8). artery. V. Protective influence of simultaneous vortex vein
116. Hayreh SS, Zimmerman MB, Podhajsky P. Hematologic occlusion. Arch Ophthalmol. 1982;100(9):148191.
abnormalities associated with various types of retinal vein 136. Newman NJ. Neuro-ophthalmology and systemic disease
occlusion. Graefes Arch Clin Exp Ophthalmol. 2002; Part I. An annual review (1994). J Neuroophthalmol.
240(3):18096. 1995;15(2):10921.
117. Hayreh SS. Management of central retinal vein occlusion. 137. MacMichael IM, Cullen JF. Pathology of ischaemic optic
Ophthalmologica. 2003;217(3):16788. neuropathy. In: Cant JS, editor. Proceedings, 2nd William
118. Hayreh SS. Venous occlusive disease: management Mackenzie Symposium on the optic nerve. London:
25years ago. Retina. 2006;26(6 Suppl):S516262. Kimpton; 1972. p. 10816.
119. Hayreh SS. The role of aspirin in non-arteritic anterior 138. Cogan DG. Neurology of the visual system. Springfield:
ischaemic optic neuropathy. Neuroophthalmology. 2010;34: Charles C. Thomas; 1966. p. 137, 173, 185188.
15. 139. McLeod D, Marshall J, Kohner EM. Role of axoplasmic
120. Dagenais GR, Yi Q, Lonn E, Sleight P, Ostergren J, Yusuf transport in the pathophysiology of ischaemic disc swell-
S. Impact of cigarette smoking in high-risk patients partici- ing. Br J Ophthalmol. 1980;64(4):24761.
pating in a clinical trial. A substudy from the Heart 140. Hayreh SS. Anterior ischaemic optic neuropathy II. Fundus
Outcomes Prevention Evaluation (HOPE) trial. Eur J on ophthalmoscopy and fluorescein angiography. Br J
Cardiovasc Prev Rehabil. 2005;12(1):7581. Ophthalmol. 1974;58(12):96480.
121. Newman NJ, Scherer R, Langenberg P, Kelman S, Feldon 141. Hayreh SS. Blood flow in the optic nerve head and factors
S, Kaufman D, etal. The fellow eye in NAION: report from that may influence it. Prog Retin Eye Res. 2001;20(5):
the ischemic optic neuropathy decompression trial follow- 595624.
up study. Am J Ophthalmol. 2002;134(3):31728. 142. Hayreh SS, Revie IH, Edwards J. Vasogenic origin of visual
122. Hayreh SS, Jonas JB, Zimmerman MB. Nonarteritic ante- field defects and optic nerve changes in glaucoma. Br J
rior ischemic optic neuropathy and tobacco smoking. Ophthalmol. 1970;54(7):46172.
Ophthalmology. 2007;114(4):8049. 143. Foulds WS. Visual disturbances in systemic disorders.
123. Botelho PJ, Johnson LN, Arnold AC. The effect of aspirin Optic neuropathy and systemic disease. Trans Ophthalmol
on the visual outcome of nonarteritic anterior ischemic Soc UK. 1969;89:12546.
optic neuropathy. Am J Ophthalmol. 1996;121(4):4501. 144. Hayreh SS, Podhajsky P. Ocular neovascularization with
124. Beck RW, Hayreh SS, Podhajsky PA, Tan ES, Moke PS. retinal vascular occlusion. II. Occurrence in central and
Aspirin therapy in nonarteritic anterior ischemic optic neu- branch retinal artery occlusion. Arch Ophthalmol.
ropathy. Am J Ophthalmol. 1997;123(2):2127. 1982;100(10):158596.
125. Salomon O, Huna-Baron R, Kurtz S, Steinberg DM, 145. Mizener JB, Podhajsky P, Hayreh SS. Ocular ischemic syn-
Moisseiev J, Rosenberg N, etal. Analysis of prothrombotic drome. Ophthalmology. 1997;104(5):85964.
and vascular risk factors in patients with nonarteritic ante- 146. Hayreh SS. Anterior ischemic optic neuropathy. IV.
rior ischemic optic neuropathy. Ophthalmology. 1999; Occurrence after cataract extraction. Arch Ophthalmol.
106(4):73942. 1980;98(8):14106.
126. Hayreh SS. Risk factors in AION. Ophthalmology. 147. Hayreh SS. Management of non-arteritic anterior ischemic
2001;108(10):17178. optic neuropathy. Graefes Arch Clin Exp Ophthalmol.
127. Abu-Amero KK, Bosley TM. Prothrombotic and athero- 2009;247(12):1595600.
sclerotic risk factors lack significance in NAION patients 148. Ellenberger Jr C, Keltner JL, Burde RM. Acute optic neu-
harbouring mitochondrial DNA mutations. Br J Ophthalmol. ropathy in older patients. Arch Neurol. 1973;28(3):
2006;90(1):11920. 1825.
128. Hayreh SS. Non-arteritic anterior ischemic optic neuropa- 149. Eagling EM, Sanders MD, Miller SJ. Ischaemic papillopa-
thy and thrombophilia. Graefes Arch Clin Exp Ophthalmol. thy. Clinical and fluorescein aniographic review of forty
2009;247(5):57781. cases. Br J Ophthalmol. 1974;58(12):9901008.
150. Repka MX, Savino PJ, Schatz NJ, Sergott RC. Clinical pro- 167. Jabs DA, Miller NR, Newman SA, Johnson MA, Stevens
file and long-term implications of anterior ischemic optic MB. Optic neuropathy in systemic lupus erythematosus.
neuropathy. Am J Ophthalmol. 1983;96(4):47883. Arch Ophthalmol. 1986;104(4):5648.
151. Guyer DR, Miller NR, Auer CL, Fine SL. The risk of cere- 168. Borruat FX, Herbort CP. Herpes zoster ophthalmicus.
brovascular and cardiovascular disease in patients with Anterior ischemic optic neuropathy and acyclovir. J Clin
anterior ischemic optic neuropathy. Arch Ophthalmol. Neuroophthalmol. 1992;12(1):3740.
1985;103(8):113642. 169. Hayreh SS, Servais GE, Virdi PS. Fundus lesions in malig-
152. Moro F, Doro D, Mantovani E. Anterior ischemic optic nant hypertension. V. Hypertensive optic neuropathy.
neuropathy and aging. Metab Pediatr Syst Ophthalmol. Ophthalmology. 1986;93(1):7487.
1989;12(13):4657. 170. Hayreh SS, Servais GE, Virdi PS. Fundus lesions in malig-
153. Giuffre G. Hematological risk factors for anterior ischemic nant hypertension. VI. Hypertensive choroidopathy.
optic neuropathy. Neuro-ophthalmol 1990;10:197203. Ophthalmology. 1986;93(11):1383400.
154. Leibowitz HM, Krueger DE, Maunder LR, Milton RC, 171. Kishi S, Tso MO, Hayreh SS. Fundus lesions in malignant
Kini MM, Kahn HA, et al. The Framingham Eye Study hypertension. I. A pathologic study of experimental hyper-
monograph: An ophthalmological and epidemiological tensive choroidopathy. Arch Ophthalmol. 1985;103(8):
study of cataract, glaucoma, diabetic retinopathy, macular 118997.
degeneration, and visual acuity in a general population of 172. Sossi N, Anderson DR. Blockage of axonal transport in
2631 adults, 1973-1975. Surv Ophthalmol. 1980;24(Suppl): optic nerve induced by elevation of intraocular pressure.
335610. Effect of arterial hypertension induced by angiotensin I.
155. Ramrattan RS, Wolfs RC, Jonas JB, Hofman A, de Jong Arch Ophthalmol. 1983;101(1):947.
PT. Determinants of optic disc characteristics in a general 173. Kishi S, Tso MO, Hayreh SS. Fundus lesions in malignant
population: The Rotterdam Study. Ophthalmology. hypertension. II. A pathologic study of experimental hyper-
1999;106(8):158896. tensive optic neuropathy. Arch Ophthalmol. 1985;103(8):
156. Wolfs RC, Ramrattan RS, Hofman A, de Jong PT. Cup-to- 1198206.
disc ratio: ophthalmoscopy versus automated measurement 174. Hayreh SS. Fluids in the anterior part of the optic nerve in
in a general population: The Rotterdam study. health and disease. Surv Ophthalmol. 1978;23(1):125.
Ophthalmology. 1999;106(8):1597601. 175. Ashton N. The eye in malignant hypertension. Trans Am
157. Beck RW, Servais GE, Hayreh SS. Anterior ischemic optic Acad Ophthalmol Otolaryngol. 1972;76(1):1740.
neuropathy. IX. Cup-to-disc ratio and its role in pathogen- 176. Garner A, Ashton N, Tripathi R, Kohner EM, Bulpitt CJ,
esis. Ophthalmology. 1987;94(11):15038. Dollery CT. Pathogenesis of hypertensive retinopathy. An
158. Hayreh SS. Anterior ischemic optic neuropathy. VIII. experimental study in the monkey. Br J Ophthalmol.
Clinical features and pathogenesis of post-hemorrhagic 1975;59(1):344.
amaurosis. Ophthalmology. 1987;94(11):1488502. 177. Dollery CT. Hypertensive retinopathy. In: Genest J, Kuchel
159. Anda RF, Williamson DF, Escobedo LG, Remington PL. O, Hamet P, Cantin M, editors. Hypertension: physiopa-
Smoking and the risk of peptic ulcer disease among women thology and treatment. 2nd ed. New York: McGraw Hill;
in the United States. Arch Intern Med. 1990;150(7): 1983. p. 72332.
143741. 178. Cove DH, Seddon M, Fletcher RF, Dukes DC. Blindness
160. Oyefeso AO, Odeyale MA. Smoking and type A behaviour. after treatment for malignant hypertension. Br Med J.
Scand J Psychol. 1991;32(1):7981. 1979;2(6184):2456.
161. Kato I, Nomura AM, Stemmermann GN, Chyou PH. A 179. Wetherill JH. Blindness after treatment for malignant
prospective study of gastric and duodenal ulcer and its rela- hypertension. Br Med J. 1979;2(6189):550.
tion to smoking, alcohol, and diet. Am J Epidemiol. 180. Hulse JA, Taylor DS, Dillon MJ. Blindness and paraplegia in
1992;135(5):52130. severe childhood hypertension. Lancet. 1979;2(8142):5536.
162. Lanas A, Hirschowitz BI. Influence of smoking on basal and 181. Pryor JS, Davies PD, Hamilton DV. Blindness and malig-
on vagally and maximally stimulated gastric acid and pepsin nant hypertension. Lancet. 1979;2(8146):803.
secretion. Scand J Gastroenterol. 1992;27(3):20812. 182. Taylor D, Ramsay J, Day S, Dillon M. Infarction of the
163. Forgays DG, Bonaiuto P, Wrzesniewski K, Forgays DK. optic nerve head in children with accelerated hypertension.
Personality and cigarette smoking in Italy, Poland, and the Br J Ophthalmol. 1981;65(3):15360.
United States. Int J Addict. 1993;28(5):399413. 183. Ledingham JG, Rajagopalan B. Cerebral complications in
164. Ischemic Optic Neuropathy Decompression Trial Study the treatment of accelerated hypertension. Q J Med.
Group. Characteristics of patients with nonarteritic anterior 1979;48(189):2541.
ischemic optic neuropathy eligible for the Ischemic Optic 184. Hayreh SS. Duke-Elder lecture: systemic arterial blood
Neuropathy Decompression Trial. Arch Ophthalmol. pressure and the eye. Eye. 1996;10:528.
1996;114(11):136674. 185. Strandgaard S, Olesen J, Skinhoj E, Lassen NA.
165. Jacobson DM, Vierkant RA, Belongia EA. Nonarteritic ante- Autoregulation of brain circulation in severe arterial hyper-
rior ischemic optic neuropathy: a case-control study of poten- tension. Br Med J. 1973;1(5852):50710.
tial risk factors. Arch Ophthalmol. 1997;115(11):14037. 186. Strandgaard S, Jones JV, MacKenzie ET, Harper AM.
166. Preechawat P, Bruce BB, Newman NJ, Biousse V. Anterior Upper limit of cerebral blood flow autoregulation in experi-
ischemic optic neuropathy in patients younger than mental renovascular hypertension in the baboon. Circ Res.
50years. Am J Ophthalmol. 2007;144(6):95360. 1975;37(2):1647.
References 313
187. Fitch W, MacKenzie ET, Harper AM. Effects of decreasing 205. Newman NJ. Perioperative visual loss after nonocular sur-
arterial blood pressure on cerebral blood flow in the baboon. geries. Am J Ophthalmol. 2008;145(4):60410.
Influence of the sympathetic nervous system. Circ Res. 206. Hayreh SS. Posterior ischaemic optic neuropathy: clinical
1975;37(5):5507. features, pathogenesis, and management. Eye. 2004;18(11):
188. Strandgaard S, MacKenzie ET, Jones JV, Harper AM. 1188206.
Studies on the cerebral circulation of the baboon in acutely 207. Landau K, Winterkorn JM, Mailloux LU, Vetter W,
induced hypertension. Stroke. 1976;7(3):28790. Napolitano B. 24-hour blood pressure monitoring in
189. Jones JV, Fitch W, MacKenzie ET, Strandgaard S, Harper patients with anterior ischemic optic neuropathy. Arch
AM. Lower limit of cerebral blood flow autoregulation in Ophthalmol. 1996;114(5):5705.
experimental renovascular hypertension in the baboon. 208. Hayreh SS, Zimmerman MB, Podhajsky P, Alward WL.
Circ Res. 1976;39(4):5557. Nonarteritic anterior ischemic optic neuropathy: role of
190. Edvinsson L, Owman C, Siesjo B. Physiological role of nocturnal arterial hypotension. Arch Ophthalmol. 1997;
cerebrovascular sympathetic nerves in the autoregulation 115(7):9425.
of cerebral blood flow. Brain Res. 1976;117(3):51923. 209. Hayreh SS. Retinal and optic nerve head ischemic disor-
191. Bill A, Linder J. Sympathetic control of cerebral blood flow ders and atherosclerosis: the role of serotonin. Prog Retin
in acute arterial hypertension. Acta Physiol Scand. Eye Res. 1999;18(2):191221.
1976;96(1):11421. 210. Weinstein JM, Feman SS. Ischemic optic neuropathy in
192. Boisvert DP, Jones JV, Harper AM. Cerebral blood flow migraine. Arch Ophthalmol. 1982;100(7):1097100.
autoregulation to acutely increasing blood pressure during 211. OHara M, OConnor PS. Migrainous optic neuropathy.
sympathetic stimulation. Acta Neurol Scand Suppl. 1977; J Clin Neuroophthalmol. 1984;4(2):8590.
64:467. 212. Katz B, Bamford CR. Migrainous ischemic optic neuropa-
193. MacKenzie ET, McGeorge AP, Graham DI, Fitch W, thy. Neurology. 1985;35(1):1124.
Edvinsson L, Harper AM. Breakthrough of cerebral auto- 213. Gonzalez-Martin-Moro J, Pilo-de-la-Fuente B, Moreno-
regulation and the sympathetic nervous system. Acta Martin P. Neuropatia optica isquemica anterior migranosa
Neurol Scand. 1977;64:489. [Migraineous anterior optic ischemic neuropathy]. Arch
194. Myers MA, Hamilton SR, Bogosian AJ, Smith CH, Wagner Soc Esp Oftalmol. 2009;84(9):4736.
TA. Visual loss as a complication of spine surgery. A review 214. Foroozan R. Visual dysfunction in migraine. Int Ophthalmol
of 37 cases. Spine. 1997;22(12):13259. Clin. 2009;49(3):13346.
195. Stevens WR, Glazer PA, Kelley SD, Lietman TM, Bradford 215. Kimura H, Masai H, Kashii S. Anterior ischemic optic neu-
DS. Ophthalmic complications after spinal surgery. Spine. ropathy associated with pulseless disease. Ophthalmologica.
1997;22(12):131924. 1995;209(6):3468.
196. Wilson JF, Freeman SB, Breene DP. Anterior ischemic 216. Horowitz J, Melamud A, Sela L, Hod Y, Geyer O. Internal
optic neuropathy causing blindness in the head and neck carotid artery hypoplasia presenting as anterior ischemic
surgery patient. Arch Otolaryngol Head Neck Surg. optic neuropathy. Am J Ophthalmol. 2001;131(5):6734.
1991;117(11):13046. 217. Biousse V, Schaison M, Touboul PJ, DAnglejan-Chatillon J,
197. Gill B, Heavner JE. Postoperative visual loss associated Bousser MG. Ischemic optic neuropathy associated with inter-
with spine surgery. Eur Spine J. 2006;15(4):47984. nal carotid artery dissection. Arch Neurol. 1998;55(5):7159.
198. Stambough JL, Dolan D, Werner R, Godfrey E. 218. Biousse V, Touboul PJ, DAnglejan-Chatillon J, Levy C,
Ophthalmologic complications associated with prone posi- Schaison M, Bousser MG. Ophthalmologic manifestations
tioning in spine surgery. J Am Acad Orthop Surg. of internal carotid artery dissection. Am J Ophthalmol.
2007;15(3):15665. 1998;126(4):56577.
199. Delattre O, Thoreux P, Liverneaux P, Merle H, Court C, 219. Koch S, Lorenzo D, Rabinstein AA, Lam B. Ischemic optic
Gottin M, etal. Spinal surgery and ophthalmic complica- neuropathy and carotid dissection. Neurology. 2005;64(5):
tions: a French survey with review of 17 cases. J Spinal 827.
Disord Tech. 2007;20(4):3027. 220. Baumgartner RW, Bogousslavsky J. Clinical manifesta-
200. Patil CG, Lad EM, Lad SP, Ho C, Boakye M. Visual loss tions of carotid dissection. Front Neurol Neurosci.
after spine surgery: a population-based study. Spine (Phila 2005;20:706.
1976). 2008;33(13):14916. 221. Ko SB, Lee K. Acute optic-nerve infarction in carotid dis-
201. Yu YH, Chen WJ, Chen LH, Chen WC. Ischemic orbital section. N Engl J Med. 2010;363(8):765.
compartment syndrome after posterior spinal surgery. 222. Fry CL, Carter JE, Kanter MC, Tegeler CH, Tuley MR.
Spine (Phila 1976). 2008;33(16):E569572572. Anterior ischemic optic neuropathy is not associated with
202. Cho HM, Park MS, Kim KB, Cho NC. Ischemic optic carotid artery atherosclerosis. Stroke. 1993;24(4):53942.
neuropathy after hemiarthroplasty for femoral neck frac- 223. Howard JH, Rechnitzer PA, Cunningham DA, Wong D,
ture. J Arthroplasty. 2009;24(8):1292. e1211e1294. Brown HA. Type A behavior, personality, and sympathetic
203. Kaeser PF, Borruat FX. Visual loss after orthopedic proce- response. Behav Med. 1990;16(4):14960.
dures. J Arthroplasty. 2011;26(2):338. 224. Fredrikson M, Blumenthal JA. Serum lipids, neuroendo-
204. Shen Y, Drum M, Roth S. The prevalence of perioperative crine and cardiovascular responses to stress in healthy Type
visual loss in the United States: a 10-year study from 1996 A men. Biol Psychol. 1992;34(1):4558.
to 2005 of spinal, orthopedic, cardiac, and general surgery. 225. Mojon DS, Hedges 3rd TR, Ehrenberg B, Karam EZ,
Anesth Analg. 2009;109(5):153445. Goldblum D, Abou-Chebl A, et al. Association between
sleep apnea syndrome and nonarteritic anterior ischemic 242. Tnz MS, Rigamonti V, Iliev ME. Simultaneous, bilateral
optic neuropathy. Arch Ophthalmol. 2002;120(5):6015. anterior ischemic optic neuropathy (AION) in polycythemia
226. Behbehani R, Mathews MK, Sergott RC, Savino PJ. vera: a case report. Klin Monatsbl Augenheilkd. 2008;
Nonarteritic anterior ischemic optic neuropathy in patients 225(5):5046.
with sleep apnea while being treated with continuous 243. Mears KA, Van Stavern GP. Bilateral simultaneous anterior
positive airway pressure. Am J Ophthalmol. 2005;139(3): ischaemic optic neuropathy associated with Sticky Platelet
51821. Syndrome. Br J Ophthalmol. 2009;93(7):8856, 913.
227. Palombi K, Renard E, Levy P, Chiquet C, Deschaux C, 244. Glueck CJ, Wang P. Ocular vascular thrombotic events: a
Romanet JP, et al. Non-arteritic anterior ischaemic optic diagnostic window to familial thrombophilia (compound
neuropathy is nearly systematically associated with obstruc- factor V Leiden and prothrombin gene heterozygosity)
tive sleep apnoea. Br J Ophthalmol. 2006;90(7):87982. and thrombosis. Clin Appl Thromb Hemost. 2009;15(1):
228. Li J, McGwin Jr G, Vaphiades MS, Owsley C. Non-arteritic 128.
anterior ischaemic optic neuropathy and presumed sleep 245. Kuhli-Hattenbach C, Scharrer I, Luchtenberg M, Hattenbach
apnea syndrome screened by the Sleep Apnea scale of the LO. Selective thrombophilia screening of patients with
Sleep Disorders Questionnaire (SA-SDQ). Br J Ophthalmol. nonarteritic anterior ischemic optic neuropathy. Graefes
2007;91(11):15247. Arch Clin Exp Ophthalmol. 2009;247(4):48590.
229. McNab AA. The eye and sleep apnea. Sleep Med Rev. 246. Giambene B, Sodi A, Sofi F, Marcucci R, Fedi S, Abbate R,
2007;11(4):26976. etal. Evaluation of traditional and emerging cardiovascular
230. Waller EA, Bendel RE, Kaplan J. Sleep disorders and the risk factors in patients with non-arteritic anterior ischemic
eye. Mayo Clin Proc. 2008;83(11):125161. optic neuropathy: a case-control study. Graefes Arch Clin
231. Nieto Enriquez J, Torres Blanch J, Badal Lafulla J. Exp Ophthalmol. 2009;247(5):6937.
Manifestaciones oculares del sindrome de apnea del sueno 247. Salomon O, Dardik R, Steinberg DM, Kurtz S, Rosenberg
[Ocular disease in sleep apnea syndrome]. Med Clin (Barc). N, Moisseiev J, et al. The role of angiotensin converting
2009;133(15):5948. enzyme and angiotensin II type 1 receptor gene polymor-
232. Slavin ML, Barondes MJ. Ischemic optic neuropathy in phisms in patients with nonarteritic anterior ischemic optic
sickle cell disease. Am J Ophthalmol. 1988;105(2):2123. neuropathy. Ophthalmology. 2000;107(9):171720.
233. Acheson JF, Sanders MD. Coagulation abnormalities in 248. Biousse V, Kerrison JB, Newman NJ. Is non-arteritic ante-
ischaemic optic neuropathy. Eye. 1994;8(1):8992. rior ischaemic optic neuropathy related to homocysteine?
234. Pianka P, Almog Y, Man O, Goldstein M, Sela BA, Br J Ophthalmol. 2000;84(5):555.
Loewenstein A. Hyperhomocystinemia in patients with 249. Heit JA. Thrombophilia: common questions on laboratory
nonarteritic anterior ischemic optic neuropathy, central assessment and management. Hematology Am Soc
retinal artery occlusion, and central retinal vein occlusion. Hematol Educ Program. 2007;2007:12735.
Ophthalmology. 2000;107(8):158892. 250. Bull GM. Meteorological correlates with myocardial and
235. Killer HE, Huber A, Portman C, Forrer A, Flammer J. cerebral infarction and respiratory disease. Br J Prev Soc
Bilateral non-arteritic anterior ischemic optic neuropathy Med. 1973;27(2):10813.
in a patient with autoimmune thrombocytopenia. Eur J 251. Bull GM, Morton J. Environment, temperature and death
Ophthalmol. 2000;10(2):1802. rates. Age Ageing. 1978;7(4):21024.
236. Srinivasan S, Fern A, Watson WH, McColl MD. Reversal 252. Clark CV. Seasonal variation in incidence of brachial and
of nonarteritic anterior ischemic optic neuropathy associ- femoral emboli. Br Med J (Clin Res Ed). 1983;287(6399):
ated with coexisting primary antiphospholipid syndrome 1109.
and Factor V Leiden mutation. Am J Ophthalmol. 253. Rogot E, Padgett SJ. Associations of coronary and stroke
2001;131(5):6713. mortality with temperature and snowfall in selected areas
237. Glueck CJ, Wang P, Bell H, Rangaraj V, Goldenberg N. of the United States, 1962-1966. Am J Epidemiol.
Nonarteritic anterior ischemic optic neuropathy: associa- 1976;103(6):56575.
tions with homozygosity for the C677T methylenetetrahy- 254. Fersini C, Manfredini R, Manfredini F, Balboni G, Fersini
drofolate reductase mutation. J Lab Clin Med. G. Chronobiologic aspects of recurrent transient ischemic
2004;143(3):18492. attack. Prog Clin Biol Res. 1987;227B:16771.
238. Nagy V, Facsko A, Takacs L, Balazs E, Berta A, Balogh I, 255. Bull GM, Brozovic M, Chakrabarti R, Meade TW, Morton
etal. Activated protein C resistance in anterior ischaemic optic J, North WR, etal. Relationship of air temperature to vari-
neuropathy. Acta Ophthalmol Scand. 2004;82(2):1403. ous chemical, haematological, and haemostatic variables. J
239. Chuah J, Meaney T. Anterior ischaemic optic neuropathy Clin Pathol. 1979;32(1):1620.
secondary to Henoch-Schonlein Purpura. Eye (Lond). 256. Keatinge WR, Coleshaw SR, Cotter F, Mattock M, Murphy
2005;19(9):1028. M, Chelliah R. Increases in platelet and red cell counts,
240. Stanger O, Weger M, Obeid R, Temmel W, Meinitzer A, blood viscosity, and arterial pressure during mild surface
Steinbrugger I, etal. Impairment of homocysteine metabo- cooling: factors in mortality from coronary and cerebral
lism in patients with retinal vascular occlusion and non- thrombosis in winter. Br Med J (Clin Res Ed).
arteritic ischemic optic neuropathy. Clin Chem Lab Med. 1984;289(6456):14058.
2005;43(10):10205. 257. Blajchman MA, Senyi AF, Hirsh J, Genton E, George JN.
241. Randhawa S, Van Stavern GP. Sticky platelet syndrome and Hemostatic function, survival, and membrane glycoprotein
anterior ischaemic optic neuropathy. Clin Experiment changes in young versus old rabbit platelets. J Clin Invest.
Ophthalmol. 2007;35(8):77981. 1981;68(5):128994.
References 315
258. Thompson CB, Jakubowski JA, Quinn PG, Deykin D, 277. Hayreh SS. Optic disc edema in raised intracranial pres-
Valeri CR. Platelet size as a determinant of platelet func- sure. V. Pathogenesis. Arch Ophthalmol. 1977;95(9):
tion. J Lab Clin Med. 1983;101(2):20513. 155365.
259. Bouhoutsos J, Morris T, Chavatzas D, Martin P. The influ- 278. Hayreh SS. Anterior ischemic optic neuropathy. V. Optic
ence of haemoglobin and platelet levels on the results of disc edema an early sign. Arch Ophthalmol. 1981;99(6):
arterial surgery. Br J Surg. 1974;61(12):9846. 103040.
260. Turitto VT, Baumgartner HR. Platelet interaction with sub- 279. Hayreh SS, Zimmerman MB. Optic disc edema in non-
endothelium in a perfusion system: physical role of red arteritic anterior ischemic optic neuropathy. Graefes Arch
blood cells. Microvasc Res. 1975;9(3):33544. Clin Exp Ophthalmol. 2007;245(8):110721.
261. Aarts PA, Bolhuis PA, Sakariassen KS, Heethaar RM, Sixma 280. Tomlinson A, Phillips CI. Ratio of optic cup to optic disc.
JJ. Red blood cell size is important for adherence of blood In relation to axial length of eyeball and refraction. Br J
platelets to artery subendothelium. Blood. 1983;62(1):2147. Ophthalmol. 1969;53(11):7658.
262. Stout RW, Crawford V. Seasonal variations in fibrinogen 281. Katz B, Spencer WH. Hyperopia as a risk factor for nonar-
concentrations among elderly people. Lancet. 1991; teritic anterior ischemic optic neuropathy. Am J Ophthalmol.
338(8758):913. 1993;116(6):7548.
263. Buchanan RR, Riglar AG. The titre of anti-centromere anti- 282. Philips B, Dralands L, Missotten L. Non-arteritic anterior
bodies: its relationship to Raynauds phenomenon and vas- ischemic optic neuropathy and refraction. Bull Soc Belge
cular occlusion. Br J Rheumatol. 1989;28(3):2216. Ophtalmol. 1995;259:1837.
264. Reilly C, Nicolau GY, Lakatua DJ, Bogdan C, Sackett- 283. Gittinger Jr JW, Lessell S, Bondar RL. Ischemic optic neu-
Lundeen L, Petrescu E, etal. Circannual rhythms of labora- ropathy associated with optic disc drusen. J Clin
tory measurements in serum of elderly subjects. Prog Clin Neuroophthalmol. 1984;4(2):7984.
Biol Res. 1987;227B:5172. 284. Sarkies NJ, Sanders MD. Optic disc drusen and episodic
265. Exton-Smith AN. Disturbances of autonomic regulation. visual loss. Br J Ophthalmol. 1987;71(7):5379.
In: Isaacs B, editor. Recent advances in geriatric medicine. 285. Michaelson C, Behrens M, Odel J. Bilateral anterior ischae-
Edinburgh: Churchill Livingstone; 1978. p. 85100. mic optic neuropathy associated with optic disc drusen and
266. Brennan PJ, Greenberg G, Miall WE, Thompson SG. systemic hypotension. Br J Ophthalmol. 1989;73(9):
Seasonal variation in arterial blood pressure. Br Med J 7624.
(Clin Res Ed). 1982;285(6346):91923. 286. Moody TA, Irvine AR, Cahn PH, Susac JO, Horton JC.
267. Hayreh SS. Pathogenesis of cupping of the optic disc. Br J Sudden visual field constriction associated with optic disc
Ophthalmol. 1974;58(10):86376. drusen. J Clin Neuroophthalmol. 1993;13(1):813.
268. Hoyt WF. Rocky Mountain Neuro-ophthalmology Society 287. Newman WD, Dorrell ED. Anterior ischemic optic neu-
Meeting 1982. ropathy associated with disc drusen. J Neuroophthalmol.
269. Beck RW, Savino PJ, Repka MX, Schatz NJ, Sergott RC. 1996;16(1):78.
Optic disc structure in anterior ischemic optic neuropathy. 288. Liew SC, Mitchell P. Anterior ischaemic optic neuropathy
Ophthalmology. 1984;91(11):13347. in a patient with optic disc drusen. Aust N Z J Ophthalmol.
270. Feit RH, Tomsak RL, Ellenberger Jr C. Structural factors in 1999;27(2):15760.
the pathogenesis of ischemic optic neuropathy. Am J 289. Cousin P, Fourmaux E, Renaud-Rougier MB, Mercie M,
Ophthalmol. 1984;98(1):1058. Pincemin D, Le Rebeller MJ. Neuropathie optique
271. Doro S, Lessell S. Cup-disc ratio and ischemic optic neu- ischmique antrieure aigu bilaterale compliquant des
ropathy. Arch Ophthalmol. 1985;103(8):11434. druses de la papille. A propos dun cas [Bilateral anterior
272. Burde RM. Optic disk risk factors for nonarteritic anterior acute ischemic optic neuropathy complicating optic nerve
ischemic optic neuropathy. Am J Ophthalmol. 1993;116(6): head drusen. Apropos of a case]. J Fr Ophtalmol.
75964. 1999;22(1):7983.
273. Wang MY, Sadun F, Levin LB, LaBree L, Feldon SE. 290. Kamath GG, Prasad S, Phillips RP. Bilateral anterior
Occurrence of familial nonarteritic anterior ischemic optic ischaemic optic neuropathy due to optic disc drusen. Eur J
neuropathy in a case series. J Neuroophthalmol. 1999; Ophthalmol. 2000;10(4):3413.
19(2):1447. 291. Purvin V, King R, Kawasaki A, Yee R. Anterior ischemic
274. Pomeranz HD, Smith KH, Hart Jr WM, Egan RA. optic neuropathy in eyes with optic disc drusen. Arch
Sildenafil-associated nonarteritic anterior ischemic optic Ophthalmol. 2004;122(1):4853.
neuropathy. Ophthalmology. 2002;109(3):5847. 292. Lam BL, Morais Jr CG, Pasol J. Drusen of the optic disc.
275. Leibowitz HM, Krueger DE, Maunder LR, Milton RC, Curr Neurol Neurosci Rep. 2008;8(5):4048.
Kini MM, Kahn HA, et al. The Framingham Eye Study 293. Lam BL, Jabaly-Habib H, Al-Sheikh N, Pezda M, Guirgis
monograph: An ophthalmological and epidemiological MF, Feuer WJ, etal. Risk of non-arteritic anterior ischae-
study of cataract, glaucoma, diabetic retinopathy, macular mic optic neuropathy (NAION) after cataract extraction in
degeneration, and visual acuity in a general population of the fellow eye of patients with prior unilateral NAION. Br
2631 adults, 1973-1975. VIII. Visual Acuity. Surv J Ophthalmol. 2007;91(5):5857.
Ophthalmol. 1980;24(Suppl):4729. 294. Carroll FD. Optic nerve complications of cataract extrac-
276. Duke-Elder S, Cook C. System of ophthalmology. In: tion. Trans Am Acad Ophthalmol Otolaryngol. 1973;
Duke-Elder S, editor. Normal and abnormal development 77(5):OP623629629.
Part 1: embryology, vol. III. St. Louis: Mosby; 1963. p. 295. Reese AB, Carroll FD. Optic neuritis following cataract
10910. extraction. Am J Ophthalmol. 1958;45(5):65962.
296. Hayreh SS, Podhajsky P, Zimmerman MB. Beta-blocker 302. Hayreh MS, Hayreh SS, Baumbach GL, Cancilla P, Martin-
eyedrops and nocturnal arterial hypotension. Am J Amat G, Tephly TR, et al. Methyl alcohol poisoning III.
Ophthalmol. 1999;128(3):3019. Ocular toxicity. Arch Ophthalmol. 1977;95(10):18518.
297. Hayreh SS. Optic disc edema in raised intracranial pres- 303. Hayreh SS, Jonas JB. Posterior vitreous detachment: clini-
sure. VI. Associated visual disturbances and their patho- cal correlations. Ophthalmologica. 2004;218(5):33343.
genesis. Arch Ophthalmol. 1977;95(9):156679. 304. The Ischemic Optic Neuropathy Decompression Trial
298. Tso MO, Hayreh SS. Optic disc edema in raised intracranial Research Group. Optic nerve decompression surgery for
pressure. III. A pathologic study of experimental papille- nonarteritic anterior ischemic optic neuropathy (NAION)
dema. Arch Ophthalmol. 1977;95(8):144857. is not effective and may be harmful. JAMA. 1995;273(8):
299. Tso MOM, Hayreh SS. Optic disc edema in raised 62532.
intracranial pressure. IV. Axoplasmic transport in exp 305. Hayreh SS, Zimmerman MB. Nonarteritic anterior isch-
erimental papilledema. Arch Ophthalmol. 1977;95(8): emic optic neuropathy: natural history of visual outcome.
145862. Ophthalmology. 2008;115(2):298305.
300. Fink BR, Byers MR, Middaugh ME. Dynamics of colchi- 306. Hayreh SS, Zimmerman B. Management of giant cell
cine effects on rapid axonal transport and axonal morphol- arteritis: our 27-year clinical study; New light on old con-
ogy. Brain Res. 1973;56:299311. troversies. Ophthalmologica. 2003;217(4):23959.
301. Hayreh SS. Arterial blood supply of the eye. In: Bernstein 307. Hayreh SS, Zimmerman B. Visual deterioration in giant
EF, editor. Amaurosis fugax. New York: Springer; 1988. cell arteritis patients while on high doses of corticosteroid
Chapter 1, p. 123. therapy. Ophthalmology. 2003;110(6):120415.
Pathogenesis of Some Controversial
Non-arteritic Anterior Ischemic Optic 15
Neuropathy Clinical Entities
There are some cases of non-arteritic anterior ischemic pathogenesis of incipient NA-AION, one has to
optic neuropathy (NA-AION) whose very existence, consider the following two basic facts:
nature and pathogeneses have been controversial and
1. Axoplasmic flow itself is not concerned with trans-
require detailed discussion.
mission of a visual impulse [7] the two are inde-
pendent processes. This is very well demonstrated
by optic disc edema in raised intracranial pressure,
Incipient NA-AION which, though due to axoplasmic flow stasis [4], is
not accompanied by any detectable axonal dysfunc-
tion per se. There are other similar examples of
A disease usually evolves from a latent phase to a
optic disc edema due to axoplasmic flow stasis
symptomatic phase. Incipient NA-AION is an example
without any visual dysfunction.
of that. Patients with NA-AION typically present with
2. While considering the effect of ischemia on neural
a history of sudden visual loss, optic disc-related visual
tissue, it is vital to appreciate that, contrary to the
field defects and optic disc edema in the involved eye.
general impression, ischemic damage does not
In 1981, I [1] reported that symptomless optic disc
follow an all or none law (discussed elsewhere
edema precedes the visual loss and may be the earliest
[1] see Chap 14). Available evidence indicates
sign of AION (NA-AION). There is a widespread
that subclinical ischemia only produces axoplasmic
belief that unless and until there is sudden visual loss,
flow stasis without interfering with transmission of
in addition to optic disc edema, it cannot be NA-AION.
visual impulses and without producing infarction
This has resulted in: (a) missing the diagnosis of
this is the case in incipient NA-AION. At the other
NA-AION when a patient presented initially with only
end of the scale is severe ischemia, which can result
asymptomatic optic disc edema, and (b) extensive and
in infarction of the optic nerve head, as is seen in
expensive neurologic and other investigations. For
arteritic AION, with complete, permanent loss of
example, initial asymptomatic optic disc edema in dia-
vision. Therefore, ischemia can cause optic nerve
betics has been designated diabetic papillopathy or
head ischemic damage of varying severity. In con-
diabetic papillitis. Similarly, initially asymptomatic
sidering the natural history of NA-AION, one must
optic disc edema in patients who take amiodarone has
not equate acute ischemia with infarction; acute
been considered amiodarone optic neuropathy.
ischemia of the neural tissue does not always pro-
We [2] recently presented a series of 60 eyes with
duce infarction and complete loss of function. There
this clinical entity, calling it, more accurately, incipi-
is now ample evidence in the literature that many
ent NA-AION, and discussed its pathogenesis based
such patients may recover a variable amount of
on the findings in those cases. It is well-established
function after an acute ischemic episode of the ner-
now that axoplasmic flow stasis is the common path-
vous system, as well as of the optic nerve [811].
way for development of optic disc edema in a variety
of conditions [35]. Ischemia of the axons is known to The factors that initiate the subclinical ischemia of the
produce axoplasmic flow stasis [3, 6], as has been optic nerve head in the first place are still not fully
demonstrated in AION [6]. To understand the understood. As discussed in Chap. 14, available

318 15 Pathogenesis of Some Controversial Non-arteritic Anterior Ischemic Optic Neuropathy Clinical Entities
evidence indicates that NA-AION is a multifactorial capillaries below the critical level is the last straw in
disease [12, 13] with several risk factors playing roles the scenario. Other factors such as defective autoregu-
and contributing to deranged optic nerve head blood lation of blood flow in the optic nerve head, vascular
flow. Those factors that influence the optic nerve head changes secondary to arterial hypertension, diabetes
blood flow are discussed at length in Chap. 5. mellitus and atherosclerosis, may also contribute to the
From the available evidence, incipient NA-AION entire process of development of incipient NA-AION,
most likely represents the earliest clinical stage in the as well as its progression to classical NA-AION.
evolution of classical NA-AION. It may progress to It seems that all NA-AION patients start with
the classical NA-AION with visual loss. There are asymptomatic optic disc edema. They are normally
several reports in the literature of the progression of seen only when they progress to classical NA-AION
incipient NA-AION to classical NA-AION [1, 2, 14 with visual loss, unless they happen by chance to be
17]. In my initial study [1], I reported progression in 2 examined during the asymptomatic phase and the
of 4 patients with incipient NA-AION after 7 and incipient AION is discovered incidentally. The patho-
12weeks respectively. I was able to find in the litera- genetic mechanism outlined above applies to classical
ture reports of seven similar cases [1417], who pro- NA-AION as well [23, 24].
gressed after 3days (one case) [16], 2weeks (one case)
[17], 46 weeks (4 cases) [14], and at a later date
(one case) [15]. Since then, in our study [2] of 54
patients (60 eyes) with incipient NA-AION, we found NA-AION Following the Use of
that 25% of the incipient progressed to classical Phosphodiesterase Type 5 (PDE5)
NA-AION (after a median time of 5.8weeks). Inhibitors
The following mechanism seems to be responsible
for the progression of incipient NA-AION to classical Since 1998, phosphodiesterase type 5 (PDE5) inhibi-
NA-AION. Almost all the eyes with incipient NA-AION tors have been used extensively for the treatment of
have no cup on the optic disc. This means that these erectile dysfunction. These agents include sildenafil
eyes have a very restricted space in the optic nerve (Viagra, Pfizer), vardenafil (Levitra, Bayer AG), and
head, surrounded by a tight Bruchs membrane and tadalafil (Cialis, Lilly-ICOS LLC). Development of
small scleral canal, resulting in crowding of the nerve NA-AION has been reported following the use of these
fibers. For swollen axons to expand in such a restricted drugs [25, 26] mostly Viagra. This has attracted a
space, they have to displace or squeeze surrounding tis- good deal of interest and has become a controversial
sues, and the only tissues they can compress are the topic which requires detailed discussion about its
capillaries lying in between them, because of the very pathogenesis. (It is discussed at length elsewhere
low pressure in the latter. Thus, axonal swelling from [2527]). Following is a brief summary, based on: (a)
subclinical ischemia produces secondary vascular the blood supply of the optic nerve head: the factors
changes in the optic nerve head [2, 5, 18]. That can set that influence the optic nerve head (ONH) blood flow
up a vicious cycle: axonal swellingcompression of and its measurement (see Chaps. 5 and 6), (b) the
capillaries lying in between themreduced blood pathogenesis of NA-AION (see Chap. 14), and (c) the
flowmore ischemia of the axonsincrease in systemic and ocular vascular effects of PDE5 inhibitors.
axoplasmic flow stasismore swelling of the
axonsfinally resulting in marked ischemia and visual
loss. In addition to that, once the capillaries in the optic
nerve head are compressed, that increases the vascular The Systemic and Ocular Vascular
resistance in them and consequently they require a Effects of PDE5 Inhibitors
higher level of blood pressure to maintain an adequate
blood flow. If the blood pressure falls below critical lev- Vascular endothelium plays an active role in vasomo-
els, as in nocturnal arterial hypotension [1922] or for tor function and regulation of blood flow by release of
any other reason, incipient NA-AION may immediately a variety of endothelial derived vasoactive agents [28]
convert to classical NA-AION. Available evidence sug- (see Chap. 5). One of those agents is nitric oxide (NO),
gests that a fall of blood pressure in the optic nerve head which in turn induces vasodilation by increasing
NA-AION Following the Use of Phosphodiesterase Type 5 (PDE5) Inhibitors 319
production of the second messenger cyclic guanosine development of NA-AION in the light of all the basic
monophosphate (cGMP). PDE5 enzyme is found scientific facts about the pathogenesis of NA-AION
within the vascular smooth muscle cells in the walls of and others factors discussed above and in Chap. 14.
systemic arteries and veins. The activity of PDE5 regu- From that information, it emerges the following fac-
lates levels of vascular smooth muscle cGMP. Viagra, tors play important roles.
a highly selective antagonist of PDE5, therefore,
enhances the effect of NO [29]. 1. A critical review of all the reported cases shows that
It is well-known that vasoconstriction of the termi- there is usually a close temporal relationship between
nal arterioles results in arterial hypertension by increas- the ingestion of these drugs and onset of NA-AION.
ing the peripheral vascular resistance. Conversely, There are also reports of repeated challenging with
dilation of the terminal arterioles decreases the periph- the drugs reproducing the same results [35, 39]. For
eral vascular resistance, resulting in lowered blood example, Morgan etal. [40] described a patient who
pressure, as is evident from the various drugs currently suffered a transient ischemic attack followed by a
used to treat arterial hypertension. PDE5 antagonists stroke in the same distribution 6 days later, each
enhance the effect of NO, which is a potent vasodilator event being associated with Viagra use. In another
and results in dilation of terminal arterioles and conse- study, Humphrey perimetry after ingestion of
quently reduced vascular resistance and fall of blood 200mg of Viagra in a healthy young woman showed
pressure. It is well-established that Viagra is associ- bilateral superior and inferonasal visual field depres-
ated with a fall of blood pressure [3032]. sion [41]. Sivaswamy and Vanstavern [42] reported
Individuals using Viagra may also be taking arterial development of NA-AION in a child who was given
hypotensive drugs for arterial hypertension or other car- Viagra to treat pulmonary hypertension.
diovascular disorders. The combination of antihyper- 2. Most reported patients are middle-aged or elderly
tensive therapy with PDE5 inhibitors has been shown to men. Arterial hypertension, diabetes mellitus,
cause more fall of blood pressure than these agents hyperlipidemia and other systemic cardiovascular
individually [31, 33]. Development of orthostatic risk factors are common in this group. As dis-
hypotension has been reported when PDE5 inhibitors cussed in Chap. 14, those factors predispose them
are used in conjunction with an alpha-blocker [33]. to NA-AION. Patients with arterial hypertension
Since NA-AION is an ischemic disorder of the and other cardiovascular disorders are routinely
ONH, our primary interest here is the evaluation of prescribed beta-blockers, ACE-inhibitors, calcium
blood flow and its autoregulation in the ONH. channel blockers or other drugs with an arterial
Unfortunately, we do not as yet have any reliable hypotensive effect. Patients with benign prostatic
method to obtain that information [34]. For example, hypertrophy too are often advised to take alpha
Pomeranz and Bhavsar [35] stated that Grunwald etal. adrenoreceptor blocking agents at bedtime; these
[36] found no significant change in the ONH blood drugs can aggravate the arterial hypotensive effect
flow with Viagra compared with placebo, implying of PDE5 inhibitors and may also produce noctur-
that Viagra does not influence the ONH blood flow. nal arterial hypotension [22, 43, 44]. Erectile dys-
But they [36] used a Laser Doppler flowmeter to mea- function is a side-effect of many arterial
sure it, and the study by Petrig etal. [37] has clearly hypotensive drugs.
shown that the Laser Doppler flowmeter does not mea- 3. Our [22, 43, 44] 24 h ambulatory blood pressure
sure the ONH blood flow reliably. Therefore, results monitoring study showed that when patients took
based on this method are not reliable [38]. an arterial hypotensive medication in the evening
or at bedtime, that aggravated the normal physio-
logical fall of blood pressure during sleep, result-
ing in marked nocturnal arterial hypotension
Role of PDE5 Inhibitors in the (Fig. 15.1). Since Viagra and other erectile dys-
Development of NA-AION function drugs are most commonly taken in the
evening or before going to bed, their arterial
One has to evaluate the role of Viagra and other PDE5 hypotensive effect is likely to add to nocturnal arte-
inhibitors (used for erectile dysfunction) in the rial hypotension.
a Verapomil for migraine

160 mg morning and bedtime and 80 mg noon
280
260
240
220
200
180
160
140
120
100
80
60
40
20
10 12 14 16 18 20 22 0 2 4 6 8 10
Time (hour)
b 160 mg morning and 80 mg noon
280
260
240
220
200
180
160
140
120
100
80
60
40
20
10 12 14 16 18 20 22 0 2 4 6 8 10
Time (hour)
Fig. 15.1 Ambulatory blood pressure monitoring records when she was taking Verapamil 160mg morning and bedtime
(based on individual readings) over a 24h period, starting from and 80mg at noon, there was a marked degree of nocturnal arte-
about 10 A.M., in a 63year-old woman who developed normal rial hypotension (b) This record shows a marked improvement
tension glaucoma while on Verapamil hydrochloride for in nocturnal arterial hypotension on stopping the bedtime dose
migraine (Reproduced from Hayreh [43]) (a) This record shows
4. In the vast majority of cases, nocturnal arterial with sildenafil was approximately 1.6-fold higher
hypotension is the precipitating risk factor for than normal, lasting for about 4h, and with tadalafil
development of NA-AION in individuals with pre- 1.6- to 1.9-fold higher than normal for at least
existing predisposing risk factors [21]. 48 h. The aqueous humor protein content was
5. A recent experimental study showed that ingestion approximately 39% higher with 100mg sildenafil.
of sildenafil as well as of tadalafil increased This transient rise in IOP, along with arterial
intraocular pressure (IOP) [45]. The raised IOP hypotension, may contribute in development of
NA-AION Following the Use of Phosphodiesterase Type 5 (PDE5) Inhibitors 321
NA-AION by further decreasing the blood flow in 1. The most common argument is that the incidence
the optic nerve head, since blood flow in the optic of development of NA-AION among millions of
nerve head depends upon the perfusion pressure person using erectile dysfunction drugs is so small
(perfusion pressure=mean blood pressure minus that it is hard to imagine that there is any cause-
the IOP). and-effect relationship between the two. This
6. Like the vast majority of NA-AION patients, most requires an explanation. NA-AION is an ischemic
patients with NA-AION following ingestion of disorder of the ONH. To understand the mechanism
Viagra discovered visual loss upon awakening in of development of ONH ischemia requires an in-
the morning [21]. depth understanding of the ONH circulation, which
7. Phillips et al. [46] showed sympathetic activation is extremely complex, and is influenced by a large
with Viagra and an associated increase in plasma number of factors, both systemic and local (see
norepinephrine levels after Viagra administration; Chap. 14). Erectile dysfunction is a side-effect of
they concluded that sympathetic activation by many arterial hypotensive drugs. ONH circulatory
Viagra may have implications for understanding impairment is produced by these factors interacting
cardiovascular events associated with Viagra use. in different combinations, and to different extents
Norepinephrine is a potent vasoconstrictor. For in different ONHs. Hence, ONH vascular insuffi-
example, in malignant arterial hypertension, ONH ciency resulting in NA-AION is a highly complex
ischemia and NA-AION are the results of diffusion process [28], and NA-AION is multifactorial in
of vasoconstrictor agents from the peripapillary origin [12, 28]. The quest for a single villain that
choroid into the ONH [47]. That may be yet another causes NA-AION is a mistaken and misleading
contributory factor in development of NA-AION endeavor, which has resulted in controversy and
following Viagra. confusion. To determine whether there is a cause-
8. Pomeranz etal. [48] concluded that: A small cup- and-effect relationship between a drug and devel-
to-disc ratio may be a risk factor for development of opment of NA-AION, the first important
NAION in association with the use of sildenafil. consideration is whether that particular drug influ-
There is a prevalent misconception that an absent or ences the factors that influence the ONH circula-
a small optic disc cup is the primary risk factor for tion. Several studies have demonstrated a fall of
development of NA-AION (see Chap. 14). Studies blood pressure with Viagra (see above). In this con-
by Beck etal. [23] and us [24] investigated this sub- nection, the findings of my study of 24h ambula-
ject in a large cohort of patients and showed that an tory BP monitoring in more than 700 individuals
absent or small cup is simply a secondary contrib- provide important evidence; that the development
uting factor, once the process of NA-AION has of a certain amount of nocturnal arterial hypoten-
started, and not a primary factor. sion is a universal, normal phenomenon yet not
everyone develops NA-AION, although nocturnal
When all the above evidence is put together, it suggests arterial hypotension is the precipitating risk factor
that Viagra and other PDE5 inhibitors can result in in development of NA-AION in at least 73% of
development of NA-AION in persons who already patients [21]. Thus, it is a matter of: (a) what pre-
have predisposing risk factors for NA-AION. As dis- disposing risk factors an individual has for develop-
cussed previously [25], all this suggests a cause-and- ment of NA-AION (see Chap. 14) and (b) how
effect relationship between the PDE5 inhibitor drugs much fall of blood pressure develops after taking
and development of NA-AION, despite the recent con- PDE5 inhibitors; these factors in concert determine
clusion by Harris etal. [49] that A causal relationship whether a person will develop NA-AION or not.
between PD5 inhibitors and NAION has not been Another probable reason for the low reported inci-
established. (Dr. Harris is a consultant to Pfizer Inc., dence of NA-AION in persons using PDE5 inhibi-
manufacturer of Viagra). tors is that patients do not always tell their
Skeptics about the cause-and-effect relationship ophthalmologists that they use these drugs, because
between the use of Viagra and development of of embarrassment and/or because they do not see
NA-AION have put forward several arguments in sup- any connection between the use of erectile dysfunc-
port of their views, including the following: tion drugs and their eye problems.
2. The other argument put forward has been that some I. A patient may be having marked nocturnal
patients had been taking Viagra, without visual loss, arterial hypotension, even not taking any blood
for a variable length of time before they finally pressure lowering medication, while the daytime
developed NA-AION, which may suggest that blood pressure is absolutely ideal (Fig.15.2).
Viagra was not the cause. All the available evidence II. It is always assumed that atherosclerosis is a dis-
indicates that persons who develop NA-AION have ease of the elderly, but atherosclerotic plaques in
systemic and local risk factors which make them the coronary arteries were seen among healthy
susceptible to develop NA-AION. This is evident young (average 22years of age) American sol-
from the fact that there is: diers killed in the Korean War [57].
i. A fairly high incidence of development of III. Blood flow autoregulation in the optic nerve
NA-AION in the fellow eye [5052]. head plays an important role in its blood flow,
ii. Some patients develop recurrence of NA-AION so that derangement of it plays an important
in the same eye, but not all [53]. Our study [53] role in various types of circulatory disorders of
showed that the only significant risk factor for the ONH; but unfortunately, we have no means
recurrence of NA-AION in the same eye is noc- of evaluating that so far.
turnal arterial hypotension. IV. I [56] recently reported an interesting case,
iii. There is a high incidence of development of who by all conventional standards was quite
NA-AION following cataract extraction in the fit, healthy and asymptomatic, and whose rou-
fellow eye [54]. tine physical examination and electrocardio-
gram revealed no abnormality; however, a
In all these conditions, there is always a time lag between
Holter monitor for a totally unrelated reason
the involvement of the two eyes or the subsequent epi-
showed periods of complete cardiac asystole,
sode, a lag which varies from a few days to several years.
with flat electrocardiogram, lasting as long as
If what the skeptics imply were correct, then all patients
10s every few minutes during sleep (see Fig.6
developing NA-AION must develop it simultaneously in
in Chap. 14). During sleep that man could have
both eyes, which is not the case at all. This indicates that
developed stroke, NA-AION and even could
a critical combination of systemic and local risk factors
have died from cardiac arrest. He required
is required in each case before NA-AION will develop in
urgent pacemaker implantation.
an eye. One of those factors is the extent of fall of blood
pressure required to produce the critical amount of Thus, the axiom should be that if one does not find
hypoperfusion in a susceptible ONH which will result in any evident risk factor that does not necessarily mean
the development of NA-AION. That may depend upon a that person actually has no risk factors [56].
whole host of local ONH and systemic risk factors (see In conclusion, all the available evidence suggests a
Chap. 5) [28, 55]. As discussed in Chap. 6, we currently cause-and-effect relationship between the ingestion of
lack the technology to evaluate fully the local risk factors erectile dysfunction drugs and development of
in the ONH and its blood flow in humans [34]. NA-AION. To reduce the incidence of development of
NA-AION following the use of these drugs, patients
3. It has also been argued that NA-AION appears to with cardiovascular risk factors, diabetes mellitus,
occur in some patients using erectile dysfunction sleep apnea, using arterial hypotensive drugs, and/or a
drugs who have no evident systemic or vascular pre- history of previous NA-AION need to be counseled
disposing risk factors. Among over 1,000 patients against their use, to prevent visual loss.
with spontaneously-occurring NA-AION, whom I
have seen over the years, there were many who
seemed perfectly fit and healthy by the usual stan-
NA-AION in Persons Using
dards and on routine physical evaluation. Our inabil-
ity to find a risk factor may not mean that it does not AMIODARONE
exist; the fallacy of this argument is discussed recently
elsewhere [56] (see Chap. 14). This is primarily The prevalent view currently is that patients taking ami-
because of serious limitations in our ability to detect odarone (anti-arrhythmic drug) develop amiodarone
all the systemic and local risk factors. For example: induced optic neuropathy. This was first reported
NA-AION in Persons Using AMIODARONE 323
Ambulatory blood pressure date

280
260
240
220
200
Heart Rate (BPM) Blood Pressure (mmHg)
180
160
140
120
100
80
60
40
20
200
180
160
140
120
100
80
60
40
20
10 12 14 16 18 20 22 0 2 4 6 8 10
Time (hour)
Fig. 15.2 Ambulatory blood pressure and heart rate monitoring blood pressure is ideal; however, the systolic blood pressure dropped
records (based on individual readings) over a 24h period, starting from about 140mmHg during the day to 90mmHg on going to
from about 11 A.M., in a 58 year old woman with bilateral sleep, and diastolic blood pressure dropped from about 80mmHg
NA-AION, and on no medication. This shows that during the day to low 40s during sleep (Reproduced from Hayreh etal. [22].)
twice in 1987. (1) Gittinger and Asdourian [58] reported bilateral vision loss from amiodarone toxic optic neu-
it in 3 eyes of 2 men receiving amiodarone, who were ropathy occurs infrequently, if at all. Turdumambetova
first diagnosed to have ischemic optic neuropathy but etal. [61] reported an incidence of 1.31.8% of amio-
the authors later concluded that We believe our patients darone optic neuropathy and stated that the fundo-
had an amiodarone-induced papillopathy. (2) Feiner scopic picture of amiodarone neuropathy is similar to
et al. [59] reported 13 patients with this condition. classic AION. Murphy and Murphy [62] admirably
According to them The clinical severity of this drug- reviewed the role of amiodarone in the development of
related optic neuropathy is milder than that characteris- optic neuropathy. In most reports it has been assumed
tically described in anterior ischemic optic neuropathy. that a cause-and-effect relationship exists between ami-
They stated that Whether this result was due solely to odarone and the development of amiodarone-related
amiodarone therapy, to the underlying poor health of optic neuropathy; so much so, indeed, that a $20 mil-
these patients, or to a combination of these two factors lion judgment was rendered against Wyeth Ayerst in
is uncertain. Since then a considerable literature has 1997 for omitting blindness as a potential adverse effect
accumulated on amiodarone induced optic neuropa- of amiodarone in its package insert.
thy, mostly anecdotal case reports. Mindel etal. [60] Over the years I have seen cases with so-called
in a closed-label amiodarone (n=837) or placebo amiodarone-related optic neuropathy. In the light
(n=832) prospective double-masked randomized trial of my research on various aspects of NA-AION,
concluded that At the doses commonly used clinically, reviewing the literature on amiodarone-related optic
neuropathy and my clinical experience of such a miodarone. Moreover, there are reports of patients
cases, I [27] recently reviewed the entire controversy, with asymptomatic optic disc edema progressing to
to ascertain whether there is a cause-and-effect rela- visual loss after amiodarone had been discontinued
tionship between amiodarone and the development of [62]. The comparatively higher prevalence of
so-called amiodarone-related optic neuropathy or asymptomatic optic disc edema in NA-AION and
if it is in fact simply NA-AION developing in patients early detection in patients taking amiodarone may
suffering from cardiac arrhythmias and has nothing be due the following reason: Classical NA-AION
to do with amiodarone. I find that the following facts patients are seen only after they have developed
argue against a cause-and-effect relationship between visual loss; however, there is much higher chance
amiodarone and development of optic neuropathy. of that being discovered in patients on amiodarone
because of their frequent ophthalmic evaluation by
1. Patients who take amiodarone have cardiovascular physicians who worry about ophthalmic side-effects
disorders, which are per se well-established risk of amiodarone.
factors for development of NA-AION [12, 20] (see 4. Most importantly, the clinical findings in these
Chap. 14). Many of these patients also have other patients are typically those of NA-AION, and not
risk factors for development of NA-AION as well, like a toxic optic neuropathy.
e.g., arterial hypertension, diabetes mellitus, hyper-
Thus, in the multifactorial scenario of NA-AION, it is
lipidemia and ischemic heart disease [6, 12, 20].
the systemic cardiovascular risk factors rather than the
They are potential candidates for NA-AION whether
amiodarone that cause NA-AION.
they are taking amiodarone or not.
2. Patients on amiodarone often also take other drugs,
e.g., beta-blockers, calcium channel blockers, ACE
inhibitors and other drugs influencing the cardio- Familial NA-AION
vascular system. Our 24-hour ambulatory blood
pressure monitoring studies have shown that There are five reports in the literature representing ten
patients on these drugs are at high risk of develop- unrelated families in which more than one member
ing marked nocturnal arterial hypotension, which is developed NA-AION [6468]. We [68] have shown
a common precipitating factor for development of that this rare entity of familial NA-AION is clinically
NA-AION [12, 19, 21, 22]. In our study at least similar to the classical non-familial NA-AION, with
73% of NA-AION patients reported discovering the the exception that familial NA-AION occurred in
visual loss on waking from sleep and the rest were younger patients and caused much more frequent
unsure of the time of onset [21]. involvement of both eyes than the classical
3. One of the arguments put forward to differentiate NA-AION.
optic neuropathy in patients taking amiodarone As discussed in Chap. 14, NA-AION is a multifac-
from NA-AION is that some of them develop torial disease. The various systemic and local factors
asymptomatic optic disc edema which may later which make a person susceptible to it are discussed at
progress to visual loss. But asymptomatic optic disc length there. These risk factors may be inherited or
edema has been known as an early sign of NA-AION acquired.
since 1981 [1]. Our recent study of 60 NA-AION
eyes with initial asymptomatic optic disc edema
described the clinical entity as incipient NA-AION
Inherited Familial Local and/or Systemic
[2]. Many of them progress to visual loss later on,
but not all. Optic disc edema for various reasons Risk Factors for NA-AION
can persist much longer than the 68weeks, usually
seen with typical NA-AION [63]). In our series There is always the possibility that some common
[1,2] of asymptomatic optic disc edema cases, not inherited local and/or systemic risk factors in a family
one of the 55 patients (60 eyes) with incipient could be responsible for more than one member in the
NA-AION (i.e., NA-AION with asymptomatic family developing NA-AION. In this context, one has
optic disc edema at the initial visit) was taking to consider the following possibilities.
Familial NA-AION 325
Inherited Absent or Small Cup of the Optic Disc (a) Presence of inherited thrombophilic factors in
NA-AION patients: Glueck etal. [72] evaluated in
There is a prevalent misconception in the ophthalmic 12 non-familial classical NA-AION patients (8
community that an absent or small cup of the optic disc with unilateral and 4 bilateral NA-AION) homozy-
plays a primary role in the development of NA-AION. gosity for the methylenetetrahydrofolate reductase
For example, Wang etal. [66] hypothesized that: One (MTHFR) C677T mutation. They claimed that
potential concern is that the familial incidence of disease patients with NA-AION were more likely than
can be explained solely by the inherited anomalous disc. controls to demonstrate this homozygosity.
Available evidence does not support this view, as is evi- Therefore, they concluded that there was a cause-
dent from the following [2, 23, 24, 63, 69, 70]. The and-effect relationship between coagulation disor-
Framingham Eye Study [69] of 4994 eyes of all ages in ders and NA-AION. In sharp contrast, Salomon
the general population reported no cup in 11%, and a etal. [73] evaluated various genetic and acquired
cup-disc ratio 0.1 in 15%, 0.2 in 27% and 0.3 in 27%. thrombophilic markers, including MTHFR C677T
Thus, in that study the cup-disc ratio was0.2 in 53% in 61 patients with classical NA-AION and found
and0.3 in 80%. A prospective study of 248 (126 with no association between NA-AION and these
NA-AION and 122 normal) subjects by Beck etal. [23] thrombophilic factors. Association between one of
showed that 19% of normal subjects had no cup, 11% the platelet glycoprotein polymorphisms and sec-
had a cup-disc ratio of 0.050.149, and 16% had a cup- ond eye involvement in NA-AION [74], and of
disc ratio of 0.15-0.249. This means that 46% of normal vascular tone (endothelial nitric oxide synthase
individuals had either an absent cup or a small cup, com- gene polymorphisms) with NA-AION [75] have
pared to 77% in NA-AION eyes. While no doubt an recently been reported; however, those observa-
absent or smaller cup is significantly more common in tions still need further confirmation.
eyes with NA-AION than in the general population, stud- (b) NA-AION usually not a thromboembolic but a
ies [23, 24, 69] have shown that not all eyes with no cup hypotensive disorder: The subject is discussed at
or a small cup develop NA-AION and, conversely, some length elsewhere (see Chap. 14). It is well estab-
eyes with NA-AION have a larger cup. The role of an lished that stroke is usually a thromboembolic dis-
absent or small cup in the pathogenesis of development order. However, recent studies have shown that
of NA-AION is discussed at length in Chap. 14. That NA-AION is usually a hypotensive disorder (as
shows that an absent or small cup is simply a secondary discussed in Chap. 14, with nocturnal hypoten-
contributing factor, once the process of NA-AION has sion playing an important role) [19, 21, 22, 76].
started, and not a primary factor [2, 2325, 71]. This is further supported by the following
In our series [68], in patients with familial evidence.
NA-AION, there was no difference in the optic disc i. While it is well established that aspirin has a
cup pattern compared to that in classical NA-AION beneficial effect in stroke (usually a throm-
eyes; hence, development of familial NA-AION can in boembolic disorder), NA-AION studies [51,
no way be attributed to that. Moreover, if an absent or 52] have shown that aspirin has no beneficial
small cup were the cause of familial NA-AION, then it effects (being usually a hypotensive disorder).
would occur universally and not so rarely. ii. As discussed above, Salomon et al. [73] did
not find any association between NA-AION
and various thrombophilic factors they evalu-
Inherited Thrombophilic Risk Factors ated. This was also pointed out by others [77].
iii. We [78] recently evaluated the role of tobacco
Because most ophthalmologists and neurologists con- smoking in a prospective study of 624 con-
sider NA-AION a thromboembolic disorder, akin to secutive patients with NA-AION, and found
stroke, some could argue that inheritance of some no association between NA-AION and smok-
thrombophilic factors could play a role in development ing. By contrast, there is a huge volume of lit-
of NA-AION in more than one member of a family. In erature showing a significant association
this context, a consideration of the following two between smoking and stroke a thromboem-
important factors is essential. bolic disorder [79].
iv. We [21] investigated seasonal variation in the NA-AION and tobacco smoking has been shown [78].
occurrence of NA-AION. There is a distinct In our study [68], there was no evidence of any other
difference in the seasonal variation in onset of acquired systemic risk factors in the three families.
NA-AION and systemic vascular accidents. As
discussed elsewhere (see Chap. 14), NA-AION
onset is significantly more frequent in the Hereditary Pattern Seen in Families
Summer (or the hot months) than in the Winter with Familial NA-AION
(or the cold months) (p=.0030). This is in con-
trast to the reported incidence of systemic vas- The majority of reported studies of familial NA-AION
cular accidents, such as myocardial infarction showed no definite pattern of familial inheritance. The
and cerebrovascular accidents, which are more pedigree reported by Berggren et al. [64] included
frequent during winter than other seasons [80 three members with NA-AION. The father married
82]. This indicates a difference in their patho- twice: none of the five siblings from his first marriage
geneses. This also is evidence that it is a serious had any eye problems. Three (a brother and 2 younger
mistake to equate NA-AION with stroke and sisters) of seven brothers and sisters from the second
other systemic vascular accidents. marriage had NA-AION. There was no family history
In the light of the discussion above, there is little evi- on the fathers or mothers side of any eye problem.
dence that thrombophilia plays any significant role in Wang etal. [66], in a mailed survey of NA-AION cases
development of NA-AION [70]. seen at Doheny Eye Institute, Los Angeles, USA found
four patients who reported relatives with NA-AION.
Of the four families, the first had 11 siblings (3 affected
Inherited Systemic Diseases with NA-AION), the second had 7 siblings (2 affected
with NA-AION) and the third and fourth families had
Many studies have shown a significant association 4 siblings each (each with 2 affected); of the nine
between development of NA-AION and the presence patients in this series, 66% were male. Manor [65]
of some heritable systemic diseases. For example, a reported NA-AION in identical female twins.
significant association between diabetes mellitus and The three familial NA-AION pedigrees in our study
NA-AION has been demonstrated [20, 50, 52, 71, 83]. [68] had different patterns of inheritance (Figs.15.3a-c).
Similarly, other systemic heritable diseases, such as In Family One, there were two generations of affected
arterial hypertension, and hyperlipidemia, can also be family members with two cases of male-to-male trans-
associated with NA-AION. It could be argued that mission (Fig. 15.3a), which suggests autosomal domi-
more than one member of a family can develop nant inheritance and excludes X-linked or mitochondrial
NA-AION if several members of a family suffer from inheritance. In Family Two, the mother and three of her
one or more of these inherited systemic diseases. In the four children had NA-AION (Fig.15.3b), which is con-
three families with NA-AION in our series [68], there sistent with either mitochondrial inheritance or auto-
was no evidence of any inheritance of such diseases somal dominant inheritance. Finally, in Family Three,
(one patient had end-stage renal disease from polycys- six males with the same maternal lineage developed
tic kidney associated with hypertension and hyperlipi- NA-AION (Fig.15.3c), suggesting mitochondrial inheri-
demia, and only one had elevated cholesterol alone). tance. A mitochondrial sequence variations study [67] in
From all this evidence, one can conclude that none of Family Three detected a previously unreported mutation
the inherited factors discussed above could have played a (G4132A) in all of the affected family members. The
role in development of familial NA-AION in our series. presence of the G4132A mutation in this family suggests
that it may have a role in the pathogenesis of familial
NA-AION; however, since it was not detected in the
other two families in that series [68], nor in the cases of
Acquired Systemic Risk Factors classical NA-AION, normal control subjects, suspected
Lebers hereditary optic neuropathy and Mitochondrial
Of the patients in our study [68], there was a history of Genome Database (Mitomap) [84], that suggests that the
smoking in 5 of the 14. No association between G4132A mutation is not a cause of classical NA-AION
Familial NA-AION 327
a
I
1 2
II
1 2 3 4 5 5 6 7
III
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25
2 3 4
c I
1 2
II
1 2 3 4
III
1 2 3 4 5 6 7 8 9 10 11 12
IV
1 2 3 4 5 6 7 8 9 10 11 12 13
Fig. 15.3 Familial NA-AION pedigrees: (a) Family One, (b) record documents a history of NAION are indicated by grey
Family Two, and (c) Family Three. Family members with symbols. Family members who were not available for examina-
NA-AION that were examined by one of the authors are indi- tion are indicated by symbols with diagonal lines (Reproduced
cated with black symbols. Family members whose medical from Hayreh etal. [68])
or Lebers hereditary optic neuropathy. Therefore, addi- In summary, our study [68] of three pedigrees with
tional studies will be essential to prove or disprove a familial NA-AION and previous reports [6467] pro-
causal relationship between this mutation (G4132A) and vides evidence that NA-AION may be heritable in
the development of familial NA-AION. Also, we cannot rare pedigrees. However, the inheritance of NA-AION
exclude the possibility that a different genetic factor may in these pedigrees does not appear to follow a single,
be the cause of disease in this family. For example, the obvious Mendelian pattern. The different patterns of
pattern of NA-AION in Family Three is also consistent inheritance observed in pedigrees with familial
with inheritance of an X-linked factor. Similarly, we can- NA-AION may be due to the presence of a range of
not exclude the possibility of autosomal recessive inheri- genetic risk factors that have not yet been
tance with pseudo-dominance in any of these pedigrees. discovered.
Conclusion such cases. During recent years, a consensus has

emerged that it is due to ischemia of the optic nerve
head; that is, it represents NA-AION. It is well estab-
This shows that clinically, the rare entity of familial
lished that the earliest fundus change after visual loss
NA-AION is similar to the classical non-familial
is almost invariably optic disc edema, which evolves in
NA-AION, with the exception that familial NA-AION
about 2months into optic atrophy very typical find-
occurs in younger patients and has much higher
ings and course for NA-AION.
involvement of both eyes than the classical non-famil-
It is important to know the following information
ial NA-AION. The role of genetic factors in familial
about this condition.
NA-AION is not known. In fact, it could be argued that
since NA-AION is a common disease, the possibility 1. The visual loss usually develops hours, days, or
of the occasional occurrence of clusters in families even weeks after systemic bleeding (8085% [89,
without any genetic abnormality cannot be ruled out. 90]) and is uncommon during or immediately after
The potential role of genetic factors in familial bleeding (1520% [89, 90]). Unfortunately, the
NA-AION may be clarified with additional research. syncopal type of transient visual loss has been
One preliminary study has suggested that the G4132A mixed with the true posthemorrhagic visual loss in
mitochondrial mutation may be associated with the the data available in the literature.
disease in at least one pedigree with familial NA-AION 2. The visual loss is usually seen after recurrent hem-
[67]. However, our study [68] indicates that this muta- orrhages, and is very uncommon after a single hem-
tion is not associated with the disease in two other orrhage in a perfectly healthy person.
familial NA-AION pedigrees nor is this mutation asso- 3. In almost all of my [88] cases and others [9194],
ciated with the much more common classical non- visual loss was noted after adequate blood transfu-
familial NA-AION. sion (hence, hemoglobin level was within normal
limits).
4. In many cases in the literature and in almost all my
Posthemorrhagic Amaurosis cases visual loss occurred, when the blood pressure
recorded during the day was within normal limits.
5. Visual loss frequently develops or progresses dur-
In this clinical entity, there is a visual loss following
ing sleep.
distant recurrent or massive systemic hemorrhages,
from a variety of medical conditions, e.g., uterine These factors strongly suggest that in most cases neither
bleeding, gastro-intestinal bleeding, hemoptysis, immediate post-hemorrhage arterial hypotension alone
epistaxis or repeated venesection, often recurrent in nor low hemoglobin level can explain the sudden devel-
nature. This condition was first described by opment of visual loss resulting from NA-AION. One has
Hippocrates [85] almost 25 centuries ago, in a descrip- to look elsewhere for an answer. Available evidence indi-
tion of black disease. He wrote ... the patient vomits cates that the following probably plays a crucial role.
black, bloody material and he has pain in his head and
does not see with his eyes.... (Translation of
Hippocrates paper by my wife from the original Greek
Release of Endogenous Vasoconstrictor
text). It seems that the first definite case of this disease
was reported by Nicholas Fontanus [86] in 1641 in his Agents After Recurrent Hemorrhages
Collection of Rare Observations describing the
visual loss developed after therapeutic venesection. In Hodge etal. [95] found that in dogs hemorrhage causes
1781, Truka de Krzowitz [87] reviewed from the litera- an increase in the concentration of circulating angio-
ture a large number of cases of visual loss after bleed- tensin, even when there is no fall of arterial mean or
ing from different causes. Since then there have been pulse pressure. Similarly, Bunag et al. [96] showed
many serial reviews [88]. that in dogs bleeding caused renal release of renin
Many theories have been postulated over the years whether or not there was measurable change in either
to explain its pathogenesis, which I [88] briefly arterial pressure or total renal blood flow, and they also
reviewed some time ago. I [88] have also seen some found that hemorrhage-induced release of renin
Posthemorrhagic Amaurosis 329
occurred more consistently in dogs fed a low-sodium [19, 22] see Fig. 15.2). The fact that, in my [88]
diet than in those fed a standard diet. These findings cases, the initial visual loss or further deterioration was
contradicted the usual view that release of renin during noticed on waking up in the morning strongly suggests
hemorrhage is entirely dependent on the associated the role of hypotension during sleep in the develop-
arterial hypotension [97]. Bunag etal. [96] concluded ment of NA-AION, as is the case in the vast majority
that neural stimuli are capable of causing release of of NA-AION cases [21], as discussed in Chap. 14.
renin in the absence of gross change in renal perfusion
pressure of flow. A considerable body of evidence has
accumulated showing that the sympathoadrenergic Failure of Autoregulation of the Blood Flow
system can influence renin release, and the subject has
been summarized by Zanchetti and co-workers [98, This happens if the perfusion pressure falls below the
99]. Hodge etal. [100] concluded that the concentra- critical level, because of one or both factors mentioned
tion of angiotensin in the blood is affected by cardio- previously (see Chap. 5).
vascular reflexes. More recent evidence [101, 102] has Thus, a fall of perfusion pressure to a very low
provided the mechanism by which non-hypotensive level, vasoconstriction of vessels in the optic nerve
hemorrhage may evoke the release of renin and other head, accompanied by increase in peripheral vascular
vasoconstrictors. In conclusion, all the available evi- resistance, and failure of autoregulation, together com-
dence indicates that activation of the sympathetic ner- promise the blood flow in the optic nerve head and
vous system and increased circulating levels of three result in the development of NA-AION. If the optic
humoral agents, angiotensin, epinephrine, and vaso- nerve head ischemia is subclinical, symptomless optic
pressin, all lead to vasoconstriction in a circumstance disc edema may develop in the eye [1, 2], as in one of
such as non-hypotensive hemorrhage, and at least one my cases of posthemorrhagic amaurosis at the initial
of the humoral factors (epinephrine) is also known to visit [88].
increase renin secretion by the kidney [101, 102]. Interindividual variation in the development of
From the evidence in the literature, and my various NA-AION after massive systemic hemorrhage could
basic, experimental and clinical studies, I [88] postu- be due to differences in individual responses and sus-
lated that a combination of the following would com- ceptibility. That is, in certain individuals, much more
promise the blood flow in the optic nerve head. angiotensin and other endogenous vasoconstrictor
agents may be liberated after an equal amount of hem-
orrhage than in others, because of interindividual dif-
Increase in Peripheral Vascular Resistance ferences in the liberation of catecholamines and other
sympathoadrenergic and vasomotor neural responses.
As discussed elsewhere [47] and in Chap. 4, the It is possible that recurrent hemorrhages may also
absence of a blood-optic nerve head barrier allows the produce increased platelet aggregation, and that, in
endogenous vasoconstrictor agents, such as angio- turn, may produce microembolism and play some role
tensin, epinephrine, and vasopressin (liberated in in the production of NA-AION. In my cases [88], there
response to recurrent hemorrhages), to leak from the was an increase in platelet counts soon after the
peripapillary choroid into the optic nerve head tissues, hemorrhage.
producing vasoconstriction. This may also be produced This hypothesis thus may help to explain most of
by vasoconstriction of the peripapillary choroidal the features of the visual loss after distant systemic
arteries (major source of blood supply to the ONH). hemorrhages. There seems little evidence to support
The vasoconstriction by itself may cause NA-AION. the theory that a mere fall in hemoglobin level causes
NA-AION. There is no evidence in the literature that a
fall in hemoglobin level causes release of angiotensin
Fall of Systemic Blood Pressure or other vasopressor agents.
In conclusion, from the available evidence, it is
This occurs immediately after the bleeding owing to postulated that release of angiotensin and other
hypovolemia or during sleep (although the blood pres- endogenous vasoconstrictor agents (secondary to recur
sure during waking hours may be within normal limits rent massive systemic hemorrhages, with or without
arterial hypotension) is most probably an important latter could not be attributed to the former. Since
factor in the production of NA-AION. Arterial hypoten- Ganssauge et al.s [106] patient recognized the
sion seems to be an additional important factor. visual field defect, the possibility that it was there
Increased platelet aggregation may also play a role. before and he was unaware of it cannot be ruled out. I
There is no available evidence to indicate that fall in have seen that in some patients with NA-AION.
hemoglobin level per se is responsible for the develop- Hosseini and Razeghinejads [105] patient woke up
ment of NA-AION. with visual loss 1week after the injection. There is an
explanation for the delay between the intravitreal
injection and development of visual loss due to
NA-AION. As I [1] reported in 1981, asymptomatic
NA-AION Following Intravitreal
optic disc edema is an early sign of NA-AION. More
Anti-VEGF Therapy recently, we described the clinical entity of Incipient
non-arteritic anterior ischemic optic neuropathy in
Some have advocated using intravitreal vascular 60 eyes [2], in which, again, NA-AION patients were
endothelial growth factor (VEGF) inhibitory drugs for seen initially with asymptomatic optic disc edema and
management of NA-AION, arguing that since it 25% of them later on progressed to classical NAION
reduces macular edema, it may similarly reduce the with a median time of progression of 5.8 (interquartile
optic disc edema and be beneficial in NA-AION. In 3.210.1) weeks after the initial diagnosis. Thus, con-
2007, Bennett etal. [103] gave one patient an intravit- trary to the universal impression, the time of visual
real injection of Bevacizumab (Avastin) a vascular loss is not always the time of onset of NA-AION it is
endothelial growth factor (VEGF) inhibitory drug, only the time when the patient notices it. In these two
3weeks after the onset of NA-AION in one eye. They cases [105, 106], the intravitreal injection most prob-
claimed visual improvement followed, and that bevaci- ably started the chain of events. It initially produced
zumab improved visual acuity in the eye by reducing Incipient non-arteritic anterior ischemic optic neu-
optic disc edema. But it is impossible to judge the ropathy; however, the patients were seen by the oph-
effectiveness of a mode of treatment from one eye thalmologist only when they developed and noticed
when about 40% of NA-AION eyes show spontaneous visual loss 12weeks later.
visual acuity improvement [9, 11]. Moreover, recently There is another, more recent case report [107]
Pece etal. [104] reported three patients with NA-AION where the authors claimed that a 70year old woman
who had intravitreal ranibizumab injection 12 days with a choroidal neovascular membrane in the macular
after onset. They found that intravitreal ranibizumab, region with normal IOP, developed NA-AION
although it reduced the optic disc edema, did not result 1 month after intravitreal injection of bevacizumab.
in any functional improvement. They attributed that to the bevacizumab. I find that
In contrast to that, there have recently been two there was no cause-and-effect relationship between the
reports of intravitreal bevacizumab injection actually intravitreal injection of bevacizumab and development
resulting in the development of NA-AION. Hosseini of NA-AION in this case, as discussed below. It was
and Razeghinejad [105] reported a 72year old woman purely a coincidence, and the explanation given by the
who woke up with visual loss 1week after intravitreal authors was not valid.
injection of bevacizumab was given for exudative Relevant to the development of NA-AION follow-
age-related macular degeneration and was diagnosed ing intravitreal injection of anti-VEGF reported above,
to have NA-AION. Ganssauge et al. [106] report a the whole subject of intravitreal injection of bevaci-
51year-old male who recognized visual field defects zumab requires detailed discussion, because of the
2weeks after injection of intravitreal bevacizumab for current enthusiasm for the use of anti-VEGF drugs in
choroidal neovascularisation, and was diagnosed as age-related macular degeneration, diabetic macular
having NA-AION. edema and other ocular conditions. Naturally the ques-
In these two cases [105, 106], it could be argued tion arises: what is the mechanism of development of
that because there was a time lag of 1 and 2 weeks NA-AION following intravitreal injection of bevaci-
respectively between the intravitreal injection of beva- zumab? The ONH circulation depends upon the perfu-
cizumab and the development of NA-AION, the sion pressure (Perfusion pressure=mean blood
NA-AION Following Intravitreal Anti-VEGF Therapy 331
pressure minus IOP) in its capillaries. Following intra- require IOP-lowering therapy. Good et al. [116] in a
vitreal injection of anti-VEGF drugs two types of rise retrospective study of 215 eyes found that sustained
of intraocular pressure have been reported. elevated IOP in patients receiving intravitreal anti-
VEGF injections is significant. Additionally, these
data suggest the possibility of a heightened risk for
further elevation of IOP in patients with pre-existing
Short-term Rise of IOP glaucoma who receive either bevacizumab or ranibi-
zumab. At the annual meeting of the Association of
Intravitreal injection increases the volume in the eye- Research and Vision in Ophthalmology in May 2009,
ball, thereby resulting in an immediate, transient rise there was another report (I do not remember the names
of IOP in all eyes. After intravitreal injection of beva- of the authors) of four cases with similar persistent
cizumab, Kim et al. [108] reported a rise of IOP to elevation of IOP following intravitreal injection of
44 mmHg (range, 487 mmHg.), Falkenstein et al. bevacizumab. At the annual meeting of the American
[109] a mean IOP of 36 mmHg, and Hollands et al. Academy of Ophthalmology in October 2010 in
[110] an IOP of 25 mmHg or higher at 30 min; one Chicago, Freund and colleagues reported development
patient required a 1week course of glaucoma medica- of elevated IOP in 22 eyes of 20 normotensive patients
tion to control the IOP. Semoun etal. [111] reported a following long-term intravitreal ranibizumab therapy
case of an acute angle-closure glaucoma that occurred for neovascular age-related macular degeneration. In
immediately after an intravitreal injection of 0.05ml this series the mean IOP was 30.5 (range 2258)
of bevacizumab. mmHg. In them the baseline IOP was 17.2 (range
1421) mmHg. All these patients required treatment
for the elevated IOP. While sustained IOP rise follow-
ing anti-VEGF therapy has been recorded by these
Severe and Sustained Rise of IOP reports, we do not have information about how com-
mon it is after intravitreal anti-VEGF therapy on a
There are several reports of this in the literature. For long-term follow-up, because of the almost invariable
example, Jalil etal. [112] reported a patient who was practice of the retina specialists to monitor IOP only
found to have an IOP of 56mmHg 3days after intrav- immediately post-injection and not serially later on; it
itreal injection of bevacizumab. Bakri et al. [113] is possible that this sustained rise of IOP following
reported sustained ocular hypertension as high as 30, intravitreal anti-VEGF therapy may be much more
34, 46, and 50 mmHg in four patients, following prevalent than reported in the literature.
0.5mg intravitreal ranibizumab for age-related macu- Similarly, there are many reports showing a sub-
lar degeneration. The elevated IOP persisted across stantial rise in IOP a few days or weeks after intravit-
several visits, requiring control with topical glaucoma real triamcinolone (which has been used in management
therapy and in two cases the addition of a systemic of NA-AION see Chap. 17) in as high as about 40%
carbonic anhydrase inhibitor. None of the patients had [117]; by contrast, oral corticosteroid therapy for
a previous history of glaucoma. Kahook et al. [114] NA-AION did not have that effect on IOP during a
reported six cases of significant and sustained eleva- short-term treatment [118].
tion of lOP after intravitreal injections of bevacizumab
for age-related macular degeneration, all requiring
IOP-lowering therapy. Adelman etal. [115] in a retro-
Importance of Rise of IOP in Relation
spective study of 116 patients found that 3.45% devel-
oped sustained elevated IOP after multiple intravitreal to NA-AION
injections of bevacizumab 1.5 mg/0.06 mL and/or
ranibizumab 0.5 mg/0.05 mL. They concluded that This is an important issue. As mentioned above, blood
persistent ocular hypertension may occur after intrav- flow in the optic nerve head capillaries depends upon the
itreal anti-VEGF injection in patients with no previous perfusion pressure; the perfusion pressure is the differ-
diagnosis of glaucoma or ocular hypertension, and that ence between the mean blood pressure and the IOP. Thus,
may persist across several visits and patients may a rise of IOP results in a fall of perfusion pressure.
Use of Intravitreal Anti-VEGF Therapy this disparity may be that retina specialists, who give
as Treatment for NA-AION intravitreal injections of anti-VEGF for age-related
macular degeneration or macular edema, are usually
This has been advocated by some. But the ONH of focused primarily on the macular changes, and may
NA-AION patients already has precarious circulation. not pay critical attention to changes in the optic disc at
Under those circumstances, even a transient rise of the same time. Any further visual deterioration follow-
IOP, letalone a prolonged, sustained high IOP, has the ing the intravitreal injection may be attributed to wors-
potential to further compromise the circulation and ening of macular pathology. Moreover, the central
result in further ONH damage and visual loss. Thus, visual acuity is not always affected in NA-AION (see
giving intravitreal injection of anti-VEGF in the man- Chap. 16), and visual fields are the only ones which are
agement of NA-AION can in fact be harmful. affected; however, no visual fields are usually plotted
by the retina specialists in patients who are being
treated by intravitreal injections of anti-VEGF for age-
Use of Intravitreal Anti-VEGF for Treatment related macular degeneration or macular edema. All
of Age-Related Macular Degeneration these confounding factors may result in missing the
and Macular Edema development of NA-AION in these cases. Also dia-
betic macular edema is the common indication for
Age-related macular degeneration is a common condi- intravitreal injections of anti-VEGF; it is well-estab-
tion, seen mostly in the elderly population. Macular lished now that NA-AION is much more common
edema is also common in diabetics. For both these condi- among diabetics than non-diabetics [71], therefore,
tions, intravitreal anti-VEGF therapy is currently the treat- their eyes are susceptible to NA-AION.
ment of choice. Therefore, tens of thousands of patients In view of all these considerations, when treating
with these conditions are being treated every year, all over age-related macular degeneration or diabetic macular
the world, with intravitreal anti-VEGF therapy. In view of edema, on a risk/benefit ratio, it may be prudent to avoid
that, it is essential to discuss the possible harmful effect of any immediate or prolonged rise of IOP by taking pre-
this rise of IOP following anti-VEGF therapy. cautionary measures and measuring IOP frequently for
In patients who do not already have had NA-AION, several weeks. For persons who have already had
the chance of developing NA-AION with this transient NA-AION in one eye, the risk of the second eye devel-
or sustained high IOP is limited mainly to those who oping NA-AION is much higher than for those who
have low perfusion pressure because of the various never had NA-AION before. A high IOP can precipitate
predisposing risk factors or have other predisposing development of NA-AION in persons with predisposing
risk factors (see Chap. 14). Persons who develop age- risk factors and can worsen the visual loss in those who
related macular degeneration fall into that category; already have NA-AION. This shows that intravitreal
there is growing evidence that chronic macular choroi- injections of anti-VEGF agents or of triamcinolone have
dal ischemia plays a role in the development of age- the potential to be harmful a factor not realized in the
related macular degeneration [119122]. This chronic current enthusiasm for the widespread use of anti-VEGF
macular choroidal ischemia usually reflects what is agents. Moreover, anti-VEGF agents do rarely have sys-
going on in the rest of the eye. Moreover, patients with temic complications, e.g., acute elevation of systemic
age-related macular degeneration generally have other blood pressure, cerebrovascular accidents, myocardial
systemic risk factors, being elderly (as the name infarctions, and even deaths have been reported in some
implies), including a high prevalence of various sys- cases [129]. They must be used with caution.
temic vascular diseases [122128]. All these factors
make patients with age-related macular degeneration
susceptible to NA-AION. Abnormal levels of high IOP
in them may precipitate the development of NA-AION. Conclusion
Given the current vogue for intravitreal anti-VEGF
therapy, the two cases of development of NA-AION There are several reports showing that intravitreal anti-
following intravitreal injection recently reported may VEGF therapy can cause severe and sustained rise of IOP
represent only the tip of the iceberg. The reasons for in some eyes. This has important implications for the
References 333
development of NA-AION in susceptible persons, such College of Physicians and Surgeons, Glasgow) London:
as the elderly with cardiovascular risk factors, with Kimpton; 1968. p. 136141.
16. Sanders MD. Ischaemic papillopathy. Trans Ophthalmol Soc
age-related macular degeneration, or with diabetes. In UK. 1971;91:36986.
persons with such risk factors whose ONH circulation is 17. Boghen DR, Glaser JS. Ischaemic optic neuropathy: the
already precarious, this rise of IOP has the potential to clinical profile and history. Brain. 1975;98(4):689708.
further compromise the circulation and result in the devel- 18. Hayreh SS. Experimental allergic encephalomyelitis. II:
retinal and other ocular manifestations. Invest Ophthalmol
opment of NA-AION. In the management of NA-AION, Vis Sci. 1981;21(2):27081.
moreover, intravitreal injection of anti-VEGF not only 19. Hayreh SS, Zimmerman MB, Podhajsky P, Alward WL.
has no beneficial effect but can in fact be harmful. Nocturnal arterial hypotension and its role in optic nerve
head and ocular ischemic disorders. Am J Ophthalmol.
1994;117(5):60324.
20. Hayreh SS, Joos KM, Podhajsky PA, Long CR. Systemic
diseases associated with nonarteritic anterior ischemic optic
References neuropathy. Am J Ophthalmol. 1994;118(6):76680.
21. Hayreh SS, Podhajsky PA, Zimmerman B. Nonarteritic
anterior ischemic optic neuropathy: time of onset of visual
1. Hayreh SS. Anterior ischemic optic neuropathy. V: optic loss. Am J Ophthalmol. 1997;124(5):6417.
disc edema an early sign. Arch Ophthalmol. 1981;99 22. Hayreh SS, Podhajsky P, Zimmerman MB. Role of noctur-
(6):103040. nal arterial hypotension in optic nerve head ischemic disor-
2. Hayreh SS, Zimmerman MB. Incipient nonarteritic anterior ders. Ophthalmologica. 1999;213(2):7696.
ischemic optic neuropathy. Ophthalmology. 2007;114(9): 23. Beck RW, Servais GE, Hayreh SS. Anterior ischemic optic
176372. neuropathy. IX: cup-to-disc ratio and its role in pathogene-
3. Hayreh SS. Fluids in the anterior part of the optic nerve in sis. Ophthalmology. 1987;94(11):15038.
health and disease. Surv Ophthalmol. 1978;23(1):125. 24. Hayreh SS, Zimmerman MB. Nonarteritic anterior ischemic
4. Tso MOM, Hayreh SS. Optic disc edema in raised intracra- optic neuropathy: refractive error and its relationship to cup/
nial pressure. IV: axoplasmic transport in experimental disc ratio. Ophthalmology. 2008;115(12):227581.
papilledema. Arch Ophthalmol. 1977;95(8):145862. 25. Hayreh SS. Erectile dysfunction drugs and non-arteritic
5. Hayreh SS. Optic disc edema in raised intracranial pressure. anterior ischemic optic neuropathy: is there a cause
V: pathogenesis. Arch Ophthalmol. 1977;95(9):155365. and effect relationship? J Neuroophthalmol. 2005;25(4):
6. McLeod D, Oji EO, Kohner EM, Marshall J. Fundus signs in 2958.
temporal arteritis. Br J Ophthalmol. 1978;62(9):5914. 26. Hayreh SS. Non-arteritic anterior ischaemic optic neuropa-
7. Fink BR, Byers MR, Middaugh ME. Dynamics of colchi- thy and phosphodiesterase-5 inhibitors. Br J Ophthalmol.
cine effects on rapid axonal transport and axonal morphol- 2008;92(12):157780.
ogy. Brain Res. 1973;56:299311. 27. Hayreh SS. Amiodarone, erectile dysfunction drugs, and
8. Hayreh SS, Podhajsky P. Visual field defects in anterior isch- non-arteritic ischemic optic neuropathy. J Neuroophthalmol.
emic optic neuropathy. Doc Ophthalmol Proc Ser. 1979;19: 2006;26(2):1545.
5371. 28. Hayreh SS. Blood flow in the optic nerve head and factors that
9. The Ischemic Optic Neuropathy Decompression Trial may influence it. Prog Retin Eye Res. 2001;20(5):595624.
Research Group. Optic nerve decompression surgery for non- 29. Beavo JA. Cyclic nucleotide phosphodiesterases: functional
arteritic anterior ischemic optic neuropathy (NAION) is not implications of multiple isoforms. Physiol Rev.
effective and may be harmful. JAMA. 1995;273(8):62532. 1995;75(4):72548.
10. Hayreh SS. Posterior ischaemic optic neuropathy: clinical 30. Zusmcan RM, Morales A, Glasser DB, Osterloh IH. Overall
features, pathogenesis, and management. Eye. 2004;18(11): cardiovascular profile of sildenafil citrate. Am J Cardiol.
1188206. 1999;83(5A):35C44C.
11. Hayreh SS, Zimmerman MB. Nonarteritic anterior ischemic 31. Mahmud A, Hennessy M, Feely J. Effect of sildenafil on
optic neuropathy: natural history of visual outcome. blood pressure and arterial wave reflection in treated hyper-
Ophthalmology. 2008;115(2):298305. tensive men. J Hum Hypertens. 2001;15(10):70713.
12. Hayreh SS. Acute ischemic disorders of the optic nerve: 32. US Food and Drug Administration. Information for health-
pathogenesis, clinical manifestations and management. care professionals: sildenafil (marketed as Viagra and
Ophthalmol Clin North Am. 1996;9(3):40742. Revatio), Vardenafil (marketed as Levitra), Tadalafil (mar-
13. Hayreh SS. Ischemic optic neuropathy. Prog Retin Eye Res. keted as Cialis). 2007 [updated 8/19/2009; cited 2010 July
2009;28(1):3462. 11];Available from: www.fda.gov/Drugs/DrugSafety/
14. Georgiads G, Konstas P, Stangos N. Rflexions issues de PostmarketDrugSafetyInformationforPatientsandProviders/
ltude de nombreux cas de pseudo-papillite vasculaire ucm124841.htm.
(Study of numerous cases of vascular pseudopapillitis). Bull 33. Kloner RA. Pharmacology and drug interaction effects of
Mm Soc Fr Ophtalmol. 1966;79:50639. the phosphodiesterase 5 inhibitors: focus on alpha-blocker
15. Foulds WS. Ischaemic optic neuropathy. In: Cant JS, editor. interactions. Am J Cardiol. 2005;96(12B):42M6M.
Proceedings, William Mackenzie Centenary Symposium on 34. Hayreh SS. Evaluation of optic nerve head circulation: review
the ocular circulation in health and disease (1968: Royal of the methods used. J Glaucoma. 1997;6(5):31930.
35. Pomeranz HD, Bhavsar AR. Nonarteritic ischemic optic 54. Hayreh SS. Anterior ischemic optic neuropathy. IV: occur-
neuropathy developing soon after use of sildenafil (Viagra): rence after cataract extraction. Arch Ophthalmol.
a report of seven new cases. J Neuroophthalmol. 1980;98(8):14106.
2005;25(1):913. 55. Hayreh SS. The blood supply of the optic nerve head and the
36. Grunwald JE, Siu KK, Jacob SS, Dupont J. Effect of sildena- evaluation of it: myth and reality. Prog Retin Eye Res.
fil citrate (Viagra) on the ocular circulation. Am J Ophthalmol. 2001;20(5):56393.
2001;131(6):7515. 56. Hayreh SS. Management of non-arteritic anterior ischemic
37. Petrig BL, Riva CE, Hayreh SS. Laser Doppler flowmetry optic neuropathy. Graefes Arch Clin Exp Ophthalmol.
and optic nerve head blood flow. Am J Ophthalmol. 2009;247(12):1595600.
1999;127(4):41325. 57. Enos WF, Holmes RH, Beyer J. Coronary disease among
38. Hayreh SS. Effect of sildenafil citrate (Viagra) on the ocular United States soldiers killed in action in Korea; preliminary
circulation (correspondence). Am J Ophthalmol. report. J Am Med Assoc. 1953;152(12):10903.
2002;133(1):16970. 58. Gittinger Jr JW, Asdourian GK. Papillopathy caused by ami-
39. Bollinger K, Lee MS. Recurrent visual field defect and isch- odarone. Arch Ophthalmol. 1987;105(3):34951.
emic optic neuropathy associated with tadalafil rechallenge. 59. Feiner LA, Younge BR, Kazmier FJ, Stricker BH, Fraunfelder
Arch Ophthalmol. 2005;123(3):4001. FT. Optic neuropathy and amiodarone therapy. Mayo Clin
40. Morgan JC, Alhatou M, Oberlies J, Johnston KC. Transient Proc. 1987;62(8):70217.
ischemic attack and stroke associated with sildenafil (Viagra) 60. Mindel JS, Anderson J, Hellkamp A, Johnson G, Poole JE,
use. Neurology. 2001;57(9):17301. Mark DB, et al. Absence of bilateral vision loss from
41. McCulley TJ, Lam BL, Marmor MF, Hoffman KB, Luu JK, amiodarone: a randomized trial. Am Heart J.
Feuer WJ. Acute effects of sildenafil (viagra) on blue-on- 2007;153(5):83742;erratum in Am Heart J. 2007 Jul;
yellow and white-on-white Humphrey perimetry. J 2154(2001):2053.
Neuroophthalmol. 2000;20(4):2278. 61. Turdumambetova G, Bredehorn T, Duncker GI. Okulare
42. Sivaswamy L, Vanstavern GP. Ischemic optic neuropathy in Nebenwirkungen der Amiodaron-Therapie (Ocular side-
a child. Pediatr Neurol. 2007;37(5):3712. effects associated with amiodarone therapy). Klin Monbl
43. Hayreh SS. Duke-Elder lecture: systemic arterial blood pres- Augenheilkd. 2005;222(6):48592.
sure and the eye. Eye. 1996;10(1):528. 62. Murphy MA, Murphy JF. Amiodarone and optic neuropa-
44. Hayreh SS. Role of nocturnal arterial hypotension in the thy: the heart of the matter. J Neuroophthalmol. 2005;25(3):
development of ocular manifestations of systemic arterial 2326.
hypertension. Curr Opin Ophthalmol. 1999;10(6): 63. Hayreh SS, Zimmerman MB. Optic disc edema in non-arter-
47482. itic anterior ischemic optic neuropathy. Graefes Arch Clin
45. Gerometta R, Alvarez LJ, Candia OA. Effects of sildenafil Exp Ophthalmol. 2007;245(8):110721.
and tadalafil on intraocular pressure in sheep: implications 64. Berggren L, Thorburn W, Fodstad H. Three cases of non-
for aqueous humor dynamics. Invest Ophthalmol Vis Sci. inflammatory ischemic optic neuropathy occurring in the
2010;51(6):313944. same family. Acta Ophthalmol (Copenh). 1974;52(5):
46. Phillips BG, Kato M, Pesek CA, Winnicki M, Narkiewicz K, 58995.
Davison D, et al. Sympathetic activation by sildenafil. 65. Manor RS. Nonarteritic ischemic optic neuropathy in identi-
Circulation. 2000;102(25):306873. cal female twins: improvement of visual outcome in one by
47. Hayreh SS, Servais GE, Virdi PS. Fundus lesions in malig- optic nerve decompression. Arch Ophthalmol. 1990;108(8):
nant hypertension. V: hypertensive optic neuropathy. 10678.
Ophthalmology. 1986;93(1):7487. 66. Wang MY, Sadun F, Levin LB, LaBree L, Feldon SE.
48. Pomeranz HD, Smith KH, Hart Jr WM, Egan RA. Sildenafil- Occurrence of familial nonarteritic anterior ischemic optic
associated nonarteritic anterior ischemic optic neuropathy. neuropathy in a case series. J Neuroophthalmol. 1999;19(2):
Ophthalmology. 2002;109(3):5847. 1447.
49. Harris A, Kagemann L, Ehrlich R, Ehrlich Y, Lopez CR, 67. Fingert JH, Grassi MA, Janutka JC, East JS, Howard JG,
Purvin VA. The effect of sildenafil on ocular blood flow. Br Sheffield VC, etal. Mitochondrial variant G4132A is associ-
J Ophthalmol. 2008;92(4):46973. ated with familial non-arteritic anterior ischemic optic neu-
50. Beri M, Klugman MR, Kohler JA, Hayreh SS. Anterior isch- ropathy in one large pedigree. Ophthalmic Genet.
emic optic neuropathy. VII: incidence of bilaterality and 2007;28(1):17.
various influencing factors. Ophthalmology. 1987; 68. Hayreh SS, Fingert JH, Stone E, Jacobson DM. Familial
94(8):10208. non-arteritic anterior ischemic optic neuropathy. Graefes
51. Beck RW, Hayreh SS, Podhajsky PA, Tan ES, Moke PS. Arch Clin Exp Ophthalmol. 2008;246(9):1295305.
Aspirin therapy in nonarteritic anterior ischemic optic neu- 69. Leibowitz HM, Krueger DE, Maunder LR, Milton RC, Kini
ropathy. Am J Ophthalmol. 1997;123(2):2127. MM, Kahn HA, et al. The Framingham Eye Study mono-
52. Newman NJ, Scherer R, Langenberg P, Kelman S, Feldon S, graph: an ophthalmological and epidemiological study of
Kaufman D, etal. The fellow eye in NAION: report from the cataract, glaucoma, diabetic retinopathy, macular degenera-
ischemic optic neuropathy decompression trial follow-up tion, and visual acuity in a general population of 2631 adults,
study. Am J Ophthalmol. 2002;134(3):31728. 19731975. Surv Ophthalmol. 1980;24(Suppl):335610.
53. Hayreh SS, Podhajsky PA, Zimmerman B. Ipsilateral recur- 70. Hayreh SS. Non-arteritic anterior ischemic optic neuropathy
rence of nonarteritic anterior ischemic optic neuropathy. Am and thrombophilia. Graefes Arch Clin Exp Ophthalmol.
J Ophthalmol. 2001;132(5):73442. 2009;247(5):57781.
References 335
71. Hayreh SS, Zimmerman MB. Nonarteritic anterior ischemic 86. Fontanus N. Observationum rariorum analecta.
optic neuropathy: clinical characteristics in diabetic patients Amstelodami: sumptibus Henrici Laurentii; 1641.
versus nondiabetic patients. Ophthalmology. 2008;115(10): 87. von Krowitz T. Historia amauroseos omnis aevi observata
181825. medica continens. Vienna: R. Graeffer; 1781.
72. Glueck CJ, Wang P, Bell H, Rangaraj V, Goldenberg N. 88. Hayreh SS. Anterior ischemic optic neuropathy. VIII: clini-
Nonarteritic anterior ischemic optic neuropathy: associa- cal features and pathogenesis of post-hemorrhagic amauro-
tions with homozygosity for the C677T methylenetetrahy- sis. Ophthalmology. 1987;94(11):1488502.
drofolate reductase mutation. J Lab Clin Med. 89. Singer K. ber Sehstrungen nach Blutverlust. Beitr
2004;143(3):18492. Augenheilkd. 1904;6:130.
73. Salomon O, Huna-Baron R, Kurtz S, Steinberg DM, 90. Locket S. Blindness associated with haemorrhage. Br J
Moisseiev J, Rosenberg N, etal. Analysis of prothrombotic Ophthalmol. 1949;33(9):54355.
and vascular risk factors in patients with nonarteritic anterior 91. Hartmann E, Parfonry J. Ccit par perte de sang amliore
ischemic optic neuropathy. Ophthalmology. 1999;106(4): par lactylcholine mais conservant un rtrcissement bina-
73942. sal du champ visuel. Bull Mm Soc Fr Ophtalmol.
74. Salomon O, Rosenberg N, Steinberg DM, Huna-Baron R, 1934;52:5667.
Moisseiev J, Dardik R, etal. Nonarteritic anterior ischemic 92. Hollenhorst RW, Wegener HP. Loss of vision after distant
optic neuropathy is associated with a specific platelet poly- hemorrhage. Am J Med Sci. 1950;219(2):20918.
morphism located on the glycoprotein Iba gene. 93. Ballen PH, Fox MJ, Weissman GS. Ischemic optic neu-
Ophthalmology. 2004;111(1):1848. ropathy secondary to intestinal hemorrhage. Ann
75. Sakai T, Shikishima K, Matsushima M, Kitahara K. Ophthalmol. 1985;17(8):4868.
Endothelial nitric oxide synthase gene polymorphisms in 94. Caballero Presencia A, Martinez Hernandez A, Diaz Guia
non-arteritic anterior ischemic optic neuropathy. Graefes E. Amaurosis following blood loss. Ophthalmologica.
Arch Clin Exp Ophthalmol. 2007;245(2):28892. 1985;191(2):11921.
76. Hayreh SS. Risk factors in AION. Ophthalmology. 95. Hodge RL, Lowe RD, Vane JR. The effect of acute increases
2001;108(10):17178. in blood volume on the concentration of circulating angio-
77. Abu-Amero KK, Bosley TM. Prothrombotic and atheroscle- tensin in dogs. J Physiol. 1965;181:5960P.
rotic risk factors lack significance in NAION patients har- 96. Bunag RD, Page IH, McCubbin JW. Neural stimulation of
bouring mitochondrial DNA mutations. Br J Ophthalmol. release of renin. Circ Res. 1966;19(4):8518.
2006;90(1):11920. 97. Huidobro F, Braun-Menendez E. Secretion of renin by
78. Hayreh SS, Jonas JB, Zimmerman MB. Nonarteritic anterior intact kidney. Am J Physiol. 1942;137(1):4755.
ischemic optic neuropathy and tobacco smoking. 98. Zanchetti A, Stella A. Neural control of renin release. Clin
Ophthalmology. 2007;114(4):8049. Sci Mol Med. 1975;2(Suppl):215s23.
79. Dagenais GR, Yi Q, Lonn E, Sleight P, Ostergren J, Yusuf S. 99. Zanchetti A, Stella A, Leonetti G, Morganti A, Terzoli L.
Impact of cigarette smoking in high-risk patients participat- Control of renin release: a review of experimental evidence
ing in a clinical trial: a substudy from the Heart Outcomes and clinical implications. Am J Cardiol. 1976;37(4):
Prevention Evaluation (HOPE) trial. Eur J Cardiovasc Prev 67591.
Rehabil. 2005;12(1):7581. 100. Hodge RL, Lowe RD, Vane JR. Increased angiotensin for-
80. Bull GM. Meteorological correlates with myocardial and mation in response to carotid occlusion in the dog. Nature.
cerebral infarction and respiratory disease. Br J Prev Soc 1966;211(5048):4913.
Med. 1973;27(2):10813. 101. Thames MD. Renin release: reflex control and adrenergic
81. Bull GM, Morton J. Environment, temperature and death mechanisms. J Hypertens. 1984;2 Suppl 1:5777.
rates. Age Ageing. 1978;7(4):21024. 102. Kopp U, DiBona GF. Catecholamines and neurosympa-
82. Clark CV. Seasonal variation in incidence of brachial and thetic control of renal function. In: Fisher JW, editor.
femoral emboli. Br Med J (Clin Res Ed). Kidney hormones. 3rd ed. London: Fisher-Duncan Ltd;
1983;287(6399):1109. 1986. p. 62160.
83. Ischemic Optic Neuropathy Decompression Trial Study 103. Bennett JL, Thomas S, Olson JL, Mandava N. Treatment of
Group. Characteristics of patients with nonarteritic anterior nonarteritic anterior ischemic optic neuropathy with intrav-
ischemic optic neuropathy eligible for the Ischemic Optic itreal bevacizumab. J Neuroophthalmol. 2007;27(3):
Neuropathy Decompression Trial. Arch Ophthalmol. 23840.
1996;114(11):136674. 104. Pece A, Querques G, Quinto A, Isola V. Intravitreal ranibi-
84. Brandon MC, Lott MT, Nguyen KC, Spolim S, Navathe SB, zumab injection for nonarteritic ischemic optic neuropathy.
Baldi P, etal. MITOMAP: a human mitochondrial genome J Ocul Pharmacol Ther. 2010;26(5):5237.
database2004 update. Nucleic Acids Res. 2005;33(Database 105. Hosseini H, Razeghinejad MR. Anterior ischemic optic
issue):D6113. neuropathy after intravitreal injection of bevacizumab. J
85. Hippocrates. Oeuvres completes dHippocrates. In Littre E, Neuroophthalmol. 2009;29(2):1601.
ed. Traduction nouvelle avec le texte grec en regard, 106. Ganssauge M, Wilhelm H, Bartz-Schmidt KU, Aisenbrey
collationne sur les manuscrits et toutes les editions: S. Non-arteritic anterior ischemic optic neuropathy
accompagnee dune introduction, de commentaires (NA-AION) after intravitreal injection of bevacizumab
medicaux, de variantes et de notes philologiques; suivie (Avastin) for treatment of angoid streaks in pseudoxan-
dune table generale des matieres. Paris: Chez J.B. Bailliere; thoma elasticum. Graefes Arch Clin Exp Ophthalmol.
193961. p. 1103. 2009;247(12):170710.
107. Bodla AA, Rao P. Non-arteritic ischemic optic neuropathy 120. Ross RD, Barofsky JM, Cohen G, Baber WB, et al.
followed by intravitreal bevacizumab injection: is there an Presumed macular choroidal watershed vascular filling,
association? Indian J Ophthalmol. 2010;58(4):34950. choroidal neovascularization, and systemic vascular dis-
108. Kim JE, Mantravadi AV, Hur EY, Covert DJ. Short-term ease in patients with age-related macular degeneration. Am
intraocular pressure changes immediately after intravitreal J Ophthalmol. 1998;125:7180.
injections of anti-vascular endothelial growth factor agents. 121. Xu W, Grunwald JE, Metelitsina TI, DuPont JC, et al.
Am J Ophthalmol. 2008;146(6):9304. Association of risk factors for choroidal neovascularization
109. Falkenstein IA, Cheng L, Freeman WR. Changes of in age-related macular degeneration with decreased foveo-
intraocular pressure after intravitreal injection of bevaci- lar choroidal circulation. Am J Ophthalmol. 2010;
zumab (Avastin). Retina. 2007;27(8):10447. 150(2):407.
110. Hollands H, Wong J, Bruen R, Campbell RJ, Sharma S, 122. Hayreh SS. Submacular Choroidal Vascular Bed Watershed
Gale J. Short-term intraocular pressure changes after intra- Zones and Their Clinical Importance. Am J Ophthalmol.
vitreal injection of bevacizumab. Can J Ophthalmol. 2010;150(6):9401; author reply 9412.
2007;42(6):80711. 123. Age-Related Eye Disease Study Group. Risk factors asso-
111. Semoun O, Blumen-Ohana E, de Preobrajensky N, ciated with age-related macular degeneration: a case-con-
Nordmann JP. Glaucome aigu par fermeture de langle trol study in the age related eye disease study: Age-Related
compliquant une injection intravitreenne de bevacizumab Eye Disease Study Report Number 3. Ophthalmology.
(Acute angle-closure glaucoma complicating an intravit- 2000;107(12):222432.
real injection of bevacizumab). J Fr Ophtalmol. 124. Hyman L, Schachat AP, He Q, etal. Hypertension, cardio-
2009;32(1):58. vascular disease, and age-related macular degeneration.
112. Jalil A, Fenerty C, Charles S. Intravitreal bevacizumab Arch Ophthalmol. 2000;118(3):3518.
(Avastin) causing acute glaucoma: an unreported complica- 125. Klein R, Klein BE, Tomany SC, Cruickshanks KJ. The
tion. Eye (Lond). 2007;21(12):1541. association of cardiovascular disease with the long-term
113. Bakri SJ, McCannel CA, Edwards AO, Moshfeghi DM. incidence of age-related maculopathy. Ophthalmology.
Persisent ocular hypertension following intravitreal ranibi- 2003;110(6):127380.
zumab. Graefes Arch Clin Exp Ophthalmol. 126. Tomany SC, Wang JJ, van Leeuwen R, etal. Risk factors for
2008;246(7):9558. incident age-related macular degeneration: pooled findings
114. Kahook MY, Kimura AE, Wong LJ, Ammar DA, Maycotte from 3 continents. Ophthalmology. 2004;111(7):12807.
MA, Mandava N. Sustained elevation in intraocular pres- 127. Hogg RE, Woodside JV, Gilchrist SE, etal. Cardiovascular
sure associated with intravitreal bevacizumab injections. disease and hypertension are strong risk factors for choroidal
Ophthalmic Surg Lasers Imaging. 2009;40(3):2935. neovascularization. Ophthalmology. 2008;115(6):
115. Adelman RA, Zheng Q, Mayer HR. Persistent ocular 104652.
hypertension following intravitreal bevacizumab and 128. Nitsch D, Evans J, Roderick PJ, Smeeth L, Fletcher AE.
ranibizumab injections. J Ocul Pharmacol Ther. Associations between chronic kidney disease and age-
2010;26(1):10510. related macular degeneration. Ophthalmic Epidemiol.
116. Good TJ, Kimura AE, Mandava N, Kahook MY. Sustained 2009;16(3):1816.
elevation of intraocular pressure after intravitreal injections 129. Wu L, Martinez-Castellanos MA, Quiroz-Mercado H,
of anti-VEGF agents. Br J Ophthalmol. 2010 Aug 11. Arevalo JF, Berrocal MH, Farah ME, etal. Twelve-month
[Epub ahead of print] safety of intravitreal injections of bevacizumab (Avastin):
117. Tao Y, Jonas JB. Intravitreal Triamcinolone. Ophthalmo results of the Pan-American Collaborative Retina Study
logica. 2010;225(1):120. Group (PACORES). Graefes Arch Clin Exp Ophthalmol.
118. Hayreh SS, Zimmerman MB. Non-arteritic anterior isch- 2008;246(1):817.
emic optic neuropathy: Role of systemic corticosteroid
therapy. Graefes Arch Clin Exp Ophthalmol. 2008;246:
102946.
119. Hayreh SS. Submacular choroidal vascular pattern-experi- For a complete bibliography and detailed review of previously
mental fluorescein fundus angiographic studies. Graefes published studies in the literature on the subject, please refer to
Arch Clin Exp Ophthalmol. 1974;192:18196. bibliography in previous publications listed here.
Clinical Features of Non-arteritic Anterior
Ischemic Optic Neuropathy 16
Non-arteritic anterior ischemic optic neuropathy population in the study by Hattenhauer etal. [2]. was
(NA-AION) is the most common type of ischemic optic primarily white, mostly of Scandinavian origin, living
neuropathy. It usually has classical clinical features in the Midwest region of the United States. With this
which make it easy to diagnose. tremendous variability from region to region and
among various ethnic groups in the world, it is diffi-
cult to get a definite, reliable incidence of NA-AION
applicable all over the world,.
Demographic Characteristics
2. Study Design: The result of a study depends very
of NA-AION much upon its design. This is evident from the wide
difference between the study by Johnson and Arnold
Incidence of NA-AION [1] and that conducted in Olmsted County,
Minnesota [2], i.e., 2.30 versus 10.2/100,000 indi-
viduals respectively.
A population-based study in the state of Missouri and
Los Angeles County, California, USA, found that Thus, no one definite incidence of the occurrence of
among persons 50 or older the estimated mean annual this disease can be given.
incidence rate of NA-AION was 2.30/100,000 popula-
tions [1]. Another epidemiological study in the popu-
lation of Olmsted County, Minnesota, USA, in persons Age
50 and older, reported the incidence of NA-AION as
10.2/100,000 individuals [2]. Such a large discrepancy
There is almost a universal impression that NA-AION
between the two studies indicates the problems of
is a disease of the middle-aged and elderly population
obtaining reliable information as well as the varying
and does not occur in young persons. In two studies
ethnic nature of the population. The incidence of
reported from the University of Iowa, one of 406
NA-AION depends not only upon the prevalence of
NA-AION patients [3] and another of 624 patients [4],
various universal risk factors, but also other factors
meanSD age was 60.113.64 and 61.012.3 respec-
causing NA-AION in the region where the study is
tively and the age range 1191 and 18100 years
conducted. For example:
respectively; in both studies, 11% of the patients were
1. Ethnicity of the Population: This obviously varies young (<45 years), 49% were middle aged (45
widely, not only from one region of the world to 64years) and 40% were elderly (65years). In another
another but also among different ethnic groups living multicenter study [5] of 420 patients, it was 66.0 8.7
in the same region. In the study by Johnson and Arnold (SD) years. In still another study [6] from Atlanta
[1], conducted in Los Angeles California and Missouri (USA), of 727 consecutive patients with NA-AION,
(USA), NA-AION was significantly more prevalent 23% were younger than 50 years (median 43 years;
among white individuals than black or Hispanics liv- range, 1349 years). Thus, the prevalent impression
ing in the United States. Incidence varied even within that NA-AION is a disease only of the elderly is not
the same ethnic group, because of other factors. The correct no age is immune from NA-AION; this

338 16 Clinical Features of Non-arteritic Anterior Ischemic Optic Neuropathy
misconception has resulted in young persons with discovery of visual loss was upon awakening from
NA-AION sometimes being misdiagnosed. sleep in the morning or a nap in 52%, and soon after
that when the patient had the first opportunity to use
vision critically in 21.5%. In the remaining 27% epi-
sodes of NA-AION, although the patients became
Gender aware of the visual loss later on during the day, they
could not say definitely when it occurred, or rule out
NA-AION is seen somewhat more often in men than in the possibility that it had been there since awakening.
women it has been reported as 62% versus 38% [5], Becoming aware of visual loss is extremely variable,
59% versus 41% [4] and 58% versus 42% [6]. This depending upon the location, severity and type of visual
would be in keeping with greater prevalence of cardio- field defect, the occupation of the patient, and, above
vascular diseases among men than women. all, the perceptiveness of the patient in detecting a
visual problem. For example, I saw a patient with supe-
rior visual field loss who only noticed visual loss while
hunting, when he tried to shoot a flying duck, but was
Racial Features sure that his vision had been normal the previous
day; another patient, a physician, was unaware of a
fairly marked visual loss in the inferior nasal sector of
It is far more common among the white population
one eye, till it was discovered during clinic examina-
than in blacks, Hispanics and other racial groups [1,
tion. Landau et al. [8], based on a study of only 24
36]. This may explain the above difference in inci-
patients claimed that the patient with AION typically
dence of NA-AION in the studies by Johnson and
notices blurred vision in the affected eye within 1 or 2h
Arnold [1], conducted in Los Angeles California and
of arising in the morning. In the multicenter Ischemic
Missouri (USA) (containing white, black and Hispanic
Optic Neuropathy Decompression Trial Study
population), and by Hattenhauer et al. [2] in the
(IONDT ) [5], it was reported that 42% (174 of 418) of
Midwest region of the United States (containing
patients recalled that the onset of visual loss occurred
primarily white, mostly of Scandinavian origin).
within 2 h of awakening. To place the difference in
incidences between this and our study [7] in proper
perspective, one has to consider the following factors:
Symptoms
Difference in the NA-AION Population
In the vast majority of cases, these are typical.
NA-AION patients usually complain of loss of vision Information in the IONDT study [5] was based on a
in the lower field of vision more often in the inferior selected population of patients with NA-AION found
nasal field than altitudinal loss (Figs.16.1ae). There to be eligible for the study, and not on consecutive,
is a sudden and painless deterioration of vision. I [7] unselected NA-AION patients, as was the case in our
investigated the time of day and seasonal variation of study [7].
onset of visual loss in NA-AION.
Differences in Examiners and how Specifically

the Patients were Questioned to Obtain this
Time of Onset of Visual Loss in NA-AION Information
This is an important piece of information which helps To get the correct information, the right person has to
in the diagnosis of NA-AION as well as in the under- ask the right question in the right manner. In our study
standing its pathogenesis. In our study [7], in a total of [7], all patients were personally seen by me and ques-
544 episodes of NA-AION, the time of day for tioned specifically and in detail about the time of the
Symptoms 339
a b
c d
Fig.16.1 Visual fields (plotted with a Goldmann perimeter) of defect (Figs. c and e). (Fig. a reproduced from Hayreh and
right eyes in Figs.a, b, d and e, and of left eye in Fig.c. They show Zimmerman [17]; Figs.be Reproduced from Hayreh [56].)
inferior nasal defect (Figs. a, b and d) and inferior altitudinal
day they became aware of visual loss, without any sug- Progressive Visual Loss
gestion of a right answer. In the IONDT Study [5], the
data were collected from 25 different centers and many
Progressive visual deterioration in some eyes with
more persons were involved (including technicians).
NA-AION is well-known. Patients usually notice it on
There must therefore have been variability in the nature
waking in the morning. The following patient of mine
and manner of questioning about the time of onset. The
recorded herself this progression very well on her com-
information being volunteered by a patient spontane-
puter. A 64-year-old diabetic woman, she became aware
ously compared to that in response to specific questions
of a spot in the lower nasal part of her right eye visual
can make all the difference the latter obviously gives
field accidentally (Fig.16.2a arrow) when she covered
far more reliable information than the former.
the left eye. She saw an ophthalmologist that day, who
The time of first discovery of visual loss in NA-AION
found visual acuity 20/20 and optic disc edema in that
is a very critical piece of information because it gives a
eye. Fig. 16.2be show the progression over the next
clue to the time of development of NA-AION and also,
8days. When first seen in our clinic on the 9thday, her
indirectly, about its pathogenesis. Our study [7] indi-
visual acuity was 20/250 and kinetic perimetry showed
cates that NA-AION most often develops during sleep;
an absolute central scotoma and inferior nasal visual
there must be something happening during sleep which
field loss (Fig.16.3) and optic disc edema (Fig.16.4).
precipitates the development of NA-AION. I have dis-
cussed in Chap. 14 the role of nocturnal arterial hypoten-
sion in the pathogenesis of NA-AION [3, 9, 10].
Some patients definitely develop NA-AION during Intermittent Blurring of Vision
the daytime, not associated with a nap. In such cases, it
is suggestive of one or both of the following two mech-
Some patients may complain of intermittent blurring
anisms for the development of NA-AION that may be
of vision when the visual field defect passes through
at work:
the fixation point because of unconscious shifting of
(a) Embolism into the posterior ciliary artery (PCA) cir- fixation between the seeing and the blind areas near the
culation (the main source of blood supply to the fixation [16, 17] (Figs.16.1ae, and 16.5).
optic nerve head [10, 11]) can produce NA-AION at
any time of the day. The eyes with the usual type of
NA-AION generally have an absent or small optic Photophobia
disc cup [1214], but those of an embolic nature
generally tend to have the normal cup size. On fluo-
rescein fundus angiography, in these cases, complete Later on, photophobia is a common complaint, partic-
occlusion of a PCA (with negative temporal artery ularly in bilateral cases.
biopsy) has been demonstrated [14a] (see Figs.3.27,
3.33 and 3.34 in Chap. 3) like that seen in arteritic
AION, but there is no evidence of giant cell arteritis
Simultaneous Onset of Visual Loss
in these patients and a negative temporal artery
biopsy. Thus, fluorescein fundus angiography is in Both Eyes
extremely useful in evaluating whether NA-AION is
embolic in nature or not. The source of the embolus Some patients may complain of simultaneous onset
may be the carotid arteries, aorta or the heart. of visual loss in both eyes; however, in my study of
(b) Our experimental study [15] in atherosclerotic dealing with about 1,350 patients with NA-AION,
monkeys has shown that in atherosclerosis sero- the perceived simultaneous visual loss in both
tonin can produce vasospasm of the PCA, result- eyes usually occurred because the patient was
ing in its occlusion and development of NA-AION. unaware of the prior visual loss in the first eye and
We speculated that the vasospasm is produced by discovered it only when the second eye became
the release of serotonin by platelet aggregation on involved. I found simultaneous bilateral onset of
atherosclerotic plaques. NA-AION extremely rare.
Symptoms 341
Fig.16.2 This is an
excellent demonstration of
progressive worsening of
vision in a 64-year-old
diabetic woman with
NA-AION. (a) Shows a small
scotoma (arrow) in the lower
nasal part of her right eye
visual field on the day she
first discovered that on
covering her left eye. (be)
Show the progressive
increase in the blur over the
next 8days
b
d
Symptoms 343
Fig.16.3 In the above patient in Fig.16.2, visual field (plotted Fig. 16.4 This fundus photograph, in the above patient in
with a Goldmann perimeter) show inferior nasal visual defect Fig.16.2 shows optic disc edema and peripapillary hemorrhages
increased risk of developing subsequent cerebrovas-

cular disease, myocardial infarction or mortality, and
smoking had no association with NA-AION. This is
evidence that it is a serious mistake to equate
NA-AION with stroke and other systemic vascular
accidents (see Chap. 14).
Visual Acuity
This has been reported differently in different studies

[5, 6, 2224]. For example, in a combined cohort of
500 consecutive NA-AION eyes in our two studies
Fig.16.5 Visual field (plotted with a Goldmann perimeter) of a
[22, 24], when patients were seen within 2weeks after
right eye with NA-AION shows absolute superior nasal visual the onset of visual loss, initial visual acuity was 20/20
defect 20/25 in 33%, better than 20/40 in 51%, and 20/200
or worse in 21%. In the IONDT study [5], initial
visual acuities in the study eyes ranged from 20/20 or
Seasonal Variations of Onset of Visual better, to light perception, with 49% of the patients
Loss in NA-AION seeing better than 20/64 and 34% of the patients see-
ing 20/200 or worse. However, the baseline findings
in that study were derived from a selected population
We [7] investigated seasonal variation of the onset of of patients with NA-AION who were eligible for the
NA-AION for 839 episodes. Their data showed that IONDT and not an unselected group. Preechawat
NA-AION onset was significantly more frequent in the et al. [6]. in a study of 727 NA-AION consecutive
Iowa (USA) Summer (or the hot months, with tem- patients, found that visual acuity was 20/40 or better
perature in 60s to low 90sF) than in the Winter (or the in 59%, and 20/200 or worse in 20%. Thus, about half
cold months, when the maximum temperature is usu- of the NA-AION eyes initially present with a visual
ally much below freezing) (p=.0030), with Spring/ acuity of 20/40 or better. This fact is not fully appreci-
Fall or the mild months intermediate. This is in con- ated in the ophthalmic community and is sometimes
trast to the reported incidence of systemic vascular responsible for missing the diagnosis of NA-AION,
accidents, such as myocardial infarction and cerebro- because of the prevalent belief that every eye with
vascular accidents, which is higher during winter than ischemia of the optic nerve head (as in NA-AION)
other seasons [1820]. must have worse than normal visual acuity. Therefore,
In order to explain the seasonal variations of myo- the presence of normal visual acuity does not rule
cardial infarction, cerebrovascular accidents and pos- out NA-AION.
sibly other vascular accidents, a number of studies
have been conducted on the effects of temperature
variations on hematologic and other vascular changes
in the body which could be contributory to develop-
Natural History of Visual Acuity
ment of vascular accidents (see Chap. 14).
Thus, there is a distinct difference in the seasonal in NA-AION
variation in onset of NA-AION and systemic vascular
accidents, which indicates a difference in their patho- The most important piece of information required in
geneses. Factors other than hematologic abnormali- NA-AION, from the points of view of both patient
ties play an important role in the development of and ophthalmologist, is the natural history of visual
NA-AION [7, 21]. Moreover, in our studies [3, 4], outcome. This is also important in the management
there was no evidence that NA-AION patients are at of NA-AION because the outcome of any treatment
Visual Fields Defects 345
has to be judged in the light of its natural history, so that of the control group was 36.8years; that marked
that natural recovery is not interpreted as a beneficial age difference accounts for the finding of more hyper-
effect of treatment. There are two prospective stud- metropia in the NA-AION group than in the control
ies that have evaluated this, one based on 125 eyes group. This is further supported by the fact that in our
[23] and the other on 386 eyes [24]; both arrived at study [14] there was no difference in refraction
the same conclusion. Both studies showed that between NA-AION and arteritic AION eyes, because
4143% of patients seen within 2weeks of onset of those groups were almost similar in age. The other
visual loss with initial visual acuity of 20/70 or two studies [25, 26] had age-matched control groups
worse, experienced improvement, while 1519% but their sample sizes were small; it is possible that
grew worse at 6 months. This completely contra- their findings might have been different with a much
dicts the highly prevalent impression that no spon- larger sample size.
taneous visual acuity improvement occurs in
NA-AION.
Visual Fields Defects
Refractive Error In contrast to visual acuity, which can be within nor-

mal limits in almost half of the eyes with NA-AION,
Beck etal. [13], in a study of 126 NA-AION and 122 visual field defects are a universal occurrence.
control eyes, found more than +1.0 diopter hyper- Therefore, perimetry is the most important and essen-
metropia in 25% of the NA-AION group compared to tial visual function test to evaluate the visual loss
only 6.5% of the control group. Katz and Spencer and the visual disability caused by NA-AION. I would
[25], on comparison of 50 patients with NA-AION to like to discuss the subject of visual field loss at length
an age-matched control group, concluded that their here. Information in the literature on the visual field
study suggested that patients with NA-AION are more defects in NA-AION is mainly based on manual
hypermetropic than the control group. Philips et al. kinetic perimetry with a Goldmann perimeter [16,
[26]. on comparison of 37 eyes with NA-AION and 74 2838] and, more recently, using automated static
eyes of a sex and age matched control group found threshold perimetry [3942].
mild hypermetropia more frequent in the NA-AION These eyes can present with a variety of optic
group than in the control group. However, Mansour nerve related visual field defects. Figures 16.1ae,
etal. [27]. in a study of 9 NA-AION eyes and 26 con- 16.3, 16.5 and 16.6 show some examples of the
trol eyes found no difference in refractive error visual field defects seen in our study [17]. Very
between the two. In our study [14] of 502 consecutive rarely I have seen an eye with nasal hemianopic loss
NA-AION patients 50 years and older, there was in NA-AION. Findings of the studies in the litera-
minus <6.0 diopters in 1%, minus 6.0 to minus 3.1 ture from 1967 to 1991 are summarized in Table16.1.
diopters in 2%, minus 3 to minus 0.1 diopters in 16%, In these studies the overall pattern of visual field
emmetropia in 24%, plus 0.1 to 3 diopters in 46%, defects in NA-AION reveals that: (a) in almost all
plus 3.1 to 5 diopters in 7%, and plus>5 diopters in the studies the visual field information was given
5%. Comparison of refraction in NA-AION patients simply as part of the general clinical features of
50years and older with an age-matched general popu- NA-AION rather than based on specific, planned
lation showed no significant difference between the and systematic investigation of visual field patterns
two groups. This latter study [14] contradicts the find- seen in NA-AION at initial presentation, (b) no men-
ings of three previous studies based on 126 [13], 50 tion was made whether those were relative or abso-
[25] and 37 [26] eyes respectively, which suggested lute defects (see below), and (c) most studies were
that patients with NA-AION are more hypermetropic based on a comparatively small number of eyes. In
than the control group. Of the three previous studies, the literature, an inferior altitudinal visual field
the study by Beck etal. [13]. had the largest number of defect is almost invariably described as the classical
eyes, i.e. 126 eyes with NA-AION. In that study, the field defect in NA-AION. Among the studies in
mean age of NA-AION group was 62.1 years while Table 16.1, where visual fields were plotted with
a b
c d
e f
Fig. 16.6 Visual fields (plotted with a Goldmann perimeter) show some examples of visual field defects seen in NA-AION.
(a) Reproduced from Hayreh [56], (b, c, h) reproduced from Hayreh and Zahoruk [72]
g h
i j
k l
m n
manual kinetic perimetry, the reported prevalence Manual Kinetic Perimetry Versus Static
rate of inferior altitudinal visual field defect varies Threshold Perimetry
from 25% to 79%.
We [17] investigated visual field abnormalities, their
pattern and prevalence at the initial presentation in Currently, visual fields are usually plotted using auto-
NA-AION in a systematic, planned study of 312 con- mated perimetry. Visual field information provided by
secutive NA-AION eyes, by manual kinetic perimetry manual kinetic perimetry performed with a Goldmann
(i.e., visual fields plotted with a Goldmann perimeter), perimeter may be very different from that by auto-
using I-2e, I-4e and V-4e targets. Our study [17] showed mated static threshold perimetry [Humphrey 302 or
that I-2e target was seen by 75%, I-4e by 91% and V-4e 242 SITA (Swedish Interactive Testing Algorithm)].
by 100%. The overall prevalence of general visual field Both types have their advantages and disadvantages,
defects in that study was 83% with I-2e, 79% with I-4e and one should be aware of those when interpreting
and 69% with V-4e, while those of scotoma(s) within the findings. Automated perimetry does not provide
the central 30 was 55%, 49% and 36% respectively. any information beyond the central 2430. Kinetic
Central scotoma was seen in 49% with I-2e, 44% with perimetry, by contrast, provides information all the
I-4e, and 29% with V-4e. Relative inferior altitudinal way to about 8090 temporally, 70 inferiorly, 6070
defect (Figs.16.1a, b, d and 16.6d, h) was most com- nasally and 5060 superiorly. This has important
mon (35% with I-2e; and 22% with I-4e), but absolute implications in characterizing the visual field defects
inferior altitudinal defect (Figs.16.1c and e) was seen in NA-AION and evaluating visual disability caused
in only 8%. By contrast, relative inferior nasal defect by NA-AION.
(Figs.16.1d, 16.3 and 16.6g, h, j, k) and absolute infe-
rior nasal visual loss (Figs.16.1a, b and 16.6k) was the
most common defect detected in NA-AION (22%), but Characterization of Visual Field Defect Based
occurred in only 3% with I-2e and 11% with I-4e. on the Two Types of Perimetry in NA-AION
Overall, loss of the nasal part of the visual field was the
most common occurrence. Thus, a combination of a Information provided by automated perimetry may
relative inferior altitudinal defect with absolute infe- result in different characterization of the pattern of
rior nasal defect is the most common pattern in the visual field defects compared to that by kinetic
NA-AION (Figs. 16.1a and b). This contradicts the perimetry. For example, in automated perimetry a
commonly held belief that inferior altitudinal visual large scotoma or defect extending all the way
field defect is typical of NA-AION. to 2430 is almost invariably interpreted as an
Table16.1 Literature on Visual Field Defects on Kinetic Perimetry with a Goldmann Perimeter in NA-AION [17])
Author(s) Year Number Inferior Superior Segmental Arcuate Hemianopic Peripheral Central Paracentral Miscellaneous
of Eyes altitudi- altitudi- defect defect defect island scotoma scotoma
or nal defect nal defect field only
patients
Visual Fields Defects
Cullen [28] 1967 12 eyes 3 1 inf nasal 1 Temporal 5 2 inferior

Ellenberger 1968 3 patients 1 1 1 sup nasal
and Netsky
[29]
Ellenberger 1973 64 eyes 34 inferior, 8 Temporal
etal. [30] 17 superior arcuate
4,,enlarged
blind spot 1
Eagling etal. 1974 24 eyes 14 2 5 5
[31]
Boghen and 1975 43 eyes 12 4 12 inf 3 inferior 3 temporal, 6 Generalized
Glaser [32] nasal, 1 sup nasal constriction 1
1 superior
Aulhorn and 1978 61 eyes 48 4 9 both sup and
Tanzil [33] inf defect
Hayreh and 1979 123 eyes 38 10 39 inf nasal, 2 superior 7 nasal, 7 53* 7*
Podhajsky 9 sup nasal, 1 inferior 3 temporal
[16] 5 inf
temporal,
3 sup
temporal
Bronner 1979 50 patients 32 4 2 temporal 11
etal. [34]
Repka etal. 1983 172 eyes 79 29 14 inferior 14 inferior 34 Others in 2
[36]
Rizzo and 1991 58 patients 27 7 2 inf nasal 1 15 Others in 6
Lessell [37]
Gerling etal. 1998 36 patients 19 3 6 Others 8
[38]
*These field defects were in addition to other defects
349
altitudinal field defect (Fig. 16.7a), when kinetic field (Fig. 16.5), 72% in the inferior field, 50% in
perimetry may show normal peripheral field beyond nasal and 12% in the temporal visual field most of
that (Fig.16.7b); this has important implications: (1) the eyes had more than one location of visual field
This may be one of the reasons that an inferior altitu- defect. While the relative inferior altitudinal defect
dinal visual field defect is almost invariably described (Figs.16.1a, b, d, 16.6d, 16.6h, and 16.7b) was most
as the classical field defect in NA-AION, but it is not common (35% with I-2e; and 22% with I-4e), that
the case in most of the eyes with NA-AION (Fig.16.7), was not true of the absolute inferior altitudinal defect
and (2) Full information about the peripheral visual (Figs.16.1a and e in only 8% with V-4e). In sharp
fields (which the automated does not provide) is criti- contrast to that, absolute inferior nasal sector visual
cal to evaluate visual disability (see below), since the loss (Figs.16.1a and b) was the most common defect
peripheral field plays a critical role in our navigating seen in NA-AION (22%), but it was seen in only 3%
and day-to-day living. In addition to that, I have also with I-2e and 11% with I-4e. Thus, this study showed
found that most patients prefer manual kinetic perim- that the most frequently seen combination is inferior
etry over automated perimetry, and patients who are and nasal absolute visual field defect in NA-AION,
too old or too young to participate in automated and much less frequently in the superior and tempo-
perimetry do well on manual kinetic perimetry. It is ral regions.
most unfortunate that highly useful manual kinetic
perimetry is rapidly being replaced by automated
perimetry. Extent and Severity of Visual Field Defect
This has functional significance, because in the region

Relative Versus Absolute Visual Field Defects of the relative field defect, there is still a variable
amount of intact visual function more in the defect
A review of the literature [17] shows that the descrip- seen with I-2e isopter than that with I-4e. For example,
tions of the various types of visual field defects in an absolute inferior altitudinal visual field defect
NA-AION almost never specify whether those were (Figs.16.1a and e) is far more visually disabling than
relative or absolute. But a distinction between relative only a relative inferior altitudinal defect, and the latter
and absolute visual field defects is essential, for the less disabling when it is only detected with I-2e rather
following important reasons: than that with I-4e (Figs.16.1a, b, d, 16.6d, e, g, h, j
and 16.7b).
Difference in Pattern of Visual Field Defect Evaluation of Extent of Visual Functional

Disability in NA-AION
In visual fields plotted with a Goldmann perimeter,
defects seen with smaller isopters are more frequent NA-AION is a common, visually disabling disease. To
and more extensive than those with larger isopters assess that disability, one needs information on both
(Figs. 16.1ae, 16.3, 16.5, 16.6 and 16.7b). Of the peripheral and central visual field defects. The type of
312 eyes in our study [17], the I-2e target was seen functional disability produced by the two types can be
by 75%, I-4e by 91% and V-4e by 100%. The preva- very different.
lence of pattern of various types of visual field defects
therefore depends upon whether it is relative (i.e. with
only I-2e and/or I-4e) or absolute (i.e. with V-4e),
and, even among the relative field defects, whether Extent of Visual Functional Disability Produced
it is only detected with I-2e or both I-2e and I-4e by Peripheral Visual Field Loss
isopters (Figs. 16.1ae, 16.3, 16.5, 16.6 and 16.7b).
In our study, the prevalence rate of any visual field It is well established that the constant tracking pro-
defects with V-4e isopter was 24% in the superior vided by the peripheral visual fields is essential for
Fig. 16.7 This shows the difference in pattern of visual field Inferior paracentral scotoma seen in (b) plotted with Goldmann
defect revealed by visual fields plotted with an automated perim- perimeter was interpreted in automated perimetry field in (a) as
eter (a) and Goldmann perimeter (b) in an eye with NA-AION. inferior altitudinal defect
b visual field that is used all the time for daily work. A
similar defect in the superior field is much less
disabling.
Clinical Significance of Central Versus

Peripheral Visual Field Loss
My long experience of dealing with NA-AION

patients has shown that loss of peripheral visual field,
particularly in the lower part, can often be far more
disabling than that produced by the central visual
field defect. For example, a person with central sco-
toma but with normal peripheral visual field
(Figs.16.6 f, l-n and 16.7b) can lead a fairly normal
life, although unable to read, write, drive or see prop-
erly in the center. By contrast, patients with bilateral
absolute inferior altitudinal defects or with markedly
Fig.16.7 (continued) constricted visual fields, but with perfectly normal
visual acuity (Figs. 16.1c and e) are markedly dis-
abled, stumbling, and unable to drive or navigate.
sensory input to our day to day activity. For example, Obviously, this has important functional and practi-
the peripheral visual field is vital for walking, driving cal implications. Disability Boards tend to deny eli-
and navigating in this world. In view of that, to gibility to persons with NA-AION who can see 20/20
assess the visual function disability produced by in both eyes, in spite of the fact that they have com-
NA-AION, it is important to have complete informa- plete loss of the lower half of visual fields in both
tion about the peripheral visual fields (particularly eyes (Figs. 16.1c and e); from a practical point of
the inferior field) and any impairment in them. As view they are legally blind. A scotoma in the lower
mentioned above, automated perimetry provides central field (Fig.16.6d, f, h, m) can be bothersome
information only up to about 2430 in the periphery. compared to that in the upper central field (Fig.16.6e),
Kinetic perimetry, by contrast, provides peripheral because we use the lower field all the time for daily
visual field information all the way to about 8090 work.
temporally, 70 inferiorly, 6070 nasally and 5060
superiorly. Thus, in NA-AION the visual field plot-
ted with manual kinetic perimetry provides far supe- Natural History of Visual Field Defects
rior information about the peripheral visual field
defects for the purpose of evaluating visual func-
in NA-AION
tional disability.
This was evaluated in a study of 332 consecutive
NA-AION eyes that had 8weeks or more of follow-
up from the initial visit [24]. Visual fields with
Extent of Visual Functional Disability Produced kinetic perimetry showed that 26% of those who
by Central Visual Field Defect were first seen2weeks after onset, with moderate
to severe visual field defect, had improvement at
It is well known that a defect in the central visual field 6 months and 27% at 1 year after the initial visit.
usually results in deterioration of visual acuity but not Two years after the initial visit, 27% of eyes with
always if it spares the fixation spot. However, it is still initial minimal to mild field defects showed worsen-
disabling to some extent because it is the inferior ing, as did 19% of those with moderate to severe
defects. Visual acuity and visual fields showed spot (Figs.16.1ae, 16.5 and 16.6g, j) or a small cen-
improvement or further deterioration mainly up to tral scotoma (Figs.16.6f and 16.7b). In such cases,
6 months, with no significant change after that. In by eccentric fixation, the patient may test much bet-
this study, during the follow-up period, the central ter over time without any actual improvement in the
visual field remained stable in 68%, improved in retinal or optic nerve function. In my studies on vari-
16%, and worsened in 16%. There was improvement ous types of ocular vascular disorders, I have found
in peripheral visual field in 17% of the eyes and that, for genuine visual acuity improvement, there
worsening in 18%. When NA-AION develops in the must be simultaneous improvement in both the cen-
second eye, there is no correlation in the visual out- tral scotoma and the visual acuity [24, 43, 44]. In our
come between the two eyes. study [24] dealing with the natural history of visual
Visual acuity and visual fields provide very differ- outcome in NA-AION, improvement was due to
ent information about the visual status in NA-AION; eccentric fixation in 26% of the eyes with visual acu-
the changes in the two can be totally independent and ity improvement. This was further supported by the
unrelated to each other. Also, the amount of visual following two findings in that study:
acuity and visual field change seen in a patient during
follow-up depends upon the time when a patient is first (a) Visual acuity of 20/70 or worse improved in 42%,
seen after the onset of visual loss. In our study [24], but if one excludes those with visual improvement
patients who were first seen more than 2 weeks due to eccentric fixation, the genuine improvement
(3weeks to about 10weeks) after the onset of visual was in about 30%. This corresponds closely to the
loss and still had optic disc edema, showed both less percentage with improved visual fields (27%) dur-
improvement and less deterioration in visual acuity ing follow-up.
and visual fields than those first seen within 2weeks of (b) The incidence of worsening of visual acuity and
onset. Obviously, the longer the time interval between visual field was the same in that study because that
onset and first evaluation, the greater the likelihood of reflects the actual state of the patients visual sta-
visual changes having occurred already, and, hence, tus. That is why I strongly emphasize that appar-
the smaller the chance of change in visual status from ent improved visual acuity without corresponding
then on. That study showed that the visual change does improvement of central visual field defects on
not always mean either improvement or deterioration perimetry or on Amsler chart can be misleading
throughout the entire course of initial follow-up but [24, 43, 44].
there may be a changing pattern; for example in some
eyes there was improvement/deterioration at one time
and the opposite at another in the same eye.
In my experience of dealing with eyes with both Development of a Variable Degree of Amblyopia
NA-AION [24] and arteritic AION [43] as well as cen- in the Eye with Worse Vision
tral retinal artery occlusion [44], I have found the fol-
lowing two phenomena, from time to time. In our studies [24] on NA-AION, I have also found
that in some patients with bilateral NA-AION, when
the second eye developed NA-AION with marked
deterioration of visual acuity, the previously involved
The Recorded Improvement in Visual Acuity eye with comparatively better visual acuity showed
does not Always Reflect Genuine Improvement spontaneous improvement. In NA-AION and other
in the Optic Nerve Function ocular vascular disorders studied in my Ocular Vascular
Clinic since 1973, I [24] have found that sometimes
It may simply be due to the patient having learned by the first eye, with poor visual acuity, may develop a
experience to read the test chart better by looking variable degree of amblyopia, even in middle aged and
around and fixating eccentrically. This applies par- elderly patients, because the patient may not use that
ticularly to an eye which has a visual field defect eye for central vision when the fellow eye has normal
passing through or just involving the central fixation visual acuity (a phenomenon similar to occlusion
amblyopia in children); however, when the fellow eye Optic Disc Edema
develops marked visual loss from NA-AION, the
amblyopic eye with comparatively better visual acuity
It is always present at the onset of visual loss. There
soon shows a variable amount of spontaneous improve-
are several misconceptions about optic disc edema in
ment. This has erroneously been attributed to some
NA-AION. The following data from our study [46],
treatments [45].
based on 749 eyes with NA-AION, provides useful
information about the various aspects of optic disc
edema in NA-AION.
Anterior Segment of the Eye
Evolution of Optic Disc Edema in NA-AION
This invariably shows no abnormality except for the
presence of relative afferent pupillary defect in unilateral It is important to comprehend that optic disc edema in
NA-AION eyes. In some there may be evidence of neo- NA-AION evolves as the time passes from onset to res-
vascular glaucoma, angle closure glaucoma or raised olution. The ophthalmoscopic appearance of optic disc
intraocular pressure due to primary open angle glaucoma edema in NA-AION can be divided into two phases.
or pseudo-exfoliation of the lens. Raised intraocular
pressure is a definite risk factor for development of
NA-AION (see Chap. 14). Thus, a thorough slitlamp Ophthalmoscopic Appearance of Optic Disc
evaluation of the anterior segment and evaluation of Edema at the Onset of NA-AION
intraocular pressure are essential for NA-AION patients.
The lens and vitreous invariably do not show any Eyes with NA-AION usually have no optic disc cup
abnormality in these eyes, except that there may be (Fig. 16.8a). When these eyes are seen soon after the
normal aging changes in the lens in old patients or onset of NA-AION and visual loss, the optic disc always
pseudoexfoliation of the lens. is edematous and the edema does not differ in color
from edema due to other causes in some cases there
may even be hyperemia of the optic disc (Fig. 16.8b)
[10, 4749]. Initially edema may be more marked in one
Optic Disc Changes part of the disc than the other (Figs.16.8b and 16.9c)
[10, 4749]. Frequently there are splinter hemorrhages
at the disc margin (Figs.16.9 and 16.10). Rarely there
Optic Disc Characteristics in NA-AION may be some peripapillary retinal detachment.
There is an almost universal impression that optic
As discussed in Chap. 14, it is well established that disc edema in NA-AION is always pale edema from
eyes with NA-AION usually have either no cup or a the very start. This has resulted in the belief that unless
very small cup, compared to the general population. the optic disc edema is pale, the disease cannot be
The importance of that in the pathogenesis of NA-AION NA-AION. This misconception is not infrequently
is discussed at length in Chap. 14. responsible for missing the diagnosis of NA-AION.
The following detailed discussion is required to explain
the basis of this misconception.
1. One common explanation given for pale edema in

Optic Disc Change During the Initial Acute
NA-AION is that the color of the optic disc depends
Phase of NA-AION upon its vascularity, and since NA-AION is an isch-
emic condition, poor circulation makes the disc in
Optic disc edema is a classical initial sign of acute these eyes pale. But my study [50], correlating the
NA-AION. color of the optic disc and vascularity on fluorescein
Optic Disc Changes 355
a angiography in 55 normal optic discs, showed that:

in a normal optic disc, redness of the optic disc is
not a true guide to its vascularity. That study found
that there was uniform redness of the disc all over in
40% and more redness in the nasal than the tempo-
ral part in the remaining 60%; however, on fluores-
cein angiography, optic disc vascularity was uniform
all over in 60% and more marked on the temporal
side in 40%. This was also very well demonstrated
by the fact that eyes with retinitis pigmentosa,
whose discs on ophthalmoscopy have the classical
waxy pale appearance, have quite normal disc vas-
cularity in the prelaminar and deeper parts of the
optic nerve head on angiography (Fig.16.11) [50].
In another study [51], we also investigated the
source of redness of a normal optic disc, and con-
cluded: the normal redness of the disc is essentially
b due to the filling of the capillaries in the surface
layer of the disc and not due to its deeper vessels.
The surface layer of the optic nerve head is supplied
by the retinal circulation (Fig.16.12) [50, 52]. Since
NA-AION is due to vascular insufficiency in the
deeper parts of the optic nerve head, supplied by the
PCA circulation [10, 4749], (and NOT the retinal
circulation), it cannot cause the optic disc edema to
be pale at onset.
2. The idea that the optic disc edema is always pale in
NA-AION may have been caused by the lumping
together of the findings of optic disc edema in both
arteritic and non-arteritic AION. However, a
chalky white optic disc edema is seen only in eyes
with arteritic AION (see Chap. 12) [47, 48, 53]
c and not NA-AION, unless the latter is caused by
embolic occlusion of the PCA [14a], which is a
rare occurrence. This difference in the ophthalmo-
scopic appearance of optic disc edema in the two
types of AION is caused by the following basic
difference in their pathogenesis, among other
things [10, 11, 48, 49].
Fig.16.8 Fundus photographs of left eye of a 53-year-old man.

(a) Shows normal optic disc before developing NA-AION. (b)
Shows optic disc edema during the acute phase of NA-AION
note edema is more marked in the inferior temporal part of the
disc. (c) Shows disc after resolution of edema and development
of pallor more marked in temporal part than nasal part
b
Fig.16.10 Fundus photographs showing optic disc edema with
a splinter hemorrhage at the temporal margin of the disc
Fig. 16.11 Fluorescein fundus angiogram of right eye of a

patient with retinitis pigmentosa shows filling of the peripapil-
lary choroid and the prelaminar region of the optic nerve head
but no filling of the retinal vessels at this stage (Reproduced
from Hayreh [50].)
Fig.16.9 Fundus photographs showing three types of optic disc

edema in NA-AION. Note that in (c) the edema is more marked
in the inferior nasal region than elsewhere
CZ
C
Retinal vein
Retinal artery R S
PCA
OD
PR
LC
PCA
Fig.16.12 Schematic representation of blood supply of the optic nerve head (Modified from Hayreh [124]). C choroid, CZ circle
of Zinn and Haller, LC lamina cribrosa, OD optic disc, PCA posterior ciliary artery, PR prelaminar region, R retina, S sclera
(a) Arteritic AION is caused by thrombosis of the Evolutionary Pattern of Optic Disc
PCA by giant cell arteritis, resulting in complete Changes in NA-AION
occlusion and severe ischemic damage and
infarction of the optic nerve head [10, 11, 48,
49, 53]. By contrast, as discussed at length in Our study [46] showed that if the patient is seen about
Chap. 14, NA-AION is NOT due to complete 23 weeks or longer after the onset, the optic disc
occlusion of the PCA but to transient hypoxia/ gradually develops progressive pallor and the edema
ischemia of the optic nerve head [10, 49], starts to resolve. Once optic disc edema has resolved,
caused by a transient fall of perfusion pressure the disc develops sectoral (Figs. 16.8c, 16.13 and
in the optic nerve head vascular bed, usually 16.14a) or generalized (Fig.16.14b) pallor. If an eye
during sleep [7]. has a small disc cup, rarely I have seen that to increase
(b) Moreover, after the optic disc edema has in size due to loss of nerve fibers.
resolved, in arteritic AION the disc mostly
develops cupping (because of severe acute
ischemic damage see Chap. 12), similar to Optic Disc Edema Undergoes the Following
that seen in glaucomatous optic neuropathy, Evolutionary Changes, till it Resolves Completely
but not in NA-AION [10, 12, 47, 48, 54, 55].
Optic disc edema in NA-AION is due to axoplasmic
flow stasis in the ischemic axons in the optic nerve
head. The speed of axonal destruction depends upon
Pattern of Optic Disc Edema in Diabetics the severity of ischemia the more marked the isch-
emic damage, the earlier and more marked the destruc-
This is discussed below in NA-AION in diabetics. tion of the axons [56]. It has been shown that once an
axon is destroyed, there is nothing left to swell and

consequently no optic disc edema [57, 58]. As the
axons are destroyed, the edema regresses correspond-
ingly and pallor develops.
Our study [46] showed that optic disc changes in
NA-AION go through a characteristic evolutionary
process during the first 23months, and the findings
depend upon the stage at which a particular eye is
seen. The evolution usually follows the following
pattern: initially the involved part of the disc (i.e.,
corresponding to the location of visual field loss) has
edema, with the rest of the disc normal or showing
much less edemaafter several days the entire disc
may show generalized edemastill later the optic
disc in the originally involved part begins to develop
pallor and the edema gradually starts to regress in
that part, so that the uninvolved part (corresponding
Fig. 16.13 Fundus photograph of a diabetic patient with to normal visual field) may have more edema than
NA-AION, 3months after onset, showing segmental pallor in the ischemic part (this pattern has caused a good deal
the upper half of the optic disc, mild edema in the inferior temporal
of confusion)then the involved part has pallor but
region and in the lower half, multiple peripapillary hemorrhages
and mild diabetic retinopathy. (Reproduced from Hayreh [47].) is not edematous any more, while the rest of the disc
may show mild edema and even some pallorthe
optic disc edema gradually resolvespallor of the
a
involved region only or the entire disc; in the latter
case, the pallor may or may not be more marked in
the involved part. There is a marked variation in this
pattern from eye to eye, depending upon (a) the
severity of ischemia and the extent of visual loss in
different eyes, and (b) the stage at which different
eyes were initially seen. Therefore, the sector of the
optic disc showing edema usually corresponds to the
location of visual field defect ONLY during the very
early stage, but NOT later on. Arnold and Hepler
[59] similarly recorded no consistent correlation
by the sector of the disc edema with the visual field
defect.
b Delayed development of optic disc edema in the
part of the disc with no demonstrable visual field defect
indicates subclinical hypoxia/ischemia of axons in that
part [56, 60]. This is caused by secondary compression
of capillaries in the uninvolved part of the optic nerve
by the swollen axons in the originally ischemic part,
because of restricted space in the optic nerve head with
Fig. 16.14 Fundus photographs showing optic disc pallor in

NA-AION. (a) Shows marked pallor of the temporal part, mod-
erate of nasal part and minimal below and above. (b) Shows
marked generalized pallor
no or only a small cup. Later on, with the axonal Effect of Severity of Visual Acuity Deterioration
destruction in the original ischemic part of the optic
nerve head, the element of secondary vascular com- Similarly, the study showed that the worse the initial
pression resolves, resulting in restoration of normal visual acuity, the shorter the time to resolution of
blood flow and axonal flow in the uninvolved part and optic disc edema (p=0.04). As in visual field loss,
disappearance of optic disc edema there. the extent of visual acuity deterioration is usually an
This evolutionary pattern of the presence, appear- index of severity of axonal damage. However, when
ance, distribution and severity of edema and pallor of a visual field defect spares the central fixation, the
the optic disc in NA-AION has resulted in a good deal extent of visual field loss and visual acuity deterio-
of confusion about the optic disc changes in NA-AION, ration do not correspond. For example, in eyes with
their relationship to visual loss, and pathogenesis of altitudinal visual field loss but sparing the fixation
NA-AION. This has even resulted in missing the diag- area [17], although half of the visual field is lost
nosis of NA-AION. (as a result of loss of half of the axons), visual acuity
is still 20/20 (Figs. 16.1a and e). That disparity
between the extent of visual field loss and visual
Time from Onset of Visual Loss to Resolution acuity deterioration in some eyes is the reason why
of Optic Disc Edema in NA-AION, there is a more significant correlation between the
and the Factors that Influence It time to resolution of optic disc edema with visual
field loss (p<0.0001) than with visual acuity dete-
Our study [46] showed that the duration of optic disc rioration (p=0.04), although both show a significant
edema in NA-AION is influenced significantly by sev- correlation with the time to resolution of optic disc
eral factors. The overall median time (25th75th percen- edema.
tile) to resolution of optic disc edema from the onset of
visual loss is 7.9 (5.811.4) weeks. It take a signifi-
cantly (p=0.003) longer time for optic disc edema to Effect of Diabetes Mellitus
resolve in diabetic patients and a significantly
(p<0.0001) shorter time in those with worse initial The study showed that in diabetics optic disc edema
visual field loss, worse initial visual acuity (p=0.0009), due to NA-AION takes significantly (p=0.003) longer
and those treated with high dose corticosteroid therapy to resolve than in non-diabetics [46, 61]. The reasons
(p<0.0001). Multi-factor analysis in that study for this are discussed below in optic disc changes in
showed that each of these factors had an independent diabetics and in incipient NA-AION.
influence on the resolution time of optic disc edema.
Naturally, the question arises as to why these factors
influence the time to resolution and what is their Effect of Corticosteroid Therapy
importance.
Interestingly, the study showed that optic disc edema
resolved significantly (p=0.0008) sooner in patients
Effect of Severity of Visual Field Loss who were treated with high dose corticosteroid ther-
apy, than in those who were not, independent of the
The severity of visual field loss is a reflection of the effect of the severity of their initial visual field loss
amount of axonal loss due to ischemia, and that in turn [46] (Fig. 16.15) or visual acuity loss on optic disc
depends upon the severity of ischemia the more edema resolution.
marked the ischemic damage, the earlier and more The presence of an ischemic penumbra in nervous
marked the destruction of the axon [56]. It has been tissue is well known. The postulated rationale for
shown that once an axon is destroyed, there is nothing giving corticosteroid therapy during the early stages
left to swell and consequently no optic disc edema [57, of NA-AION is that it would resolve optic disc
58]. So, it is no surprise that the study showed that the edema faster, leading to improved circulation in the
worse and more marked the initial visual field loss, the vessels of the optic nerve head [22, 62]. The axons
earlier the optic disc edema resolved (p<0.0001). that are still in the phase of reversible ischemic
100%
90%
80%
Cumulative Probility ODE Resolved (%)
70%
60%
50%
40%
No di abetes, No steroid, Mid VF defect
30% No di abetes, No steroid, Severe VF defect
No di abetes, Steroid within 2 weeks, Mild VF defect
20% No di abetes, Steroid within 2 weeks, Severe VF defect
With diabetes, No steroid, Mid VF defect
10% With diabetes, No steroid, Severe VF defect
With diabetes, Steroid within 2 weeks, Severe VF defect
0%
0 4 8 12 16 20 24 28
Weeks From Onset of AION
Fig.16.15 Plots of the fitted cumulative distributions of time to therapy, and severity of initial field defect (Reproduced from
optic disc edema resolution estimated from fitting the multi-fac- Hayreh and Zimmerman [46])
tor model for various combinations of diabetes mellitus, steroid
damage (i.e., anatomically alive but not function- evidence that corticosteroids are effective in many
ing), may recover function with improved circula- non-inflammatory diseases. For example, oral [65]
tion and not suffer irreversible axonal damage. or intravitreal [6669] corticosteroid therapy reduces
Naturally, the question arises as to why corticoster- macular edema due to various causes. Although the
oid therapy shortened the optic disc edema resolution exact mechanism by which corticosteroids act in all
time. Optic disc edema in NA-AION is due to a com- these conditions is not known, it seems most proba-
bination of (a) axoplasmic flow stasis, and (b) increased ble that they alter capillary permeability and reduce
capillary permeability and fluid leakage, as shown by fluid leakage. As discussed below, fluorescein
fluorescein angiography (Fig.16.16). So far, the effect angiography invariably shows leakage of fluorescein
of systemic corticosteroids on axoplasmic flow stasis in the optic nerve head when the disc is edematous
has not been studied in the optic nerve head. In one in NA-AION (Fig. 16.16b), but not in normal or
postmortem, invitro study of the effect of cortisol on atrophic discs a proof of increased capillary per-
axoplasmic flow in prefrontal and temporal cortical meability in optic disc edema. The more rapid reso-
neurons of four aged human brains, a bell-shaped lution of optic disc edema with corticosteroid
effect was found: a stimulating effect at low concentra- therapy suggests a reduction in capillary leakage,
tions, and a depressing effect at high concentrations similar to that in macular edema. Therefore, the
[63]. Whether that has any relevance to in vivo finding that corticosteroid therapy significantly
axoplasmic flow stasis seen in NA-AION remains (p=0.0006) shortens the duration of optic disc
unknown. edema (Fig.16.15) has therapeutic implications, as
Foulds [64] postulated that corticosteroid therapy shown by the significantly better visual outcome fol-
in acute NA-AION reduces optic disc edema by lowing systemic corticosteroid therapy, discussed in
reducing the capillary permeability. There is ample Chap. 17 [22].
Fig.16.16 (a) Fundus

photograph (left) and
a
fluorescein angiogram (right)
of right eye with NA-AION.
Fundus photograph shows
optic disc edema. Fluorescein
angiogram at that time
showed non-filling of
temporal part of the
peripapillary choroid and
adjacent optic disc and of
upper half of the choroidal
watershed zone (arrow), with
normal filling of both the
medial and lateral posterior
ciliary arteries (Reproduced
from Hayreh [14a].).
(b) Fluorescein angiogram
showing late staining of the
optic disc in an eye with
NA-AION
Optic Disc Pallor neuropathy typically corresponds to the resultant visual

field defect; this is the prevalent impression. However,
our study [46] found there was often little or no such
Once optic disc edema resolves in NA-AION, a vari-
correlation. For example, optic discs with uniform,
able severity and distribution of disc pallor is invari-
generalized pallor were frequently associated with only
ably seen (Figs.16.8c, 16.13 and 16.14).
altitudinal, quadrantic or other types of visual field
defects, and visual acuity anywhere from normal to
Relationship Between Distribution and Severity markedly deteriorated. The fact that in such cases the
of Optic Disc Pallor on Resolution of Optic Disc optic disc shows optic disc edema in parts of the optic
Edema, and those of Visual Field Defects and disc at an early stage without any corresponding visual
Visual Acuity Deterioration field defect, and later generalized pallor, indicates some
diffuse optic nerve fiber ischemic damage in an area
Arnold and Hepler [59] stated late segmental optic where there is clinically no detectable visual loss. This
atrophy in nonarteritic anterior ischemic optic seems to be due to the inability of perimetry to show
visual field loss when there is subthreshold, diffuse very early arterial phase of dye filling in the fundus.
loss of nerve fibers in the optic nerve. This is very well My studies have shown that there is almost invariably
demonstrated by OCT and GDX studies in NA-AION a filling defect/delay in the prelaminar region and in
by Dr. Randy Kardon (Personal communication). OCT the peripapillary choroid and/or choroidal watershed
was obtained with Zeiss Meditec Stratus OCT3 using a zones (Fig. 16.16a; also see Figs. 3.39, 3.41, 3.45,
standard peripapillary circular scan (fast retinal nerve 3.52, 3.54, 3.573.59 in Chap. 3) at onset of NA-AION
fiber layer scan) which is an average of three sequential [10, 14a]. Others [59, 70] have shown the same. In the
scans. The GDx was obtained with variable corneal rare case where NA-AION is due to embolism into the
compensation (VCC). Figure16.17 is a representative PCA, the part of the choroid supplied by the occluded
case from Dr. Kardons series. In this case both kinetic PCA or short PCA does not fill (see Figs.3.27, 3.33,
and static (Fig.16.17a) perimetry showed a complete 3.35 in Chap. 3). On angiography, the optic disc with
superior altitudinal field loss and no detectable thresh- optic disc edema always shows dye leaking from the
old decrease in the lower field; however, it clearly capillaries in the optic nerve head and late staining
showed the following two findings: (1) OCT (Fig.16.16b). Fluorescein leakage may be due to two
(Fig.16.17b) and GDX (Fig.16.17c) both showed sim- factors: (1) ischemic insult to the capillaries in the
ilar thinning of both the superior and inferior nerve optic nerve head, and (2) venous stasis produced by the
fiber bundles, but (2) nevertheless, the optic disc pallor capillary compression [71]. Foulds [64] also pointed
(Fig.16.17d) in the right eye was mainly in the inferior out that increased capillary permeability due to anoxic
temporal region. On occasions, I have seen a perfectly capillary damage was an important factor in the devel-
normal-looking optic disc after resolution of optic disc opment of optic disc edema in NA-AION. Therefore,
edema with well documented visual field defects. there are primary and secondary changes in the optic
The implication of these observations is that in nerve head which produce optic disc edema in
NA-AION: (a) the nerve fiber damage is often more NA-AION the primary change being ischemic axo-
extensive than demonstrated by the visual field defects plasmic flow stasis and the secondary vascular changes
or even the visual acuity, and (b) the presence and dis- and fluid leakage. Late optic disc staining (Fig.16.16b)
tribution of optic disc pallor does not always reflect the is due to optic disc capillary leakage, but from the
extent and location of nerve fiber loss. Because of diagnostic point of view it is a non-specific finding of
these confounding factors, one should be cautious in optic disc edema, present in optic disc edema from dif-
interpreting the severity and distribution of visual and ferent causes, and not of diagnostic importance.
nerve fiber loss from the severity and distribution of
optic disc pallor.
Role of Fluorescein Fundus Angiography
in Arteritic AION
Fluorescein Fundus Angiography
Fluorescein angiography is extremely helpful in the
differentiation of arteritic AION from NA-AION.
It is most unfortunate that this extremely useful and
Since the former is due to thrombotic occlusion of the
readily available test is not used as often as it should
PCA, as discussed in Chap. 12, fluorescein angiogra-
be in evaluation of AION.
phy shows evidence of occlusion of the PCA, i.e.,
absent filling of the area of the choroid as well as of the
optic disc supplied by the occluded artery; but in
Role of Fluorescein Fundus Angiography
NA-AION no evidence of occlusion of the PCA is seen
in NA-AION except in an occasional case of embolic NA-AION
(see Figs.3.27, 3.33, 3.35 in Chap. 3). This informa-
It provides important and useful information when tion is available to the ophthalmologist within a few
performed during the initial stages of the disease. minutes and helps to differentiate the two types of
Telltale impaired circulation and its location in AION. Since arteritic AION is a dire ophthalmic emer-
NA-AION on angiography are seen only during the gency, and it is essential to start aggressive high dose
Fluorescein Fundus Angiography 363
Fig.16.17 A 42-year old woman developed NA-AION in her bundles in spite of the fact that there was no detectable threshold
right eye. Two years later, both kinetic and static (a) perimetry decrease in the lower field (a). The optic disc showed pallor
showed a complete superior altitudinal field loss only and visual mainly in the inferior temporal region (d). (Reproduced from
acuity 20/20. At the same time, OCT (b) and GDX (c) showed a Hayreh and Zimmerman [46])
similar thinning of both the superior and inferior nerve fiber
b
Microns 154 73 58
112 53
300
41 43
200 42 55
60 58
51
100
OD
0 95
0 20 40 60 80 100 120 140 160 180 200 220 240 S
65 T N 50
TEMP SUP NAS INF TEMP
I
Microns 56
300
Signal Strength (Max 10) 8
200
100
134 153 158
111 103
0
OS
0 20 40 60 80 100 120 140 160 180 200 220 240 53 68
TEMP SUP NAS INF TEMP 43 87

77 175
161
Microns
148
300 S
200 69 N T 86
I
100
138
0
0 20 40 60 80 100 120 140 160 180 200 220 240 Signal Strength (Max 10) 8
TEMP SUP NAS INF TEMP

OD OS
Fluorescein Fundus Angiography 365
d OD OS
corticosteroid therapy immediately, to prevent any fur- Bilateral NA-AION

ther visual loss, this immediately available information
is invaluable.
Reports of the cumulative probability of the fellow
eye developing NA-AION have varied among different
studies: 25% within 3 years (Fig. 16.19) in 438
patients [75], 17% in 5years in 431 patients [76] and
Other Fundus Changes in NA-AION 15% over 5years in 326 patients [77]; however, dif-
ferent criteria were used to determine the probability,
The presence of a few splinter hemorrhages on optic which may explain the differences. According to one
disc or immediate peripapillary region is common in study [75], the risk is greater in men (Fig.16.20), par-
association with the optic disc edema (Figs.16.9 and ticularly young (<45 years) diabetic men; this sub-
16.10); those resolve spontaneously with resolution of group had 1.56 times the risk of young diabetic
the optic disc edema. Diabetics tend to have more peri- women, 2.56 times the risk of women who are either
papillary retinal hemorrhages than non-diabetics (see nondiabetics or who are not young, and 1.64 times the
below) [61, 72]. Occasionally, I have seen mild serous risk of both older men and nondiabetic men
retinal detachment between the optic disc and macula (Figs.16.21 and 16.22). This is an intriguing finding,
and that may even extend to the macular region to pro- but we have no explanation for it. According to
duce macular edema, which resolves with resolution another study [77], an increased incidence is associ-
of optic disc edema (Fig. 16.18). This has also been ated with poor baseline visual acuity and diabetes
reported by other authors [73, 74]. Because of the optic mellitus. The risk of the second eye getting involved
disc edema, there is a certain amount of retinal venous by NA-AION was evaluated by us [61] in 655 patients
engorgement. In some eyes, as the optic disc edema (206 diabetics and 449 nondiabetics) and that showed
resolves, some lipid deposits are seen in the peripapil- a significantly (p=0.003) greater risk in diabetics
lary or macular region, particularly in patients with than in nondiabetics; our study [61] also showed that
elevated lipids. These lipid deposits persist longer than the median (25th75th percentile) time to involvement
optic disc edema, so they are a good footprint for the of the fellow eye by NA-AION was 6.9 (0.416.9)
presence of optic disc edema in the recent past, and years in diabetics and 9.1 (1.819.0) years in
this is helpful towards the diagnosis of NA-AION nondiabetics.
when the patient is seen after the resolution of optic There are reports in the literature of simultaneous
disc edema. onset of bilateral NA-AION. In my study of about
Bilateral NA-AION 367
OCT Image Fundus Image
Scanned Image
S
T %
1 5
Fig.16.18 Fundus photograph (a) and OCT (b) of right eye with NA-AION and serous retinal detachment between the optic disc
and the macula (Reproduced from Hayreh [125])
c
STRATUS OCT
Proportional Process Report - 4.0.5 (0076)
MORRIS, THOMAS Scan Type: 7MM post. pole OD- OD
Scan Date: 6/22/2006
DOB: 4/18/1958, ID: 02-39472-8, Gender unknow
Scan Length: 7.0mm
OCT Image Fundus Image

S
Scanned Image T X
1 5
Signature:
Physician: Randy Kardon Univ of Iowa Ophthalmology & Visual Scienc...
1,350 patients with NA-AION, the reported simul- develop sudden, severe arterial hypotension, e.g. dur-
taneous visual loss in both eyes due to NA-AION ing hemodialysis or surgical shock.
was usually reported by patients who had been
unaware of the NA-AION in the first eye until the
second eye got involved. The pattern of optic disc
edema during the initial stages of NA-AION is dif- Recurrence of NA-AION in the Same Eye
ferent from that later on (see above), and that is help-
ful to time the onset in bilateral NA-AION. Such recurrences mentioned in the literature are often
Simultaneous bilateral onset of NA-AION is more a progression of NA-AION during the acute
extremely rare, occurring only in patients who stage rather than actual new episodes after the first
Recurrence of NA-AION in the Same Eye 369
100 100
80 80
Cumulative percent
Cumulative percent
60 60
40 40
20 20
0 0
0 12 24 36 48 60 72 84 96 108 120 0 12 24 36 48 60 72 84 96 108 120
Months to bilaterality Months to bilaterality
ARTER ARTERITIC NON-ARTERITIC AGE < 45 4564 >=4564
Fig. 16.19 A graphic representation of cumulative incidence Fig. 16.21 A graphic representation of cumulative incidence
rates (Kaplan-Meier estimates) of developing bilateral arteritic rates (Kaplan-Meier estimates) of developing bilateral NA-AION
and non-arteritic AION. The curve for arteritic AION ends at in patients of different age groups (i.e. <45, 4564, and65years
6.1 months because in no patients with giant cell arteritis did old) (Reproduced from Beri etal. [75])
arteritic AION subsequently developed (Reproduced from Beri
etal. [75])
100 100
80 80
Cumulative percent
Cumulative percent
60 60
40 40
20 20
0 0
0 12 24 36 48 60 72 84 96 108 120 0 12 24 36 48 60 72 84 96 108 120
Months to bilaterality Months to bilaterality
SEX MALE FEMALE DISEASE DM only HT only DM and HT
Others Others
Fig. 16.20 A graphic representation of cumulative incidence
rates (Kaplan-Meier estimates) of developing bilateral NA-AION Fig. 16.22 A graphic representation of cumulative incidence
in men and women (Reproduced from Beri etal. [75]) rates (Kaplan-Meier estimates) of developing bilateral NA-AION
in patients with diabetes mellitus, arterial hypertension, both
diabetes and hypertension and other systemic disease or idio-
pathic group (Reproduced from Beri etal. [75])
episode has resolved completely. In our study of 829
NA-AION eyes [78], on a median follow-up of
3.1years (inter-quartile range of 1.17.2years), 7.5% 2.7%0.7%, at 1 year 4.1%0.9%, and 2 years
patients or 6.4% eyes developed more than one epi- 5.8%1.1%. There was no difference in the presence
sode of NA-AION in the same eye (i.e. recurrence of or absence of an optic disc cup between the groups
NA-AION). Four patients developed more than two without and with recurrence of NA-AION in the same
episodes of NA-AION anywhere from three to more eye. There was an association between age at first
than seven in an eye. The Kaplan-Meier survival curve onset and recurrence (p=0.069), in particular, patients
(Fig. 16.23), showed the cumulative percentage of who had a recurrence in both eyes were those who had
recurrence of NA-AION from first episode to second their first onset of AION at a significantly younger age
episode at 3 months 1.0%0.4%(SE), at 6 months than patients with no recurrence (p=0.007). There was
0.40 her waking hours it was perfectly normal (mean systolic

and diastolic pressures 129/73mmHg) that is, she had
severe nocturnal arterial hypotension. Her intraocular
pressure on a 3hourly plotting during the waking hours
0.30 was between 15 and 20mmHg in both eyes. The amount
Recurrence Disrtribution
of fall of blood pressure during the night depends upon

the depth of sleep [9]. It is very likely that, when this
lady was sound asleep at home in her own bed (and not
0.20
disturbed by the repeated inflation and deflation of the
arm blood pressure monitoring cuff every 20min in a
hospital bed) her diastolic blood pressure may have
0.10
been even lower than 35 mmHg. Moreover, the
35mmHg diastolic blood pressure was in the brachial
artery, whereas it would be about half of that or even
less in the capillaries of the optic nerve heads. A num-
0.00 ber of studies have shown that intraocular pressure is
Year 0 2 4 6 8 10 12 14 16 18 higher during sleep than during waking hours [7982].
(n) 594 345 238 171 121 83 63 48 34 19
Therefore, this patients intraocular pressure was most
Fig.16.23 Kaplan-Meier curve showing the incidence of recur- probably about 1920 mmHg or even higher during
rence of NA-AION over a 20-year period of follow-up, and the sleep. From this information, one can deduce that there
number of patients that were followed at various time intervals were probably periods during sleep when there was no
(Reproduced from Hayreh etal. [78].)
blood flow in the optic nerve head capillaries during the
diastolic phase of the cardiac cycle, resulting in a pulsa-
no significant association between the recurrence of tile blood flow in the optic nerve head. She discovered
NA-AION and the systemic conditions that were visual loss in most of her episodes on waking up in the
examined, except for nocturnal diastolic arterial morning, which is the usual pattern in NA-AION [7]. In
hypotension. Patients with a recurrence of NA-AION addition to all that, she was diabetic and had had ele-
had a significantly lower mean nighttime minimum vated cholesterol for years. This case reinforces the
diastolic blood pressure (p=0.003) and greater mean findings about the role of nocturnal arterial hypotension
percentage drop during sleep in diastolic blood pres- in the development of a first episode of NA-AION as
sure (p=0.011) than did the patients who did not suffer well as its recurrence [3, 7, 9, 10, 83].
recurrence. Compared to the development of NA-AION In another case in our study [78], there was a strong
in the second eye (see above), the development of more suggestion of recurrent microembolization as the main
than one episode of NA-AION in the same eye is much risk factors in the development of four episodes in the
less common, being 5.8%1.1% on a follow-up of right and two episodes in the left eye. However, this
2years after the first episode. Thus, nocturnal arterial seems to be a very rare cause. If patients have elevated
hypotension emerges as an important causative fac- cholesterol and associated plaques in the carotid arter-
tor for persons developing more than one episode of ies, it is generally considered that the associated isch-
AION in an eye. emic disorders are due to embolism; however, our
The following case report in our study is an instruc- experimental studies [15, 84] have suggested that
tive example of this [78]. A 57-year old hypertensive release of serotonin (a potent vasoconstrictor) by plate-
woman, who was taking multiple arterial hypotensive lets deposited on plaques can also cause vascular
therapies more than once a day, developed more than occlusion by producing a spasm, and that may play a
seven episodes of NA-AION in the right eye and more role in the recurrence of NA-AION in some cases.
than six in the left eye, over a follow-up period of The question arises: why does NA-AION recur so
15years in my clinic, rendering her legally blind with much less often in the same eye than in the fellow eye?
extensive visual loss in both eyes. On 24-hour ambula- A number of hypotheses have been put forward to
tory blood pressure monitoring, her lowest blood pres- explain the infrequent recurrence of NA-AION in the
sure during sleep was barely 77/35mmHg while during same eye, and these include the following:
NA-AION in Diabetics Versus Non-diabetics 371
Inability of a Patient to Recognize Further compressed by the swollen axons, a fall of BP during
Visual Loss After the Initial Visual Loss sleep (nocturnal arterial hypotension) may further
reduce the perfusion pressure in those capillaries, lead-
ing to reduced blood flow or even no circulation during
It has been postulated that a patient might be less likely sleep, and consequent ischemia of the ONH. That
to recognize further visual loss in the same eye after would explain why nocturnal arterial hypotension is
having suffered visual loss once [85, 86]. In our study not only a precipitating factor for NA-AION in vulner-
[78] this happened only rarely. Beck et al. [85]. also able ONHs but also shows a significant correlation
discounted this as a valid reason for the rare finding of with progressive visual field deterioration in NA-AION,
recurrence of NA-AION in the same eye. particularly in arterial hypertensives on hypotensive
therapy [9, 83].
As a corollary to that, some ophthalmologists have
put forward the hypothesis that initial destruction of
Shunting Blood from the Damaged Part nerve fibers in a crowded optic disc may provide more
of the Optic Nerve Head (ONH) to the space for the surviving fibers to swell into, thus mini-
Normal Part mizing the effect of the mechanical factor that may
aggravate an initially mild ischemic insult, rendering it
It has been postulated that once a portion of the optic irreversible [86]. However, this hypothesis does not
disc is damaged by ischemia, the blood supply is then stand up to critical scrutiny. In glaucomatous optic
shunted away from this area to the still-functioning neuropathy and arteritic AION [47, 87], there is an
part of the disc, and that is sufficient to protect the increase in cup size with the loss of axons, but not in
remainder of the disc from a second ischemic episode NA-AION [12]. This is because destruction of the
[85]. Based on my multiple studies for more than axons in NA-AION is accompanied by secondary glio-
4decades on various aspects of the ONH circulation in sis in the ONH, without development of any cupping
health and disease, I can find no evidence to support or increase in cup size, so that no additional space
this speculation. Similarly the concept of so-called becomes available for the surviving axons.
luxury perfusion, postulated by some, is not valid. Thus, we do not yet have a satisfactory explanation
why NA-AION recurs so much less often in the same
eye than in the fellow eye.
Initial Destruction of Nerve Fibers in

a Crowded Optic Disc Providing More NA-AION in Diabetics Versus
Space for the Surviving Fibers Non-diabetics
To explain why optic discs without a cup are far more Comparison of various aspects of NA-AION in diabet-
vulnerable to NA-AION, it has been [13, 14] postu- ics and non-diabetics revealed interesting information
lated that in such discs there is crowding of the optic in our study [61] of 655 consecutive NA-AION patients
nerve fibers in a restricted space, surrounded by the (931 eyes) 206 patients with diabetes and 449 with-
inelastic Bruchs membrane. When the axons swell out. Comparison of various clinical features of
due to axoplasmic flow stasis during the incipient stage NA-AION between diabetics and non-diabetics
of NA-AION [56, 60], there is no space for them to showed no significant difference in age, but diabetics
expand (because of the absence of a cup and rigid had slightly more women than men (45% vs. 38%;
Bruchs membrane around the optic canal). p=0.078), and a higher prevalence of arterial hyper-
Consequently they compress the intervening capillar- tension (p<0.0001), ischemic heart disease
ies; this further compromises the ONH blood flow, (p=0.0001), transient ischemic attacks (p=0.0003),
thereby setting up a vicious circle: ischemiaaxonal and second eye involvement by NA-AION (p=0.003).
swellingcompression of the capillaries in the Initial visual acuity did not differ significantly between
ONHmore ischemia. When the ONH capillaries are diabetics and non-diabetics; however, of those seen
within 2 weeks of onset of NA-AION, diabetics had disc, and much more numerous peripapillary retinal
less severe visual field defect (p=0.010). At 6months hemorrhages than in non-diabetics [10, 61, 72]
after onset, there was no significant difference in visual (Figs. 16.13, 16.24a, c, 16.25a, c, 16.26 and 16.27).
acuity and visual field improvement between diabetics When the optic disc edema resolves spontaneously,
and non-diabetics. The time to optic disc edema reso- these prominent telangiectatic disc vessels and retinal
lution was significantly (p=0.003) longer in diabetics hemorrhages also resolve spontaneously (Figs.16.24b, d
than non-diabetics. and 16.25b, d). This pattern of optic disc edema and
fundus changes in diabetics during the acute phase of
NA-AION has confused the issue and often resulted in
a good deal of controversy, and the following two types
Pattern of Optic Disc Edema of misdiagnoses.
1. These findings have been mistaken for proliferative
and Fundus Findings in Diabetics Versus
diabetic retinopathy associated with optic disc neovas-
Non-diabetics cularization, and mistakenly treated with panretinal
photocoagulation [10, 61, 72]. Then the spontaneous
This is of great practical clinical importance. My stud- resolution of optic disc and fundus changes have been
ies of NA-AION since 1973 have shown that in diabet- attributed to the supposed beneficial effect of pan-
ics of all ages, during the initial stages of NA-AION, retinal photocoagulation [61, 72]; in fact, panretinal
the optic disc edema is usually (though not always) photocoagulation is not only not indicated in this dis-
associated with characteristic and diagnostic prominent, ease, but produces unnecessary extensive retinal dam-
dilated and frequently telangiectatic vessels over the age and associated visual complications.
Fig.16.24 Fundus photographs (a, b) of the right eye and (c, d) engorged retinal veins, and in the right macular region lipid depos-
of the left eye, of a 19 year-old white male juvenile diabetic with its. Visual acuity was 20/25 in the right eye and 20/40 in the left
NA-AION (Reproduced from Hayreh and Zahoruk [72]) (a, c) On eye, with enlargement of blind spot in both eyes. (b, d) On resolu-
first visit to our clinic: Fundus photographs, (a) of the right eye tion of optic disc edema: Fundus photographs, (b) of the right eye
and (c) of the left eye, show massive optic disc edema with marked and (d) of the left eye, show normal-looking optic discs, no abnor-
telangiectatic vessels on the optic disc, many retinal hemorrhages, mal vessels on the discs, and no retinal hemorrhages
Pattern of Optic Disc Edema and Fundus Findings in Diabetics Versus Non-diabetics 373
Fig. 16.25 Fundus photographs of both eyes of a 51-year old and many retinal hemorrhages. Visual acuity was 20/20 in the left
woman with adult onset diabetes mellitus. She developed bilateral and 20/15 in the right eye. Both eyes had inferior nasal visual field
NA-AION, first in the right eye (a, b) and 8months later in the left defect. (b, d) On resolution of optic disc edema: Fundus photo-
eye (c, d) (Reproduced from Hayreh and Zahoruk [72]) (a, c) On graphs show no optic disc edema but mild temporal pallor, no
first visit to our clinic: Fundus photographs show massive optic abnormal vessels on optic disc, and no retinal hemorrhages in the
disc edema with marked telangiectatic vessels on the optic disc, right eye (b) and a few resolving hemorrhages in the left eye (d)
Fig. 16.27 Fundus photograph of the left eye of a diabetic

Fig. 16.26 Fundus photograph of the left eye of a diabetic patient with NA-AION 16days after onset, with visual acuity of
patient with NA-AION when seen at initial visit 20/40. This shows optic disc edema which was maximum in the
superior nasal half, with superficial peripapillary retinal hemor-
rhages and early diabetic retinopathy. Fig.16.13 is 3months
after onset (Reproduced from Hayreh [47])
2. This pattern of optic disc edema and fundus changes
in NA-AION in diabetics has also often been misdi-
(b) Some diabetic patients with NA-AION or incip-
agnosed as a distinct clinical entity called diabetic
ient NA-AION also concurrently have diabetic
papillopathy or diabetic papillitis [8896], caus-
macular edema with poor visual acuity; this has
ing confusion and controversy. The following two
added to the confusion. Diabetic macular edema,
factors in diabetics are responsible for this
per se, however, does not produce any optic
misunderstanding:
nerve related visual field defects.
(a) Incipient NA-AION [60] (see below) initially
occurs more frequently in diabetics than non- We have discussed at length elsewhere why diabetic
diabetics. papillopathy is actually NA-AION in diabetics, and
Incipient NA-AION 375
not a distinct clinical entity [46, 60, 72, 97],. Once one time to resolution of the optic disc edema in the pro-
considers all these factors, it becomes apparent that the gressed group was 5.8 weeks versus 9.6 weeks in
term and diagnosis of diabetic papillopathy are not those that did not. Patients who progressed to classical
correct and confuse the issue. NA-AION were significantly younger than those who
Studies have shown no difference in NA-AION in did not. Patients with incipient NA-AION had a sig-
optic disc morphology [13] or fluorescein angio- nificantly (p<0.0001) greater prevalence of diabetes
graphic [46] findings between the diabetic and non- mellitus than classical NA-AION. Thus, incipient
diabetic NA-AION eyes, except that, as discussed NA-AION is a distinct clinical entity.
above, diabetic NA-AION eyes usually had telangiec- The following two case summaries are representa-
tatic vessels during the acute phase of NA-AION (as tive of a common clinical pattern in incipient
shown in Figs. 16.13, 16.2416.28), which leaked NA-AION.
fluorescein.
Case 1
Incipient NA-AION
A 52-year old white male diabetic was first seen in my
In 1981, I [56] reported that symptomless optic disc clinic with a history of loss of the lower half of the
edema precedes the visual loss and may be the earliest visual field in his left eye (Fig.16.1e). On examination
sign of AION (NA-AION). There is an almost universal he was found to have classical NA-AION with optic
belief that unless and until there is sudden visual loss in disc edema and inferior altitudinal defect and visual
addition to optic disc edema, it cannot be NA-AION. acuity of 20/20. Two months later, at a routine follow-
This has resulted in: (a) missing the diagnosis of up visit, examination of his right eye showed optic disc
NA-AION when a patient presents initially with asymp- edema involving the upper part of the optic disc with
tomatic optic disc edema, and (b) extensive and expen- some telangiectatic vessels (Fig.16.28a, b). On ques-
sive neurologic and other investigations. For example, tioning he denied any subjective visual symptoms at all
initial asymptomatic optic disc edema in diabetics has in that eye. On examination, visual acuity in the right
been designated as diabetic papillopathy or diabetic eye was 20/15, and visual fields showed no defect at all
papillitis (see above).Similarly, initially asymptomatic even with I-1e isopter, except for some enlargement of
optic disc edema in patients who take amiodarone has the blind spot with I-1e (Fig.16.29). The patient was a
been considered amiodarone optic neuropathy (see traveling salesman; because of the permanent absolute
below). As discussed in this chapter, both these condi- inferior altitudinal defect in his left eye, he was very
tions are in fact NA-AION. apprehensive that if the same visual defect developed in
I [56] was able to trace seven similar cases previ- his right eye, he would be crippled professionally. When
ously reported in the literature [32, 64, 98, 99], and I seen 2weeks later, optic disc edema in his right eye was
[10] later named this clinical entity incipient slightly more marked (Fig. 16.28c), still without any
NA-AION. Recently we [60] reported the clinical subjective or objective visual loss. In consultation with
features and other aspects of incipient NA-AION the Endocrinology Department of our University
based on 54 such patients (60 eyes). In that study, the Hospital, he was treated with systemic steroid therapy
mean (SD) age of the patients was 58.715.9 years (according to the protocol described in Chap. 17) while
(range 1685years). There was no significant (p=0.43) that department monitored his diabetes. One week later
difference in mean age at onset/diagnosis between the his optic disc edema was resolving and 3weeks later
incipient NA-AION and classical NA-AION. At the there was only mild optic disc edema. Six weeks after
initial visit, all eyes with incipient NA-AION had that, when he was off steroid therapy, he woke up in the
optic disc edema without any visual loss attributable morning with a complete loss of vision in the lower
to NA-AION. In 55%, the fellow eye had classical field of his right eye. Next day in the clinic, examina-
NA-AION, in 25% incipient progressed to classical tion showed visual acuity of 20/20 and absolute inferior
NA-AION (after a median time of 5.8 weeks), and nasal sector defect, sparing fixation (Fig.16.1b), with
20% developed classical NA-AION after resolution of recurrence of optic disc edema; 1week later the visual
a first episode of incipient NA-AION. The median field defect had progressed to an absolute inferior
a b
c d
Fig. 16.28 Fundus photographs and fluorescein angiogram of nent, dilated vessels over the edematous part of the optic disc.
the right eye of a diabetic patient (Reproduced from Hayreh SS Visual field on this visit is shown in Fig.16.29. (c) Two 2weeks
[55].) (a, b) On the day asymptomatic optic disc edema was first later: optic disc edema and prominent, dilated vessels in his right
discovered: (a) Fundus photograph. (b) Fluorescein fundus angio- eye are slightly more marked, and the eye was still completely
gram during the late phase; both show optic disc edema involving asymptomatic. (d) On resolution of optic disc edema: Pallor of
the superior temporal region of the optic disc. In (a) note promi- the upper half of the optic disc. Visual field shown in Fig.16.1c
altitudinal defect, bisecting fixation (Fig.16.1c). About

3months later, optic disc edema had resolved, with
development of pallor of the upper half of the disc and
a normal lower half (Fig.16.28d). There was no visual
recovery in either eye on follow-up.
Case 2
A 19 year-old white male juvenile diabetic was referred

to my clinic for evaluation of diabetic retinopathy. On
questioning he said that he had experienced vague blur-
ring of vision for 34h every morning in both eyes for
Fig.16.29 Visual field (plotted with a Goldmann perimeter) of
about 1month and had no other visual complaint. On
the right eye with incipient NA-AION shows mild enlargement examination, the visual acuity was 20/25 in the right and
of the blind spot only (Reproduced from Hayreh [56]) 20/40 in the left eye, and visual fields were normal
Incipient NA-AION 377
except for an enlarged blind spot in both eyes. Thus, in Case 1, incipient NA-AION was discovered
Ophthalmoscopy revealed bilateral massive optic disc accidentally during a routine follow-up of classical
edema with marked telangiectatic vessels on it, and NA-AION in the fellow eye, while in Case 2 incipient
punctate hemorrhages in the surrounding retina NA-AION developed simultaneously in both eyes. In
(Fig.16.24a, c). Both eyes showed evidence of early dia- both cases, the visual field defects were typically those
betic retinopathy, with mild macular edema and some seen in NA-AION, i.e. the most common was the inferior
lipid deposits in the right macular region. A complete nasal defect and the next the inferior altitudinal defect.
evaluation in the Neurology Department of our University When considering incipient NA-AION, one has to
Hospital revealed no neurologic abnormality to explain bear in mind that almost all diseases go through an evo-
the bilateral optic disc edema. Eleven days later, his left lutionary process, passing from asymptomatic to symp-
eye developed blurred vision in the lower half and when tomatic phases. It is only rarely that the clinician is
seen in our clinic a week after that the visual acuity was fortunate enough to observe the evolution of a disease
20/25 in the right and 20/30 in the left eye. Visual fields before it becomes symptomatic, since it is the onset of
in the left eye showed inferior altitudinal defect, sparing symptoms that normally brings the patient to a physician.
fixation, with I-2e isopter, a large inferior scotoma with Similarly, NA-AION patients are usually first seen when
III-4e and an inferior nasal defect with I-4e isopter they develop visual loss. In incipient NA-AION there is:
(Fig.16.6h), and in the right eye only an enlarged blind (a) optic disc edema, with a segmental pattern similar to
spot. Optic disc edema was more marked than before. that seen in the early stages of classical NA-AION, i.e.,
He was diagnosed to have bilateral NA-AION. In con- often either involving only one part of the disc or more
sultation with the Endocrinology Department of our marked in one part than the other (Fig.16.8b), (b) no sub-
University Hospital, he was treated with systemic ste- jective or objective visual loss attributable to classical
roid therapy (according to the protocol described in NA-AION though some eyes had an enlarged blind
Chap. 17) while that department monitored his diabetes. spot corresponding to the optic disc edema, and (c) ocu-
Six days later his visual acuity was 20/20 in both eyes. lar, orbital, neurologic and systemic evaluation reveals
Both fundi were normal after 7weeks (Figs.16.24b, d). no other cause for the asymptomatic optic disc edema.
The enlarged blind spot in the right eye gradually By contrast, classical NA-AION patients first present
resolved to normal. When optic disc edema resolved in with a typical clinical picture, i.e., a history of sudden
both eyes, the left eye showed almost normal visual field visual loss, usually discovered on waking up in the morn-
(Fig.16.30) comparison of visual fields in Figs.16.6h ing [7], segmental optic disc edema initially [46], and
and 16.30 shows marked visual field recovery. optic disc related visual field defects [17] with or without
defective visual acuity [24]. Thus, available evidence
indicates that incipient NA-AION represents an asymp-
tomatic, early phase in the evolution of the disease pro-
cess. This is further supported by the fact that in this
study, of the 60 eyes with incipient NA-AION, over half
(55%) had classical NA-AION simultaneously or at
another time in the fellow eye; of the 55 eyes with fol-
low-up, in 25% incipient NA-AION progressed to classi-
cal NA-AION, and 20% developed classical NA-AION
in a second, later episode after the resolution of the epi-
sode of incipient NA-AION.
Since the main feature of incipient NA-AION is
asymptomatic optic disc edema, naturally the question
arises: how to differentiate that from optic disc edema
due to other causes? Optic disc edema is seen in a whole
host of conditions, for example, unilateral optic disc
edema in eyes with central retinal vein occlusion, ocular
Fig.16.30 Visual fields (plotted with a Goldmann perimeter) of
the left eye with incipient NA-AION shows almost normal visual
hypotony, posterior uveitis, optic neuritis, optic disc vas-
field, with mild inferior nasal constriction with I-4e (Reproduced culitis [65, 100], neuroretinitis, vitreous traction syn-
from Hayreh and Zahoruk [72]) drome, orbital compressive lesions, Lebers hereditary
optic neuropathy, pseudopapilledema, optic nerve infil- flow itself is not concerned with transmission of a visual
trative lesion, and sometimes with raised intracranial impulse [105] the two are independent processes.
pressure. The most common cause of bilateral optic disc This is very well demonstrated by optic disc edema in
edema is raised intracranial pressure, and less often raised intracranial pressure, which, though due to axo-
malignant arterial hypertension [101] or any one of the plasmic flow stasis [102], is not accompanied by any
aforementioned causes of unilateral optic disc edema. In detectable axonal dysfunction per se. There are other
some of the above mentioned conditions, optic disc similar examples of optic disc edema due to axoplasmic
edema is almost always associated with visual loss, e.g., flow stasis without any visual dysfunction.
in optic neuritis, neuroretinitis, and Lebers hereditary Pathogenesis of incipient NA-AION, is discussed
optic neuropathy. The differential diagnosis of other in Chap. 15.
causes of asymptomatic optic disc edema from incipient
NA-AION is relatively easy. Central retinal vein occlu-
sion always has markedly engorged retinal veins and reti-
Progression of Incipient NA-AION
nal hemorrhages, especially in the periphery, at the time
of associated optic disc edema. In ocular hypotony the to Classical NA-AION
intraocular pressure is always very low and there is often
associated macular edema. Posterior uveitis eyes have As mentioned above, incipient NA-AION most likely
cells in the vitreous and may also have low intraocular represents the earliest clinical stage in the evolution of
pressure. Optic disc vasculitis [100] eyes, in addition to classical NA-AION. In my initial study in 1981, I [58]
optic disc edema, often have visual symptoms due to reported progression in two of four patients with incip-
marked enlargement of the blind spot, cells in the vitre- ient NA-AION after 7 and 12weeks respectively. I was
ous, peripapillary hemorrhages, retinal venous engorge- able to find in the literature reports of seven more simi-
ment and other findings. In eyes with compression of the lar cases [32, 64, 98, 99], who progressed after 3days
retrobulbar optic nerve, in addition to optic disc edema, (one case) [99], 2weeks (one) [32], 46weeks (four)
there are choroidal folds and other evidence of orbital [98], and at a later date (one) [64]. Almog and
mass, e.g., proptosis. Raised intracranial pressure usually Goldstein [106] described 23 patients aged 4774years
causes bilateral optic disc edema, although that can with asymptomatic disc edema; nine of their patients
sometimes initially be unilateral or asymmetric, and is progressed to classical NA-AION after 16.8 weeks
almost invariably associated with systemic and neuro- (range 280weeks). The conversion rate to NA-AION
logic signs and symptoms. Similarly, malignant arterial was 40% in patients who had had NA-AION in the fel-
hypertension is associated with not only markedly ele- low eye, 31% in patients with diabetes, and 43% in
vated blood pressure but also other hypertensive fundus patients with diabetic retinopathy.
changes [101] and systemic symptoms. Thus, with com- The pathogenesis of progression of incipient
plete ocular, orbital, neurologic and systemic evaluations, NA-AION to classical NA-AION is discussed in Chap. 15.
one can usually differentiate these possibilities from Based on the evidence of our study [60] and other
incipient AION. The presence of (a) classical NA-AION studies, I believe that all NA-AION patients start with
in the fellow eye, or (b) progression of incipient NA-AION asymptomatic optic disc edema. Normally they are seen
to classical NA-AION, or (c) development of classical only when they progress to classical NA-AION with
NA-AION later on in that eye provides firm evidence. In visual loss, unless they happen by chance to be exam-
my original paper in 1981 [56], I concluded that: In the ined (as in case 1 above) and the incipient AION is dis-
differential diagnosis of symptomless optic disc edema covered incidentally. From the clinical point of view, it
seen in the elderly and diabetic middle-aged persons, is important to answer the following two questions:
AION (incipient NA-AION) must be borne in mind as a
strong possibility if intracranial, orbital, and ocular and
hematologic and systemic diseases are ruled out. Is Every Eye with Incipient NA-AION Doomed
Axoplasmic flow stasis is the common pathway for to Visual Loss?
the development of optic disc edema in a variety of con-
ditions [71, 102, 103]. Ischemia of the axons is known Our study [60] showed that 25% progressed to classical
to produce axoplasmic flow stasis [103, 104], and that NA-AION and 20% developed classical NA-AION
has been demonstrated in AION [104]. Axoplasmic later on, after the resolution of incipient NA-AION.
Special Types of NA-AION 379
The ipsilateral recurrence rate of NA-AION in these nerve head. In such circumstances, some secondary
eyes was much higher than that seen in classical factors predispose it to conversion to classical
NA-ION (20% versus 6.4% [78]). In the rest there was NA-AION with visual loss. The potential secondary
no subjectively or objectively detectable visual loss (on risk factors so far known are: (1) absence of optic
visual acuity testing and perimetry). However, there is disc cup [13, 14, 60] (its role is discussed in Chap.
always the possibility of some subthreshold loss of 14), and (2) importantly nocturnal arterial hypoten-
nerve fibers in the optic nerve, as ordinary visual testing sion [9, 15, 83] (its role is discussed in Chap. 14). Of
(i.e., visual acuity and perimetry) is not capable of rul- these two factors, the only one we can sometimes
ing it out, as has recently been demonstrated by OCT control is prevention of nocturnal arterial hypoten-
and GDX studies in NA-AION by Dr. Randy Kardon sion (which usually acts as the last straw in the
discussed above [46]. Therefore, from the practical scenario [7]); so it must be controlled wherever pos-
point of view, not all patients with incipient NA-AION sible. Management of incipient NA-AION is dis-
are necessarily doomed to visual loss. On the other cussed in Chap. 17.
hand, patients with incipient NA-AION who recover
completely appear to be at higher risk for the develop-
ment of symptomatic NA-AION at a later date, com-
pared with the risk of recurrent NA-AION after an Misconceptions About Incipient NA-AION
attack of symptomatic NA-AION [78]. This increased
risk may be due to fewer (if any) optic nerve axons
Since this entity is more common in diabetics than non-
being destroyed during the attack of incipient
diabetics, it has often been misdiagnosed as diabetic
NA-AION, thus resulting in continued crowding of
papillopathy or even proliferative diabetic retinopa-
nerve fibers in the optic nerve head and/or persistence
thy and treated with panretinal photocoagulation
of other predisposing risk factors in the multifactorial
which is not indicated and can be harmful [60, 72].
scenario.
What are the Potential Risk Factors which Special Types of NA-AION
may Predispose Incipient NA-AION to Progress
to Classical NA-AION?
Familial NA-AION
In other words, what can we do in such cases to pre-
vent their progressing to classical NA-AION? As NA-AION almost invariably occurs in an apparently ran-
discussed in Chap. 14, NA-AION is due to deranged dom, sporadic fashion; however, there have been five
blood flow in the optic nerve head, and the latter is reports which identified the disease in more than one
influenced by multiple factors (discussed in Chap. 5) member of a family in ten unrelated families [110114].
[107, 108]. Unfortunately, we do not currently have It is essential to find out if clinical features of familial
the means to evaluate many of those risk factors, as NA-AION are similar or different in any way from the
discussed elsewhere [109] (see Chap. 14); for exam- classical NA-AION.
ple, abnormalities in the autoregulation of optic
nerve head blood flow and endothelial derived vaso-
active agents play an important role in causing Clinical Features of Familial NA-AION
deranged blood flow in the optic nerve head, but we
cannot evaluate or manage them. This is the dilemma Gender
in the management of NA-AION. The only thing we
know is that incipient NA-ION results from the vari- The gender pattern in the reported cases was as follows.
ous risk factors producing hypoxiaswollen axons Berggren etal. [110] in one family with NA-AION had
from axoplasmic flow stasisasymptomatic optic one male and two females. In the series of Wang etal.
disc edema. By definition, in incipient NA-AION, [112]., six of the nine patients were males. Manor [111]
swollen nerve fibers are not destroyed, but they only reported NA-AION in identical female twins. In our
cause compression of the capillaries in the optic series [114], of 17 patients, there were 14 males and
three females (Figs.15.3ac in Chap. 15). Thus, of these Incidence of Familial NA-AION Among
31 patients with familial NA-AION, there were 21 males NA-AION Patients
(67%) and ten females (33%). In classical NA-AION, in
one study of 624 consecutive patients with NA-AION, In my Ocular Vascular Clinic at the University of Iowa
the ratio was 59% males versus 41% females [4] and in since 1973, I have seen about 1,350 patients with
another study [6] 58% versus 42% respectively. Thus NA-AION. During that period, only two families were
there seems to be no significant gender difference among found to have more than one member affected with
the familial and non-familial NA-AION. NA-AION. This shows that the incidence of familial
NA-AION is extremely rare.
Age of Onset
In the reported cases it was as follows: Berggren etal.s NA-AION and Phosphodiesterase-5
[110] patients were 53, 54 and 58years old, and Manors (PDE5) Inhibitors
[111] identical twins developed NA-AION at 45 and
49years respectively. In the series of Wang etal. [112] Since 1998, phosphodiesterase type 5 (PDE5) inhibitors
age range was 3281 (meanSD 55.014.9) years, with have been used extensively for the treatment of erectile
55% occurring before the age of 60years. Among all the dysfunction. These agents include sildenafil (Viagra,
patients in the three families in our series [114], age Pfizer), vardenafil (Levitra, Bayer AG), and tadalafil
range was 2867 (mean 47.38.6) years. In 624 con- (Cialis, Lilly-ICOS LLC). Development of NA-AION
secutive patients with classical non-familial NA-AION, has been reported following the use of these drugs [115,
the age range at onset of NA-AION was between 11 and 116] mostly Viagra. This has attracted a good deal of
91 (mean 60.113.6) years, with 11% below 45years, interest and has become an important and controversial
49% aged 4564years, and 40%65years [4]. Thus, the topic a detailed discussion is given in Chap. 15.
average age of onset of familial NA-AION is younger The clinical findings in NA-AION associated with
than that seen in classical non-familial NA-AION. phosphodiesterase type 5 (PDE5) inhibitors use are no dif-
ferent from those seen in classical NA-AION [115, 116].
Bilaterality of NA-AION
In all the three patients of Berggren etal. [110] both eyes Amiodarone and NA-AION
were involved. Both twins seen by Manor [111] had
bilateral NA-AION. Wang etal. [112] reported bilateral
NA-AION in 67% of their nine patients. In our series Amiodarone is used for cardiac arrhythmias. A large
[114], eight of the 17 (47%) patients with familial number of reports dealing with ocular side-effects of
NA-AION had bilateral involvement. Bilateral involve- amiodarone have been published, one of which has been
ment in classical non-familial NA-AION, by contrast, is described as amiodarone-related optic neuropathy
much lower (25% within 3years [75], and 19% [76] and since 1987 [117121]. Most patients using amiodarone
14.7% [77] within 5years). The time interval between do not develop visual symptoms. I have seen many
the involvement of two eyes, in our series [114], when patients on amiodarone with and without ophthalmic
known (in five patients), varied from 5days to 15years. problems. Amiodarone keratopathy is a well-known
complication of amiodarone which can cause visual
halos, glare or mildly blurred vision. My experience of
Ophthalmic Clinical Findings dealing with patients who developed optic neuropathy
while on amiodarone indicates that so-called amio-
These findings showed that all patients with familial darone induced optic neuropathy is simply NA-AION,
NA-AION, like the classical non-familial NA-AION, unrelated to amiodarone; the reasons for this conclusion
had the typical mode of onset of visual loss symptoms are discussed elsewhere [122] (see Chap. 15). Most
[7], visual field defects [17], and others signs and clini- importantly, the clinical features of the optic neuropathy
cal findings of classical non-familial NA-AION dis- in patients taking amiodarone are typical of NA-AION.
cussed above. Some of them initially develop asymptomatic optic disc
edema which may later progress to visual loss but not On examination in my Clinic that day, the visual acu-
always, as is the case with incipient NA-AION [60]. In ity was bare light perception without projection in the left
the multifactorial scenario of NA-AION, it is the sys- eye and 20/20 in the right eye. On plotting the visual
temic cardiovascular risk factors rather than amiodarone fields with a Goldmann perimeter, in the left eye a small
that cause NA-AION. Some ophthalmologists have island of field was seen in the inferior temporal periph-
advocated discontinuation of amiodarone use in patients eral region with V-4e isopter only, and the visual field in
with cardiac arrhythmias when optic neuropathy devel- the right eye was perfectly normal. The intraocular pres-
ops, under the misconception that amiodarone causes sure was 15 and 10mmHg in the right and left eye respec-
optic neuropathy. However, since there is no cause-and- tively. The left fundus showed the optic disc markedly
effect relationship between optic neuropathy and amio- swollen and hyperemic, with no other abnormality
darone, and amiodarone is a well-established treatment (Fig.16.31a). Right fundus was normal except for a small
for severe cardiac arrhythmias, there is no justification cotton-wool spot about a disc diameter superotemporal
for that recommendation.
a
Posthemorrhagic Amaurosis
In this clinical entity, there is a visual loss following

distant recurrent or massive systemic hemorrhages,
from a variety of medical conditions, e.g., uterine
bleeding, gastro-intestinal bleeding, hemoptysis,
epistaxis or repeated venesection, often recurrent in
nature. Its pathogenesis is discussed elsewhere [123]
and in Chap. 15.
During recent years, a consensus has emerged that
it is due to ischemia of the optic nerve head; that is, it
represents NA-AION. The following three examples
demonstrate that very well.
Case 1
b
A 56-year-old white man developed perforation of a
peptic ulcer and was admitted to hospital with blood
pressure of 90/60mmHg and hemoglobin 6.4g/dL. He
required 15 units of blood transfusion preoperatively
and following that his blood pressure was 130/70mmHg
and hemoglobin 16.3g/dL. He had partial gastrectomy
performed next day. Five day after that, his hemoglo-
bin was 12.9 g/dL and he was maintaining a blood
pressure of 110130mmHg systolic and 7080mmHg
diastolic. His 4 a.m. blood pressure every day was low-
est, being about 110/65 mmHg. Four days later (i.e.,
10days after the initial visit), he woke up with a shadow
in the lower half of the visual field of the left eye. At 4
a.m. and 8 a.m., while in the hospital, his blood pres-
sure was 110120/6070mmHg. Two days after that,
Fig.16.31 Fundus photographs of case 1. (a) On the day he lost
he woke up with a marked loss of vision. At 4 a.m. that vision in the left eye, there is marked optic disc edema. (b) Five
morning his blood pressure was 110/60mmHg. weeks after that, it shows optic atrophy
to the disc. Hematologic examination that day revealed a

hemoglobin of 12.6 g/dL, hematocrit 38%, platelets
444,000/cm, with no other abnormality. The patient was
followed by me periodically as an outpatient. Five weeks
later the left disc became atrophic (Fig.16.31b). When
last seen 2years later, the right eye was normal and the
left eye had bare light perception with optic atrophy.
Case 2
A 59-year-old white man, with a history of recurrent

bleeding peptic ulcer for 35years, one day developed
black stools during the night and woke up with blurred
vision in the superior nasal quadrant of the left eye. His
blood pressure that day was 128/82mmHg. Next day, b
he noticed further worsening of the visual field defect
in the left eye. He was admitted to his local hospital
that day for management of recently developed black
stools, blood loss and anemia. His blood pressure was
128/78mmHg at the time of admission. He had hemo-
globin of 7.7g/dL which within a few hours dropped
to 6.6g/dL. He was transfused with 4 units of packed
red blood cells which brought his hemoglobin to
11.0g/dL soon after that and 13.2g/dL 2days later.
Frequent monitoring of his blood pressure showed that
while during the day it was about 130/70mmHg, dur-
ing the night it dropped to a fairly low level, being
94/60mmHg at 34 a.m. He had been discovered to be
Fig.16.32 Fundus photographs of case 2. (a) Three days after
diabetic about 6months earlier and was well controlled the loss of vision in the left eye, shows marked optic disc edema
with oral antidiabetic therapy. and a cottonwool spot (arrow). (b) Five weeks after that, shows
He was seen by his local ophthalmologist the day optic atrophy
after the onset of blurred vision in his left eye, when the
visual acuity was 20/30 in the left eye with optic disc The patient continued to have progressive deterio-
edema while the right eye showed no abnormality. He ration of vision and visual fields in both eyes. The left
was seen in my clinic 3days after that. At that time, the eye, 2 weeks after the initial visit showed markedly
visual acuity was 20/20 in both eyes. On plotting the constricted visual fields so that with I-2e only paracen-
visual fields with a Goldmann perimeter, the right eye tral 6 field was present and with V-4e only 25 field
showed no visual field defect but the left eye showed a was present, with a visual acuity of 20/60. In another
markedly constricted field with loss of nasal half of the week it deteriorated to 20/200 and only 10 field tem-
field with all the isopters. The right fundus showed mod- poral to fixation was recorded with V-4e only. One
erate optic disc edema, involving 2/3 of the disc in the month later, the visual acuity in the left eye was 20/300
superior nasal area, with no other abnormality. The left and 5 temporal island field with V-4e only. Two weeks
fundus showed marked optic disc edema with a cotton after that, the left disc showed generalized marked pal-
wool spot (arrow) (Fig. 16.32a). Fluorescein fundus lor with no edema (Fig.16.32b).
angiography of the left eye revealed delayed and patchy The right eye, 1 week after the initial visit to my
choroidal and optic disc filling, with late staining of both clinic, developed a temporal defect with I-2e and I-4e
optic discs involving the entire disc in the left eye and although the visual acuity had remained 20/20 cen-
mostly 3/4 of the right disc in the superonasal region. trally. Two weeks later the right disc looked almost
normal. Eleven weeks after his initial visit, the visual a

acuity was 20/20 in the right and 20/400 in the left,
with only a tiny island of visual field in the inferior
temporal region in the left eye and stable visual field in
the right eye. Fundus examination revealed optic atro-
phy in the left eye and normal-looking disc in the right.
Detailed neurologic, systemic and cardiovascular eval-
uation revealed no other abnormality.
At last examination 15months after the initial visit,
the visual acuity in the left eye had deteriorated to hand
motion only and no visual field could be plotted. The
patient was still having recurrent gastrointestinal
bleeding; the last episode had occurred a month before.
The right eye showed no change.
Case 3
A 57-year-old white man, at 3 a.m. started to have

black stools. He was admitted to the hospital at 11 a.m. b
next day. At the time of admission, his blood pressure
was 124/64mmHg and his hemoglobin level 8g/dL.
He had 700mL of blood transfusion which raised the
hemoglobin that day to 11.1 g/dL, hematocrit 33%,
and platelet count 323,000/mm3. He was found to have
multiple duodenal ulcers. He was discharged from the
hospital after 4days. Two days after that, in the morn-
ing when he woke up he noticed blurred vision in the
inferior part of the visual field of the left eye. It gradu-
ally progressed over the following days to involve the
superior field, so that 4days later the vision was blurred
all over the left eye. He saw his local ophthalmologist
that day who found his visual acuity in the left eye only
count fingers, marked relative afferent pupillary defect,
a temporal island of visual field, and marked optic disc
edema in that eye; he gave a retrobulbar injection of
depot corticosteroids but that did not improve the
vision.
He was referred to my clinic 6days after that. At
that time his visual acuity was counting fingers at about Fig.16.33 Fundus photographs of case 3. (a) Ten days after the
6ft eccentrically in a temporal island. Visual field plot- onset of visual loss in the left eye, shows marked optic disc
edema. (b) Five weeks after that, shows optic atrophy
ting with a Goldmann perimeter showed the loss of
central and entire nasal half field of vision; the patient
could see I-4e in the temporal field, from the blind spot revealed sluggish choroidal circulation and late disc
temporally in the left eye, and normal fields in the right staining. That day his hemoglobin was 13.0 g/dL,
eye. The intraocular pressure was 20 and 19mmHg in hematocrit 39% and platelet count 429,000/mm3. He
the right and left eye respectively. The left fundus was found to have no other hematologic or cardiovas-
showed optic disc edema (Fig.16.33a), and the right cular abnormality. He was seen again 5 weeks later,
fundus was normal. Fluorescein fundus angiography when there was no change in his visual status. The left
optic disc had developed generalized pallor with no 6. That all eyes with NA-AION initially have pale
edema (Fig. 16.33b). When last seen 3 months after optic disc edema.
that, the left eye showed no further change and the As discussed above, disc pallor actually starts to
right eye was normal. develop only 23weeks after the onset of visual
Conclusion: These examples show that these cases loss; before that there is no pale optic disc
have the classical presentation, findings and course edema.
seen in NA-AION, i.e., discovering visual loss on wak- 7. That inferior altitudinal defect is the classical
ing up in the morning, initially optic disc edema and diagnostic visual field defect in NA-AION.
optic nerve related visual field defects, and optic disc As discussed above, a study [17] of 312 NA-AION
edema resolving in about 2months followed by optic eyes showed that inferior nasal field defect is the
atrophy. As discussed in Chap. 15, there is a time lag most common defect.
between the development of initial hemorrhages accom- 8. That all eyes with NA-AION have poor visual
panied by low hemoglobin, hematocrit and blood pres- acuity at onset.
sure and the onset of NA-AION in this clinical entity. In a combined cohort of 500 consecutive NA-AION
eyes in our two studies [22, 24], when patients
were seen within 2weeks after the onset of visual
loss, initial visual acuity was 20/2020/25 in 33%,
Misconceptions About NA-AION better than 20/40 in 51%. Preechawat etal. [6] in
a study of 727 NA-AION consecutive patients,
The subject of NA-AION is plagued with multiple mis- found that visual acuity was 20/40 or better in
conceptions, resulting in controversy, confusion and 59%,
misdiagnosis. Following are the major misconceptions. 9. That steroid therapy has no role in the manage-
ment of NA-AION.
1. That NA-AION and cerebral stroke are similar in In our study [22] (as discussed in Chap. 17) of 696
nature. NA-AION eyes (364 treated versus 332 controls) the
As discussed in Chap. 14, cerebral stroke is a throm- treated group showed significantly more visual acu-
boembolic disorder whereas NA-AION is primarily ity improvement than the control group (70% versus
a hypotensive disorder. 41% in eyes seen within 2 weeks of onset of
2. That NA-AION is due to occlusion of the PCA. NA-AION).
As discussed in Chap. 14, there is no actual occlu- 10. That smoking is a risk factor for development of
sion of the PCA in NA-AION but only transient NA-AION.
hypoperfusion or non-perfusion of the vessels in the Two large prospective studies have shown that this
optic nerve head in the vast majority of these eyes. is not true [4, 77].
3. That absence of optic disc cup is the main cause of 11. That aspirin reduces the risk of second eye
development of NA-AION. involvement by NA-AION.
As discussed in Chap. 14, an absent or small cup is sim- Two large studies have disproved this belief [76,
ply a secondary contributing factor, once the process of 77].
NA-AION has started, and not a primary factor. 12. That all patients with NA-AION should be inves-
4. That there is no spontaneous visual improvement tigated for thrombophilia.
in NA-AION. As discussed in Chap. 14, NA-AION is not a
Two large prospective natural history studies have thromboembolic disorder in the vast majority of
shown that visual acuity improves spontaneously in cases. Therefore, there is little justification to inves-
41%43% of the eyes [23, 24]. tigate all NA-AION patients for thrombophilia.
5. That NA-AION is not seen in young persons. 13. That the pathogenesis of NA-AION is not
As discussed above, studies [3, 4, 6] based on large known.
cohorts of NA-AION patients have disproved this As discussed in Chap. 14, the pathogenesis of
myth. NA-AION is complex but is not unknown.
References 385
References 17. Hayreh SS, Zimmerman B. Visual field abnormalities in

nonarteritic anterior ischemic optic neuropathy: their pattern
and prevalence at initial examination. Arch Ophthalmol.
1. Johnson LN, Arnold AC. Incidence of nonarteritic and 2005;123(11):155462.
arteritic anterior ischemic optic neuropathy. Population- 18. Bull GM. Meteorological correlates with myocardial and
based study in the state of Missouri and Los Angeles cerebral infarction and respiratory disease. Br J Prev Soc
County, California. J Neuroophthalmol. 1994;14(1):3844. Med. 1973;27(2):10813.
2. Hattenhauer MG, Leavitt JA, Hodge DO, Grill R, Gray DT. 19. Bull GM, Morton J. Environment, temperature and death
Incidence of nonarteritic anterior ischemic optic neuropa- rates. Age Ageing. 1978;7(4):21024.
thy. Am J Ophthalmol. 1997;123(1):1037. 20. Clark CV. Seasonal variation in incidence of brachial and
3. Hayreh SS, Joos KM, Podhajsky PA, Long CR. Systemic femoral emboli. Br Med J (Clin Res Ed). 1983;
diseases associated with nonarteritic anterior ischemic 287(6399):1109.
optic neuropathy. Am J Ophthalmol. 1994;118(6):76680. 21. Hayreh SS. Non-arteritic anterior ischemic optic neuropathy
4. Hayreh SS, Jonas JB, Zimmerman MB. Nonarteritic ante- and thrombophilia. Graefes Arch Clin Exp Ophthalmol.
rior ischemic optic neuropathy and tobacco smoking. 2009;247(5):57781.
Ophthalmology. 2007;114(4):8049. 22. Hayreh SS, Zimmerman MB. Non-arteritic anterior isch-
5. Ischemic Optic Neuropathy Decompression Trial Study emic optic neuropathy: role of systemic corticosteroid ther-
Group. Characteristics of patients with nonarteritic anterior apy. Graefes Arch Clin Exp Ophthalmol. 2008;
ischemic optic neuropathy eligible for the Ischemic Optic 246(7):102946.
Neuropathy Decompression Trial. Arch Ophthalmol. 23. Ischemic Optic Neuropathy Decompression Trial Research
1996;114(11):136674. Group. Optic nerve decompression surgery for nonarteritic
6. Preechawat P, Bruce BB, Newman NJ, Biousse V. Anterior anterior ischemic optic neuropathy (NAION) is not effective
ischemic optic neuropathy in patients younger than and may be harmful. JAMA. 1995;273(8):62532.
50years. Am J Ophthalmol. 2007;144(6):95360. 24. Hayreh SS, Zimmerman MB. Nonarteritic anterior ischemic
7. Hayreh SS, Podhajsky PA, Zimmerman B. Nonarteritic optic neuropathy: natural history of visual outcome.
anterior ischemic optic neuropathy: time of onset of visual Ophthalmology. 2008;115(2):298305.
loss. Am J Ophthalmol. 1997;124(5):6417. 25. Katz B, Spencer WH. Hyperopia as a risk factor for nonar-
8. Landau K, Winterkorn JM, Mailloux LU, Vetter W, teritic anterior ischemic optic neuropathy. Am J Ophthalmol.
Napolitano B. 24-hour blood pressure monitoring in 1993;116(6):7548.
patients with anterior ischemic optic neuropathy. Arch 26. Philips B, Dralands L, Missotten L. Non-arteritic anterior
Ophthalmol. 1996;114(5):5705. ischemic optic neuropathy and refraction. Bull Soc Belge
9. Hayreh SS, Zimmerman MB, Podhajsky P, Alward WL. Ophtalmol. 1995;259:1837.
Nocturnal arterial hypotension and its role in optic nerve 27. Mansour AM, Shoch D, Logani S. Optic disk size in isch-
head and ocular ischemic disorders. Am J Ophthalmol. emic optic neuropathy. Am J Ophthalmol. 1988;106(5):
1994;117(5):60324. 5879.
10. Hayreh SS. Acute ischemic disorders of the optic nerve: 28. Cullen JF. Ischaemic optic neuropathy. Trans Ophthalmol
pathogenesis, clinical manifestations and management. Soc UK. 1967;87:75974.
Ophthalmol Clin North Am. 1996;9(3):40742. 29. Ellenberger Jr C, Netsky MG. Infarction in the optic nerve. J
11. Hayreh SS. Anterior ischaemic optic neuropathy. I. Neurol Neurosurg Psychiatry. 1968;31(6):60611.
Terminology and pathogenesis. Br J Ophthalmol. 1974; 30. Ellenberger Jr C, Keltner JL, Burde RM. Acute optic neu-
58(12):95563. ropathy in older patients. Arch Neurol. 1973;28(3):
12. Hayreh SS. Pathogenesis of cupping of the optic disc. Br J 1825.
Ophthalmol. 1974;58(10):86376. 31. Eagling EM, Sanders MD, Miller SJ. Ischaemic papillopa-
13. Beck RW, Servais GE, Hayreh SS. Anterior ischemic optic thy. Clinical and fluorescein aniographic review of forty
neuropathy. IX. Cup-to-disc ratio and its role in pathogen- cases. Br J Ophthalmol. 1974;58(12):9901008.
esis. Ophthalmology. 1987;94(11):15038. 32. Boghen DR, Glaser JS. Ischaemic optic neuropathy. The
14. Hayreh SS, Zimmerman MB. Nonarteritic anterior isch- clinical profile and history. Brain. 1975;98(4):689708.
emic optic neuropathy: refractive error and its relationship 33. Aulhorn E, Tanzi IM. Comparison of visual field defects in
to cup/disc ratio. Ophthalmology. 2008;115(12):227581. glaucoma and in acute anterior ischemic optic neuropathy.
14a. Hayreh SS. Inter-individual variation in blood supply of Doc Ophthalmol Proc Ser. 1978;19:739.
the optic nerve head. Its importance in various ischemic 34. Bronner A, Gerhard JP, Risse JF, Bigerel A. La neuropathie
disorders of the optic nerve head, and glaucoma, low-ten- optique ischemique anterieure aigue. Etude clinique,
sion glaucoma and allied disorders. Doc Ophthalmol. perimetrique et angiographique de 50 observations. [Acute
1985;59(3):21746. anterior ischaemic optic neuritis. Clinical, visual field and
15. Hayreh SS, Piegors DJ, Heistad DD. Serotonin-induced angiographic study of 50 cases]. Rev Otoneuroophtalmol.
constriction of ocular arteries in atherosclerotic monkeys. 1979;51(5):47984.
Implications for ischemic disorders of the retina and optic 35. Reuscher A, Chromek W, Kommerell G. Gesichtsfeldausfalle
nerve head. Arch Ophthalmol. 1997;115(2):2208. und Fluoreszenzangiographie bei der Apoplexia papillae.
16. Hayreh SS, Podhajsky P. Visual field defects in anterior [Visual field defects and fluorescein angiography in acute
ischemic optic neuropathy. Doc Ophthalmol Proc Ser. anterior ischemic optic neuropathy]. Klin Monbl
1979;19:5371. Augenheilkd. 1978;173(1):6974.
36. Repka MX, Savino PJ, Schatz NJ, Sergott RC. Clinical pro- 55. Hayreh SS, Jonas JB. Optic disc morphology after arteritic
file and long-term implications of anterior ischemic optic anterior ischemic optic neuropathy. Ophthalmology.
neuropathy. Am J Ophthalmol. 1983;96(4):47883. 2001;108(9):158694.
37. Rizzo III JF, Lessell S. Optic neuritis and ischemic optic 56. Hayreh SS. Anterior ischemic optic neuropathy. V. Optic disc
neuropathy. Overlapping clinical profiles. Arch Ophthalmol. edema an early sign. Arch Ophthalmol. 1981;99(6):103040.
1991;109(12):166872. 57. Hayreh SS. Pathogenesis of Oedema of the Optic Disc
38. Gerling J, Meyer JH, Kommerell G. Visual field defects in (Papilloedema) [thesis for PhD]. London, England: London
optic neuritis and anterior ischemic optic neuropathy: dis- University; 1965.
tinctive features. Graefes Arch Clin Exp Ophthalmol. 58. Hayreh SS. Pathogenesis of oedema of the optic disc. Doc
1998;236(3):18892. Ophthalmol. 1968;24(2):289411.
39. Traustason OI, Feldon SE, Leemaster JE, Weiner JM. 59. Arnold AC, Hepler RS. Fluorescein angiography in acute
Anterior ischemic optic neuropathy: classification of field nonarteritic anterior ischemic optic neuropathy. Am J
defects by Octopus automated static perimetry. Graefes Arch Ophthalmol. 1994;117(2):22230.
Clin Exp Ophthalmol. 1988;226(3):20612. 60. Hayreh SS, Zimmerman MB. Incipient nonarteritic anterior
40. Arnold AC, Hepler RS. Natural history of nonarteritic ante- ischemic optic neuropathy. Ophthalmology. 2007;114(9):
rior ischemic optic neuropathy. J Neuroophthalmol. 1994; 176372.
14(2):669. 61. Hayreh SS, Zimmerman MB. Nonarteritic anterior ischemic
41. Boone MI, Massry GG, Frankel RA, Holds JB, Chung SM. optic neuropathy: clinical characteristics in diabetic patients
Visual outcome in bilateral nonarteritic anterior ischemic versus nondiabetic patients. Ophthalmology. 2008;115(10):
optic neuropathy. Ophthalmology. 1996;103(8):12238. 181825.
42. WuDunn D, Zimmerman K, Sadun AA, Feldon SE. 62. Hayreh SS. Anterior ischaemic optic neuropathy. III.
Comparison of visual function in fellow eyes after bilateral Treatment, prophylaxis, and differential diagnosis. Br J
nonarteritic anterior ischemic optic neuropathy. Ophthalmology. Ophthalmol. 1974;58(12):9819.
1997;104(1):10411. 63. Dai J, Buijs R, Swaab D. Glucocorticoid hormone (cortisol)
43. Hayreh SS, Zimmerman B, Kardon RH. Visual improve- affects axonal transport in human cortex neurons but shows
ment with corticosteroid therapy in giant cell arteritis. Report resistance in Alzheimers disease. Br J Pharmacol. 2004;
of a large study and review of literature. Acta Ophthalmol 143(5):60610.
Scand. 2002;80(4):35567. 64. Foulds WS. Ischaemic optic neuropathy. In: Cant JS, ed..
44. Hayreh SS, Zimmerman MB. Central retinal artery occlusion: Proceedings, William Mackenzie Centenary Symposium on
visual outcome. Am J Ophthalmol. 2005;140(3):37691. the Ocular Circulation in Health and Disease (1968 : Royal
45. Sergott RC, Cohen MS, Bosley TM, Savino PJ. Optic nerve College of Physicians and Surgeons, Glasgow). London:
decompression may improve the progressive form of nonar- Kimpton; 1968:136141.
teritic ischemic optic neuropathy. Arch Ophthalmol. 65. Hayreh SS. Optic disc vasculitis. Br J Ophthalmol.
1989;107(12):174354. 1972;56(9):65270.
46. Hayreh SS, Zimmerman MB. Optic disc edema in non-arter- 66. Jonas JB, Sofker A. Intraocular injection of crystalline corti-
itic anterior ischemic optic neuropathy. Graefes Arch Clin sone as adjunctive treatment of diabetic macular edema. Am
Exp Ophthalmol. 2007;245(8):110721. J Ophthalmol. 2001;132(3):4257.
47. Hayreh SS. Anterior ischaemic optic neuropathy. II. Fundus 67. Greenberg PB, Martidis A, Rogers AH, Duker JS, Reichel E.
on ophthalmoscopy and fluorescein angiography. Br J Intravitreal triamcinolone acetonide for macular oedema due
Ophthalmol. 1974;58(12):96480. to central retinal vein occlusion. Br J Ophthalmol.
48. Hayreh SS. Anterior ischemic optic neuropathy. New York: 2002;86(2):2478.
Springer; 1975. 68. Jonas JB, Kreissig I, Degenring RF. Intravitreal triamcino-
49. Hayreh SS. Anterior ischemic optic neuropathy. Arch lone acetonide as treatment of macular edema in central reti-
Neurol. 1981;38(11):6758. nal vein occlusion. Graefes Arch Clin Exp Ophthalmol.
50. Hayreh SS. Blood supply of the optic nerve head and its role 2002;240(9):7823.
in optic atrophy, glaucoma, and oedema of the optic disc. Br 69. Martidis A, Duker JS, Greenberg PB, Rogers AH, Puliafito
J Ophthalmol. 1969;53(11):72148. CA, Reichel E, etal. Intravitreal triamcinolone for refractory
51. Hayreh SS, Weingeist TA. Experimental occlusion of the diabetic macular edema. Ophthalmology. 2002;109(5):
central artery of the retina. I. Ophthalmoscopic and fluores- 9207.
cein fundus angiographic studies. Br J Ophthalmol. 70. Arnold AC, Badr MA, Hepler RS. Fluorescein angiography
1980;64(12):896912. in nonischemic optic disc edema. Arch Ophthalmol. 1996;
52. Hayreh SS. The blood supply of the optic nerve head and the 114(3):2938.
evaluation of it: myth and reality. Prog Retin Eye Res. 71. Hayreh SS. Optic disc edema in raised intracranial pressure.
2001;20(5):56393. V. Pathogenesis. Arch Ophthalmol. 1977;95(9):155365.
53. Hayreh SS, Podhajsky PA, Zimmerman B. Ocular manifes- 72. Hayreh SS, Zahoruk RM. Anterior ischemic optic neuropa-
tations of giant cell arteritis. Am J Ophthalmol. 1998; thy. VI. In juvenile diabetics. Ophthalmologica. 1981;
125(4):50920. 182(1):1328.
54. Danesh-Meyer H, Savino PJ, Spaeth GL, Gamble GD. 73. Tomsak RL, Zakov ZN. Nonarteritic anterior ischemic optic
Comparison of arteritis and nonarteritic anterior ischemic neuropathy with macular edema: visual improvement and
optic neuropathies with the Heidelberg retina tomograph. fluorescein angiographic characteristics. J Neuroophthalmol.
Ophthalmology. 2005;112(6):110412. 1998;18(3):1668.
References 387
74. Hedges III TR, Vuong LN, Gonzalez-Garcia AO, Mendoza- 93. Regillo CD, Brown GC, Savino PJ, Byrnes GA, Benson WE,
Santiesteban CE, Amaro-Quierza ML. Subretinal fluid from Tasman WS, etal. Diabetic papillopathy: patient characteristics
anterior ischemic optic neuropathy demonstrated by optical and fundus findings. Arch Ophthalmol. 1995;113(7):88995.
coherence tomography. Arch Ophthalmol. 2008;126(6): 94. Inoue M, Tsukahara Y. Vascular optic neuropathy in diabe-
81215. tes mellitus. Jpn J Ophthalmol. 1997;41(5):32831.
75. Beri M, Klugman MR, Kohler JA, Hayreh SS. Anterior isch- 95. Ignat F, Barascu D, Perovic I, Munteanu A. Leziuni ale ner-
emic optic neuropathy. VII. Incidence of bilaterality and vului optic in diabet. [Optic nerve lesions in diabetes mel-
various influencing factors. Ophthalmology. 1987;94(8): litus]. Oftalmologia. 2002;54(3):3943.
10208. 96. Vaphiades MS. The disk edema dilemma. Surv Ophthalmol.
76. Beck RW, Hayreh SS, Podhajsky PA, Tan ES, Moke PS. 2002;47(2):1838.
Aspirin therapy in nonarteritic anterior ischemic optic neu- 97. Hayreh SS. Diabetic papillopathy and nonarteritic anterior
ropathy. Am J Ophthalmol. 1997;123(2):21217. ischemic optic neuropathy. Surv Ophthalmol. 2002;
77. Newman NJ, Scherer R, Langenberg P, Kelman S, Feldon S, 47(6):6002.
Kaufman D, etal. The fellow eye in NAION: report from the 98. Georgiads G, Konstas P, Stangos N. Rflexions issues de
ischemic optic neuropathy decompression trial follow-up ltude de nombreux cas de pseudo-papillite vasculaire
study. Am J Ophthalmol. 2002;134(3):31728. [Study of numerous cases of vascular pseudopapillitis].
78. Hayreh SS, Podhajsky PA, Zimmerman B. Ipsilateral recur- Bull Mm Soc Fr Ophtalmol. 1966;79:50639.
rence of nonarteritic anterior ischemic optic neuropathy. Am 99. Sanders MD. Ischaemic papillopathy. Trans Ophthalmol
J Ophthalmol. 2001;132(5):73442. Soc UK. 1971;91:36986.
79. Frampton P, Da Rin D, Brown B. Diurnal variation of 100. Oh KT, Oh DM, Hayreh SS. Optic disc vasculitis. Graefes
intraocular pressure and the overriding effects of sleep. Am Arch Clin Exp Ophthalmol. 2000;238(8):64758.
J Optom Physiol Opt. 1987;64(1):5461. 101. Hayreh SS. Hypertensive retinopathy. Ophthalmol Clin
80. Brown B, Morris P, Muller C, Brady A, Swann PG. North Am. 1998;11(4):53558.
Fluctuations in intra-ocular pressure with sleep: I. Time 102. Tso MOM, Hayreh SS. Optic disc edema in raised intracra-
course of IOP increase after the onset of sleep. Ophthalmic nial pressure. IV. Axoplasmic transport in experimental
Physiol Opt. 1988;8(3):2468. papilledema. Arch Ophthalmol. 1977;95(8):145862.
81. Wildsoet C, Eyeson-Annan M, Brown B, Swann PG, 103. Hayreh SS. Fluids in the anterior part of the optic nerve in
Fletcher T. Investigation of parameters influencing intraocu- health and disease. Surv Ophthalmol. 1978;23(1):125.
lar pressure increases during sleep. Ophthalmic Physiol Opt. 104. McLeod D, Oji EO, Kohner EM, Marshall J. Fundus signs
1993;13(4):35765. in temporal arteritis. Br J Ophthalmol. 1978;62(9):5914.
82. Buguet A, Py P, Romanet JP. 24-hour (nyctohemeral) and 105. Fink BR, Byers MR, Middaugh ME. Dynamics of colchi-
sleep-related variations of intraocular pressure in healthy cine effects on rapid axonal transport and axonal morphol-
white individuals. Am J Ophthalmol. 1994;117(3): ogy. Brain Res. 1973;56:299311.
3427. 106. Almog Y, Goldstein M. Visual outcome in eyes with
83. Hayreh SS, Podhajsky P, Zimmerman MB. Role of noctur- asymptomatic optic disc edema. J Neuroophthalmol.
nal arterial hypotension in optic nerve head ischemic disor- 2003;23(3):2047.
ders. Ophthalmologica. 1999;213(2):7696. 107. Hayreh SS. Factors influencing blood flow in the optic
84. Hayreh SS. Retinal and optic nerve head ischemic disorders nerve head. J Glaucoma. 1997;6(6):41225: erratum in J
and atherosclerosis: the role of serotonin. Prog Retin Eye Glaucoma 1998;7(1):71.
Res. 1999;18(2):191221. 108. Hayreh SS. Blood flow in the optic nerve head and factors
85. Beck RW, Savino PJ, Schatz NJ, Smith CH, Sergott R. that may influence it. Prog Retin Eye Res. 2001;
Anterior ischaemic optic neuropathy: recurrent episodes in 20(5):595624.
the same eye. Br J Ophthalmol. 1983;67(10):7059. 109. Hayreh SS. Management of non-arteritic anterior ischemic
86. Hamed LM, Purvin V, Rosenberg M. Recurrent anterior optic neuropathy. Graefes Arch Clin Exp Ophthalmol.
ischemic optic neuropathy in young adults. J Clin 2009;247(12):1595600.
Neuroophthalmol. 1988;8(4):23948. 110. Berggren L, Thorburn W, Fodstad H. Three cases of non-
87. Hayreh SS, Jonas JB. Optic disc morphology after AAION: inflammatory ischemic optic neuropathy occurring in the same
author reply. Ophthalmology. 2002;109(7):1199201. family. Acta Ophthalmol (Copenh). 1974;52(5):58995.
88. Lubow M, Makley Jr TA. Pseudopapilledema of juvenile 111. Manor RS. Nonarteritic ischemic optic neuropathy in iden-
diabetes mellitus. Arch Ophthalmol. 1971;85(4):41722. tical female twins: improvement of visual outcome in one
89. Appen RE, Chandra SR, Klein R, Myers FL. Diabetic papil- by optic nerve decompression. Arch Ophthalmol.
lopathy. Am J Ophthalmol. 1980;90(2):2039. 1990;108(8):10678.
90. Barr CC, Glaser JS, Blankenship G. Acute disc swelling in 112. Wang MY, Sadun F, Levin LB, LaBree L, Feldon SE.
juvenile diabetes. Clinical profile and natural history of 12 Occurrence of familial nonarteritic anterior ischemic optic
cases. Arch Ophthalmol. 1980;98(12):218592. neuropathy in a case series. J Neuroophthalmol. 1999;
91. Pavan PR, Aiello LM, Wafai MZ, Briones JC, Sebestyen JG, 19(2):1447.
Bradbury MJ. Optic disc edema in juvenile-onset diabetes. 113. Fingert JH, Grassi MA, Janutka JC, East JS, Howard JG,
Arch Ophthalmol. 1980;98(12):21935. Sheffield VC, etal. Mitochondrial variant G4132A is asso-
92. Algan M, Ziegler O, Drouin P. La neuropathie optique du ciated with familial non-arteritic anterior ischemic optic
sujet diabetique. [Optic neuropathy in diabetic subjects]. neuropathy in one large pedigree. Ophthalmic Genet.
Diabte Mtab. 1993;19(5):3959. 2007;28(1):17.
114. Hayreh SS, Fingert JH, Stone E, Jacobson DM. Familial 121. Mindel JS, Anderson J, Hellkamp A, Johnson G, Poole JE,
non-arteritic anterior ischemic optic neuropathy. Graefes Mark DB, etal. Absence of bilateral vision loss from amio-
Arch Clin Exp Ophthalmol. 2008;246(9):1295305. darone: a randomized trial. Am Heart J. 2007;153(5):837
115. Hayreh SS. Erectile dysfunction drugs and non-arteritic 42; erratum in Am Heart J. 2007 Jul;2154(2001):
anterior ischemic optic neuropathy: is there a cause and 2053.
effect relationship? J Neuroophthalmol. 2005;25(4):2958. 122. Hayreh SS. Amiodarone, erectile dysfunction drugs, and
116. Hayreh SS. Non-arteritic anterior ischaemic optic neuropa- non-arteritic ischemic optic neuropathy. J Neuroophthalmol.
thy and phosphodiesterase-5 inhibitors. Br J Ophthalmol. 2006;26(2):1545.
2008;92(12):157780. 123. Hayreh SS. Anterior ischemic optic neuropathy. VIII.
117. Gittinger Jr JW, Asdourian GK. Papillopathy caused by Clinical features and pathogenesis of post-hemorrhagic
amiodarone. Arch Ophthalmol. 1987;105(3):34951. amaurosis. Ophthalmology. 1987;94(11):1488502.
118. Feiner LA, Younge BR, Kazmier FJ, Stricker BH, 124. Hayreh SS. Anatomy and physiology of the optic nerve
Fraunfelder FT. Optic neuropathy and amiodarone therapy. head. Trans Am Acad Ophthalmol Otolaryngol.
Mayo Clin Proc. 1987;62(8):70217. 1974;78:OP24054.
119. Turdumambetova G, Bredehorn T, Duncker GI. Okulare 125. Hayreh SS. Ischemic optic neuropathy. Prog Retin Eye
Nebenwirkungen der Amiodaron-Therapie. [Ocular side- Res. 2009;28:3462.
effects associated with amiodarone therapy]. Klin Monbl
Augenheilkd. 2005;222(6):48592.
120. Murphy MA, Murphy JF. Amiodarone and optic neuropa- For a complete bibliography and detailed review of previously
thy: the heart of the matter. J Neuroophthalmol. 2005; published studies in the literature on the subject, please refer to
25(3):2326. bibliography in previous publications listed here.
Management of Non-arteritic Anterior
Non-arteritic anterior ischemic optic neuropathy instructive and educational for a proper perspective on
(NA-AION) is a major cause of seriously impaired many such treatments, which gain wide popularity but are
vision among the middle-aged and elderly population; finally found to be harmful. I am aware that some may
in addition, contrary to popular belief, it also occurs in disagree with my giving all this detailed historical infor-
young persons, though less commonly. Nevertheless, mation about a method of treatment which has been
almost all publications on NA-AION contain the proven to be not only of no benefit but harmful, and aban-
following two statements: doned. It is that very fact which makes it instructive to
consider this procedure in detail, to stress the point that a
1. The pathogenesis of NA-AION is unknown.
treatment without a scientific rationale, in spite of wide
2. There is no known treatment for NA-AION.
popularity, may finally prove harmful for the reasons dis-
Given the current state of our knowledge of NA-AION, cussed in this historical review.
both statements are wrong. The pathogenesis of Sergott etal. [2] first published an account of optic
NA-AION is discussed at length in Chap. 14, which nerve sheath decompression in NA-AION in 1989. They
concludes the pathogenesis of NA-AION is complex performed optic nerve sheath decompression in 14
but not, as often stated, unknown. NA-AION eyes with progressive visual loss and three
The management of NA-AION has been highly con- eyes with nonprogressive visual loss. They claimed that
troversial. Various treatments have been advocated Optic nerve sheath decompression surgery improved
from time to time [1], but were found to be useless and visual function for 12 of 14 patients with progressive
even, in some cases, harmful. To evaluate the validity of nonarteritic ischemic optic neuropathy (NAION). Visual
any treatment, there are two essential considerations: recovery was maintained in all patients during a follow-
up period of 618months (average, 11months). Seven
1. It must have a scientific rationale; any treatment
patients had experienced a previous NAION in the eye
without that, irrespective of how much enthusiasm
that was not operated on that did not improve spontane-
it generates, and how attractive it may seem, will
ously. Surprisingly, 2 of these 7 eyes with long-standing
not actually work, as was demonstrated by the saga
decreased vision demonstrated some visual improve-
of optic nerve sheath decompression in NA-AION
ment after surgery on the contralateral, acutely affected
(as discussed below).
eye. Spontaneous visual improvement did not occur in
2. The treatment of a disease must arise from a full
an age- and sex-matched control group of 12 patients
understanding of its pathogenesis. The following is
with similar entry-level visual acuity and field loss. Only
a discussion of the main treatments for NA-AION
1 of 3 patients with sudden, nonprogressive visual loss
which have been advocated during the recent past.
secondary to NAION improved after surgery. In a con-
trol group with nonprogressive NAION, 2 of 15 eyes (14
patients) demonstrated spontaneous improvement.
Optic Nerve Sheath Decompression They concluded: Optic nerve sheath decompression
improves visual loss due to progressive NAION, a disor-
I have decided to give a detailed historical review of this der without any previously effective therapy. Because
particular treatment method because its story is highly of this conclusion, the paper was published on an

390 17 Management of Non-arteritic Anterior Ischemic Optic Neuropathy
expedited basis in the Archives of Ophthalmology, so intracranial pressure is due to axoplasmic flow sta-
that this new means of treatment for NA-AION was sis [5]. Optic disc edema due to intracranial hyper-
immediately available to the ophthalmic community. tension, even of a very severe degree, does not by
On reading the paper, based on my studies dealing itself cause visual dysfunction. When visual dys-
with various basic aspects of NA-AION, I felt that function is seen in these cases, it is secondary to
optic nerve sheath decompression had no scientific ischemic changes [6]; an eye with optic disc edema
rationale in NA-AION and could be harmful, and I blacks out briefly when the blood pressure drops.
wrote a letter to that effect to the editor of the Archives This is further evidence that the mechanism is isch-
of Ophthalmology [3]. I reproduce here the whole emia. Axoplasmic flow plays no role in the conduc-
original text of that letter [3], because it deals with the tion of nerve impulses. Thus, there is not a shred of
reasons why I concluded that decompression had no evidence that axoplasmic flow stasis in optic disc
scientific rationale, and also with several other contro- edema per se causes any clinically detectable visual
versial aspects of NA-AION. loss. The visual loss in AlON is due to ischemia of
Following is the original text of that letter to the the nerve fibers and not to axoplasmic flow stasis.
Editor [3]: Can normal CSF pressure with normal intraocular
pressure produce axoplasmic flow stasis and optic disc
I was very interested to read the expedited publica- edema? We have no evidence for this. The authors
tion in the December 1989 issue of the ARCHIVES statement that drainage of the CSF might allow
by Sergott etal. [2] claiming that surgical optic nerve recovery of fast axoplasmic transport, with subsequent
sheath decompression improved visual function in improvement of vision has no scientific basis.
patients with progressive nonarteritic anterior isch- Axoplasmic flow stasis in NA-AION is due to isch-
emic optic neuropathy (NA-AION). They hypothe- emia and is in no way related to the CSF pressure.
sized that in NA-AION, progressive visual loss after Thorough neurologic investigation, including
the initial ischemic event could be due to interference CSF pressure measurement, of patients with
with rapid axoplasmic transport produced by CSF NA-AION reveals no such abnormality. While it is
[cerebrospinal fluid] pressure within the anatomi- true that swollen axons (due to axoplasmic flow sta-
cally restricted confines of the perineural optic nerve sis from any cause) in the restricted space in the
space. Drainage of the CSF might allow recovery of optic disc may develop secondary capillary vascu-
fast axoplasmic transport with subsequent improve- lar changes and ischemia, resulting in further isch-
ment in vision. This theory raises some very impor- emic damage, the important factor to combat is the
tant issues that need comment before they are primary cause of axoplasmic flow stasis, which in
accepted as scientifically valid concepts. NA-AION is ischemia and not CSF pressure. The
My comments are based on the following: (1) authors further added, If the initial event in pro-
my prospective studies on NA-AION in more than gressive NAION [NA-AION] is indeed secondary
600 patients; (2) my studies of the blood supply of to an ischemic, vaso-occlusive process, optic nerve
the optic nerve head for the past 35 years; (3) sheath decompression may be effective because it
experimental and clinical studies on the pathogen- decreases the perineural pressure exerted by the
esis of AlON (20years), and the pathogenesis of CSF. This pressure under normal circumstances
optic disc edema in raised intracranial pressure may not interfere with optic nerve blood flow; how-
and other conditions (30years); (4) studies of the ever, in progressive NAION this subdural CSF
role of axoplasmic flow in optic disc edema due to might produce additional compromise of vascular
various causes as well as in acute ocular hyperten- perfusion and impair axoplasmic transport. This
sion; and (5) experimental studies on decompression statement is simply not supported by the basic
of the optic nerve sheath in optic disc edema due hemodynamics of the optic nerve head.
to raised intracranial pressure [4]. Based on this Decompression of the sheath of the optic nerve can-
wide experience, I raise the following questions not influence ischemia of the optic disc in NA-AION
about the hypothesis of Sergott etal. [2]. because the optic disc is for the most part supplied
In NA-AION, is the visual loss due to blockage of by the peripapillary choroid. Our fluorescein fun-
the axoplasmic flow? Optic disc edema in raised dus angiographic studies, in more than 500 patients
Optic Nerve Sheath Decompression 391
seen during the very early stages of acute NA-AION, papillopathy is totally erroneous, and their claim
have overwhelmingly shown peripapillary choroi- that AION in diabetic patients is an often self-lim-
dal filling defects. Thus, the authors statement that ited condition with visual recovery is contradicted
optic nerve sheath decompression represents a by our large prospective study on the subject. Sergott
rational neurophysiologic approach in NA-AION etal. [2] after surgery, reported visual improvement
has no scientific validity. in two patients in the contralateral eye with long-
What does a distended optic nerve sheath repre- standing decreased vision due to NA-AION and
sent? The authors stated, At surgery, the optic nerve attributed it once again to a decreased perineural
sheaths were distended, similar in appearance to the CSF pressure around the previously damaged optic
nerve sheath we have observed in pseudotumor cerebri nerve which had not been operated on. As discussed
Incision of the retrobulbar meningeal sheath resulted in above, the assumption that perineural CSF pressure
the release of a considerable amount of clear CSF such has anything to do with the visual loss in NA-AION
as observed with pseudotumor cerebri. Incisions into is entirely mistaken. I have seen similar visual recov-
multiple areas of the nerve sheath and lysis of presumed ery without any treatment whatsoever. Thus, the
arachnoidal adhesions within the subdural space natural history of NA-AION is more variable than is
resulted in additional CSF drainage, similar to the oper- generally thought; a few patients progress and a few
ative findings associated with chronic papilledema spontaneously recover some vision. The authors may
caused by pseudotumor cerebri. I studied the normal contend that they had a group of control cases in
anatomy of the optic nerve sheath in 80 human subjects their series, but their sample size is far too small,
[7] and also in vivo in more than 200 normal rhesus both in the treated and control groups, to give accu-
monkeys. I found that just behind the globe the sheath is rate information, and, as the authors pointed out, the
normally bulbous and much larger in diameter than the study was neither randomized nor double-masked.
retrobulbar optic nerve. Since the dural sheath is made There is another confounding factor that one must
up of collagen tissue and not of elastic tissue it cannot keep in mind when recording visual acuity and visual
distend abnormally at a level of CSF pressure com- fields patients show a certain learning curve so that
patible with life. The arachnoidal adhesions are a com- the first visual acuity and visual fields may not be
mon normal finding. The amount of CSF released on representative of the actual visual loss in all patients.
incision of the retrobulbar part of the sheath depends on Variability of visual functions on repeated testing in
the capacity of the bulbous part of the sheath and on the NA-AION is a not uncommon finding.
extent of communication between the sheath and the Thus, I have serious reservations about the claims
intracranial subarachnoid space through the optic canal of Sergott etal. [2] and I find no scientific basis for
region, which is highly variable [7]. Thus, the authors the hypothesis they put forward to explain their
have described as abnormal perfectly normal findings claim. My biggest concern, however, is that the
on the retrobulbar part of the sheath of the optic nerve. authors may have unwittingly and with the best of
What is the natural history of visual loss in intentions opened a Pandoras box with this publica-
NA-AION? There is no doubt that there is a progres- tion for these reasons: (1) NA-AION is a very com-
sive form of NA-AION, but it is quite unusual; the mon disease in middle-aged and elderly people. (2)
incidence in our prospective series is far lower than Patients who have suffered a dramatic visual loss
that cited by Sergott etal. [2]. We have seen sponta- are desperate people who will try anything, and are
neous improvement of vision of the type claimed by therefore most vulnerable. New surgical procedures
the authors in some cases, without any treatment. always have a special glamour, and many patients
Over the years, we have given the option of systemic with AION will happily submit themselves to such
corticosteroid therapy to these patients; some of the procedures, whether their visual loss is progressive
treated patients improved, a few from finger count- or not. (3) Surgical decompression of the sheath of
ing visual acuity and marked visual field defects to the optic nerve is not altogether a benign procedure;
20/40 or better, with almost normal visual fields. The in addition to complications associated with general
statement by Sergott etal. [2] that patients showing anesthesia, serious blinding complications are pos-
beneficial influence of corticosteroid therapy in my sible, especially in patients with AION who are par-
previous study [8] could have had diabetic ticularly vulnerable to further ocular ischemic
disorders. (4) We know that surgical procedures are imaging technique) in 25 patients after optic nerve
a very lucrative business. The combination of all sheath decompression with progressive NA-AION,
these factors, added to the understandable enthusi- presented a paper at the annual meeting of the American
asm of even honest ophthalmologists to undertake a Academy of Ophthalmology, concluding that: These
new procedure, could create a serious ethical prob- data demonstrate that eyes with acute NAION have
lem. Although Sergott et al. [2] have stressed that impaired blood flow when compared with the contral-
surgical decompression of the sheath should be ateral control group. Furthermore, they suggest that
considered only for that small group of patients ONSD may improve blood flow to the ischemic optic
with progressive visual loss, I fear that proviso will nerve halting the progression of visual loss and in some
be quickly forgotten, as ethics and caution are cases improving visual function. I was asked by the
pushed aside by expediency and fashion. Academy to discuss this study at the meeting. An
expert in color Doppler imaging technique, was my
Heidemann and Kaufman [9] also questioned the valid-
co-author (Dr. Kirk W. Beach); we found problems
ity of axoplasmic flow playing any role in optic nerve
with the interpretation of the data, and concluded that
sheath decompression, as postulated by Sergott etal.
their data did not support their conclusion [16].
[2]. They shared my misgivings about the beneficial
Nevertheless, the procedure soon gained world-wide
effects of optic nerve sheath decompression claimed
favor not only in progressive but also in all types of
by Sergott etal. [2].
NA-AION. Others jumped on the bandwagon, claim-
ing a beneficial effect of optic nerve sheath decompres-
Since that time more definite information has emerged,
sion also [1720]. Manor [17] claimed improvement of
based on large studies, about the following statements
visual outcome by optic nerve decompression in one
that I made in my original letter to the editor.
eye. Kelman etal. [18] claimed that they did optic nerve
1. We have seen spontaneous improvement of vision of sheath decompression in seven patients with NA-AION
the type claimed by the authors in some cases, with- and Visual acuity improved markedly in all patients;
out any treatment. The two large prospective studies they concluded that their experience supports the pos-
[10, 11] have now shown that there is a spontaneous sible beneficial effect of optic nerve sheath decompres-
visual acuity improvement in 4143% of cases. sion in patients with nonarteritic anterior ischemic
2. Over the years, we have given the option of sys- optic neuropathy. Spoor etal. [19] similarly performed
temic corticosteroid therapy to these patients; some optic nerve sheath decompression on four patients (five
of the treated patients improved. Our study [12] on eyes) with visual loss secondary to NA-AION. They
a large cohort of NA-AION patients, discussed claimed that 4 of the 5 eyes had marked improvement
below at length, has shown that there is a significant in visual function after the operation, and concluded
visual acuity (p=0.001) and visual field (p=0.005) that Optic nerve sheath decompression is an effective
improvement with systemic corticosteroid therapy. treatment for patients with nonarteritic ischemic optic
3. Their claim that AION in diabetic patients is an neuropathy and progressive visual loss. In another
often self-limited condition with visual recovery study, these authors [20] did optic nerve sheath decom-
is contradicted by our large prospective study on the pression in 23 patients with progressive NA-AION and
subject. Recent studies [13, 14] have confirmed 15 patients with nonprogressive NA-AION, and
that statement. claimed that Optic nerve sheath decompression
improves visual acuity but has little effect on overall
This letter was not well-received by the neuro-ophthal- visual function in patients with progressive NAION.
mology community nor by the surgeons who wanted to In contrast to all the claims made for beneficial
do optic nerve sheath decompression, who all believed effects of optic nerve sheath decompression, there fol-
that this was a definitive treatment for NA-AION a lowed a few studies, which did not find any beneficial
disease which had no other known treatment at that time. effect. For example, Jablons etal. [21] performed optic
I was blamed for denying this apparently hopeful treat- nerve sheath fenestration in 26 eyes with progressive
ment to these desperate patients who had lost vision. NA-AION and failed to substantiate the sanguine
Later, Sergott and colleagues [15], based on their visual outcome in recently reported series of patients
retrobulbar circulation study (using the color Doppler undergoing optic nerve sheath decompression. Glaser
Aspirin 393
etal. [22] performed optic nerve sheath fenestration in knowledge of these basic facts results in advocating
21 eyes with progressive NA-AION, and concluded treatments which not only confer no benefit but may
that surgical procedures indicate no beneficial effect even be harmful. Optic nerve sheath decompression in
on visual morbidity. Yee etal. [23] performed optic NA-AION is a classical example of that. The contro-
nerve sheath decompression in 18 eyes with progres- versy about my role in debunking optic nerve sheath
sive and nonprogressive NA-AION, and found no ben- decompression still lingers on [24].
eficial effect.
Finally, a multicenter, randomized, single-masked,
clinical trial [10] was conducted by the United States
National Institutes of Health to assess the safety and Aspirin
efficacy of optic nerve decompression surgery, com-
pared with careful follow-up alone in patients with It is a common practice among ophthalmologists and
NA-AION. The study began in October 1992, and was neurologists to advocate the use of aspirin (or occasion-
terminated in October 1994, on the recommendation ally, other anti-platelet-aggregating drugs) for patients
of its Data and Safety Monitoring Committee. The with NA-AION. This is because of the common per-
study included 244 patients with NA-AION, of whom ception among ophthalmologists and neurologists that
119 had optic nerve sheath decompression surgery and NA-AION and cerebral stroke are similar in nature
125 had no surgery and a careful follow-up. The crite- pathogenetically and in management. This has resulted
ria to include patients in the study were a definite diag- in major controversy on the pathogenesis and manage-
nosis of NA-AION, best-corrected visual acuity in the ment of NA-AION. In view of this, the crucial question
affected eye of 20/64 or worse and duration of symp- is: is NA-AION really a thromboembolic disorder in
toms less than 14days at the baseline examination in the vast majority? The answer to that lies in etiology
persons aged 50years or older. A comparison of visual and pathogenesis of NA-AION. This is discussed else-
acuity between patients who had optic nerve sheath where (see Chap. 14). NA-AION is usually a hypoten-
decompression and those who did not have any sur- sive disorder. I recently discussed in detail the role of
gery, showed that at 6 months 32.6% of the surgery aspirin in NA-AION elsewhere [25]. The following
group improved compared with 42.7% of the careful pieces of evidence further show that, unlike stroke,
follow-up group. The odds ratio for three or more lines NA-AION is usually not a thromboembolic disorder.
better, adjusted for baseline visual acuity and diabetes,
was 0.74 (95% confidence interval [CI], 0.391.38). 1. First and foremost, if NA-AION were a thromboem-
There was a significantly greater risk of losing three or bolic disorder, fluorescein fundus angiography dur-
more lines of vision at 6months in patients who had ing the early acute stage, at the onset of visual loss,
optic nerve sheath decompression (23.9%) compared must invariably show evidence of complete occlu-
to those without any surgery (12.4%). The 6-month sion of the posterior ciliary artery (Figs. 17.1 and
adjusted odds ratio for three or more lines worse was 17.2; also see Figs. 3.243.26, 3.283.37 in Chap. 3),
1.96 (95% CI, 0.874.41). There was no difference in which supplies the optic nerve head [27, 28]. My
treatment effect between patients with progressive practice has been to do fluorescein fundus angiogra-
NA-AION and all others. The study concluded that the phy in all cases with fresh onset of NA-AION more
results indicate that optic nerve decompression sur- than a thousand patients so far. Fluorescein fundus
gery for NAION is not effective, may be harmful, and angiography soon after the onset of NA-AION shows
should be abandoned. The spontaneous improvement only a delayed and slow filling of the peripapillary
rate is better than previously reported. choroid and/or choroidal watershed zones (mostly
Thus, this study confirmed what I had stated 5years due to atherosclerosis and/or arteriosclerotic changes
earlier, based on my basic studies on NA-AION, that in the ocular and ophthalmic arteries) (Figs.17.3 and
optic nerve sheath decompression can be harmful in 17.4; also see Figs. 3.39, 3.41, 3.45, 3.52, 3.54 and
NA-AION. Again for the rational management of a 3.573.59 in Chap. 3), but NO permanent occlusion
disease, the first essential is to comprehend the basic at all [2729]. This provides definite proof that
facts about its pathogenesis, because those provide NA-AION is not a thromboembolic occlusive
the scientific rationale for treatment. Lack of disorder.

itic AION, shows normal filling of the area supplied by the lat-
eral PCA (including the temporal of the optic disc) but no
filling of the area supplied by the medial PCA (including the
nasal of the disc). (Reproduced from Hayreh [26])
Fig. 17.3 Fundus photograph (above) and fluorescein angio-

gram (below) of right eye with NA-AION. Fundus photograph
Fig. 17.2 Fluorescein fundus angiogram of right eye with shows optic disc edema. Fluorescein angiogram at that time
embolic NA-AION and carotid artery disease (negative tempo- showed non-filling of temporal part of the peripapillary choroid
ral artery biopsy for giant cell arteritis), showing normal filling and adjacent optic disc and of upper half of the choroidal water-
of the area supplied by the lateral PCA (including the temporal shed zone (arrows), with normal filling of the medial and lateral
half of optic disc), but no filling of the area supplied by the posterior ciliary arteries. (Reproduced from Hayreh [27])
medial PCA (including the nasal half of optic disc). (Reproduced
from Hayreh [27])
2. The severity and the duration of ischemia determine nonperfusion or hypoperfusion of the optic nerve
the severity of ischemic optic nerve head damage, head circulation, usually occurring during sleep,
and that in turn determines the extent of recovery of there is usually much less severe and less extensive
visual function following the acute episode. In clas- optic nerve head damage than in arteritic AION due
sical NA-AION, because there is only transient to giant cell arteritis, in which there is thrombotic
Aspirin 395
a giant cell arteritis would cause optic nerve head

damage similar to that seen in arteritic AION.
3. The beneficial effect of aspirin in cerebrovascular
accident (usually a thromboembolic disorder) is well-
established. Two large studies [32, 33] have investi-
gated the role of aspirin in NA-AION. In the first
[32], the role of aspirin was evaluated in a long-term,
comprehensive study in 431 patients with unilateral
NA-AION. That revealed no long-term benefit from
aspirin in reducing the risk of NA-AION in the fel-
low eye on a 5-year follow-up. Another prospective
study [33] of 173 patients with unilateral NA-AION,
found no association between regular aspirin use and
b incidence of new NA-AION in the fellow eye on a
5-year follow-up. Thus, two comprehensive studies,
involving 604 patients with unilateral NA-AION and
a 5-year follow-up, showed no beneficial effect of
aspirin. This is not surprising since NA-AION is
NOT a thromboembolic but a hypotensive disorder
[28, 29, 34], and aspirin has no effect on the blood
pressure or nocturnal arterial hypotension.
4. There is a huge volume of literature showing a sig-
nificant association between smoking and cerebro-
vascular accident (a thromboembolic disorder) [35].
Yet two large prospective studies [33, 36] have shown
no association between smoking and NA-AION.
5. While an association has been reported between
thrombophilic risk factors and cerebrovascular
Fig.17.4 Two fluorescein fundus angiograms at 6-s interval in accident, no significant association has been found
the dye circulation (15s after the injection of fluorescein), shows
between NA-AION and thrombophilic risk factors
no filling of the temporal, superior and inferior peripapillary in the vast majority [3739], because NA-AION is
choroid and superior choroidal watershed zone (arrow), and the hypotensive and not thrombotic.
optic disc, with normal filling of both the medial and lateral pos- 6. There is a distinct difference in the seasonal varia-
tion between onset of NA-AION and that of cere-
there is complete filling of the peripapillary choroid (main brovascular accident and myocardial infarction.
source of blood supply to the optic nerve head) and the water- Our study [34] showed that NA-AION onset was
shed zone significantly more frequent in summer (or hot
months) than in winter (or cold months) (p=0.0030).
This is in contrast to the reported incidence of myo-
cardial infarction and cerebrovascular accidents,
which develop more frequently during winter than
occlusion of the posterior ciliary artery [2729]. other seasons [4042]. The reasons for the seasonal
Two large studies [10, 11] have shown that about variations of myocardial infarction and cerebrovas-
40% of eyes with NA-AION show spontaneous cular accidents are discussed at length elsewhere
visual acuity improvement. In arteritic AION (with [34]. Moreover, our study [43] of 406 NA-AION
complete occlusion of the posterior ciliary artery), patients showed that, in those without arterial
by contrast, no such visual improvement is seen hypertension and diabetes mellitus, there is no sig-
[30, 31]. Thrombotic or embolic occlusion of the nificant increase in mortality, and no increase in
posterior ciliary artery due to causes other than cerebrovascular disease.
7. Transcranial Doppler in NA-AION patients, com- 750 patients, Beck etal. [32] had already stressed
pared to age-matched controls, did not reveal an that the findings of their study in 431 patients lim-
increased incidence of embolic events [44]. its the feasibility of conducting a randomized trial
of aspirin use in treating nonarteritic anterior isch-
Thus, there is overwhelming evidence that NA-AION,
emic optic neuropathy because of the sample size
unlike stroke, is not a thromboembolic disorder, and
that would be required to detect a beneficial treat-
that it is a serious mistake to manage NA-AION like
ment effect. In addition, the routine use of aspirin
stroke the primary basis for advocating the use of
for many individuals older than 50 years of age
aspirin in NA-AION. There is no scientifically valid
would have a detrimental effect on recruitment for a
rationale for it.
placebo-controlled trial.
Stiebel-Kalish and colleagues [45], on a review of
In conclusion, aspirin has no beneficial effect in ordi-
the literature, concluded that: Data suggest a benefit
nary NA-AION. However, in a rare case where NA-AION
of aspirin in prevention of second eye NAION, but the
is embolic in nature (Fig.17.2), with the embolus com-
evidence is weak. The conclusions of this study must
ing from plaques in the carotid artery or the heart, use of
be put in proper perspective as to (1) their validity, and
aspirin would be indicated. Therefore, in some cases of
(2) most importantly the rationale for the role of aspi-
NA-AION, evaluation for a source of embolism may be
rin in NA-AION.
indicated. In this respect, my studies have shown that the
1. The result of any such review depends upon the presence of plaque in the carotid arteries is far more
sample size, the nature of the studies selected and important than the degree of stenosis.
(sometimes) personal biases. The authors seem to
base their conclusion on four studies with very dif-
ferent sample sizes and natures. Their conclusion of
a beneficial effect of aspirin in NA-AION is drawn Levodopa
solely from two small studies, based on retrospec-
tive review of charts of 131 [46] and 56 [47] patients Johnson et al. [49, 50] in a retrospective study of
and a follow-up of 2years. The authors did not con- patients with NA-AION, concluded that levodopa
sider the conclusions of two large, comprehensive appears to be beneficial in the treatment of recent-onset
studies based on 431 [32] and 173 [33] patients NAION [50]. However, this conclusion was not justi-
respectively, and a follow-up of 5years both of fiable, as discussed in detail elsewhere [51, 52].
which came to the exact same conclusion, that aspi- Johnson et al.[50] claimed that 10 of 13 (76.9%)
rin has no beneficial effect in NA-AION. It is puz- NA-AION patients with visual acuity of 20/40 or
zling that Stiebel-Kalish and colleagues [45] did worse, when treated within 45 days of onset with
not take into consideration the findings of the two levodopa, showed a significant improvement in visual
latter comprehensive studies [32, 33] which cannot acuity at 6 months. As mentioned earlier, the first
be called weak evidence. Moreover, Salomon essential in evaluating any therapy for a disease is to
etal. [47] (one of the two studies cited by the authors determine whether it has a scientific rationale, and
in support of their conclusion) miscited [48] the whether the study design is scientifically sound. The
study by Beck etal. [32] when advocating adminis- authors [49, 50] assumed that with levodopa and carbi-
tration of aspirin. Beck et al. [32] actually con- dopa, following acute ischemia, the axons of the optic
cluded that there is little or no long-term (5years) nerve in NA-AION are still surviving and suffering
benefit in using aspirin to reduce the risk of NAION from ischemia for up to 45days (their inclusion crite-
in the fellow eye. rion [50]) or for longer than 6 months [49] and that
2. As discussed above [25], there is no scientifically their function can be restored by drugs. There is no
valid rationale for the use of aspirin in NA-AION. scientific evidence to support that. It is difficult to
Thus, the conclusion by Stiebel-Kalish and col- imagine how levodopa or, for that matter, any drug can
leagues [45] that Data suggest a benefit of aspirin resurrect permanently destroyed axons. Thus, there is
in prevention of second eye NAION is unwarranted a flaw in the basic rationale of the study. Johnson etal.
and misleading. Regarding their call for conducting [50] explained their claimed visual improvement from
a multicenter randomized controlled trial, recruiting levodopa by arguing (based on studies in the literature
Levodopa 397
on amblyopia and Parkinsons disease) that: Levodopa 20/200 into logMAR. The main claim of this study is
crosses the bloodbrain and blood-retinal barriers to improvement in visual acuity, which is easy to make
increase brain and retinal dopamine levels... Dopamine but often may be illusory. From my experience of pro-
may promote visual recovery by enhancing neuronal spectively studying about 1,350 patients with NA-AION
plasticity or function in either the retina, lateral and even larger number of patients with various retinal
geniculate body, or visual cortex. Dopamine also may vascular and optic nerve disorders (testing the visual
alter the metabolic milieu of the retina and vitreous, acuity carefully in every patient myself), I have found
thereby preventing ischemic neuronal injury. The fun- a number of confounding factors and artifacts in the
damental fact is that NA-AION is due to acute isch- visual acuity obtained on routine testing, particularly
emic damage to the axons in the optic nerve head. It is when a patient has a central scotoma or a visual field
a disorder neither of the retina, nor of the lateral genic- defect which passes through or involves the fixation
ulate body nor the visual cortex, with no vitreous point (see Figs. 16.116.5 and 16.9 in Chap. 16). Not
abnormalities. That simple fact invalidates the argu- uncommonly the initial visual acuity tends to be poorer
ment for using levodopa in NA-AION. From the thera- than it actually is, because, having suffered a sudden
peutic point of view, the authors equated NA-AION visual loss, the patient is emotionally stressed and per-
with amblyopia and Parkinsons disease (as indicated forms poorly. Also, apparent improvement in visual
above) in this study, but there is nothing in common acuity may simply represent a learning experience on
between NA-AION and the other two diseases, the part of the patient; over time the patient may learn
morphologically, etiologically or pathogenetically; by experience to read the testing chart much better by
NA-AION is an acute ischemic disorder of the optic fixating eccentrically. It also depends upon how knowl-
nerve head where the axons are destroyed permanently, edgeable the examiner is about the nature of visual
immediately, or within a few days, by acute ischemia; defect and how to test to get the best visual acuity.
that is not at all the case in the other two conditions. I have often found that on testing the visual acuity of
Amblyopia is not an ischemic disorder and the optic these patients, I can get much better (often three or
nerve axons are intact. Parkinsons disease is a totally more lines) visual acuity than that recorded even by an
different ballgame. To argue that all the three diseases experienced technician only 15min earlier in another
would respond identically to levodopa represents a clinic. In view of this problem, it is essential to corre-
poor understanding of the different disease processes late the visual acuity improvement with the type of
of NA-AION, amblyopia, and Parkinsons disease. central visual field defect and the changes in it on fol-
Thus, it is evident that the use of levodopa in the low-up. Visual acuity improvement without corre-
treatment of NA-AION or its claimed beneficial effect sponding central visual field improvement usually
on visual acuity has no scientific rationale. represents an artifact [11, 30, 53]. When judging
The conclusions of Johnson etal. [50] were based improvement in visual fields, simply using a differ-
on a retrospective retrieval of information from charts ence in mean deviation of 3.0dB or more is not mean-
of 37 patients, selected from 112 NA-AION patients ingful, because overall visual field improvement has
seen by the authors over the previous 9 years. The little correlation with change in visual acuity. Moreover,
study was retrospective, unplanned, nonrandomized, the study by Johnson etal. [50] did not show any sig-
based on a small sample size with unbalanced treated nificant improvement in the visual fields overall any-
and untreated groups, with a serious possibility of bias way, further evidence that the visual acuity improvement
in selection and evaluation of patients and visual acuity claimed was probably not genuine. Johnson etal. [50]
measurement (because neither the patients, nor the gave extensive statistics and p-values, but experienced
physicians and testers were masked as to treatment biostatisticians have found a number of confounding
assignment). There were also problems with statistical factors in their statistical methods and data analyses
data analyses [51]. The authors claimed that 76.9% [51], and that invalidated their claim.
showed improvement of visual acuity of three or more In conclusion, the claims made by Johnson et al.
lines; however, 5 of the 10 patients in that group had a [49, 50] in their studies have no scientific rationale and
baseline visual acuity of 20/400 or worse, and the contain flaws; their conclusion that levodopa appears
claimed improvement may be due to an artifact caused to be beneficial in the treatment of recent-onset
by conversion of Snellen visual acuity of less than NAION cannot be accepted.
Systemic Corticosteroid Therapy systemic corticosteroids improved visual outcome.

Unfortunately, that clinical trial was not funded by the
U.S. National Institute of Health, because of a firm but
Miller and Smith [54] in 1966, first treated 6 of 11
unsubstantiated belief among neuro-ophthalmology
patients with corticosteroid therapy. They stated:
reviewers that corticosteroid therapy has no role in
Steroids and anticoagulants have been used in these
NA-AION. (Unfortunately, as is evident from the dis-
patients but the final evaluation of their efficacy awaits
cussion below, in spite of tremendous advances in our
further study. Foulds [55] treated 13 of 24 patients in
knowledge of NA-AION during the past 4 decades,
1969 with NA-AION with systemic corticosteroids; on
that basic philosophy of neuro-ophthalmologists on
comparing the treated with the untreated patients, he
the subject has not changed at all).
reported significant visual improvement in 85% (11 of
Lacking extramural funding, I decided on a pro-
13) of the treated cases compared to 45% (5 of 11)
spective patient choice controlled study instead of
of the untreated patients. Based on a preliminary study
the conventional randomized study the next best
of 14 patients with NA-AION, seen from 1970 to 1972
choice. Every fresh NA-AION patient seen in my
at the Edinburgh Royal Infirmary, Scotland, I [8]
clinic was given a free and informed patient choice.
reported visual acuity improvement in 75% of the eight
The decision was left entirely up to the patient, to opt
patients with NA-AION and in 17% of the six untreated
for corticosteroid therapy or no treatment, in consulta-
patients.
tion with their physicians or other sources. I and my
In 1973, I decided to conduct a systematic study on
staff had no in-put at all into their choice. I specifi-
the role of corticosteroid therapy in NA-AION, at the
cally told all patients that I really did not know whether
Ocular Vascular Clinic of the University of Iowa
the treatment was beneficial, ineffective or harmful. I
Hospitals and Clinics, for the following reasons.
collected the data for 28 years, completely masked
about visual outcomes and numbers of patients in each
1. Studies on small numbers of patients by Foulds [55]
group. The primary objective of this study was to eval-
and myself [8] showed encouraging results of ben-
uate in a large cohort of NA-AION patients the role of
eficial effects of systemic corticosteroid therapy in
systemic corticosteroid therapy during its acute phase.
NA-AION.
The study finally included 613 consecutive NA-AION
2. Miller and Smith [54], who used corticosteroid
patients (696 eyes) seen in my clinic [12].
therapy in 6 of their 11 patients, commented that
We [12] recently reported its finding. Following
the final evaluation of their efficacy awaits further
was the corticosteroid treatment protocol in this
study.
study: Initially 80mg of Prednisone daily for 2weeks,
3. Also, during my discussions with ophthalmologists
then tapered down to 70 mg for 5 days, 60 mg for
during professional meetings in North and South
5days, then cut down by 5mg every 5days, to noth-
America, Europe and Asia, I found that many oph-
ing. In this patient randomization study, there were
thalmologists were treating these patients with sys-
696 eyes with NA-AION; of these patients, 51% vol-
temic corticosteroid therapy empirically, for lack of
untarily opted for systemic corticosteroid therapy
any other alternative treatment, and all giving dif-
while 49% opted for no treatment. The study showed
ferent doses for different lengths of time.
that at 6months after the onset of NA-AION, of the
4. The whole subject of the role of corticosteroid ther-
eyes with initial visual acuity of 20/70 or worse, seen
apy in NA-AION was highly controversial, with
within 2 weeks of onset, there was visual acuity
some having strong opinions about the lack of any
improvement in 69.8% (95% confidence interval
rationale for corticosteroid therapy in NA-AION. I
(CI): 57.379.9%) in the treated group, compared to
felt that a comprehensive study in a large cohort of
40.5% (95% CI: 29.252.9%) in the untreated group
patients with NA-AION was essential, to determine
(odds ratio of improvement: 3.39; 95% CI:1.627.11;
whether this therapy was beneficial, ineffective or
p=0.001). Comparison of visual field defects at
harmful.
6months from onset of NA-AION, among those seen
In view of the above, in 1973, I planned a large, multi- within 2weeks of NA-AION onset, with moderate to
center randomized clinical trial to investigate system- severe initial visual field defect, showed improve-
atically in a large cohort of NA-AION patients whether ment in 40.1% (95% CI: 33.147.5%) of the treated
Systemic Corticosteroid Therapy 399
group versus 24.5% (95% CI: 17.732.9%) of the She had an episode of headache when the vision in the
untreated group (odds ratio: 2.06, 95% CI: 1.243.40; right eye blurred. Figures 17.6a and 17.7a show the
p=0.005). In both treated and untreated groups, the visual field defects in the right and left eye respectively
visual acuity and visual fields kept improving for up at her first visit to my clinic; her visual acuity was 20/200
to about 6months from onset of NA-AION and very and 20/50 respectively. When seen 11days later, there
little thereafter. A comparison of treated versus was marked deterioration in the visual fields in both
untreated groups also showed that optic disc edema eyes (Figs.17.6b and 17.7b) and the visual acuity was
resolved significantly (p=0.0006) faster in the treated counting fingers at one foot in the right eye and 20/100
group [56]. in the left eye. She decided to go on corticosteroid
This study showed that treating NA-AION eyes therapy at that visit. When seen 1week later, there was
during the acute phase with high-dose systemic corti- marked improvement in the visual field (Figs. 17.6c
costeroids results in a significantly higher probability and 17.7c) in both eyes, and the visual acuity was
of improvement in visual acuity (p=0.001) and visual 20/50 in the right eye and 20/25 in the left eye. Two
fields (p=0.005), compared to an untreated group. weeks after the start of therapy, visual acuity was
Both visual acuity and visual fields improved for up 20/30 in the right eye and 20/20 in the left eye, with a
to 6 months after onset of NA-AION and no more normal visual field in the left eye (Fig. 17.7d). Four
after that. months after the start of therapy, in the right eye the
Striking visual improvement occurred in some eyes visual acuity was 20/25 with a normal visual field
following corticosteroid therapy, as is illustrated by the (Fig.17.6d). I followed this lady for 21years, and her
following examples (3 eyes of two patients). vision remained stable.
Case 1: A 64-year old man noticed blurred vision This study, however, has not been at all well-
in his left eye in the morning at breakfast time, 12days received by the neuro-ophthalmology community.
before he was seen in my clinic. Next morning he Neuro-ophthalmologists do not accept the findings of
noticed that the vision disappeared. When seen in this study and are still of the opinion that corticoster-
my clinic, his visual acuity was bare counting fingers oid therapy has no role or rationale in the treatment of
in a temporal island and Fig.17.5a shows the visual NA-AION. I have learned from multiple e-mails, let-
field defect at that visit. Evaluation showed no evi- ters and telephone calls from NA-AION patients that
dence of giant cell arteritis. The fluorescein fundus they have been told by neuro-ophthalmologists
angiographic findings in Fig. 17.4a definitely ruled that corticosteroid therapy will not benefit them and
out arteritic AION (see Chap. 16). He opted for corti- that there is no known treatment available. This study
costeroid therapy 6 days later. Visual fields in [12] recently has prompted much debate [1, 5760],
Figs.17.5bf show the course after the start of corti- and has become a center of controversy; therefore, it
costeroid therapy Fig.17.5b after 1week, Fig.17.5c is crucial to discuss it at length to place the whole
after 3 weeks, Fig. 17.5d after 3 months, Fig. 17.5e subject in proper perspective. A full discussion of this
after 6months and Fig.17.5f finally. One week after treatment regimen [12] is absolutely vital because
the start of corticosteroid therapy, the visual acuity there has not previously been any effective treatment
was counting fingers at 4 ft, after 3 weeks 20/200 for NA-AION, and this study [12] provides hope to
20/300, after 3months 20/70, after 6months it fluctu- these patients.
ated between 20/40 and 20/70, depending upon The following major objections have been raised by
eccentric fixation or not, due to the presence of a supe- neuro-ophthalmologists to this study:
rior paracentral scotoma touching fixation. The final
1. There is no scientific rationale for the use of corti-
visual acuity achieved varied between 20/30 and
costeroid therapy in NA-AION.
20/40. I followed this patient for 14 years and he
2. There was no conventional randomization in this
maintained his improved vision for all that time, with
study.
a normal right eye.
3. The data were not collected in a masked fashion.
Case 2: A 59year old woman developed decreased
vision in the right eye 1 month, and in the left eye We discussed all these issues at length in our original
1week, before she was seen in my clinic. She suffered article [12]. Following are my responses to further
from arterial hypotension, migraine and hyperlipidemia. clarify these objections:
Fig.17.5 Visual fields of Case 1 before (a) and at different time intervals after the start of corticosteroid therapy (b after 1week, c
after 3weeks, d after 3months, e after 6months and f finally)
Fig.17.6 Visual fields of Case 2 right eye, before (a at initial visit and b 11days later) and at different time intervals after the start
of corticosteroid therapy (c after 1week, d after 4months)
Scientific Rationale for Visual accumulation and consequent axonal swelling in

Improvement with Corticosteroid the optic nerve head [5, 64, 65]; that manifests as
optic disc edema, which is seen invariably in all
Therapy in NA-AION
NA-AION eyes initially.
3. It has been shown that, in the majority of NA-AION
Naturally, the question arises, why did corticosteroid
eyes, the optic disc has no cup at all or only a small
therapy help to improve the visual acuity and visual
cup [66, 67]. Thus, there is crowding of the nerve
fields of NA-AION patients? This is discussed at
fibers as they pass through a restricted space in the
length elsewhere [12].
rigid opening in Bruchs membrane and the small
To comprehend that, one has to consider some rel-
scleral canal. The importance of this factor is that the
evant basic aspects of NA-AION.
swollen axons in the restricted and unyielding space
1. NA-AION is due to ischemia of the optic nerve within the optic nerve head have to expand at the
head, which is primarily supplied by the posterior cost of other tissues in that space. The only thing that
ciliary artery circulation [6163]. they can compress to expand is the fine capillaries
2. Ischemia of axons in NA-AION results in axoplasmic lying among them; that results in secondary vascular
flow stasis [64, 65], which in turn causes axoplasmic changes [5]. A vicious circle may, therefore, be set
Fig.17.7 Visual fields of Case 2, left eye, before (a at initial visit and b 11days later) and at different time intervals after the start
of corticosteroid therapy (c after 1week, d after 2weeks)
up, in which compression of capillaries may further due to dye leaking from the capillaries in the optic
aggravate ischemia, particularly when perfusion nerve head (Fig.17.8). Fluorescein leakage may be
pressure in the capillaries falls for any reason (as for due to two factors: (1) ischemic insult to the capil-
example, during nocturnal arterial hypotension) [34, laries in the optic nerve head, and (2) venous stasis
68, 69]. This is supported by the fact that in at least produced by the capillary compression [5]. Foulds
73% of episodes of NA-AION, visual loss was dis- [55] also pointed out that increased capillary per-
covered first upon awakening or a first opportunity to meability due to anoxic capillary damage was an
use vision critically after sleeping, because of a fall important factor in the development of optic disc
of blood pressure during sleep [34] the remaining edema in NA-AION. Kaur et al. [70] recently
patients did not know exactly when it happened. reviewed the effect of ischemia/hypoxia on the
4. On fluorescein fundus angiography, the optic disc bloodoptic nerve barrier. There is disruption of the
with edema in NA-AION always shows late staining bloodoptic nerve barrier in hypoxia/ischemia,
which may be due to several mechanisms, includ- the effect of systemic corticosteroid therapy on optic
ing increased production of vascular endothelial disc edema in NA-AION, by comparing the rate of
growth factor (VEGF), nitric oxide, free radicals, resolution of optic disc edema in the treated group
inflammatory mediators and hemodynamic altera- (343 eyes) versus the untreated group (380 eyes).
tions. VEGF is known as a vascular permeability Those treated with corticosteroid therapy within
enhancing factor [71, 72]. 2 weeks after onset of NA-AION had significantly
(p=0.0006) faster optic disc edema resolution than the
Therefore, primary and secondary changes take place
untreated cases. This indicates a reduction in capillary
in the optic nerve head to produce optic disc edema in
leakage, similar to that seen in macular edema with
NA-AION the primary change being ischemic axo-
corticosteroid therapy.
plasmic flow stasis in the axons and the secondary
Thus, the most likely scenario that emerges to explain
change, vascular changes and fluid leakage.
the beneficial effect of corticosteroid therapy on visual
Foulds [55] postulated that corticosteroid therapy in
outcome in NA-AION seems to be as follows: The
acute NA-AION reduces optic disc edema by reducing
faster resolution of optic disc edema with corticoster-
the capillary permeability. There is ample evidence
oid therapy (compared to the untreated patients) [56]
that corticosteroids work in many non-inflammatory
progressive decrease of compression of the capillar-
diseases. Corticosteroids have a specific effect on
ies in the optic nerve headbetter blood flow in the
reducing vascular permeability [73]. For example, a
capillariesimproved circulation in the optic nerve
large number of studies have shown that corticosteroid
headimproved function of the surviving but not
therapy reduces macular edema from various causes.
functioning hypoxic axons. It is like a person who
The reduction is due to alteration of capillary permea-
while being starved of food, is too weak to work; but
bility and reduction of fluid leakage. As discussed
once he is fed, he resumes his normal work. There is a
above, fluorescein angiography shows leakage of fluo-
possibility that corticosteroids may have beneficial
rescein in the optic nerve head when the disc is edema-
effects from some other unknown mechanisms; one of
tous in NA-AION (Fig. 17.8), but not in normal or
those mentioned has been inhibition of damage by free
atrophic discs a proof of increased capillary perme-
radicals. Therefore, in NA-AION corticosteroid ther-
ability in optic disc edema. Our study [56] investigated
apy DOES have a scientifically valid rationale.
The False Analogy of Stroke with NA-AION
As discussed above, currently there is a common per-

ception among neuro-ophthalmologists and neurolo-
gists that stroke and NA-AION are identical
pathogenetically, and therefore their management
should be identical. It has been argued that since corti-
costeroid therapy does not benefit stroke, there is no
reason to believe it works in NA-AION. This philoso-
phy is evident from comments by Biousse [59], who
stated: A number of well-designed studies in the neu-
rologic literature have shown that corticosteroids do
not improve outcome of patients with acute arterial or
venous cerebral ischemia. Indeed it has even been sug-
gested that corticosteroid therapy can be detrimental to
patients with acute cerebral ischemia and should not
be prescribed. There is fundamental flaw in this argu-
ment. As discussed above, the two conditions are
Fig.17.8 Fluorescein angiogram showing late staining of the pathogenetically totally different clinical entities, and
optic disc in an eye with NA-AION therefore must be different in management. Stroke is
usually a thromboembolic disorder, whereas, as dis- reviewers, based on a firm but unsubstantiated convic-
cussed at length elsewhere [28, 29, 34, 68, 69], tion, held that corticosteroid therapy has no role in
NA-AION is predominantly a hypotensive disorder NA-AION. Therefore, lacking extramural funding, I
caused by transient hypoperfusion or nonperfusion of decided on a prospective patient choice controlled
the optic nerve head circulation. study instead of the conventional randomized study
Furthermore, the severity and the duration of isch- the next best choice.
emia, and the amount of tissue involved by ischemia The study finally included 613 consecutive
determine the functional recovery. In classical NA-AION patients (696 eyes) seen in my clinic [12].
NA-AION, there is not usually marked ischemia,
because it is usually due to transient nonperfusion or
hypoperfusion of the optic nerve head circulation,
What Are the Crucial Criteria Required
occurring almost invariably only during sleep [34],
for Conventional Randomization?
and it involves only a small amount of tissue in the
optic nerve head. That is why two large prospective
It is to have treated and untreated groups comparable
studies have shown that about 40% of eyes with
at baseline in demographic and clinical characteristics.
NA-AION show significant spontaneous visual acuity
In this study [12], as discussed below, that was indeed
improvement [10, 11]. By sharp contrast, in arteritic
the case, as is evident from the following:
AION, which is a thrombotic disorder with severe
ischemic damage, there is practically no visual 1. Of the patients with 696 eyes with NA-AION, 51%
improvement [30, 31]. Similarly, cerebral stroke, being voluntarily opted for systemic steroid therapy.
a thromboembolic disorder, causes comparatively Forty-nine percent opted for no treatment. Unknown
much severer ischemic damage and, importantly, to me while doing the study, the number of patients
involves a much larger mass of tissue than NA-AION in the treated and untreated groups turned out to be
does in the optic nerve head. Therefore, obviously, similar. This would not have occurred if there had
corticosteroid therapy has no beneficial effect in been any bias in selection.
stroke. 2. There was no significant difference between the
Thus, there is overwhelming evidence that two groups in visual acuity, visual fields or systemic
NA-AION, unlike stroke, is not a thromboembolic diseases, except that patients who opted for treat-
disorder, and the two are totally unrelated pathogeneti- ment were slightly younger (59.2 vs. 62.0 years)
cally and consequently in their management. To equate and had a lower prevalence of arterial hypertension
NA-AION and stroke is a fundamental mistake, as is (34% vs. 43%). To determine if differences in age
the consequent assumption that because corticosteroid and arterial hypertension, and also if presence of
therapy is not beneficial in stroke, it is not beneficial in transient ischemic attacks/cerebrovascular acci-
NA-AION. To equate stroke and NA-AION is to call dents and diabetes mellitus influenced the visual
apples oranges. outcome in this study, they were accounted for in
the statistical analysis by including them as covari-
ates in the logistic regression model. (This is simi-
lar to the analysis performed for epidemiological
There Was No Conventional
studies, which do not involve randomization to
Randomization in this Study treatment groups.) From the statistical analysis, we
established that differences in age, arterial hyper-
The reason why a conventional randomized trial was tension, transient ischemic attacks/cerebrovascular
not conducted is discussed in the original paper [12] accidents and diabetes mellitus showed no signifi-
and above. To reiterate, my original planned large, cant association with the primary outcome of visual
multicenter randomized clinical trial, in the early acuity (age at onset p=0.817; hypertension
1970s, to investigate systematically the role of corti- p=0.589; transient ischemic attacks/cerebrovascu-
costeroid therapy in NA-AION in a large cohort of lar accidents p=0.929; diabetes mellitus p=0.516)
NA-AION patients was not funded by the U.S. National or visual field improvement (age at onset p=0.746;
Institute of Health, because neuro-ophthalmology hypertension p=0.271; transient ischemic attacks/
cerebrovascular accidents p=0.829; diabetes mellitus betics totally out of context to this study [12]. The
p=0.972); nor did they alter the finding of a signifi- evaluation of optic disc edema was done by direct
cantly greater likelihood of improvement in visual ophthalmoscopy and later on from stereoscopic
outcome with corticosteroid therapy. Thus, the optic disc photographs and fluorescein angiograms
slight differences in age and in prevalence of arte- all in a masked fashion, so that there is no ques-
rial hypertension did not make any significant dif- tion of bias in the optic disc edema evaluation.
ference in the visual outcome. Therefore, concerns Moreover, we [12] discussed at length what steps
with respect to differences between the groups that were taken to mask visual acuity and visual field
may be due to the study not being randomized had evaluation in the study (see below).
been adequately addressed by the statistical 3. In spite of the above data and my response [57] to
analysis. their original criticism of our study [12] Biousse
[58, 59] further criticized our [12] study by stating
the following:
(a) The untreated group had more vascular risk
Response to Criticism by Biousse and Colleagues
factors (especially diabetes) As discussed in
About Our Study [12]
detail above and in the original manuscript [12],
contrary to her assertion, a detailed statistical
In spite of the above information, unfortunately,
analysis clearly showed that there was NO dif-
Biousse and colleagues [1, 58, 59] insist that there
ference in the frequency of diabetes or other
are several flaws in our study [12], which invalidate
vascular risk factors between the two groups. I
our conclusions. I feel it is important to discuss
fail to understand where that idea originated. It
their comments fully to place them in proper per-
is simply not true.
spective. Following are my [57] responses to their
(b) She went on to state [58] that most statistical
comments:
tests (such as those used in this paper) cannot
1. They [1] stated: Their patients were not random- support an absence of difference, as is argued. It
ized, and the untreated group had more vascular is also important not to place too much impor-
risk factors making it difficult to interpret their tance on only the p-value. From the scientific
results. Both statements are incorrect. In our study point of view, this rejection of well-established
[12], the patients were indeed randomized, but for statistical methods contradicts standard scien-
the reasons discussed above and in the original tific practice is a completely invalid statement.
paper [12], we were forced to use patient random- 4. Biousse [59] stated: A number of well-designed
ization, rather than conventional randomization. studies in the neurologic literature have shown that
The critical aim of conventional randomization, corticosteroids do not improve outcome of patients
as discussed above, is to have comparable treated with acute arterial or venous cerebral ischemia.
and untreated groups at baseline in demographic Indeed it has even been suggested that corticoster-
and clinical characteristics. Our [12] patient ran- oid therapy can be detrimental to patients with acute
domization study fulfilled all the vital criteria of cerebral ischemia and should not be prescribed. As
conventional randomization, as discussed above discussed above, there is fundamental flaw in equat-
in detail. ing stroke and NA-AION with regard to their patho-
2. They [1] stated: In addition, the earlier resolution geneses and treatment, and it is like comparing
of disc edema seen in the treated group is particu- apples and oranges.
larly subject to selection bias, since diabetics were
much less likely to be treated, and the authors own Therefore, in conclusion, when all these facts are put
data demonstrate that edema persists longer in dia- together, one can conclude that this patient choice
betics. This statement is invalid for two reasons. controlled study fulfilled the most crucial criteria of a
First, as discussed above, there was no difference conventional randomization clinical trial, i.e. the
(p=0.126) in the diabetic patients between the treated and untreated groups were comparable in demo-
treated and untreated groups. Second, they had cited graphic and clinical characteristics at baseline. Therefore,
our study [56] dealing with optic disc edema in dia- the results of this study must be valid. Unfortunately,
there is a tremendous amount of unnecessary genuflex- of the diagnosis, because these patients were mixed
ion to conventional randomization or class I stud- with other patients seen in the entire ophthalmology
ies; Those studies are not always without flawed department. Finally, the fields were graded and judged
results. by three neuro-ophthalmologists, independently, in
masked fashion, without any knowledge of the patients.
Thus, the entire visual field information is based on
completely masked collection and interpretation of the
The Study Data Were Not Collected data.
in a Masked Fashion This discussion, ought to answer all the objections
to our study raised by critics [12]. Two additional fac-
I expected this criticism right from the start of the tors have played an important role in the controversy
study. Therefore, I took all possible steps to ensure that about corticosteroid therapy in the treatment of
the data were collected in a masked fashion, as well as NA-AION.
during data analysis. These steps are discussed at
length in the paper [12].
The following are proof that the collected data had
no bias. Pathogenesis of NA-AION
It seems the primary reason for controversy on

Visual Acuity NA-AION management is lack of understanding of its
pathogenesis. It is invariably stated, as Biousse [59]
The results of our study [12] exactly mirrored the rele- did, that the pathogenesis of NA-AION is unknown.
vant findings of the Ischemic Optic Neuropathy This contention is based on a lack of an in-depth
Decompression Trial, a randomized and masked knowledge of: (1) blood supply of the ONH [6163]
study [10]; in that study, in the untreated eyes seen (see Chap. 3) (2) blood flow in the optic nerve head
within 2weeks after the onset of NA-AION with VA of and the factors that influence it [74] (see Chap. 5), (3)
20/70 or worse, visual acuity improved in 43% of various local and systemic risk factors [28, 29] (see
untreated eyes In our study [12], in an identical group Chap. 14), and (4) the role of nocturnal arterial
of NA-AION patients, in the untreated group visual hypotension (see Chap. 14). Most of the postulated
acuity improved in 41%. In both studies, the visual acu- theories on pathogenesis suffer from these fundamen-
ity kept improving only for the first 6 months. Had tal flaws. The pathogenesis of NA-AION and the flaws
there been any bias in the collection of visual acuity in the various postulated theories are discussed at
data, the results of our study [12] would not have been length in Chap. 14. Based on currently available scien-
absolutely identical. This is the most convincing proof tific information, it can be concluded that available
that the visual acuity data in my study were unbiased. evidence shows that the pathogenesis of NA-AION is
Moreover, the odds ratio for improvement in visual complex but not, as often stated, unknown [29].
acuity from initial visit in the corticosteroid treated
group relative to that in the natural history group was
4.45 (95% CI: 2.039.75; p=0.0002) at 3months, 3.39
(95% CI: 1.627.11; p=0.001) at 6 months and 4.06 Treating Elderly Patients with High
(95% CI: 1.928.57) at 1year. This is not at all a minor Dose Systemic Corticosteroids
difference when the two groups were basically similar.
Many physicians are nervous of treating patients, par-
ticularly the elderly, with high dose corticosteroid
Visual Fields therapy. Over 45years, I have treated several thousand
patients with giant cell arteritis, scleritis, uveitis, reti-
As discussed at length in the original paper [12], these nal vasculitis and a host of other acute ophthalmic con-
were plotted by perimetrists who were totally unaware ditions, with high dose corticosteroid therapy. From
Intravitreal Triamcinolone Acetonide Therapy 407
that experience, I have found that fear about corticos- however, that study had some notable flaws [77].
teroid therapy is overblown, provided the patients are Briefly, these include:
followed closely and meticulously. In our corticoster-
1. Their study was based on only four eyes.
oid therapy study [12] in NA-AION, 27% were diabet-
2. Two large natural history studies have shown spon-
ics and they were managed without any problem.
taneous visual acuity improvement in 4143% of
eyes with NA-AION [10, 11].
3. More importantly, none of the eyes in the study
When and How to Treat NA-AION by Kaderli et al. [76] showed improvement in
visual fields and all had altitudinal visual field
Patients with Corticosteroid Therapy defects. Studies have shown that in NA-AION
and arteritic AION apparent visual acuity
Given the mechanism and rationale of visual improve- improvement without visual field improvement is
ment with corticosteroid therapy discussed above, it due to the patient learning to fixate eccentrically,
would seem that the sooner the treatment is started, the rather than being a genuine visual improvement
better the chance of improvement; the longer the [11, 30, 53]. Kaderli etal.s [76] patients had alti-
axonal ischemia persists, the more axons are likely to tudinal visual field defects passing through fixa-
be damaged permanently. tion; it is well-known that such patients soon
learn to fixate eccentrically and can see much bet-
ter without actual improvement. This eccentric
fixation may explain why in their study the visual
Secret of Corticosteroid Therapy acuity of the patients apparently improved, while
the visual fields did not.
Over a period of almost five decades having treated
It is impossible to judge the effectiveness of intravit-
several thousand patients with corticosteroid therapy
real modes of treatment in studies containing only four
for a variety of conditions (see above), I have found
eyes when 4143% of NA-AION eyes show spontane-
that the most effective way to use corticosteroid ther-
ous visual acuity improvement. Most importantly,
apy is to hit hard at the beginning and then taper down.
intravitreal injection in NA-AION eyes can be harm-
Corticosteroid therapy has been given timidly for
ful. Optic nerve head circulation depends upon the per-
NA-AION in some studies, too small a dose, for too
fusion pressure (mean blood pressure minus intraocular
short a period. This timidity has led to the prevailing
pressure). Intravitreal injection increases the volume in
misconception that corticosteroid therapy does not
the eyeball, thereby resulting in a transient rise of
help patients with NA-AION. The treatment protocol
intraocular pressure initially. In addition, there are
for systemic corticosteroid therapy used in our study
many reports showing a substantial number of patients
[12] is essential to get the required benefit, rather than
(as much as 4050%) developing a rise in intraocular
an inadequate small dose for too short a time.
pressure a few days or weeks after intravitreal triamci-
nolone, or even developing glaucoma requiring treat-
ment. In NA-AION, with already precarious optic
nerve head circulation, even a small rise in intraocular
Intravitreal Triamcinolone
pressure for any reason can further compromise the
Acetonide Therapy circulation and result in further visual loss. Oral corti-
costeroid therapy for NA-AION in our study [12], by
It could be argued that intravitreal corticosteroid ther- contrast, had no effect on intraocular pressure during a
apy may be safer and equally effective in NA-AION. short-term treatment. Thus, one cannot equate oral and
There have recently been two contradictory small stud- intravitreal corticosteroid therapy in NA-AION.
ies on this topic. Jonas etal. [75] in 3 patients, found Development of triamcinolone-associated crystalline
that it had no beneficial effect on visual acuity. Kaderli maculopathy has also been reported [78]. All this indi-
etal. [76] in four eyes, reported visual acuity improve- cates that intravitreal corticosteroid therapy may actu-
ment, but without any improvement in visual fields; ally be harmful in NA-AION.
Use of Intravitreal Vascular Endothelial asymptomatic optic disc edema is an early sign of
Growth Factor Inhibitory Therapy NA-AION. More recently, we [13] described the clini-
cal entity of Incipient nonarteritic anterior ischemic
optic neuropathy in 60 eyes, in which, again,
Currently, the use of vascular endothelial growth fac- NA-AION patients were seen initially with asymptom-
tor (VEGF) for a variety of ocular disorders is very atic optic disc edema and 25% of them later progressed
popular. This has also been tried in NA-AION. A case to classical NAION, with a median time of progression
report by Bennett etal. [79] claimed visual improve- of 5.8 (interquartile 3.210.1) weeks from the initial
ment after an intravitreal injection of Bevacizumab diagnosis. Thus, contrary to the universal impression,
(Avastin) 3weeks after the onset of NA-AION in one the time of visual loss is not always the time of onset of
eye. The authors claim that bevacizumab a VEGF NA-AION it is only the time when the patient first
inhibitory drug improved visual acuity in the eye by notices it or develops visual symptoms. In these two
reducing optic disc edema. But it is impossible to judge cases [81, 82] the intravitreal injection most probably
the effectiveness of a mode of treatment from one eye started the chain of events. It initially produced
when 4143% of NA-AION eyes show spontaneous Incipient nonarteritic anterior ischemic optic neuropa-
visual acuity improvement [10, 11]. Moreover, recently thy; however, the patients were seen by the ophthal-
Pece etal. [80] reported three patients with NA-AION mologist only when they noticed visual loss 12weeks
who had intravitreal ranibizumab injection 12 days later.
after onset. They found that intravitreal ranibizumab, I [84] have discussed at length elsewhere the role of
although it reduced the optic disc edema, did not result short-term, severe, sustained ocular hypertension fol-
in any functional improvement (most probably due to lowing intravitreal injections of anti-VEGF or triamci-
rise of intraocular pressure see Chap. 15). nolone in the development of NA-AION (see Chap. 15).
There have recently been two disturbing reports This shows that use of intravitreal injections for treat-
where intravitreal bevacizumab injection actually ment of NA-AION has the potential of being harmful
resulted in the development of NA-AION. Hosseini rather than helpful. Also, the pathogenesis of develop-
and Razeghinejad [81] reported a 72-year old woman ment of NA-AION following their use in age-related
who woke up with visual loss 1week after intravitreal macular degeneration and other conditions is discussed
injection of bevacizumab and was diagnosed to have in Chap. 15. As pointed out previously in Chap. 15, age-
NA-AION. Ganssauge etal. [82] report a 51-year-old related macular degeneration and diabetic macular
male who recognized visual field defects 2 weeks edema occur in persons with predisposing vascular risk
after injection of intravitreal bevacizumab for choroidal factors for development of NA-AION. A rise of intraoc-
neovascularisation, and was diagnosed as having ular pressure in such persons may precipitate develop-
NA-AION. Recently, also a neuro-ophthalmologist ment of NA-AION. Development of NA-AION
wrote me about seeing a patient who developed following intravitreal injection of anti-VEGF drugs may
NA-AION following intravitreal injection of an anti- be far more common than only the three cases men-
VEGF drug in an eye with pseudoxanthoma elasticum tioned above (see Chap. 15). In view of all these consid-
and choroidal neovascularisation. In the first two cases erations, when treating age-related macular degeneration
[81, 82], it could be argued that because there was a or diabetic macular edema with intravitreal injections of
time lag of 1week and 2weeks respectively between anti-VEGF or triamcinolone, on a risk/benefit ratio, it
the intravitreal injection of bevacizumab and the devel- may be prudent to avoid any immediate or prolonged
opment of NA-AION, the latter could not be attrib- rise of intraocular pressure by taking precautionary
uted to the former. Since Ganssauge etal.s [82] patient measures and measuring intraocular pressure frequently.
recognized the visual field defect, the possibility that For persons who already have had NA-AION in one
it was there before and he was unaware of it cannot be eye, the risk of the second eye developing NA-AION is
ruled out (see Chap. 16). That is not unknown in much higher than for those who have never had
patients with NA-AION. There is an explanation for NA-AION. Thus, a high intraocular pressure can pre-
the delay between the intravitreal injection and devel- cipitate development of NA-AION in persons with pre-
opment of visual loss due to NA-AION in these two disposing risk factors and can worsen the visual loss in
cases [81, 82]. In 1981, I [83] reported that those who already have NA-AION. So intravitreal
Management of Incipient NA-AION 409
injections of anti-VEGF agents or of triamcinolone have reduction of blood flow in the capillary bed, result-
the potential to be harmful a factor not realized in the ing in ischemic damage. Moreover, these agents
current enthusiasm for their use. Moreover, anti-VEGF can also cause vasoconstriction in the optic nerve,
agents do rarely have systemic complications, e.g., acute making it more susceptible to ischemic damage.
elevation of systemic blood pressure, cerebrovascular Thus, this mode of treatment is actually harmful.
accidents, myocardial infarctions; even deaths have 3. Use of vasodilators actually lowers the systemic
been reported in some cases [85]. blood pressure. Currently almost all of the drugs
used to treat arterial hypertension are vasodilators.
Lowering the blood pressure can reduce the perfu-
sion pressure in the optic nerve head and that makes
Miscellaneous Other Treatments it susceptible to NA-AION (see Chap. 14).
Advocated from Time to Time 4. Invasive procedures, e.g., transvitreal optic neuro-
tomy, vitrectomy and release of epipapillary vitre-
ous traction, transcorneal electrical stimulation,
It is well-known that a disease which has no treatment
retinal laser photocoagulation, transcorneal electri-
has many treatments each hailed with enthusiasm but
cal stimulation, and heparin-induced extracorporeal
then found to be ineffective, or even harmful. NA-AION
LDL/fibrinogen precipitation, have no scientific
has been a classical example of that. Atkins and col-
rationale and ignore the basic pathogenesis of
leagues [1] have recently listed and discussed the vari-
NA-AION. Some of them can be harmful.
ous treatments which have been advocated from time to
5. The use of brimonidine tartrate was advocated under
time. Those include anticoagulants, thrombolytics,
the belief that it is a neuroprotector. However, there
vasodilators, pressor agents (norepinephrine), brimoni-
is no evidence that it is, or that it has any beneficial
dine tartrate, hyperbaric oxygen, diphenylhydantoin,
role in NA-AION. In fact, the company making bri-
heparin-induced extracorporeal LDL/fibrinogen pre-
monidine tartrate started a multicenter clinical trial
cipitation, transvitreal optic neurotomy, vitrectomy and
in NA-AION which was quickly abandoned.
release of epipapillary vitreous traction, transcorneal
electrical stimulation, and retinal laser photocoagula-
tion. Almost all of these modes of treatment lack scien-
tific rationale for management of NA-AION. I have the Management of Incipient NA-AION
following comments about them.
1. Treatments such as aspirin, anticoagulants and throm- If a patient is diagnosed as having incipient AION, the
bolytics were advocated under the misconception physician should immediately make every attempt to
that NA-AION, like stroke, is a thromboembolic dis- pinpoint risk factors and try to modify them promptly,
order, which, as discussed in Chap. 14, is not true. if possible, to reduce the risk of progression to classical
2. The use of pressor agents (norepinephrine) has NA-AION. Measures should include: (if possible,
been advocated and widely practiced under the always with consultation with patients physician) shift-
belief that raising systemic blood pressure by these ing the taking of blood pressure lowering medicines
agents would improve the blood flow in the tissues from night or evening to morning, stopping any drugs
(e.g., optic nerve head in the case of NA-AION), that could cause a decrease in blood pressure during
and would thereby be beneficial to prevent isch- sleep (sleep medications, sedatives, alcohol, pain medi-
emia. I discussed the role of pressor agents else- cations, alpha 1 blockers used for benign prostatic
where in relation to optic nerve ischemia [86]. hypertrophy in men and bladder problems in women,
These agents elevate blood pressure by constricting erectile dysfunction drugs [87, 88]), and immediate
the terminal arterioles. Constriction of the terminal evaluation for sleep apnea, if the history indicates it. It
arterioles has a paradoxical effect proximal to the would also be advisable to try to lower intraocular pres-
arterioles there is increased blood pressure (which sure, if it is high or borderline high, but it is advisable to
gives a false sense of security of having high blood avoid using beta-blocker eye drops, which produce
pressure) but distal to the terminal arterioles (due nocturnal arterial hypotension [89], a risk factor known
to constriction of the arterioles) there is marked to precipitate the development of NA-AION.
Other Important Considerations view of this, management of nocturnal arterial

in Management of NA-AION hypotension is an important step both in the manage-
ment of NA-AION and in the prevention of its devel-
opment in the second eye. Therefore, I strongly
When dealing with my patients, I am always reminded recommend that when a patient is at risk of developing
of what Sir William Osler stated: The good physician ocular and optic nerve head ischemic and vascular dis-
treats the disease; the great physician treats the patient orders, or has the following: (a) NA-AION or history
who has the disease. of NA-AION in one eye, (b) active giant cell arteritis,
The usual advice given by ophthalmologists and (c) normal-tension glaucoma, (d) occlusion or severe
neuro-ophthalmologists to NA-AION patients is that stenosis of internal carotid artery, (e) low central reti-
nothing can be done to help them. Having dealt with nal artery pressure or (f) chronic optic disc edema due
about 1,350 patients with NA-AION and having inves- to any cause, the treating physician should be made
tigated various aspects of NA-AION over the years, I aware of the potential risks of intensive arterial
find that is an inadequate and incorrect response. Apart hypotensive therapy, particularly giving that drug in
from treating the disease, my primary objective has the evening.
always been to treat the patient who has the disease. In
this connection, there are important measures which
can be taken, and patients must be informed. Management of Nocturnal Arterial Hypotension
After conducting 24-h ambulatory blood pressure

monitoring in more than 700 patients with NA-AION
Reduction of Risk Factors and glaucomatous optic neuropathy, I have found that
marked nocturnal arterial hypotension seems to play
NA-AION is a multifactorial disease and many risk fac- an important role in both NA-AION and glaucoma-
tors contribute to it (see Chap. 14). The correct strategy tous optic neuropathy. During these studies, NA-AION
is to try to reduce as many risk factors as possible to patients who have suffered visual loss as a result of
reduce the risk of NA-AION in the second eye [32, 33, abnormal nocturnal arterial hypotension have asked
90] or any further episode in the same eye [91]. In Chap. me about management of it to prevent any further
14 various predisposing risk factor have been discussed. problems. For information on that, I reviewed the
Every patient with NA-AION must be educated about management of orthostatic hypotension, which has
those risk factors and advised to reduce them or take some similar features to marked nocturnal arterial
measures to deal with them, e.g., hyperlipidemia, sleep hypotension. The drugs currently most used for
apnoea, diabetes mellitus, among others. orthostatic hypotension are fludrocortisone and
As discussed in Chap. 14, nocturnal arterial midodrine.
hypotension is a major risk factor in NA-AION Fludrocortisone is a synthetic adrenocortical steroid
patients who already have predisposing risk factors. possessing very potent mineralocorticoid properties
Since the 1960s many highly potent drugs with arte- and high glucocorticoid activity; it is used only for its
rial hypotensive effects have emerged, to treat arterial mineralocorticoid effects. It is the first line of treat-
hypertension, other cardiovascular diseases, benign ment for orthostatic hypotension. Its most adverse
prostatic hyperplasia and other diseases; those drugs reactions are caused by the drugs mineralocorticoid
are currently widely used. It may not be coincidental activity (retention of sodium and water) and include
that the incidence of NA-AION has progressively arterial hypertension, edema, cardiac enlargement,
increased since the 1960s, so that it has now become a congestive heart failure, potassium loss, and
common visually disabling disease. This suggests that hypokalemic alkalosis.
NA-AION may actually be emerging as an iatrogenic Midodrine is a vasopressor/antihypotensive agent.
disease, stemming from the aggressive use of those An alpha1-agonist, it exerts its actions via activation of
potent arterial hypotensive agents now available. In the alpha-adrenergic receptors of the arteriolar and
Other Important Considerations in Management of NA-AION 411
venous vasculature, producing an increase in vascular visual field, or both. (Evaluation and extent of visual
tone and elevation of blood pressure. The most poten- functional disability produced by the central and
tially serious adverse reaction associated with Midodrine peripheral visual field loss is discussed at length in
hydrochloride therapy is marked elevation of supine Chap. 16.)
arterial blood pressure (supine arterial hypertension).
Systolic pressure may go as high as about 200mmHg.
Pathak et al. [92] investigated adverse drug reac- Visual Disability Due to Scotomas in the Central
tions to fludrocortisone and/or midodrine used for Part of the Visual Field
orthostatic hypotension. The study showed a high fre-
quency of adverse drug reactions (especially serious This depends upon the site and size of the scotoma. A
and unexpected adverse drug reactions) with antihy- central scotoma results in a fall in visual acuity the
potensive drugs. They concluded that it strongly sug- larger the scotoma, the worse the visual acuity. These
gests the need for a better evaluation of the safety patients should be taught how to use eccentric fixation
profile of antihypotensive drugs and improvement in to see better in the near; I have been surprised to find
the summary of product characteristics. how often this helps the patient. A paracentral sco-
Maule etal. [93] in a review article, also discussed toma which spares the central fixation does not inter-
the drugs used in the management of orthostatic fere with the central visual acuity but can be equally
hypotension. Despite the wide use of fludrocortisone disabling in some cases, particularly when located in
and midodrine, multicentre, randomised and controlled the inferior field, because that is the field of vision
studies for the treatment of orthostatic hypotension are which is most commonly used. Unfortunately, in
still scarce and limited to a few agents and groups of NA-AION paracentral scotomas tend to be more com-
patients. The pharmacological management of orthos- mon in the inferior than superior field of vision. To
tatic hypotension substantially improves the quality of improve the central visual acuity, a variety of very
life of patients, but it may be problematic because of good low vision devices now available may be helpful
associated side-effects. to these patients, to let them read and see better
Therefore, unfortunately, we do not have a safe and centrally.
effective drug to treat the nocturnal arterial hypoten-
sion which plays an important role in the optic nerve
head ischemic disorders [68, 69] (both NA-AION and Visual Disability Due to Peripheral
glaucomatous optic neuropathy). Visual Field Defect
It is well established that the constant tracking pro-

vided by the peripheral visual fields is essential for
How to Cope with the Visual Disability
sensory input to our day to day activity. For example,
Caused by NA-AION the peripheral visual field is vital for walking, driving
and navigating in this world. In view of that, to
This is a very important issue. Unfortunately, many assess the visual function disability produced by
ophthalmologists somehow see their role as limited to NA-AION, it is important to have complete informa-
prescribing a treatment or doing a procedure; if that is tion about the peripheral visual fields and any impair-
not available for a disease, the patient is simply told ment in them. Manual kinetic perimetry provides the
that nothing can be done. That is an inadequate way most reliable information about this (all the way to
to deal with people who have developed permanent about 8090 temporally, 70 inferiorly, 6070 nasally
visual disability. Most importantly, they need coun- and 5060 superiorly), but unfortunately the currently
seling on how to cope with the visual disability and to commonly used automated perimetry does not provide
lead as full a life as possible. The visual disability any information beyond 2430 in the periphery. A
may be due to scotomas in the central visual field or person with central scotoma but with normal periph-
due to loss of one or the other part of the peripheral eral visual field can lead a fairly full, mobile life,
though not able to read, write, drive and see properly in

the center. For example, a patient with a large absolute
central scotoma in one eye and no vision in the other
eye is able to walk around without any help and is
relatively happy with his visual function (Fig. 17.9).
By contrast, patients with bilateral inferior altitudinal
defects (Fig. 17.10) or with markedly constricted
visual fields are notably disabled, stumbling, and
unable to drive or navigate, even when they have
perfectly normal visual acuity due to sparing central
fixation. Obviously, this has important functional and
practical implications. Disability Boards tend to deny
eligibility to persons with NA-AION who can see
20/20 in both eyes, in spite of the fact that they have
complete loss of lower half of visual fields in both eyes
Fig.17.9 Visual fields (plotted with a Goldmann perimeter) in
(Fig. 17.10); from a practical point of view they are an eye with NA-AION show a huge absolute central scotoma
legally blind and unable to drive. The ophthalmologist with normal peripheral visual field with V4e isopter
may need to act as an expert advocate for such patients.
Such patients may need to use a white cane to help
them to navigate. extremely important, even though the ophthalmolo-
Thus, in the management of NA-AION patients, gist may not be able to restore their lost vision.
counseling them how to reduce predisposing risk Referral of some of them to a low-vision clinic can be
factors and how to cope with the visual disability are helpful.
Fig.17.10 Visual fields (plotted with a Goldmann perimeter) in a patient with NA-AION show bilateral inferior altitudinal defect.
References 413
References ple visual function measurements. Arch Ophthalmol.

1991;109(5):66771.
19. Spoor TC, Wilkinson MJ, Ramocki JM. Optic nerve sheath
1. Atkins EJ, Bruce BB, Newman NJ, Biousse V. Treatment of decompression for the treatment of progressive nonarteritic
nonarteritic anterior ischemic optic neuropathy. Surv ischemic optic neuropathy. Am J Ophthalmol. 1991;111(6):
Ophthalmol. 2010;55(1):4763. 7248.
2. Sergott RC, Cohen MS, Bosley TM, Savino PJ. Optic nerve 20. Spoor TC, McHenry JG, Lau-Sickon L. Progressive and
decompression may improve the progressive form of nonar- static nonarteritic ischemic optic neuropathy treated by optic
teritic ischemic optic neuropathy. Arch Ophthalmol. nerve sheath decompression. Ophthalmology. 1993;
1989;107(12):174354. 100(3):30611.
3. Hayreh SS. The role of optic nerve sheath fenestration in 21. Jablons MM, Glaser JS, Schatz NJ, Siatkowski RM, Tse DT,
management of anterior ischemic optic neuropathy. Arch Kronish JW. Optic nerve sheath fenestration for treatment of
Ophthalmol. 1990;108(8):10635. progressive ischemic optic neuropathy. Results in 26 patients.
4. Hayreh SS. Pathogenesis of oedema of the optic disc (papil- Arch Ophthalmol. 1993;111(1):847.
loedema): a preliminary report. Br J Ophthalmol. 1964;48: 22. Glaser JS, Teimory M, Schatz NJ. Optic nerve sheath fenes-
52243. tration for progressive ischemic optic neuropathy. Results in
5. Hayreh SS. Optic disc edema in raised intracranial pressure. second series consisting of 21 eyes. Arch Ophthalmol.
V. Pathogenesis. Arch Ophthalmol. 1977;95(9):155365. 1994;112(8):104750.
6. Hayreh SS. Optic disc edema in raised intracranial pressure. 23. Yee RD, Selky AK, Purvin VA. Outcomes of optic nerve
VI. Associated visual disturbances and their pathogenesis. sheath decompression for nonarteritic ischemic optic neu-
Arch Ophthalmol. 1977;95(9):156679. ropathy. J Neuroophthalmol. 1994;14(2):706.
7. Hayreh SS. The sheath of the optic nerve. Ophthalmologica. 24. Kelman SE. Intravitreal triamcinolone or bevacizumab for
1984;189(12):5463. nonarteritic anterior ischemic optic neuropathy: Do they
8. Hayreh SS. Anterior ischaemic optic neuropathy. III. merit further study? J Neuroophthalmol. 2007;27(3):1613.
Treatment, prophylaxis, and differential diagnosis. Br J 25. Hayreh SS. The role of aspirin in non-arteritic anterior
Ophthalmol. 1974;58(12):98189. ischaemic optic neuropathy. Neuro-Ophthalmol. 2010;
9. Heidemann SR, Kaufman D. Axonal transport and anterior 34(1):15.
ischemic optic neuropathy. Arch Ophthalmol. 1991;109(7): 26. Hayreh SS. Anterior ischemic optic neuropathy. Arch
9167. Neurol. 1981;38(11):6758.
10. The Ischemic Optic Neuropathy Decompression Trial 27. Hayreh SS. Inter-individual variation in blood supply of the
Research Group. Optic nerve decompression surgery for optic nerve head. Its importance in various ischemic disor-
nonarteritic anterior ischemic optic neuropathy (NAION) is ders of the optic nerve head, and glaucoma, low-tension
not effective and may be harmful. JAMA. 1995; glaucoma and allied disorders. Doc Ophthalmol.
273(8):62532. 1985;59(3):21746.
11. Hayreh SS, Zimmerman MB. Nonarteritic anterior ischemic 28. Hayreh SS. Acute ischemic disorders of the optic nerve:
optic neuropathy: natural history of visual outcome. pathogenesis, clinical manifestations and management.
Ophthalmology. 2008;115(2):298305. Ophthalmol Clin North Am. 1996;9(3):40742.
12. Hayreh SS, Zimmerman MB. Non-arteritic anterior isch- 29. Hayreh SS. Ischemic optic neuropathy. Prog Retin Eye Res.
emic optic neuropathy: role of systemic corticosteroid ther- 2009;28(1):3462.
apy. Graefes Arch Clin Exp Ophthalmol. 2008;246(7): 30. Hayreh SS, Zimmerman B, Kardon RH. Visual improve-
102946. ment with corticosteroid therapy in giant cell arteritis. Report
13. Hayreh SS, Zimmerman MB. Incipient nonarteritic anterior of a large study and review of literature. Acta Ophthalmol
ischemic optic neuropathy. Ophthalmology. 2007;114(9): Scand. 2002;80(4):35567.
176372. 31. Hayreh SS, Zimmerman B. Visual deterioration in giant cell
14. Hayreh SS, Zimmerman MB. Nonarteritic anterior ischemic arteritis patients while on high doses of corticosteroid ther-
optic neuropathy: clinical characteristics in diabetic patients apy. Ophthalmology. 2003;110(6):120415.
versus nondiabetic patients. Ophthalmology. 32. Beck RW, Hayreh SS, Podhajsky PA, Tan ES, Moke PS.
2008;115(10):181825. Aspirin therapy in nonarteritic anterior ischemic optic neu-
15. Flaharty PM, Sergott RC, Lieb W, Bosley TM, Savino PJ. ropathy. Am J Ophthalmol. 1997;123(2):2127.
Optic nerve sheath decompression may improve blood flow 33. Newman NJ, Scherer R, Langenberg P, Kelman S, Feldon S,
in anterior ischemic optic neuropathy. Ophthalmology. Kaufman D, etal. The fellow eye in NAION: report from the
1993;100(3):297302. discussion 303295. ischemic optic neuropathy decompression trial follow-up
16. Hayreh SS, Beach KW. Optic nerve sheath decompression study. Am J Ophthalmol. 2002;134(3):31728.
may improve blood flow in anterior ischemic optic neuropa- 34. Hayreh SS, Podhajsky PA, Zimmerman B. Nonarteritic
thy: discussion. Ophthalmology. 1993;100(3):3035. anterior ischemic optic neuropathy: time of onset of visual
17. Manor RS. Nonarteritic ischemic optic neuropathy in identi- loss. Am J Ophthalmol. 1997;124(5):6417.
cal female twins: improvement of visual outcome in one by 35. Dagenais GR, Yi Q, Lonn E, Sleight P, Ostergren J, Yusuf S.
optic nerve decompression. Arch Ophthalmol. 1990;108(8): Impact of cigarette smoking in high-risk patients participat-
10678. ing in a clinical trial. A substudy from the Heart Outcomes
18. Kelman SE, Elman MJ. Optic nerve sheath decompression Prevention Evaluation (HOPE) trial. Eur J Cardiovasc Prev
for nonarteritic ischemic optic neuropathy improves multi- Rehabil. 2005;12(1):7581.
36. Hayreh SS, Jonas JB, Zimmerman MB. Nonarteritic anterior 54. Miller GR, Smith JL. Ischemic optic neuropathy. Am J
ischemic optic neuropathy and tobacco smoking. Ophthalmol. 1966;62(1):10315.
Ophthalmology. 2007;114(4):8049. 55. Foulds WS. Visual disturbances in systemic disorders. Optic
37. Salomon O, Huna-Baron R, Kurtz S, Steinberg DM, Moisseiev neuropathy and systemic disease. Trans Ophthalmol Soc
J, Rosenberg N, etal. Analysis of prothrombotic and vascular UK. 1969;89:125146.
risk factors in patients with nonarteritic anterior ischemic 56. Hayreh SS, Zimmerman MB. Optic disc edema in non-arter-
optic neuropathy. Ophthalmology. 1999;106(4):73942. itic anterior ischemic optic neuropathy. Graefes Arch Clin
38. Abu-Amero KK, Bosley TM. Prothrombotic and atheroscle- Exp Ophthalmol. 2007;245(8):110721.
rotic risk factors lack significance in NAION patients har- 57. Hayreh SS. Non-arteritic anterior ischemic optic neuropa-
bouring mitochondrial DNA mutations. Br J Ophthalmol. thy: role of systemic corticosteroid therapy. Surv Ophthalmol.
2006;90(1):11920. 2010;55(4):399400.
39. Hayreh SS. Non-arteritic anterior ischemic optic neuropathy 58. Biousse V, Bruce BB, Newman NJ. Non-arteritic anterior
and thrombophilia. Graefes Arch Clin Exp Ophthalmol. ischemic optic neuropathy. Role of systemic corticosteroid
2009;247(5):57781. therapy. Surv Ophthalmol. 2010;55(4):4001.
40. Bull GM. Meteorological correlates with myocardial and 59. Lee AG, Biousse V. Should steroids be offered to patients
cerebral infarction and respiratory disease. Br J Prev Soc with nonarteritic anterior ischemic optic neuropathy? J
Med. 1973;27(2):10813. Neuroophthalmol. 2010;30(2):1938.
41. Bull GM, Morton J. Environment, temperature and death 60. Hayreh SS. Role of steroid therapy in nonarteritic anterior
rates. Age Ageing. 1978;7(4):21024. ischemic optic neuropathy. J Neuroophthalmol. 2010;
42. Clark CV. Seasonal variation in incidence of brachial and 30(4):388.
femoral emboli. Br Med J (Clin Res Ed). 1983; 61. Hayreh SS. Blood supply of the optic nerve head and its role
287(6399):1109. in optic atrophy, glaucoma, and oedema of the optic disc. Br
43. Hayreh SS, Joos KM, Podhajsky PA, Long CR. Systemic J Ophthalmol. 1969;53(11):72148.
diseases associated with nonarteritic anterior ischemic optic 62. Hayreh SS. The 1994 Von Sallman lecture: the optic nerve
neuropathy. Am J Ophthalmol. 1994;118(6):76680. head circulation in health and disease. Exp Eye Res.
44. Kosmorsky G, Straga J, Knight C, Dagirmanjian A, Davis 1995;61(3):25972.
DA. The role of transcranial Doppler in nonarteritic isch- 63. Hayreh SS. The blood supply of the optic nerve head and the
emic optic neuropathy. Am J Ophthalmol. 1998; evaluation of it: myth and reality. Prog Retin Eye Res.
126(2):28890. 2001;20(5):56393.
45. Stiebel-Kalish H, Hasanreisoglu M, Leibovici L. Aspirin 64. Hayreh SS. Fluids in the anterior part of the optic nerve in
following non-arteritic ischaemic optic neuropathy: a sys- health and disease. Surv Ophthalmol. 1978;23(1):125.
tematic review and meta-analysis. Neuro-Ophthalmology. 65. McLeod D, Marshall J, Kohner EM. Role of axoplasmic
2010;34(1):149. transport in the pathophysiology of ischaemic disc swelling.
46. Kupersmith MJ, Frohman L, Sanderson M, Jacobs J, Br J Ophthalmol. 1980;64(4):24761.
Hirschfeld J, Ku C, etal. Aspirin reduces the incidence of 66. Beck RW, Servais GE, Hayreh SS. Anterior ischemic optic
second eye NAION: a retrospective study. J Neuroophthalmol. neuropathy. IX Cup disc ratio role pathogenesis.
1997;17(4):2503. Ophthalmology. 1987;94(11):15038.
47. Salomon O, Huna-Baron R, Steinberg DM, Kurtz S, Seligsohn 67. Hayreh SS, Zimmerman MB. Nonarteritic anterior ischemic
U. Role of aspirin in reducing the frequency of second eye optic neuropathy: refractive error and its relationship to cup/
involvement in patients with non-arteritic anterior ischaemic disc ratio. Ophthalmology. 2008;115(12):227581.
optic neuropathy. Eye (Lond). 1999;13(Pt 3a):3579. 68. Hayreh SS, Zimmerman MB, Podhajsky P, Alward WL.
48. Beck RW, Hayreh SS. Role of aspirin in reducing the fre- Nocturnal arterial hypotension and its role in optic nerve
quency of second eye involvement in patients with non-arter- head and ocular ischemic disorders. Am J Ophthalmol.
itic anterior ischaemic optic neuropathy (Correspondence). 1994;117(5):60324.
Eye (Lond). 2000;14(Pt 1):118. 69. Hayreh SS, Podhajsky P, Zimmerman MB. Role of noctur-
49. Johnson LN, Gould TJ, Krohel GB. Effect of levodopa and nal arterial hypotension in optic nerve head ischemic disor-
carbidopa on recovery of visual function in patients with ders. Ophthalmologica. 1999;213(2):7696.
nonarteritic anterior ischemic optic neuropathy of longer 70. Kaur C, Foulds WS, Ling EA. Bloodretinal barrier in
than six months duration. Am J Ophthalmol. 1996;121(1): hypoxic ischaemic conditions: basic concepts, clinical fea-
7783. tures and management. Prog Retin Eye Res. 2008;27(6)):
50. Johnson LN, Guy ME, Krohel GB, Madsen RW. Levodopa 62247.
may improve vision loss in recent-onset, nonarteritic ante- 71. Dobrogowska DH, Lossinsky AS, Tarnawski M, Vorbrodt
rior ischemic optic neuropathy. Ophthalmology. 2000;107(3): AW. Increased bloodbrain barrier permeability and endothe-
5216. lial abnormalities induced by vascular endothelial growth
51. Beck RW. Does Levodopa improve visual function in factor. J Neurocytol. 1998;27((3):16373.
NAION? Ophthalmology. 2000;107(8):14314. 72. Croll SD, Ransohoff RM, Cai N, Zhang Q, Martin FJ,
52. Hayreh SS. Does Levodopa improve visual function in Wei T, etal. VEGF-mediated inflammation precedes angio-
NAION? Ophthalmology. 2000;107(8):14348. genesis in adult brain. Exp Neurol. 2004;187(2):
53. Hayreh SS, Zimmerman MB. Central retinal artery occlu- 388402.
sion: visual outcome. Am J Ophthalmol. 2005;140(3): 73. Nussenblatt RB, Whitcup SM, Palestine AG. Uveitis: funda-
37691. mentals and clinical practice. St Louis: Mosby; 1996.
References 415
74. Hayreh SS. Blood flow in the optic nerve head and factors 85. Wu L, Martinez-Castellanos MA, Quiroz-Mercado H,
that may influence it. Prog Retin Eye Res. 2001;20(5): Arevalo JF, Berrocal MH, Farah ME, et al. Twelve-month
595624. safety of intravitreal injections of bevacizumab (Avastin):
75. Jonas JB, Spandau UH, Harder B, Sauder G. Intravitreal tri- results of the Pan-American Collaborative Retina Study
amcinolone acetonide for treatment of acute nonarteritic Group (PACORES). Graefes Arch Clin Exp Ophthalmol.
anterior ischemic optic neuropathy. Graefes Arch Clin Exp 2008;246(1):817.
Ophthalmol. 2007;245(5):74950. 86. Hayreh SS. Posterior ischaemic optic neuropathy: clinical
76. Kaderli B, Avci R, Yucel A, Guler K, Gelisken O. Intravitreal features, pathogenesis and management. Eye (Lond).
triamcinolone improves recovery of visual acuity in nonar- 2004;18(11):1188206.
teritic anterior ischemic optic neuropathy. J Neuroophthalmol. 87. Hayreh SS. Erectile dysfunction drugs and non-arteritic
2007;27(3):1648. anterior ischemic optic neuropathy: Is there a cause and
77. Hayreh SS. Intravitreal triamcinolone for nonarteritic ante- effect relationship? J Neuroophthalmol. 2005;25(4):2958.
rior ischemic optic neuropathy. J Neuroophthalmol. 88. Hayreh SS. Non-arteritic anterior ischaemic optic neuropa-
2008;28(1):778. author reply 789. thy and phosphodiesterase-5 inhibitors. Br J Ophthalmol.
78. Sarraf D, Vyas N, Jain A, Bui A, Kertes PJ, Freund KB, 2008;92(12):157780.
et al. Triamcinolone-associated crystalline maculopathy. 89. Hayreh SS, Podhajsky P, Zimmerman MB. Beta-blocker
Arch Ophthalmol. 2010;128(6):68590. eyedrops and nocturnal arterial hypotension. Am J
79. Bennett JL, Thomas S, Olson JL, Mandava N. Treatment of Ophthalmol. 1999;128(3):3019.
nonarteritic anterior ischemic optic neuropathy with intravit- 90. Beri M, Klugman MR, Kohler JA, Hayreh SS. Anterior isch-
real bevacizumab. J Neuroophthalmol. 2007;27(3):23840. emic optic neuropathy. VII. Incidence of bilaterality
80. Pece A, Querques G, Quinto A, Isola V. Intravitreal ranibi- and various influencing factors. Ophthalmology. 1987;
zumab injection for nonarteritic ischemic optic neuropathy. 94(8)):10208.
J Ocul Pharmacol Ther. 2010;26(5):5237. 91. Hayreh SS, Podhajsky PA, Zimmerman B. Ipsilateral recur-
81. Hosseini H, Razeghinejad MR. Anterior ischemic optic neu- rence of nonarteritic anterior ischemic optic neuropathy. Am
ropathy after intravitreal injection of bevacizumab. J J Ophthalmol. 2001;132(5):73442.
Neuroophthalmol. 2009;29(2):1601. 92. Pathak A, Raoul V, Montastruc JL, Senard JM. Adverse drug
82. Ganssauge M, Wilhelm H, Bartz-Schmidt KU, Aisenbrey S. reactions related to drugs used in orthostatic hypotension: a
Non-arteritic anterior ischemic optic neuropathy (NA-AION) prospective and systematic pharmacovigilance study in
after intravitreal injection of bevacizumab (Avastin) for France. Eur J Clin Pharmacol. 2005;61(56):4714.
treatment of angoid streaks in pseudoxanthoma elasticum. 93. Maule S, Papotti G, Naso D, Magnino C, Testa E, Veglio F.
Graefes Arch Clin Exp Ophthalmol. 2009;247(12): Orthostatic hypotension: evaluation and treatment. Cardiovasc
170710. Hematol Disord Drug Targets. 2007;7(1):6370.
83. Hayreh SS. Anterior ischemic optic neuropathy. V. Optic
disc edema an early sign. Arch Ophthalmol. 1981;
99(6):103040.
84. Hayreh SS. Management of non-arteritic anterior ischemic For a complete bibliography and detailed review of previously
optic neuropathy. Graefes Arch Clin Exp Ophthalmol. published studies in the literature on the subject, please refer to

Animal Model of Non-arteritic Anterior
To have an animal model simulating a human disease already knew that the PCA supplied the ONH, I cut
is helpful, sometimes critical, in obtaining information different PCAs in turn and recorded the changes in the
about the pathogenesis, management and other aspects optic disc and ONH on ophthalmoscopy, fluorescein
of a disease, since it is not always either possible or fundus angiography, silicon rubber infusion and histo-
ethical to explore many aspects of a disease in the pathology. The optic disc developed edema and later
human. on optic atrophy. Fluorescein fundus angiography
My studies [1] showed in 1969 that the main source showed no filling of the area of the optic disc and chor-
of blood supply to the optic nerve head (ONH) is by oid supplied by the cut PCA. In the ONH, in the region
the posterior ciliary artery (PCA) circulation. Because of the occluded PCA, silicon rubber infusion studies
of my interest in anterior ischemic optic neuropathy showed absence of filling of vessels. Histopathological
(AION), I then conducted a study [2] to produce AION study revealed marked ONH damage in the region sup-
experimentally in rhesus monkeys (Fig.18.1). Since I plied by the cut PCA (Fig.18.2). Since, in these eyes,
cutting of the PCA produced massive infarction of the
involved ONH, the duration of their optic disc edema
was much shorter than is usually seen in non-arteritic
AION (NA-AION) in the human. This agrees with
what our study [3] on the evolution of optic disc edema
in NA-AION showed, i.e., the more extensive and
severe the visual loss (i.e., loss of axons), the quicker
the resolution of optic disc edema. My study [2] pro-
duced what is clinically seen in arteritic AION, where
giant cell arteritis causes complete occlusion of the
involved PCA.
Unfortunately, we do not have as yet an animal
model exactly simulating the NA-AION seen in the
human, in spite of claims made to that effect, as dis-
cussed below.
Bernstein and colleagues, with the objective of hav-
ing an animal model of NA-AION to study in detail
various aspects of this disease, and possible therapeu-
tic interventions, have created two animal models
one in rodents (rAION model) [49] and another
one in non-human primates (pNAION model) [10],
and have claimed that those resemble human
NA-AION. Their rAION model, in view of their
Fig. 18.1 Fundus photograph of a rhesus monkey, showing
chalky white optic disc edema 2days after occlusion of the PCA claim, is also being used by other investigators for
(Reproduced from Hayreh and Baines [2]) other related studies [11, 12]. I will discuss these two

418 18 Animal Model of Non-arteritic Anterior Ischemic Optic Neuropathy
a b
Fig.18.2 Photomicrographs of the ONH and retrolaminar optic region, producing an appearance resembling cavernous degen-
nerve in rhesus monkeys. (a) 36days after occlusion of all the eration. (Massons trichrome stain) (Reproduced from Hayreh
PCAs (Massons trichrome stain) showing atrophy and degen- [29]) (c) 42days after occlusion of the lateral PCA (Massons
erative changes in retrolaminar region. (Reproduced from trichrome stain) showing atrophy and degenerative changes
Hayreh and Baines [2]) (b) Higher magnification of a part of (a) involving the temporal half of the optic nerve, with nasal half
showing marked degeneration of neural tissue in retrolaminar normal (Reproduced from Hayreh and Baines [2])
models in detail to explain why the claim of Bernstein retinal, choroidal, and iris capillary beds filling with
and colleagues that these animal models resemble India ink. The eyes were dissected, and postfixed in
human NA-AION is not justified [13]. 4% PF-PBS for 24h. The retina, along with the chor-
Bernstein et al. [4] in 2003, produced a rodent oidal vasculature, was dissected immediately adjacent
photoembolic stroke model in rats and called it to the optic nerve, and flat mounted with an aqueous
rAION. To produce that model, they first adminis- mounting medium. Optic nerve tissue was fixed in
tered intravenously 90% Rose Bengal. After that, the glutaraldehyde-paraformaldehyde and embedded in
rat optic nerve was directly treated with an argon Epon.
laser light (514 nm) or frequency-doubled yttrium- On clinical examination, they stated that 1day after
aluminum garnet laser (YAG, light 535 nm). They rAION induction, the ON was pale and swollen and
stated that to confirm the ischemic nature of the laser- ON edema decreased 3 days after induction and
induced optic nerve lesion, 30min after rAION induc- resolved by 5days after induction. The 5-day nerve
tion, the rats were anesthetized and through the left appeared grossly normal. However, the ON at
ventricle were rapidly perfused with 5mL India ink. 37days after induction was pale and shrunken. They
The mouse was decapitated 10s after the beginning of concluded that The ON changes in rAION mirrored
perfusion. According to them, this resulted in the the clinical course of AION, But our study [3] dealing
18 Animal Model of Non-arteritic Anterior Ischemic Optic Neuropathy 419
with the evolution of optic disc edema in NA-AION in B. Bernstein etal.s [4] reference number 24, by Beck
the human showed, as mentioned above, that the more and colleagues [16], deals with the role of cup to
marked the optic nerve damage, the quicker the resolu- disc ratio and was not a histological study.
tion of optic disc edema and development of optic
atrophy. So in the rAION model, the optic nerve suf- Thus, their claim that rAION resembles human
fered massive ischemic damage, totally unlike that AION is not justified. While, no doubt, they produced
seen in NA-AION in humans. This is very well dem- anterior ischemic optic neuropathy, it does not resem-
onstrated by the presence of massive amount of infarc- ble the human NA-AION; the severity of ONH isch-
tion of the optic nerve in the longitudinal histological emic damage in rAION model is similar to our model
section of the nerves in the rAION model [7, 8]. On of AION [2] produced in rhesus monkeys by cutting
microscopic evaluation of the optic nerves, the capil- the posterior ciliary artery in both models there was
laries of the optic nerve of rAION animals showed a severe optic nerve head damage (like arteritic AION)
severe reduction in the filling of many of the ONH cap- and consequently optic disc edema resolving in a short
illaries. Bernstein et al. stated that: Thus, the ON period (as discussed above). The findings from rAION
defect induced by the RB-laser technique is a true isch- cannot be applied to the human NA-AION.
emic lesion. and they concluded that long-term his- Chen etal. [10] described a A Primate Model of
tological changes in rAION were similar to those seen Nonarteritic Anterior Ischemic Optic Neuropathy
in the human disease. They give two references (their (pNAION). They produced this model in rhesus
references 3 and 24) in support of this statement, but monkeys ((maccaca mulatta) by injecting Rose Bengal
unfortunately both are inappropriate to their conclu- dye intravenously and then intravascularly photoacti-
sion for the following reasons: vated in the optic disc with a neodymium-Yttrium alu-
minum garnet (Nd:YAG). They described this as a
A. Bernstein etal.s [4] reference number 3 deals with photoembolic mechanism to affect preferentially the
a histological evaluation by Levin and colleagues small vessels on the optic disc. They concluded that
[14, 15] of one optic nerve. I have discussed this The pNAION model of ON damage resembled human
particular study [14, 15] in detail in Chap. 14. NAION (NA-AION) clinically, angiographically, elec-
Briefly the development of NA-AION in that optic trophysiologically, and histologically.
nerve was pathogenetically not the same type as Relevant to this study: as is evident from the various
seen in classical human NA-AION, for the follow- chapters of this book, I have conducted comprehensive
ing reasons: basic, experimental and clinical research studies deal-
1. Detailed histological study of that same optic ing with the pathogenesis, clinical and angiographic
nerve by Tesser and Niendor [15] (discussed in aspects of NA-AION (see Chaps. 1416). Based on that
Chap. 14), showed that the optic nerve ischemic knowledge, I have analyzed this claim of a pNAION
damage was actually located in the central part animal model and its relevance to the human NA-AION.
of the intraorbital part and not in the ONH, which This is important because findings from such a model
is the classical location in human NA-AION. may automatically be applied to the human NA-AION,
2. In classical human NA-AION, the ischemic dam- resulting in misleading information.
age is located in the area supplied by the PCA
circulation (i.e., ONH), but in the optic nerve 1. Site of vascular insufficiency in human NAION:
examined by Levin and colleagues [14, 15] the As discussed in Chap. 3, the prelaminar, laminar
ischemic damage was in the central part of the and retrolaminar regions of the ONH are supplied
intraorbital part of the optic nerve supplied by primarily by the PCA circulation. By contrast, the
the intraneural branches of the central retinal surface nerve fiber layer of the ONH is supplied by
artery, and not the PCA circulation. Thus, the the retinal arterial circulation (Fig.18.3). Both arter-
optic nerve examined by Levin and colleagues itic and nonarteritic AION are due to vascular insuf-
[14, 15] did not have classical NA-AION, in ficiency in the PCA circulation only, without
spite of their claim, but was pathogenetically a involving the retinal circulation in the surface
rare variety of NA-AION. Moreover, it was nerve fiber layer at all. In pNAION the surface layer
totally different from the rAION model. of the optic disc is also involved. Clinically, this
CZ
C
Retinal vein
Retinal artery R S
PCA
OD
PR
LC
PCA
Fig.18.3 Schematic representation of blood supply of the optic nerve head. Abbreviations: C choroid, CZ circle of Zinn and Haller,
LC lamina cribrosa, OD optic disc, PCA posterior ciliary artery, PR prelaminar region, R retina, S sclera (Modified from Hayreh [30])
information is best supplied by fluorescein fundus past 34 years, I have reproduced multiple angio-
angiography. My fluorescein fundus angiographic grams showing the filling defect/delay in the prelam-
studies in about 1,200 eyes with NA-AION at its inar region of the optic nerve head and peripapillary
onset have almost invariably shown a filling defect/ choroid (Figs.18.4 and 18.5; and for example also
delay in the prelaminar region and in the peripapil- Figs. 3.39, 3.41, 3.45, 3.52, 3.54, 3.573.59) [17,
lary choroid and/or choroidal watershed zones 18], as also did Arnold and colleagues [19, 20]. In
(Figs.18.4 and 18.5; also see Figs.3.39, 3.41, 3.45, the literature, there are several studies dealing with
3.52, 3.54, 3.573.59). Typically, in NA-AION there the measurement of blood flow in NA-AION, show-
is no filling defect in the surface layer of the optic ing deranged blood flow in the PCAs (the main
disc supplied by the retinal circulation (Fig. 18.3) source of blood supply to only the deeper parts of
[17, 18]. Similarly, Arnold and colleagues [19, 20] the ONH [21, 22]). Above all, like NA-AION, arter-
showed filling defects in the prelaminar region in itic AION is also due to ischemia of the deeper part
NA-AION. An evaluation of five angiograms taken of the ONH; in arteritic AION, occlusion of the
by Arnold and colleagues [19, 20] during the early PCA has invariably been demonstrated on: (1) fluo-
phase of fluorescein filling in their two papers rescein fundus angiography (see Figs. 3.24, 3.25,
showed a choroidal peripapillary or watershed zone 3.28-3.34, 3.36,3.37, 3.43, 3.49, and 49,50 in chap-
defect in all the five eyes with NA-AION. However, ter 3) [23, 24], (2) many postmortem histopathologi-
Bernstein etal. [21] stated: Dr. Hayrehs statement cal studies (see Chap. 12), and (3) by our experimental
that NAION is due to vascular insufficiency in the occlusion of the PCA [2]. That provides ironclad
deeper part of the ONH supplied by the posterior evidence of the vascular insufficiency in the deeper
ciliary artery only is also disputable. Dr. Hayreh part of the ONH supplied by the posterior PCA only
provides no cited evidence for this statement. But in AION. This definitely contradicts the statement
this is not true at all; in a series of papers during the by Bernstein etal. [21].
18 Animal Model of Non-arteritic Anterior Ischemic Optic Neuropathy 421
Fig.18.4 Two fluorescein fundus angiograms at 6s interval in

the dye circulation (15s after the injection of fluorescein), shows
no filling of the temporal, superior and inferior peripapillary
choroid and superior choroidal watershed zone (arrow), and the
optic disc, with normal filling of both the medial and lateral pos-
there is complete filling of the peripapillary choroid (main
source of blood supply to the ONH) and the watershed zone.
These angiograms illustrate that in NA-AION there is only Fig. 18.5 Fundus photograph (above) and fluorescein angio-
delayed filling of the peripapillary choroid and the watershed gram (below) of right eye with NA-AION. Fundus photograph
zone and no occlusion of the PCA shows optic disc edema. Fluorescein angiogram at that time
showed non-filling of temporal part of the peripapillary choroid
and adjacent optic disc and of upper half of the choroidal water-
shed zone (arrows), with normal filling of both the medial and
2. The nature of vascular insufficiency and patho-
lateral posterior ciliary arteries (Reproduced from Hayreh [17])
genesis of human NA-AION: In arteritic AION,
it is well-established now that there is thrombotic
occlusion of the PCA caused by giant cell arteri- the ONH vessels, without any occlusion of those
tis [18, 23, 24] (as discussed in Chap. 12). vessels, as clearly shown by fluorescein fundus
Contrary to a widespread misconception (as dis- angiography in eyes with NA-AION in Figs.18.4
cussed in Chap. 14) NA-AION in the vast major- and 18.5, and also by multiple angiograms in
ity is NOT a thromboembolic but a hypotensive Chap. 3 (Figs. 3.39, 3.41, 3.45, 3.52, 3.54,
disorder, caused by transient hypoperfusion of 3.573.59).
In the light of these basic facts about human NA-AION, there was faster resolution of optic disc edema
let us examine the validity the claim that The pNAION in arteritic AION than in NA-AION. Therefore,
model of ON damage resembled human NAION. the argument by the authors [10] that the fast
resolution of optic edema in pNAION com-
1. pNAION was produced by a thromboembolic
pared to that in the human NAION was related
mechanism to affect preferentially the small vessels
to difference in life span is not valid.
on the optic disc [21]. This means there was throm-
(c) Peripapillary retinal hemorrhages: In that
bosis of small vessels in the surface nerve fiber layer,
study [10], there were marked amounts of peri-
prelaminar region, lamina cribrosa region and the
papillary retinal hemorrhages in pNAION,
immediate retrolaminar region (Fig.18.3) depend-
which is the result of thrombotic occlusion of
ing upon the depth of penetration of the Nd:YAG
the small retinal veins in the surface layer of
laser. As mentioned above, in human NA-AION
the optic disc (Fig. 18.3), produced by their
there is typically NO vascular occlusion of the sur-
technique. That is not seen in typical human
face nerve fiber layer, only transient hypoperfusion
NA-AION except in diabetics [26].
or non-perfusion in the peripapillary region (which
(d) Fluorescein fundus angiographic findings:
supplies the prelaminar and rest of the ONH
According to Chen etal. [10], leakage of fluo-
Fig.18.3), usually during sleep [25]. Once the small
rescein dye from optic disc vessels within 1day
ONH vessels are thrombosed and occluded, there is
after induction was evidence that they had pro-
no circulation at all in the thrombosed capillaries
duced NA-AION similar to that seen in the
and fine vessels in the ONH for many days at the
human. However, it is well-established that optic
least. Thus, in pNAION the circulatory situation is
disc staining due to leakage of fluorescein from
exactly like that seen in arteritic AION this is fur-
the optic disc vessels is a universal, non-specific
ther confirmed by the optic disc and other findings
finding of optic disc edema due to any cause,
in pNAION (see below). The site and type of occlu-
and has no diagnostic importance whatsoever.
sion in this animal model has no similarity to that
(e) Histopathological studies: It is well known
seen clinically in human NA-AION.
that optic nerve fiber damage in ONH from any
2. The authors state that in that study [10]: pNAION-
cause results in secondary development of
induced ON edema occurred within 1 day after
optic disc pallor, loss of retinal ganglion cells
induction and resolved over 2 to 3 weeks. The
and disruption of normal architecture in the
edema was visible as optic disc swelling and was
optic nerve on histopathology. Similar histo-
similar to the edema in human NAION, as were the
pathological changes were seen when I experi-
peripapillary retinal hemorrhages that appeared at
mentally produced AION [2], resembling
about the same time (citing the paper by Beck
arteritic AION (Fig.18.2). Knox etal. [27] in
etal. [16]). There are several serious problems with
their mixture of arteritic and non-arteritic
this statement, including the following:
AION cases, had similar findings. Thus, the
(a) The paper by Beck etal. [16] contains no such histopathological studies of Chen etal. [10] do
description. not lend any support to the contention that their
(b) Optic disc edema: In that study [10], 1 day animal model actually represents the human
after induction of pNAION, the disc was pale NA-AION.
and+swollen. This is a classical finding of
arteritic AION [23, 24] but NEVER present in In the light of all the evidence, the conclusion by Chen
typical human NA-AION. The presence of pale etal. [10] that The model is clinically, angiographi-
optic disc edema 1 day after induction of cally, electrophysiologically, and histopathologically
pNAION is a definite manifestation of severe similar to human NAION is not valid. It simply repre-
ischemia caused by thrombosis of small ves- sents, rather, a model of ONH damage caused by severe
sels on the optic disc (see above). Our study ischemic insult to the surface layer of the optic disc
has shown that the more extensive the ONH and adjacent part of the ONH (Fig.18.3), totally differ-
ischemic damage, the faster the resolution of ent in nature from the human NA-AION, but more like
optic disc edema [3]. That is why, in my studies, arteritic AION. Studies based on such a model have
References 423
the potential to introduce misleading information on the best interest of Science, and provides checks and
human NA-AION and even harm patients. balances. I agree with his comments that: there are no
The following is a summary of some of the main easy or complete answers to the in vivo study of
responses from Bernstein [28] to my comments given NAION. Complete elucidation of the advantages and
above [13]. He rightly stated that: The new models disadvantages of these models, their mechanisms and
allow us, for the first time, to measure precisely the their contribution to the study of NAION will require
effect of sudden, isolated axonal ischemia invivo on multiple investigators, in multiple labs, years of effort
retinal ganglion cells, total retina, and optic nerve. The to achieve. However, I have good reasons to doubt
models have opened up new avenues of research into their claim that their animal models will provide infor-
inflammatory and vascular responses to NAION. In mation about NA-AION and to fear that they may lead
spite of their earlier conclusion that The pNAION us in the wrong direction. His comment that NA-AION
model of ON damage resembled human NAION clini- is a currently untreatable disorder is not correct;
cally, angiographically, electrophysiologically, and once again our large, prospective study [31] has shown
histologically, he now stated: We have never stated that treating NA-AION eyes during the acute phase
nor claimed that our new models of NAION are abso- with high-dose systemic corticosteroids results in a
lutely identical with the human disease. However, he significantly higher probability of improvement in
contended that: By focusing only on whether or not visual acuity (p=0.001) and visual fields (p=0.005),
this model or that model are the true model of compared to an untreated group (see Chap. 17).
NAION, he (Hayreh) ignores exciting new findings Cioffi [32] has produced chronic localized ischemia
and relevant data obtainable from multiple approaches. of the anterior part of the optic nerve in primates by
I disagree with him here; to apply findings from an delivering endothelin-1 to the retrobulbar space. He
animal model that is not truly valid for the human concluded; Chronic optic nerve ischemia causes
NA-AION can be misleading and has the potential to demonstrable and localized damage of the optic nerve,
introduce misinformation which can do more harm without intraocular pressure elevation. There is prefer-
than good in the long run. Their model resembles arter- ential loss of large retinal ganglion cell axons in ani-
itic AION and NOT NA-AION the two differ in mals with significant axonal loss. Ischemia-induced
many ways in their pathophysiology, clinical features focal axonal loss is similar to human glaucoma and
and management. According to Bernstein [28], my may represent a differential regional vulnerability.
statement that only during the very early phase of However, that animal model is designed to study glau-
[fluorescein] dye filling in the fundus [provides infor- comatous optic neuropathy and is not relevant to the
mation about NAION] . . . after that the information is clinically seen human NA-AION.
lost. is not correct, and he goes on to state: Who
determined that absolute limit? My absolute limit is
based on fluorescein angiography results from over a Conclusion
thousand eyes with NA-AION and that has stood the
test of time.
Unfortunately, we do not have as yet an animal model
Bernstein [28] states: Dr. Hayreh makes a number
exactly duplicating the NA-AION seen in the human,
of additional absolute statements in his critique. We
in spite of claims made to that effect.
would hope that he is not making these statements to
suppress additional experimentation or to support the
validity only of a particular approach. Although there
are merits to his AAION model, they have not proven
sufficient to explain many of the mechanisms or References
responses in NAION or to suggest new treatments for
patients with this currently untreatable disorder. I 1. Hayreh SS. Blood supply of the optic nerve head and its role
have always been an enthusiastic supporter of new in optic atrophy, glaucoma, and oedema of the optic disc. Br
J Ophthalmol. 1969;53(11):72148.
avenues of research, so long as they do not claim to do 2. Hayreh SS, Baines JA. Occlusion of the posterior ciliary
or be more than they are. The field of medicine is full artery. III. Effects on the optic nerve head. Br J Ophthalmol.
of such wrecks. An honest and candid criticism is in 1972;56(10):75464.
3. Hayreh SS, Zimmerman MB. Optic disc edema in non-arter- 17. Hayreh SS. Inter-individual variation in blood supply of the
itic anterior ischemic optic neuropathy. Graefes Arch Clin optic nerve head. Its importance in various ischemic disorders
Exp Ophthalmol. 2007;245(8):110721. of the optic nerve head, and glaucoma, low-tension glaucoma
4. Bernstein SL, Guo Y, Kelman SE, Flower RW, Johnson MA. and allied disorders. Doc Ophthalmol. 1985;59:21746.
Functional and cellular responses in a novel rodent model of 18. Hayreh SS. Acute ischemic disorders of the optic nerve:
anterior ischemic optic neuropathy. Invest Ophthalmol Vis pathogenesis, clinical manifestations and management.
Sci. 2003;44:415362. Ophthalmol Clin North Am. 1996;9:40742.
5. Goldenberg-Cohen N, Guo Y, Margolis F, Cohen Y, Miller 19. Arnold AC, Hepler RS. Fluorescein angiography in acute
NR, Bernstein SL. Oligodendrocyte dysfunction after induc- nonarteritic anterior ischemic optic neuropathy. Am J
tion of experimental anterior optic nerve ischemia. Invest Ophthalmol. 1994;117:22230.
Ophthalmol Vis Sci. 2005;46(8):271625. 20. Arnold AC, Badr MA, Hepler RS. Fluorescein angiography
6. Bernstein SL, Guo Y, Slater BJ, Puche A, Kelman SE. in nonischemic optic disc edema. Arch Ophthalmol.
Neuron stress and loss following rodent anterior ischemic 1996;114:2938.
optic neuropathy in double-reporter transgenic mice. Invest 21. Bernstein SL, Kelman SE, Miller NR. Animal model for
Ophthalmol Vis Sci. 2007;48(5):230410. nonarteritic anterior ischemic optic neuropathy. J
7. Zhang C, Guo Y, Miller NR, Bernstein SL. Optic nerve Neuroophthalmol. 2008;28:801.
infarction and post-ischemic inflammation in the rodent 22. Hayreh SS. The blood supply of the optic nerve head and the
model of anterior ischemic optic neuropathy (rAION). Brain evaluation of it myth and reality. Prog Retin Eye Res.
Res. 2009;1264:6775. 2001;20:56393.
8. Zhang C, Guo Y, Slater BJ, Miller NR, Bernstein SL. Axonal 23. Hayreh SS. Anterior ischaemic optic neuropathy II. Fundus
degeneration, regeneration and ganglion cell death in a on ophthalmoscopy and fluorescein angiography. Br J
rodent model of anterior ischemic optic neuropathy (rAION). Ophthalmol. 1974;58:96480.
Exp Eye Res. 2010;91(2):28692. 24. Hayreh SS, Podhajsky PA, Zimmerman B. Ocular manifesta-
9. Slater BJ, Mehrabian Z, Guo Y, Hunter A, Bernstein SL. tions of giant cell arteritis. Am J Ophthalmol. 1998;125:
Rodent anterior ischemic optic neuropathy (rAION) induces 50920.
regional retinal ganglion cell apoptosis with a unique tempo- 25. Hayreh SS, Podhajsky PA, Zimmerman B. Non-arteritic
ral pattern. Invest Ophthalmol Vis Sci. 2008;49(8):36716. anterior ischemic optic neuropathy time of onset of visual
10. Chen CS, Johnson MA, Flower RA, Slater BJ, Miller NR, loss. Am J Ophthalmol. 1997;124:6417.
Bernstein SL. A primate model of nonarteritic anterior isch- 26. Hayreh SS, Zimmerman MB. Non-arteritic anterior isch-
emic optic neuropathy. Invest Ophthalmol Vis Sci. emic optic neuropathy: clinical characteristics in diabetics
2008;49(7):298592. versus non-diabetics. Ophthalmology. 2008;115(10):1818
11. Danylkova NO, Alcala SR, Pomeranz HD, McLoon LK. 25. Epub 2008 May 27.
Neuroprotective effects of brimonidine treatment in a rodent 27. Knox DL, Kerrison JB, Green WR. Histopathologic studies
model of ischemic optic neuropathy. Exp Eye Res. of ischemic optic neuropathy. Trans Am Ophthalmol Soc.
2007;84(2):293301. 2000;98:20320.
12. Goldenberg-Cohen N, Dadon-Bar-El S, Hasanreisoglu M, 28. Bernstein S. A primate model of nonarteritic anterior ischemic
Avraham-Lubin BC, Dratviman-Storobinsky O, Cohen Y, optic neuropathy. (Correspondence) Invest Ophthalmol Vis
et al. Possible neuroprotective effect of brimonidine in a Sci. http://www.iovs.org/cgi/eletters/49/7/2985#7919.
mouse model of ischaemic optic neuropathy. Clin Exp 29. Hayreh SS. Anterior ischemic optic neuropathy. New York:
Ophthalmol. 2009;37(7):71829. Springer; 1975.
13. Hayreh SS. A primate model of nonarteritic anterior 30. Hayreh SS. Anatomy and physiology of the optic nerve
ischemic optic neuropathy. (Correspondence) Invest head. Trans Am Acad Ophthalmol Otolaryngol. 1974;78:
Ophthalmol Vis Sci. http://www.iovs.org/cgi/eletters/49/7/ OP24054.
2985#7919. 31. Hayreh SS, Zimmerman MB. Non-arteritic anterior ischemic
14. Levin LA, Louhab A. Apoptosis of retinal ganglion cells in optic neuropathy: role of systemic corticosteroid therapy.
anterior ischemic optic neuropathy. Arch Ophthalmol. 1996; Graefes Arch Clin Exp Ophthalmol. 2008;246:102946.
114(4):48891. 32. Cioffi GA. Ischemic model of optic nerve injury. Trans Am
15. Tesser RA, Niendorf ER, Levin LA. The morphology of an Ophthalmol Soc. 2005;103:592613.
infarct in nonarteritic anterior ischemic optic neuropathy.
Ophthalmology. 2003;110(10):20315.
16. Beck RW, Servais GE, Hayreh SS. Anterior ischemic optic For a complete bibliography and detailed review of previously
neuropathy. IX. Cup-to-disc ratio and its role in pathogene- published studies in the literature on the subject, please refer to
sis. Ophthalmology. 1987;94(11):15038. bibliography in previous publications listed here.
Classification and Incidence of Posterior
I was the first to describe the clinical entity of posterior is usually hard to make with certainty, it is difficult to
ischemic optic neuropathy (PION) in 1981 [1], based ascertain its true incidence. In the literature, informa-
on my studies on the blood supply of the optic nerve tion on the incidence of PION consists almost entirely
(see Chap. 3) and clinical studies of ischemic optic of anecdotal reports of one or two patients, except for
neuropathies. It is very distinct from anterior ischemic three studies; (1) by Isayama etal. [6] in 14 patients, (2)
optic neuropathy pathogenetically and clinically (see by Sadda etal. [4] in 72 patients, and (3) my study [2]
Chaps. 14 and 16). Since then, an extensive literature in 43 patients (53 eyes). As far as I can judge, all the
has accumulated on PION. More recently, I [2] pub- cases in Isayama et al.s [2] series had non-arteritic
lished a detailed account of its clinical features, patho- PION. Among the 72 patients in Sadda etal.s [4] series,
genesis, and management, based on a detailed study of 53% had non-arteritic, 8% arteritic and 39% surgical
53 consecutive eyes with PION seen in my clinic. PION. In my series [2] of 43 patients (53 eyes), 65%
Etiologically, PION is not one clinical entity; there- had non-arteritic, 28% arteritic and 7% surgical PION.
fore, it is important to classify it into its various types The clinical findings in any PION study depend upon
for a full understanding of their pathogeneses and not only the number of patients with the different types
management, discussed in the subsequent chapters. of PION but also on the study design and the referral
pattern. For example, my study [2] and Sadda etal.s [4]
have some fundamental differences, and consequently
very different findings. In my study [2], all patients
Classification of PION
were referred by ophthalmologists from Iowa and adja-
cent parts of other states in the United States; the data
Etiologically, it can be classified into the following from all the patients was collected systematically by
three types: me, in a uniform manner and to a uniform standard.
1. Arteritic PION: is due to giant cell arteritis. Also there were only a few surgical PION cases in my
2. Non-arteritic PION: is due to various causes other series, because they were all local patients. The study
than giant cell arteritis. by Sadda et al. [4], by contrast, was a retrospective,
3. Surgical PION: is attributable to a surgical proce- multicenter chart review study of 72 patients with
dure. This has also been called postoperative [3] PION, with some of the data obtained from examina-
or perioperative [4]. I prefer the term surgical tion by the referring ophthalmologists (which the
PION [2, 5] because it is more inclusive. authors state were often incomplete), and follow-up
information for patients who did not return for follow-
up obtained from a review of medical records of other
physicians, or from the closest living relative if the
Incidence of PION patients was deceased. Since my study was based on
consecutive patients, seen by one observer and in one
When I first described PION as a distinct clinical entity, clinic part of a large University of Iowa Hospitals and
I stressed that it is a diagnosis of exclusion [1].Since the Clinics complex it probably reflects more closely the
diagnosis of PION, and especially non-arteritic PION, real frequency of the various types of PION.

426 19 Classification and Incidence of Posterior Ischemic Optic Neuropathy
References 5. Hayreh SS. Neuropatas pticas Isqumicas. In: Ginebreda

JA, Dalmau BS, editors. Neuropatas pticas diagnstico y
tratamiento. Barcelona: LXXVIII Ponencia Oficial de la
1. Hayreh SS. Posterior ischemic optic neuropathy. Sociedad Espaola de Oftalmologa; 2002. p. 20937.
Ophthalmologica. 1981;182(1):2941. 6. Isayama Y, Takahashi T, Inoue M, Jimura T. Posterior isch-
2. Hayreh SS. Posterior ischaemic optic neuropathy: clinical emic optic neuropathy. III. Clinical diagnosis. Ophthalmo
features, pathogenesis, and management. Eye. 2004;18(11): logica. 1983;187(3):1417.
1188206.
3. Roth S, Barach P. Postoperative visual loss: still no answers
yet. Anesthesiology. 2001;95(3):5757.
4. Sadda SR, Nee M, Miller NR, Biousse V, Newman NJ, For a complete bibliography and detailed review of previously
Kouzis A. Clinical spectrum of posterior ischemic optic neu- published studies in the literature on the subject, please refer to
ropathy. Am J Ophthalmol. 2001;132(5):74350. bibliography in previous publications listed here.
Pathogenesis of Posterior Ischemic
Optic Neuropathy 20
Posterior ischemic optic neuropathy (PION) is due to the pathogenesis of PION. For descriptive purposes, the
acute ischemia of the posterior part of the optic nerve. posterior part of the optic nerve can be divided into
For a logical understanding of the pathogenesis of PION, intraorbital, intracanalicular and intracranial parts.
it is essential to have a thorough knowledge of (a) the
blood supply of the posterior part of the optic nerve, (b)
histopathological findings in PION and (c) risk factors. Intraorbital Part
This is further subdivided by point of entry of the cen-

Blood Supply of the Posterior Part tral retinal artery in the optic nerve into (a) anterior and
of the Optic Nerve (b) posterior segments (Fig.20.1).
(a) Anterior Segment
This is discussed at length in Chap. 3. However, a brief This is between the optic nerve head and the site of
description of the blood supply of the posterior part of the entry of the central retinal artery into the nerve.
optic nerve is essential here to make it easy to understand This part has two vascular systems for its supply.
C
R S Col. Br.
PCA
D
A
Pia
ON
PR
LC
SAS
CRV
PCA CRA
Fig.20.1 Schematic representation of blood supply of the optic optic nerve; PCA = posterior ciliary artery; PR = prelaminar region;
nerve. A = arachnoid; C = choroid; CRA = central retinal artery; R = retina; S = sclera; SAS = subarachnoid space (modified from
Col. Br. = Collateral branches supplying the optic nerve pial plexus; Hayreh [1])
CRV = central retinal vein; D = dura; LC = lamina cribrosa; ON =

428 20 Pathogenesis of Posterior Ischemic Optic Neuropathy
(i) A peripheral centripetal vascular system is (b) Posterior Segment

present in all cases and consists of the pial vas- This is primarily supplied by the peripheral cen-
cular plexus, supplied by multiple pial branches tripetal vascular system formed by the pial vascu-
originating from the peripapillary choroid, lar plexus, supplied by multiple small collateral
circle of Haller and Zinn, central retinal artery, arteries usually arising directly from the ophthal-
ophthalmic artery and other orbital arteries mic artery and less often from other orbital arteries
[27] (Figs.20.120.3). [5] (Figs.20.1 and 20.2). In about 10% of the optic
(ii) An axial centrifugal vascular system is pres- nerves there may be an axial centrifugal vascular
ent in 75% of the nerves, supplied by 18 system extending backward for a variable distance,
intraneural branches of the central retinal artery formed by intraneural branches of the central reti-
[2, 3, 6] (Figs.20.1 and 20.3). nal artery [2, 3, 6, 8] (Fig.20.3).
Intracanalicular Part
Rec. Br. CZ This has only the peripheral centripetal system, sup-
era
Scl CZ
plied almost entirely by fine collateral branches from
the ophthalmic artery when it lies under the optic nerve
[5, 7, 8] (Fig.20.2).
MPCA LPCA
Awai [9] investigated the angioarchitecture of the
monkey intracanalicular optic nerve. The vascular
Col. Br. architecture consisted of branches from the ophthalmic
Col. Br.
Med. Mus. artery. Some branches derived from the chiasmal part
CAR of the ophthalmic artery, some from its intracanalicular
part, and others from its intraorbital part. On the sur-
OA face of the intracanalicular optic nerve and in its inner
portion, the vascular distribution was sparse. A clear
pial plexus was not observed. However, it was ascer-
Col. Br.
tained that there existed a few anastomoses between the
branches from the ophthalmic artery on the surface of
Optic Canal the optic nerve, and between the intraneural branches
Col. Br. within the optic nerve: no end artery appeared to exist.
Col. Br.
Optic Nerve
Intracranial Part
ICA This once again has only a pial vascular plexus, sup-
Ant. Sup.Hyp.
Art. plied by a variable and inconsistent number of fine
Optic Chiasma branches coming from various surrounding arteries,
Optic Tract including the anterior superior hypophyseal, anterior
cerebral, anterior communicating and ophthalmic
Fig. 20.2 Diagrammatic representation of origin, course and arteries [7, 8, 10] (Fig.20.2).
branches of ophthalmic artery, as seen from above. Ant. Sup.
Hyp. Art. = anterior superior hypophyseal artery; CAR = central
retinal artery; Col. Br. = Collateral branches supplying the optic
nerve pial plexus; CZ = circle of Zinn and Haller; ICA = internal
Histopathological Findings in PION
carotid artery; LPCA = lateral posterior ciliary artery; Med.
Mus. = medial muscular artery; MPCA = medial posterior cili-
ary artery; OA = ophthalmic artery; Rec. Br. CZ = recurrent pial
There are several histopathological reports on PION
branches from peripapillary choroid/CZ (Reproduced from [1115]. Each is based on evaluation of only one
Hayreh [6]) patient, except that of Isayama and Takahashi [11]
Histopathological Findings in PION 429
Fig. 20.3 Diagram (based on camera lucida drawings) showing the central retinal artery in humans. CRA = central retinal artery;
one of the intraneural branches of the central retinal artery running CZ = circle of Zinn and Haller; PCA = posterior ciliary artery (From
backward in the axial part of the optic nerve posterior to the central Hayreh [2])
retinal artery. From one of the specimens in my anatomical study on
which is based on 12 cases. They [11] found that the of tunnel-like central vision (see Chap. 21). These varia-
ischemic lesions were located in the transverse, periph- tions in the optic nerve involvement in PION depend
eral, altitudinal or axial areas of the optic nerve. upon its blood supply pattern, particularly with respect
Schobel etal. [15] found complete loss of axons in the to the role of the central retinal artery (see above). If the
right nerve in their patient with bilateral PION, whereas entire blood supply is by centripetal branches from the
the central portion of the nerve fiber of the left nerve pial plexus, then the central part is the most distal part of
was intact, with complete loss of axons in the periph- supply by the terminal capillaries of these branches
eral part. In contrast to that, Weinstein et al. [12] in (Fig.20.1). The terminal capillary bed is most vulnera-
their case with PION found that transverse sections ble to hypoperfusion and ischemia in the German lit-
through 17mm behind the eyeball showed only patchy erature this phenomenon has been aptly named
necrosis, most prominent centrally, with relative spar- wipfeldre, which literally means withering away of the
ing of the peripheral part. In the remaining reports outermost twigs of a tree because nutrition is not reach-
infarction of the entire optic nerve was found. ing them. This makes the central part of the posterior
optic nerve a watershed zone, rendering it more vulner-
able to ischemia than the more peripheral part. On the
other hand, when the posterior part of the optic nerve
Why Is the Peripheral Part of the Optic has a centrifugal vascular system supplied by the central
retinal artery (Fig.20.3) and the circulation in the cen-
Nerve Sometimes Spared in PION, and
tral retinal artery is not yet compromised because of
Sometimes the Central Part of the Nerve? blood flow autoregulation, that may prevent ischemic
damage to the central part of the nerve, although the
Histopathological studies in PION have shown that in peripheral part supplied by the fine pial capillary plexus
some nerves infarction involves the entire thickness of with low pressure is damaged.
the nerve, in others only the central part, in still others It is well-established that AION is due to ischemia
only the peripheral part [1115]. Similarly, clinically an of the optic nerve head region supplied by the posterior
eye with PION may be totally blind, or may show mostly ciliary artery. However, as is evident from the descrip-
central visual field loss, or there may be a marked gen- tion above, the posterior part of the optic nerve is sup-
eralized constriction of the visual field with preservation plied by multiple branches arising from multiple
sources [37] (Fig. 20.2). Therefore, PION, unlike 22], and marked arterial hypotension [19]. There are
AION, does not represent an ischemic disorder of any anecdotal case reports of non-arteritic PION associated
one known artery and does not have any specific loca- with migraine [2325], systemic lupus erythematosus
tion in the posterior part of the optic nerve. Any local- [17, 26], polyarteritis nodosa [27, 28], Wegeners granu-
ized segment or part of the posterior optic nerve may lomatosis [29], carotid artery stenosis or occlusion [11,
be involved. Thus, pathogenetically and clinically 22], carotid artery dissection [3032], pulseless disease
AION and PION are very different disease entities, and [33], rupture of intracranial aneurysm [12, 34], extra-
it is unfortunate that ophthalmologists and neurologists dural hematoma [35], subdural hematoma [36], poste-
lump them together as ischemic optic neuropathy. rior-draining dural cavernous sinus fistula [37], head
injury [11], endoscopic sinus surgery [38], emboli [11],
aplastic anemia [11], sickle cell SS disease [39], hemo-
dialysis [40], chronic renal-failure [41], following use of
Pathogeneses of Various Types of PION sildenafil [42], following correction of malignant hyper-
tension [43] and following aspergillus fumigatus infec-
tion [12]. Sadda etal. [24] found a variety of associated
Arteritic PION
systemic diseases in their series of non-arteritic PION
patients.
This is due to giant cell arteritis (GCA); a number of Systemic diseases seen in my series [44] of 28
cases with this type of PION have been reported in the patients with non-arteritic PION were as follows: arte-
literature [1518]. In the older literature they were rial hypertension (13 patients), diabetes mellitus (6),
called retrobulbar optic neuritis due to GCA [16]. ischemic heart disease (7), other cardiac diseases
Although the most common artery involved by GCA is (including valvular disease, atrial fibrillation, patent
the posterior ciliary artery, resulting in the development foramen ovale, congestive heart failure and echocar-
of arteritic AION, GCA can sometimes also involve the diographic abnormalities, in 6), arterial hypotension (3
ophthalmic artery or its other orbital branches. Arteritic orthostatic in one), thyroid disease (3), rheumato-
PION must be due to involvement of one or more of the logic diseases (3: 2 of them had systemic lupus erythe-
collateral branches supplying the posterior part of the matosus), cerebrovascular accidents (5), carotid artery
optic nerve (Fig. 20.2). Arteritic PION is much less disease (stenosis and/or plaque in 9), peripheral vascu-
common than arteritic AION. For example, in our study lar disease (1), vasospastic diseases (migraine in 4 and
of 123 eyes with visual loss due to GCA, arteritic AION Raynauds disease in 1), gastrointestinal ulcer (4), and
was seen in 94 eyes and arteritic PION in only 7 [18]. hypercholesterolemia (11). Nine patients were smok-
ers (3415 pack years). Magnetic resonance imaging
and/or computer tomography was done in 16 patients
to rule out any other cause of visual loss. Eight had
Non-arteritic PION
temporal artery biopsy to rule out GCA. When the
prevalence of major systemic diseases in non-arteritic
Non-arteritic PION, like non-arteritic AION (see Chap. PION patients of my study [44] was compared with
14), is a multifactorial disease. Therefore, one has to those expected in the age-matched control population
look into the various risk factors which can possibly in the US Caucasian population, it showed a signifi-
play a role in its development. cantly higher prevalence of arterial hypertension
(p=0.022), ischemic heart disease (p=0.026), cere-
brovascular disease (p=0.006), carotid artery and
Systemic Risk Factors peripheral vascular disease (p<0.0001), diabetes mel-
litus (p=0.014), migraine (p=0.039) and gastrointesti-
An association between non-arteritic PION and a variety nal ulcers (p=0.011) in the non-arteritic PION patients.
of systemic diseases has been reported in the literature, This association does not necessarily mean there is a
mostly in anecdotal case reports. These diseases include cause-and-effect relationship to non-arteritic PION,
diabetes mellitus [11, 1921], arterial hypertension [11, although those diseases may constitute risk factors for
19, 22], arteriosclerosis [11, 19], atherosclerosis [19, the development of non-arteritic PION. However, there
Pathogeneses of Various Types of PION 431
are some cases where one can find a causal relation- PION [44, 48] because it is more inclusive. Surgical
ship. For example, four of my patients developed non- PION tends to cause bilateral, massive visual loss or
arteritic PION following an attack of migraine; some even complete blindness which is usually permanent;
patients with severe nocturnal arterial hypotension therefore, as well as being tragic for the patient, it has
woke up with visual loss due to non-arteritic PION; great medicolegal importance. A large number (mostly
severe arterial hypotension in association with a surgi- anecdotal) of surgical PION case reports have appeared
cal procedure (see below) or hemodialysis can result in in the literature, almost invariably associated with pro-
non-arteritic PION; and one of my patients developed longed systemic surgical procedures for a variety of
non-arteritic PION only on the side with 100% occlu- conditions, including spinal and other orthopedic sur-
sion of the internal carotid artery. gical procedures [44, 4954], radical neck dissection
In my study [44] erythrocyte sedimentation rate [1315, 5560], venous graft in extremities [61, 62],
(ESR) and C-reactive protein (CRP) levels were evalu- coronary artery bypass [63], hip surgery [63], nasal
ated in all patients. Patients with non-arteritic PION, had surgery [64, 65], thoracotomy for hemothorax [58],
a median ESR of 25 (Range 189) mm/h, and for CRP penetrating thoracoabdominal injury [66], breast aug-
all had <0.5mg/dl, except for two (1.1mg in one and mentation and abdominal liposuction [67], and pros-
10.6mg in another due to recent myocardial infarction tatectomy [68]; also it has been reported following
and congestive heart failure in both the temporal artery cataract surgery [44, 69], eyelid blepharoplasty [70],
biopsy was negative). In patients with arteritic PION, by strabismus surgery [71], and orbital surgery [44].
comparison, the median ESR was 90.5 (range 13130) Sadda etal. [24] reported 28 patients following a vari-
mm/h, and median CRP 4.75 (Range 1.512.2) mg/dl. ety of procedures.
Thus, non-arteritic PION, like non-arteritic AION Several reviews have dealt with surgical PION.
[45, 46] (see Chap. 14) is associated with multiple sys- Dunker et al. [59], in a study of 46 surgical PION
temic diseases. This suggests that the disease is multi- cases reported in the literature and 7 by them, con-
factorial in nature, with a variety of systemic diseases cluded that patients undergoing spinal surgery with
(see above) and other vascular risk factors predispos- prolonged intraoperative hypotension and postopera-
ing a person to PION. tive anemia and facial swelling are at risk of develop-
ing PION from hypovolemic hypotension. Gill and
Heavner [72], based on seven studies representing 102
Local Risk Factors in the Optic Nerve cases with postoperative visual loss associated with
spinal surgery, reviewed age, sex, comorbidities, diag-
There may be local risk factors in the optic nerve pre- nosis, operative time, blood loss, systolic blood pres-
disposing an optic nerve to develop non-arteritic PION. sure, lowest hematocrit, and visual deficits and
For example, the vascular pattern of the posterior part improvement. Ischemic optic neuropathy, especially
of the optic nerve, and changes in the arteries supply- surgical PION, was the most common diagnosis found
ing this part of the optic nerve, such as arteriosclerosis, in the studies. The average age of the patients ranged
atherosclerosis, spasm, compression and defective from 46.5years to 53.3years with the majority having
autoregulation of the optic nerve blood flow, may play at least one comorbidity. Operative time ranged on
a role. Some of these are considered later in the discus- average from 385 min to 410 min with a median in
sion dealing with pathogenesis of various types of one case series of 480min, average blood loss ranged
visual field defects. from 3.5 to 4.3l and no visual improvement was seen
Finally, one or the other of the risk factors may act in the majority of the cases. Patients with a large
as the precipitating factor (last straw) to produce amount of blood loss producing hypotension and ane-
non-arteritic PION. mia along with prolonged operative times may be at a
greater risk in developing visual disturbances. An
acute anemic state may have an additive or synergistic
Surgical PION effect with other factors (medical comorbidities) lead-
ing to visual disturbances. However, their study failed
This entity has also been called postoperative [47] or to provide definitive causative factors for postopera-
perioperative [24]. I have used the term surgical tive visual loss.
Pathogenesis of Surgical PION superior ophthalmic vein has been reported in the
immediate postoperative period in surgical PION after
For a logical understanding of this, one has to consider prolonged, prone position lumbar laminectomy, sug-
the main risk factors responsible for that. These include gesting an increase in orbital venous pressure during
the following: surgery [74].
(b) Arterial Hypotension
1. Severe and Prolonged Arterial Hypotension: This In these surgical PION patients, a fall of arterial blood
is due to prolonged general anesthesia, surgical pressure is well-documented. Some anesthesiologists
trauma and massive blood loss. Some orthopedic give vasopressor agents to increase blood pressure in
surgeons deliberately produce arterial hypotension patients with arterial hypotension during surgery; that
during surgery to reduce bleeding. can be dangerous, because these agents increase blood
2. Hemodilution and Anemia: This is from adminis- pressure by constricting the terminal arterioles.
tration of a large amount of intravenous fluids to Constriction of the terminal arterioles has a paradoxi-
compensate for the blood loss. cal effect proximal to the arterioles there is increased
3. Raised Intraorbital Venous Pressure: This is due to blood pressure (which gives a false sense of security)
orbital and periorbital edema, chemosis, and rarely but distal to the terminal arterioles (due to constriction
even direct orbital compression by prone position. of the arterioles) there is marked reduction of blood
During spinal surgery the Trendelenburg position is flow in the capillary bed, resulting in ischemic dam-
often used, which can further contribute to increased age. Moreover, these agents can also directly cause
orbital venous pressure and edema. Cheng etal. [73] vasoconstriction in the optic nerve, making it more
found that prone position during anesthesia increases susceptible to ischemic damage.
intraocular pressure, which is most likely a reflec- Increased orbital venous pressure by the above
tion of the increased orbital venous pressure. An mechanism(s), along with simultaneous arterial
increase in orbital venous pressure may also be the hypotension, is a dangerous combination, resulting in
result of radical neck dissection. One of my patients a fall of perfusion pressure (arterial pressure minus
with surgical PION developed it following the venous pressure) and consequently reduced blood flow
development of marked orbital edema after surgical to the optic nerve and ischemia. The presence of other
repair of an orbital floor fracture. risk factors discussed above may also contribute.
4. Systemic Cardiovascular Disease and Several authors have equated visual loss due to
Autoregulatory Dysfunction: These must make surgical PION to that seen in patients with post-hem-
these patients more susceptible to surgical PION. orrhagic amaurosis, i.e. visual loss after recurrent
5. Compression of the Capillary Pial Plexus of the systemic hemorrhages [75] (this is fully discussed in
Optic Nerve: In some cases raised orbital pressure Chap. 15). The two are actually very different in
may directly cause this. nature because, in post-hemorrhagic amaurosis, visual
loss develops: (a) hours, days or even weeks after sys-
Based on the available evidence, I postulated the fol-
temic bleeding, (b) very rarely after a single hemor-
lowing pathogenesis of surgical PION [44]: It develops
rhage, (c) when hemoglobin and blood pressure may
due to decreased blood flow in the posterior part of the
be within normal limits, and (d) usually during sleep;
optic nerve. Normally, blood flow depends upon the
or it may worsen during sleep [75]. To explain post-
perfusion pressure, which depends upon the difference
hemorrhagic amaurosis I [75] postulated that release
between the arterial and venous pressures. A fall of per-
of angiotensin and other endogenous vasoconstrictor
fusion pressure is caused by the following mechanism:
agents (secondary to recurrent massive systemic hem-
(a) Rise of Orbital Venous Pressure orrhage, with or without arterial hypotension) is most
This is caused by the following sequence of events: probably a major factor, with arterial hypotension
marked orbital edema increased intraorbital pres- being an additional important risk factor, and possi-
sure increased orbital venous pressure. The bly increased platelet aggregation. Although post-
Trendelenburg position may also contribute to that. In hemorrhagic amaurosis and surgical PION are not
radical neck dissection, ligation of the jugular veins identical clinically, they may share these factors in
can result in raised orbital venous pressure. Dilated their pathogeneses.
Why Is PION Much Less Common Than AION? 433
Pathogenesis of Various Types of Visual the preservation of a tunnel-like central field in PION
Field Defects Seen in PION patients, in spite of loss of the peripheral visual field
(see Chap. 21).
Since each collateral branch supplies a localized
The various types of visual field defects seen in PION area of variable size in the posterior optic nerve
are described in the chapter dealing with the clinical (Figs.20.1 and 20.2), occlusion of different collaterals
findings in PION (Chap. 21). To understand their can result in a variety of visual field defects, depending
pathogenesis, it is essential to have a full understand- upon their area of supply and the origin of the optic
ing of (a) the arrangement of the optic nerve fibers in nerve fibers from the retina in that region. Isayama and
the posterior part of the optic nerve (discussed in Takahashi [11] in their histopathological study in
Chap. 2) and (b) its vascular pattern (see above and PION found that the ischemic lesions could be located
Chap. 3). in the transverse, peripheral, altitudinal or axial areas
of the nerve. That would explain the different types of
visual field defects seen in PION.
Arrangement of the Optic Nerve Fibers
in the Optic Nerve
Why Is PION Much Less Common
This is discussed fully in Chap. 2. They rearrange Than AION?
themselves as they travel posteriorly in the optic nerve;
for example, the macular fibers lie in the temporal part Since 1973, I have seen about 1,350 patients with non-
of the optic nerve head but lie in the central part of the arteritic AION and more than 150 with arteritic AION
optic nerve posteriorly [76]. Hoyt and Luis [77] found in my Ocular Vascular Clinic. We know that both types
in the Java monkey, however, that the macular fibers of AION are due to ischemia of the optic nerve head.
mix freely with those from the peripheral retina and in During the same period, on the other hand, I have seen
no portion of the optic nerve except the distal optic only about 50 patients with definite PION. This may
nerve are they confined exclusively to a single area or be due to the very different nature of the blood supply
bundle. The implication of this is that segmental and blood flow of the optic nerve head compared to
ischemia of the optic nerve head is likely to produce a that of the posterior part of the optic nerve (Fig.20.1),
visual field defect very different from that produced by as is evident from the following:
segmental ischemia in the posterior part of the optic
nerve.
Nature of the Blood Supply

Vascular Pattern of the Posterior Part
The optic nerve head is supplied by the posterior cili-
of the Optic Nerve ary artery circulation which is an end-arterial system
without any anastomoses with surrounding arteries
As discussed above, the axial region in the posterior [78, 79]. The posterior part of the optic nerve is sup-
part of the optic nerve is a usually a watershed zone, plied by the continuous pial vascular plexus, which is
which makes it more susceptible to ischemic damage not an end-arterial system because the various col-
than the peripheral part. That may explain the common lateral branches (Figs. 20.1 and 20.2) supplying it
occurrence of a central visual field defect in PION (see anastomose freely with one another in the pial plexus
Chap. 21). In about 10% of the nerves, by contrast, the [2, 3, 7]. The anastomoses in the pial plexus may have
axial region of the posterior part of the optic nerve is a protective influence if one or another collateral
supplied by an intraneural branch of the central retinal branch alone is involved, which is not the case in the
artery [2, 3] (Fig.20.3), and that helps to protect the end-arterial posterior ciliary artery distribution in the
axial region from ischemic damage; that may explain optic nerve head.
Factors Influencing the Perfusion Pressure 9. Awai T. Angioarchitecture of intracanalicular part of mon-
key optic nerve. Jpn J Ophthalmol. 1986;30(1):6373.
and the Blood Flow 10. Blunt MJ, Steele EJ. The blood supply of the optic nerve and
chiasm in man. J Anat. 1956;90(4):48693.
11. Isayama Y, Takahashi T. Posterior ischemic optic neuropa-
The factors which influence the perfusion pressure thy. II. Histopathology of the idiopathic form.
(consequently the blood flow) in the optic nerve head Ophthalmologica. 1983;187(1):818.
are different from those in the posterior part of the 12. Weinstein JM, Morris GL, ZuRhein GM, Gentry LR.
optic nerve. In the optic nerve head, perfusion pressure Posterior ischemic optic neuropathy due to Aspergillus
fumigatus. J Clin Neuroophthalmol. 1989;9(1):713.
is equal to the difference between the mean arterial 13. Marks SC, Jaques DA, Hirata RM, Saunders Jr JR. Blindness
pressure and the intraocular pressure. On the other following bilateral radical neck dissection. Head Neck.
hand, in the posterior part of the optic nerve, the perfu- 1990;12(4):3425.
sion pressure is equal to the difference between the 14. Nawa Y, Jaques JD, Miller NR, Palermo RA, Green WR.
Bilateral posterior optic neuropathy after bilateral radical
mean arterial pressure and the venous pressure, with neck dissection and hypotension. Graefes Arch Clin Exp
intraocular pressure playing no role. Intraocular pres- Ophthalmol. 1992;230(4):3018.
sure is always much higher than the orbital venous 15. Schobel GA, Schmidbauer M, Millesi W, Undt G. Posterior
pressure. Therefore, with the same level of mean arte- ischemic optic neuropathy following bilateral radical neck
dissection. Int J Oral Maxillofac Surg. 1995;24(4):2837.
rial blood pressure, the optic nerve head is more vul- 16. Hollenhorst RW. Effect of posture on retinal ischemia from
nerable to develop lower perfusion pressure than the temporal arteritis. Arch Ophthalmol. 1967;78(5):56977.
posterior part of the optic nerve. 17. Hayreh SS. Posterior ischemic optic neuropathy.
Thus a combination of these and other factors may Ophthalmologica. 1981;182(1):2941.
18. Hayreh SS, Podhajsky PA, Zimmerman B. Ocular manifes-
make the optic nerve head more vulnerable to ischemic tations of giant cell arteritis. Am J Ophthalmol. 1998;
disorders than the rest of the optic nerve. Moreover 125(4):50920.
and most importantly, since PION is a diagnosis of 19. Isayama Y, Takahashi T, Inoue M, Jimura T. Posterior isch-
exclusion [17], it is possible that some PION cases emic optic neuropathy. III. Clinical diagnosis. Ophthalmo
logica. 1983;187(3):1417.
may be misdiagnosed or undiagnosed, giving a false 20. Pedrotti MM, Basso AL. Diabetic retrobulbar optic neuropa-
impression that it is less common than it actually is. thy. Metab Pediatr Syst Ophthalmol. 1986;9(24):947.
21. Bertram B, Reim H, Reim M. Beidseitige posteriore ischa-
mische Optikusneuropathie bei einer Jugendlichen mit
Diabetes mellitus mit dekompensierter Blutzuckereinstellung.
(Bilateral posterior ischemic optic neuritis in an adolescent
References with diabetes mellitus with decompensated blood glucose
homeostasis). Klin Monatsbl Augenheilkd. 1995;206(1):
3945.
1. Hayreh SS. Anatomy and physiology of the optic nerve 22. Inoue M, Tsukahara Y. Vascular optic neuropathy in diabe-
head. Trans Am Acad Ophthalmol Otolaryngol. tes mellitus. Jpn J Ophthalmol. 1997;41(5):32831.
1974;78:OP24054. 23. Lee AG, Brazis PW, Miller NR. Posterior ischemic optic
2. Hayreh SS. A study of the central artery of the retina in neuropathy associated with migraine. Headache.
human beings. [Thesis for the Master of Surgery]. 1996;36(8):50610.
Chandigarh: Panjab University; 1958. 24. Sadda SR, Nee M, Miller NR, Biousse V, Newman NJ,
3. Singh (Hayreh) S, Dass R. The central artery of the retina. II. Kouzis A. Clinical spectrum of posterior ischemic optic neu-
A study of its distribution and anastomoses. Br J Ophthalmol. ropathy. Am J Ophthalmol. 2001;132(5):74350.
1960;44:28099. 25. Foroozan R, Marx DP, Evans RW. Posterior ischemic optic
4. Hayreh SS. The blood supply of the optic nerve and its clini- neuropathy associated with migraine. Headache.
cal significance. Acta XIX Conc Ophthalmol New Delhi. 2008;48(7):11359.
1962;2:11949. 26. Jabs DA, Miller NR, Newman SA, Johnson MA, Stevens
5. Hayreh SS. The ophthalmic artery: III. Branches. Br J MB. Optic neuropathy in systemic lupus erythematosus.
Ophthalmol. 1962;46(4):21247. Arch Ophthalmol. 1986;104(4):5648.
6. Hayreh SS. The central artery of the retina. Its role in the 27. Saraux H, Le Hoang P, Laroche L. Neuropathie optique isch-
blood supply of the optic nerve. Br J Ophthalmol. eemique aigue anterieure et posterieure au cours dune peri-
1963;47:65163. arterite noueuse. Traitement par plasmapherese. (Anterior
7. Hayreh SS. Blood supply and vascular disorders of the optic and posterior acute ischemic optic neuropathy related to
nerve. An Inst Barraquer. 1963;4:7109. poly-arteritis nodosa). J Fr Ophtalmol. 1982;5(1):5561.
8. Isayama Y, Hiramatsu K, Asakura S, Takahashi T. Posterior 28. Hutchinson CH. Polyarteritis nodosa presenting as posterior
ischemic optic neuropathy. I. Blood supply of the optic ischaemic optic neuropathy. J R Soc Med. 1984;77(12):
nerve. Ophthalmologica. 1983;186(4):197203. 10436.
References 435
29. Blaise P, Robe-Collignon N, Andris C, Rakic JM. Wegeners 48. Hayreh SS. Neuropatas pticas Isqumicas. In: Ginebreda
granulomatosis and posterior ischemic optic neuropathy: JA, Dalmau BS, editors. Neuropatas pticas diagnstico y
atypical associated conditions. Eur J Intern Med. tratamiento. Barcelona: LXXVIII Ponencia Oficial de la
2007;18(4):3267. Sociedad Espaola de Oftalmologa; 2002. p. 20937.
30. Rivkin MJ, Hedges 3rd TR, Logigian EL. Carotid dissection 49. Katz DM, Trobe JD, Cornblath WT, Kline LB. Ischemic
presenting as posterior ischemic optic neuropathy. Neurology. optic neuropathy after lumbar spine surgery. Arch
1990;40(9):1469. Ophthalmol. 1994;112(7):92531.
31. Tsai RK, Sun CY. Spontaneous dissection of internal carotid 50. Myers MA, Hamilton SR, Bogosian AJ, Smith CH, Wagner
artery presenting as isolated posterior ischaemic optic neu- TA. Visual loss as a complication of spine surgery. A review
ropathy. Br J Ophthalmol. 1997;81(6):513. of 37 cases. Spine. 1997;22(12):13259.
32. Kawabe K, Nagaoka T, Iguchi H, Ikeda K, Iwasaki Y. Optic 51. Stevens WR, Glazer PA, Kelley SD, Lietman TM, Bradford
nerve MRI enhancement in posterior ischaemic optic neu- DS. Ophthalmic complications after spinal surgery. Spine.
ropathy due to internal carotid artery dissection. Clin Neurol 1997;22(12):131924.
Neurosurg. 2010;112(4):3502. 52. Roth S, Nunez R, Schreider BD. Unexplained visual loss
33. Kiyosawa M, Asano T. A case of posterior ischemic optic after lumbar spinal fusion. J Neurosurg Anesthesiol.
neuropathy associated with pulseless disease. Nippon Ganka 1997;9(4):3468.
Gakkai Zasshi. 1982;86(11):20505. 53. Alexandrakis G, Lam BL. Bilateral posterior ischemic optic
34. Choudhari KA, Pherwani AA. Sudden visual loss due to poste- neuropathy after spinal surgery. Am J Ophthalmol.
rior ischemic optic neuropathy following craniotomy for a rup- 1999;127(3):3545.
tured intracranial aneurysm. Neurol India. 2007;55(2):1635. 54. Lee LA, Lam AM. Unilateral blindness after prone lumbar
35. Gerber CJ, Neil-Dwyer G, Kennedy P. Posterior ischaemic spine surgery. Anesthesiology. 2001;95(3):7935.
optic neuropathy after a spontaneous extradural haematoma. 55. Milner GA. A case of blindness after bilateral neck dissec-
J Neurol Neurosurg Psychiatry. 1992;55(7):630. tion. J Laryngol Otol. 1960;74:8805.
36. Kretz A, Preul C, Fricke HJ, Witte OW, Terborg C. Unilateral 56. Kirkali P, Kansu T. A case of unilateral posterior ischemic
optic neuropathy following subdural hematoma: a case optic neuropathy after radical neck dissection. Ann
report. J Med Case Reports. 2010;4(1):19. Ophthalmol. 1990;22(8):2978.
37. Hashimoto M, Ohtsuka K, Suzuki Y, Hoyt WF. A case of 57. Pazos GA, Leonard DW, Blice J, Thompson DH. Blindness
posterior ischemic optic neuropathy in a posterior-draining after bilateral neck dissection: case report and review. Am J
dural cavernous sinus fistula. J Neuroophthalmol. 2005; Otolaryngol. 1999;20(5):3405.
25(3):1769. 58. Warner ME, Warner MA, Garrity JA, MacKenzie RA,
38. Huang TW, Liu CM, Cheng PW, Yang CH. Posterior isch- Warner DO. The frequency of perioperative vision loss.
emic optic neuropathy following endoscopic sinus surgery. Anesth Analg. 2001;93(6):141721.
Otolaryngol Head Neck Surg. 2003;129(4):44850. 59. Dunker S, Hsu HY, Sebag J, Sadun AA. Perioperative risk
39. Perlman JI, Forman S, Gonzalez ER. Retrobulbar ischemic factors for posterior ischemic optic neuropathy. J Am Coll
optic neuropathy associated with sickle cell disease. J Surg. 2002;194(6):70510.
Neuroophthalmol. 1994;14(1):458. 60. Worrell L, Rowe M, Petti G. Amaurosis: a complication of
40. Buono LM, Foroozan R, Savino PJ, Danesh-Meyer HV, bilateral radical neck dissection. Am J Otolaryngol.
Stanescu D. Posterior ischemic optic neuropathy after hemo- 2002;23(1):569.
dialysis. Ophthalmology. 2003;110(6):12168. 61. Wessels IF. Posterior ischemic optic neuropathy during gen-
41. Aydin P, Yilmaz G, Kilinc M, Akbulut A. Posterior ischemic eral surgery. Am J Ophthalmol. 1987;104(5):5556.
optic neuropathy in chronic renal failure. Neuroophthalmology. 62. Remigio D, Wertenbaker C. Post-operative bilateral vision
2000;23:180. loss. Surv Ophthalmol. 2000;44(5):42632.
42. Su DH, Ang PS, Tow SL. Bilateral posterior ischemic optic 63. Rizzo 3rd JF, Lessell S. Posterior ischemic optic neuropathy
neuropathy associated with use of sildenafil. J during general surgery. Am J Ophthalmol. 1987;
Neuroophthalmol. 2008;28(1):75. 103(6):80811.
43. Choi JH, Choi KD, Kim JS, Kim HJ, Lee JE, An SJ. 64. Savino PJ, Burde RM, Mills RP. Visual loss following intra-
Simultaneous posterior ischemic optic neuropathy, cerebral nasal anesthetic injection. J Clin Neuroophthalmol.
border zone infarction, and spinal cord infarction after cor- 1990;10(2):1404.
rection of malignant hypertension. J Neuroophthalmol. 65. Enriquez JN, Guerrero JL. Posterior ischaemic optic neu-
2008;28(3):198201. ropathy after intranasal anesthetic injection: case report with
44. Hayreh SS. Posterior ischaemic optic neuropathy: clinical OCT findings. Neuroophthalmology. 2009;33(4):199201.
features, pathogenesis, and management. Eye. 2004;18(11): 66. Asensio JA, Forno W, Castillo GA, Gambaro E, Petrone P.
1188206. Posterior ischemic optic neuropathy related to profound
45. Hayreh SS, Joos KM, Podhajsky PA, Long CR. Systemic shock after penetrating thoracoabdominal trauma. South
diseases associated with nonarteritic anterior ischemic optic Med J. 2002;95(9):10537.
neuropathy. Am J Ophthalmol. 1994;118(6):76680. 67. Rath EZ, Falick Y, Rumelt S. Posterior ischemic optic neu-
46. Hayreh SS. Acute ischemic disorders of the optic nerve: ropathy following breast augmentation and abdominal lipo-
pathogenesis, clinical manifestations and management. suction. Can J Ophthalmol. 2009;44(3):3467.
Ophthalmol Clin North Am. 1996;9(3):40742. 68. Weber ED, Colyer MH, Lesser RL, Subramanian PS.
47. Roth S, Barach P. Postoperative visual loss: still no answers Posterior ischemic optic neuropathy after minimally inva-
yet. Anesthesiology. 2001;95(3):5757. sive prostatectomy. J Neuroophthalmol. 2007;27(4):2857.
69. Luscavage LE, Volpe NJ, Liss R. Posterior ischemic optic 75. Hayreh SS. Anterior ischemic optic neuropathy. VIII.
neuropathy after uncomplicated cataract extraction. Am J Clinical features and pathogenesis of post-hemorrhagic
Ophthalmol. 2001;132(3):4089. amaurosis. Ophthalmology. 1987;94(11):1488502.
70. Kordic H, Flammer J, Mironow A, Killer HE. Perioperative 76. Duke-Elder S, Wybar KC. Systems of ophthalmology, vol.
posterior ischemic optic neuropathy as a rare complication 2. London: Kimpton; 1961. p. 6435.
of blepharoplasty. Ophthalmologica. 2005;219(3):1858. 77. Hoyt WF, Luis O. Visual fiber anatomy in the infragenicu-
71. Retout A, Charlin JF, Brasseur G, Didier T. Amaurose lors late pathway of the primate. Arch Ophthalmol. 1962;
de la chirurgie du strabisme. A propos dun cas. (Amaurosis 68:94106.
during surgery of strabismus. Apropos of a case). Bull Soc 78. Hayreh SS. In vivo choroidal circulation and its watershed
Ophtalmol Fr. 1989;89(2):2479. zones. Eye. 1990;4(Pt 2):27389.
72. Gill B, Heavner JE. Postoperative visual loss associated with 79. Hayreh SS. Posterior ciliary artery circulation in health and
spine surgery. Eur Spine J. 2006;15(4):47984. disease: the Weisenfeld lecture. Invest Ophthalmol Vis Sci.
73. Cheng MA, Todorov A, Tempelhoff R, McHugh T, Crowder 2004;45(3):74957, 748.
CM, Lauryssen C. The effect of prone positioning on
intraocular pressure in anesthetized patients. Anesthesiology.
2001;95(6):13515.
74. Reddy A, Foroozan R, Edmond JC, Hinckley LK. Dilated
superior ophthalmic veins and posterior ischemic optic neu- For a complete bibliography and detailed review of previously
ropathy after prolonged spine surgery. J Neuroophthalmol. published studies in the literature on the subject, please refer to
Clinical Features of Posterior Ischemic
Optic Neuropathy 21
As mentioned earlier (see Chap. 19), posterior isch- Visual Acuity

emic optic neuropathy (PION) can be classified into
three types: (1) arteritic PION due to giant cell arteri-
This may vary from normal to no light perception
tis, (2) non-arteritic PION due to causes other than
Table21.1 gives the information of visual acuity in my
giant cell arteritis, and (3) surgical PION attributable
series [2] of 53 eyes with PION. Five of the eyes in that
to a surgical procedure. Therefore, their clinical fea-
series had also experienced amaurosis fugax before
tures may show differences.
visual loss (3 of them with arteritic PION).
Demographic Characteristics Visual Field Defects
Clinically, patients with arteritic and non-arteritic In PION there are optic nerve-related visual field defects.
PION typically present with acute, painless visual Table 21.2 gives details of the types of visual field
loss in one or both eyes, sometimes discovering it defects seen in 45 eyes with PION in my series [2].
upon waking up in the morning. In some eyes it may This shows that a central visual field defect is the most
initially be progressive. Patients with surgical PION common type (Figs. 21.121.3) 84% in non-arteritic
usually develop visual loss during a prolonged, major PION and 69% in arteritic PION. However, in contrast to
surgical procedure; they discover it as soon as they that, a marked generalized peripheral constriction, with
are alert, which may be several days after surgery. only a small normal central residual field, was seen in 3
Surgical PION usually tends to cause bilateral mas- eyes (1 eye with non-arteritic PION and 2 eyes with arter-
sive visual loss or even complete blindness which is itic PION) (Fig. 21.4). When only a peripheral island
usually permanent [1, 2]. In my series [2] of 43 visual field was present, in the 5 non-arteritic PION eyes,
patients (53 eyes), there was non-arteritic PION in it was in the temporal periphery in 2, inferior temporal
66% of the eyes, arteritic PION in 26% and surgical periphery in 1 and inferior nasal periphery in 2; in the 2
PION in 8%. Non-arteritic PION was seen in 17 arteritic PION eyes it was located in the inferior periph-
women and 11 men (both eyes in 7), arteritic PION in ery in 1 and inferior temporal periphery in 1.
10 women and 2 men (both eyes in 2), and surgical
PION in 2 women and one man (both eyes in 1). Age
range was 2090 (median 61.5, interquartile 5270) Comparison of Visual Field
years in non-arteritic PION, 6283 (median 73.4,
Defects in Non-Arteritic PION
interquartile 7179) years in arteritic PION, and sur-
gical PION was seen in persons 17, 55 and 82years and Non-Arteritic AION
old. Of the 12 patients with arteritic PION, 3 had
occult giant cell arteritis [3] with no systemic symp- To determine whether the pattern of visual field
toms or signs of GCA but all arteritic PION patients defects in non-arteritic AION and PION can help us
had a positive temporal artery biopsy. to differentiate them, I [2] compared the fields

438 21 Clinical Features of Posterior Ischemic Optic Neuropathy
Table21.1 Visual acuity at the initial visit in the three types of Table21.2 Visual field defects at the initial visit in arteritic and
PION non-arteritic PION
Non-arteritic Arteritic Surgical Type of visual field defect Non-arteritic Arteritic
Visual acuity PION PION PION PION PION
20/2020/25 6 4 0 Total eyesa 32 13
20/3020/40 1 2 0 Superior altitudinal defect 3 0
20/5020/70 3 1 0 Inferior altitudinal defect 0 1
20/80 1 0 0 Central scotoma alone 6 3
20/20020/400 5 3 0 Superior nasal paracentral 1 2
scotoma
Counting 14 2 1
fingers Inferior nasal paracentral 1 0
scotoma
Hand motion 2 0 0
Central scotoma with other 9 2
Light perception 2 1 0 field defects
No light 1 1 3 Centrocecal scotoma alone 5 0
perception
Marked generalized 1 2
Total eyes 35 14 4 constriction only with no
Reproduced from Hayreh [2] central scotoma
Nasal peripheral loss 1 0
plotted with a Goldmann perimeter (with I-2e, I-4e Temporal peripheral loss 2 0
and V-4e targets) in the two types of neuropathy in Only peripheral island field 5 2
my clinic. remaining
Inferior nasal defect 4 2
Superior temporal defect 1 1
Visual Field Defects in Non-arteritic AION Superior nasal defect 2 0

Peripheral constriction with 3 0
central defect
I analyzed the data about the pattern of central and
Reproduced from Hayreh [2]
peripheral visual field defects in 312 consecutive eyes a
Visual acuity no light perception or too poor to record visual
with non-arteritic AION at the initial visit to my clinic fields in 3 non-arteritic PION eyes and 1 arteritic PION eye
[4]. That study showed an absolute inferior nasal sec-
tor defect in 36% of cases in non-arteritic AION; in
contrast to that, absolute inferior altitudinal defects the 16 eyes with only a residual peripheral island field
were seen in only 8% (see Chap. 16). Thus, contrary to remaining with V-4e isopter, none was in the nasal
the prevalent impression, an absolute inferior nasal region, but in 14 eyes it was present in the temporal
sector defect is the most common type of visual field region.
defect in non-arteritic AION rather than an inferior
altitudinal defect. Other types of visual field defect are
much less common. Overall, loss of nasal part of the
visual field is the most common occurrence in non- Visual Field Defects in Non-arteritic PION
arteritic AION. This is because: (a) A nasal field defect
(i.e., nasal quadrant+nasal step+nasal vertical) was These patterns in non-arteritic PION are very differ-
seen in 44%; whereas the corresponding temporal field ent from those in non-arteritic AION, as shown in
defect (i.e., temporal quadrant+temporal wedge) Table 21.2 above. When evaluating the visual field
occurred in only 11%. (b) A nasal vertical, almost defects in AION and PION and comparing their
hemianopic, visual field defect was seen in 2%, but no prevalence in different published studies, it is impor-
temporal vertical defect was seen in that study. (c) Of tant to bear in mind that we cannot compare visual
Comparison of Visual Field Defects in Non-Arteritic PION and Non-Arteritic AION 439
a b
c d
Fig.21.1 Four visual fields showing varying sizes and densities (b) an 87year-old woman; and (c) and (d) fields in the left and
of central scotoma with normal peripheral visual fields in non- right eyes respectively of a 52year-old man. Visual fields plotted
arteritic PION. Right eye fields of (a) a 74year-old woman and with automated perimeter (Reproduced from Hayreh [2])
field defects plotted with the automated 24 or 30 normal peripheral fields beyond the central 24 or
perimetry now commonly used with those plotted 30 (Fig. 21.5). That has functional importance,
with a Goldmann perimeter. This is because, among because that field is helpful for the patients day to
other problems, a large inferior scotoma extending day navigation. Thus, in conclusion, inferior nasal
beyond the central 24 or 30 may erroneously be or inferior altitudinal field defects are typical of
classified on automated perimetry as an inferior alti- AION but are rare in PION. Central visual field loss,
tudinal defect. I have seen that happen time and by contrast, is the most common type of defect in
again (Fig. 21.5). This differentiation is important PION. Loss of the peripheral field with an intact
because eyes with the so-called inferior altitudinal central field is far more common in PION than in
defect on automated perimetry may have perfectly non-arteritic AION.
a b
Fig. 21.2 Visual fields of both eyes of a 74 year old woman visual acuity of 20/300 in both eyes. Visual fields plotted with
with non-arteritic PION. (a) and (b) are visual fields of left and automated perimeter (Reproduced from Hayreh [2])
right eyes respectively, 2days after the onset of visual loss, with
Fig.21.3 Visual field of left eye of an 81year old man with non-arteritic PION on the day of the visual loss, with visual acuity only
count fingers. Visual fields plotted with automated perimeter (Reproduced from Hayreh [2])
Optic Disc and Fundus Changes study [2], disc pallor usually developed in 68weeks,
although sometimes it was as soon as only 34weeks,
Initially there is no fundus or optic disc abnormality while in other cases it was over 8 weeks. There
on ophthalmoscopy and fluorescein fundus angiog- was no evident difference in the optic disc pallor
raphy. Unlike non-arteritic AION, the optic disc in among the different types of PION. In my study [2],
non-arteritic PION does not commonly have an rarely the optic disc developed cupping in non-
absent cup. The disc usually develops pallor which arteritic PION; this was also reported by Sonty and
is usually more marked in the temporal part. In my Schwartz [5].
Diagnosis of PION 441
a b
Fig.21.4 Visual fields of (a) right and (b) left eyes of a 79year- fields in both eyes. Visual fields plotted with automated perimeter
old woman with arteritic PION, showing remaining, markedly (Reproduced from Hayreh [2])
constricted central visual fields, with complete loss of peripheral
The Rest of the Ophthalmic Evaluation initially a chalky white swelling [612] and later on
most optic discs develop cupping [615], indistin-
guishable from that seen in glaucomatous optic neu-
External and slit-lamp evaluation of the anterior seg-
ropathy, except that the disc rim is pale, whereas it is
ment, intraocular pressure measurements, and ophthal-
of normal color in glaucomatous optic neuropathy. In
moscopy reveal no abnormality in these eyes, except
arteritic AION, fluorescein angiography shows evi-
for age-related lens changes, and the presence of rela-
dence of occlusion of the involved posterior ciliary
tive afferent pupillary defect in the involved eye in
artery [612], but in arteritic PION there is no abnor-
uniocular PION.
mality on angiography.
Differentiation of Arteritic from Diagnosis of PION

Non-arteritic PION
Since the diagnosis of PION, and especially non-
Unlike arteritic AION, initial ophthalmoscopic and arteritic PION, is usually hard to make with certainty,
fluorescein fundus angiographic findings are of no it is difficult to ascertain its true incidence. When I
help to differentiate arteritic from non-arteritic PION, first described this as a disease entity, I stressed that
because in both types they show no abnormality; for PION is a diagnosis of exclusion [16]. It should be
that reason, a diagnosis of giant cell arteritis is essen- made only after all other possibilities have been
tially based on their systemic symptoms and signs, carefully ruled out, e.g., macular and retinal lesions,
and elevated erythrocyte sedimentation rate and non-arteritic AION, retrobulbar optic neuritis, com-
C-reactive protein levels (see Chap. 11). Therefore, pressive optic neuropathy, other optic disc and optic
with all PION patients 50years and older, it is essen- nerve lesions, neurological lesions, hysteria, even
tial to rule out giant cell arteritis. By contrast, ophthal- malingering, and a host of other lesions. In my study
mic findings are valuable in the differentiation of [2] none of patients showed any evidence of any of
arteritic from non-arteritic AION [6, 7], as discussed those lesions during follow-up. The findings of my
in Chap. 12, because in arteritic AION usually there is study [2], as well as those of others [1, 17], indicate
Fig. 21.5 This shows the difference in pattern of visual field perimeter was interpreted in automated perimetry field in (a) as
defect revealed by visual fields plotted with an automated perim- inferior altitudinal defect. Visual fields plotted with automated
eter (a) and Goldmann perimeter (b) in an eye with NA-AION. perimeter
Inferior paracentral scotoma seen in (b) plotted with Goldmann
References 443
b impression that surgical PION is more common than

it actually is.
References
1. Sadda SR, Nee M, Miller NR, Biousse V, Newman NJ,

Kouzis A. Clinical spectrum of posterior ischemic optic neu-
ropathy. Am J Ophthalmol. 2001;132(5):74350.
2. Hayreh SS. Posterior ischaemic optic neuropathy: clinical
features, pathogenesis, and management. Eye. 2004;
18(11):1188206.
3. Hayreh SS, Podhajsky PA, Zimmerman B. Occult giant cell
arteritis: ocular manifestations. Am J Ophthalmol.
1998;125(4):5216, 893.
4. Hayreh SS, Zimmerman B. Visual field abnormalities in
nonarteritic anterior ischemic optic neuropathy: their pattern
and prevalence at initial examination. Arch Ophthalmol.
2005;123(11):155462.
5. Sonty S, Schwartz B. Development of cupping and pallor in
Fig.21.5 (continued) posterior ischemic optic neuropathy. Int Ophthalmol.
1983;6(3):21320.
that a combination of the following is highly sugges- 6. Hayreh SS. Anterior ischaemic optic neuropathy: differen-
tiation of arteritic from non-arteritic type and its manage-
tive of PION: ment. Eye. 1990;4(1):2541.
1. Sudden onset of visual deterioration, with or with- tis: our 27-year clinical study: new light on old controver-
out deterioration of central visual acuity, sies. Ophthalmologica. 2003;217(4):23959.
2. Optic nerve-related visual field defects in the 8. Hayreh SS. Anterior ischaemic optic neuropathy. II: fundus
involved eye, particularly the two typical types of on ophthalmoscopy and fluorescein angiography. Br J
Ophthalmol. 1974;58(12):96480.
visual field defects discussed above [central sco- 9. Hayreh SS. Anterior ischemic optic neuropathy. New York:
toma with almost normal peripheral field, or loss of Springer; 1975.
peripheral field with intact central field], 10. Hayreh SS. Ischemic optic neuropathy. Int Ophthalmol.
3. The presence of a relative afferent pupillary defect 1978;1:918.
11. Hayreh SS. Ischemic optic neuropathy. Prog Retin Eye Res.
in the involved eye in patients with a perfectly nor- 2009;28(1):3462.
mal fellow eye, 12. Hayreh SS, Podhajsky PA, Zimmerman B. Ocular manifes-
4. An initially normal optic disc and fundus on tations of giant cell arteritis. Am J Ophthalmol.
ophthalmoscopy and fluorescein fundus 1998;125(4):50920.
13. Hayreh SS. Pathogenesis of cupping of the optic disc. Br J
angiography, Ophthalmol. 1974;58(10):86376.
5. No other ocular, orbital or neurological abnormality 14. Danesh-Meyer HV, Savino PJ, Sergott RC. The prevalence of
to explain the visual loss, and cupping in end-stage arteritic and nonarteritic anterior isch-
6. Development of optic disc pallor, usually within emic optic neuropathy. Ophthalmology. 2001;108(3):5938.
15. Danesh-Meyer H, Savino PJ, Spaeth GL, Gamble GD.
68weeks. Comparison of arteritis and nonarteritic anterior ischemic
optic neuropathies with the Heidelberg Retina Tomograph.
The optic discs in eyes with non-arteritic AION usu- Ophthalmology. 2005;112(6):110412.
ally have an absent cup (see Chap. 16); this is not the 16. Hayreh SS. Posterior ischemic optic neuropathy.
Ophthalmologica. 1981;182(1):2941.
case in PION. 17. Isayama Y, Takahashi T, Inoue M, Jimura T. Posterior isch-
The diagnosis of surgical PION, on the other emic optic neuropathy. III: clinical diagnosis.
hand, is relatively straightforward; these patients Ophthalmologica. 1983;187(3):1417.
discover dramatic visual loss as soon as they are alert
enough after a major surgical procedure, and have
clinical findings detailed above. Because surgical
For a complete bibliography and detailed review of previously
PION is easy to diagnose and often results in published studies in the literature on the subject, please refer to
medico-legal cases, there may be an erroneous bibliography in previous publications listed here.

Management of Posterior Ischemic
Optic Neuropathy 22
As mentioned earlier (Chap. 19), posterior ischemic In my study [6], all patients with non-arteritic PION,
optic neuropathy (PION) is of three types: (1) arteritic were given the option to try corticosteroid therapy, if
PION due to giant cell arteritis, (2) non-arteritic PION they wished (patient randomization); it was fully
due to other causes, and (3) surgical PION, attributable explained in advance that there was no known treat-
to a surgical procedure. Therefore, the management of ment for this condition and that we had no knowledge
PION depends upon the type of PION. In all cases whether corticosteroid therapy was beneficial, harmful
other than surgical PION, the most important first step or ineffective. All the pros and cons of the therapy were
in persons aged 50years or older is always to rule out discussed. The patients were advised to consult their
GCA, as in AION [13]. physician or any other person before they made a deci-
sion on whether to start the corticosteroid therapy or
not. We had no input at all into their decision; this
was a similar procedure to the one we used for corti-
Management of Arteritic PION costeroid therapy for non-arteritic AION [7] and fol-
lowed the same protocol of treatment, as discussed in
In this type of PION, treatment is basically the manage- Chap. 17. Of 28 patients with non-arteritic PION, 14
ment of giant cell arteritis by early diagnosis and opted to try the corticosteroid therapy, and it was started
prompt and aggressive high-dose systemic corticos- the day of the initial visit to the clinic or soon thereaf-
teroid therapy, to prevent further visual loss, as dis- ter. The initial dose was usually 80mg oral Prednisone
cussed fully in Chap. 13. The corticosteroid therapy daily, except for three who were given a single intrave-
regimen I use in these patients is exactly similar to that nous mega-dose of corticosteroids initially this was
for arteritic AION, as discussed in detail elsewhere completely random, without any particular reason to
[35] and in Chap. 13. In my study [6], patients with do so. Usually 80mg Prednisone was given for 2weeks
arteritic PION, like arteritic AION [3], showed no sig- and then gradually tapered off, with the whole corti-
nificant visual improvement with corticosteroid ther- costeroid therapy lasting for about 2months.
apy. However, corticosteroid therapy has to be continued
to prevent further visual loss (see Chap. 13).
Effect of Corticosteroid Therapy
on Visual Outcome in Non-arteritic
Management of Non-arteritic PION PION in My Study [6]
Since there was no known treatment for this type of The following criteria were used for visual acuity
PION and no information available as to whether corti- improvement:
costeroid therapy is beneficial to these patients or not, I (a) Visual acuity improvement must be at least 2
[6] investigated this in my study. I will discuss this study lines of Snellen chart.
in detail here because its results are encouraging, given (b) There must be concomitant improvement in the
that this is a disease with no known treatment so far. central 50 of visual fields, to rule out apparent

446 22 Management of Posterior Ischemic Optic Neuropathy
visual acuity improvement from simple eccentric were excluded from data analysis for visual acuity
fixation or learning experience. improvement. Of the remaining 12 eyes with visual
(c) The visual improvement must have been main- acuity of 20/70 or worse, 4 (33%) improved.
tained on follow-up and not be just fluctuating or The treated eyes showed a significant improvement
transient. from baseline acuity (p=0.031), with mean acuity at
final follow-up being significantly better than the
This was evaluated in 32 of the 35 eyes with non-arter- untreated eyes (p=0.023). Thus, the findings of this
itic PION, because the remaining 3 eyes of 2 patients study showed that eyes of patients treated with high
had a follow-up of less than 2weeks and were excluded dose systemic corticosteroid therapy showed significant
from this evaluation. Patients with 16 of the 32 eyes visual acuity improvement compared to untreated eyes.
opted for high dose corticosteroid therapy and the rest
opted for no treatment. Visual improvement was evalu-
ated on follow-up by the change in visual acuity, as Visual Acuity Stable or Worse in Non-arteritic PION
well as central and peripheral visual fields (plotted
with a Goldmann perimeter). Visual acuity and visual Of the 16 eyes in the treated group, 5 remained stable
field evaluation techniques were similar to those used (20/2020/25 in 2, 20/80 in 1, hand motion in 1 and no
in the non-arteritic AION corticosteroid therapy study light perception in 1), and 2 deteriorated (from 20/40
[7] (see Chap. 17). to 20/400, and 20/6020/400). Of the 16 eyes in
Time Interval between Onset of Visual Loss and untreated group, 9 remained stable (20/2520/25 in 4,
Start of Corticosteroid Therapy: This was as follows. and count fingers in 5), and 2 deteriorated (from 20/200
(1) In 9 eyes with improvement: the therapy was started to count fingers).
the same day in 5, and 2, 7 and 9days later in 2, 1 and Time Interval Between Start of Corticosteroid
1 respectively. (2) In 2 eyes with deterioration: it was Therapy and Onset of Visual Improvement: It was
started 27 days after the onset when the patient was one day after the start of therapy in 4, and after 7,9,13
first seen. (3) In 5 eyes with stable visual acuity: of and 16days, with no data in one eye. In the 4 untreated
20/20, 20/25, 20/80, hand motion and no light percep- eyes with visual acuity improvement, it was seen
tion, the therapy was started 1, 27, 7, 5 and 24 days 4,6,43 and 53days after the onset of PION.
respectively after onset.
Visual Field Change

Visual Acuity Change
Mean visual field grade at baseline did not differ sig-
Visual Acuity Improvement in Non-arteritic PION nificantly between the corticosteroid treated and
untreated eyes (p=1.000) indicating that the two
Of the eyes that had visual acuity at baseline 20/40 or groups were similar to begin with. Figure22.2 shows
worse, mean visual acuity (log MAR) at baseline did the changes in the visual fields in eyes with non-arter-
not differ significantly between the corticosteroid itic PION (both treated and untreated groups) during
treated and untreated eyes (p=0.165), indicating that follow-up. The eyes of the untreated group did not
the two groups had similar visual acuity to begin with. show a significant improvement from baseline
In the treated group of 16 eyes, 9 showed improvement (p=0.465). In contrast to that, the eyes of the treated
by 2 lines of Snellen chart. Two of the 16 eyes had group showed a significant improvement from baseline
initial visual acuity of 20/2020/25, and therefore could (p<0.001) and it was significantly greater (p=0.030)
not have shown any significant improvement; they were than in the untreated eyes (see below). The data also
excluded from data analysis on visual acuity improve- indicated a better mean visual field grade at final fol-
ment. Of the remaining 14 eyes with initial visual acuity low-up for the treated eyes compared to the untreated
of 20/40 or worse, 9 eyes (64%) showed visual improve- eyes (p=0.064). Thus, the findings of this study showed
ment. Among the 16 eyes of untreated patients, 4 that the eyes of patients treated with high-dose sys-
showed visual acuity improvement (Fig.22.1). Again, 4 temic corticosteroid therapy showed significant visual
of the 16 eyes, with initial visual acuity of 20/2020/25, field improvement compared to untreated eyes.
Management of Non-arteritic PION 447
Fig.22.1 Two graphs showing

change from initial to final
visual acuity in 16 eyes of
patients treated with high-dose
corticosteroid therapy (upper
graph) and 16 eyes of those not
treated (lower graph)
Table 22.1 summarizes the findings about the visual The following 3 examples demonstrate the amount
field response to corticosteroid therapy in PION. of visual improvement seen after corticosteroid ther-
An addendum to the study [6] also reported another apy in non-arteritic PION.
case of a 46year old woman with non-arteritic PION, CASE 1: A 74-year old white woman with non-
with an initial visual acuity of 20/80 and an absolute arteritic PION was seen in my clinic 2days after the
centrocecal scotoma and normal peripheral fields. onset of visual loss, with visual acuity of 20/300 in
With high dose corticosteroid therapy started 6 days both eyes, and visual field defects as shown in
after onset, her visual acuity improved to 20/20 and her Fig. 22.3a and b in the left and right eyes respec-
visual fields were normal. tively. She was started on 80 mg Prednisone daily
Treated Non-Arteritic PION Untreated Non-Arteritic PION

normal normal
(2) (3)
mild mild
defect defect
Final visual vield grade

Final visual field grade
moderate moderate
defect defect
marked marked
defect (2) defect (3)
severe severe
defect defect (4)
defect
severe
defect
marked
defect
moderate
defect
mild
normal
defect
severe
defect
marked
defect
moderate
defect
mild
normal
Initial visual field grade Initial visual field grade
Fig.22.2 Two graphs showing change from initial to final visual fields in 15 eyes of patients treated with high-dose corticosteroid
therapy (left graph) and 15 eyes of those not treated (right graph). Normal normal visual fields (Reproduced from Hayreh [6])
that day. The visual field of the left eye was normal field is shown in Fig.22.4a. He was started on 80mg
13 days after the start of the therapy (Fig. 22.3c), Prednisone daily that day. The next day the visual
with visual acuity of 20/20. Thirty-four days after field (Fig.22.4b) had improved, with visual acuity
the start of the therapy the visual field of the right of 20/30. Fifteen days later the visual field showed
eye had improved markedly (Fig. 22.3d) (compare marked improvement in the central field (Fig.22.4c),
22.3b and 22.3d) and the visual acuity was 20/40. with visual acuity of 20/25. The total duration of
After 6months follow-up the visual fields and acu- corticosteroid therapy was 2months. On a follow-
ity remained stable. up of 19 months the visual fields and acuity were
CASE 2: An 81-year old white man was seen on stable.
the day of visual loss due to non-arteritic PION. His CASE 3: An 18year old black woman with sys-
visual acuity was counting fingers and the visual temic lupus, while in our hospital, woke up with
Table22.1 Visual field response to corticosteroid therapy in non-arteritic and arteritic PION
Type of PION Visual field change Central field only Peripheral field Both central and
during follow-up (eyes) only (eyes) peripheral field
defects (eyes)
Corticosteroid treated Improved 4 2 6
non-arteritic PION groupa Stable 2 4 1
(15 eyes) Worse 0 0 2
Untreated non-arteritic Improved 2 3 3
PION groupa (15 eyes) Stable 3 2 3
Worse 0 0 4
Arteritic PION groupb Improved 2 1 1
(13 eyes) Stable 1 2 9
Worse 0 0 0
Reproduced from Hayreh [6]
a
One eye in both the treated and untreated groups had unreliable fields for evaluating change
b
One eye had unreliable fields for evaluating change
Management of Non-arteritic PION 449
c d
Fig.22.3 Visual fields of both eyes of case 1 with non-arteritic PION, showing marked improvement with high-dose steroid therapy
marked visual loss in her left eye. She had had non- Conclusion
arteritic PION in her right eye 19months earlier, but
at the time she was accused of malingering as all the The statistical data analysis proved that the non-arter-
ophthalmic findings were negative. She was seen in itic PION eyes of patients treated with high-dose sys-
my clinic that day and found to have non-arteritic temic corticosteroid therapy showed significant visual
PION in that eye, with visual acuity of counting fin- acuity and visual field improvement, compared to
gers in both eyes and visual field defects as shown in untreated eyes. Also, the magnitude of the improve-
Figs. 22.5a and b in the right and left eyes respec- ment in visual acuity and visual fields was much
tively. Her right optic disc was atrophic and the left greater in the treated group than the untreated group.
was normal. She was started on high-dose corticos- Thus, the findings of this study indicate a beneficial
teroid therapy that day. Two weeks later her visual effect on visual function from high-dose systemic
acuity in the left was 20/40 and the visual field corticosteroid therapy during the very early stages of
showed marked improvement (Fig.22.5c), and 2days the disease (Figs. 22.322.5). However, the study
after that the visual acuity was 20/20. When the showed that spontaneous improvement in visual acu-
visual field was plotted 3weeks later, it was normal ity and visual field may also occur to some extent, in
(Fig.22.5d). some eyes, without corticosteroid therapy. Sadda
b c
Fig.22.4 Three visual fields of the left eye of case 2 with non-arteritic PION, showing marked improvement with high-dose steroid
therapy (Reproduced from Hayreh [6])
etal. [8] reported that visual acuity improved in 34%, testing varied considerably, and (e) no follow-up
remained stable in 28% and worsened in 38%, but visual acuity data were ascertainable for 22% of
they did not mention whether any of their patients patients. Another confounding factor, often forgot-
were treated or not. The authors rightly cautioned ten when reporting visual acuity improvement, is that
against drawing any definite conclusion from their unless there is a corresponding improvement in the
data about this visual acuity change, because the central 5 of the visual fields, an improvement in
visual acuity data in their series had several potential visual acuity alone may simply represent the patients
drawbacks, including these: (a) visual acuity testing learning to fixate eccentrically [3, 9, 10].
was not standardized, (b) it was not with best refrac- Killer et al. [11] claimed visual improvement in
tion, (c) in some cases initial and follow-up visual two cases following optic nerve sheath decompres-
acuity measurements were obtained from different sion. As in non-arteritic AION, optic nerve sheath
offices, (d) the times to initial and final visual acuity decompression has no scientific rationale whatever
Management of Surgical PION 451
Fig.22.5 Visual fields of right (A) and left (B-D) eyes of case 3 with old non-arteritic PION in the right eye and new in the left eye.
High-dose steroid therapy improved the visual field in the left eye markedly (compare b with d)
in non-arteritic PION [12, 13] (see Chap. 17). Management of Surgical PION
Moreover, as noted above, some PION patients show
visual improvement spontaneously, without any There were only 4 eyes (3 patients) in my study [6]. Two
treatment. patients (three eyes with no light perception) were
Treating arteritic PION patients with high-dose cor- treated with intravenous corticosteroid therapy immedi-
ticosteroid therapy is mandatory to prevent any further ately on the discovery of visual loss, but neither of them
visual loss. None of the 14 eyes in my study [6] treated recovered any vision. The untreated eye also did not
with high-dose corticosteroid therapy showed any sig- improve.
nificant visual acuity change. Table 22.1 shows the Basically, the management of surgical PION
changes in the central and peripheral visual fields in amounts to using adequate prophylactic measures to
these eyes during follow-up; however, the mild prevent its development, because once the visual loss
improvement seen in visual fields in some was not suf- occurs, it is usually bilateral, severe and irreversible. In
ficient to improve visual function very much. these cases, no treatment has been found to be effective
Importantly, since systemic risk factors may play a to recover or improve the lost vision. Prophylactic mea-
part in the development of non-arteritic PION, as in sures during surgery include avoiding arterial hypoten-
non-arteritic AION (see Chap. 14), one should try to sion, excessive fluid replacement and hemodilution,
reduce as many risk factors as possible to minimize the pressure on the eyeball and orbit and dependent posi-
risk of second eye involvement (see Chap. 17). tion of the head, as well as shortening the duration of
surgery to the minimum. Since systemic cardiovascu- of a large study and review of literature. Acta Ophthalmol
lar risk factors may predispose a patient to a higher risk Scand. 2002;80(4):35567.
4. Hayreh SS, Zimmerman B. Visual deterioration in giant cell
of developing surgical PION, it may be advisable to arteritis patients while on high doses of corticosteroid ther-
weigh those factors in the decision to perform surgery. apy. Ophthalmology. 2003;110(6):120415.
tis. Our 27-year clinical study: new light on old controver-
sies. Ophthalmologica. 2003;217(4):23959.
Conclusion 6. Hayreh SS. Posterior ischaemic optic neuropathy: clinical
features, pathogenesis, and management. Eye.
2004;18(11):1188206.
My study [6] showed that visual prognosis varies with 7. Hayreh SS, Zimmerman MB. Non-arteritic anterior isch-
emic optic neuropathy: role of systemic corticosteroid ther-
the type of PION. Eyes with non-arteritic PION in apy. Graefes Arch Clin Exp Ophthalmol. 2008;
patients treated with high-dose corticosteroid therapy 246(7):102946.
at onset showed significantly greater visual improve- 8. Sadda SR, Nee M, Miller NR, Biousse V, Newman NJ,
ment than the untreated patients. Patients with arteritic Kouzis A. Clinical spectrum of posterior ischemic optic neu-
ropathy. Am J Ophthalmol. 2001;132(5):74350.
PION, if treated urgently and aggressively with high 9. Hayreh SS, Zimmerman MB. Central retinal artery occlu-
dose-corticosteroid therapy, showed no change in sion: visual outcome. Am J Ophthalmol. 2005;140(3):
vision, but it prevented further visual loss. Patients 37691.
with surgical PION usually suffer severe, irreversible 10. Hayreh SS, Zimmerman MB. Nonarteritic anterior ischemic
optic neuropathy: natural history of visual outcome.
visual loss and do not respond to corticosteroid ther- Ophthalmology. 2008;115(2):298305.
apy; visual loss often tends to be bilateral, as has also 11. Killer HE, Forrer A, Blumer BK. Nonarteritic posterior isch-
been reported by others [8]. Prevention is the only emic optic neuropathy treated with optic nerve sheath
remedy. decompression: report of two cases. Neuro-Ophthalmology.
1998;19(2):1016.
12. Hayreh SS. The role of optic nerve sheath fenestration in
management of anterior ischemic optic neuropathy. Arch
Ophthalmol. 1990;108(8):10635.
References 13. The Ischemic Optic Neuropathy Decompression Trial
Research Group. Optic nerve decompression surgery for
1. Hayreh SS. Anterior ischaemic optic neuropathy. nonarteritic anterior ischemic optic neuropathy (NAION) is
Differentiation of arteritic from non-arteritic type and its not effective and may be harmful. JAMA. 1995;
management. Eye. 1990;4(1):2541. 273(8):62532.
2. Hayreh SS, Podhajsky PA, Zimmerman B. Ocular manifes-
tations of giant cell arteritis. Am J Ophthalmol. 1998;
125(4):50920. For a complete bibliography and detailed review of previously
3. Hayreh SS, Zimmerman B, Kardon RH. Visual improve- published studies in the literature on the subject, please refer to
ment with corticosteroid therapy in giant cell arteritis. Report bibliography in previous publications listed here.
Index
A other fundus changes, 366

Animal model, 417423 gender, 338
Anterior ischemic optic neuropathy (AION), 2 incidence, 337
arteritic, 2 intermittent blurring, 340
non-arteritic, 2 optic disc changes, 354362
types, 2 in diabetes mellitus, 359, 371375
Arterial blood pressure role, optic nerve head optic disc edema, 354357
blood flow, 9193 optic disc pallor, 361362
arterial hypertension, 9192 photophobia, 340
arterial hypotension, 9293 progressive vision loss, 340
Autoregulation of blood flow, 8588 racial features, 338
mechanisms, 86 recurrence, 368371
Autoregulation of optic nerve head blood flow, 8688 refractive error, 345
breakdown, 87 seasonal variations, 344
deranged autoregulation, role of, 8788 simultaneous bilateral onset, 340344
range, 8687 symptoms, 338344
time of onset of visual loss, 338340
B visual acuity, 344345
Blood-optic nerve barrier, 7982 natural history of, 344345
anatomical considerations, 79 visual disability, 350
hypoxia/ischemia, effect of, 82 visual field defects, 338339, 345354
optic nerve head blood barrier, 80 natural history of, 352353
Blood-optic nerve head barrier, clinical importance Clinical features, PION, 437443
hypertensive optic neuropathy, 8081 arteritic vs. non-arteritic PION, 441
methanol toxic optic neuropathy, 81 demographic characteristics, 437
NA-AION, 81 diagnosis, 441, 443
Blood supply of optic nerve, 3574. See also Optic nerve, fundus changes, 440
blood supply non-arteritic PION vs. non-arteritic
of posterior part, 427428 AION, 437439
intracanalicular part, 428 optic disc changes, 440
intracranial part, 428 temporal artery biopsy, 437
intraorbital part, 427429 visual acuity, 437, 438
visual field defects, 437443
C Controversial NA-AION, 317333
Central retinal artery occlusion, 216 following amiodarone, 322324
Cilioretinal artery occlusion, 217218 amiodarone optic neuropathy, 322324
Circle of Haller and Zinn, 5960 following erectile dysfunction drugs, 318323
Clinical evaluation of patients, 145161 familial NA-AION, 324327
Clinical features, NA-AION, 337384 incipient NA-AION, 317318, 330
age, 337338, 380 following intravitreal anti-VEGF, 330332
amblyopia development, 353354 following phosphodiesterase5 inhibitors,
anterior segment, 354 318323
bilateral, 340, 366369, 380 posthemorrhagic amaurosis, 328
corticosteroid therapy effect on optic disc edema, 359361
demographic characteristics, 337338 D
fluorescein fundus angiography, 362366 Diplopia, 220

DOI: 10.1007/978-3-642-11852-4, Springer-Verlag Berlin Heidelberg 2011
454 Index
E neuro-radiological tests, 192193

Endothelial derived vasoactive agents, 8891 occult giant cell arteritis, 180
causes of abnormalities, 8889 ophthalmic emergency, 174175
experimental endothelin1 studies, 9091 ophthalmic ischemic lesions, 211221
known agents, 88 amaurosis fugax, 211
ocular ischemic disorders, role in, 8990 anterior ischemic optic neuropathy, 201202, 212216
regulation of blood flow, role in, 89 anterior segment ischemic lesions, 219
arteritic anterior ischemic optic neuropathy,
G 201202, 212216
GCA. See Giant cell arteritis (GCA) arteritic posterior ischemic optic neuropathy, 216
Giant cell arteritis (GCA), 163264 central retinal artery occlusion, 216
age distribution, 163164 cerebral ischemic lesions producing visual loss, 221
American College of Rheumatology criteria, 174 choroidal ischemic lesions, 218
arteritic AION cilioretinal artery occlusion, 217218
clinical features, 174176 cotton-wool spots, 218
corticosteroid sparing therapy, 251254 extra-ocular motility disorders, 220
adalimumab (Humira), 252253 miscellaneous ophthalmic manifestations, 220221
aspirin, 253254 ocular ischemic syndrome, 219
azathioprine, 252 orbital complications, 220
cyclosporine A, 252 pupillary abnormalities, 219
enbrel (Etanercept), 252 ophthalmic manifestations, 199221
infliximab (Remicade), 252 classification, 211
methotrexate, 251252 ophthalmologist perspective, 179182
corticosteroid therapy, 228259 pathophysiology, 167168
alternate day therapy, 235 polymyalgia rheumatica manifestation, 180
duration of therapy, 230232, 235236 racial differences, 166167
guiding rules to prevent relapses, 236237 recurrences/relapses, 168
intravenous, 228229, 242, 249250 rheumatologist perspective, 179182
monitoring criteria, 232235, 236 seasonal variation, 164
morbidity and mortality, 254256 signs, 175, 176
oral, 229230, 242, 249250 systemic features, 174176
relapse, 236237 temporal artery abnormality, 175, 176
resistance to corticosteroid therapy, 238 temporal artery biopsy, 181182, 189191
side-effects, 250251 thrombocytosis, 184186
tapering down, 230232 visual loss
unique personal experience, 257259 anterior segment ischemic lesions, 219
vision deterioration while on corticosteroids, 243250 central retinal artery occlusion, 216
vision improvement with corticosteroids, 238242 cerebral ischemic lesions producing visual loss, 221
c-reactive protein, 183184 choroidal ischemic lesions, 218219
definite diagnosis, 182193 cilioretinal artery occlusion, 217218
diagnostic tests, 182193 extra-ocular motility disorders, 220
Doppler tests, 191192 miscellaneous ophthalmic manifestations, 220221
duration, 168 ocular involvement incidence, 202208
epidemiology, 164166 ocular ischemic syndrome, 219
erythrocyte sedimentation rate (Westergren), 182183 optic nerve ischemic lesions, 211216
fluorescein fundus angiography, 193 orbital complications, 220
gender distribution, 164 pupillary abnormalities, 219
genetic predisposition, 167 retinal ischemic lesions, 216218
other hematologic tests, 187189 second-eye involvement, 208210
juvenile, 164 site of lesion, 199202
management controversy to prevent visual loss, 179182
management to prevent blindness, 227260 I
controversy, 227228 Intraocular pressure role, optic nerve head blood flow, 9495
corticosteroid sparing therapy, 251254 Ischemic optic neuropathy, 13
corticosteroid therapy, 227250, 254259 relative frequency of various types, 23
misconceptions, 257 terminology, 1
patient education, 256257 types, 12
permanent visual loss, predictors for, 256
personal experience, 257259 M
preventing vision loss, 227260 Management, NA-AION, 389412
self-limited nature, 237238 aspirin, 393396
Index 455
bevacizumab, 408 size, 131133

corticosteroids, systemic, 398407 terminology, 131
disability, coping, 411412 Optic disc cup size, 137138
incipient NA-AION, management, 409 configuration and depth, 137138
levodopa, 396397 cup/disc ratio, 138
miscellaneous advocated treatments, 409 relation to optic disc size, 137
nocturnal arterial hypotension management, 410411 Optic disc edema, 155161
optic nerve sheath decompression, 389393 bilateral optic disc edema with visual field loss, 158
risk factors reduction, 410411 blood tests, 160161
triamcinolone, intravitreal, 407 electrophysiologic testing, 160
VEGF inhibitors, 408409 main ancillary tests in outlier cases, 158161
Misconceptions, NA-AION, 384 no optic disc edema with visual loss, 158
Multifactorial pathogenesis, NA-AION, 278, 307 procedures, 161
unilateral optic disc edema
N with visual field loss, 155157
NA-AION. See Non-arteritic anterior ischemic optic neuropa- without visual field loss, 157158
thy (NA-AION) Optic nerve, 733
Neuroretinal rim, 135137 angioarchitecture, 7074
pallor, 137 blood supply, 3574
shape, 136137 intracanalicular part, 6869
size, 135136 intracranial part, 6970
Non-arteritic anterior ischemic optic neuropathy intraorbital part, 6368
(NA-AION), 265308, 317333, optic nerve head, 3642
337384, 389412, 417423 fibers, 2833
blood-optic nerve head barrier, clinical importance, 81 topographic localization, 2933
clinical features (see Clinical features, NA-AION) head, 721
controversial (see Controversial NA-AION) lamina cribrosa region, 1415
management of (see Management, NA-AION) prelaminar region, 814
misconceptions, 384 surface nerve fiber layer, 8
multifactorial pathogenesis, 278, 307 intracanalicular part, 27
ocular risk factors (see Ocular risk factors, NA-AION) intracanalicular part blood supply, 6869
pathogenesis (see Pathogenesis, NA-AION) intracranial part, 2728
risk factors (see Systemic risk factors, NA-AION) intracranial part blood supply, 6970
types, special, 379384 intraorbital part, 2124
following amiodarone, 380381 intraorbital part blood supply, 6368
amiodarone optic neuropathy, 380381 anterior segment, 64
in diabetics, 372375, 376 posterior segment, 64
familial, 379 venous drainage, 6468
incipient, 375379 sheath, 2427
following phosphodiesterase5 inhibitors, 380 Optic nerve head blood flow
posthemorrhagic amaurosis, 381384 arterial blood pressure role, 9193
arterial hypertension, 9192
O arterial hypotension, 9293
Occult giant cell arteritis, 168169 factors influencing, 85102
Ocular risk factors, NA-AION, 297307 arterial blood pressure, 9193
blood-optic nerve head barrier, 307 autoregulation of blood flow, 8588
cup/disc ratio, 297302 central retinal venous pressure, 96100
intraocular pressure, 303 endothelial derived vasoactive agents, 8891
optic disc drusen, 303 other factors, 96101
optic disc edema, 304305 intraocular pressure, 9495
optic nerve head blood flow autoregulation, 306307 pathophysiology of factors, 85
optic nerve head vascular disorders, 306 posterior ciliary artery watershed zone location, 98101
post-cataract extraction, 304 resistance to blood flow, 9596
posterior ciliary artery watershed zone location, 305306 measurement methods, 114124
posterior vitreous detachment, 307 14C2-deoxyglucose (2-DG) method, 122123
refractive error, 302303 color Doppler imaging, 116118
Optic disc, 131141 fluorescein fundus angiography, 119121
definition of, 131 Heidelberg retina flowmeter, 116
histological findings, 141 hydrogen clearance method, 122
retinal arterioles, diameter of, 139 iodoantipyrine autoradiography, 123
shape, 133135 laser Doppler flowmetry, 114115
456 Index
laser speckle flowgraphy, 115116 incidence, 425

magnetic resonance imaging, 121 management
microsphere method, 123 arteritic, 2, 425, 445
morphologic studies, 123124 corticosteroids therapy, 445446
oxygen tension method, 122 non-arteritic, 2, 425, 445449
pulsatile ocular blood flow, 119 surgical, 2, 425, 451452
scanning laser Doppler flowmetry, 115 visual acuity, 446447
scanning laser fluorescein angiography, 121 visual field change, 446451
temperature measurement, 121122 pathogenesis (see Pathogenesis, PION)
transcranial Doppler ultrasound, 119 types, 2
Optic nerve head blood supply, 3663
controversies, 6063 R
inter-individual variations, 60 Resistance to blood flow, 9596
lamina cribrosa region, 3940 experimental vasomotor studies, 9596
prelaminar region, 3639 rheological properties of blood, 95
retrolaminar region, 4042 state and caliber of the vessels, 95
surface nerve fiber layer, 36 Retinal nerve fiber layer, 139141
venous drainage, 63 defects, 140
evaluation, 140
P thickness in different part of optic disc, 141
Parapapillary chorioretinal atrophy, 138139 Risk factors, NA-AION, 281307. See also
Pathogenesis, NA-AION, 265308 Systemic risk factors, NA-AION
histopathology of optic nerve, 265270
historical review, various hypotheses, 270278 S
multifactorial pathogenesis, 278, 280307 Stroke, 221, 253255, 276, 277, 295297, 319,
neural ischemia severity, 307308 322, 325, 326, 344, 384, 393, 396,
site of lesion, 270 403405, 418
vascular insufficiency mechanism in optic nerve head, and NA-AION differences, 403404
278280 Systemic risk factors, NA-AION, 284297
Pathogenesis, PION, 427434 atherosclerosis, 292293
arteritic, 430 cardiac valvular disease, 293
compared to AION, 433434 cardiovascular autoregulation defect, 293294
histopathology, 428430 hematologic disorders, 294296
non-arteritic, 430431 internal carotid artery disease, 293
risk factors, 430431 malignant arterial hypertension, 287288
surgical, 431432 migraine, 293
visual field defects, 433 nocturnal arterial hypotension, 289292
PCAs. See Posterior ciliary arteries (PCAs) seasonal variations, 296297
PION. See Posterior ischemic optic neuropathy (PION) sleep apnea syndrome, 294
Polymyalgia rheumatica, 164168, 173174 systemic arterial hypotension, 288292
clinical features, 173174 Type A behavior, 294
diagnosis, 173174 vasospastic disorders, 293
epidemiology, 164166
gender distribution, 164 T
genetic predisposition, 167 Temporal arteritis, 163264. See also Giant cell
pathophysiology, 167168 arteritis (GCA)
seasonal variation, 164
Posterior ciliary arteries (PCAs), 4259 V
terminology, 4344 Vascular endothelial growth factor (VEGF), 406407
invivo segmental supply, 4555 Visual loss patients
invivo supply, 4445 follow-up evaluation, 155
watershed zones, 5559 initial evaluation of, 145155
locations, 5559 detailed history, 145147
significance, 55 detailed ophthalmic evaluation, 147154
Posterior ischemic optic neuropathy (PION), 2, 425, 427434, hematological evaluation, 154
437443, 445452 systemic evaluation, 154155
classification, 425 temporal artery biopsy, 154, 155
clinical features (see Clinical features, PION) visual function evaluation, 148152

Ischemic Optic Neuropathies - S. Hayreh

Încărcat de

Informații document

Drepturi de autor

Formate disponibile

Partajați acest document

Partajați sau inserați document

Opțiuni de partajare

Vi se pare util acest document?

Este necorespunzător acest conținut?

Drepturi de autor:

Formate disponibile

Ischemic Optic Neuropathies - S. Hayreh

Încărcat de

Drepturi de autor:

Formate disponibile

Sohan Singh Hayreh

ISBN 978-3-642-11849-4 e-ISBN 978-3-642-11852-4

Library of Congress Control Number: 2011928229

Springer-Verlag Berlin Heidelberg 2011

Cover design: eStudioCalamar, Figueres/Berlin

Printed on acid-free paper

Springer is part of Springer Science+Business Media (www.springer.com)

Justus von Liebig (1846)

Iowa City, Iowa, USA  Sohan Singh Hayreh

1 Terminology and Types of Ischemic Optic Neuropathy . . . . . . . . . . . 1

Part I Basic Sciences

2 Structure of the Optic Nerve . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7

3 Blood Supply of the Optic Nerve . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35

 4 Blood-Optic Nerve Barrier . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 79

5 Factors Influencing the Optic Nerve Head Blood Flow . . . . . . . . . . . . . 85

6 Measurement of the Optic Nerve Head Blood Flow . . . . . . . . . . . . . . . 111

Part II Clinical Sciences

7 Optic Disc . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 131

8 Clinical Evaluation of Patients with Ischemic Optic Neuropathies . . . 145

9 General Characteristics of Giant Cell Arteritis . . . . . . . . . . . . . . . . . . . 163

10 Clinical Features of Giant Cell Arteritis . . . . . . . . . . . . . . . . . . . . . . . . . 173

11 Diagnostic Criteria for Giant Cell Arteritis . . . . . . . . . . . . . . . . . . . . . . 179

12 Ophthalmic Manifestations of Giant Cell Arteritis . . . . . . . . . . . . . . . . 199

13 Management of Giant Cell Arteritis to Prevent Visual Loss . . . . . . . . 227

14 Pathogenesis of Classical Non-arteritic Anterior Ischemic Optic

15 Pathogenesis of Some Controversial Non-arteritic

16 Clinical Features of Non-arteritic Anterior

17 Management of Non-arteritic Anterior Ischemic

18 Animal Model of Non-arteritic Anterior Ischemic

19 Classification and Incidence of Posterior Ischemic

20 Pathogenesis of Posterior Ischemic Optic Neuropathy . . . . . . . . . . . . . 427

21 Clinical Features of Posterior Ischemic Optic Neuropathy . . . . . . . . . 437

22 Management of Posterior Ischemic Optic Neuropathy . . . . . . . . . . . . . 445

Terminology Infarction of the optic nerve [27]

S.S. Hayreh, Ischemic Optic Neuropathies, 1

 nterior Ischemic Optic

S.S. Hayreh, Ischemic Optic Neuropathies, 7

may determine the shape of the cup. For example,

Fig.2.2 Longitudinal section

Fig.2.3 Longitudinal section

Fig.2.5 Bundles of the

The prelaminar region consists of optic nerve fibers

Fig.2.7 Glial elements

Cilio-retinal artery Fig.2.9 Fundus photograph of a human eye showing obliter-

Fig.2.11 Glial portion of

Fig.2.13 Prelaminar region

Fig.2.14 Glial membrane

Fig.2.17 Transverse section

Fig.2.18 Edema of the optic

Fig.2.19 High-power view

Fig.2.21 Optic nerve head

Fig.2.26 Transverse section of the lamina

Fig.2.27 Transverse section

Fig.2.30 Longitudinal section of the optic

Fig. 2.33 Transverse section of a human retrolaminar optic

Fig.2.36 A high-power view of a part of Fig. 2.35, showing

Fig.2.35 Longitudinal section of the retrolaminar part of the

Fig.2.37 Longitudinal connective-tissue fibers near the central

Fig.2.38 Transverse section of the optic nerve, showing fibrous

Fig.2.39 Fibrous astrocytes

stretching and distension of the sheath due to raised

Fig.2.41 A group of oligodendrocytes

canal. The ophthalmic artery is always attached to the

Iowa City, Iowa, USA Sohan Singh Hayreh

4 Blood-Optic Nerve Barrier . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 79

Terminology Infarction of the optic nerve [27]

S.S. Hayreh, Ischemic Optic Neuropathies, 1

nterior Ischemic Optic

S.S. Hayreh, Ischemic Optic Neuropathies, 7

Optic Nerve Fibers

opographic Localization of the Nerve

S.S. Hayreh, Ischemic Optic Neuropathies, 35

Terminology of PCAs and its Implications

In Vivo Supply by the PCAs

L ocations of Watershed Zones Between

ajor Areas of Controversy in Blood

enous Drainage of the Optic

Anterior Segment supplied by multiple small collateral arteries usually

lood Supply of the Intracanalicular

lood Supply of the Intracranial Part