WIDEFIELD FLUORESCEIN

ANGIOGRAPHY IN PATIENTS WITHOUT

PERIPHERAL DISEASE
A Study of Normal Peripheral Findings
ANKOOR R. SHAH, MD,*† ASHKAN M. ABBEY, MD,*† YOSHIHIRO YONEKAWA, MD,*†
Downloaded from http://journals.lww.com/retinajournal by BhDMf5ePHKav1zEoum1tQfN4a+kJLhEZgbsIHo4XMi0hCywCX1AWnYQp/IlQrHD3yRlXg5VZA8vVuITsjh2taYhBJ2pPUCJxI5bqFxzXwcA= on 09/08/2018

SARA KHANDAN, MD,* JEREMY D. WOLFE, MD,*† MICHAEL T. TRESE, MD,*†

GEORGE A. WILLIAMS, MD,*† ANTONIO CAPONE, JR., MD*†

Purpose: Widefield photography and angiography provide access up to 200-degrees of

the retinal periphery. The range of normal peripheral findings has not been characterized,
yet is relevant to studies addressing putative peripheral retinal vascular pathology.
Methods: This study was an observational retrospective cohort study. Adult patients
with epiretinal membrane or choroidal nevi who underwent imaging with Optos 200 MA/
200Tx were included. Dye transit times, peripheral arteriovenous shunting, presence of
vessels crossing the horizontal raphe, right angle vessels, terminal networks, absence of
capillary detail, ground glass hyperfluorescence, peripheral drusen, and microaneurysms
were evaluated.
Results: Fifty-eight eyes of 31 patients met inclusion criteria. Mean peripheral arterial
filling time was 8.65 ± 2.54 seconds (range 3–15 seconds). One or more peripheral anom-
alies were noted in all patients (P , 0.01). The prevalences of findings were: arteriovenous
shunting (0.00%), vessels crossing the horizontal raphe (44.83%), right angle vessels
(70.69%), terminal networks (77.59%), absence of capillary detail (98.28%), ground glass
hyperfluorescence (87.93%), drusen (34.48%), and microaneurysms (41.38%).
Conclusion: There was a high prevalence of peripheral vascular anatomic variations in
eyes expected to have normal peripheral retinal vasculature. These findings may provide
a reference for future studies addressing putative pathologic peripheral angiographic
findings.
RETINA 36:1087–1092, 2016

W idefield (WF) fundus photography and angiog-

raphy continue to increase in popularity. Periph-
eral pathology studied thus far spans multiple
ophthalmic subspecialties including pediatric ophthal-
mology, uveitis, ocular oncology, and retina. Specific
diseases that have been studied with WF imaging include
retinopathy of prematurity,1 familial exudative vitreore-
tinopathy,2 Coats disease,1 peripheral retinal vasculitis,3
From the *Associated Retinal Consultants, Royal Oak, Michigan;
and †Department of Ophthalmology, Oakland University William
Beaumont School of Medicine, Beaumont Eye Institute, Royal
Oak, Michigan.
None of the authors have any financial/conflicting interests to
disclose.
A. R. Shah and A. M. Abbey contributed equally to the devel-
opment of the article and are listed as co-first authors.
Reprint requests: Antonio Capone Jr., MD, 3535 W 13 Mile
Road, Suite 344, Royal Oak, MI 48073; e-mail: acaponejr@arcpc.net
age-related macular degeneration,4 central serous cho-
rioretinopathy,5 retinal pigment epithelium adenomas,6
diabetic retinopathy,7,8 and pathologic myopia.9
Any critical discussion of pathologic WF retinal
vascular findings must include a comparison with
patients who have normal peripheral retinal anatomy
as a reference standard. To our knowledge, published
data speaking to WF fundus angiography findings in
normal patients—that is, patients lacking any identifi-
able retinal pathology—are lacking. To allow for anal-
ysis of normal subjects without subjecting otherwise
healthy patients to the risks of angiography,10 we re-
viewed peripheral fundus findings in diseases that are
believed to have no peripheral pathology, namely epi-
retinal membranes (ERM) and choroidal nevi (exclud-
ing the nevus itself). We believe the findings detailed
below will serve as a reference base for future studies

1087
1088 RETINA, THE JOURNAL OF RETINAL AND VITREOUS DISEASES  2016  VOLUME 36  NUMBER 6

speaking to pathologic angiographic features of the Table 2. Vascular Architecture

peripheral retina. Mean ± SD Unit
Completion of peripheral arterial 8.65 ± 2.54 Seconds
Materials and Methods filling

Data Acquisition Category Number Percentage

Arteriovenous shunting Present 0 0.00
This study was approved by the Western Institu- Absent 58 100.00
tional Review Board. We performed a retrospective Vessels crossing the Present 26 44.83
chart review from the billing records of all patients horizontal raphe Absent 32 55.17
who received Optos (Optos 200 MA/200Tx, Optos Right angle vessels Present 41 70.69
PLC, Dunfermline, Scotland, United Kingdom) WF Absent 17 29.31
Terminal networks Present 45 77.59
fluorescein angiography for ERM (without retinal Absent 13 22.41
holes or tears) or choroidal nevus (not involving the
far retinal periphery) at Associated Retinal Consul-
tants, Royal Oak, MI, through June 2014. Inclusion was a discrepancy between the image graders, the
criteria required aforementioned diagnoses, age .18 images were reviewed again with both evaluators
years old, and imaging with fluorescein angiogram with present to render a consensus determination. The
the Optos WF camera. Exclusion criteria included pa- prevalence data were calculated by determining the
tients with concomitant diseases with a recognized percentage of our sample population with a given
potential for peripheral retinal pathology specifically, peripheral finding as the numerator, with the total study
diabetic retinopathy, sickle cell retinopathy, uveitis, ret- population as the denominator. Categorical variables
inal vein occlusions, premature birth history, and path- were analyzed using the Pearson chi-square test for
ologic myopia. As such, all patients known to have independence or Fisher’s exact test, and the binomial
a systemic diagnosis of diabetes mellitus or sickle cell test with the null hypothesis set at 50% was used to
disease with or without any known retinopathy were examine whether proportions occurred significantly
excluded. We documented demographic data including more or less than half the time. Statistical tests were
age, sex, diagnosis, and comorbidities. A patient was 2-tailed and significance was defined as P , 0.05. Sta-
considered “hypertensive” for subgroup analysis if they tistical analyses were performed using Stata version 9.0
were taking at least one hypertensive medication or if (StataCorp, LP, College Station, TX).
they identified themselves as hypertensive in their med-
ical history. We also recorded dye transit times, qualities
of vascular architecture (vasculature crossing the hori- Results
zontal raphe, formation of right angle vessels, presence
of terminal networks), and other notable peripheral find- A total of 58 eyes of 31 patients met inclusion
ings (absence of capillary detail, peripheral diffuse and/ criteria. Four patients were imaged unilaterally. The
or focal hyperfluorescence, presence of microaneur- mean patient age was 66.01 ± 12.51 years (range: 42–
ysms, and presence of peripheral drusen). 92 years). The proportion of eyes sorted by sex,
Two readers (ARS, AMA) independently evaluated
all Optos images to determine whether each of the
above findings was present for a given patient. If there

Table 1. Baseline Demographics

Mean ± SD Unit
Age 66.01 ± 12.51 Years
Category Eyes Percentage
Sex Male 23 39.66
Female 35 60.34
Diagnosis ERM 31 53.45
Choroidal nevus 27 46.55
HTN Yes 29 50.00
No 29 50.00 Fig. 1. Ultra-WF fluorescein angiogram of a right eye with the inferior
arcade (blue arrows) clearly crossing the horizontal meridian (orange
HTN, hypertension. dashed line) as it progresses into the periphery.
 WIDEFIELD ANGIOGRAPHY OF NORMAL RETINA  SHAH ET AL 1089

Fig. 2. Right angle vessels on ultra-WF fluorescein angiography. A. Terminal networks and microaneurysms are often seen adjacent to these vessels. B.
Right angle vessels (blue arrows) turn away from areas of far peripheral capillary nonperfusion. C. A more extreme example of a right angle vessel (blue
arrows) that turns back in the direction from which it originated. Orange arrows highlight peripheral drusen.

diagnosis, and known hypertensive status are listed in
Table 1.
The mean peripheral arterial filling time (from the
start of arterial phase to completion of peripheral
arterial filling) was 8.65 ± 2.54 seconds (range: 3–15
seconds). Arteriovenous shunting was not seen in the
periphery of any study eye (Table 2). Vessels cross-
ing the horizontal raphe (Figure 1) were present in
44.83% of eyes. Peripheral vascular changes, such as
right angle vessels (Figure 2) and terminal networks
(Figure 3) were noted in 70.69% and 77.59% of eyes
respectively, and were more likely to be present
than not (P , 0.01) (Table 2).
Additional peripheral findings included areas absent
of capillary detail (Figure 4) in 98.28%, ground glass
hyperfluorescence (Figure 5) in 87.93%, peripheral
drusen (Figure 6A) in 34.48%, and microaneurysms
(Figure 6B) in 41.38% (Table 3). Surprisingly, one or
more of these changes were noted in most study eyes
(P , 0.01).
Subgroup analysis was performed to determine the
difference in the proportion of peripheral findings across
different groups. This was first performed based on
ophthalmic diagnosis (Table 4). We further divided the
groups based on the presence or absence of systemic
hypertension and age (.65 or #65) (Table 4). In the
hypertensive group, there was a smaller proportion that
had right angle vessels (P , 0.01) and terminal networks
(P , 0.01). For those .65 years, there were a smaller
proportion of patients with right angle vessels (P ,
0.05) and ground glass hyperfluorescence (P , 0.01).
Discussion
As experience with WF grows, so does awareness of
peripheral findings in a broad range of retinal

Fig. 3. Terminal networks on ultra-WF fluorescein angiography. A. These networks frequently have associated microaneurysms (blue arrow). B. They are often
seen adjacent to an area of peripheral capillary nonperfusion. C. The magnified, inverted image highlights the branched pattern of capillaries within the networks.
1090 RETINA, THE JOURNAL OF RETINAL AND VITREOUS DISEASES
Fig. 4. Capillary nonperfusion
on ultra-WF fluorescein angi-
ography delineated by double-
sided arrows (A and B). Note
the networks and micro-
aneurysms adjacent to the area
of nonperfusion (A).
pathologies. Paradoxically, there is little descriptive
information available regarding peripheral features in
normal eyes. Orlin et al11 described a set of 33 control
eyes in their study of white without pressure, but many
of those controls had conditions (uveitis, vein occlu-
sions, and myopia) with known associated peripheral
changes. Kaneko et al9 evaluated eyes with pathologic
myopia as compared with 42 emmetropic eyes from
a population of patients with age-related macular
degeneration or central serous chorioretinopathy.
However recent studies have suggested that there are
peripheral findings in these disease processes that may
limit their usefulness as normals.4,5 Nonetheless, some
of their findings relating to microaneurysms, periph-
eral avascular areas, and retinal vascular changes do
mirror our findings.
Our baseline findings of the peripheral vascular
architecture challenge some common assumptions in
vascular anatomy. For instance, arterial phase filling is
classically taught as lasting approximately 2 seconds;
however, when the peripheral arterial circulation is
Fig. 5. Ground glass hyper-
fluorescence on ultra-WF fluo-
rescein angiography (A and B).
 2016  VOLUME 36  NUMBER 6
included our study suggests the normal range using 2
SDs is approximately between 4 seconds and 14
seconds. Moreover, the conventional notion is that
retinal vessels typically respect the horizontal raphe.
Although this is generally the case for the postequa-
torial retinal vasculature, approximately 45% of eyes
in the current series had vessels crossing the hori-
zontal raphe peripherally.
Several of the other vascular architectural findings
noted in this study are unique to the normal retinal
periphery. A sharp, right angle vascular course was
noted commonly. Terminal networks are also common
in the far periphery and are best appreciated approx-
imately 60 seconds into the angiogram. We suspect
these areas mark a watershed zone in circulation; thus,
they may reflect an area of overall reduced perfusion.
The prevalence of missing capillary details and
presence of ground glass hyperfluorescence were high
in study eyes. This pattern was noted for its prominent
appearance in the periphery, yet its significance in
normal patients and pathologic conditions is not well
 WIDEFIELD ANGIOGRAPHY OF NORMAL RETINA  SHAH ET AL 1091

Fig. 6. Drusen and micro-

aneurysms on ultra-WF fluo-
rescein angiography. A. Drusen
(blue arrowheads) tend to have
brighter fluorescence in the far
periphery and are independent
of peripheral vasculature. B.
Microaneurysms (blue arrow-
heads) tend to stem from
vessels.

established. This pattern of peripheral change is hypertensive population we examined was partially,
similar to the “Pattern two” as described by Orlin if not completely, controlled. Future large-scale stud-
et al11 when evaluating patients with white without ies are warranted to determine the clinical significance
pressure. We entertained the possibility this might of far peripheral microaneurysms and drusen. Differ-
have been an artifactual finding, perhaps a consequence entiating the two findings may be challenging. The
of a shadowing effect of the imaging system. Yet, vascular association of the microaneurysms aside, dru-
closer evaluation of the vessels passing through the sen are distinguished by the fact that they are both
areas of absent capillary detail reveals no commensu- independent of the vasculature and generally demon-
rate dimming, as one would expect from a shadowing strate brighter fluorescence when in the far periphery.
artifact. It remains unclear whether these areas lack There are several limitations of this study. First, it is
capillary detail because of capillary 1) nonformation a retrospective study with the inherent biases known to
or 2) a secondary loss of capillaries. The latter case such reviews. Second, we used a surrogate for normal
would more classically be defined as capillary nonper- eyes when evaluating the periphery. However, absent
fusion. We hypothesize that the ground glass hyper- exposing asymptomatic patients with normal emme-
fluorescence may reflect areas of peripheral retinal tropic eyes to the risks of fluorescein angiography
pigment epithelium thinning, resulting in greater trans- there is likely no better alternative. Third, one variable
mission of subjacent choroidal hyperfluorescence.12 we could not account for because of unavailable A-
Microaneurysms and peripheral drusen were also scan measurements was degree of myopia in pseudo-
noted commonly in our patient cohort. The hypothesis phakic eyes. However, as there is no known evidence
that these findings would increase with age or with of increased myopia in patients with ERM and
hypertensive status was not confirmed in subgroup choroidal nevi, one would expect a relatively normal
analysis. One possibility is that hypertensive patients distribution of refractive errors in our study cohort.
were identified based on medical history or use of Fourth, ERM can often have secondary changes on
a hypertensive medication—thus ensuring that the central retinal vasculature because of tractional
forces. These tractional forces may alter the tortu-
osity of peripheral retinal vasculature causing right
Table 3. Peripheral Findings angle vessels or altering the number of vessels
Category Number Percentage crossing the horizontal raphe as reported; but this
does not seem to be the case as the frequency of
Capillary nonperfusion Present 57 98.28
Absent 1 1.72 these findings by diagnoses (i.e., ERM vs. Nevus)
Ground glass Present 51 87.93 seems to be similar. Moreover, it is less likely that
hyperfluorescence Absent 7 12.07 tractional changes alter the development of the other
Microaneurysms Present 24 41.38 factors examined (i.e., terminal networks, capillary
Absent 34 58.62 nonperfusion, ground glass hyperfluorescence, mi-
Peripheral drusen Present 20 34.48
Absent 38 65.52 croaneurysms, and peripheral drusen). Future studies
would likely benefit from including measurements of
1092 RETINA, THE JOURNAL OF RETINAL AND VITREOUS DISEASES  2016  VOLUME 36  NUMBER 6

Table 4. Vascular and Peripheral Findings by Subgroup

HTN Age Diagnosis
Subgroup Present Absent .65 #65 ERM Nevus
Number of eyes 29 29 32 26 33 25
Crossing the horizontal raphe, % 51.72 37.93 43.75 46.15 39.39 52.00
Right angle vessels, % 51.72 89.66 59.38 84.62 72.73 68.00
Terminal networks, % 58.62 96.55 68.75 88.46 81.82 72.00
Capillary nonperfusion, % 96.55 100.00 96.88 100.00 96.97 100.00
Ground glass hyperfluorescence, % 79.31 96.55 65.63 100.00 84.85 92.00
Microaneurysms, % 27.59 55.17 37.50 42.31 36.36 48.00
Peripheral drusen, % 31.03 37.93 31.25 30.77 36.36 32.00

refractive status and axial length. Last, our study
population was not large enough to allow for broad
statements regarding population-based findings.
In summary, we describe peripheral fundus features
and angiographic findings in eyes of patients with
choroidal nevi and ERM, who would be presumed to
have an otherwise normal retinal periphery. Unique
aspects of the peripheral retinal vasculature were noted
with moderate prevalence. Additionally, peripheral
areas with absent capillary details are far more
common than one might initially suspect. These
findings may provide a reference point for studies
addressing pathologies with peripheral retinal vascular
findings.
Key words: fluorescein angiography, peripheral ret-
ina, retinal imaging, widefield imaging.
Acknowledgments
A. Capone had full access to all of the data in the
study and takes responsibility for the integrity of the
data and the accuracy of the data analysis.
References
1. Kang KB, Wessel MM, Tong J, et al. Ultra-widefield imaging
for the management of pediatric retinal diseases. J Pediatr
Ophthalmol Strabismus 2013;50:282–288.
2. Kashani AH, Learned D, Nudleman E, et al. High prevalence
of peripheral retinal vascular anomalies in family members of
patients with familial exudative vitreoretinopathy. Ophthalmol-
ogy 2014;121:262–268.
3. Hong BK, Nazari Khanamiri H, Rao NA. Role of ultra-
widefield fluorescein angiography in the management of uve-
itis. Can J Ophthalmol 2013;48:489–493.
4. Tan CS, Heussen F, Sadda SR. Peripheral autofluorescence
and clinical findings in neovascular and non-neovascular age-
related macular degeneration. Ophthalmology 2013;120:
1271–1277.
5. Pang CE, Shah VP, Sarraf D, Freund KB. Ultra-widefield
imaging with autofluorescence and indocyanine green angiog-
raphy in central serous chorioretinopathy. Am J Ophthalmol
2014;158:362–371 e2.
6. Shah SP, Jain A, Coffee RE, McCannel TA. Optos Panoramic
200MA ultrawide-field imaging of peripheral RPE adenoma.
Semin Ophthalmol 2009;24:37–39.
7. Kernt M, Pinter F, Hadi I, et al. Diabetic retinopathy: compar-
ison of the diagnostic features of ultra-widefield scanning laser
ophthalmoscopy Optomap with ETDRS 7-field fundus photog-
raphy. Ophthalmologe 2011;108:117–123.
8. Patel RD, Messner LV, Teitelbaum B, et al. Characterization of
ischemic index using ultra-widefield fluorescein angiography
in patients with focal and diffuse recalcitrant diabetic macular
edema. Am J Ophthalmol 2013;155:1038–1044.
9. Kaneko Y, Moriyama M, Hirahara S, et al. Areas of nonper-
fusion in peripheral retina of eyes with pathologic myopia
detected by ultra-widefield fluorescein angiography. Invest
Ophthalmol Vis Sci 2014;55:1432–1439.
10. Yannuzzi LA, Rohrer KT, Tindel LJ, et al. Fluorescein angiog-
raphy complication survey. Ophthalmology 1986;93:611–617.
11. Orlin A, Fatoo A, Ehrlich J, et al. Ultra-widefield fluorescein
angiography of white without pressure. Clin Ophthalmol 2013;
7:959–964.
12. Panda-Jonas S, Jonas JB, Jakobczyk-Zmija M. Retinal pigment
epithelial cell count, distribution, and correlations in normal
human eyes. Am J Ophthalmol 1996;121:181–189.