G Model

NEUCHI-669; No. of Pages 10 ARTICLE IN PRESS

Neurochirurgie xxx (2014) xxxxxx

Disponible en ligne sur

ScienceDirect
www.sciencedirect.com

Rapport : Douleurs lombaires postopratoires

Pathophysiological characterisation of back pain generators in failed

back surgery syndrome (part B)
Identication physiopathologique des gnrateurs potentiels de la composante
douloureuse lombaire dans les lombo-radiculalgies postopratoires (partie B)
P. Rigoard a,b,c, , S. Blond d , R. David a,b , P. Mertens e,f
a
Department of Neurosurgery, Poitiers University Hospital, 86021 Poitiers cedex, France
b
N3Lab: Neuromodulation & Neural Networks, Poitiers University Hospital, 86021 Poitiers cedex, France
c
Inserm CIC 802, Poitiers, 86021 Poitiers cedex, France
d
Department of Neurosurgery, Lille University Hospital, 59037 Lille cedex, France
e
Department of Neurosurgery, Lyon University hospital, 69677 Lyon cedex, France
f
Laboratory of Anatomy, Faculty of Medicine, 69677 Lyon cedex, France

a r t i c l e i n f o a b s t r a c t

Article history: Introduction. Low back surgery, including as many type of spine procedures as the multitude of failed
Received 2 July 2014 back surgery syndrome (FBSS) etiologies, is not always the answer for patients with chronic low back pain.
Accepted 8 October 2014 Paradoxically, although a patient is considered to present FBSS because he has already undergone spinal
Available online xxx
surgery, any new symptom in the back or deterioration of back pain must not be immediately attributed
to FBSS, but could be related to another cause independently of the initial mechanical problem. The aim
Keywords: of this paper is to extensively review the potential back pain generators in FBSS patients and to discuss
Low back pain
their respective roles and interactions in back pain pathophysiology.
Failed back surgery syndrome
Back pain generators
Methods. Literature searches included an exhaustive review of 643 references and 74 book chapters
Pathophysiology updated by searching the major electronic databases from 1930 to August 2013.
Aetiologic treatment Results. Nociceptive bres innervating any of the back anatomical structures can all play a part in the
Multidisciplinary management pathogenesis of the low back pain component in FBSS. The main spinal pain generators are not only
myofascial syndrome or muscle spasm but also the facets, the disc complex or a sagittal imbalance and
should therefore be carefully reviewed. Only after these steps and appropriate imaging, would it be
justied to irremediably diagnose the patient with a refractory chronic condition, requiring no further
spine surgery and to propose palliative pain treatment options.
Conclusion. Clinical investigations of the low back pain component in FBSS patients should be based
on meticulous dissection of all potential triggers that could be a source of the nociceptive pain charac-
teristics and possibly amenable to further aetiological treatment. Clinicians should therefore rene pain
management strategies to ensure that the chronic nature of the pain becomes the guiding principle for
multidisciplinary assessement.
2014 Elsevier Masson SAS. All rights reserved.

r s u m

Mots cls : Introduction. La chirurgie lombaire ne constitue pas toujours la solution univoque et durable pour les
Lombalgies patients souffrant de lombalgies chroniques rfractaires. Paradoxalement, ds lors quun patient a subi
Lombo-radiculalgies postopratoires une chirurgie rachidienne, toute douleur lombaire est considre comme partie intgrante dun syndrome
Gnrateurs de lombalgies squellaire tiquet lombo-radiculalgies postopratoires (LRPO), correspondant au FBSS des Anglo-
Physiopathologie
Saxons, alors que celle-ci peut tre lie une autre cause indpendante du problme mcanique initial.
Traitement tiologique
Le but de cet article est de dcrire et danalyser les gnrateurs potentiels de lombalgies chez les patients
Approche muti-disciplinaire
souffrant de LRPO et de discuter leurs rles et interactions dans la physiopathologie de ces douleurs
postopratoires.

Corresponding author. Unit rachis et neurostimulation, service de neurochirurgie, 2, rue de la Miltrie, 86021 Poitiers cedex, France.
E-mail address: philipperigoard@yahoo.fr (P. Rigoard).

http://dx.doi.org/10.1016/j.neuchi.2014.10.104
0028-3770/ 2014 Elsevier Masson SAS. All rights reserved.

Please cite this article in press as: Rigoard P, et al. Pathophysiological characterisation of back pain generators in failed back surgery
syndrome (part B). Neurochirurgie (2014), http://dx.doi.org/10.1016/j.neuchi.2014.10.104
G Model
NEUCHI-669; No. of Pages 10 ARTICLE IN PRESS
2 P. Rigoard et al. / Neurochirurgie xxx (2014) xxxxxx

Mthodes. Les recherches bibliographiques ont inclus une revue exhaustive de 643 rfrences et 74
chapitres de livres mis jour en consultant les principales bases de donnes lectroniques de 1930 aot
2013.
Rsultats. Une connaissance approfondie de lanatomie rachidienne parat essentielle pour la com-
prhension des mcanismes lsionnels lorigine de douleurs lombaires postopratoires. Chaque bre
nociceptive innervant les ligaments des disques intervertbraux, les facettes articulaires ou la muscula-
ture paravertbrale peut dtenir un rle cl dans la pathogense de la composante lombaire des LRPO. Les
principaux gnrateurs de douleurs rachidiennes par excs de nociception ne proviennent pas seulement
dun syndrome myofascial ou de spasmes musculaires, mais aussi dhyperpressions facettaires, dune
dgnrescence secondaire du complexe disco-vertbral ou dun dsquilibre sagittal. Ils doivent tre
examins scrupuleusement et de manire systmatise. Cest seulement aprs cette tape et lappui
dimageries appropries, quil parat possible dinscrire un patient dans le stade de la chronicit rfrac-
taire et de conclure labsence de toute nouvelle indication chirurgie rachidienne vise tiologique.
Cette dcision est irrversible et aboutira proposer des traitements palliatifs pour tenter de juguler
une partie des douleurs.
Conclusion. Une dissection physiopathologique minutieuse de tous les gnrateurs potentiels de la com-
posante lombaire rsiduelle aprs chirurgie rachidienne est ncessaire pour prtendre un traitement
tiologique, en cas de douleurs rfractaires composante nociceptive. La nature chronique et la com-
plexit smiologique de ces douleurs justient une interaction multidisciplinaire pour bncier dune
analyse multi-angulaire et dune prise en charge optimale.
2014 Elsevier Masson SAS. Tous droits rservs.

1. Introduction Back pain can be best understood from a biopsychosocial per-

Chronic back pain should not be considered to be a disease
related to failed back surgery syndrome (FBSS) patients, but rather
as a symptom and a part of this syndrome. The characteristic
inuence of position [1], reported by almost all of these patients,
reects the complex pathophysiology involving both nociceptive
and neuropathic factors. Back pain is anatomically dened as pain
situated below the shoulder blades and as far as the gluteal folds
[2,3] and can arise from various anatomical structures, including
ligaments, muscles and fascia, facet joints, spinal nerve roots, ver-
tebral periosteum, intervertebral disc and blood vessels [4]. The
most common form of back pain in this context is low back pain
(LBP). Low back pain is generally situated around the lumbosacral
vertebrae and iliac crests, between the 12th rib and the coccyx,
corresponding to the most common level of spine surgery. Pain
commonly radiates to the legs, or less often, to the waist or hips
[5].
In general and independently of any previous spine surgery,
somatic causes of back pain can be divided into three groups:
mechanical disorders of the spine (such as a disc prolapse or disc
protrusion, spinal stenosis, spinal deformity or spondylolisthesis),
non-mechanical diseases of the spine (such as tumours or infec-
tion), and organ diseases affecting the urogenital or gastrointestinal
tracts [5,6] (see Referred pain in part A [7]). Although the patient
is considered to present FBSS because he or she has already under-
gone spinal surgery, any new symptom in the back or deterioration
of back pain must not be immediately attributed to FBSS, but could
be related to another cause independently of the initial mechanical
problem. This justies a systematic approach based on appropriate
imaging and meticulous clinical dissection of all potential pain
generators, which must be carefully reviewed, one by one, before
drawing any conclusions.
Despite all these efforts, no precise organic cause can be identi-
ed for a signicant proportion of FBSS patients. However, it would
be wrong to assume that pain is therefore due to psychological
factors. Assessment of the pain threshold and a rigorous neurolog-
ical examination, starting with the DN4 questionnaire [8] can be
helpful to characterize co-existing neuropathic features and guide
the treatment plan [9]. Finally, all modern models of chronic back
pain incorporate both organic and psychosocial variables, or at least
acknowledge that these variables play a role in the chronic pain
condition [9,10].
spective, taking several matters into consideration: the organic
level, the individual psychological level and the social envi-
ronment level. This discussion could be considered to be a
digression from pathophysiology to psychopathology, but brain
and soul are one and the same in this complex process: the
nal pain integrator and sometimes the rst pain generator. . .
It seemed important to mention this aspect in the introduction
before focusing on identication of organic potential spinal pain
generators.
2. Back pain generators in FBSS patients
An extensive knowledge of spinal anatomy is essential to the
understanding of potential sources of pain [11], especially in the
presence of complex symptoms and signs, as in the case of FBSS.
There are several potential anatomical sources of pain in the low
back. Nociceptive bres innervate the ligaments of the interver-
tebral disc complex, facet joints, and paravertebral musculature.
They can all play a part in the pathogenesis of the low back pain
(LBP) component in FBSS (Fig. 1). Within these anatomical struc-
tures, nociceptive back pain can be the end-result of mechanical or
chemical irritation of nociceptive bres.
Accurate diagnosis of pain generators in the spine is vital to
ensure that no further surgery is required to resolve a persistent
biomechanical problem. It is important for interventional or phar-
macological treatments used to treat spinal generators in FBSS
patients to target the specic source of nociceptive pain.
Pain derived from the axial spine must be distinguished from
that of perisistent pain that is primarily muscular in origin known
as myofascial pain syndrome. Similarly, pain syndromes from
facet arthropathy are referred to as facet syndromes. The exist-
ence of facet syndrome holds some controversy. The distinction
between degeneration and pain derived from facet arthropathy
versus degenerative changes in the intervertebral disc may be
impossible. Nevertheless, some cases of LBP can be attributed to
the complex interaction of these structures at articulating sur-
faces. Finally, much has been written about the nature and cause of
discogenic LBP. Although the role of the disc space is traditionally
considered to be predominant, the role of the paired posterior facet
joints and their interactions with the anterior column should not
be underestimated [12].

Please cite this article in press as: Rigoard P, et al. Pathophysiological characterisation of back pain generators in failed back surgery
syndrome (part B). Neurochirurgie (2014), http://dx.doi.org/10.1016/j.neuchi.2014.10.104
G Model
NEUCHI-669; No. of Pages 10 ARTICLE IN PRESS
P. Rigoard et al. / Neurochirurgie xxx (2014) xxxxxx
Fig. 1. The complexity and interaction of potential back pain generators.
La complexit des interactions entre les gnrateurs potentiels de lombalgies.
Personal drawing/P. Rigoard.
Each of these pain syndromes is considered to respond to ther-
apy directed toward the presumed source of pain, as soon as the
pain generator has been precisely identied. The aim of this paper
is to extensively review the potential back pain generators in FBSS
patients and to discuss their respective roles and interactions in
back pain pathophysiology.
3. Methods
3.1. Literature searches
All references and book chapters were initially identied (Fig. 2)
from a systematic review of pathophysiology, anatomy and physi-
ology textbooks available in the following medical libraries: 1. Paris
Medical Library (Universit Descartes, Paris 5, rue de l cole de
Mdecine, 75006 Paris, France), 2. Paris Anatomy Library (Anatomy
Laboratory, Universit des Saints-Pres, Paris 6e , France), 3. Poitiers
Anatomy Library (Department of Morphology, Poitiers Medical Col-
lege, rue de la Miltrie, 86000 Poitiers, France), 4. UIC Library of
Health Sciences (University of Illinois at Chicago, 1912 Polk St.,
Chicago, tats-Unis), 5. Dorsch Neuroscience Library (Institute of
Neurology and Neuropsychiatry, 712 S Wood St., Chicago, tats-
Unis), performed by some of the authors (RD, PR). This list of
book chapters was updated by searching the following electronic
databases from 1930 to August 2013: MEDLINE (Ovid), MEDLINE
InProcess (Ovid), EMBASE (Ovid), Cochrane Library (Cochrane Cen-
tral Register of Controlled Trials (CENTRAL)), Health Technology
Assessment (HTA) databases. The search strategy was developed
in order to maximise sensitivity of article identication and was
not restricted by language, or any other limits. Registries of ongo-
ing controlled trials (metaRegister of Controlled Trials ISRCTN
database, metaRegister of Controlled Trials, UK Clinical Research
Network Portal, World Health Organisation International Clini-
cal Trials Research Portal and ClinicalTrials.gov) on low back pain
research were searched for information on current or recently com-
pleted studies. Citation lists in the chapters and papers consulted
and recent systematic reviews were checked for additional ref-
erences. Two reviewers (RD and PR) independently scanned all
Please cite this article in press as: Rigoard P, et al. Pathophysiological characterisation of back pain generators in failed back surgery
syndrome (part B). Neurochirurgie (2014), http://dx.doi.org/10.1016/j.neuchi.2014.10.104
3
titles and abstracts and identied potentially relevant articles for
retrieval. Full-text copies of papers were obtained, analysed and
summarized.
3.2. Keywords
The following keywords were used for these searches: patho-
physiology, pain, nociception, neuropathic pain, inammatory
pain, central sensitization, chronic pain, anatomy, neuroanatomy,
low back pain, failed back surgery syndrome, pain generators,
triggers, facet joints, sacroiliac joint, muscle pain, myogenic syn-
drome, spine, biomechanics, molecular mechanisms, discogenic
pain, spine instability, spinal fusion, sagittal imbalance, dorsal horn,
bulbospinal projections, ascending tracts, pain receptors, noci-
ceptors, CNS afferents, CNS efferents, neurobiology, neuromatrix,
gate control, pathological pain, physiological pain, pain threshold,
pain tolerance, back pain, mechanisms of action, neurostimulation,
spinal cord stimulation, peripheral nerve stimulation, motor cortex
stimulation, cordotomy, myelotomy, DREZ, pain pathways.
3.3. Literature analysis
Some well-written textbooks on pain management, pathophysi-
ology and anatomy were particularly useful as well as the following
references [11,1315]. The synopsis of these books on low back
pain largely contributed to this review and, whenever possible, the
reader is referred to the original publications for a more detailed
discussion.
3.3.1. Muscle pain after spine surgery
3.3.1.1. Anatomy and physiology [16,17]. The muscles that directly
control the movements of the vertebral column are divided into
categories according to their relationship to the spine: postverte-
bral and prevertebral muscles. Postvertebral muscles are further
subdivided into three groups: deep, intermediate, and supercial
(Fig. 3).
The deep muscles consist of short muscles that connect adjacent
spinous processes: posterior and transverse. Intermediate muscles
also connect transverse processes to posterior spinous processes.
Supercial muscles are collectively called the erector spinae and
are long muscles without segmental attachments. The prevertebral
muscles are the four abdominal muscles, three of which encircle the
abdominal region: external oblique, internal oblique, and transver-
sus abdominis. The fourth muscle is rectus abdominis, which lies
vertically and anteriorly in the midline. The spinal muscles produce
body movements by generating bending forces and torques, and
resist external forces. By reinfocing the ligamentar structures they
ensure stability. Depending on the activity necessary, muscles gen-
erate force isometrically as well as isotonically. Intrinsic systems
(the spindle and golgi apparatus) guide all muscle contractions and
have a sensory component. Neural mechanisms of muscle trophic-
ity are inherent, then modied by training and repetitive use.
Ladin explained that muscles balance the external movements
in response to external loads [18]. For the basis of physiological
activity, the other vertebral and paravertebral structures equili-
brate external forces and balance muscle forces. Infringement of
this underlying principle is the foundation for LBP and dysfunction.
3.3.1.2. Pathophysiology and clinical aspects. Normal neuromus-
culoskeletal function can be altered by various muscle injuries
involved in FBSS; notably muscle trauma due to spine surgery
and adaptive disorders related to new biomechanical constraints
superimposed on psychological factors which can inuence pain
perception (fatigue, anxiety, etc.) and fear of recurrence of injury
[17].
G Model
NEUCHI-669; No. of Pages 10 ARTICLE IN PRESS
4 P. Rigoard et al. / Neurochirurgie xxx (2014) xxxxxx

Fig. 2. Literature searches methodology.

Mthodologie de recherche documentaire.

Intricate muscle mechanisms are compounded by coupling
motions of the spine changed by any posterior instrumentation
or decompression, such as rotation and exion. The intricacies of
muscle activity are also compound by simultaneous relaxation of
antagonists while agonists contract at set speed and force. Faulty
neuromuscular activity may cause other tissues of the functional
unit (intervertebral disc, facets and ligaments) to be exposed to
further mechanical trauma leading to impairment, disability and
worsening of pain. The muscle trophicity can also be compromised
by repeated surgery, leading the vertebral column to potential and
progressive unstability [19]. A vicious circle can occur when addi-
tional loads cause displacements spine and induce new tensions.
When muscles and tendons are exposed to single or recurrent
episodes of biomechanical overloading, increased impulse activ-
ity of their nociceptors establishes the neurophysiological basis
of muscle pain. It is probable that an increase in chemical sub-
stances generated by overused, misused and tramatised muscles
is the foundation for this increased activity.
Muscle pain implies the origin of nociception from striated
muscle, but also from muscle fascia and tendinous insertions. Ter-
minating in laminae I, II, and IV in the dorsal horn, we can assume
that muscle afferent bres, including nociceptive bres converging
from musculoskeletal nociceptors, transmit the pain signals to the
brain via spinothalamic projections [20].
Finally, three different pain generators involving the muscu-
loskeletal system must be considered: muscle spasm, myogenic (or
myofascial) pain and peripheral fatigue.
3.3.1.2.1. Muscle spasm. Muscle spasm is a consequence
of inammation of tissues in the vertebral column. Muscle
spasm is rarely discussed, dened, or documented, and its neuro-
physiological basis has not been elucidated. In a review of a large
number of publications, the term is mentioned but never dened.
It can be assumed that isometric contraction of adjacent muslces
can be caused by inammation within the functional unit to prevent
further movements that would aggravate the pain. However, elec-
tromyographic verication of this protective spasm has never been
investigated. It is indicated that ischaemia can cause muscle spasm
to turn from a mechanical or an inammatory incident to a painful
condition. As the intramuscular pressure is insufciently elevated,
it is demonstrated in by various studies that spasm pain cannot
be attributed to increased muscle tension but rather contraction
under ischaemic indisposition [21].
3.3.1.2.2. Myogenic pain. Myogenic or myofascial pain syn-
drome is distinguished by muscles that are in a shortened or
contracted state. These muscles will have increased tone and stiff-
ness, and will contain trigger points (TrP). Identied on palpation of
the muscles, Trp are rm, tender, 3 to 6 mm nodules [22]. Palpation
of TrP provokes radiating, aching-type pain in localised reference
zones. Other injuries to tissues of the verteral functional unit can
produce myogenic pain (Fig. 4).
3.3.1.2.3. Peripheral fatigue. There is a chemical basis to
peripheral fatigue. Mononeuronal ring or muscle glycogen deple-
tion may trigger fatigue [2325]. Fatigue impairs the sequence
of neuromusculoskeletal function in the paravertebral muscles

Please cite this article in press as: Rigoard P, et al. Pathophysiological characterisation of back pain generators in failed back surgery
syndrome (part B). Neurochirurgie (2014), http://dx.doi.org/10.1016/j.neuchi.2014.10.104
G Model
NEUCHI-669; No. of Pages 10 ARTICLE IN PRESS
P. Rigoard et al. / Neurochirurgie xxx (2014) xxxxxx 5

Fig. 3. Anatomy of the paravertebral muscles.

Anatomie des muscles paravertbraux.
Graphic conception: K. Nivole, R. David and P. Rigoard.

making it a factor in the production of episodes of acute low
back pain, perodically decompensating FBSS. A change in muscle
recruitment also takes place in FBSS patients as fatigue indicates
a signicant decrease in motor performance. A paradoxical sig-
nicant increase in contraction of internal oblique and latissimus
dorsi muscles has also been documented [2628]. In conclusion,
a better dissection and understanding of the muscular aspects
is helpful to analyse the lumbosacral pain component in FBSS
patients.
3.3.2. Facet joint pain and spinal instability
Facet joints were rst identied by Goldthwait in 1911 [29]
where they were recognized as potential sources of back pain. The
term facet syndrome was rst used by Ghormley [30], to describe
a lumbosacral pain with or without sciatic pain, often occuring
abruptly after a twisting or rotary strain of the lumbosacral region.
Spine surgery induces major changes in biomechanical loads on
these structures, after spine decompression or stabilization [11].
This mechanical reorganization of loading redistribution occurs
particularly in adjacent spinal segments, above and around the
anterior or posterior instrumentation.
3.3.2.1. Anatomy and physiology [11,14]. The facet joints are syno-
vial joints found between the posterior elements of the adjacent
vertebrae that share compressive loads and other biomechanical
forces with the intervertebral disc [31] (Fig. 5). Translatory shear
motion, lateral exion and rotation are limited by angulation of
the facets [32,33]. Such limitation provides stability of the func-
tional unit along with the disc annual bres and ligaments [34].
Compressive forces are minimised by lubrication of the joint. The
mechanical structure of the cartilage makes this lubrication possi-
ble and is secreted into the joint by compressive forces of gravity
and muscle contraction [14]. Friction is minimised by lubrication
in coating the joint surfaces with its adhesive properties.
Innervation of the facet joints is suplied by the posterior pri-
mary division of the nerve roots containing afferent and efferent
nerves as well as sympathetic nerve bres (Fig. 6A). During ex-
ion activities, it has been assumed that the proprioceptive function

Please cite this article in press as: Rigoard P, et al. Pathophysiological characterisation of back pain generators in failed back surgery
syndrome (part B). Neurochirurgie (2014), http://dx.doi.org/10.1016/j.neuchi.2014.10.104
G Model
NEUCHI-669; No. of Pages 10 ARTICLE IN PRESS
6 P. Rigoard et al. / Neurochirurgie xxx (2014) xxxxxx

Fig. 4. Neural basis of myogenic pain. Noxious impulses are transmitted to the brain through spinothalamic tract at the spinal cord level. Nociceptive signals converge with
visceral afferents (from vessels) at the level of the dorsal horn and can induce ischemia via reex vasospasm.
Bases neurales des douleurs dorigine musculaire. Les inux nociceptifs sont vhiculs vers le cortex via le faisceau spino-thalamique au niveau mdullaire. La convergence des bres
sensitives somesthsiques et des affrences viscroceptives au niveau des cornes postrieures de la mlle peut expliquer certaines manifestations ischmiques musculaires, secondaires
un vasospasme rexe.
Graphic conception: K. Nivole, R. David and P. Rigoard, adapted from Cailliet, 2003 [14].

of facet joints plays an important role in spontaneous reduction
of myoelectric undertaking in the lumbar erector spinae muscles
[32,33].
3.3.2.2. Pathophysiology. There are a number of nerve end-organs
that function as proprioceptors and nociceptors in the presence
of mechanical injury inammation [35]. Proprioceptor end-organs
provide information with regards to movement and have been
detected within the tendons, deep back muscles, interspinous lig-
aments and facets. The nerve end-organs can also be found in the
posterior ligamentous structures. They are known to be a signi-
cant site of low back pain due to this innervation. It has been shown
that pain induced by a noxious injection into the joint is relieved
by subsequent injection of an analgesic agent [36].

Fig. 5. Anatomy of the spine.

Anatomie du rachis.
Graphic conception: K. Nivole, R. David and P. Rigoard.

Please cite this article in press as: Rigoard P, et al. Pathophysiological characterisation of back pain generators in failed back surgery
syndrome (part B). Neurochirurgie (2014), http://dx.doi.org/10.1016/j.neuchi.2014.10.104
G Model
NEUCHI-669; No. of Pages 10 ARTICLE IN PRESS
P. Rigoard et al. / Neurochirurgie xxx (2014) xxxxxx 7

Fig. 6. Facet innervation. A. The facet joint receives a superior innervation from the descending medial branch of the upper nerve root and an inferior innervation from the
ascending lateral branch of the lower nerve root. B. The inuence of facet inammation on muscle spasm. Facet joint inammation activates wide dynamic range (WDR)
neurons at the dorsal horn level and can lead to activation of anterior horn cells, which in turn, cause contraction of the extrafusal muscle bers and muscle spasm. C.
Degenerative facet disease and foraminal stenosis. Facet hypertrophy associated with disc collapse result in narrowing the formanen and in possible nerve root compression.
Innervation facettaire. A. La facette articulaire recoit une double innervation : suprieure par la branche mdiale descendante de la racine nerveuse sus-jacente et infrieure, par la
branche latrale ascendante de la racine nerveuse sous-jacente. B. Linammation articulaire postrieure est responsable dune activation de la population neuronale WDR , au sein
de la corne postrieure. Cette population stimule son tour les neurones de la corne antrieure de la mlle entranant une contraction des bres musculaires extrafusales, responsable
du spasme musculaire. C. Pathologie dgnrative des facettes et stnose foraminale. Une hypertrophie facettaire peut rduire la taille du foramen lorsquelle est combine une perte
de la hauteur discale, lie la dgnrescence arthrosique.
Graphic conception: K. Nivole, R. David and P. Rigoard. Adapted from Cailliet, 2003 [14].

Degenerative lumbar facets exposed to ageing and cumulative
biomechanical loading are where substance P (SP) has been iden-
tied in the nerve bres. Infusion of SP into facet joints speeds up
the degenerative process. Proteolytic and collagenolytic enzymes
that cause degradation of the cartilaginous matrix have been linked
to these inammatory mediators. The presence of algogenic neu-
ropeptides, such as SP and CGRP, and evidence of nociceptive
afferents in facets and periarticular tissues (Fig. 6B), thus suppor-
ting a role for these structures as spinal pain generators [37].
anteriorly from vertebral spurs and posteriorly from facet arthritic
changes (Fig. 6C). The facets may be the origin on reproduction
of the patients symptoms [38,39]. Injection of an analgesic agent
into the facet joint is considered to be diagnostic and sometimes
therapeutic [4044] but relief from an analgesic injection does not
guarantee that the facet is the cause, as the facet is innervated by
two nerve root levels and not just one specic nerve root branch
[41]. Continued degeneration of the facets results in more instabil-
ity of the functional unit, which may cause listhesis.

3.3.2.3. Degenerative facet disease and foraminal stenosis. There is 3.3.3. Discogenic and radicular pain after spine surgery
simultaneous narrowing of the posterior elements in the presence In addition to probable decompensation of an independent
of discogenic disease that narrows the anterior compartment of primary disease of the spine after one or more spinal operative
the functional unit, particularly with bony hypertrophy where the procedures, the notion of surgical failure may include one of the
facets undergo degenerative changes. This causes stenosis of the following diagnoses [45]:
foramen with possible entrapment of the nerve root at that level
(Fig. 6C). recurrent disc herniation;
peridural scarring;
3.3.2.4. Clinical aspects. It remains difcult and controversial to nerve root compressed by scarring;
indentify facets as the source of pain and impairment. Active and deformity of the dural sac;
passive hyperextension of the patients low back with simulta- herniation of adjacent disc;
neous lateral exion to both sides is the standard test for verifying spinal instability;
pain arising from the facets. This undertaking imitates narrowing facetectomy;
of the foramina by the facets causing nerve root entrapment. The pseudomeningocele;
foramina become narrowed as a result of degenerative changes: arachnoid cysts;

Please cite this article in press as: Rigoard P, et al. Pathophysiological characterisation of back pain generators in failed back surgery
syndrome (part B). Neurochirurgie (2014), http://dx.doi.org/10.1016/j.neuchi.2014.10.104
G Model
NEUCHI-669; No. of Pages 10 ARTICLE IN PRESS
8 P. Rigoard et al. / Neurochirurgie xxx (2014) xxxxxx

arachnoiditis;
foraminal restenosis.

Note that 7 of these 11 items are directly and closely related

to disc space anatomy, biology and biomechanics, and all 11 are
related to adjacent spinal nerve roots.

3.3.3.1. Radicular pain. Spinal radicular pain and its pathophysi-

ology is an ongoing subject of research. Nerve root compression
with dysfunction, ischemia, inammation and biomechanical inu-
ences are all suggested aetiologies [46]. Spinal roots are devoid
of a well developed endoneurial blood-nerve barrier, which may
make them more susceptible to symptomatic compression injury
and more vulnerable to endoneural oedema formation compared to
peripheral nerves [47]. Increased vascular permeability caused by
mechanical nerve root compression can induce endoneural oedema
[47]. In addition, capillary blood ow can be impeded by increased
endoneural uid pressure (caused by endoneural oedema) and
could cause intraneural brosis [48]. It has been shown that per-
ineural brosis renders nerve roots hyperaesthetic and increasingly
sensitive to compressive forces [47,49]. As opposed to the nor-
mal progressive sensory and then motor dysfunction often seen
with peripheral nerve compression, nerve root ischaemia or venous
stasis can also generate pathological biochemical changes that pro-
duce pain [47].

3.3.3.2. Discogenic pain. It is thought that axial back pain is often

brought about by abnormalities involving the vertebral bodies and
intervertebral discs. It is predominantly described as a deep, mid-
line, aching pain. Positional in nature, the pain is generally worse
when upright rather than supine [22]. It is thought that this type
of pain is generated by the visceral afferents that innervate the
intervertebral disc.
3.3.3.3. Innervation of the disc [11]. The intervertebral disc is inner-
vated via sinu-vertebral nerves (SVN), which were rst described
by Lushka [50] in the mid-1800s (Fig. 7A). According to Lushka, this
nerve derived from spinal nerves with a connection to the sym-
pathetic system (Fig. 7A and B). The SVN arises from two roots,
one from the ventral primary ramus and the other from ramus
communicans. The SVN enters the neural canal through the inter-
vertebral foramen (Fig. 7A). Inside the canal, the nerve ramies into
two branches. The major branch ascends rostrally along the lateral
border of the posterior longitudinal ligament. A second division
passes medially and dorsally to innervate the intervertebral disc at
the level of origin of the parent nerve. Bogduk et al. [51] demon-
strated by anatomical studies that SVN, through its two primary
branches, innervates two separate intervertebral levels. The infe-
rior branch innervates the superior aspect of the disc at the level of
entry, and the rostral branch provides innervation to the disc one
level above. The SVN provides the majority of innervation to the
spinal canal. The ligaments, dura mater, and blood vessels, includ-
ing the posterior longitudinal ligament and posterior aspect of the
annulus brosis, are innervated by this network of nerve endings
(Fig. 7B). It is noteworthy that the dura has been demonstrated to
receive nociceptive innervation from the nerve of Lushka, whereas
the nerve root itself has not been shown to receive any signicant
innervation [52].
3.3.3.4. Chemical factors. The nucleus pulposus of the interverte-
bral disc (Fig. 5) is composed of collagen, mucopolysaccharides,
and water. Painful discs have a lower pH than non-painful discs
in humans [31].
In some cases, inammatory factors may be responsible for pain
and epidural steroid injections can provide relief. In capsular and
joint nerve bres of the facets and the disc, neuropeptides (SP, VIP
Fig. 7. Intervertebral disc innervation. A. Collateral branches of the spinal nerves
at the dorsolumbar level represented on a cross-sectional view. The disc annulus is
mainly innervated by the sinu-vertebral nerve (SVN) and its branches to the pos-
terior longitudinal ligament. B Illustrative view of the disc and dura innervation by
the SVN, which is part of sympathetic nervous system
Innervation du disque intervertbral. A. Coupe transversale reprsentant les principales
branches collatrales des nerfs spinaux au niveau pri-vertbral. Lannulus discal est
principalement innerv par le SVN et ses branches jusquau ligament longitudinal
postrieur. B. Innervation discale vgtative. Vue illustrative de linnervation du disque
et de la dure-mre par le SVN, qui fait partie intgrante du systme nerveux sympathique
Graphic conception: K. Nivole, R. David and P. Rigoard. Adapted from Cailliet, 2003 [14].
and calcitonin gene-related peptide [CGRP]) are present [53] and
could be responsible for progressive deterioration of the motion
segment structures, especially the intervertebral disc. As in the
pathophysiology of facet joint degeneration, these factors may be
released in response to noxious biochemical forces and environ-
mental factors (i.e. biomechanical stress, microtrauma) and induce
the synthesis of inammatory agents (cytokines, prostaglandin
E2) and degradative enzymes (protease, collagenase). In the gray
matter of the dorsal spinal cord, initial nociceptive afferences are
probably modulated by SP [7]. There has been suggestion that phos-
pholipase A2 (PLA2) may play a dual role, predisposing to disc
degeneration and sensitising annular nerve bres [54]. In human
disc herniations, NItric oxide (NO) concentration is raised when
the hydrostatic pressure of the disc is increased by biomechanical
stressors [55]. In cells of the nucleus pulposus, proteoglycan syn-
thesis is inhibited by NO. This leads to reduction in water content,
disc degeneration and proteoglycan loss [55].
Lastly, proinammatory cytokines can also add to discogenic
pain by sensitising nociceptors and their effect on disc degeneration
by suppressing proteoglycan synthesis and increasing disc matrix
degradation. Neural injury in the CNS causes cytokines to be pro-
duced and may play a part in spinal neural hypersensitisation and
chronic neuropathic pain [56,57].

Please cite this article in press as: Rigoard P, et al. Pathophysiological characterisation of back pain generators in failed back surgery
syndrome (part B). Neurochirurgie (2014), http://dx.doi.org/10.1016/j.neuchi.2014.10.104
G Model
NEUCHI-669; No. of Pages 10 ARTICLE IN PRESS
P. Rigoard et al. / Neurochirurgie xxx (2014) xxxxxx
3.3.3.5. Biomechanics. Over time, the annulus brosis undergoes
changes. Molecular cascades, dehydration and annular tears are
further exacerbated by ssuring and disruption of annular lamellae
[58,59]. Radial lesions are macroscopic deformities extending radi-
ally from the nucleus to the periphery [60]. The inner two thirds of
the annulus and the nucleus pulposus are free of nerve endings. The
supporting structures of the intervertebral disc are innervated by
a concise network of nerve bres. A plexus of nerves from the SVN
and the gray rami of the lumbar sympathetic trunks are contained
in the lateral aspect of the vertebral column and the posterior longi-
tudinal ligament [61]. Nerve bres spread through the longitudinal
ligament deep into the outer third of the annulus brosis [62]. It
is assumed that when the outer third of the annulus is exposed
to painful stimulus, the intervertebral disc may take on the role
of a pain trigger. A painful response is produced with provocation
discography of a diseased disc space when elevated intradisc pres-
sure is communicated via the injured annular channels to the nerve
endings in the outer third of the annulus [49,63].
3.3.4. Atypical back pain generators
Many other congenital or acquired conditions of the lum-
bosacral spine may be direct the causes of mechanical low back
pain, or may even decompensate the low back pain component
after previous spine surgery. These conditions, listed below, will be
reviewed in more detail in another paper [64]:
spondylolisthesis and spinal instability;
spinal stenosis;
scoliosis and spinal deformities;
iliolumbar syndrome;
piriformis syndrome;
sacroiliac disease causing LBP;
sacralization of transverse process;
spinal infections;
tumours.
4. Conclusion
Low back surgery, including as many spine procedures as the
multitude of FBSS etiologies, is not always the answer for patients
with chronic low back pain [65]. However, claiming that persis-
tent LBP after surgery, or even before surgery (so-called virgin
chronic LBP), is simply due to muscle inammation or degenera-
tion remains controversial since chronic LBP can also be considered
as a neuropathic pain condition, for which the superimposition of
focal and specic nociceptive pain generators can decompensate a
fragile steady-state in this chronic condition. Taking these aspects
into account, clinical investigations of the low back pain compo-
nent in FBSS patients should be based on meticulous dissection of
all potential triggers that could be a source of the nociceptive pain
characteristics and possibly amenable to further aetiological treat-
ment. The main spinal pain generators are myofascial syndrome,
the facets and the disc complex and should therefore be carefully
reviewed. Only after these steps and appropriate imaging, would
it be justied to irremediably diagnose the patient with a refrac-
tory chronic condition and propose palliative pain treatment
options. Lastly, as with all neuropathic pain syndromes, it must
be remembered that chronic LBP represents a transformation from
nociceptive pain into neuropathic pain that must take the patients
underlying perception into account [66]. The overall neuropathic
pain syndrome should be addressed whilst combining interven-
tions targeted at the nociceptive-mediated biochemical pain in the
treatment of this condition. Clinicians should therefore rene pain
management strategies to ensure that the chronic nature of the pain
becomes the guiding principle for multidisciplinary assessement.
Please cite this article in press as: Rigoard P, et al. Pathophysiological characterisation of back pain generators in failed back surgery
syndrome (part B). Neurochirurgie (2014), http://dx.doi.org/10.1016/j.neuchi.2014.10.104
9
Disclosure of interest
Dr Rigoard is a consultant for Medtronic Inc. and received hon-
oraria for medical training from St Jude Medical, research grants
from Medtronic Inc & St Jude Medical.
All other authors reported no conict of interest for this paper.
Acknowledgements
The authors would like to thank Mr Lee Wesley for reviewing
this manuscript, Nancy Ladmirault for her technical help, the N3Lab
for its assistance and Poitiers University Hospital (Department of
Research Management, Mr Carles De Bideran, Ms Sarah Guyon) for
their support.
References
[1] Waddell G. The Back Pain Revolution. Churchill Livingstone; 1998 [pp. 9199,
192194, 211216].
[2] Barolat G, Massaro F, He J, Zeme S, Ketcik B. Mapping of sensory responses to
epidural stimulation of the intraspinal neural structures in man. J Neurosurg
1993;78(2):2339.
[3] Rubin DI. Epidemiology and risk factors for spine pain. Neurol Clin
2007;25(2):35371.
[4] Deyo RA, Rainville J, Kent DL. What can the history and physical examination
tell us about low back pain? JAMA 1992;268(6):7605.
[5] Deyo RA. Early diagnostic evaluation of low back pain. J Gen Intern Med
1986;1(5):32838.
[6] Bassam BA. Low back syndromes. The challenge of accurate diagnosis and man-
agement. Postgrad Med 1990;87(4):20915.
[7] Patrick Mertens, Serge Blond, Romain David, Philippe, Rigoard. Anatomy, phys-
iology and neurobiology of the nociception: a focus on Low Back Pain (Part A).
Neurochirurgie 2014 [Accepted for publication (Hors-srie)].
[8] Bouhassira D, Attal N, Alchaar H, Boureau F, Brochet B, Bruxelle J, et al. Com-
parison of pain syndromes associated with nervous or somatic lesions and
development of a new neuropathic pain diagnostic questionnaire (DN4). Pain
2005;114(12):2936.
[9] Gatchel RJ. Pespectives on pain: a historical overview. In: Psychosocial factors
in pain: clinical perspectives. New York [u.a.]: Guilford Press; 1999. p. 317.
[10] Turk D, Flor H. Chronic pain: a biobehavioral perpective. In: Psychosocial factors
in pain: clinical perspectives. New York [u.a.]: Guilford Press; 1999.
[11] Resnick DK, Haid RW. American Association of Neurological Surgeons. Surgical
management of low back pain. Rolling Meadows, Ill: American Association of
Neurological Surgeons; 2001.
[12] Highsmith JM, Rodts Jr G. Facet-Mediated versus Discogenic Pain: Differences
in Presentation and Treament. Chapter 8. In: Surgical management of low back
pain. DK Resnick, RW Haid Jr, JC Wang. Rolling Meadows Ill: American Associ-
ation of Neurological Surgeons; 2001.
[13] Cailliet R. Low back pain syndrome. Philadelphia: F.A. Davis; 1995.
[14] Cailliet R. Low Back Disorders: A Medial Enigma. Wolters Kluwer Health; 2003.
[15] Lautenbacher S, Fillingim RB. Pathophysiology of pain perception. New York:
Kluwer Academic/Plenum; 2004.
[16] Cailliet R. Musculoskeletal Pain. Chapter 4. In: Low back pain syndrome.
Philadelphia: F.A. Davis; 1995.
[17] Muscular Control of the Low Back. Chapter 5. In: Low Back Disorders. A Medial
Enigma R Cailliet. Wolters Kluwer Health; 2003.
[18] Ladin Z. Use of musculoskeletal models in the diagnosis and treatment of low
back pain. Chapter 24. In: Multiple Muscle Systems; Biomechanics and Move-
ment Organization. Springer Verlag. J.M. Winters and S.L-Y. Woo; 1990.
[19] Lucas D, Bresler B. Stability of ligamentous spine. In: Report No.: Biomechanics
Laboratory Report; 1961. p. 40.
[20] Nieuwenhuys R, Voogd J, Huijzen van C. The human central nervous system.
New York: Springer; 2008.
[21] Emre M, Mathies H. Muscle spasms pain. Parthenon: Park Ridge I.L.; 1988.
[22] Simons DG, Travell JG. Travell & Simons myofascial pain and dysfunction: the
trigger point manual. Williams & Wilkins: Baltimore; 1999.
[23] Edwards RH. Human muscle function and fatigue. Ciba Found Symp
1981;82:118.
[24] Bigland-Ritchie B, Cafarelli E, Vllestad NK. Fatigue of submaximal static con-
tractions. Acta Physiol Scand Suppl 1986;556:13748.
[25] Kukulka C. Human skeletal muscle fatigue. Contemporary perspectives in reha-
bilitation. Philadelphia: FA Davis. Currier DP, Nelson RM; 1992.
[26] Macdonald AJ. Abnormally tender muscle regions and associated painful move-
ments. Pain 1980;8(2):197205.
[27] Bigland-Ritchie B, Johansson R, Lippold OC, Woods JJ. Contractile speed and
EMG changes during fatigue of sustained maximal voluntary contractions. J
Neurophysiol 1983;50(1):31324.
[28] Sparto PJ, Parnianpour M, Marras WS, Granata KP, Reinsel TE, Simon S.
Neuromuscular trunk performance and spinal loading during a fatiguing iso-
metric trunk extension with varying torque requirements. J Spinal Disord
1997;10(2):14556.
G Model
NEUCHI-669; No. of Pages 10 ARTICLE IN PRESS
10 P. Rigoard et al. / Neurochirurgie xxx (2014) xxxxxx
[29] Goldthwait JE. The lumbo-sacral articulation: an explanation of many cases of
lumbago sciatica and paraplegia. Boston Med Surg J 1911;164:36572.
[30] Ghormley RK. Low back pain with special reference to the articular facets, with
presentation of an operative procedure. J Am Med Assoc 1933;101:17737.
[31] Fujiwara A, Tamai K, An HS, Kurihashi T, Lim TH, Yoshida H, et al. The relation-
ship between disc degeneration, facet joint osteoarthritis, and stability of the
degenerative lumbar spine. J Spinal Disord 2000;13(5):44450.
[32] Adams M, Hutton W. The mechanical function of the lumbar apophyseal joints.
Spine 1983;8:32730.
[33] Taylor J, Twomey LT. Structure and function of lumbar zygapophyseal (facet)
joints: a review. J Orthop Med 1992;14:71.
[34] Sato H, Kikuchi S. The natural history of radiographic instability of the lumbar
spine. Spine 1993;18(14):20759.
[35] Minaki Y, Yamashita T, Ishii S. An electrophysiological study on the mechanore-
ceptors in the lumbar spine and adjacent tissues. Neuro-Orthop 1996;164:365.
[36] Mooney V, Robertson J. The facet syndrome. Clin Orthop 1976;115:14956.
[37] Bidros D, Steinmetz M, Resnick DK. Pathophysiology and Treatment of Low
Back Pain. Chapter 2. In: Surgical management of low back pain. DK Resnick,
RW Haid Jr, JC Wang. Rolling Meadows Ill: American Association of Neurological
Surgeons; 2001.
[38] Maigne R, Le Corre F, Judet H. [Low back pain of dorse-lumbar origin: surgical
treatment of postrlor articular capsule excision (authors transl)]. Nouv Presse
Med 1978;7(7):5658.
[39] Laros G. Differential diagnosis of low back pain. In: Contemporary conservative
care for painful spinal disorders. Philadelphia: Lea & Febiger. Mayer TG, Mooney
V, Gatchel RJ;; 1991. p. 12230.
[40] Bogduk N, Long DM. The anatomy of the so-called articular nerves and their
relationship to facet denervation in the treatment of low-back pain. J Neurosurg
1979;51(2):1727.
[41] Bogduk N, Long DM. Percutaneous lumbar medial branch neurotomy: a modi-
cation of facet denervation. Spine 1980;5(2):193200.
[42] Carrera GF. Lumbar facet joint injection in low back pain and sciatica: descrip-
tion of technique. Radiology 1980;137(3):6614.
[43] Lynch MC, Taylor JF. Facet joint injection for low back pain. A clinical study. J
Bone Joint Surg Br 1986;68(1):13841.
[44] Marks RC, Houston T, Thulbourne T. Facet joint injection and facet nerve block:
a randomised comparison in 86 patients with chronic low back pain. Pain
1992;49(3):3258.
[45] Aldrete JA. Failed Back Surgery Syndrome. Chapter 96. In: Pain management
Waldman SD. Philadelphia: PA: Elsevier/Saunders; 2011.
[46] Wheeler AH, Murrey DB. Spinal pain: pathogenesis, evolutionary mechanisms,
and management. Chapter 30. In: The neurological basis of pain. New York:
McGraw-Hill M Pappagallo; 2005. p. 42152.
[47] Rydevik BL. The effects of compression on the physiology of nerve roots. J
Manipulative Physiol Ther 1992;15(1):626.
[48] Payan DG, McGillis JP, Goetzl EJ. Neuroimmunology. Adv Immunol
1986;39:299323.
Please cite this article in press as: Rigoard P, et al. Pathophysiological characterisation of back pain generators in failed back surgery
syndrome (part B). Neurochirurgie (2014), http://dx.doi.org/10.1016/j.neuchi.2014.10.104
[49] Mooney V, Presidential address. International Society for the Study of
the Lumbar Spine Dallas, 1986. Where is the pain coming from? Spine
1987;12(8):7549.
[50] Luschka H, Available from: http://books.google.fr/books?id=XiRCAAAAcAAJ
[51] Bogduk N, Tynan W, Wilson AS. The nerve supply to the human lumbar inter-
vertebral discs. J Anat 1981;132(Pt 1):3956.
[52] Sunderland S, Bradley KC. Stress-strain phenomena in human spinal nerve
roots. Brain 1961;84(1):1204.
[53] Manchikanti L, Pampati V, Singh V, Beyer C, Damron K, Fellows B. Evaluation
of the role of facet joints in persistent low back pain in obesity: a controlled,
prospective, comparative evaluation. Pain Physician 2001;4(3):26672.
[54] Pedersen HE, Blunck CF, Gardner E. The anatomy of lumbosacral posterior
rami and meningeal branches of spinal nerve (sinu-vertebral nerves); with
an experimental study of their functions. J Bone Joint Surg Am 1956;38-A(2):
37791.
[55] Yamashita T, Cavanaugh JM, el-Bohy AA, Getchell TV, King AI. Mechanosen-
sitive afferent units in the lumbar facet joint. J Bone Joint Surg Am
1990;72(6):86570.
[56] Sorkin L, Xiao W, Wagner R. TNF-alpha applied to the sciatic nerve trunk elicits
background ring in nociceptive primary afferents bers. In: IASP Abstr Vanc
IASP Press 8th World Congr Pain. 1996.
[57] Wagner R, Myers RR. Endoneurial injection of TNF-alpha produces neuropathic
pain behaviors. Neuroreport 1996;7(18):2897901.
[58] Vernon-Roberts B, Pirie CJ. Healing trabecular microfractures in the bodies of
lumbar vertebrae. Ann Rheum Dis 1973;32(5):40612.
[59] Vernon-Roberts B, Pirie CJ. Degenerative changes in the intervertebral discs
of the lumbar spine and their sequelae. Rheumatol Rehabil 1977;16(1):
1321.
[60] Kirkaldy-Willis WH, Wedge JH, Yong-Hing K, Reilly J. Pathology and pathogen-
esis of lumbar spondylosis and stenosis. Spine 1978;3(4):31928.
[61] Groen GJ, Baljet B, Drukker J. Nerves and nerve plexuses of the human vertebral
column. Am J Anat 1990;188(3):28296.
[62] Hirsch C, Ingelmark BE, Miller M. The anatomical basis for low back pain. Stud-
ies on the presence of sensory nerve endings in ligamentous, capsular and
intervertebral disc structures in the human lumbar spine. Acta Orthop Scand
1963;33:117.
[63] Vanharanta H, Sachs BL, Spivey MA, Guyer RD, Hochschuler SH, Rashbaum RF,
et al. The relationship of pain provocation to lumbar disc deterioration as seen
by CT/discography. Spine 1987;12(3):2958.
[64] Assaker R, Zairi F. Failed back surgery syndrome: to re-operate or not to
re-operate? A retrospective review of patient selection and failures. Neu-
rochirurgie 2014 [Accepted for publication (Hors-srie)].
[65] Hussain A, Erdek M. Interventional pain management for failed back surgery
syndrome. Pain Pract 2014;14(1):6478.
[66] Hainline B. Neuropathic Pain: Neuropsychiatric. Diagnostic, and Management
Considerations. Chapter 23. In: Pain management Waldman SD. PA: Else-
vier/Saunders: Philadelphia; 2011.