+ All Categories
Home > Documents > Discography Review

Discography Review

Date post: 30-Sep-2014
Category:
Upload: ramoob
View: 125 times
Download: 14 times
Share this document with a friend
Popular Tags:
21
Lumbar Discography: A Comprehensive Review of Outcome Studies, Diagnostic Accuracy, and Principles Steven P. Cohen, M.D., Thomas M. Larkin, M.D., Steven A. Barna, M.D., William E. Palmer, M.D., Andrew C. Hecht, M.D., and Milan P. Stojanovic, M.D. Background and Objectives: Since its advent more than 50 years ago, the use of discography has been mired in controversy. The purpose of this review is to provide a clinical overview of lumbar discography and discogenic back pain, with special emphasis on determining the accuracy of discography and whether or not the procedure improves outcomes for surgery. Methods: Material for this review was obtained from a MEDLINE search conducted from 1951 thru Septem- ber 2004, bibliographic references, book chapters, and conference proceedings. Results: Based on a large number of comparative studies, plain discography is less accurate than magnetic resonance imaging in diagnosing lumbar herniated nucleus pulposus and comparable or slightly more sensitive in detecting degenerative disc disease. For disc degeneration, CT discography remains the gold standard for diagnosis. There are very few studies comparing surgical outcomes between patients who have undergone preoperative provocative discography and those who have not. What little evidence exists is conflicting. Before disc replacement surgery, approximately half the studies have used preoperative discography. A comparison of outcomes did not reveal any significant difference between the 2 groups but none of the studies was controlled, and they used different outcome measures, follow-up periods, and surgical techniques. Because all intradiscal electrothermal therapy (IDET) studies have used discography before surgery, no conclusions can be drawn regarding its effects on outcome. Conclusions: Although discography, especially combined with CT scanning, may be more accurate than other radiologic studies in detecting degenerative disc disease, its ability to improve surgical outcomes has yet to be proven. In the United States and Europe, there are inconsistencies in the use of lumbar discography such that it is routinely used before IDET, yet only occasionally used before spinal fusion. Reg Anesth Pain Med 2005;30:163-183. Key Words: Degenerative disc disease, Discogenic pain, Discography, Internal disc disruption, Intervertebral disc, Lumbar spine. D iscography was first described in 1948 as a diagnostic tool for herniated nucleus pulposus (HNP). 1 Since that time, the development of sim- pler, safer, and more accurate imaging modalities have largely supplanted discography as an investi- gative technique for nerve root compression. Yet, provocative lumbar discography continues to be a popular, if controversial, means for diagnosing axial low back pain (LBP) caused by internal disc disrup- tion (IDD). This is because, unlike magnetic reso- nance imaging (MRI) or computed tomography (CT) scanning, discography is not just an imaging modality but a provocative test purported to corre- late symptoms with pathology. Although some studies have shown a high degree of correlation between discography results and histologic find- ings 2,3 and discography and surgical outcomes, 4,5 others have failed to show such a relationship. 6,7 Anatomy The intervertebral disc is composed of 3 major components: the nucleus pulposus (NP), annulus fibrosus (AF), and vertebral endplates (VE). Above From the Pain Management Center, Johns Hopkins Medical Institutions, Baltimore, MD; Pain Management Center, Walter Reed Army Medical Center, Washington, DC; MGH Pain Center, Dept. of Anaesthesia and Critical Care, Massachusetts General Hospital, Harvard Medical School, Boston, MA; Department of Anesthesiology, Walter Reed Army Medical Center, Washing- ton, DC; Department of Radiology, Massachusetts General Hospi- tal, Harvard Medical School, Boston, MA; and Department of Or- thopedic Surgery, Mount Sinai School of Medicine, New York, NY. Accepted for publication October 1, 2004. Reprints: Steven P. Cohen, M.D., Department of Anesthesiol- ogy and Critical Care Medicine, Johns Hopkins Medical Institu- tions, 550 North Broadway, Suite 301, Baltimore, MD 21205. E-mail: [email protected] © 2004 by the American Society of Regional Anesthesia and Pain Medicine. 1098-7339/05/3002-0008$30.00/0 doi:10.1016/j.rapm.2004.10.006 Regional Anesthesia and Pain Medicine, Vol 30, No 2 (March–April), 2005: pp 163–183 163
Transcript
Page 1: Discography Review

LoP

SW

D(phgppltn(

IRDHAttt

umbar Discography: A Comprehensive Reviewf Outcome Studies, Diagnostic Accuracy, andrinciples

teven P. Cohen, M.D., Thomas M. Larkin, M.D., Steven A. Barna, M.D.,illiam E. Palmer, M.D., Andrew C. Hecht, M.D., and Milan P. Stojanovic, M.D.

Background and Objectives: Since its advent more than 50 years ago, the use of discography has been miredin controversy. The purpose of this review is to provide a clinical overview of lumbar discography and discogenicback pain, with special emphasis on determining the accuracy of discography and whether or not the procedureimproves outcomes for surgery.

Methods: Material for this review was obtained from a MEDLINE search conducted from 1951 thru Septem-ber 2004, bibliographic references, book chapters, and conference proceedings.

Results: Based on a large number of comparative studies, plain discography is less accurate than magneticresonance imaging in diagnosing lumbar herniated nucleus pulposus and comparable or slightly more sensitive indetecting degenerative disc disease. For disc degeneration, CT discography remains the gold standard for diagnosis.There are very few studies comparing surgical outcomes between patients who have undergone preoperativeprovocative discography and those who have not. What little evidence exists is conflicting. Before disc replacementsurgery, approximately half the studies have used preoperative discography. A comparison of outcomes did notreveal any significant difference between the 2 groups but none of the studies was controlled, and they used differentoutcome measures, follow-up periods, and surgical techniques. Because all intradiscal electrothermal therapy (IDET)studies have used discography before surgery, no conclusions can be drawn regarding its effects on outcome.

Conclusions: Although discography, especially combined with CT scanning, may be more accurate than otherradiologic studies in detecting degenerative disc disease, its ability to improve surgical outcomes has yet to be proven.In the United States and Europe, there are inconsistencies in the use of lumbar discography such that it is routinelyused before IDET, yet only occasionally used before spinal fusion. Reg Anesth Pain Med 2005;30:163-183.

Key Words: Degenerative disc disease, Discogenic pain, Discography, Internal disc disruption, Intervertebraldisc, Lumbar spine.

mlsbio

A

cfi

otE

P

iscography was first described in 1948 as adiagnostic tool for herniated nucleus pulposus

HNP).1 Since that time, the development of sim-ler, safer, and more accurate imaging modalitiesave largely supplanted discography as an investi-ative technique for nerve root compression. Yet,rovocative lumbar discography continues to be aopular, if controversial, means for diagnosing axialow back pain (LBP) caused by internal disc disrup-ion (IDD). This is because, unlike magnetic reso-ance imaging (MRI) or computed tomographyCT) scanning, discography is not just an imaging

From the Pain Management Center, Johns Hopkins Medicalnstitutions, Baltimore, MD; Pain Management Center, Waltereed Army Medical Center, Washington, DC; MGH Pain Center,ept. of Anaesthesia and Critical Care, Massachusetts Generalospital, Harvard Medical School, Boston, MA; Department ofnesthesiology, Walter Reed Army Medical Center, Washing-

on, DC; Department of Radiology, Massachusetts General Hospi-

al, Harvard Medical School, Boston, MA; and Department of Or-hopedic Surgery, Mount Sinai School of Medicine, New York, NY.

Regional Anesthesia and Pain Medicine, Vol 30

odality but a provocative test purported to corre-ate symptoms with pathology. Although sometudies have shown a high degree of correlationetween discography results and histologic find-ngs2,3 and discography and surgical outcomes,4,5

thers have failed to show such a relationship.6,7

natomy

The intervertebral disc is composed of 3 majoromponents: the nucleus pulposus (NP), annulusbrosus (AF), and vertebral endplates (VE). Above

Accepted for publication October 1, 2004.Reprints: Steven P. Cohen, M.D., Department of Anesthesiol-

gy and Critical Care Medicine, Johns Hopkins Medical Institu-ions, 550 North Broadway, Suite 301, Baltimore, MD 21205.-mail: [email protected]© 2004 by the American Society of Regional Anesthesia and

ain Medicine.

1098-7339/05/3002-0008$30.00/0doi:10.1016/j.rapm.2004.10.006

, No 2 (March–April), 2005: pp 163–183 163

Page 2: Discography Review

aiVtjtrtTbt

wasitptabn(dsrlttc

ctroL

tplpnrfctTfifdalb

ts

tanpmnptila

Eo

tosnot

Ftasits(c

164 Regional Anesthesia and Pain Medicine Vol. 30 No. 2 March–April 2005

nd below the disc lie the vertebral bodies. The discs attached to the adjacent vertebral bodies by theE centrally and the ligamentous attachments of

he AF peripherally. These components form aoint-like structure that allows for movements inhe sagittal, horizontal, and coronal planes.8 Poste-iorly, it is supported by the other components ofhe “3-joint structure,” the 2 zygapophysial joints.ogether, these joints function to support and sta-ilize the spine and prevent injury by limiting mo-ion in all planes of movement.

The healthy adult disc is essentially avascular,ith its nutrition being supplied through the VE

nd AF via diffusion. The nucleus itself has no bloodupply. The annulus contains blood vessels only ints most superficial lamellae. Nutrients that passhrough the endplates come from the arteries sup-lying the vertebral bodies. Anabolic functions ofhe healthy disc are maintained by chondrocytesnd fibroblasts; catabolic functions are performedy the matrix metalloproteinase enzymes collage-ase (which degrades collagen) and stromelysinwhich degrades proteoglycans). This balance is pHependent, requiring oxygen, glucose, and sub-trates. The pH of the NP is between 6.9 and 7.1,eflecting anaerobic metabolism secondary to itsack of oxygen.9,10 Any number of factors can con-ribute to a breakdown in the functional capacity ofhe disc, including inflammatory mediators,hanges in pH, and nutritional deficiencies.8

The nerve supply to the intervertebral disc isomplex. The most significant source of innervationo the lumbar spine is derived from the L2 dorsaloot ganglion.11 This is the basis for the eradicationf lumbar discogenic back pain after blockade of the2 nerve roots.12

Functionally, the innervation to the lumbar in-ervertebral discs stems from 2 extensive nervelexi that accompany the posterior and anteriorongitudinal ligaments.8 These are known as theosterior and anterior plexuses. The sinuvertebralerve, formed from the confluence of a somaticoot from the ventral ramus and an autonomic rootrom the gray ramus communicantes (which re-eives its autonomic input from the sympatheticrunk), is the largest branch in the posterior plexus.his plexus is a diffuse network of interconnectingbers receiving somatic and autonomic branches

rom multiple spinal levels.8,13 In addition to theiscs, the posterior plexus innervates the ventralspect of the dura mater, the posterior longitudinaligament, and the posterior portion of the vertebralodies.The anterior plexus receives contributions from

he anterior branches of the rami communicantes,

mall medioventral branches of the sympathetic s

runk, and perivascular nerve plexi. The posteriornd anterior plexi are connected via a less promi-ent conglomeration of nerves known as the laterallexus, formed by branches of the grey rami com-unicantes.8 Several different types of nonvascularerve endings have been described including sim-le, cluster, partially, and fully encapsulated.14 Al-hough the exact role of each type of nerve endings unknown, it is speculated that under nonpatho-ogical conditions they primarily function as mech-noreceptors (Fig 1).15

merging Theories on the Pathogenesisf Discogenic Pain

In the 1940s and 50s, the disc itself was thoughto be devoid of any nerve supply and thus incapablef being a potential pain generator.16,17 Subsequenttudies have since refuted this assertion. In theormal human disc, sensory nerves extend into theuter third of the annulus. In the degenerated disc,he innervation is deeper and more extensive, with

ig 1. Schematic drawing of the nerve plexi surroundinghe vertebral body (VB) and intervertebral disc (ID); 1nd 7 represent the anterior and posterior plexuses, re-pectively. The deep, extensive penetration of the nervesndicates degeneration has occurred. 2, sympatheticrunk; 3, rami communicantes; 4, ventral ramus of thepinal nerve; 5, dorsal ramus; 6, sinuvertebral nerves.Modified from Groen et al.13 Drawing courtesy of Spe-ialist Jennifer Sempsroft, US Army.)

ome nerve fibers penetrating into the nucleus pul-

Page 3: Discography Review

pvc

weclpifibthic

tesgwgpsnTaOsduoAttt

wlsotTclscee2a

ttisrttmlfstrrepipgdt

w adjaherea

Lumbar Discography • Cohen et al. 165

osus.15,18,19 It is now generally accepted that inter-ertebral discs can be and frequently are a signifi-ant source of back pain.13,14,20-22

Mechanical Changes. In the normal disc,eight bearing is uniformly distributed across the

ntire disc plane.23,24 This is accomplished by me-hanical interplay between the AF and NP. When aoad is placed on the disc, the NP acts as a noncom-ressible mass, with its gelatinous contents bulgingn the axial, sagittal, and coronal planes. The intactbers of the annulus tend to resist this outwardulging, resulting in an even distribution of forcehroughout the entire surface of the disc.8 In theealthy spine, the AF is not unduly stressed because

ts broad surface area translates into the nucleusarrying the bulk of the load.Age-related changes can lead to a breakdown in

he normal mechanical process of load bearing. Asarly as the second decade of life, the nutritionalupply to the endplate diminishes because of aradual reduction in blood flow.25 This, combinedith either an acute event such as a torsional load

reater than the disc is capable of handling or re-etitive low-level stress, can lead to 2 possibleources of injury. The first is a break in the conti-uity of the VE, also known as a microfracture.26

he second is a tear in the lamellar structure of thennulus (i.e., annular tear) from torsional overload.nce this occurs, the ability of the nucleus to dis-

ipate a compressive load changes. Unlike normaliscs, a compressive load in degenerated discs is notniformly distributed. Instead, the preponderancef the weight bearing burden is borne by theF.23,24 Although this can be maintained for short

ime periods, if the endplate fracture does not heal,he repetitive stress on the annulus eventually leads

Table 1. Modic Changes on MRI in

Category of Signal IntensityChange* T1-Weighted MR Image

Type I† Decreased signal intensity

Type II Increased signal intensity

Type III Decreased signal intensity

Adapted from Modic et al.28

*Refers to signal intensity changes in the vertebral body marro†Type I changes have been observed to progress to type II, w

o the development of tears. These tears reduce the t

eight bearing capacity of the annulus because theamellar fibers that are torn no longer function asupport apparatus. This results in even more stressn the remaining fibers, which leads to furtherearing and an eventual loss of annular integrity.27

he disc is then predisposed to herniation of itsontents and subsequent diminution in height. Theoss of disc height may then degenerate into furtherhortening of the disc, replete with the pathologichanges commonly seen in degenerative disc dis-ase (DDD) such as Modic changes, sclerosis of thendplates, and bridging osteophytes (Tables 1 and).28,29 In severe cases, this manifests as autofusionnd ankylosis of adjacent segments.30

Chemical Changes. The chemical changeshat occur within the disc are 2-fold. First, the frac-ure of an endplate can lead to the introduction ofnflammatory cytokines into the nucleus. This up-ets the delicate nutritional balance of the disc,esulting in reduced oxygen diffusion, a rise in lac-ate levels, and a decrease in pH. These factors serveo slow metabolic and reparative processes and aug-ent catabolic processes via an increase in metal-

oproteinase and chondrocyte activity, leading tourther disc degradation.10,31 Under certain circum-tances, cytokines can be a direct source of pain. Ifhe annulus is intact, the patient should not expe-ience pain because inflammatory mediators cannoteach sensory nerve endings in the outer disc. How-ver if an annular tear is present, these chemicalain mediators can now reach nociceptors, produc-ng LBP.32 The situation may be further com-ounded by perivascular or nonvascular nerve in-rowth penetrating as deep as the NP, which is wellescribed in DDD.15 Pain may also be produced byhe sensitization and irritation of nerve endings in

nts With Degenerative Disc Disease

2-Weighted MR Image Histopathologic Changes

creased signal intensity Disruption and fissuring of theendplate and vascularizedfibrous tissue within theadjacent marrow

o- or slightlyhyperintense signal

Endplate disruption withyellow marrow replacementin the adjacent vertebralbody

ecreased signalintensity

Extensive bony sclerosis,indicative of dense wovenbone within the vertebralbody rather than themarrow

cent to the end plates of degenerative intervertebral discs.s type II changes tend to remain stable.

Patie

T

In

Is

D

he endplate.27 This model of chemical nociception

Page 4: Discography Review

imrn(kcb

pfatald“oaplgicwacpta

sIbtdoadboedotow

TF

vepttaasCrsmlsm8w(d

edTpdpehpadtd

166 Regional Anesthesia and Pain Medicine Vol. 30 No. 2 March–April 2005

s supported by numerous studies showing disc im-unoreactivity to substance P and calcitonin gene-

elated peptide,33,34 as well as elevated levels ofitric oxide,31 prostaglandin E2,31,35 interleukinIL)-2,31 IL-6,35 IL-8,35 phospholipase A2,36-40 leu-otriene B4,41 thromboxane B2,41 and tumor ne-rosis factor-� (TNF-�)42,43 in diseased interverte-ral discs.Taken together, these factors work in concert to

rovide a mechanical and biochemical explanationor the basis of discography. As one might expect,cute exacerbations of pain are best explained byhe preponderance of inflammatory mediatorsround the sensitized disc. In this scenario, evenow pressures may provoke intense pain duringiscography. This is the basis for the concept ofchemically sensitized” discs (see technical aspectsf discography, Table 4). When the disruption is lesscute, the disc may tolerate greater amounts ofressure and react to stimulation only at higheroads, at which point the nerve fibers of the de-raded annulus are stretched to the point of painnduction. This scenario describes the “mechani-ally sensitized” disc. In practical terms, a patientith a chemically sensitive disc might experience

lmost constant LBP, whereas someone with a me-hanically sensitized disc may have pain only withrolonged sitting or lifting. Another way to considerhese 2 concepts is that a chemically sensitive disc is

Table 2. Categorization of Degenerative Disc DiseaseBased on MRI

Category Nucleus SignalProlapseDetection

Bone MarrowSignal

A No signal loss No prolapse No intensitychange

B No signal loss Prolapse No intensitychange

C No signal loss No prolapse Intensitychange

D No signal loss Prolapse Intensitychange

E Moderate signalloss

No prolapse No intensitychange

F Moderate signalloss

Prolapse No intensitychange

G Moderate signalloss

No prolapse Intensitychange

H Moderate signalloss

Prolapse Intensitychange

I Total signal loss No prolapse No intensitychange

J Total signal loss Prolapse No intensitychange

K Total signal loss No prolapse Intensitychange

L Total signal loss Prolapse Intensitychange

Adapted from Frobin et al.29

nalogous to allodynia, whereas a mechanically t

ensitive disc is more comparable to hyperalgesia.n the latter stages of disc disease, the annulusecomes functionally incompetent. In these cases,he injection of contrast fails to generate high intra-iscal pressures. Depending on the patient, pathol-gy, and interpreter, these discograms may be reads positive, negative, or indeterminate. Normaliscs resist pain with stimulation because they lackoth the chemical sensitization and mechanicalverloading seen in diseased discs. In practice, thesexamples represent an idealized diagnostic para-igm that does not take into account the multitudef genetic, social, cultural, and psychological factorshat affect pain perceptions. To optimize treatmentutcomes, these factors must be taken into accounthen performing any pain-provoking procedure.

he Controversy Over Discography:alse-Positive Pain Provocation

Originally used as a diagnostic tool for HNP, pro-ocative discography is no longer used to routinelyvaluate radiculopathy, having been largely re-laced by the advent of noninvasive, more sensitiveests such as CT scanning and MRI. The evidencehat these modalities are not only safer but moreccurate than plain discography in detecting herni-ted nuclear material is irrefutable. In a prospectivetudy by Jackson et al.44 comparing myelography,T myelography, plain discography, and CT discog-aphy with surgical findings in 231 discs explored aturgery, the authors found CT discography to be theost accurate test (87%) and discography to be the

east accurate (58%). Disco-CT ranked highest inensitivity for HNP (92%) compared with 78% foryelo-CT, 72% for CT, 70% for myelography, and

1% for plain discography. For specificity, CT-discoas also the most accurate (81%) followed by CT

76%), CT-myelo (76%), myelography (70%), andiscography (31%).In the first phase of a 2-part experiment, Yasuma

t al.45 studied 181 lower thoracic and lumbar ca-aver discs discographically and histologically.heir findings revealed 32 true-positive, 15 false-ositive, 122 true-negative, and 12 false-negativeiscograms. Discograms were designated as falseositive when the injected contrast was noted toxtend beyond the peripheral vertebral margin, butistologic sectioning of the disc was negative forrotrusion. False-negatives were defined as a neg-tive discogram despite a histologically confirmedisc protrusion. In the second phase of this study,he authors retrospectively analyzed the cases of 77iscography patients who were subsequently found

o have a herniated disc during surgical exploration.
Page 5: Discography Review

Ti5ppfincfcmm

opatfiOipbitil

rfpfvL6agpisocldp

shstgw34pta

swfbslwcatpgImtflabiotodrsst7idtastfotasIdGitgitpafsa

Lumbar Discography • Cohen et al. 167

he discograms were falsely interpreted as negativen 32% of the 59 patients with a protruding disc and6% of the 18 patients with a prolapse.45 In arevious study, Yasuma et al.46 found that false-ositive discograms were more likely to occur whenssures or cysts were present in a degenerated an-ulus but did not establish continuity with the nu-lear cavity. MRI, which is more sensitive than CTor disc pathology, is also more accurate than dis-ography in diagnosing acute disc herniation.47 Inost studies, discography has been found to beore sensitive than specific.Proponents of discography argue that it is the

nly diagnostic modality that attempts to correlateathology with symptoms. This point seems reason-ble given the fact that close to two thirds of asymp-omatic subjects have been found to have abnormalndings on MRI scans of their lumbar spines.48

pponents of discography question both the signif-cance of discographic pathology and the validity ofrovoked symptoms. These criticisms are supportedy the relative lack of specificity of discography, thenherent difficulty in validating provoked symp-omatology, and the large number of studies show-ng false-positive discograms in patients withoutow back symptoms.

The first study to question the validity of discog-aphy was published in 1968 by Holt,49 who foundalse-positive results in 37% of 30 asymptomaticrisoners. Over 20 years later, Walsh et al.50 per-ormed CT discography in 10 asymptomatic maleolunteers and 7 “control” patients with chronicBP, 2 of whom were female. In the control group,5% of the 20 discs injected were radiologicallybnormal, with all 7 patients having at least 1 de-enerated disc. In the study group without backain, CT discograms were interpreted as abnormaln 17% of the 35 discs injected and half of the 10ubjects. However, none of these 5 patients met thether criteria for a positive discogram, which was aoncordant pain response coupled with pain-re-ated behavior (e.g., grimacing, guarding, with-rawing, verbalizing, and so on). Thus, the false-ositive rate in this study was 0%.In 1999 and 2000, Carragee et al.51-53 published a

eries of studies attempting to identify patients atigh risk for false-positive discograms. In the firsttudy, 8 patients without baseline low back symp-oms who had undergone recent iliac crest bonerafting for non-thoracolumbar spine proceduresere studied with provocative discography of theirmost caudal discs.51 Among the 8 study subjects,experienced severe LBP concordant with their

ostoperative bone graft discomfort during injec-ion of at least 1 disc. All symptomatic discs had an

bnormal morphologic appearance. In the second b

tudy, 26 individuals without low back symptomsere divided into 3 groups: those who were pain

ree (n � 10), those with chronic neck and arm painut no back symptoms (n � 10), and those withomatization disorder (n � 6).52 Each patient had ateast their 3 most caudad lumbar discs injected,ith 5 patients having the L2-3 disc done as a

ontrol. The incidence of false-positive discography,s defined by moderate or severe pain during con-rast injection in at least 1 disc, was 10% in theain-free group, 40% in the chronic cervical painroup, and 83% in the somatization disorder group.n this study, none of the 31 radiographically nor-al discs were associated with pain during injec-

ion. In the last study, 3-level discography was per-ormed in 47 patients who had undergone single-evel discectomy.53 Twenty of these subjects weresymptomatic, and 27 patients had recurrent lowack or leg symptoms. In the asymptomatic partic-pants, positive injections occurred in 8 of 20 (40%)perated discs. In the “control” patients with symp-oms, positive injections occurred in 17 of the 27perative discs (63%), with the pain being concor-ant in 15 of these. All positive discs containedadiographic abnormalities. Consistent with theirecond study, patients with normal psychometriccores were less likely to have false-positive injec-ions than those with abnormal scores (43% vs0%). Possible causes of false-positive discogramsnclude inadvertent annular injection, contrast-in-uced irritation of nervous tissue, endplate deflec-ion resulting from suboptimal needle placement,nd stimulation of pressure receptors when exces-ively high pressures are generated.54,55 Whereashe findings of Carragee et al.51,52 and others arerequently cited as evidence against the routine usef lumbar discography before operative interven-ion, an inherent flaw in these studies is that theuthors cannot consider true pain concordance inubjects without pre-existing low back symptoms.n the past 2 decades, this criterion has become aefining characteristic of a positive discogram.iven the high propensity for false-positive findings

n patients with previous back surgery, psychopa-hology, or somatization symptoms, positive disco-rams should be viewed with caution in thesendividuals. To optimize specificity, we suggest ob-aining 2 adjacent control discs, a recommendationreviously put forth by Endres and Bogduk56 in anttempt to improve accuracy in patients at high riskor false-positive discography. Although it mayeem intuitive that this would enhance specificitynd/or improve outcomes, these issues have yet to

e addressed in the literature.
Page 6: Discography Review

168 Regional Anesthesia and Pain Medicine Vol. 30 No. 2 March–April 2005

Table 3. Studies Comparing Lumbar Discography With CT Scanning or MRI in PatientsWith Degenerative Disc Disease

Author, Year Number of Subjects Nature of Study Results Comments

Sandhu, 200062 53 patients with LBP,133 discograms

Retrospective analysiscomparing discographywith vertebral endplatesignal changes on MRI.

No significant correlation betweendiscography and endplatesignal changes.

41% of discs with (�) endplatechanges were (�)discograms vs. 27% without.Among (�) discograms,only 23% exhibited T2-weighted MRI endplatechanges.

Brightbill, 19942 7 patients with LBP Clinical case seriesinvolving pts withdiscrepancy betweendiscography and MRIwho underwent surgeryand were found to haveinternal disc disruption.

All 7 subjects had normal MRimaging and (�) discography.

Did not consider surgicaloutcomes.

Yu, 198963 8 cadavers, 36 discs Compared MRI anddiscography againstcryomicrotomy anatomicsectioning for detectingannular tears.

Discography identified 15 radialfissures, 10 of which wereseen on MRI. Two of the 15annular fissures were missedon cryomicrotomy.

Included a newborn and 2-year-old. Considered onlyradial tears of annulus.Could not correlate findingswith symptoms.

Kakitsubata, 200364 24 discs from 5cadavers

Compared MRI and MRdiscography withanatomic correlation fordetecting annular tears.

Sensitivity of MR-discogr was100%, 57%, and 21% forradial, transverse andconcentric, tears in annulus,respectively, vs. 67%, 71%,and 21% for conventional MRI.

Could not correlate findingswith symptoms.

Zucherman, 198865 18 patients with LBPwith or withoutradicularsymptoms

Clinical case series. In mostcases discography wasfollowed by CT scanning.

All patients had normal MRI andabnormal discograms

Normal MRI and abnormaldiscograms were basis forinclusion.

Horton, 199266 25 patients with non-radicular LBP,involving 63discograms

Comparative study betweenMRI and discography fordiscogenic LBP.

19 patients had (�) discograms.Of the different MRI patterns,only ‘dark/torn’, ‘dark/bulged’ or‘speckled/flat’ were more likelyto be associated with (�)rather than (�) discograms.

MRI findings classified bypattern, not presence orabsence of pathology.

Collins, 199059 29 patients, 73discograms

Compared MRI anddiscography in pts withaxial LBP.

57 discs were abnormal ondiscogr, with 13 producingconcordant pain in 12 pts.Discogr findings correlated withMRI in 90% of cases. 4 discsshowed degeneration ondiscogr w/ nml MRI, and 4 hadabnormal MRI w/ nmldiscography.

The 12 pts with (�)discograms underwentspinal fusion, with 9reporting clinicalimprovement at 9-month f/u.

Schneiderman,198761

36 pts with LBP, withor without leg pain,101 discograms

Compared MRI anddiscography.

MRI accurate in assessing discmorphology in 100 of 101levels. Of 52 discs with nmlMRI, only 1 had (�)discogram. Of 49 discs withdecreased MR signal, only 2discograms nml.

Used only T2-weighted MRI.CT-discography used on 39levels.

Milette, 199967 45 pts, 132 discs Evaluated MRI anddiscography results in ptswith chronic LBP.

On MRI, 71% of discs showednml contour, and 64% nmlsignal intensity. Only 40% ofdiscograms wereradiographically nml. Discogrdemonstrated stage 2 and 3disc disruptions in 26% of discsw/nml contour on MRI, and13% of discs with both nmlcontour and signal.

Used only T2-weighted MRI.Study was designed toassess differences betweendisc protrusions, bulges, &loss of signal intensity onMRI, not to compareimaging studies.

Linson, 199060 50 pts, 97 discs Compared MRI anddiscography in pts withchronic LBP.

91% correlation for discdegeneration between MRI anddiscogr.

5 of the 6 discs with negativecorrelation were read as nmlby MRI and abnormal bydiscography. No mention ofcontrol discs during discogr.

Simmons, 199158 164 pts, 371 discs Compared CT-discographyand MRI in pts withchronic LBP with orwithout radiculopathy.

55% correlation based on pts,80% based on discs.

MRI nml and discogramabnormal in 34 discs.Discogram nml and MRIabnormal in 60 discs. 37%of discs abnormal on MRIwere asymptomatic ondiscogr. Did not includeoutcomes in 76 pts who

underwent surgery.
Page 7: Discography Review

CR

pmdcflMa4didb

nsAbo

tprt6gnds

tion; HIZb

Lumbar Discography • Cohen et al. 169

orrelation of Discography Withadiologic Studies

Although there is strong evidence that discogra-hy, with or without CT scanning, may overesti-ate the prevalence of clinically significant internal

isc disruption, it is equally clear that, even whenorrectly performed and analyzed, discography mayail to detect disc pathology seen with other radio-ogic studies. In a study by Gibson et al.57 comparing

RI and discography in the diagnosis of DDD,greement between the 2 techniques was found in4 of 50 discs studied. In the 6 discs in which aiscrepancy occurred, evidence of IDD was missedn 5 discograms and 1 MRI. In the 5 cases in whichiscography failed to detect disc pathology, 2 were

Table 3.

Author, Year Number of Subjects Nature of Study

Gibson, 198657 22 pts, 50 discs Compared MRI anddiscography in pts wmechanical LBP.

Yoshida, 200268 23 pts, 56 discs Examined correlationbetween MR imagespain response ondiscography.

Birney, 199269 90 pts, 264 discs Examined correlationbetween MRI anddiscography for DDDHNP. Compared surfindings with discogr57 pts.

Osti, 199270 33 pts, 114 discs Compared MRI anddiscography in pts wLBP.

Schellas, 199671 63 pts, 100 discswith HIZ on T2MRI in pts withLBP and/orradicular pain.

Retrospective analysisanalyzing thesignificance of HIZ zin predicting (�)discography.

Loneragan, 199472 18 pts with chronicLBP thought to bediscogenic (43discs).

Compared MRI and CTdiscogr in the diagnoof DDD and HNP.

Aprill, 199273 41 pts (105 discs)with chronic LBPwith or w/oradicularsymptoms.

Compared HIZ on T2-weighted MRI with Cdiscography.

NOTE. Discography refers to discograms performed without CT scanniAbbreviations: DDD, degenerative disc disease; IDD, internal disc disrup

ack pain; nml, normal; pts, patients.

ecause of incorrect placement of the discography i

eedle in the annulus, with failure to detect earlyigns of degeneration accounting for the other 3.lthough the authors recorded symptoms producedy disc injections, discography results were basedn radiologic findings only.In a similar study by Simmons et al.,58 the au-

hors compared CT-discography with MRI in 164atients with chronic LBP, some of whom sufferedadicular symptoms. Correlation between the 2echniques was seen in 55% of cases. In 13% (n �0) of discs, MR images were abnormal but disco-rams normal. Among all the discs classified as ab-ormal on MRI, 37% (n � 108) were asymptomaticuring disc injection. In 7% of discs (n � 34), MRIhowed normal and discography abnormal find-

tinued)

Results Comments

Agreement between studies in 44of 50 discs.

Discography results based onradiographic findings only aspts were sedated. In the 6discs that didn’t correlate,MRI was superior todiscography.

Sensitivity, specificity, positivepredictive value, and negativepredictive value of T2-MRIwere 94%, 71%, 59%, and97%, respectively.

Did not specifically comparediscography and MRI. T2-weighted MRI superior togadolinium-enhancedimages.

Agreement between MRI anddiscogr in 86% of discs. MRImore accurate for HNP;Discogr slightly superior to MRIfor DDD (MRI missed 1 disc,discogr 100% sensitive).

Considered pts with LBP andradicular pain. Surgicalfindings correlated withdiagnostic studies at 63 of76 levels.

All 54 discs identified asabnormal on MRI showedabnormal discogram patterns.6 of the 60 discs identified asnml on MRI were abnormal ondiscogr. Of the 39 discs thatprovoked concordant pain ondiscogr, 27 were abnormal onMRI. 33 of the 39asymptomatic discs by discogrhad nml MRI signals, with 24having nml discographicpatterns.

Six of 46 discs classified asdegenerate on MRI wereasymptomatic atdiscography. Concludeddiscography is moreaccurate than MRI fordetecting annular pathology.Pt population not well-defined.

All 100 discs with HIZ werediscographically abnormal, with87 showing concordant pain. In17 asymptomatic control pts,MRI scans revealed only 1 HIZdisc.

37 pts had prior back surgery.Also included pts withradiculopathy.

MRI missed 3 of 10 discs withearly degenerative changes,and 1 of 3 herniations.

In no cases did MRI offer moreinformation than CT-discogr.

In all pts who exhibited an HIZ onMRI, CT-discogr revealedeither a grade 3 or 4 annulardisruption. A grade 3 or 4disruption was also present in34 pts w/o an HIZ.

Concordant pain provocationwith discogr was present in38 of 40 HIZ discs, and 22of 78 discs w/o an HIZ. CT-discogr performed in only 41out of 500 pts in whom MRIwas examined.

, high-intensity zone on MRI; HNP, herniated nucleus pulposus; LBP, low

(Con

ith

and

andgicalin

ith

ones

-sis

T-

ng.

ngs. In 21 of these 34 discs, injection of contrast

Page 8: Discography Review

iciidceMiccfp

rbi9wnedeotrocwet(o

CO

p

dwtbicsrttcc

pfibsfmgvdtwachbdepptcc

170 Regional Anesthesia and Pain Medicine Vol. 30 No. 2 March–April 2005

nto the disc elicited exact pain reproduction. In aomparative study evaluating MRI and discographyn patients with axial LBP, Collins et al.59 found thatmaging characteristics for the 2 diagnostic proce-ures correlated in 65 of 73 discs. In the other 8ases, 4 disc levels showed evidence of early degen-ration on discography but appeared normal onRI, whereas the other 4 showed decreased signal

ntensity on T2-weighted MR images but were dis-ographically normal. In the 12 patients with con-ordant pain on discography, spinal fusion was per-ormed. At their 9-month follow-up, 9 of 12atients reported clinical improvement.Several studies have shown a much stronger cor-

elation between MRI and discography. In a studyy Linson and Crowe,60 findings between the 2nvestigative modalities were in agreement in 91 of7 discs studied in 50 patients. In the 6 discs inhich a discrepancy was present, 5 were read asormal by MRI but abnormal by discography. In anarlier study by Schneiderman et al.,61 MRI andiscography positively correlated in 100 of 101 lev-ls. Overall, discography appears to be comparabler slightly more sensitive for the detection of IDDhan MRI or CT scanning, especially with regard toadial annular fissures. In discs read as degenerativen MRI, approximately 15% will fail to provokeoncordant pain on discography. The main problemith correlative studies is that when an incongruity

xists, it is impossible to determine whether or nothe discrepancy is caused by a lack of sensitivityfalse negatives) or specificity (false positives) inne of the diagnostic procedures (Table 3).

orrelation With Lumbar Spinal Fusionutcomes

Axial back pain is one of the most challenging

Table 4. Interpreta

Disc ClassificationIntradiscal Pressure at Pain

Provacation P

Chemical Immediate onset of pain occurringas �1 mL of contrast isvisualized reaching the outerannulus, or pain provocation at�15 PSI above openingpressure

Mechanical Between 15 and 50 PSI aboveopening pressure

Indeterminate Between 51 and 90 PSI �Normal �90 PSI N

Abbreviation: PSI, pounds per square inch.Modified from Derby et al.84

roblems confronted by physicians. There are many e

isorders besides DDD that can cause axial LBP,ith 2 of the more common ones being facet ar-

hropathy and myofascial pain. In a porcine studyy Indahl et al.,74 the authors determined that theres significant overlap between the neuromuscularonnections of the intervertebral discs, zygapophy-eal joints, and paraspinal muscles, such that theelative contributions to LBP of each of these struc-ures may be difficult to estimate. It is thus likelyhat in many, if not most patients suffering fromhronic LBP, several different pain generators existoncurrently.Several authors have attempted to determine the

revalence of discogenic pain in patients sufferingrom back symptoms. In one of the most cited stud-es, Schwarzer et al.75 found the incidence of IDD toe 39% in 92 patients with chronic LBP. Notwith-tanding the fact that many patients with DDD suf-er from other concomitant causes of back pain that

ay not respond to operative intervention, the sur-ical treatment of DDD is itself mired in contro-ersy. Although outcome studies for spinal arthro-esis vary widely, they are generally acknowledgedo be less beneficial than surgery for radicular pain,ith success rates ranging from less than 50% to

lmost 90%.76 These quoted success rates must beonsidered in light of the fact that no fusion studiesave ever been conducted under controlled,linded conditions. Moreover, a recent Cochraneatabase review concluded there was no scientificvidence supporting any form of surgical decom-ression or fusion in the treatment of DDD.77 Theresence of other possible pain generators in pa-ients with discogenic pain, along with the mixedlinical outcomes even when arthrodesis is techni-ally successful,77,78 are factors that must be consid-

f Disc Stimulation

verity Pain Type Interpretation

Concordant Positive

Concordant Positive (but other pain generatorsmay be present; furtherinvestigation is warranted)

Concordant Further investigation warrantedor

ureNA Negative

tion o

ain Se

7/10

7/10

7/10o painpress

red when evaluating clinical studies examining the

Page 9: Discography Review

cc

dcicdowsobsu

Ctatopwepto

endtprtfapretroppmvmgr

oMLb4

dp2dmffltepcot

CER

tsitmfiacta

bsldrf

arsnsawwfarca

D

Lumbar Discography • Cohen et al. 171

orrelation between discography results and surgi-al outcomes.Several investigators have attempted to correlate

iscography results with surgical findings and out-omes. Colhoun et al.4 evaluated surgical outcomesn 162 patients who underwent preoperative dis-ography for axial LBP. In the 137 patients whoseiscography provoked pain, 89% had a favorableutcome (mean follow-up period for study patientsas 3.6 years). In the 25 patients whose discs

howed morphological abnormalities but no prov-cation of symptoms, only 52% reported significantenefit. Of note, the authors did not measure pres-ure during discography and the surgical treatmentssed were predominantly spinal fusions.Not all outcome studies have been as positive as

olhoun’s.4 In a prospective study by Esses et al.,79

he authors assessed the role of external spinal fix-tion in predicting the success of subsequent ar-hrodesis in patients with chronic LBP. Thirty-twof the 35 subjects underwent provocative discogra-hy before fixator placement. Of the 21 patientshose discography showed radiographically degen-

rated discs, 17 experienced complete or significantain relief with the external fixator. In the 13 pa-ients for whom results were available, 9 went on tobtain significant relief after spinal fusion.In the 11 patients with normal-appearing discs, 8

xperienced complete or significant relief with spi-al fixation, with 6 of the 8 who went on to surgeryeriving clinical benefit from their fusion. Amonghe 15 patients with concordant pain on discogra-hy, 8 experienced complete relief and 5 significantelief after external spinal fixation. In the 10 pa-ients with concordant pain who underwent spinalusion, 3 obtained complete pain relief, whereasnother 3 experienced significant relief. Of the 17atients with non-concordant or no pain on discog-aphy, 5 experienced complete relief and 7 experi-nced significant relief after fixator placement. Inhe 12 patients with no or nonconcordant discog-aphy pain who went on to definitive surgery, 3btained complete pain relief, with another 6 re-orting significant relief. The follow-up period afterosterior spinal fusion is not mentioned in theanuscript. Although determining the predictive

alue of discography on surgical success was not theain objective of Esses’ study, their outcomes sug-

est positive discograms do not predict therapeuticesponse.

The predictive value of provocative discographyn surgical outcome was also assessed in a study byadan et al.6 involving 73 patients with chronic

BP. Thirty-two patients underwent spinal fusionased on pain provocation during discography, with

1 patients having surgery without the benefit of i

iscography. In the discography group, 75.6% ofatients had satisfactory outcomes at a minimum-year follow-up versus 81.2% in the group whoid not have preoperative discography. In sum-ary, the lack of strong evidence for the use of

usion to treat DDD,77,78 and the methodologicalaws in the existing studies makes interpretation ofhe data exceedingly difficult. For the data that doesxist, the results are mixed as to whether or notreoperative discography improves surgical out-omes in patients with discogenic LBP. The resultsf studies evaluating discography as a predictiveool for surgery are presented in Table 5.

orrelation With Intradiscallectrothermal Therapy and Disceplacement Surgical Outcomes

In recent years, the advent of intradiscal electro-hermal therapy (IDET) and total disc replacementurgery have generated intense interest in the med-cal community. As a treatment for discogenic pain,he outcomes for IDET vary widely, ranging frominimal benefit86 to success rates approaching 80%

or single-level procedures.87 Most studies reportmprovement rates in the 50% range.88,89 Becausell published studies have used preprocedure dis-ography as a screening test for IDET candidates,he effect the procedure has on outcomes cannot bessessed.The concept of disc arthroplasty was introduced

y Fernstrom in 1966,90 who described using atainless steel ball as a vertebral spacer to restoreost disc height and preserve motion. Over the lastecade, there has been renewed interest in disceplacement surgery, with most studies comingrom Europe.

The studies assessing disc replacement surgeryre almost equally divided between those that haveoutinely used preoperative discography as acreening test and those that have not. Unfortu-ately, methodological flaws in these uncontrolledtudies, wide variability in outcome criteria, and thebsence of any direct comparisons between patientsho underwent preoperative discography and thoseho did not preclude any meaningful conclusions

rom being drawn. Based on the limited data that arevailable, it does not appear that preoperative discog-aphy confers any significant benefit in patients beingonsidered for total disc replacement. These studiesre summarized in Table 6.

iscography Versus Bony Vibration Test

Proponents of discography claim the procedure is

ndispensable, being the only diagnostic tool that
Page 10: Discography Review

potcs

opps

I t to theator; ID

172 Regional Anesthesia and Pain Medicine Vol. 30 No. 2 March–April 2005

urports to correlate symptoms with pathol-gy.104,105 However, this distinction no longer holdsrue. The bony vibration test (BVT), which entailsompressing a blunt, vibrating object (usually the

Table 5. Clinical Studies Evaluating Effect o

Study, Year Number of Patients Type of Stu

Gill, 199280 53 patients withpredominantly axial LBPand IDD at L5-S1.

Retrospective sinvolving L5-fusion.

Colhoun, 19884 162 patients with axialLBP.

Prospectiveobservationastudy.

Parker, 199681 23 patients withmechanical LBP andpositive discography.

Prospective casseries involvspinal fusioninstrumentati

Esses, 198979 35 patients with chronicLBP and FBSS whounderwent ESF beforearthrodesis.

Prospective stuevaluating efESF prior tofusion. 32 ptunderwent pdiscography.

Kostuik, 197982 350 pts with painfulscoliosis whounderwent spinalinstrumentation.

Retrospective s

Wetzel, 19947 48 pts with axial LBP whounderwent lumbararthrodesis followingprovocativediscography.

Retrospective s

Madan, 20026 41 pts who underwentspinal arthrodesiswithout pre-opdiscography and 32 ptswho underwent surgerybased on (�)discography.

Not indicated

Knox, 199383 22 pts who underwentanterior spinal fusion fordiscogram-concordantLBP

Retrospective s

Derby, 199984 96 pts who underwentdiscography for LBP

Retrospective s

Vamvanij, 199885 56 pts with discogenicLBP confirmed by CT-discography whounderwent 1 of 4 fusiontechniques.

Not indicated

NOTE. Type I discogram designated as internal disc disruption (IDD) withI and III denote the presence of annular disruption with spread of contras

Abbreviations: DDD, degenerative disc disease; ESF, external spinal fix

haft of an electric toothbrush) against the skin 5

verlying successive spinous processes to provokeain, has been purported to be a quick, safe, andainless method for diagnosing discogenic pain. In atudy by Yrjama and Vanharanta106 conducted in

perative Discography on Surgical Outcomes

Results Comments

50% of pts with type Idiscogram and normal MRIscans improved vs 75% ofpts with types II or IIIdiscogram and abnormalMRI.

Abnormal discogram was thebasis for surgery. Averagefollow-up 20 months.

Of 137 pts in whomdiscogram revealed DDDand provoked concordantpain, 89% had favorableoutcome. Only 52% ofthose pts in whom discogrshowed DDD but provokedno pain had a favorableoutcome.

Mean follow-up 3.6 years.

39% of pts reported goodoutcomes, 13% fairoutcomes, and 48% hadpoor results.

Abnormal discogram wasbasis for surgery. Meanfollow-up 47 months.

Among the 15 pts withconcordant pain ondiscography, 13experienced significant orcomplete pain relief withESF. 6 of these 10 pts hadsignificant relief afterarthrodesis.

Study not designed toevaluate predictive valueof discography on surgicalsuccess. Not all ptsunderwent preoperativediscography.

Pre-operative discographyimproved success ratefrom 65%-70% to 85%.

Used L5-S1 discography todetermine whetheranterior or posteriorinstrumentation should beused.

At first follow-up (mean 5.3wks), 66% had satisfactoryoutcome. At final follow-up(mean 35 months), 46%had satisfactory outcome.

CT-discography used in allbut 1 pt. Not all pts had acontrol disc (26 pts hadsingle-level discography).

81% of pts who had surgerybased on MRI and clinicalfindings had satisfactoryoutcome vs. 76% of ptswho underwent arthrodesisbased on (�) discogram.

Mean follow-up 2.4 years indiscography group and 2.8years in MRI/clinicalgroup.

Poor results in all 5 pts with2-level fusions. In single-level fusions, 35% of ptshad good results, 18% fair,and 47% poor outcomes.

Strong correlation betweensubjective (clinicalimprovement) andobjective (fusion success)results.

In pts with chemicallysensitized discs (�6/10concordant pain @ � 15psi above openingpressure, n � 36), successrates were 89% forinterbody/combined fusion,20% for posterior inter-transverse fusion and 12%for no surgical rx.

Mean follow-up for surgicalpts 28 months. Nodifference betweenoutcomes for interbody/combined fusion andposterior intertransversefusion for surgical sampleas a whole.

Overall rate of pt satisfaction46%.

Success rate for pts who hadanterior lumbar fusion withcage and facet fusion63%. Success rates forthe other 3 groups rangedfrom 36% to 46%.

avasation of contrast associated with concordant pain reproduction. Typesperiphery and epidural space, respectively.D, internal disc disruption; LBP, low back pain; pts, patients.

f Preo

dy

tudyS1

l

eingand/oron.dyfect ofspinals alsore-op

tudy

tudy

tudy

tudy

out extr

7 patients with LBP, the authors found the sensi-

Page 11: Discography Review

trpd�iBaiacig

PBspL

LA

io

ut pati

Lumbar Discography • Cohen et al. 173

ivity and specificity of BVT to be 0.71 and 0.63,espectively, compared with provocative discogra-hy. When patients with failed back surgery syn-rome and painful herniated discs were excluded (n

40), the sensitivity jumped to 0.96 and the spec-ficity to 0.72. In 2 follow-up studies comparingVT findings with MRI and CT-discography, theuthors found similarly high sensitivities and spec-ficities when failed back surgery syndrome patientsnd those with complete annular tears were ex-luded.107,108 The main problem with these studiess that they fail to distinguish patients with disco-

Table 6. Summary of Outcome Data forPreoperative Di

Study, YearPreoperativeDiscography?

Number of DisReplacement

Patients

Blumenthal, 200391 Yes 57

Shim, 200392 In pts with DDDat more than 1level.

46

Van Ooij, 200393 Yes 27

Tropiano, 200394 No 53

McAfee, 200395 Yes 41

Kim, 200396 Not mentionedexcept for (�)discogrambeing acontraindication.

11 pts withjuxtafusional

DDD

Zeegers, 199997 In 36 of 50 pts. 50

Bertagnoli, 200298 No 108

Hochshuler, 200299 Yes 56

Mayer, 2002100 No 34

Jin, 2003101 No 45

Zigler, 2003102 Not routinely 28

Enker, 1993103 Yes 6

NOTE. Table does not include studies lacking information aboAbbreviation: f/u, follow-up; pts, patients.

enic pain from those with radicular symptoms. v

resently, there are no published articles correlatingVT findings with surgical outcomes. Until thesetudies are conducted, it is unlikely BVT will sup-lant discography as a diagnostic tool for discogenicBP.

iterature Review on the Technicalspects of Discography

Indications and Pain Referral Patterns. Thendications for discography have been previouslyutlined in numerous guidelines and re-

ar Disc Replacement Surgery Based onphy Screening

Type ofSurgery Outcomes

otal discreplacement

63% of pts improved at 2-year follow-up(based on � 20 pts improvement onVAS score).

artial discreplacement

Mean VAS score 8.5 pre-op, 3.1 at 1-yearfollow-up. 11% had excellent and 67%good results.

otal Good outcome obtained in 12 of 26 pts,with variable f/u period (range 1 month-10 years).

otal 87% of pts “entirely satisfied”. MeanOswestry disability score 56 pre-op, 14at last f/u (mean 1.4 years).

otal Mean VAS score 73.5 pre-operatively,and 30.4 at 1–3 year f/u.

otal Of the 5 pts followed for � 6 months,mean Oswestry disability indexdecreased from 64% to 24%.

otal 32 of 46 pts followed for 2 years had apositive clinical result. Did not provideseparate data for pts havingdiscography.

otal Results “excellent” in 91%, “good” in 7%,“fair” in 2%, and poor in no (0%)patients at 3-month to 2-year follow-up.

otal 52.7% improvement in mean VAS scoresat 6-week follow-up. In the 22 ptsfollowed for � 1 year, improvements inVAS and Oswestry scores weremaintained.

otal Mean VAS score decreased from 6.3 pre-op to 3.4 at 1-year follow-up (not all ptsfollowed for 1 yr). 61% of pts“completely satisfied,” and 22%“satisfied.”

artial Mean Oswestry disability index decreasedfrom 52.2% to 16.5% in the 30 ptsseen at their 6-month f/u. 87% of ptswere clinically improved.

otal Decrease in VAS score fromapproximately 7.8 to 5.6 after 6months.

otal 4 of 6 patients had satisfactory results(1 excellent, 2 good, 1 fair).

ent selection criteria.

Lumbscogra

c

T

P

T

T

T

T

T

T

T

T

P

T

T

iews.56,105,109-118 These include the following:

Page 12: Discography Review

gtttepwtd(detrwic

peifrtn

wchrtpfo

csrcdtttr

atdtvtfi

rgcimpu(pmp

crStesetcimpi

t

174 Regional Anesthesia and Pain Medicine Vol. 30 No. 2 March–April 2005

1. Evaluation of patients with unremitting spinalpain, with or without extremity pain, ofgreater than 4 months duration that has beenunresponsive to conservative therapy.

2. Patients with persistent symptoms in whomother diagnostic tests have failed to revealclear confirmation of a suspected disc as thesource of pain.

3. Evaluation of persistent pain in the postoper-ative patient whose symptoms may be arisingfrom intervertebral disc degeneration, recur-rent herniation, or pseudoarthrosis of a spinalfusion.

4. To determine the number of levels to be fusedin spinal surgery and/or determination of theprimary symptomatic disc.

5. If the patient prima facie satisfies the criteriafor treatment by intradiscal electrothermaltherapy, in which case pain is provoked todetect discogenic pain. In these cases, CT dis-cography may be undertaken to assess discmorphology to determine whether or not an(IDET) electrode can be navigated in the discand, if so, where it should be placed.

6. For assessment of minimally invasive surgicalcandidates to confirm contained disc hernia-tions or to investigate dye distribution beforechemonucleolysis or other percutaneous de-compression procedures.

Current International Spinal Injection Societyuidelines recommend that 2 control levels be ob-ained for adequate interpretation of disc stimula-ion.56 However, many authors do not adhere tohese strict guidelines.105,110,119 Recently, Ohnmeisst al.120 described patterns of pain provocation forositive discograms. For L3-4 injections, discogramsere more likely to be positive if patients described

heir pain as involving the lumbar region with ra-iation into the anterior but not posterior thigh71.4%). For L4-5 discs, the most common painescription was that of lumbar pain with morequivalent proportions of anterior and posteriorhigh pain (�63%). In L5-S1 discogenic pain, theeferral pattern tended to involve the lumbar regionith posterior thigh or leg pain (�75%). Pain lim-

ted to the low back and buttocks was usually asso-iated with the absence of disc pathology (58.3%).There are numerous ways to perform discogra-

hy, and no one method for needle placement hasver been shown to be superior. Interspinous andnterlaminar approaches have been advocated afterusion with paraspinal bone graft but are otherwisearely used because of the necessity of dural pene-ration. At L5-S1, a modified double-needle tech-

ique is often necessary because of the iliac crests, t

hich prohibit a trajectory aimed directly at theenter of the disk. Entry into this level is subject toigher failure rates, which may be substantiallyeduced by using a curved-needle technique.121 Al-hough recommended by some,56,105,122 an ap-roach from the side contralateral to a painful re-erral pattern has not been shown to affect the ratef false-positive discograms.123

Because of the possibility of inadvertent intrathe-al administration, only nonionic contrast materialhould be injected. Nonionic dye also minimizes theisk of allergic reaction, which can be delayed untilontrast material is systemically absorbed from theisc. In patients with known prior contrast reac-ions, premedication is necessary. Another option ishe use of gadolinium as the contrast agent.124 Forhe technical aspects of discography, readers areeferred to other reviews.56,105,110,119,121,122,125,126

Disc Stimulation and Interpretation. Oncell needles have been properly positioned, the pa-ient is prepared for the injections. The sequence ofisc injection depends on the discographer’s expec-ation for pain provocation at each level. Once se-ere pain is produced, the patient may be less ableo tolerate or judge subsequent injections. There-ore, the level considered likely to cause the mostntense pain should be injected last.

Normal discs take �1 mL of contrast before firmesistance is reached (high-pressure endpoint). De-enerated discs generally accept larger volumes ofontrast and show moderate or soft resistance tonjection (low-pressure endpoint). If the disc com-

unicates with the epidural space because of com-lete annular disruption, unlimited contrast vol-me can be injected with little to no resistancevolume endpoint). When severe symptoms arerovoked, the injection must be stopped even ifinimal contrast has been injected (pain end-

oint).127

Injection pressures can be measured using spe-ialized, commercially available devices, and areecommended by the International Spinal Injectionociety.56 Others have not endorsed the use ofhese devices,110,118,125 claiming that an experi-nced discographer can determine relative pres-ures using only a syringe. Proponents of manom-try state that a pressure recording device allows forhe objective designation of painful discs as chemi-ally or mechanically sensitive.84,87,112 Although its true that manometry can provide an objective

easure of disc sensitivity, it does not ensure that aatient will respond promptly and accurately to discnjection.

If manometry is used, one should first determinehe opening pressure of the disc, which in practical

erms is the pressure at which contrast is first visu-
Page 13: Discography Review

asmahpva5pa�iacasmPbpern

hsslttor

dtwcrc

pcftsudaltnwptc

lstd

bctissaps

ddaftdtdsqu

Lumbar Discography • Cohen et al. 175

lized entering the disc (usually �15 pounds perquare inch [PSI]). Once opening pressure is deter-ined, the disc is slowly filled with contrast, in

liquots of 0.5 to 1 mL. The endpoint is usually aigh-pressure endpoint at around 100 PSI or theoint at which pain is experienced. In cases of se-erely degenerative discs, it may not be possible tochieve pressures above 50 PSI. If either 100 PSI or0 PSI above opening pressure is reached withoutain being noted, the disc is considered negativend can serve as a control. Generating pressures100 PSI greatly increases the risk of iatrogenic

njury. If pain is noted at low pressures (�15 PSIbove opening pressure), the disc is consideredhemically sensitive. If pain is noted between 15nd 50 PSI above opening pressure, the disc is con-idered mechanically sensitive and the discogramay or may not be positive. If pain is noted at �50

SI above opening pressure, “the response cannote considered clinically significant.”56 Pain at higherressures may be caused by mechanical irritation,nd-plate deflection, or the stimulation of pressureeceptors. There are no studies assessing whether orot the use of manometry affects surgical outcomes.Numerous discography classification systems

ave been advocated, both with and without pres-ure monitoring. No single set of criteria is univer-ally accepted as standard (Tables 4 and 7). In aarge, multicenter study by Derby et al.,84 the au-hors showed that patients with chemically sensi-ized discs (i.e., concordant pain at �15 PSI abovepening pressure) were more likely to obtain goodesults with interbody fusion surgery.

Some authors have advocated videotaping toocument patient reactions during disc injec-ions.110,126 Videotaped findings are later correlatedith fluoroscopy or compared with simultaneous

hanges in physiological parameters such as heartate and blood pressure. If relevant, the videotapean then be made available to the referring surgeon.Image Interpretation. For morphologic inter-

retation of the fluoroscopic images, although dis-ography is clearly inferior to CT-discography, use-ul information may nevertheless be garnered fromhe overall appearance of the contrast spread. Amall central “cotton ball” accumulation of contrastsually indicates a normal painless disc. Normaliscs may also appear to have a central bilobularppearance, like a “hamburger in a bun.” A thinine of contrast that extends to the posterior edge ofhe disc is consistent with an annular tear, and it isot unusual for the patients to complain of painhen the contrast moves posteriorly. More diffuseosterior contrast spread indicates a large annularear. Spread of contrast into the epidural space is

onsistent with a complete tear of the annulus and g

oss of integrity of the outer annular wall. Diffusepread in all directions, with or without spread intohe epidural space, indicates a severely degeneratedisc.Fluoroscopy can show morphological differences

etween normal and degenerated discs but poorlyharacterizes the locations of annular tears andheir relationships to nerve roots. Disc architectures much better assessed by CT because of its cross-ectional acquisition of thin slices parallel to the discpaces. Thus, many annular tears seen on CT im-ges are invisible at fluoroscopy. For this reason, weerform procedures requiring a functional annulusuch as nucleoplasty only after CT discography.128

The most commonly used terminology of discisease was adapted for MRI. This terminology isifficult to apply to CT discography because it onlyddresses contour abnormalities, such as diffuse orocal disc bulge, protrusion, extrusion, and seques-ration. Internal disc abnormalities that cannot beetected on MR become obvious on CT becausehey are filled with contrast material. The Dallasiscogram scale was originally developed to de-cribe and grade annular tears and has subse-uently undergone several modifications.110,127 Bysing this modified scheme, disc morphology is

Table 7. Interpretation of Discograms

Unequivocal discogenic pain1. Stimulation of the target disc reproduces concordant pain.2. Pain intensity is rated as being at least 7 on a 10-point

visual analog scale.3. Pain is reproduced at a pressure of less than 15 PSI

above opening pressure.4. Presence of 2 adjacent control discs that are nonpainful.

Definite discogenic pain1. Stimulation of target disc reproduces concordant pain.2. Pain intensity is rated as being at least 7 on a 10-point

visual analog scale.3. Pain is reproduced at a pressure of less than 15 PSI

above opening pressure.4. Presence of 1 adjacent control disc that is nonpainful.

or1. Stimulation of the target disc reproduces concordant pain.2. Pain intensity is rated as being at least 7 on a 10-point

visual analog scale.3. Pain is reproduced at a pressure of less than 50 PSI

above opening pressure.4. Presence of 2 adjacent control discs that are nonpainful.

Probable discogenic pain1. Stimulation of the target disc reproduces concordant pain.2. Pain intensity is rated as being at least 7 on a 10-point

visual analog scale.3. Pain is reproduced at a pressure of less than 50 PSI

above opening pressure.4. Presence of 1 adjacent control disc that is nonpainful,

and a second adjacent control disc that producesnonconcordant pain at a pressure of greater than 50PSI.

Abbreviation: PSI, pounds per square inch.Adapted from Endres et al.56

raded from 0 to 7 (Table 8 and Fig 2).110

Page 14: Discography Review

rmdvpcepssm

C

t

oweweraetsmc

t0dmbaAdajpsdp

uhdctcd

T

TT

T

T

T

T

Ffif

176 Regional Anesthesia and Pain Medicine Vol. 30 No. 2 March–April 2005

The final interpretation of lumbar discographyeflects the synthesis of disc stimulation data andorphology. Only rarely is pain reproduced in a

isc that is morphologically, manometrically, andolumetrically normal. In these cases, the discogra-her must carefully consider the patient’s entirelinical picture, including historical and physicalxamination findings, the presence of spinal andsychological pathology, additional radiologic andupplemental diagnostic tests, and the technicaluccess of the discogram before deciding on a treat-ent course.

omplications

Because of the poor blood supply of the interver-ebral discs, discitis is the most feared complication

Table 8. CT Class

ype 1: The discogram is normal manometrically, volumetricallycontrast to be centrally located in both axial and sagittal pro

ype 2: Identical to type 1 except it is positive for pain reproducype 3: The annular tears lead to a radial fissure. This group caType 3a: The radial fissure is posteriorType 3b: The fissure radiates posterolaterally andType 3c: The fissure extends lateral to a line drawn from the c

articulating process.ype 4: When the radial fissure reaches the periphery of the an

to bulge.ype 5: When the outer annular fibers rupture, nuclear material

contacting either the dura or a nerve root.ype 6: When the extruded fragment is no longer in continuity w

volumetrically, and radiologically, these discograms are alwpressure is generated against the free fragment to cause st

ype 7: The end stage of the degenerative process is internal dare abnormal manometrically and volumetrically, and familiausually fills the entire interspace in a chaotic fashion. The Cannular tears.

Adapted from Bernard Jr et al.110 and Sachs et al.127

ig 2. Modified Dallas discogram scheme for the classi-cation of annular tears by CT discography. (Modified

rrom Sachs et al.127 Drawings by Jee Hyun Kim.)

f discography. Any patient who complains oforsening pain 1 week postprocedure warrants re-

valuation. At minimum, the postdiscographyorkup should include a focused history, physical

xamination, and laboratory screening tests. Eryth-ocyte sedimentation rate and C-reactive proteinre the most sensitive indicators of discitis, but el-vation usually does not occur until 3 weeks afterhe procedure.129,130 If the ESR is �50, then a bonecan or MRI is indicated to rule out discitis. Theost common etiologic agent in postprocedure dis-

itis is Staphylococcus aureus.130

In a study by Fraser et al.,131 the authors foundhe incidence of discitis was reduced from 2.7% to.7% by using a through-and-through double-nee-le technique. In a review by Guyer and Ohn-eiss,132 the authors found an incidence of discitis

etween 0.1% and 0.2%, with most of the studiesnalyzed not administering prophylactic antibiotics.fter a 2-part study assessing the efficacy of intra-iscal cefazolin in sheep and humans, Osti et al.133

dvocated mixing 1 mg/mL cefazolin with the in-ected contrast as a safe and effective means ofreventing discitis. A recent in vitro study by Kles-ig et al.134 supports the prophylactic use of intra-iscal antibiotics in lieu of systemic therapy foratients undergoing provocative discography.The prophylactic use of antibiotics is by no means

niversal. Willems et al.135 conducted a compre-ensive review of the incidence of postdiscographyiscitis without the use of prophylactic antibiotics inonjunction with 200 of their own patients. Amonghe 4,981 patients included in the analysis, the in-idence of discitis was 0.25% (0.09% of 12,770iscograms). The authors concluded that the low

n of Discography

ogically, and produced no pain. The CT discogram showeds.

rther subdivided into:

of the disk tangential to the lateral border of the superior

brosus, nuclear material protrudes causing the outer annulus

trude beneath the posterior longitudinal ligament directly

interspace it is said to be sequestrated. Manometrically,normal. Concordant pain may be reproduced only if enoughon of pain-sensitive structures.uption, in which multiple annular tears occur. The discogramsmay or may not be reproduced. Radiologically, the contrastogram shows contrast extravasation throughout multiple

ificatio

, radioljectiontion.n be fu

enter

nulus fi

may ex

ith theays abimulatiisc disrr painT-disc

isk of discitis when the double-needle technique is

Page 15: Discography Review

ultuihd(

iwbe0ansptFvhatscbvILprct

pevtpToa4pailpcstjpti

C

cvTddplpsdrsbys

F

cwTannsw

a

t

p

bA

f

oos

Lumbar Discography • Cohen et al. 177

sed does not warrant the routine use of prophy-actic antibiotics. Considering the inherent difficul-ies in treating discitis and low doses of antibioticssed, the authors believe the use of prophylactic

ntradiscal antibiotics is justified. Collectively, weave performed over 2,000 discograms using intra-iscal antibiotics, with a 0% incidence of discitisunpublished data, Cohen and Larkin, 1999-2004).

Can discography cause or worsen existing LBP ornjure the intervertebral disc? In a study assessinghether discography is associated with long-termack symptoms in asymptomatic subjects, Carrageet al.136 found the incidence of persistent LBP to be% in patients without chronic neck pain or som-tization disorder, 20% in patients with chroniceck and shoulder pain, and 66% in patients withomatization disorder 1 year after lumbar discogra-hy. However, in 2 other studies assessing the long-erm effect of discography, both Johnson137 andlanagan and Chung138 found no evidence of inter-ertebral disc injury. The results of Carragee et al.136

ave not been supported by subsequent studies,nd it is likely their findings are germane only tohe subpopulations they tested. In a follow-uptudy by Carragee et al.,139 the authors found psy-hological distress and preexisting chronic pain toe strong predictors of future LBP in asymptomaticolunteers undergoing experimental discography.n contrast, painful disc injection did not predictBP or any other functional outcome measure. Asreviously alluded to, persistent pain after discog-aphy can result from high pressures causing mi-roinjury to the disc(s), which may be more likelyo occur without the use of manometry.

In a study analyzing 146 lumbar discograms in 52atients, Tallroth et al.140 found that 2% of patientsxperienced nausea, 4% convulsions, and 6% se-ere back pain during the procedures. A day afterhe procedure was performed, 10% of patients re-orted a severe headache and 81% worsening LBP.he headaches were attributed to neuraxial leakagef contrast. In a large, retrospective study assessingdverse outcomes in 1,357 patients undergoing,400 cervical discograms, Zeidman et al.141 re-orted complications occurring in 0.6% of patientsnd 0.16% of disc injections. The complicationsncluded 7 cases of discitis and 1 abscess. Prophy-actic antibiotics were only administered to thoseatients at high risk for infection. Other possibleomplications of discography include meningitis,pinal headache, subdural or epidural abscess, in-rathecal hemorrhage, arachnoiditis, nerve root in-ury, paravertebral muscle pain and contusions,ostprocedural pain exacerbation, vasovagal reac-ions, allergic reactions, and damage to the disc

ncluding but not limited to herniation.54,105,142-145 d

ost Analysis of Lumbar Discography

To accurately assess the cost-effectiveness of dis-ography, one needs to account for a multitude ofariables that inevitably play into the calculation.hese include but are not limited to the screening ofiscography patients to eliminate patients withoutiscogenic LBP and those at high risk for false-ositive results; the rate of positive discography; theikelihood patients with positive discograms willroceed to surgery; the type and success rate of theurgical procedure done in patients with positiveiscograms; and the total cost of discography thateflects the number of discs injected, the use of CTcanning, and whether or not sedation is used. Ta-le 9 represents a very simplified cost-benefit anal-sis of lumbar discography as a screening test forpinal fusion.

uture Areas of Investigation

Wherein lies the future of clinical research on dis-ography? The main issue that needs to be resolved ishat effect discography has on surgical outcomes.his includes not only open surgical procedures butlso minimally invasive procedures such as IDET anducleoplasty. Once this question is answered, theext step would be to determine how best to improvepecificity without compromising sensitivity. Oneay to accomplish this would be to conduct a ran-

Table 9. Cost-Benefit Analysis of LumbarDiscography as a Screening Test for Spinal Fusion

1. Average cost of 3-level lumbar discography withoutcomputed tomography or conscious sedation—$1,140

2. Average cost of lumbar spinal fusion surgery—$33,8003. Total cost of discography plus spinal fusion 100 patients with

suspected discogenic pain and an expected positivediscographic confirmation rate of 83%—($33,800 � 83) �($1140 � 100) � $2,919,400

4. Total cost of the same 100 patients proceeding directly tosurgery—($33,800 � 100) � $3,380,000

5. Total savings by using lumbar discography as a presurgicalscreening tool—($3,380,000 � $2,919,400)/100 � $460,000,or $4,600 per patient.

NOTE. This analysis is based on the following assumptions:1. All patients presenting for discography have abnormal MRIs

nd would otherwise undergo spinal fusion.2. All patients have failed more conservative efforts to treat

heir pain.3. Disc stimulation is performed according to the standards

resented in Table 4.4. The rate of positive confirmatory disc stimulation is 83%,

ased on a previously published study conducted at Walter Reedrmy Medical Center.123

5. All patients with positive discography will undergo spinalusion.

6. The costs of discography and lumbar spine fusion are basedn the reimbursement rates Walter Reed receives from the Dept.f Veterans Affairs, and loosely resembles the DRG paymentchedule of TRICARE (Fall 2004).

omized, prospective study comparing surgical out-

Page 16: Discography Review

cdArtshsotrfca

ittr(wopjpasmdart(csio2arpttmj1mt

C

adts

rhrtatpctc

eaeppc

lsloerspraemui

tLmpbtispwatdoeihipi

178 Regional Anesthesia and Pain Medicine Vol. 30 No. 2 March–April 2005

omes in patients who have undergone preoperativeiscography screening with and without manometry.nother might be to determine whether or not cor-

elating discographic findings with behavioral charac-eristics (e.g., facial expressions and withdrawal re-ponse) and/or physiological parameters (change ineart rate) has any effect on accuracy. Behavioraligns have been previously reported to enhance thebjectivity of the physical examination in LBP pa-ients.146,147 Although previously used during discog-aphy,50,110,126 their effect on accuracy has not beenormally studied. The association between heart ratehanges and pain response during disc stimulation isn area we are currently investigating.Another promising area of discography research

nvolves the injection of analgesic substances intohe disc, thereby rendering the procedure therapeu-ic. Recent animal studies have shown that nerveoots exposed to exogenous tumor necrosis factor-�TNF-�) mimic the histologic changes observedith herniated discs43 and that the administrationf intramuscular or intravenous anti–TNF-� canrevent nucleus pulposus-induced spinal nerve in-ury.148 In a prospective open-label study by Karp-inen et al.,149 the authors infused 2 mg/kg of thenti–TNF-� agent infliximab in 10 patients withevere sciatica secondary to disc herniation. After 3onths, 90% of the study group reported �75%

ecrease from baseline pain. TNF-� inhibitors maylso prove effective in discogenic back pain. In aetrospective review of 20 patients who receivedhe TNF-� antagonist etancercept subcutaneouslyaverage number of doses 1.8, range 1-5) for cervi-al or lumbar discogenic pain, Tobinick and Brit-chgi-Davoodifar150 reported a significant decreasen the Oswestry Disability Index, from a mean scoref 54.5 at baseline to 9.8 at latest follow-up (mean30 days). A larger follow-up study by the sameuthors conducted in patients with both axial andadicular low back and neck pain yielded similarlyromising results.151 The doses used in these clinicalrials were the same as those used to treat rheuma-oid arthritis and inflammatory bowel disease anday be associated with significant side effects. In-

ecting these drugs intradiscally at 1/50th or/100th of the systemic dose would be expected toinimize the incidence of side effects and maximize

he benefit.

onclusions

So where do we stand with discography, and whatre its prospects for the future? Although it is true thatiscography, with or without CT, is the only imagingool that allows a clinician to relate pathology with

ymptoms, the significance of this remains unclear. In h

egards to symptomatic disc herniations, the en-anced accuracy and safety of MRI has made discog-aphy an archaic procedure for this purpose. For in-ernal disc disruption, CT-discography is a moreccurate diagnostic tool than plain MRI,2,60,69,72 al-hough T2-weighted MR images may detect someathology missed on plain discograms.57 Based onadaver studies, discography seems to be more sensi-ive for detecting radial rather than transverse or con-entric tears in the AF.63,64

When assessing discographic pathology, it is nec-ssary to correlate provoked pain with anatomicbnormalities. Disc stimulation should always bevaluated in the context of the high risk for false-ositive discograms in some patients.52,53 In suchatients, we would recommend injecting 2 adjacentontrol discs.The critical element for the future of discography

ies in its ability to improve patient outcomes as acreening test, and presently the evidence for this isacking. In the 2 studies directly comparing fusionutcomes between patients who underwent preop-rative discography and those who had not, theesults are conflicting.4,6 For IDET, all publishedtudies have routinely used preoperative discogra-hy so that no conclusions can be drawn. With disceplacement surgery, although the studies are splits to whether discography was used during preop-rative assessment, the differences in outcomeeasures, length of follow-up, surgical techniquesed, and methodological flaws render any compar-

sons futile.What is clear is that there is a wide discrepancy in

he use of lumbar discography to treat discogenicBP. The routine use of discography as a screeningeasure before IDET, a minimally invasive surgical

rocedure, and the sporadic use of the procedureefore spinal arthrodesis is inconsistent. The au-hors speculate that given the low risk and uniquenformation discography provides, the procedurehould be cautiously used before invasive surgicalrocedures designed to treat discogenic LBP excepthen the evidence implicating a particular disc(s)

s the pain generator is overwhelming. However,his recommendation presumes that the results ofisc stimulation must be considered in the contextf other diagnostic screening tests and with consid-ration to the propensity of patients with preexist-ng psychopathology and somatization symptoms toave false-positive pain provocation. Clinical stud-

es are needed to better elucidate the role discogra-hy will assume in the diagnosis of disc pathologyn the future and to determine what effect, if any, it

as on the surgical treatment of discogenic pain.
Page 17: Discography Review

Lumbar Discography • Cohen et al. 179

References

1. Lindblom K. Diagnostic puncture of intervertebraldisks in sciatica. Acta Orthop Scand 1948;17:231-239.

2. Brightbill TC, Pile N, Eichelberger RP, Whitman MJr. Normal magnetic resonance imaging and abnor-mal discography in lumbar disc disruption. Spine1994;19:1075-1077.

3. Adams MA, Dolan P, Hutton WC. The stages of discdegeneration as revealed by discograms. J Bone JointSurg (Br) 1986;68:36-41.

4. Colhoun E, McCall IW, Williams L, Cassar PullicinoVN. Provocation discography as a guide to planningoperations on the spine. J Bone Joint Surg (Br) 1988;70:267-271.

5. Whitecloud TS 3rd, Seago RA. Cervical discogenicsyndrome. Results of operative intervention pa-tients with discography. Spine 1987;12:313-316.

6. Madan S, Gundanna M, Harley JM, Boeree NR,Sampson M. Does provocative discography screen-ing of discogenic back pain improve surgical out-comes? J Spinal Disord Tech 2002;15:245-251.

7. Wetzel FT, LaRocca SH, Lowery GL, Aprill CN. Thetreatment of lumbar spinal pain syndromes diag-nosed by discography. Lumbar arthrodesis. Spine1994;19:792-800.

8. Bogduk N. Clinical Anatomy of the Lumbar Spine andSacrum. 3rd ed. Edinburgh, Scotland: Churchill Liv-ingstone; 1997:205-212.

9. Holm S, Maroudas A, Urban JP, Selstam G,Nachemson A. Nutrition of the intervertebral disc:Solute transport and metabolism. Connect Tissue Res1981;8:101-119.

10. Ohshima H, Urban JP. The effect of lactate and pHon proteoglycan and protein synthesis rates in theintervertebral disc. Spine 1992;17:1079-1082.

11. Aoki Y, Takahashi Y, Ohtori S, Moriya H, TakahashiK. Distribution and immunocytochemical charac-terization of dorsal root ganglion neurons innervat-ing the lumbar intervertebral disc in rats: A review.Life Sci 2004;74:2627-2642.

12. Nakamura S, Takahashi K, Takahashi Y, YamagataM, Moriya H. The afferent pathways of discogeniclow-back pain. J Bone Joint Surg (Br) 1996;78:606-612.

13. Groen GJ, Baljet B, Drukker J. Nerves and nerveplexuses of the human vertebral column. Am J Anat1990;188:282-296.

14. Malinsky J. The ontogenetic development of nerveterminations in the intervertebral discs of man. ActaAnat 1959;38:96-113.

15. Freemont AJ, Peacock TE, Goupille P, Hoyland JA,O’Brien J, Jayson MI. Nerve ingrowth into diseasedintervertebral disc in chronic back pain. Lancet 1997;350:178-181.

16. Wiberg G. Back pain in relation to nerve supply ofthe intervertebral disc. Acta Orthop Scand 1947;19:211-221.

17. Ikari C. A study of the mechanism of low back pain.The neurohistological examination of the disease.

J Bone Joint Surg 1954;36A:195.

18. Coppes MH, Marani E, Thomeer RT, Oudega M,Groen GJ. Innervation of annulus fibrosis in lowback pain. Lancet 1990;336:189-190.

19. Coppes MH, Marani E, Thomeer RT, Groen GJ.Innervation of “painful” lumbar discs. Spine 1997;22:2342-2349.

20. Ehrenhaft JL. Development of the vertebral columnas related to certain congenital and pathologicalchanges. Surg Gynecol Obstet 1943;76:282-292.

21. Hirsch C, Ingelmark BE, Miller M. The anatomicalbasis for low back pain. Acta Orthop Scand 1963;33:1-17.

22. Jackson HC, Winkelmann RK, Bickel WM. Nerveendings in the human lumbar spinal column andrelated structures. J Bone Joint Surg 1966;48A:1271-1281.

23. Adams MA, McNally DS, Dolan P. ‘Stress’ distribu-tions inside intervertebral discs. The effects of ageand degeneration. J Bone Joint Surg 1996;78B:965-972.

24. Adams MA, McNally DS, Wagstaff J, Goodship AE.Abnormal stress concentrations in the lumbar inter-vertebral discs following damage to the vertebralbodies: Cause of disc failure? Eur Spine J 1993;1:214-221.

25. Boos N, Weissbach S, Rohrbach H, Weiler C, SprattKF, Nerlich AG. Classification of age-related changesin lumbar intervertebral discs. Spine 2002;27:2631-2644.

26. Yoganandan N, Maiman DJ, Pintar F, Ray G,Myklebust JB, Sances A Jr, Larson SJ. Micro-trauma in the lumbar spine: A cause of low backpain. Neurosurgery 1988;23:162-168.

27. Heggeness MH, Doherty BJ. Discography causes endplate deflection. Spine 1993;18:1050-1053.

28. Modic MT, Masaryk TJ, Ross JS, et al. Imaging ofdegenerative disk disease. Radiology 1988;168:177-186.

29. Frobin W, Brinckmann P, Kramer M, Hartwig E.Height of lumbar discs measured from radiographscompared with degeneration and height classifiedfrom MR images. Eur Radiol 2001;11:263-269.

30. Vernon-Roberts B. Age-related and degenerativepathology of intervertebral discs and apophysealjoints. In: Jayson MV, ed. The Lumbar Spine and BackPain. Edinburgh, Scotland: Churchill Livingstone;1992:17-41.

31. Kang JD, Georgescu HI, McIntyre-Larkin L,Stefanovic-Racic M, Donaldson WF 3rd, Evans CH.Herniated lumbar intervertebral discs spontane-ously produce matrix metalloproteinases, nitric ox-ide, interleukin-6, and prostaglandin E2. Spine1996;21:271-277.

32. Bogduk N, McGuirk B. Pain Research and ClinicalManagement. Medical Management of Acute and ChronicLow Back Pain. An Evidence Based Approach. Vol 13.Amsterdam, Holland: Elsevier Science BV; 2002:119-122.

33. Konttinen YT, Gronblad M, Antti-Poika I, Seitsalo S,

Santavirta S, Hukkanen M, Polak JM. Neuroimmu-
Page 18: Discography Review

180 Regional Anesthesia and Pain Medicine Vol. 30 No. 2 March–April 2005

nohistochemical analysis of peridiscal nociceptiveneural elements. Spine 1990;15:383-386.

34. Ashton IK, Roberts S, Jaffray DC, Polak JM,Eisenstein SM. Neuropeptides in the human in-tervertebral disc. J Orthop Res 1994;12:186-192.

35. Burke JG, Watson RW, McCormack D, Dowling FE,Walsh MG, Fitzpatrick JM. Intervertebral discswhich cause low back pain secrete high levels ofproinflammatory mediators. J Bone Joint Surg 2002;84B:196-201.

36. Lu GH, Li J, Zhou JN. Inflammatory-inducing effectof lumbar disc tissue: An experimental study (inChinese). Hunan Yi Ke Da Xue Xue Bao 2001;26:531-533.

37. Tanaka N, Ishida T, Hukuda S, Horiike K. Purifica-tion of a low-molecular-weight phospholipase A(2)associated with soluble high-molecular-weightacidic proteins from rabbit nucleus pulposus and itscomparison with a rabbit splenic group IIa phospho-lipase A(2). J Biochem (Tokyo) 2000;127:985-991.

38. Miyahara K, Ishida T, Hukuda S, Horiike K,Okamoto M, Tojo H. Human group II phospholipaseA2 in normal and diseased intervertebral discs. Bio-chim Biophys Acta 1996;1316:183-190.

39. Saal JS, Franson RC, Dobrow R, Saal JA, White AH,Goldthwaite N. High levels of inflammatory phos-pholipase A2 activity in lumbar disc herniations.Spine 1990;15:674-678.

40. Franson RC, Saal JS, Saal JA. Human disc phospho-lipase A2 is inflammatory. Spine 1992;17(6 suppl):S129-132.

41. Nygaard OP, Mellgren SI, Osterud B. The inflamma-tory properties of contained and noncontained lum-bar disc herniation. Spine 1997;22:2484-2488.

42. Olmarker K, Rydevik B. Selective inhibition of tu-mor necrosis factor-alpha prevents nucleus pulpo-sus-induced thrombus formation, intraneuraledema, and reduction of nerve conduction velocity:Possible implications for future pharmacologic treat-ment strategies of sciatica. Spine 2001;26:863-869.

43. Igarashi T, Kikuchi S, Shubayev V, Myers RR. Ex-ogenous tumor necrosis factor-alpha mimics nu-cleus pulposus-induced neuropathology. Molecular,histologic, and behavioral comparisons in rats. Spine2000;25:2975-2980.

44. Jackson RP, Becker GJ, Jacobs RR, Montesano PX,Cooper BR, McManus GE. The neuroradiographicdiagnosis of lumbar herniated nucleus pulposus: I. Acomparison of computed tomography (CT), CT-my-elography, discography, and CT-discography. Spine1989;14:1356-1361.

45. Yasuma T, Ohno R, Yamauchi Y. False-negative dis-cograms. J Bone Joint Surg 1988;70A:1279-1290.

46. Yasuma T, Makino E, Saito S. A significance of thelimited form of discography—A comparison withhistological findings of the lumbar discs in the au-topsy cases (in Japanese). Seikeigeka 1988;39:311-318.

47. Yussen PS, Swartz JD. The acute lumbar disc herni-ation: Imaging diagnosis. Semin Ultrasound CT MR

1993;14:389-398.

48. Jensen MC, Brant-Zawadzki MN, Obuchowski N,Modic MT, Malkasian D, Ross JS. Magnetic reso-nance imaging of the lumbar spine in people with-out back pain. N Engl J Med 1994;331:69-73.

49. Holt EP Jr. The question of lumbar discography.J Bone Joint Surg 1968;50A:720-726.

50. Walsh TR, Weinstein JN, Sprat KF, Lehmann TR,Aprill C, Sayre H. Lumbar discography in normalsubjects. A controlled, prospective study. J BoneJoint Surg 1990;72A:1081-1088.

51. Carragee EJ, Tanner CM, Yang B, Brito JL, TruongT. False-positive findings on lumbar discography.Reliability of subjective concordance assessmentduring provocative disc injection. Spine 1999;24:2542-2547.

52. Carragee EJ, Tanner CM, Khurana S, Hayward C,Welsh J, Date E, Truong T, Rossi M, Hagle C. Therates of false-positive lumbar discography in selectpatients without low back symptoms. Spine 2000;25:1373-1380.

53. Carragee EJ, Chen Y, Tanner CM, Truong T, Lau E,Brito JL. Provocative discography in patients afterlimited lumbar discectomy. Spine 2000;25:3065-3071.

54. Collis JS, Gardner WJ. Lumbar discography. Ananalysis of one thousand cases. J Neurosurg 1962;19:452-461.

55. Almen T. Experimental investigations with iohexoland their clinical relevance. Acta Radiol 1983;366(suppl):9-19.

56. Endres S, Bogduk N. Practice guidelines and proto-cols: Lumbar disc stimulation. Presented at The In-ternational Spinal Injection Society 9th Annual Sci-entific Meeting, Boston, MA, Sept 14-16, 2001, pp56-75.

57. Gibson MJ, Buckley J, Mawhinney R, MulhollandRC, Worthington BS. Magnetic resonance imagingand discography in the diagnosis of disc degenera-tion. A comparative study of 50 discs. J Bone JointSurg 1986;68B:369-373.

58. Simmons JW, Emery SF, McMillin JN, Landa D,Kimmich SJ. Awake discography. A comparisonstudy with magnetic resonance imaging. Spine 1991;16(6 Suppl):S216-221.

59. Collins CD, Stack JP, O’Connell DJ, Walsh M,McManus FP, Redmond OM, Ennis JT. The role ofdiscography in lumbar disc disease: A comparativestudy of magnetic resonance imaging and discogra-phy. Clin Radiol 1990;42:252-257.

60. Linson MA, Crowe CH. Comparison of magneticresonance imaging and lumbar discography in thediagnosis of disc degeneration. Clin Orthop 1990;250:160-163.

61. Schneiderman G, Flannigan B, Kingston S, ThomasJ, Dillin WH, Watkins RG. Magnetic resonance im-aging in the diagnosis of disc degeneration: Corre-lation with discography. Spine 1987;12:276-281.

62. Sandhu HS, Sanchez-Caso LP, Parvataneni HK,Cammisa FP Jr, Girardi FP, Ghelman B. Association

between findings of provocative discography and
Page 19: Discography Review

Lumbar Discography • Cohen et al. 181

vertebral endplate signals as seen on MRI. J SpinalDisord 2000;13:438-443.

63. Yu S, Haughton VM, Sether LA, Wagner M. Com-parison of MR and discography in detecting radialtears of the annulus: A postmortem study. AJNRAm J Neuroradiol 1989;10:1077-1081.

64. Kakitsubata Y, Theodorou DJ, Theodorou SJ,Trudell D, Clopton PL, Donich AS, Lektrakul N,Resnick D. Magnetic resonance discography in ca-davers: Tears of the annulus fibrosus. Clin Orthop2003;407:228-240.

65. Zucherman J, Derby R, Hsu K, Picetti G, Kaiser J,Schofferman J, Goldthwaite N, White A. Normalmagnetic resonance imaging with abnormal discog-raphy. Spine 1988;13:1355-1359.

66. Horton WC, Daftari TK. Which disc as visualized bymagnetic resonance imaging is actually a source ofpain? Correlation between magnetic resonance im-aging and discography. Spine 1992;17(6 Suppl):S164-171.

67. Milette PC, Fontaine S, Lepanto L, Cardinal E,Breton G. Differentiating lumbar disc protrusions,disc bulges, and discs with normal contour but ab-normal signal intensity. Magnetic resonance imag-ing with discographic correlations. Spine 1999;24:44-53.

68. Yoshida H, Fujiwara A, Tamai K, Kobayashi N, SaikiK, Saotome K. Diagnosis of symptomatic disc bymagnetic resonance imaging: T2-weighted and gad-olinium-DPTA-enhanced T1-weighted magneticresonance imaging. J Spinal Disord Tech 2002;15:193-198.

69. Birney TJ, White JJ Jr, Berens D, Kuhn G. Compar-ison of MRI and discography in the diagnosis oflumbar degenerative disc disease. J Spinal Disord1992;5:417-423.

70. Osti OL, Fraser RD. MRI and discography of annulartears and intervertebral disc degeneration. A pro-spective clinical comparison. J Bone Joint Surg 1992;74B:431-435.

71. Schellhas KP, Pollei SR, Gundry CR, Heithoff KB.Lumbar disc high-intensity zone. Correlation ofmagnetic resonance imaging and discography. Spine1996;21:79-86.

72. Loneragan R, Khangure MS, McCormick C,Hardcastle P. Comparison of magnetic resonanceimaging and computed tomographic discographyin the assessment of lumbar disc degeneration.Australas Radiol 1994;38:6-9.

73. Aprill C, Bogduk N. High-intensity zone: A diagnos-tic sign of painful lumbar disc on magnetic reso-nance imaging. Br J Radiol 1992;65:361-369.

74. Indahl A, Kaigle AM, Reikeras O, Holm SH. Inter-action between porcine lumbar intervertebral disc,zygapophyseal joints, and paraspinal muscles. Spine1997;22:2834-2840.

75. Schwarzer AC, Aprill CN, Derby R, Fortin J, Kine G,Bogduk N. The prevalence and clinical features ofinternal disc disruption in patients with chronic low

back pain. Spine 1995;17:1878-1883.

76. Junge A, Frohlich M, Ahrens S, Hasenbring M,Sandler A, Grob D, Dvorak J. Predictors of bad andgood outcome of lumbar spine surgery. A prospec-tive clinical study with 2 years’ follow up. Spine1996;21:1056-1064.

77. Gibson JN, Waddell G, Grant IC. Surgery for degen-erative lumbar spondylosis. Cochrane Database SystRev 2000;2:CD001352.

78. Deyo RA, Nachemson A, Mirza SK. Spinal-fusionsurgery—The case for restraint. N Engl J Med 2004;350:722-726.

79. Esses SI, Botsford DJ, Kostuik JP. The role of exter-nal spinal skeletal fixation in the assessment of low-back disorders. Spine 1989;14:594-600.

80. Gill K, Blumenthal SL. Functional results after an-terior lumbar fusion at L5-S1 in patients with nor-mal and abnormal MRI scans. Spine 1992;17:940-942.

81. Parker LM, Murrell SE, Boden SD, Horton WC. Theoutcome of posterolateral fusion in highly selectedpatients with discogenic low back pain. Spine 1996;21:1909-1917.

82. Kostuik JP. Decision making in adult scoliosis. Spine1979;4:521-525.

83. Knox BD, Chapman TM. Anterior lumbar interbodyfusion for discogram concordant pain. J Spinal Disord1993;6:242-244.

84. Derby R, Howard MW, Grant JM, Lettice JJ, VanPeteghem PK, Ryan DP. The ability of pressure-controlled discography to predict surgical and non-surgical outcomes. Spine 1999;24:364-372.

85. Vamvanij V, Fredrickson BE, Thorpe JM, StadnickME, Yuan HA. Surgical treatment of internal discdisruption: An outcome study of four fusion tech-niques. J Spinal Disord 1998;11:375-382.

86. Spruit M, Jacobs WC. Pain and function after intra-discal electrothermal treatment (IDET) for symp-tomatic lumbar disc degeneration. Eur Spine J 2002;11:589-593.

87. Saal JA, Saal JS. Intradiscal electrothermal treat-ment (IDET) for chronic discogenic low back pain: Aprospective outcome study with minimum 2-yearfollow-up. Spine 2002;27:966-973.

88. Cohen SP, Larkin T, Abdi S, Chang A, Stojanovic M.Risk factors for failure and complications of intra-discal electrothermal therapy: A pilot study. Spine2003;28:1142-1147.

89. Karasek M, Bogduk N. Twelve-month follow-up ofa controlled trial of intradiscal thermal anuloplastyfor back pain due to internal disc disruption. Spine2000;25:2601-2607.

90. Fernstrom U. Arthroplasty with intercorporeal en-doprosthesis in herniated disc and in painful disc.Acta Chir Scand Suppl 1966;357:154S-159S.

91. Blumenthal SL, Ohnmeiss DD, Guyer RD, Hoch-schuler SH. Prospective study evaluating total discreplacement: Preliminary results. J Spinal DisordTechnol 2003;16:450-454.

92. Shim CS, Lee SH, Park CW, Choi WC, Choi G, Choi

WG, Lim SR, Lee HY. Partial disc replacement with
Page 20: Discography Review

1

1

1

1

1

1

1

1

1

1

1

1

1

1

1

1

1

1

1

1

1

1

1

1

1

1

1

1

182 Regional Anesthesia and Pain Medicine Vol. 30 No. 2 March–April 2005

the PDN prosthetic disc nucleus device. J SpinalDisord Technol 2003;16:324-330.

93. van Ooij A, Oner FC, Verbout AJ. Complications ofartificial disc replacement: A report of 27 patientswith the SB Charite disc. J Spinal Disord Technol2003;16:369-383.

94. Tropiano P, Huang RC, Girardi FP, Marnay T. Lum-bar disc replacement: Preliminary results with Pro-Disc II after a minimum follow-up period of 1 year.J Spinal Disord Technol 2003;16:362-368.

95. McAfee PC, Fedder IL, Saiedy S, Shucosky EM,Cunningham BW. SB Charite disc replacement: Re-port of 60 prospective randomized cases in a U.S.center. J Spinal Disord Technol 2003;16:424-433.

96. Kim WJ, Lee SH, Kim SS, Lee C. Treatment ofjuxtafusional degeneration with artificial disc re-placement (ADR): Preliminary results of an ongoingstudy. J Spinal Disord Technol 2003;16:390-397.

97. Zeegers WS, Bohnen LM, Laaper M, Verhaegen MJ.Artificial disc replacement with the modular type SBCharite III: 2-year results in 50 prospectively studiedpatients. Eur Spine J 1999;8:210-217.

98. Bertagnoli R, Kumar S. Indications for full pros-thetic disc arthroplasty: A correlation of clinical out-come against a variety of indications. Eur Spine J2002;11(Suppl 2):S131-136.

99. Hochschuler SH, Ohnmeiss DD, Guyer RD, Blumen-thal SL. Artificial disc: Preliminary results of a prospec-tive study in the United States. Eur Spine J 2002;11(Suppl 2):S106-110.

00. Mayer HM, Wiechert K, Korge A, Qose I. Minimallyinvasive total disc replacement: Aurgical techniqueand preliminary clinical results. Eur Spine J 2002;11(Suppl 2):S124-130.

01. Jin D, Qu D, Zhao L, Chen J, Jiang J. Prosthetic discnucleus (PDN) replacement for lumbar disc hernia-tion. J Spinal Disord Technol 2003;16:331-337.

02. Zigler JE, Burd TA, Vialle EN, Sachs BL, RashbaumRF, Ohnmeiss DD. Lumbar Spine arthroplasty:Early results using ProDisc II: A prospective ran-domized trial of arthroplasty versus fusion. J SpinalDisord Technol 2003;16:352-361.

03. Enker P, Steffee A, McMillin C, Keppler L, Biscup R,Miller S. Artificial disc replacement: Preliminary re-port with a 3-year minimum follow-up. Spine 1993;18:1061-1070.

04. Schellhas KP, Pollei SR. The role of discography inthe evaluation of patients with spinal deformity.Orthop Clin North Am 1994;25:265-273.

05. Tehranzadeh J. Interventional procedures in mus-culoskeletal radiology: Discography 2000. RadiolClin North Am 1998;36:463-495.

06. Yrjama M, Vanharanta H. Bony vibration stimula-tion: A new, non-invasive method for examiningintradiscal pain. Eur Spine J 1994;3:233-235.

07. Yrjama M, Tervonen O, Kurunlahti M, VanharantaH. Bony vibration stimulation test combined withmagnetic resonance imaging. Can discography bereplaced? Spine 1997;22:808-813.

08. Vanharanta H, Ohnmeiss DD, Aprill CN. Vibration

pain provocation can improve the specificity of MRI

in the diagnosis of symptomatic lumbar discrupture. Clin J Pain 1998;14:239-247.

09. Executive Committee of the North American SpineSociety. Position statement on discography. Spine1988;13:1343.

10. Bernard TN Jr. Lumbar discography followed bycomputed tomography. Refining the diagnosis oflow back pain. Spine 1990;15:690-707.

11. Abdullah AF, Ditto EW, Byrd EB, Williams R. Ex-treme-lateral lumbar disc herniations. Clinical syn-drome and special problems of diagnosis. J Neuro-surg 1974;41:229-234.

12. Perey O. Contrast medium examination of the in-tervertebral disc of the lower lumbar spine. ActaOrthop Scand 1951;20:237-334.

13. Keck C. Discography: Technique and interpretation.Am Arch Surg 1960;80:580-585.

14. Edwards WC, Orme TJ, Orr-Edwards G. CT discog-raphy: Prognostic value in the selection of patientsfor chemonucleolysis. Spine 1987;12:792-795.

15. Ohnmiess DD, Guyer RD, Hochschuler SH. Laserdisc decompression: The importance of proper pa-tient selection. Spine 1994;19:2054-2058.

16. Trosier O, Cypel D. Discography: An element ofdecision: Surgery vs. chemonucleolysis. Clin Orthop1986;206:70-78.

17. Singh V, Piryani S, Liao K. Evaluation of percuta-neous disc decompression using coablation inchronic back pain with and without leg pain. PainPhysician 2002;6:273-280.

18. Guyer RD, Ohnmeiss DD. NASS: Lumbar discogra-phy. Spine J 2003;3(3 Suppl):11S-27S.

19. Tarver JM, Rathmell JP, Alsofrom GF. Lumbar dis-cography. Reg Anesth Pain Med 2001;26:263-266.

20. Ohnmeiss DD, Vanharanta H, Ekholm J. Relationbetween pain location and disc pathology: A studyof pain drawings and CT/discography. Clin J Pain1999;15:210-217.

21. Kumar N, Agorastides ID. The curved needle tech-nique for accessing the L5/S1 disc space. Br J Radiol2000;73:655-657.

22. Schellhas KP. Diskography. Neuroimaging Clin NorthAm 2000;10:579-596.

23. Cohen SP, Larkin T, Fant GV, Oberfoell R,Stojanovic M. Does needle insertion site affect dis-kography results? A retrospective analysis. Spine2002;27:2279-2283.

24. Falco FJ, Moran JG. Lumbar discography using gad-olinium in patients with iodine contrast allergy fol-lowed by postdiscography computed tomographyscan. Spine 2003;28:E1-4.

25. Fenton DS, Czervionke LF. Image-Guided Spine Inter-vention. Philadelphia: Saunders; 2003:227-256.

26. Raj PP, Lou L, Serdar E, Staats PS. Lumbar discog-raphy. From clinical insights on MD consult online.Available at: http://www.mdconsult.com. AccessedJanuary 2004.

27. Sachs BL, Vanharanta H, Spivey MA, Guyer RD,Videman T, Rashbaum RF, Johnson RG, Hoch-

schuler SH, Mooney V. Dallas discogram descrip-
Page 21: Discography Review

1

1

1

1

1

1

1

1

1

1

1

1

1

1

1

1

1

1

1

1

1

1

1

1

Lumbar Discography • Cohen et al. 183

tion. A new classification of CT/discography inlow-back disorders. Spine 1987;12:287-294.

28. Cohen SP, Williams S, Kurihara C, Griffith S, LarkinTM. Nucleoplasty with or without intradiscal elec-trothermal therapy (IDET) as a treatment for lum-bar herniated disc. J Spinal Disord Technol 2005;18(Suppl 1):S119-S124.

29. Guyer RD, Collier R, Stith WJ, Ohnmeiss DD, Hoch-schuler SH, Rashbaum RF, Regan JJ. Discitis afterdiscography. Spine 1988;13:1352-1354.

30. Silber JS, Anderson DG, Vaccaro AR, Anderson PA,McCormick P. Management of postprocedural dis-citis. Spine J 2002;2:279-287.

31. Fraser RD, Osti OL, Vernon-Roberts B. Discitis afterdiscography. J Bone Joint Surg 1987;69B:26-35.

32. Guyer RD, Ohnmeiss DD. Lumbar discography. Po-sition statement from the North American SpineSociety Diagnostic and Therapeutic Committee.Spine 1995;20:2048-2059.

33. Osti OL, Fraser RD, Vernon-Roberts B. Discitis afterdiscography. The role of prophylactic antibiotics.J Bone Joint Surg 1990;72B:271-274.

34. Klessig HT, Showsh SA, Sekorski A. The use ofintradiscal antibiotics for discography: an in vitrostudy of gentamycin, cefazolin and clindamycin.Spine 2003;28:1735-1738.

35. Willems PC, Jacobs W, Duinkerke ES, De KleuverM. Lumbar discography: should we use prophylac-tic antibiotics? A study of 435 consecutive disco-grams and a systematic review of the literature.J Spinal Disord Technol 2004;17:243-247.

36. Carragee EJ, Chen Y, Tanner CM, Hayward C, RossiM, Hagle C. Can discography cause long-term backsymptoms in previously asymptomatic subjects?Spine 2000;25:1803-1808.

37. Johnson RG. Does discography injure normal discs?An analysis of repeat discograms. Spine 1989;14:424-426.

38. Flanagan MN, Chung BU. Roentgenographicchanges in 188 patients 10-20 years after discogra-phy and chemonucleolysis. Spine 1986;11:444-448.

39. Carragee EJ, Barcohana B, Alamin T, van den HaakE. Prospective controlled study of the developmentof lower back pain in previously asymptomatic sub-jects undergoing experimental discography. Spine

2004;29:1112-1117.

40. Tallroth K, Soini K, Antti-Poika I, Konttinen YT,Honkanen V, Yrjonen T. Premedication and shortterm complications in iohexol discography. AnnChir Gynaecol 1991;80:49-53.

41. Zeidman SM, Thompson K, Ducker TB. Complica-tions of cervical discography: Analysis of 4400diagnostic disc injections. Neurosurgery 1995;37:414-417.

42. Roosen K, Bettag W, Fiebach O. Complications ofcervical discography (in German). ROFO FortschrGeb Rontgenstr Nuklearmed 1975;122:520-527.

43. Junila J, Niinimaki T, Tervonen O. Epidural abscessafter lumbar discography. A case report. Spine 1997;22:2191-2193.

44. deSeze S, Levernieux J. Les accidents de la discog-raphie. Rev Rhum Mal Osteoartic 1952;19:1027-1033.

45. Goldie I. Intervertebral disc changes after discogra-phy. Acta Chir Scand 1957;113:438-439.

46. Keefe FJ, Wilkins RH, Cook WA. Direct observationof pain behavior in low back pain patients duringphysical examination. Pain 1984;20:59-68.

47. Spratt KF, Lehmann TR, Weinstein JN, Sayre HA. Anew approach to the low-back physical examina-tion. Behavioral assessment of mechanical signs.Spine 1990;15:96-102.

48. Olmarker K, Rydevik B. Selective inhibition of tu-mor necrosis factor-alpha prevents nucleus pulpo-sus-induced thrombus formation, intraneuraledema, and reduction of nerve conduction velocity.Possible implications for future pharmacologictreatment strategies of sciatica. Spine 2001;26:863-869.

49. Karppinen J, Korhonen T, Malmivaara A, PaimelaL, Kyllonen E, Lindgren KA, Rantanen P, TervonenD, Niinimaki J, Seitsalo S, Hurri H. Tumor necrosisfactor-alpha monoclonal antibody, infliximab, usedto manage severe sciatica. Spine 2003;28:750-754.

50. Tobinick EL, Britschgi-Davoodifar S. PerispinalTNF-alpha inhibition for discogenic pain. Swiss MedWkly 2003;133:170-177.

51. Tobinick E, Davoodifar S. Efficacy of etanercept de-livered by perispinal administration for chronic backand/or neck disc-related pain: A study of clinicalobservations in 143 patients. Curr Med Res Opin

2004;20:1075-1085.

Recommended