Fusion Lombaire

7/26/2019 Fusion Lombaire

1/13

SPINE Volume 24, Number 17, pp 1820 18321999, Lippincott Williams & Wilkins, Inc.

The Cochrane Review of Surgery for Lumbar DiscProlapse and Degenerative Lumbar Spondylosis

J. N. Alastair Gibson, MD, FRCS, Inga C. Grant, MSc, and Gordon Waddell, DSc, MD, FRCS

Study Design. A Cochrane review of randomized con-

trolled trials.

Objectives. To collate the scientific evidence on surgi-

cal management for lumbar disc prolapse and degenera-

tive lumbar spondylosis.

Summary of Background Data.Surgical investigations

and interventions account for as much as one third of the

health care costs for spinal disorders, but the scientific

evidence for most procedures still is unclear.

Methods.A highly sensitive search strategy identified

all published randomized controlled trials. Cochrane

methodology was used for meta-analysis of the results.

Results. Twenty-six randomized controlled trials of

surgery for lumbar disc prolapse and 14 trials of surgery

for degenerative lumbar spondylosis were identified.

Methodologic weaknesses were found in many of the

trials. Only one trial directly compared discectomy and

conservative management. Meta-analyses showed that

surgical discectomy produces better clinical outcomes

than chemonucleolysis, which is better than placebo.

Three trials showed no difference in clinical outcomes

between microdiscectomy and standard discectomy, but

in three other studies, both produced better results than

percutaneous discectomy. Three trials showed that in-

serting an interposition membrane after discectomy does

not significantly reduce scar formation or alter clinical

outcomes. Five heterogeneous trials on spinal stenosis

and degenerative spondylolisthesis permit very limited

conclusions. There were nine trials of instrumented ver-sus noninstrumented fusion: Meta-analysis showed that

instrumentation may facilitate fusion but does not im-

prove clinical outcomes.

Conclusions.There is now strong evidence on the rel-

ative effectiveness of surgical discectomy versuschemo-

nucleolysis versus placebo. There is considerable evi-

dence on the clinical effectiveness of discectomy for

carefully selected patients with sciatica caused by lumbar

disc prolapse that fails to resolve with conservative man-

agement. There is no scientific evidence on the effective-

ness of any form of surgical decompression or fusion for

degenerative lumbar spondylosis compared with natural

history, placebo, or conservative management. The Co-

chrane reviews will be updated continuously as other

trials become available. [Key words: chemonucleolysis,Cochrane review, decompression, degenerative disc dis-

ease, discectomy, fusion, instrumented fusion, lumbar

disc prolapse, lumbar spondylosis, meta-analysis, out-

comes, randomized controlled trials, spinal stenosis, sur-

gery, systematic review] Spine 1999;24:18201832

In all studies of lumbar spine disorders, 10% of patientsaccount for more than 80% of the total health care andsocial costs, and the 1% of patients who undergo surgeryare the most expensive group. Although surgical investi-gations and interventions account for up to one third ofthe health care costs,13 the scientific evidence for mostprocedures is unclear. Surgical practice and any pro-posed new developments should have a proper scientificbasis, which should include randomized controlled tri-als (RCTs).

This article is based on two Cochrane reviews of sur-gical management for lumbar disc prolapse25 and degen-

erative lumbar spondylosis,26

which address these ques-tions: 1) What evidence exists concerning the clinicaleffectiveness of lumbar spine surgery? and 2) What evi-dence exists concerning alternative forms and techniquesof lumbar spine surgery?

The primary rationale for surgery in any form for discprolapse is to relieve nerve root irritation or compressioncaused by herniated disc material. It is important to de-fine the optimal management for specific types of pro-lapse. For example, different surgical procedures areprobably appropriate if disc material is sequestratedrather than contained by the outer layers of the anu-

lus fibrosus.Chemonucleolysis using chymopapain was advocatedat one point for contained lumbar disc prolapse (i. e.,when there was no sequestration of fragment[s] into thespinal canal),59 but the trial by Schwetschenau et al58

instigated a moratorium by the Food and Drug Admin-istration, and the drug was not released again for generaluse until 1982. Now, 17 years later, the role and effec-tiveness of chemonucleolysis still is disputed by spinalsurgeons, making a systematic review of all availableevidence especially valuable.

Several nonsystematic reviews consider the relativemerits of microdiscectomy and automated percutaneousdiscectomy. In both treatments, smaller wounds are saidto promote faster patient recovery with earlier hospitaldischarge,36,52 but there still is no clear evidence aboutthe clinical outcomes of these procedures. In the firstclinical series of 420 patients who had laser discectomy,a 78% good or fair result was reported, but there was nocontrol group, and outcome assessment was not blind-ed.11 Subsequent studies have shown quite variable results.

There are two published meta-analyses on the surgicalmanagement of spinal stenosis50,68 and one on the sur-gical management of degenerative spondylolisthesis.42

These reviews suggest that decompression alone is the

safest procedure and may give better results for patients

From The University Department of Orthopaedic Surgery, PrincessMargaret Rose Orthopaedic Hospital, Edinburgh, Scotland.Funding was provided by the Medical Research Council, UK.Acknowledgment date: February 9, 1999.Acceptance date: February 9, 1999.Device status category: 9.

1820


2/13

with spinal stenosis, those with less than 8 years of clin-ical history, and elderly unfit patients. These studies alsosuggest that decompression combined with instrumentedfusion may have some place for patients with degenera-tive spondylolisthesis, or for those with more than 15years of history, but that there is a higher complicationrate. However, these reviews were based mainly on un-controlled, retrospective case series that reported widely

varying results, and thus do not permit firm conclusions.After more than 80 years, there still is considerable

dispute as to whether lumbar fusion is an appropriateand effective method of managing back pain in patientswho have degenerative lumbar spondylosis with nonerve root or cauda equina compression. Surgical tech-nique varies widely. The nature and role of instabilityis poorly defined, and the clinical indications for surgeryuncertain.62 Indeed, when reported satisfactory clinicaloutcomes range from 16% to 95%,67 it is particularlyapposite to perform meta-analysis of the availableevidence.

In recent years, there has been an explosion of surgicaland commercial interest in widely varying methods ofinstrumented fusion in both Europe and the UnitedStates. The aforementioned meta-analysis of publishedcase series of degenerative spondylolisthesis42 suggestedthat fusion with pedicle screws produced a higher fusionrate (93% vs. 86%) than fusion without instrumenta-tion, although the difference was not statistically signifi-cant and produced no difference in clinical outcomes(86%vs. 90%). Randomized controlled trials of fusionwith instrumentation have appeared only in the past 6years, and the results have not yet been collated. There is

even less available scientific information about othermethods of fusion, whether anterior or posterior.

Materials and Methods

All randomized or quasi-randomized (methods of allocatingpatients between treatments that are not strictly random, forexample, by date of birth, hospital record number, or alterna-tion) controlled trials were identified in a search of computerdatabases including MEDLINE (1966 to December 1998),EMBASE, BIOSIS, Dissertation Abstracts, Index to UK Theses,and the Cochrane Controlled Trials Register, and by handsearching of relevant surgical journals, citation tracking, and

correspondence with experts. The details of the search strate-gies and specific search terms are provided in the Co-chrane reviews.25,26

All trials pertinent to the management of 1) lumbar discprolapse and 2) degenerative lumbar spondylosis and the asso-ciated pathologies or clinical syndromes of back pain, instabil-ity, spinal stenosis, and degenerative spondylolisthesis were in-cluded. Trials of patients with spinal fractures, tumors,infection, and childhood deformity or RCTs investigating theeffectiveness of steroid injections and local anesthetic as well asvarious postoperation management regimens were excluded.The quality of assignment concealment was rated in terms ofthree grades: A (clearly yes), B (unclear), and C (clearly no).Withdrawal, blinding of patients and observers, and intentionto treat analyses were assessed according to the standard Co-

chrane method. These data and the nature and timing of theoutcome measures were tabulated (Tables 1 and 2).

The most frequently reported outcome measures (found infour or more studies) were used for meta-analysis. In fact, forlumbar disc prolapse, only three outcomes were reported con-sistently: the patients rating of success, a surgeons rating ofsuccess, and the need for a second procedure (treatment fail-ure); for degenerative lumbar spondylosis, only four outcomeswere reported consistently: the patient and surgeon ratings of

success, the need for a second procedure (treatment failure),and the attainment of spinal fusion. To pool the results, ratingsof excellent, good, and fair were classified as success,whereas ratings of poor, unimproved, and worse were catego-rized as failure.

Random effects odds ratios (ORs) are quoted unless it isotherwise stated. Details of all the analyses are available in theCochrane reviews, and only the most important meta-analysesare presented here as figures. It should be noted that the currentORs and figures of successful surgical outcome appear to differfrom the same data in the Cochrane reviews because of theCochrane convention of presenting all results negatively asfailure. The following levels of evidence were used: A, Strong

(consistent evidence in multiple high-quality RCTs); B, Mod-erate (consistent evidence in one high-quality RCT or severallow-quality RCTs); C, Limited (inconsistent evidence in one ormore RCTs); and D, Nil (no scientific evidence in any trial).

Results

In all, 26 RCTs of lumbar disc prolapse (Table 1) and 14RCTs of degenerative lumbar spondylosis (Table 2) wereidentified and included in the review.

Methodologic Quality of the Studies

Lumbar Disc Prolapse.Many of the trials had defects ofdesign, which were generally worse in the trials of sur-

gery than in trials of chemonucleolysis. Some of the trialsincluded very small numbers of patients. Methods andpublished details of randomization often were poor, andthere was a lack of randomization concealment. The na-ture of the surgical interventions often made doubleblinding impossible. Blinded assessment of outcome gen-erally was feasible, yet frequently not even attempted.There were few proper clinical outcomes,15 and the mostcommon surgical outcomes were crude ratings by pa-tients or surgeons. Some of the assessments were madeby the operating surgeon, or by a resident or fellow be-holden to the primary investigator.

Although 22 of the 26 trials had follow-up rates of atleast 90%, there often was considerable early code breakor crossover of patients, which was not always properlyallowed for in the analysis or presentation of results.Only 4 of the 26 trials presented 2-year follow-up resultsas recommended for surgical studies, although 2 of thesetrials also did present 10-year results. The defects of trialdesign introduced some potential for bias. Most of theconclusions in this review are about short-term, 6- to12-month outcomes, and there is a general lack of infor-mation on long-term outcomes.

Degenerative Lumbar Spondylosis.Many of the trials had

major defects of design. Six of the 14 trials each had

1821Cochrane Review of Lumbar Surgery Gibson et al


3/13

Table 1. Characteristics of Included StudiesLumbar Disc Prolapse

Study Method Participants Interventions Outcomes

Benoist 19931

(Paris, France)Randomized from independently

generated listAllocation concealment: BBlinding: doubleLost to follow-up: 34/118 at 1 yr

118 patients; 80 m, 38 f; age 2170 yrLumbar disc herniation radicular

painUnsuccessful conservative treatment

(6 wk)

Exp: chymopapain(2000 units)

Ctl: chymopapain(4000 units)

Patient ratingSurgeon rating1 yr

Bontoux 19904

(Poitiers, France)

Randomization from random

tableAllocation concealment: ABlinding: assessorLost to follow-up: 0/80 at 6 mo

80 patients

Sciatica for 2 mo

Exp: chymopapain

(4000 units)Ctl: trimacinolonehexacetonide (70mg)

Independent

observer rating2nd procedurerequired

6 moBourgeois 19886

(Paris, France)Randomized by drawing lotsAllocation concealment: CBlinding: doubleLost to follow-up: 0/60 at 6 mo

60 patients; 40 m, 20 f; age 2662 yrSciatica for 6 wk


Ctl: triamcinolonehexacetonide (80mg)

Independentobserver rating2nd procedurerequired

6 moBromley 19848

(New Jersey)Randomized from random tableAllocation concealment: ABlinding: doubleLost to follow-up: 0/30 at 17 mo

30 patients; 15 m, 15 f; age 2163 yrFailed conservative therapy (including

2 wk bed rest)Myelogram: confirming a single

herniated disc

Exp: collagenase (600units/mL)

Ctl: normal saline

Patient rating17 mo

Chatterjee 199510

(Liverpool, UK)Randomization method not

statedAllocation concealment: BBlinding: assessorLost to follow-up0/71 at 6 mo

71 patients; 39 m, 32 f; age 2067 yrContained disc herniation at a single

levelUnsuccessful conservative treatment

(minimum 6 wk)

Exp:automatedpercutaneouslumbar discectomy

Ctl: microdiscectomy

Repeat surgery(microdiscectomy)requiredfollowing failedAPLD

Cost analysisreported byStevenson etal61

Crawshaw 198412

(Nottingham, UK)Randomization method not

statedAllocation concealment: BBlinding: NilLost to follow-up: 2/52 at 1 yr

52 patients; age 1560 yrRoot involvement at a single levelFailed conservative treatment

(minimum 3 mo)

Exp: discectomy(standard)


Independentobserver rating

6 mo

Dabezies 198814

(MulticenterUS)Randomization method not

statedAllocation concealment: ABlinding: doubleLost to follow-up: 9/173 at 6 mo

173 patients; 112 m, 61 f; age 1870 yrProven classic lumbar disc syndrome

with unilateral single-levelradiculopathy

Failed conservative treatment

(minimum 2 wk strict bed rest)

Exp: chymopapain (8mg in 2 mL)

Ctl: cysteine-edetate-iothalamate

Surgeon rating 2ndprocedurerequired

1 yr

Ejeskar 198316

(Gothenburg, Sweden)Randomization method not

statedAllocation concealment: BBlinding: assessorLost to follow-up: 0/29 at 1 yr

29 patients; 22 m, 7 f; age 1973 yrObvious signs symptoms of a

herniated discSevere symptoms for longer than 4

mo positive myelogram



Patient ratingSurgeon rating 2nd

procedurerequired

1 yrFeldman 198618

(Paris, France)Randomized by drawing lotsAllocation concealment: BBlinding: doubleLost to follow-up: 0/39

39 patientsSymptoms resistant to 1 mo

conservative treatment


Ctl: saline (2 mL)

Patient ratingSurgeon rating2nd procedure

required1 yr

Fraser 19822224,27

(Adelaide, Australia)Randomization method not

statedAllocation concealment: ABlinding: doubleLost to follow-up: 0/60 at 2 yr;

4/60 at 10 yr

60 patients; 39 m, 21 f; age 1969 yrFailed conservative treatment

(unknown duration) withinpreceding 6 mo

Myelogram demonstrating

posterolateral herniated disc atsingle level

Exp: chymopapain (8mg in 2 mL)

Ctl: saline (2 mL)


required2, 10 yr

Hedtmann 199230,72

(Bochum, Germany)Randomization method not

statedAllocation concealment: BBlinding: nilLost to follow-up: 0/100

100 patients; 65 m, 35 fContained disc at 1 levelFailed conservative treatment

(minimum 6 wk)


Ctl: collagenase (400ABC units)

Surgeon rating1 yr2nd procedure

required3 yr

Henriksen 199631

(Copenhagen,Denmark)

Randomized by closed envelopetechnique

Allocation concealment: ABlinding: single

79 patients; age 3048 yrSingle level nerve root compromise

Exp: microdiscectomyCtl: discectomy

(standard)

Back pain scoreLeg pain scoreTime to discharge

Javid 198334

(multicenterUS)Randomization from permuted

blocksAllocation concealment: ABlinding: doubleLost to follow-up: 2/108 at 6 mo

108 patients; 63 m, 45 f; age 3641 yrPeriod of study 19811982Positive myelogram for single disc

herniationFailed conservative treatment

(minimum 6 wk)


Ctl: saline

Patient ratingPhysician rating2nd procedure

requiredCode break

6 mo

1822 Spine Volume 24 Number 17 1999


4/13

Table 1. (Continued)


Jensen 199635

(Hilleroed, Denmark)Randomization method not

statedAllocation concealment: BBlinding: single assessor

118 patients; 53 m, 46 f; age 1975Myelogram or CT verified disc

prolapse

Exp: implantation offree fat graftfollowingdiscectomy

Ctl: nil followingdiscectomy

Patient assessmentCT

1 yr

Lagarrigue 1994

38

(Toulouse, France) Randomized by drawing of lotsAllocation concealment: BBlinding: assessorLost to follow-up: 0/80 at 15 mo

80 patientsDisc hernia treated conservatively for3 mo

Exp: microdiscectomyCtl: discectomy(standard)

Surgeon rating15 mo

Lavignolle 198739

(Bordeaux, France)Randomization method not

statedAllocation concealment: BBlinding: nilLost to follow-up: 0/358 at 2 yr

358 patientsHernia without major neurological

deficit

Exp: microdiscectomyCtl: chymopapain

(4000 units)

Surgeon ratingIndependent

observer rating2nd procedure

requiredCost analysis

2 yrMacKay 199540

(Michigan)Randomization method not

statedAllocation concealment: BBlinding: assessorLost to follow-up: 36/154

154 patients; 106 m, 48 f; age 1479 yrRadiographically proven single-level

herniationUnsuccessful nonoperative treatment

(minimum 6 wk)

Exp: free-fat graft/gelfoam

Ctl: nil

Independentobserverassessment

MRI scar formation1 yr

Mayer 199344

(Berlin, Germany)Randomization method not

statedAllocation concealment: BBlinding: nilLost to follow-up: 0/40 at 2 yr

40 patients; 26 m, 14 f; age 1263 yrPrevious unsuccessful conservative

treatment (time period not stated)Only small noncontained disc

herniations included

Exp: percutaneousendoscopicdiscectomy

Ctl: microdiscectomy

Patient ratingSurgeon ratingRepeat surgery

2 yr

Muralikuttan 199249

(Belfast, UK)Randomized from computer-

generated randomization listAllocation concealment: BBlinding: nilLost to follow-up: 6/92 at 1 yr

92 patients; 55 m, 37 f; age 1960 yrNerve root pain with/without back

painFailed conservative treatment

(minimum 4 wk, including 2 wk bedrest)



2nd procedurerequired

Cost analysis1 yr

North 199551

(Baltimore, Maryland)Randomization method not

statedAllocation concealment: B3/30 had not reached 6 mo

review

30 patientsFailed back surgery syndrome

Exp: spinal cordstimulation

Ctl: reoperation

Crossover6 mo

Petrie 199653

(Multicenter, Europe)

Randomization by closed

envelope techniqueAllocation concealment: ABlinding: single assessor

213 patients; 134 m, 79 f; age mean 40

yr Ctl; 38 yr ExpDiscectomy patients

Exp: ADCON-L

antiadhesionbarrier gel

Ctl: nil

Patient rating

MRI scar score6 mo

Revel 199355

(Paris, France)Randomization from perrmuted

blocksAllocation concealment: BBlinding: nilLost to follow-up: 2/141 at 6 mo

141 patients; 94 m, 47 f; age 2165 yrUnsuccessful conservative treatment

(minimum 30 days)Proven disc herniation at one

vertebral level

Exp: automatedpercutaneousdiscectomy


Patient rating2nd procedure

required6 mo

Schwetschenau 197643,58

(Washington DC)Randomization method not

statedAllocation concealment: ABlinding: doubleLost to follow-up: 0/66 at 1 yr

66 patients; 44 m, 22 f; age 3240 yrOne or more clinical signs of

herniated lumbar disc positivemyelogram

Failed conservative treatment(including at least 3 wk bed rest)

Exp: chymopapain (4mg)

Ctl: sodiumiothalamate (4 mg)

Surgeon rating2nd procedure

required(discectomy)

1 yr

Tulberg 199365,66

(Stockholm, Sweden)Randomization method not

stated

Allocation concealment: BBlinding: nilLost to follow-up: 0/60 at 1yr

60 patients; 39 m, 21 f; age 1764 yrSingle lumbar disc herniation

Failed conservative treatment(minimum 2 mo)

Exp: microdiscectomyCtl: discectomy

(standard)

Back pain score2nd procedure

required1 yr

van Alphen 198969

(Amsterdam, Holland)Quasi-randomized by

alternationAllocation concealment: BBlinding: nilLost to follow-up: 1/151 at 1 yr

151 patients; 99 m, 52 f; age 1845 yrProven disc herniation (myelography)Failed conservative treatment

(including 2 wk minimum bed rest)

Exp: discectomy(standard)Ctl: chymopapain(4000 units)


required1 yr

Weber 198370,71

(Oslo, Norway)Randomized by closed envelope

techniqueAllocation concealment: ABlinding: nilLost to follow-up: 5/126 at 1 yr

126 patients; 68 m, 58 f; age 2555 yr5th lumbar /or 1st sacral root lesionFailed conservative treatment

(minimum 2 wk)


Ctl: conservativetreatment

Independentobserver rating

Cost analysisreported byMalter et al40

1, 4, and 10 yr

Ctl control; Exp experimental; M male; F female; APLD automated percutaneous lumbar discectomy.



5/13

Table 2. Characteristics of Included StudiesDegenerative Lumbar Spondylosis


Bridwell 19937

(St. Louis, Missouri)Randomization method not statedAllocation concealment: CBlinding: nilLost to follow-up: 1/44 at 2 yr

44 patients; 10 m, 34 f; age 4479 yrSpinal claudication symptomsCAT scan/myelogram/MRI

evaluation

Exp: posterolateral fusion with/without instrumentation

Ctl: no fusion

Spondylolisthesisprogression

2nd procedure required2 yr

Carragee 19979

(Stanford, California)Randomized from sealed envelopes

containing random numbersAllocation concealment: A

Blinding: nilLost to follow-up: 2 at 4.5 yr

42 patients; 26 m, 16 f; age 1951 yrGrade I or II isthmic

spondylolisthesis

Exp: a) Smokers withinstrumented arthrodesis decompressive laminectomy

b) Nonsmokers with graftalone decompressivelaminectomy

Ctl: same groups withoutdecompressive laminectomy

Back pain scalePatient ratingFusion

3 yr

Emery, 199517

(Cleveland, Ohio)Randomization method not statedAllocation concealment: BBlinding: nilLost to follow-up: 5/84 at 22 mo

84 patientsHeterogeneous clinical conditions

Exp: fusion Luque plate/screwsystem

Ctl: posterolateral fusion withautogenous graft

Fusion22 mo

Fischgrund 199719,20

(Royal Oak,Michigan)

Randomized by closed envelopetechnique

Allocation concealment: ABlinding: assessorLost to follow-up: 8/76 at 2.4 yr

76 patients; 17 m, 59 f; age 5286 yrDegenerative spondylolisthesis and

spinal stenosis

Exp: fusion Steffee plateCtl: posterolateral fusion only

Back pain scaleLeg pain scaleSurgeon ratingFusionProgression of

spondylolisthesis2 yr

France 199721

(multicenterUSA)Randomization method not statedAllocation concealment: BBlinding: nilLost to follow-up: 12/83 at 40 mo

83 patients; 58 m, 25 f; age 1976 yrHeterogeneous spinal conditions

Exp: fusion Steffee plateCtl: posterolateral fusion only

Back pain scalePatient ratingFusion

2 yrGrob 199529

(Zurich, Switzerland)Quasi-randomized by date of

admissionAllocation concealment: CBlinding: nilLost to follow-up: 0/30 at 28 mo

45 patients; 21 m, 24 f; age 4887 yrStenosis accepted as a mid-sagittal

spinal canal diameter of less than11 mm

Exp: decompression witharthrodesis (both mono multisegmental)

Ctl: decompression withoutarthrodesis

Patient ratingSurgeon rating2nd procedure required28 mo

Herkowitz 199132

(Royal Oak,Michigan)

Quasi-randomized by alternatelyassigning to treatments

Allocation concealment: BBlinding: nilLost to follow-up: 0/50 at 3 yr

50 patients; 36 f, 14 m; age 5284 yrSpinal stenosis associated with

degenerative lumbarspondylolisthesis

Exp: decompression fusionCtl: decompression

Back pain scaleLeg pain scaleSurgeon ratingFusionProgression of

spondylolisthesis3 yr

McGuire 199345

(multicenterUS)Randomization method not statedAllocation concealment: BBlinding: nilLost to follow-up 1/27 at 2 yr

27 patients; 23 m, 4 f; age 2442 yrSymptomatic grade I to II

spondylolisthesis refractory toconservative care

All treated with laminectomy

nerve root decompression

Exp: fusion Steffee plateCtl: posterolateral fusion

Fusion2nd procedure required2 yr

Moller 199646,47

(Stockholm,Sweden)(Full report andevaluation pending)

Evaluation pending 43 patients; age 1855 yrIsthmic spondylolisthesis (all

grades)

Exp: instrumented fusionCtl: posterolateral fusion alone

Clinical outcome data viaProfessor Melot

Mooney 199048

(Irvine, California)Randomization method not statedAllocation concealment: BBlinding: doubleLost to follow-up: 11/206 at 1 yr

195 patientsPatients undergoing initial attempts

at interbody spinal fusion(anterior or posterior approach)

Exp: electromagnetic brace 8 hrper day

Ctl: placebo brace 8 hr per day

Surgeon rating1 yr

Postacchini 199354

(Rome, Italy)Randomized by allocating

according to pathology and thenassigning alternately

Allocation concealment: CBlinding: nilLost to follow-up: 3/70 at 3.7 yr

70 patients.34 m, 36 f; age 4379 yrCentral lumbar stenosis

Exp: multiple laminotomyCtl: laminectomy

Patient ratingSurgeon ratingProgression of

spondylolisthesisOperating timeBlood loss3.7 yr

Rogozinski 199556

(Florida)

Randomization method not stated

Allocation concealment: BBlinding: nilLost to follow-up: nil at 2 yr

26 patients

Condition not specified

Exp: Implanted bone stimulation

wireCtl: no implant

Fusion

Thomsen 199764

(multicenter,Denmark)

Randomization method not statedAllocation concealment: BBlinding: assessorLost to follow-up: 3/129 at 2 yr

130 patients; 60 m, 69 f; age 2067yr

Severe low back pain of variousetiology

Exp: fusion CD pedicle screwrod system

Ctl: posterolateral fusion

Functional scalePatient ratingFusion2nd procedure required2 yr

Zdeblick 199373,74

(Wisconsin)Randomization by random number

generatorAllocation concealment: CBlinding: nilLost to follow-up: 1/124

124 patients; age 2080 yrLumbar or lumbosacral fusion for

degenerative conditions of thespine

Exp: fusion semi-rigid/rigidfixation

Ctl: fusion with autogenous bonegraft

Surgeon rating2nd procedure required16 mo (mean)

Zdeblick 199675

(Wisconsin)Randomization method not statedAllocation concealment: BBlinding: assessorLost to follow-up: nil

39 patientsL5S1 degenerative disc disease

with normal proximal discs

Exp: TSRH PLIFExp: ALIFCtl: TSRH

Pain requiring medicationFusion2nd procedure requiredBlood lossCost analysis

CH control; Exp experimental; M male; F female; TSRH Texas Scottish Rite Hospital; PLIF posterior lumbar interbody fusion; ALIF anteriorlaparoscopically assisted interbody fusion.



6/13

fewer than 50 patients. Many of the trials included het-erogeneous pathologies and clinical syndromes, and thesubgroups often were too small to give meaningful re-sults. Published details of randomization often werepoor, but there appeared to be a general lack of aware-ness concerning the importance of the randomizationmethod. In only two studies was there a clear attempt atallocation concealment.9,19 In eight trials, the method of

allocation was not described. One trial was considered tobe quasi-randomized because the patients were allocatedaccording to their date of admission to the hospital.29

Four trials clearly were open to potential selectionbias.7,29,54,73

Of the 14 trials, 10 had the recommended follow-upfor surgical studies of at least 2 years. All but one had afollow-up rate of at least 90%, and that one trial gavedifferent patient outcomes after best and worst case anal-yses.21 There usually was little or no attempt at blinding,and the primary investigator or the operating surgeonoften made the outcome assessment. Only one trial was

double blinded,48

and three others used an indepen-dent assessor.19,64,73

There was a general lack of patient-oriented, clinicaloutcomes.15 Ten trials gave technical surgical outcomessuch as fusion, spondylolisthesis progression, or need forreoperation. Clinical outcomes were mainly crude rat-ings on a 3- or 4-point scale: five trials gave a surgeonsrating and six gave a patients rating. Ten trials madesome attempt at assessing pain relief, although most useddifferent methods: Four trials published actual pain scalemeasurements,9,19,21,32 and one assessed functional out-come using a validated assessment scale.64

Efficacy of Treatment

Lumbar Disc Prolapse. Of the 26 RCTs identified, 15involved some form of chemonucleolysis, 10 includeddifferent forms or techniques of surgical discectomy, andone was unclassified. It was not possible to analyze pa-tients according to duration of their symptoms, previousconservative management, type of disc prolapse, or indi-cations for surgery because few of the trials provided thisdata in usable form. Four trials in this review attemptedto estimate costs,39,41,49,61 and two of these estimatedcost effectiveness,41,61 although their method has been

severely criticized.28

Chemonucleolysis (15 Trials)Five trials compared the efficacy of chemonucleolysis us-ing chymopapain versus placebo. These trials had thehighest-quality scores in this review, with generally ade-quate randomization, double blinding, and independentoutcome assessment. In all the trials, chymopapain wasinjected by standard technique.

The combined results from the five trials, comparingdata from 446 patients with an average follow up of97%, are summarized in Table 1. The meta-analysisclearly showed that chymopapain was more effective

than placebo, whether outcomes were rated by the pa-

tients (OR, 4.14; 95% confidence interval [CI], 2.04,8.42), by surgeons conducting the study, or by indepen-dent observers (OR, 2.51; 95% CI, 1.33, 4.76; Figure 1).Fewer patients after chymopapain injection proceeded toopen discectomy (OR, 2.42; 95% CI, 1.46, 3.99).

Another five trials compared chemonucleolysis and

surgical discectomy. In each case, a set dose of chymo-papain was injected by standard technique and com-pared with standard discectomy. In all the trials, therewas a poor description of the randomization method,and the nature of these studies precluded blinding ofthe patients.

The combined results from the five trials, which com-pared data from 680 patients with an average follow-upof 97%, are summarized in Table 1. It should be notedthat the test for homogeneity was significant in thisgroup of trials. Most of the trials and all of the analysesshowed consistently poorer results with chemonucleoly-

sis, although not all were statistically significant. Twotrials showed a worse result at 1 year as rated by thepatients (OR, 1.65; 95% CI, 0.81, 3.38). Three trialsshowed a poorer result at 1 year as rated by the surgeons(fixed OR, 1.96; 95% CI, 1.31, 2.92, and random OR,2.81; 95% CI, 0.96, 8.23; Figure 2).

Approximately 30% of the patients with chemonucle-olysis had further disc surgery within 2 years, and meta-analysis showed that a second procedure was more likelyafter chemonucleolysis (OR, 15.1; 95% CI, 5.6, 41.2).However, surgery after failed chemonucleolysis is notstrictly comparable with repeat surgery after failed disc-ectomy. There was some suggestion that the results ofdiscectomy after failed chemonucleolysis were poorerthan those after primary discectomy, but there was in-sufficient data to allow meta-analysis and, in any event,the patient subgroups so derived are unlikely to be com-parable. The main meta-analysis showed that the finaloutcome for patients treated by chemonucleolysis, in-cluding the effects of further surgery if chemonucleolysisfailed, remained poorer than for those treated by pri-mary discectomy.

No statistically significant differences were demon-strated between low-dose and standard-dose chymopa-pain,1 between chymopapain and collagenase,30 or be-

tween chymopapain and steroid injection.4,6 Although

Figure 1. Meta-analysis of chymopapainversusplacebo: successat 3 to 12 months as rated by independent surgical assessors.



7/13

one trial suggested that collagenase was more effectivethan placebo, that was a small study that had a 40%code break by 8 weeks.8

Lavignolle et al39 concluded that the initial shorterhospital stay produced little cost advantage in favor ofchemonucleolysis. A single study suggested that chemo-

nucleolysis was more expensive than discectomy, mainlybecause of the secondary costs associated with readmis-sion for surgery in those with failed chemonucleolysis.49

However, the average duration of hospital stay in thatstudy was 20 to 21 days for the primary treatment and50 days for reoperation. These figures would now be sounusual that the cost estimates in that study are proba-bly meaningless.

Discectomy (10 Trials)Only one trial directly compared surgical management oflumbar disc prolapse with any form of natural history,con-

servative management, or placebo. All the other trials ofdiscectomy compared two modes of invasive therapy. We-ber70,71 compared long-term outcomes of management bydiscectomy with initial conservative management followedby surgery if conservative therapy failed. The trial was notblinded, andthere was considerable crossover. Both patientand observer ratings showed that discectomy was signifi-cantly better than conservative therapy at 1 year, butthere were no significant differences in outcomes at 4 and10 years. Regardless of treatment, impaired motor functionhad a good prognosis, whereas sensory deficits remained inalmost half of the patients. Malter et al41 reanalyzed We-bers data and suggested that discectomy was highly costeffective, at approximately $29,000 per quality-adjustedlife year gained.

Three trials, with data from 219 patients, comparedmicrodiscectomy with standard discectomy. Use of themicroscope lengthened the surgical procedure, but didnot appear to make any significant difference in periop-erative bleeding or other complications, length of in-patient stay, or the formation of scar tissue. Henriksen31

and Lagarrigue38 provided clinical outcome data, butthese were not comparable and could not be pooled.However, clinical outcomes and duration of sick leavewith microdiscectomy were broadly similar to those with

standard discectomy.

Three trials, with data on 485 patients, considered theeffect that different types of interposition membrane hadon the formation of intraspinal scarring after discec-tomy, as assessed by magnetic resonance imaging or en-hanced computed tomography scan. One trial claimedless scarring and better clinical outcomes after insertionof an antiadhesion barrier gel, but the published resultswere highly selected.53 Meta-analysis suggested that

there might be some reduction in moderate to severe scarformation, but that this reduction did not reach statisti-cal significance. Meta-analysis showed no difference inclinical outcomes. The barrier did not influence the re-operation rate in the single trial reporting this data in avery small number of patients.53

Two trials compared automated percutaneous discec-tomy with microdiscectomy. These trials were not di-rectly comparable because one used modified forcepsand an automated cutter with suction,44 whereas theother used an automated suction nucleotome alone.10

Mayer et al44 showed that clinical outcomes were com-

parable with those of microdiscectomy, but consideredthat only 10% to 15% of patients needing surgical treat-ment might be suitable for automated percutaneous disc-ectomy. In contrast, Chatterjee et al10 reported only29% satisfactory results for automated percutaneousdiscectomy compared with 80% for microdiscectomy.Revel et al55 found automated percutaneous discectomyto be inferior to chemonucleolysis.

This review did not identify any published RCTs oflaser discectomy.

Degenerative Lumbar Spondylosis. Fourteen published

RCTs describing all forms of surgical management fordegenerative lumbar spondylosis were identified. Therewere no RCTs on the efficacy of any form of surgery fordegenerative lumbar spondylosis compared with naturalhistory, placebo, or any form of conservative manage-ment, and all of the trials compared two or more surgicaltechniques. Three abstracts were included for complete-ness, although they did not include full data.17,56,74 Anumber of other abstracts did not provide sufficient in-formation for proper evaluation (see the Cochrane re-views for details of trials under evaluation25,26).

From a surgical perspective, the trials fell into twobroad groups (Table 3): The first group considered tech-niques of dealing with spinal stenosis and nerve compres-sion, whereas the second considered techniques of fu-sion. In the first group, one trial compared differenttechniques of decompression for spinal stenosis, and fourtrials compared combinations of decompression with orwithout fusion. In the second group, nine trials com-pared fusion with or without instrumentation, and twotrials considered the role of electrical stimulation (directcurrent and pulsed electromagnetic stimulation) in pos-terolateral fusion. Two trials each compared three tech-niques and were included in both the first and secondgroups. Analysis was complicated by the fact that the

trials dealt with varied pathologies including degenera-

Figure 2. Meta-analysis of discectomyversuschymopapain: suc-cess at 1 year as rated by surgeons.



8/13

tive disc disease, spinal stenosis, isthmic spondylolisthe-sis, and degenerative spondylolisthesis. Some of the trialsdid not distinguish these and presented only total resultsfrom mixed pathologies. The techniques of fusion andinstrumentation also varied greatly.

Many trials provided limited information on selectedcomplications, but these were not comparable betweentrials. Two trials provided comparative information onoperating time and blood loss, and three trials provided

information on progression of spondylolisthesis. A costanalysis was performed in one trial,74 although themethod used in this trial has been severely criticized.28

Techniques for the Decompression of Spinal and

Nerve Root StenosisNo RCTs could be identified dealing with the efficacy ofsurgical decompression for degenerative lumbar spondy-losis or spinal stenosis. Readers are referred instead tometa-analyses of largely retrospective case studies of sur-gical management for spinal stenosis50,68 and for degen-erative spondylolisthesis.42

One trial considered techniques of decompression forspinal stenosis by comparing laminectomy with multiplelaminotomy.54 This trial had several confounding fac-tors. Nine of the 35 patients scheduled for laminotomyactually had a laminectomy for technical reasons, andseveral patients in each group also had an intertransversearthrodesis for degenerative spondylolisthesis. The trialdid not demonstrate any difference in clinical outcomesor spondylolisthesis progression between the two treat-ment methods.

Three trials considered whether some form of pos-terolateral fusion, with or without instrumentation, wasa useful adjunct to decompression alone.7,29,32 These tri-

als provided data on a total of 139 patients with 99%

follow-up at 2 to 3 years. Meta-analysis showed no dif-ference in outcomes between fusion of any type and lam-inectomy (OR, 0.80; 95% CI, 0.31, 2.10), as rated by thesurgeon at 18 to 24 months.

Grob et al29 considered fusion with and without in-strumentation in patients with degenerative spinal steno-sis with no evidence of instability, and found no differ-ence in clinical outcomes or relief of pain, provided theposterior elements were preserved at the time of opera-

tion. The other two trials considered the role of adjunctfusion in spinal stenosis associated with single or two-level degenerative spondylolisthesis.

Herkowitz et al32 studied noninstrumented fusion,showing that fusion produced significantly less self-reported back pain and leg pain and significantly bettersurgeon ratings of outcome. Bridwell et al7 studied bothinstrumented and noninstrumented fusion and showedthat patients who experienced a successful fusion hadless spondylolisthesis progression and a better patientrating of improvement. However, these results were sig-nificant only in the group with an instrumented fusion,

whereas noninstrumented fusion showed no such benefit.Carragee et al9 compared the results of fusion alone or

fusion plus laminectomy and decompression for patientswith an isthmic L5S1 spondylolisthesis. Their trial wasconfounded by the fact that nonsmokers had fusion bybone grafting alone, whereas smokers had their fusionsupplemented by instrumentation. In neither group didthe addition of decompression to the arthrodesis appearto improve the clinical outcomes.

Techniques of FusionNo RCTs could be identified dealing with the efficacy orclinical outcomes of surgical fusion for degenerative lum-

bar spondylosis and back pain. Readers are referred in-

Table 3. RCTs in Degenerative Lumbar SpondylosisSurgical Subgroups

StenosisStenosis

Spondylolisthesis Isthmic Spondylolisthesis Mixed Disease Degenerative Disc

Grob 19951. Decompression2. Facet screws3. CD/AO

Herkowitz 19911. Decompression2. Lateral Mass

Carragee 19971. Lateral mass

(nonsmokers)a) No decompressionb) Decompression

2.

CD (smokers)a) No Decompressionb) Decompression

Postacchini 19931. Laminectomy2. Multiple laminotomies

Zdeblick 19961. Decompression TSRH

2. TSRH PLIF

3.

ALIF

Unspecified

Bridwell 19931. Decompression2. Lateral mass3. Steffee

Thomsen 19971. Decompression

lateral mass2. CD

Zdeblick 19931. Decompression lateral mass2. Luque3. TSRH

Mooney 19901. Fusionvariable2. Electromagnetic

stimulationMcGuire 1993

1. Decompression lateral mass

2. Steffee

Emery 19951. Decompression lateral mass2. Luque

Rogozinski 19951. Lateral mass pedicle

screws2. Electrical stimulation

Fischgrund 19971. Decompression lateral mass

2. Steffee

France 19971. Decompression lateral mass2. Steffee

CD/AO Cotrel-Dubousset/AO fixation system; TSRH Texas Scottish Rite Hospital; PLIF posterior lumbar interbody fusion; ALIF anterior laparoscopicallyassisted interbody fusion.



9/13

stead to the systematic review of the largely retrospectivecase studies by Turner et al.67

Nine trials addressed the question of whether instru-mentation improved the outcome of posterolateral fu-

sion in a total of 649 patients with a mean 95% fol-low-up of 16 months to 4.5 years (median, 2 years).However, this was a very heterogeneous group of trials(Table 3). Only one trial dealt purely with patients whohad degenerative disc disease and back pain,75 althoughone other trial presented results separately for this sub-group of patients.73 One trial evaluated patients whohad degenerative lumbar spinal stenosis without insta-bility. Two trials assessed patients with isthmic spon-dylolisthesis, and another two trials involved degenera-tive spondylolisthesis. Three trials comprised patientswith mixed diagnoses, and only one of these presented

the results for each group separately.

73

A wide variety ofinstrumentation systems were used, and only three trialsused the same system (Steffee plates). There was a widevariety of outcome measures, the most common beingtechnical surgical outcomes, fusion rates, progression ofspondylolisthesis, and reoperation rates. It should benoted that the test for homogeneity was significant in allthe meta-analyses of this group of trials.

Meta-analysis of the data available from seven trialssuggested that solid fusion was more likely if instrumen-tation had been used (OR, 2.32; 95% CI, 0.95, 5.68;Figure 3). Six RCTs provided clinical outcome data oninstrumented as compared with noninstrumented fusion,including one study from the ongoing assessment list.46

Preliminary meta-analysis of all six trials suggested thatinstrumented fusion might produce slightly better clini-cal outcomes (OR, 1.36; 95% CI, 0.77, 2.42; Figure 4),although this difference did not reach statistical signifi-cance. However, only one trial73 showed a significantdifference in outcomes, and that contrasted with the re-sults of all the other trials. This trial was the only onebased entirely on crude outcome ratings by the operatingsurgeon, who claimed 92% excellent and good resultsfrom instrumented fusion. The other five trials all usedvarious patient-centered measures of pain or functional

status, and none showed any significant difference.

Meta-analysis of the 338 patients in these five trialsshowed no difference in pain or disability between instru-mented and noninstrumented fusion (OR, 1.05; 95% CI,0.68, 1.62). Because of the heterogeneity among the tri-als, it was not possible to draw any conclusions about theresults of instrumented fusion for any particular patho-logic condition, or about any particular instrumentationsystem. A single trial comparing anterior and posterior

interbody fusion75

showed no significant difference inany outcome in the small number of patients studied.

Two trials assessed whether electrical stimulationcould enhance fusion. Mooney et al48 found significantlyimproved fusion with pulsed electromagnetic stimula-tion in lumbar interbody fusion, but the improvement inclinical outcomes was not significant. In contrast, Ro-gozinski et al56 reported inconclusive results with im-planted direct current stimulation as an adjunct to in-strumented posterolateral fusion.

Discussion

These reviews identified 26 RCTs, but there were manyweaknesses of trial design and conclusions must bedrawn with caution.

Lumbar Disc ProlapseThe evidence supports chemonucleolysis as a possibletherapeutic option, as a minimally invasive, intermediatestage between conservative management and open sur-gical intervention, which has a lower complication rateand might save about 70% of patients from requiringopen surgery. However, the final outcome from chemo-nucleolysis, followed by surgery if chemonucleolysis

fails, remains poorer than the outcome from primarydiscectomy. It is therefore a matter of debate about therelative balance of possibly avoided surgery, relativerisks and complication rates, clinical outcomes over thenext year or so, and the potential impact on the lifetimenatural history of disc trouble.5

Only one trial directly compared surgical discectomywith conservative management.70,71 That trial is quotedwidely as a comparison of discectomy and conservativemanagement, and as showing that surgery produced a

Figure 3. Meta-analysis of fusion rates in trials of instrumentedfusion versusgraft alone (mixed disease).

Figure 4. Meta-analysis of good clinical outcomes in trials ofinstrumented fusion versusgraft alone (mixed disease).



10/13

temporary benefit in clinical outcomes at 1 year but nodifference on longer term follow-up at 4 and 10 years.The current authors believe that this is an inaccurateinterpretation of the results (see also Bessette et al2 for acritique of this trial). Webers70 study was actually anRCT of a subgroup of patients with uncertain indica-tions for surgery: Of all the 280 patients in this subgroup,67 were considered to have definite indications for sur-

gery; 87 patients improved with conservative manage-ment; and only the intermediate 126 patients were ran-domized in the RCT.

The intervention consisted of primary discectomycompared with initial conservative management fol-lowed by discectomy as soon as clinically considered nec-essary if the patient failed to improve. Of this conser-vative group, 26% actually came to surgery (i. e.,crossed over), although there was an intention-to-treatanalysis. The trial showed clearly that discectomy pro-duced better clinical outcomes at 1 year. It also showedthat if the clinical indications are uncertain, postponing

surgery for further assessment of clinical progress maydelay recovery but will not produce long-term harm.However, the current review produced other, better

evidence for the effectiveness of surgical discectomy.There is strong evidence showing that discectomy ismore effective than chemonucleolysis, and that chemo-nucleolysis is more effective than placebo; ergo, discec-tomy is more effective than placebo. This is entirely con-sistent with systematic reviews of non-RCT studies ondiscectomy33,60 and many clinical studies showing con-sistently that 65% to 90% of patients experience good orexcellent outcomes, particularly for the relief of sciatica

lasting at least 6 to 24 months, compared with 36% ofconservatively treated patients.33

It is not possible to draw any conclusions about indi-cations for surgery from the current review of RCTs, butthese other reviews33,60 demonstrate the need for carefulselection of patients. All of the evidence confirms clinicalexperience and teaching that the primary benefit of disc-ectomy is to provide more rapid relief of sciatica in pa-tients who have failed to resolve with conservative man-agement, even if there is no clear evidence whethersurgery alters the long-term natural history or prognosisof the underlying disc disease. Webers70 results suggestthat surgery makes no difference at 10 years, but the theresult of Gogan and Fraser27 suggest that better clinicaloutcomes are maintained at 10 years. Nevertheless,whatever the long-term effects, the short- to medium-term clinical outcomes after chymopapain have been suf-ficiently consistent for discectomy to survive the test oftime in widespread clinical practice for more than half acentury. There is still a lack of scientific evidence on theoptimal timing of surgery.

This review also provides evidence that the clinicaloutcomes of microdiscectomy are comparable with thoseof standard discectomy. In principle, the microscope pro-vides better illumination and facilitates teaching. These

trials suggest that use of the microscope lengthens the

surgical procedure, but despite previous claims, they didnot show any significant difference in perioperativebleeding, length of in-patient stay, or the formation ofscar tissue. However, it is not possible to draw any firmconclusions about complication rates from these trials.For the time being, unless or until better scientific evi-dence is available, automated percutaneous discectomyand laser discectomy should be regarded as re-

search techniques.Although a number of these trials reported the pro-

portion of patients who returned to work, it is not pos-sible to draw any firm conclusions about how effectiveany of these surgical treatments are for return to work.Scheer et al57 reached a similar conclusion from a sys-tematic review of RCTs. Taylor et al63 reviewed 19 un-controlled case studies and suggested that 82% of pa-tients were able to return to their previous level ofemployment after primary back surgery compared with59% after multiple back surgery.

There is a need for better evidence on the optimal

selection and timing of discectomy, on the possible roleof chemonucleolysis as a intermediate stage betweenconservative and surgical management, and on the rela-tive clinical outcomes, morbidity, costs, and cost effec-tiveness of microdiscectomy versus the standard proce-dure. High-quality RCTs are required to determinewhether there is any role for automated percutaneousdiscectomy or laser discectomy. There is a major need forlong-term studies concerning the effects of surgery on thelifetime natural history of disc disease.

Degenerative Lumbar Spondylosis

There is a serious lack of scientific evidence supportingsurgical management for degenerative lumbar spondylo-sis. This review of the published literature over the past30 years found only 14 RCTs that dealt entirely withquestions of alternative surgical techniques, and most ofthese reported relatively short-term, technical, surgicaloutcomes rather than patient-centered outcomes of pain,disability, and capacity for work. Currently, there is noadequate scientific evidence on more fundamental ques-tions about the efficacy or long-term effects of either sur-gical decompression or fusion. This is of particular con-cern given the magnitude of the clinical problem andthe number and costs of surgical procedures beingperformed.

The few and diverse trials on spinal stenosis and nervecompression permit very limited conclusions. The scien-tific evidence available currently does not permit anyfirm recommendations on the indications or techniquesof decompression.

Although there have been nine trials of instrumentedversusnoninstrumented fusion, these trials are very het-erogeneous both clinically and statistically, and any at-tempt to combine and interpret the results must be cau-tious and tentative. These trials dealt with very diversepathologic conditions and with different indications for

surgery. Moreover, the results were not always presented



11/13

separately for each subgroup. Most of the trials useddifferent instrumentation systems. Many of these trialswere limited by low-quality methods, inadequate ran-domization, lack of blinding, and potential for bias.Some of the trials were published in abstract form onlyand thus provide very little detailed data.

With these limitations in mind, in very general terms,instrumented fusion appears to lead to a higher fusion

rate, although the advantage is of borderline statisticalsignificance, and assessing fusion in the presence of met-alwork poses problems that few of these trials consid-ered.3,37 In general, instrumentation does not appear tomake any difference to clinical outcomes. It is not possi-ble to draw any conclusions from this review about therelative morbidity or complications, except that instru-mentation obviously is associated with unique complica-tions. On the evidence available currently, it is not pos-sible to draw any conclusions about the possible role ofinstrumented fusion for any particular pathologic condi-tion, or about any particular instrumentation system.

This review of surgery for degenerative lumbar spon-dylosis highlights the urgent need for scientific evidenceon the clinical efficacy and cost effectiveness of surgicaldecompression and/or fusion for specific pathologic andclinical syndromes associated with degenerative lumbarspondylosis. This will require high-quality RCTs thatcompare these surgical treatments with natural historyand conservative management.

Conclusions

What Evidence Exists on the Clinical Effectiveness ofLumbar Spine Surgery?

1.There is strong evidence (Strength A) that chemo-nucleolysis with chymopapain produces better clinicaloutcomes than placebo.2.There is considerable evidence (Strength A) of theclinical effectiveness of discectomy for carefully se-lected patients with sciatica caused by lumbar discprolapse. Discectomy provides faster relief from theacute attack (Strength A), although any positive ornegative effects on the lifetime natural history of discproblems are unclear (Strength C).3.There is no acceptable evidence (Strength D) of theefficacy of any form of decompression for degenera-

tive lumbar spondylosis or spinal stenosis.4.There is no acceptable evidence (Strength D) of theefficacy of any form of fusion for degenerative lumbarspondylosis, back pain, or instability.

What Evidence Exists on Alternative Forms andTechniques of Lumbar Spine Surgery?

1.There is strong evidence (Strength A) that surgicaldiscectomy produces better clinical outcomesthan chemonucleolysis.

2.There is moderate evidence (Strength B) that theclinical outcomes of microdiscectomy are comparable

with those of standard discectomy.

3.There is moderate evidence (Strength B) that au-tomated percutaneous discectomy produces poorerclinical results than standard discectomy or chemo-nucleolysis.

4. There is no acceptable evidence (Strength D) oflaser discectomy.

5. There is limited and inconclusive evidence(Strength C) of the relative efficacy of different doses

of chymopapain, chymopapain compared with colla-genase, and collagenase compared with placebo.

6. It is possible that some form of interposition mem-brane may produce a slight reduction in the formationof severe scar tissue after discectomy (Strength B).There is moderate evidence that clinical outcomes areunchanged (Strength B) and limited evidence(Strength C) that an interposition membrane does notinfluence the reoperation rate.

7. There is limited evidence (Strength C) that adjunctfusion to supplement decompression for degenerativespondylolisthesis produces less progressive slip and

better clinical outcomes than decompression alone.8.There is limited evidence (Strength C) that fusion

alone may be as effective as combined decompressionand fusion for patients with Grade I or II isthmicspondylolisthesis and no significant neurology.

9.There is strong evidence that instrumented fusionmay produce a higher fusion rate (Strength A, al-though that needs to be qualified by the difficulty ofassessing fusion), but does not improve clinical out-comes (Strength A), and other evidence (Strength D)that it may be associated with higher complica-tion rates.

10.There is limited evidence (Strength C) that there isno clear difference in clinical outcomes between ante-rior or posterior techniques of fusion.11. There is conflicting evidence (Strength C) thatsome forms of electrical stimulation may enhance fu-sion rates, but probably (Strength C) stimulation doesnot influence clinical outcomes.

The quality of surgical RCTs clearly still needs to beimproved, particularly in relation to adequate random-ization, blinding, duration of follow-up, and better clin-ical outcome measures. There are major gaps in knowl-edge concerning the costs, cost effectiveness, and work

outcomes of all forms of surgical management for lum-bar disc prolapse and degenerative lumbar spondylosis.The Cochrane reviews will be updated continuously asother trials become available. The authors will bepleased to receive information about any other RCTs onsurgical management for lumbar disc prolapse or degen-erative lumbar spondylosis.

Acknowledgments

The authors thank Professor W. J. Gillespie, Dr. HelenHandoll, and Mrs. Kathryn Quinn of the Department ofOrthopaedic Surgery, University of Edinburgh, for ad-

vice and support, Professor A. Nachemson for assistance



12/13

in the literature search and retrieval, Dr. M. Szpalski andProfessor M. Melot for collating the clinical outcomedata on instrumented versus noninstrumented fusion,and the editorial staff of the Back Review Group of theCochrane Database of Systematic Reviews.

References

1. Benoist M, Bonneville JF, Lassale B, et al. A randomized, double-blind study

to compare low-dose with standard-dose chymopapain in the treatment of her-

niated lumbar intervertebral discs. Spine 1993;18:2834.

2. Bessette L, Liang MH, Lew RA, Weinstein JN. Classics in spine surgery

revisited. Spine 1996;21:259 63.

3. Blumenthal SL, Gill K. Can lumbar spine radiographs accurately determine

fusion in postoperative patients? Correlation of routine radiographs with a sec-

ond surgical look at lumbar disc fusions. Spine 1993;18:11869.

4. Bontoux D, Alcalay M, Debiais F, et al. [Treatment of lumbar disk hernia by

intra-disk injection of chymopapain or triamcinolone hexacetonide: Compara-

tive study of 80 cases] Traitement des hernies discales lombaires par injection

intradiscale de chymopapaine ou dhexacetonide de triamcinolone: Etude com-

paree de 80 cas. Rev Rhum Mal Osteoartic 1990;57:32731.

5. Bouillet R. Complications of discal hernia therapy: Comparative study re-

garding surgical therapy and nucleolysis by chymopapain. Acta Orthop

Belg 1983;49(Suppl):4877.

6. Bourgeois P, Benoist M, Palazzo E, et al. [Multicenter randomized double-

blind study of triamcinolone hexacetonide versus chymopapain in the treatmentof disk lumbosciatica: Initial results at 6 months] Etude en double aveugle ran-

domisee multicentrique de lhexacetonide de triamcinolone versuschymopapaine

dans le traitement de la lombosciatique discale: Premiers resultats a six mois. Rev

Rhum Mal Osteoartic 1988;55:7679.

7. Bridwell KH, Sedgewick TA, OBrien MF, Lenke LG, Baldus C. The role of

fusion and instrumentation in the treatment of degenerative spondylolisthesis

with spinal stenosis. J Spinal Disord 1993;6:46172.

8. Bromley JW, Varma AO, Santoro AJ, Cohen P, Jacobs R, Berger L. Double-

blind evaluation of collagenase injections forherniatedlumbardiscs.Spine1984;

9:4868.

9. Caragee EJ. Single-levelposterolateral arthrodesis, with or without posterior

decompression, for the treatment of isthmic spondylolisthesis in adults: A pro-

spective, randomised study. J Bone Joint Surg [Am] 1997;79:117580.

10. Chatterjee S, Foy PM, Findlay GF. Report of a controlled clinical trial com-

paring automated percutaneous lumbar discectomy and microdiscectomy in the

treatment of contained lumbar disc herniation. Spine 1995;20:7348.11. Choy DS, Ascher PW, Saddekni S, et al. Percutaneous laser disc decompres-

sion: A new therapeutic modality. Spine 1992;17:94956.

12. Crawshaw C, Frazer AM, Merriam WF, Mulholland RC, Webb JK. A com-

parison of surgery and chemonucleolysis in the treatment of sciatica: A prospec-

tive randomized trial. Spine 1984;9:1958.

13. CSAG Clinical Standards Advisory Group Report of Back Pain. London:

HMSO, 1994:6572.

14. Dabezies EJ, Langford K, Morris J, Shields CB, Wilkinson HA. Safety and

efficacy of chymopapain (Discase) in the treatment of sciatica due to a herniated

nucleus pulposus: Results of a randomized, double-blind study. Spine 1988;13:

5615.

15. Deyo RA, Battie M, Beuerskens AJHM, et al. Outcome measures for low

back pain research: A proposal for standardized use. Spine 1998;23:200313.

16. Ejeskar A, Nachemson A, Herberts P, et al. Surgery versus chemonucleolysis

for herniated lumbar discs: A prospective study with random assignment. Clin

Orthop 1983;174:236 42.17. Emery SE, Stephens GC, Bolesta MJ, et al. Lumbar fusion with and without

instrumentation: A prospective study. Orthop Trans 1995;19:362.

18. Feldman J, Menkes CJ, Pallardy G, et al. [Double-blind study of the treat-

ment of disc lumbosciatica by chemonucleolysis] E tude en double-aveugle du

traitement de la lombosciatique discale par chimionucleolyse. Rev Rhum Mal

Osteoartic 1986;53:14752.

19. Fischgrund J, Mackay M, Herkowitz H, Brower R, Montgomery D, Kurz L.

Degenerative lumbar spondylolisthesis with spinal stenosis: A prospective, ran-

domized study,comparing decompression andfusionwith andwithout posterior

pedicular instrumentation. Orthop Trans 1997;21:158.

20. Fischgrund JS, Mackay M, Herkowitz HN, Brower R, Montgomery DM,

Kurz LT. Degenerative lumbar spondylolisthesis with spinal stenosis: A prospec-

tive, randomized study comparing decompressive laminectomy and arthrodesis

with and without spinal instrumentation. Spine 1997;22:280712.

21. France JC, Yaszemski MJ, Lauerman WC, et al. A randomized prospective

study of posterolateral lumbar fusion: Outcomes with and without pedicle screw

instrumentation. Spine 1999;24:55360.

22. Fraser RD.Chymopapain for thetreatment of intervertebral disc herniation:

A preliminary report of a double-blind study. Spine 1982;7:60812.

23. Fraser RD.Chymopapain for thetreatment of intervertebral disc herniation:

The final report of a double blind study. Spine 1984;9:8158.

24. Fraser RD, Sandhu A, Gogan WJ. Magnetic resonance imaging findings 10

years after treatment for lumbar disc herniation. Spine 1995;20:71014.

25. Gibson JNA, Grant IC, Waddell G. Surgery for lumbar disc prolapse (Co-

chrane review). In: the Cochrane Library, Issue 1, 1999 Oxford: Up-

date Software.

26. Gibson JNA, Waddell G, Grant IC. Surgery for degenerative lumbar spon-

dylosis (Cochrane review). In: the Cochrane Library, Issue 1, 1999 Oxford:

Update Software.

27. Gogan WJ, Fraser RD. Chymopapain: A 10-year, double-blind study.

Spine 1992;17:38894.

28. Goosens M, Evers S. Economic evaluation of back pain interventions. In:

Nachemson A, Jonsson E, eds. SBU Report on Back Pain. Stockholm: Swedish

Council on Technology Assessment in Health Care, 1999.

29. Grob D, Humke T, Dvorak J. Degenerative lumbar spinal stenosis: Decom-

pression with and without arthrodesis. J Bone Joint Surg [Am] 1995;77:1036

41.

30. Hedtmann A, Fett H, Steffen R, Kramer J. [Chemonucleolysis using chymo-

papain and collagenase: 3-year results of a prospective randomized study] Che-

monukleolyse mit Chymopapain und Kollagenase. 3-Jahres-Ergebnisse einer

prospektiv-randomisierten Studie. Z Orthop Ihre Grenzgeb 1992;130:3644.

31. Henrikson L, Schmidt V, Eskesen V, Jantzen E. A controlled study of micro-

surgical versus standard lumbar discectomy. Br J Neurosurg 1996;10:28993.

32. Herkowitz HN, Kurz LT. Degenerativelumbar spondylolisthesis with spinalstenosis: A prospective study comparing decompression with decompression and

intertransverse process arthrodesis. J Bone Joint Surg [Am] 1991;73:8028.

33. Hoffman RM, Wheeler KJ, Deyo RA. Surgery for herniated lumbar discs: A

literature synthesis. J Gen Intern Med 1993;8:48796.

34. Javid MJ, Nordby EJ, Ford LT, et al. Safety and efficacy of chymopapain

(chymodiactin) in herniated nucleus pulposus with sciatica. JAMA 1983;249:

248994.

35. Jensen T, Asmussen K, Berg-Hansen E, et al. First-time operation for lumbar

disc herniation with or without free fat transplantation: Prospective triple-blind

randomized study with reference to clinical factors and enhanced CT scan 1 year

after operation. Spine 1996;21:10726.

36. Kahanovich N, Viola K, McCulloch JA. Limited surgical discectomy and

microdiscectomy: A clinical comparison. Spine 1989;14:7981.

37. Kant AP,Daum WJ, Dean SM,UchidaT. Evaluation of lumbarspine fusion:

Plainradiographsversus directsurgical exploration and observation.Spine 1995;

20:23137.38. Lagarrigue J, Chaynes P. Comparative study of disk surgery with or without

microscopy: A prospective study of 80 cases. Neurochirurgie 1994;40:11620.

39. Lavignolle B, Vital JM, Baulny D, Grenier F, Castagnera L. [Comparative

study of surgery and chemonucleolysis in the treatment of sciatica caused by a

herniateddisk]Etudescomparees de la chirurgieet de la chimionucleolyse dans le

traitement de la sciatique par hernie discale. Acta Orthop Belg 1987;53:2449.

40. MacKay MA, Fischgrund JS, Herkowitz HN, Kurz LT, Hecht B, Schwartz

M. The effect of interposition membrane on the outcome of lumbar laminectomy

and discectomy. Spine 1995;20:17936.

41. Malter AD, Larson EB, Urban N, Deyo RA. Cost effectiveness of lumbar

discectomy for the treatment of herniated inter vertebral disc. Spine 1996;21:

1048955.

42. Mardjetko SM, Connolly PJ, Shott S. Degenerative lumbar spondylosis: A

meta-analysis of literature 19701993. Spine 1994;20(Suppl):S225665.

43. Martins AN, Ramirez A, Johnston J, Schwetschenau PR. Double-blind eval-

uation of chemonucleolysis for herniated lumbar discs: Late results. J Neuro-surg 1978;49:81627.

44. Mayer HM, Brock M. Percutaneous endoscopic discectomy: Surgical tech-

nique and preliminary results compared to microsurgical discectomy. J Neuro-

surg 1993;78:21625.

45. McGuire RA, Amundson GM. The use of primary internal fixation in spon-

dylolisthesis. Spine 1993;18:166272.

46. MollerH, Hedlund R. Fusionor conservativetreatment in adult spondylolis-

thesis: A prospective randomized study. Acta Orthop Scand 1994;65(Suppl):12.

47. MollerH, Hedlund R. Fusionor conservativetreatment in adult spondylolis-

thesis: A prospective randomized study. Acta Orthop Scand 1996;69(Suppl):13.

48. Mooney V. A randomized double-blind prospective study of the efficacy of

pulsed electromagnetic fields for interbody lumbar fusions. Spine 1990;15:708

12.

49. Muralikuttan KP, Hamilton A, Kernohan WG, Mollan RA, Adair IV. A

prospective randomized trial of chemonucleolysis and conventional disc surgery

in single-level lumbar disc herniation. Spine 1992;17:3817.

50. Niggemeyer O, Strauss JM, Schulitz KP. Comparison of surgical procedures



13/13

for degenerative lumbar spinal stenosis: A meta-analysis of the literature from

1975 to 1995. Eur Spine J 1997;6:4239.

51. North RB, Kidd DH, Piantadosi S. Spinal cord stimulation versus reopera-

tion for failed back surgery syndrome: A prospective randomized study design.

Acta Neurochir 1995;64:106 8.

52. Onik G, MooneyV, MaroonJC, et al.Automated percutaneousdiscectomy:

A prospective multi-institutional study. Neurosurgery 1990;2:22833.

53. Petrie JL, Ross JS. Use of ADCON-L to inhibit postoperative peridural fi-

brosis and related symptoms following lumbar disc surgery: A preliminary re-

port. Eur Spine J 1996;5(Suppl):S107.

54. Postacchini F, Cinotti G, Perugia D, Gumina S. The surgical treatment of

central lumbarstenosis: Multiple laminotomy compared with total laminectomy.J Bone Joint Surg [Br] 1993;75:386 92.

55. Revel M, Payan C, Vallee C, et al. Automated percutaneous lumbar discec-

tomy versus chemonucleolysis in the treatment of sciatica. Spine 1993;18:17.

56. Rogozinski A, Rogozinski C. Efficacy of implanted bone growth stimulation

in instrumented lumbosacral spinal fusion. Orthop Trans 1995;19:362.

57. Scheer SJ, Radack KL, OBrien DR. Randomised controlled trials in indus-

trial low back pain relating to return to work. Part 2: Discogenic low back pain.

Arch Phys Med Rehabil 1996;77:118997.

58. Schwetschenau PR, Ramirez A, Johnston J, Barnes E, Wiggs C, Martins AN.

Double-blind evaluation of intradiscal chymopapain for herniated lumbar discs:

Early results. J Neurosurg 1976;45:6227.

59. Smith L. Enzyme dissolution of nucleus pulposus in humans. JAMA 1964;

187:13740.

60. Stevens CD, Dubois RW, Larequi-Lauber T, Vader J-P. Efficacy of lumbar

discectomy and percutaneous treatments for lumbar disc herniation. Soz-

Pravedtivmed 1997;42:36779.61. Stevenson RC, McCabe CJ, Findlay AM. An economic evaluation of a clin-

ical trial to compare automated percutaneous lumbar discectomy with microdis-

cectomy in the treatment of contained lumbar disc herniation. Spine 1995;20:

73942.

62. Szpalski M, Gunzburg R, eds. LumbarSegmentalInstability: Fact or Fiction?

Proceedings of the Fourth Brussels International Spine Symposium, Brussels,

Belgium, November 2122, 1997.

63. Taylor ME. Return to work following back surgery: A review. Am J Ind

Med 1989;16:7988.

64. Thomsen K, Christensen FB, Eiskjaer SP, Hansen ES, Fruensgaard S, Bunger

CE.The effect of pedicle screw instrumentationon functional outcome andfusion

rates in posterolateral lumbar spinal fusion: A prospective randomized clinical

study. Spine 1997;22:281322.

65. Tullberg T, Isacson J, Weidenhielm L. Does microscopic removal of lumbar

disc herniation lead to better results than the standard procedure? Results of a

one-year randomized study. Spine 1993;18:247.

66. Tullberg T, Rydberg J, Isacsson J. Radiographic changes after lumbar disc-

ectomy: Sequential enhanced computed tomography in relation to clinical obser-

vations. Spine 1993;18:84350.

67. Turner JA, Ersek M, Herron L, et al. Patient outcomes after lumbar spinal

fusions. JAMA 1992;268:90711.

68. Turner JA, Ersek M, Herron L, Deyo R. Surgery for lumbar spinal stenosis:

Attempted meta-analysis of the literature. Spine 1992;17:18.

69. van Alphen HA, Braakman R, Bezemer PD, Broere G, Berfelo MW. Chemo-

nucleolysis versus discectomy: A randomized multicenter trial. J Neuro-surg 1989;70:86975.

70. Weber H. Lumbar disc herniation: A controlled, prospective study with ten

years of observation. Spine 1983;8:13140.

71. Weber H. Lumbardisc herniation: A Prospective Study of Prognostic Factors

Including a Controlled Trial. University of Oslo, MD Thesis, 1978.

72. Wittenberg RH, Oppenl SC, Steffen R. Intradiscal therapy with chymopa-

pain or collagenase: Five-year results of a prospective randomized investigation.

Orthop Trans 19961997;20:10645.

73. Zdeblick TA. A prospective, randomized study of lumbar fusion: Prelimi-

nary results. Spine 1993;18:98391.

74. Zdeblick TA, Ulschmid S. An outcomes and cost analysis of pedicle screw

fusions. Orthop Trans 19967;20:756.

75. Zdeblick TA, Ulschmid S, Dick JC. The surgical treatment of L5S1 degen-

erative disc disease: A prospective randomized study of laparoscopic fusion.

Orthop Trans 19967;20:75.

Address reprint requests to

J. N. A. Gibson, MD, FRCSPrincess Margaret Rose Orthopaedic Hospital

Edinburgh EH10 7EDScotland

E-mail: [email protected]


Fusion Lombaire

Documents

Transcript of Fusion Lombaire