WIXLIB

Evidence Based MedicineinSpinal SurgeryRoger Härtl, MDAssociate Professor of NeurosurgeryChief of Spinal SurgeryNeurosurgeon to the GIANTS Football teamBrain & Spine CenterWeill Cornell Medical CollegeNew York, NY, USA

Disclosure– Synthes Consultant– Baxter Fellowship funding / Consultant– Spinewave Consultant– BrainLab Consultant– Lanx Consultant– AOSpine AOSpine NA Board Fellowship funding Research funding

Back Pain Degenerative Discdisease Disc herniation Spinal stenosis Congenital anomalies– Spondylolisthesis “Slippage” Trauma– Sprains and Strains– Fractures Facet-joint pain Sacro-iliac joint pain Neoplasm, infection,referred pain

Dr. Rauschning

Types of low back pain Radicular pain; Sciatica– Herniated disc– Foraminal stenosis Neurogenic claudication– spinal stenosis Chronic low back pain– DDD– “instability” Referred pain: hips, knees

OrthopedicsFor 95%of ySurgeryCornellSpineCenterPhysiatry /PainManagementComplementaryMedicineNeurology

Non-operativemanagement At least 6 weeks Limited bedrest, early mobilization Exercises / PT– Aerobic, stretching, isometric

100 patients with low backpain

100 patients with low backpain

100 patients with low backpainOROR

What happens if nonoperative treatment fails?

The Role of Surgery- Short answer -Neck / back painRadiculopathy pain goingdown the legMyelopathy /Cauda equina Spinal cord ornerve injury

The Role of Surgery- Short answer - Factors that favor Surgery– Clinical findings and MRI findings fit– Failure of non-operative treatment– Severe Pain– Neurological deficit Weakness Bowel / bladder incontinence– Leg or arm pain or weakness

What is Evidence BasedMedicine?The use of clinical methods anddecision making that have beenthoroughly tested by properlycontrolled peer-reviewedmedical research.

Evidence Based MedicineClass I– Prospective, Randomized, Controlled TrialsClass II– Non-Randomized, Prospective ControlledTrials– Observational StudiesClass III– Case Series– Case Reports– Expert Opinion

Evidence Based MedicineLevel 1 Recommendation–Class I EvidenceLevel 2 Recommendation–Class II EvidenceLevel 3 Recommendation–Class III Evidence

Treatment AStudy SampleOutcome ARandomizationTreatment BOutcome BDiagram demonstrating the basic design of a RANDOMIZED CONTROLLED TRIAL. Thestudy sample is randomized to different treatments, and the outcomes are prospectivelydetermined.The Importance of Study Design in the Spine Literature, Pearson A.,et.al.Seminars in Spine Surgery, Vol 21 (4), Dec 2009

Treatment AOutcome ATreatment BOutcome BDiagram demonstrating the basic design of an OBSERVATIONAL COHORT STUDY. Thetreatment is chosen by the patient and physician rather than through randomization. Thestudy groups are defined by treatment and outcomes are compared. Cohort studies canbe prospective or retrospective.The Importance of Study Design in the Spine Literature, Pearson A. et.al.Seminars in Spine Surgery, Vol 21 (4), Dec 2009

Evidence Based Medicine:Spine Lumbar HNP (3)Lumbar stenosis (3)Degenerative spondylolisthesis (5)Axial LBP (4)Artificial disc (3)

LUMBARHERNIATEDNUCLEUSPULPOSUS

Case example: HerniatedLumbar Disk 43 y/o male with 2 months ofpain radiating into right foot No relieve by physical therapyand medication No back pain, weakness orurinary incontinence No “red flags”

S1L5L4L3

EBM-Spine: Lumbar HNPPeul WC et al NEJM 2007DesignMulticenter RCT with ITT AnalysisPatients141 Patients- early surgery (median: 1.9 weeks)142 Patients- non-operative managementResultsEarly surgery resulted in faster recoveryNo difference in outcomes in 1 yearLimitationsHigh cross-over rates11% of surgery Æ conservative39% of conservative Æ surgeryBlinding not possibleFollow-up only 1 year

EBM Spine: Lumbar HNPAtllas SJ, et. al. Spine 2005 (Maine Lumbar Spine Study)DesignProspective Cohort Study (observational)Patients235 Surgery272 ConservativeResultsSurgery: Improved in pain, function and satisfactionoutcomes at 1, 5 and 10 years.No difference in work status, surgery vs. conservative.Benefit of surgery narrowed between the two groups overtime but still statistically different at 10 years.LimitationsImaging not requiredMail in questionnaire rather than actual clinical exam.

EBM Spine: Lumbar HNPSpine Patient Outcomes Trial (SPORT)Weinstein JN , et.al. JAMA 2006, Spine 2008Design2 Combined Trials (Due to protocol non-adherence)RCT- 501 PatientsObservational Cohort- 743Patients1244 totalResultsSurgery resulted in greater improvement comparedwith non-operative treatment at 4 years.LimitationsCross over (40% of surgery group, 45% of nonoperative). This precluded meaningful analysis of thedata on an ITT basis because the 2 groups were verysimilar in treatment received at 2 years.

LUMBARSTENOSIS

EBM Spine: StenosisThe Finnish Spinal Stenosis StudySimotas A.C., Clin. Orthopedic Relat Res 2001DesignRCT with ITT AnalysisPatients94 Patients, (50 Surgical, 44 Non-surgical)ResultsSurgery better in ODI, leg and back pain. Greater difference at 1year than at 2 yearsCrossover rate 10% (low) in either direction.Level I evidence favoring surgery but not in walking abilityLimitationsSmall number of patients20% of surgery group had instrumented fusion (variation insurgical management)

EBM Spinal: StenosisMaine Lumbar Study Atlas SJ et al, Spine 2005DesignProspective observational Cohort10 year follow-upPatients148 Patients- (81 Surgical, 67 Nonsurgical)ResultsLevel 2 evidence that decompression MAY provide betteroutcomes over nonsurgical treatment.LimitationsCross over to surgery 39%Non-randomized: more severe patients to surgery.Few patients with mild symptoms were treated with surgery

EBM Spine: StenosisSport Trial for Lumbar Spinal StenosisWeinstein J, et. al., NEJM 2008, Spine 2010DesignRCT with prospective observational CohortPatients654 Patients (289 RCT, 365 Observational)ResultsLevel 2 evidence to suggest that surgery results in betteroutcome at 2 years and maintained at 4 years.LimitationsHigh cross over– 33% of surgery group to non-surgery group– 43% from non-surgery group had surgerySurgical treatment variable (11% had a fusion)Non-surgical treatment not specified

DEGENERATIVESPONDYLOLISTHESIS

EBM:Degenerative SpondylolisthesisSurgical vs. Nonsurgical Treatment for LumbarDegenerative SpondylolisthesisWeinstein J. et. al. NEJM 2007, JBJS 2009DesignRCT with prospective observational cohort(304 RCT, 303 Observational Cohort)Patients521 Patients Follow-up, (372 Surgery, 149 No-surgery)ResultsSurgery patients (laminectomy with 1 level fusion) hadsubstantially greater pain relief and improvement in function at 4years.LimitationsHigh level of cross over, difficult to interpret ITT analysis36% of surgery group, 49% of non-operative groupNon-operative treatment not standardizedSurgical treatment not standardized(fusion posteriorly or circumferentially with or withoutinstrumentation)

EBM:Degenerative SpondylolisthesisThe Surgical Management of Degenerative LumbarSpondylolisthesis: A Systemic Review.Martin CR et.al. Spine 2007DesignLiterature Review: RCT and comparative observational studies inEnglish, German and French (1966-2005)Patients13 Studies of 578 patientsResultsFusion is more effective than laminectomy in achieving asatisfactory outcomeInstrumentation increased fusion rateDecompression only had the least satisfactory outcomeLimitationsSome studies included non-consecutive patientsSome had undefined follow-upNo standardized outcome measure was used consistentlyStrenghtsComprehensive review on degenerative spondylolisthesis

EBM: Degenerative Spondylolisthesis”Degenerative Lumbar Spondylolisthesis with SpinalStenosis” Kornblum, et.al. Spine 2008DesignA Prospective Long Term Study “Comparing Fusion and Pseudoarthrosis”Patients58 Patients with laminectomy and non-instrumented fusionResultsGood or excellent outcome in86% fusion56% non-union25/47 (53%) developed non-unionStrengthsFollow-up was long (5-14 years)LimitationsSmall numberNon-standardized outcome measure19% (11 patients) lost to follow-upSingle center, secondary analysis

EBM:Degenerative SpondylolisthesisSurgical Treatment of Spinal Stenosis withSpondylolisthesis: Cost Effectiveness after 2 yearsTosteson AN et al, Ann Internal Medicine 2008DesignProspective Cohort StudyPatients601 Patients (randomized and observational cohort)368 Surgery (fusion in 93% / 78% instrumentation)233 Non-surgeryResultsA trend toward improved cost effectiveness with circumferentialinstrumented fusionSurgery results in better improvement of healthStrengthsMulticenter studyLarge number of patientsRCT and observational patientsValidated outcome measure usedLimitationsNon-operative care not specifiedCosts relied upon self-reported utilization dataFollow-up limited to 2 years

AXIALLOW BACK PAIN

EBM: Axial Low Back PainLumbar Fusion Versus Treatment of Chronic Low Back Pain:A Multicenter Randomized Controlled Trial From the SwedishSpine Study Group Fritzell P, et al Spine 2001DesignMulticenter RCT with 2 year follow, ITT AnalysisPatients292 (Fusion 222, Non-operative 72)ResultsFusion may lead to better outcomeStrengthsMulticenter RCT small dropout (5 patients)LimitationsNo standardization in either groupIndustry fundingAsymmetry of group sizes 75 (due to design as multiple fusionarms)

EBM: Axial Low Back PainRandomized Controlled Trial to Compare SurgicalStabilization of the Lumbar spine with IntensiveRehabilitation for Patients with Chronic Low Back Pain:The MRC Spine Stabilization Trial. Fairbank et al BMJ 2005DesignMulti-centre RCT with 2 year follow-upPatients349 Patients (179 Surgery, 170 Rehab)ResultsImprovement above rehabilitation in ODI (4.1) with surgery (barelystatistically significant)StrengthsMulticenter RCTMultiple outcome measures (ODI, walking test, SF36, work status)LimitationsHigh crossover (28% non-operative to rehabilitation)Included redo’s and spondylolisthesisFlexible stabilsations included as fusion

EBM: Axial Low Back PainRandomized Clinical Trial of Lumbar Instrumented Fusionand Cognitive Intervention in Patients with Chronic Low BackPain and Disc Degeneration. Brox et al Spine 2003DesignRCT with 1 year follow-upPatients64 patientsResultsBoth groups improved significantly and equallyStrengthsBlinding of physical therapy evaluatorStandardized nonsurgical treatmentLimitationsShort follow-upSmall numbersLack of no treatment armFailure of treatment in assigned group (4/37 of the surgery groupand 2/27 in non-surgical)

EBM Spine: Axial LBPLumbar Instrumented Fusion Compared with Cognitive Interventionand Exercises in Patients with Chronic Low Back Pain After PreviouSurgery for Disc Herniation: A Prospective Randomized ControlledStudy. Brox et al Pain 2006DesignNationwide (Norway) RCT with 1 year follow-up and ITT AnalysisPatients60 PatientsResultsNo DifferenceStrengthsRCTValidated outcome measuresBlinding of PT evaluatorLimitationsShort follow-up (1year)Small numbersLack of no treatment arm7/29 Did not have surgery, 2/31 Did not have non-surgical

ARTIFICIALDISC

EBM: Artifical DiscResults of Prospective, Randomized, Multicenter Food & DrugAdministration Investigational Device Exemption Study of Pro Disc-LTotal Disc Replacement vs. Circumferential Fusion for the Treatmentof 1 Level Degenerative Disc Disease Zigler et al Spine, 2007DesignMulticenter randomized controlled non-inferiority trial with 2 year follow-upPatients161 pro-disc patients, 75 fusion patientsResultsStudy suggests that short-term outcomes of artificial disc replacements aresimilar to or marginally better than fusionStrengthsRCTLimitationsIndustry fundingAbout 10% lost to follow-up in each groupFDA revision of success criteria