Proceedings of the NASS 19th Annual Meeting / The Spine Journal 4 (2004) 3S–119S 69S

PATIENT SAMPLE: There were 12 men and 9 women. 10 and 11 patientswere operated on each of the levels the L4–L5 and L5-S1 respectively.Data over the 95% confidence interval for the accuracy of the RSA examina-tions was in rotation 2,01, 0,64 and 0,80 degrees and for translations 0,70,0,45 and 0,68 mm for the three planes. Median age at surgery was 36,2years (range 20–59).OUTCOME MEASURES: Radiostereometric analysis (RSA) was per-formed at discharge and 5 years postoperatively in supine and standingposition.METHODS: The relative rotations and translations in relation to the 3cardinal axes were calculated. Motions between the supine and standingpositions were calculated, i.e. the inducible movement over the segment.RESULTS: L4–L5-level. No differences in inducible movements werefound between the operated and non-operated controls at discharge or atthe 5-year follow up. L5-S1-level. At discharge the inducible forwardtranslation of L5 on sacrum was greater in the operated segments thanin the control segments (p�0.02, Man-Whitney’s test). After 5 years thisdifference was numerically about the same, but did not reach statisticalsignificance (p�0.09, Man-Whitney’s test). After 5 years the axial compres-sion over the operated L5-S1 segments had decreased with about 1 mm(p�0.02, Wilcoxon signed ranks test).CONCLUSIONS: The reason for increased anterior slipping of L5 betweensupine and standing after lumbar discectomy is not known. It could be thatdecreased tension due to removal of disc material and the partial resectionof the interspinal ligaments are responsible. At 5 years this differencepersisted but was not statistically significant perhaps because of a type IIerror. Nonetheless this increased anterior motion was small suggesting acomparatively wide interindividual variability. Lumbar discectomy mayincrease the rate of degenerative changes over time. We found that theinducible axial compression between L5 and S1 decreased with time, whichsupports this hypothesis. Our observation suggests that the disc might havelost more of its elasticity over the 5 years of observation compared tothe control as a sign of a higher degree of degeneration. In the controls therewas, however no reduction at all over 5 years suggesting that age relateddisc degeneration normally is a slow process, which perhaps does notstart until the patients have reached a higher age than included in our patientpopulation. Standard discectomy in patients with lumbar disc herniationhas small influence on the segmental motion over a 5-year period.DISCLOSURES: No disclosures.CONFLICT OF INTEREST: No Conflicts.

doi: 10.1016/j.spinee.2004.05.135

P51. Early diagnosis of pediatric spondylolysis: significance of MRIand its biomechanical corroborationKoichi Sairyo1, Yoichiro Takata2, Shinsuke Katoh2, Natsuo Yasui2, VijayGoel1, Sasidhar Vadapalli1, Akiyoshi Masuda1, Ashok Biyani3, NabilEbraheim3; 1University of Toledo, Toledo, OH, USA; 2University ofTokushima, Tokushima, Japan; 3Department of Orthopedic Surgery,Medical College of Ohio, Toledo, OH, USA

BACKGROUND CONTEXT: Since early stage spondylolysis can achieveosseous healing with conservative treatment, it is important to diagnose thisdisorder as early as possible. Presently, there is no well established andreliable diagnostic tool for the early diagnosis.PURPOSE: In this study, we present the significance of high signal changesof pedicles on T2 weighted MRI for its early diagnosis, and discuss thepathomechanism of the signal changes based on the stress data from thefinite element analyses of a spinal segment.STUDY DESIGN/SETTING: Review of MRI and biomechanical analyses.PATIENT SAMPLE: Thirty-seven pediatric patients with spondylolysiswere included.Sixty-eight defectswere examined and their stagesas revealedon CT scans were recorded.OUTCOME MEASURES: Not applicable.METHODS: High signal changes (HSC) of the pedicles on axial T2-weighted MRI were also compared to the CT based stages of the defect. Usinga 3-demensional non-linear finite element model of the L3–5 segment,

Fig. 1. CT scan and T2-MRI of early stage spondylolysis.

stress distributions in the pars and pedicle regions were analyzed in responseto 400 N compression and 10.6 Nm moment. Stresses were correlated tothe high signal changes on the MRI.RESULTS: Based on CTs, 68 pars defects were classified as follows: 8very early, 24 late early, 16 progressive and 20 terminal stages. The defectsin the very early stages showed faint and discontinuous hairlines at thepars.Aclear hairline was visible in the lateearly stage, andgap wassignificantfor defects in the progressive stage. The terminal stage defects indicatedpseudoarthrosis. All defects in very early- and late early-stages (100%)showed HSC on T2-weighted MRI at the ipsilateral pedicle (The rightpedicle in Figure). Among 16 progressive-stages, eight (50%) showed HSC,while no defects of the terminal stage (0%) were found to have HSC. Stressresults from FE model indicated that pars interarticularis showed thehighest value in all loadingmodes, and thepedicle showed the secondhighest.Especially, in the axial rotation, the stresses in the pedicle region wereclose to the data in the pars interarticularis region.CONCLUSIONS: The correlation between the high stresses in the pedicleand the corresponding HSC on MRI suggest that signal changes in MRIcould be used as an indicator for early diagnosis of spondylolysis. The HSCof the pedicle provided useful information to diagnose the spondylolysis inthe early stage, even when these defects were not clearly identifiable on CTs.MRIs, as proposed in this study for pediatric patients, can be used as areliable tool to diagnose the early stage spondylolysis. Simultaneously, thesame MRIs will enable the detection of two other major pediatric lumbardisorders, herniated nucleus pulposus and apophyseal ring tear.DISCLOSURES: Device or drug: Status: Approved for this indication.Device or drug: Status: Approved for this indication. Device or drug: Status:Approved for this indication.CONFLICT OF INTEREST: No conflicts.

doi: 10.1016/j.spinee.2004.05.136

P46. Painful postoperative neuritis is common after lumbardecompression in patients with hepatitis B and CNicholas Ahn, MD1, Uri Ahn, MD2, Jason Datta, MD3, Brian Ipsen,MD3, Harpreet Basran, MD3, Zachary Post3, Thomas Salsbury3, WilliamReed, Jr., MD4, Alexander Bailey, MD4, William Hopkins, MD4, GlennAmundson, MD4; 1Heartland Hand and Spine Orthopaedic Center/University of Missouri-Kansas City, Overland Park, KS, USA; 2NewHampshire Spine Institute, Bedford, NH, USA; 3University of Missouri-Kansas City, Kansas City, MO, USA; 4Heartland Hand and SpineOrthopaedic Center, Overland Park, KS, USA

BACKGROUND CONTEXT: Demyelinating conditions and peripheralneuropathy are well known to occur in patients with Hepatitis B or C.PURPOSE: This study was conducted to determine whether results oflumbar decompression are inferior in hepatitis B� or C� patients andif these patients are more likely to experience painful postoperative radiculi-tis because of underlying neurogenic disease.STUDY DESIGN/SETTING: We performed a retrospective review ofpatients who had undergone primary lumbar decompression surgery forspinal stenosis between 2001 and 2003. All cases were performed in oneof seven different hospitals which included one university/acamedic centerand six community hospitals.

Proceedings of the NASS 19th Annual Meeting / The Spine Journal 4 (2004) 3S–119S70S

OUTCOME MEASURES: n/aMETHODS: Recombinant adenoviruses expressing human BMP-2, BMP-6,BMP-9, and green fluorescent protein (GFP) (as a control) were constructedusing the AdEasy system. Skeletally mature New Zealand White Rabbitsunderwent postero-lateral spinal arthrodesis. The bone marrow was firstaspirated from the tibia under general anaesthesia. 5cc of bone marrowaspirate (BMA) was then infected with AdBMP-2, AdBMP-6, AdBMP-9,or AdGFP (as control) (∼109 pfu per infection) for 20 minutes. Postero-lateral arthrodesis at two levels (i.e., L2-3; L5–6) was performed bilaterallyusing adenovirus-transduced autogenous BMA soaked collagen sponges(randomized for implantation levels, and 2 rabbits per treatment). Exoge-nous BMP production by marrow cells was confirmed by RT-PCR. Rabbitswere sacrificed at 4 weeks and 8 weeks post-arthrodesis and the lumbarspines were evaluated with gross, radiographic, CT and histologicalexaminations.RESULTS: In AdGFP-transduced BMA implant sites, no evidence ofspinal fusion was seen at either time point. At both 4 and 8 weeks, nomotion was detected with manual palpation across the levels treated withBMA-transduced with all three BMPs. X-Ray analysis revealed solid fusionacross all AdBMP-treated levels at 8 weeks. Quantitative CT scans at 4weeks confirmed robust bridging bone across the treated level with all threeAdBMP constructs. Histologic analyses suggest mature “fusion masses”achieved with BMA-AdBMPs.CONCLUSIONS: Our recent in vitro studies have demonstrated BMP-2,BMP-6, and BMP-9 to be most potent osteogenic factors, particularly earlyduring osteogenesis. In this study we demonstrate the ability of bone marrowaspirates infected for clinically-relevant times with adenoviruses expressingBMP-2, BMP-6 and BMP-9 to induce spinal fusion. Fusion was seen byone month after treatment. Bone marrow aspirate has been clinically usedto promote osteogenesis and contains stem cells that have the ability todifferentiate along an osteoblastic cell lineage. Autogenous bone marrowaspirate is easily obtained and appears to be a promising carrier for adenovi-ral vectors expressing osteogenic genes to promote spinal fusion.DISCLOSURES: No disclosures.CONFLICT OF INTEREST: Author (TH) Grant Research Support:NASS Research Grant 2003.

doi: 10.1016/j.spinee.2004.05.138

P41. Partial corpectomy with titanium mesh reconstruction for

PATIENT SAMPLE: 752 patients who had undergone primary lumbardecompression surgery for spinal stenosis between 2001 and 2003 wereretrospectively studied. All patients had at least 6 month follow-up. Eighteenof these patients were hepatitis B�, and 13 were hepatitis C�.OUTCOME MEASURES: Preoperative and postoperative Oswestry andVAS data was collected. Records were also examined to identify patientswith painful neuritis after surgery.METHODS: Information on age, sex, smoking history, diabetes, alcohol-ism, number of levels decompressed, postoperative infection, and intraoper-ative dural leak, as factors which may have contributed to postoperativeradiculitis were also collected in all patients. Logistic regression analyseswere used to determine if hepatitis positive patients were at significantlyhigher risk of developing postoperative radiculitis. In addition, linear regres-sion analysis was used to compare improvement in pain and functionalscores in patients with and without hepatitis. Confounding variables werecorrected for in the above analyses.RESULTS: Postoperative neuritis occurred in 6/18 of the hepatitis B�

patients and was permanent in 4. Postoperative neuritis occurred in 8/13of the hepatitis C� patients and was permanent in 7. Postoperative neuritisoccurred in 19/723 patients without hepatitis and was permanent in 8.Logistic regression analyses demonstrated that the odds of postoperativeneuritis, both transient and permanent was significantly higher in patientswith hepatitis B and/or C when compared to patients without hepatitis(p�0.01; OR>10 for hepatitis B�, OR>25 for hepatitis C�). In addition,linear regression demonstrated that the improvement in Oswestry and VASscores after surgery were significantly lower in patients with hepatitis Band/or C when compared to patients without hepatitis (p�0.01). Paired t-testsin hepatitisC� patients demonstratedsignificant but reduced improvement inVAS after surgery, but no significant improvement in Oswestry scores.CONCLUSIONS: Postoperative neuritis is much more common afterlumbar decompression in patients who are hepatitis B� and especiallyin those who are hepatitis C�. In addition, postoperative improvement inpain and functional scores are significantly diminished in the hepatitis B�

and C� population. Patients with hepatitis B or C must be warned ofdiminished success rates after lumbar decompressive surgery before opera-tive intervention is considered.DISCLOSURES: No disclosures.CONFLICT OF INTEREST: No Conflicts.

doi: 10.1016/j.spinee.2004.05.137

P96. Spinal fusion using bone marrow aspirate transduced withrecombinant adenoviral vectors expressing BMP-2, BMP-6 andBMP-9Michael Sun, MD1, Frank Phillips2, Quan Kang1, Vishal Mehta, MD1,Scott Stacy, MD1, Peter Sturm1, Purnendu Gupta3, Rex Haydon3, Tong-Chuan He, MD, PhD4; 1University of Chicago Medical Center, IL,USA; 2Rush University / Rush- Presbyterian-St. Luke’s Medical Center,Chicago, IL, USA; 3University of Chicago, Chicago, IL, USA;4University of Chicago Medical Center, Chicago, IL, USA

BACKGROUND CONTEXT: Targeted gene therapy approaches to spinalfusion may circumvent some of the difficulties and limitations associatedwith use of recombinant proteins. Successful spinal fusion using adenoviraldelivery of osteoinductive genes has typically required expansion of deliv-ery cells for several weeks prior to implantation, long viral transductiontimes, or the use of supplementary osteoinductive agents. These factorsmay compromise the ability to easily apply these techniques in the clini-cal setting.PURPOSE: This study was designed to determine whether postero-lateralspinal fusion can be reproducibly achieved using autogenous fresh bonemarrow aspirate transduced with adenoviral vectors expressing BMP-2,BMP-6 or BMP-9 for 20 minutes.STUDY DESIGN/SETTING: An in vivo study in New Zealand rabbitscomparing the ability of fresh bone marrow transduced with adenoviralvectors expressing BMPs-2, BMP-6, or BMP-9 to promote spinal fusionPATIENT SAMPLE: n/a

cervical spondylosis trauma or rrevision surgeryAntonio Castellvi, MD1, Scott Farley, DO1*, Deborah Clabeaux, RN1,Alex Castellvi2; 1Florida Orthopaedic Institute, Tampa, FL, USA;2Florida Orthopaedic Institute, FL, USA

BACKGROUND CONTEXT: Complications from posterior cervical ap-proaches to the cervical spine have generated interest in anterior cervicalapproaches that successfully decompress the cervical spine by addressinganterior spinal cord pathology.PURPOSE: In this retrospective study, the efficacy and complications ofpartial anterior cervical corpectomy (PACC) reconstructed with autographpacked titanium cages was evaluated for the treatment of multilevel cervicalspondylosis, trauma and failed primary surgery.STUDY DESIGN/SETTING: A retrospective study was performed todetermine cervical fusion rates, amount of cage subsidence, and the degreeto which cervical lordosis was restored after a one, two or three level partialcorpectomy was performed.PATIENT SAMPLE: Between 2000–2003, 38 consecutive patients under-went anterior partial cervical corpectomy with autograph packed titaniumcage reconstruction for cervical spondylosis, trauma or a failed primarycervical procedure. There were 16 men and 22 women. The average agewas 50 years (range 18–73).The average follow-up was 18 months. Inclusioncriteria included all patients with multilevel cervical disc disease, traumaor failed primary cervical spine procedures. Exclusion criteria were infectionor tumor.OUTCOME MEASURES: Preoperative and postoperative static, andflexion/extension radiographs were reviewed to assess fusion, cervical lor-dosis, and graft subsidence. CT scans with reconstructions were performed