Eur Spine J (2011) 20 (Suppl 4):S421–S464 DOI 10.1007/s00586-011-1950-z

ABSTRACTS

ORAL PRESENTATIONS EUROSPINE 2011

1 DURAL LESIONS IN DECOMPRESSION FOR LUMBAR SPINAL STENOSIS—INCIDENCE AND EFFECT ON OUTCOME B Stro¨mqvist, B Jo¨nsson, F Stro¨mqvist Dept Orthopedics, Clinical Sciences, LUND, Sweden Background context: There are few prospective studies on dural lesions in lumbar spinal stenosis surgery regarding incidence and effect on long-term outcome. Purpose: To determine the incidence of peroperative dural lesions in spinal stenosis surgery and their effect on one-year postoperative outcome. Study design/setting: Prospective cohort study (Swedish Spine Register) with a one-year follow-up. Patient sample: A five-year material of patients undergoing surgery for lumbar spinal stenosis. In total 3 699 patients were included, 935 of which underwent a fusion in conjunction with a decompressive operation and 2 764 who had decompressive surgery only. 44% were males. Mean age was 65 for fused and 68 for non-fused patients. Outcome measures: The Swedish Spine Register protocol including patient based pre- and postoperative data such as consumption of analgesics, walking distance, VAS pain, SF-36, Oswestry score and global outcome assessment. Methods: The incidence was calculated in patients with and without fusion and outcome comparison between patients with and without a peroperative dural lesion using Student’s t-test was performed. Results: The overall incidence of peroperative dural lesion was 7.4%, 8.5% of patients undergoing decompressive surgery only and 5.5% patients undergoing decompressive surgery + fusion (p \ 0.001). Smoking and high age increased the risk for sustaining a preoperative dural lesion whereas previous surgery did not. At one-year follow-up there was no difference in any patient reported outcome measures between patients sustaining and not sustaining a dural lesion preoperatively. Conclusions: Decompressive lumbar spinal stenosis surgery was afflicted with a peroperative dural lesion in 7.4% (5.5–8.4%) but the occurrence of a lesion did not influence the one-year postoperative outcome negatively. 2 VALIDATION OF A CT BASED CLASSIFICATION OF LUMBOSACRAL SEGMENT ABNORMALITIES MJ Stenning, A Issac, A Torrie, J Hutchinson, J Hutchinson Orthopaedics, North Bristol NHS trust, Verwood, UK Introduction: Lumbosacral segment abnormalities are common with a reported incidence of 4–30%. They are classically described as

being best imaged on Ferguson radiographs and current classifications, such as Castellvi’s, are based on these radiographs. Given its superior spatial resolution, CT is considered the best imaging technique for characterisation of these abnormalities. The purpose of this study is to describe and validate a CT based classification of lumbosacral segment abnormalities. Method: 400 CT scans were retrospectively reviewed, a classification devised and incidence of abnormalities recorded. 5 types of abnormality were identified. Type 0 is normal; Type 1 describes an asymmetrical shortening of the iliolumbar ligament; Type 2’s have the transverse process of L5 within 2 mm of the sacrum but not forming a joint; Type 3’s have formed a diarthrodial joint, with 3A’s showing no evidence of degeneration and 3B’s displaying degenerative changes; In type 4’s the transverse process and sacrum have fused; Type5’s have involvement of L4. In order to validate the classification, 40 scans were selected with a full cross section of types. 4 independent observers classified each scan in 2 separate sessions, 2 weeks apart. Results: In the study population there was an abnormality in 54.5% of individuals. In order to validate the classification the intra-observer and inter-observer ratings were analysed. The kappa values for the intra-observer results were between 0.69 and 0.88, indicating substantial agreement (using the Landis and Koch kappa interpretation). The results for inter-observer ratings also gave a combined score of over 0.7 for both sessions, again indicating substantial agreement. Conclusion: A CT classification of lumbosacral segment abnormalities, which is both straight forward to use and repeatable, has been produced. The incidence of these abnormalities is higher in our population of CT scans compared to previous published series using plain radiographs.

3 THE EFFECT OF BMP-2 IN POSTERIOR LUMBAR INTERBODY FUSION. A PROSPECTIVE RANDOMIZED CONTROLLED CLINICAL AND RADIOLOGICAL TRIAL J Sys, J Michielsen, A Rigaux, J Weyler Orthopaedic Surgery, Sint Blasius Hospital, Dendermonde, Belgium Introduction: In this prospective randomised controlled trial, our objective was to assess both the clinical and radiological effect of recombinant human bone morphogenic protein (rhBMP)-2 on an absorbable collagen carrier (inductosTM) in instrumented posterior lumbar interbody fusion (PLIF) with polyetheretherketone (PEEK) cages (TelamonTM). Materials and methods: Forty patients were recruited for the study fulfilling strict entry requirements and were randomised with a 1:1 ratio. Patients completed the Oswestry Disability Index (ODI), the Short-Form 36 (SF-36), and the Visual Analogue Score (VAS) preoperatively and postoperatively at 3, 6, 12, and 24 months, respectively. At each of these moments, radiographs of the lumbar spine were taken. CT-scans of the fused segment with reconstructions in both the coronal and the sagittal plane were taken postoperatively at 3, 6, and 12 months, respectively. Posterior stabilisation was achieved with pedicle screws and interbody fusion was aimed at with PEEK cages, filled with

123

S422 rhBMP-2 in the study group and autologous bone in the control group. Results: Baseline demographic data (age, sex, smoking history, preoperative outcome measures) showed no statistical difference between groups. There were no significant clinical differences (VAS, ODI, and SF-36) between the 2 groups at 2 years post-operatively. At 3 months, end plate resortion was noted around the cages filled with BMP-2 in a significant number of patients. No cage migration nor subsidence was observed. At 2 years, CT-scans showed osseous healing with no difference between groups. Complications were similar in both groups. There were no revision procedures. Detailed results will be available soon. Conclusion: End plate resorption around cages filled with rhBMP-2 on an absorbable collagen carrier did not result in different fusion rates or clinical differences when compared to cages filled with autograft bone in posterior lumbar interbody fusion.

4 LONGITUDINAL RADIOGRAPHIC EVALUATION OF MATURITY OF THE GRAFTED FUSION BONE IN SUCCESSFUL ARTHRODESIS PATIENTS AFTER PLIF USING INTERBODY CARBON CAGE—A PROSPECTIVE 5-YEAR STUDY T Kanemura, Y Ishikawa, A Matsumoto, Y Sakai, Z Ito, A Muramoto, S Imagama Spine Center, Konan Kosei Hospital, Konan-city, Japan Background: To date, there have been no prospective studies regarding long-term longitudinal radiographic changes in the interbody bone fusion site after posterior lumbar interbody fusion (PLIF) using interbody cages. Using plain radiographs and helical CT scans, the study aimed to show the maturity of the grafted fusion bone of successful arthrodesis patients after PLIF using interbody carbon cage in long-term longitudinal radiographic evaluation for [5 years. Methods: For 154 consecutive patients who underwent PLIF using interbody carbon cages at 1 and 2 levels, a prospective longitudinal radiographic evaluation including plain radiographs and CT scans was performed for [5 years after surgery. On the basis of the radiographic assessment at 12 months after surgery, 128 patients were selected as subjects with temporary successful arthrodesis. Regarding the method of grafted bone, patients were divided into local bone with iliac crest bone group (ICB) and without iliac crest bone group (LB).Radiographic findings in the interbody bone fusion site were determined by observing a contrast between radiographic densities of the bone and 4 square carbon cage struts (cross sign), continuous bony bridging, extension of bridging bone fusion density, and remodeling status of the grafted bone to the trabecular bone. Interpretation of radiographs and CT scans were graded on a 4-point scale. Results: The average grades for all radiographic and CT scan assessments increased for 5 years after surgery, and differences between these grades at each time interval compared to the previous interval were statistically significant for 3 years after surgery. The overall analysis for comparisons in the grafted bone status between ICB and LB group by ANOVA showed no significant differences of the longitudinal patterns of the maturity of the grafted fusion

123

Eur Spine J (2011) 20 (Suppl 4):S421–S464 bone. Only 51.5% of LB group and 70% of ICB group showed [50% trabecular bone formation in the original bone graft area at 2 years after surgery. Nevertheless, the proportion of the levels that showed [50% trabecular bone formation significantly increased 77.1% of LB group and 90% of ICB group at 3 years after surgery (p \ 0.01). Conclusions: The maturity of the grafted fusion bone of successful arthrodesis patients after PLIF using interbody carbon cage continued beyond 3 years after surgery despite the method of bone graft. Therefore, final assessment regarding success might also be made at least 3 years after surgery.

5 NUCLEOTOMY WITH ADDITIONAL TRANSPEDICULAR DYNAMIC STABILIZATION VERSUS NUCLEOTOMY ALONE FOR THE TREATMENT OF LUMBAR DISC PROLAPSE—A 10 YEAR FOLLOW-UP M Putzier, P Strube, C Gross, E Hoff Klinik fu¨r Orthopa¨die, Charite´—Universita¨tsmedizin Berlin, Berlin, Germany Aim of the study: Clinical and radiological long-term evaluation of the effects of transpedicular dynamic stabilization after nucleotomy. Background: Short- and mid-term investigations have shown that additional dynamic stabilization is appropriate to prevent progression of initial segment degeneration after nucleotomy and associated with superior clinical outcome compared to nucleotomy alone. Long term data are missing. Methods: 84 patients (group N) with symptomatic disc prolapse and initial degeneration (Modic = I) of the lumbar spine underwent nucleotomy between 01/1999 and 08/2002. Additional dynamic stabilization was performed in 35 subjects (group D). Clinical (ODI, VAS) and radiological (plain, extension-flexion X-ray, MRI) parameters were collected before, at 3 month postoperatively as well as at medium and long term (mean follow-up 2.8 respectively 10.2 years) follow-up. Results: In group D 29/35 patients (83%) and in group N 38/49 patients (78%) were available for examination at the final follow-up. Reoperation rate in group D was 34% (10/29) due to implant failure or disc prolapse/degeneration in adjacent segments. In group N 5/40 (13%) underwent further operation due to re-prolapse or progression of degeneration of the operated segment. In the remaining patients we found a significant improvement of ODI and VAS pre- vs. all postoperative follow-ups in both groups with similar results at the final follow-up. In group D we observed no progression of disc degeneration in the dynamically stabilized segment, but rate of disc degeneration adjacent to dynamic stabilization was not significantly different to that of the nucleotomized level in group N. Residual ROM of the dynamically stabilized segments was minimal at the long term follow-up, while the nucleotomized patients showed a physiologic ROM. Conclusion: At a long term follow-up, additional dynamic stabilization does not lead to a clinical benefit in patients with symptomatic disc prolapse and initial disc degeneration. In combination with a high rate of necessary reoperations we do not recommend this surgical strategy for this indication.

Eur Spine J (2011) 20 (Suppl 4):S421–S464 6 FEELING BETTER IS GOOD, BUT FEELING GOOD WOULD BE BETTER AF Mannion, TF Fekete, F Lattig, FS Kleinstu¨ck, UM Mutter, F Porchet, D Jeszenszky, D Grob Spine Center, Schulthess Klinik, Zu¨rich, Switzerland Introduction: Patient-rated outcome of surgery is often assessed in terms of the perceived change compared with the pre-operative state, i.e. it concerns the concept of ‘‘improvement’’. Adjectival scales (e.g. better, unchanged, etc.) as well as change scores (e.g. a 15-point reduction on a 0–100 scale) are commonly used for these purposes. However, such constructs do not indicate whether an acceptable symptom state has actually been reached, a factor that likely governs future care-seeking. This study sought to compare the concepts ‘‘feeling good’’ (acceptable state) and ‘‘feeling better’’ (improvement) as indicators of surgical success. Methods: All patients in our spine unit undergoing surgery for painful lumbar degenerative disorders (from 2005 to 2010) were asked to complete a 12-month follow-up questionnaire containing a question about the acceptability of symptoms: ‘‘if you had to spend the rest of your life with the symptoms you have right now, how would feel about it?’’. Responses were given on a 5-point Likert scale from ‘‘very satisfied’’ to ‘‘very dissatisfied’’. They also rated the global outcome of surgery on a 5-point Likert scale (operation ‘‘helped a lot’’ through to ‘‘made things worse’’). Results: Questionnaires were returned by 2945/3193 (92%) patients. Global outcomes were as follows: operation helped a lot, 44.4%; helped, 29.1%; helped only little, 14.9%; did not help, 9.9%; made things worse, 1.7%. Ratings for acceptability of the symptom-state were: very satisfied, 24.8%; somewhat satisfied, 19.1%; neither satisfied nor dissatisfied, 16.3%; somewhat dissatisfied, 18.9%; very dissatisfied, 20.8%. Cross-tabs analysis showed that even in the group reporting that the operation helped or helped a lot, only just over half of them (58%) were somewhat/very satisfied with their current symptom state; 19% were ambivalent, and 23% were dissatisfied. Conclusion: The concepts of ‘‘improvement’’ and ‘‘acceptable state’’ are both important to consider. Whilst global treatment outcome ratings and change scores show the achievement of a good response to treatment, they tend to paint a more optimistic picture than when the proportion of patients achieving an acceptable state is examined. In striving to continuously improve the quality of care in spine surgery, a more critical objective may be the achievement of a state considered acceptable by the patient.

S423 major proinflammatory cytokines that are present in the IVD during degeneration (IL-1b, TNF-a) as well as a typical TLR ligand (LPS) can stimulate mRNA and protein expression of three major Toll-like receptors: TLR2, TLR3 and TLR4. Human IVD cells were isolated from biopsies of patients undergoing spinal surgery and stimulated with different concentrations of IL-1b, TNF-a or LPS for different time periods before measuring changes in mRNA and protein levels (relative to untreated controls) by real-time RT-PCR or immunoblotting. Statistical analysis was performed by Mann–Whitney U Test (two-tailed) with a significance level of p \ 0.05. Time course experiments (2, 6 and 18 h) demonstrated that, for all conditions and TLRs, effects at the late time point were most pronounced. In detail, treatment with IL-1b increased gene expression of TLR2 in a dose-dependent manner, with the strongest effect with 10 ng/ml (18 fold); this effect could be confirmed on the protein level. Stimulation with TNF-a increased gene expression of TLR2 (120 fold: Fig. 1), TLR3 (8 fold: Fig. 2) and TLR4 (twofold) at 100 ng/ml (dose-dependent), with a corresponding protein expression pattern. LPS significantly increased TLR2 gene expression (17 fold) and protein expression at 0.01 lg/ml (inverse dose-dependency). This study revealed that two of the major proinflammatory cytokines that are present during symptomatic IVD degeneration increase gene and protein expression of TLR2 and TLR3. Interestingly, LPS (a known TLR4 ligand) did not alter TLR4 expression, but TLR2 and TLR3 expression. From the presented data, it is unclear whether IL1b, TNF-a and LPS actually activate the investigated TLRs. Therefore, future studies are planned in which we aim to analyze whether these inflammatory mediators function as TLR ligands in vitro, thus causing TLR activation and downstream activation of NF-kB.

Fig. 1 7 REGULATION OF TOLL-LIKE RECEPTOR GENE AND PROTEIN EXPRESSION IN HUMAN INTERVERTEBRAL DISC CELLS K Wuertz, M Klawitter, L Quero, T Liebscher, A Nerlich, N Boos Spine Research Group, CABMM, University of Zurich, Zurich, Switzerland During the past years, Toll-like receptors (TLRs), which are known to be implicated in innate immunity and inflammation, have been shown to be highly expressed during osteoarthritis or rheumatoid arthritis. While it is known that symptomatic intervertebral discs (IVDs) are characterized by increased levels of proinflammatory cytokines, it is unclear which receptors and pathways mediate inflammation in the IVD. Therefore, the aim of this study was to investigate whether

Fig. 2

123

S424

Eur Spine J (2011) 20 (Suppl 4):S421–S464

8 ASSOCIATION BETWEEN INTERLEUKIN 18 RECEPTOR GENES (IL18R1 AND IL18RAP) AND LUMBAR DISC DEGENERATION A Omair, BA Lie, O Reikeras, JI Brox Orthopaedics, Oslo University Hospital-Rikshoapitalet, Oslo, Norway Background: Lumbar disc degeneration (LDD) represents a major cause of low back pain. Though the etiology is complicated and yet unexplained, familial predisposition and genetic factors have been suggested to contribute substantially. Associations between LDD and genes related to disc structure, inflammation and matrix degradation have previously been reported in different study populations. Purpose: To examine structural, inflammatory, matrix modifying, vitamin D receptor and osteoarthritic genes and their association with surgical candidates and surgical patients with LDD, in light of their previously proposed susceptibility. Patient sample: 146 unrelated Norwegian patients with disc degeneration and low back pain. The group comprised of 80 females and 66 males with a mean age of 53.2 years (SD = 8.8). All the patients were clinically and radiologically evaluated to have disc degeneration and low back pain of a minimum duration of 1 year. 27 patients received conservative treatment, 27 underwent discectomy and 73 had lumbar fusion. The control group comprised of 188 healthy Norwegian individuals. Methods: All patients completed questionnaires with information regarding duration of back pain, age of onset, ethnicity, family history and smoking. Genotyping of the DNA from both cases and controls was performed for 20 single-nucleotide polymorphisms (SNPs) from collagens, aggrecan, interleukins, VDR, MMP-3 and COX-2 genes and 7 SNPs conferring risk to osteoarthritis. Individual SNPs, as well as haplotype association was assessed using Haploview. Results: The neighbouring genes IL18R1 and IL18RAP polymorphisms (rs1420100 and rs2287037), showed a marginally increased allelic risk for developing LDD (OR 1.36 [95% CI 0.99–1.87] and OR 1.33 [95% CI 0.98–1.81]). Homozygosity of these risk alleles was significantly associated with LDD (p = 0.023 and p = 0.027). The non-risk alleles at these SNPs were situated in a negatively associated haplotype with LDD (p = 0.008). Carriage of at least one non-risk allele at both loci highly reduces the risk of developing LDD (OR 0.51 [95% CI 0.33–0.80]; p = 0.003). Stratification of the LDD patients according to smoking status showed that the haplotype conferred significantly reduced risk among non-smoking patients (p = 0.007). Conclusions: Our findings support a polygenic and multifactorial nature of LDD and suggest the possible role of an inflammatory mediator in the etiology of advanced stage of LDD and associated low back pain. Table 1 Association analysis of the structural, inflammatory, matrix degradative, vitamin D and osteoarthritic genes with LDD SNP

Gene

a

Risk Allele

RAF cases

RAF control

OR CI)

(95% p value

rs1042631

AGC-1

T

0.212

0.176

1.27 (0.86–1.86) 0.231

rs1516797

AGC-1

T

0.641

0.622

1.09 (0.79–1.50) 0.602

A

0.401

0.370

1.14 (0.83–1.56) 0.414

rs10735810 VDR rs731236

VDR

A

0.603

0.568

1.16 (0.85–1.58) 0.353

rs2075555

COL1A1

G

0.873

0.864

1.08 (0.69–1.70) 0.735

rs917055

COL2A1

A

0.168

0.168

1.00 (0.67–1.51) 0.993

rs2056156

COL3A1

C

0.441

0.420

1.09 (0.80–1.48) 0.584

rs696990

COL9A1

C

0.210

0.173

1.27 (0.86–1.87) 0.224

123

Table 1 continued SNP

Gene

Risk Allele

RAFa cases

RAF control

OR CI)

rs7533552

COL9A2

T

0.757

0.731

1.14 (0.80–1.63) 0.450

rs61734651 COL9A3

C

0.942

0.939

1.05 (0.56–1.97) 0.873

rs1463035

COL11A1 T

0.810

0.797

1.08 (0.74–1.59) 0.676

rs1676486

COL11A1 A

0.229

0.217

1.07 (0.74–1.55) 0.719

rs2072915

COL11A2 T

0.731

0.714

1.09 (0.77–1.54) 0.618

rs2071375

IL1A

C

0.700

0.670

1.15 (0.82–1.60) 0.415

rs2287037

IL18R1

C

0.668

0.596

1.36 (0.99–1.87) 0.056

rs1420100

IL18RAP

A

0.493

0.422

1.33 (0.98–1.81) 0.067

rs1420106

1L18RAP

A

0.253

0.239

1.08 (0.76–1.54) 0.675

rs917997

IL18RAP

T

0.262

0.235

1.16 (0.81–1.65) 0.426

rs72520913 MMP-3

A

0.497

0.470

1.11 (0.82–1.51) 0.502

rs5277

COX-2

C

0.836

0.832

1.02 (0.68–1.54) 0.913

rs4140564

PTGS-2

A

0.949

0.931

1.35 (0.72–2.56) 0.342

rs12885713 CALM-1

C

0.451

0.406

1.20 (0.88–1.64) 0.246

rs11718863 DVWA

A

0.842

0.824

1.14 (0.76–1.71) 0.537

rs7639618

DVWA

C

0.842

0.824

1.14 (0.76–1.71) 0.537

rs288326

FZRP

A

0.123

0.109

1.15 (0.72–1.84) 0.567

rs7775

FZRP

C

0.083

0.059

1.42 (0.79–2.56) 0.243

rs143383

GDF-5

A

0.651

0.629

1.10 (0.79–1.52) 0.567

a

(95% p value

Risk allele frequency

Table 2 Association analyses of estimated haplotypes of IL18R1 (rs2287037) and IL18RAP (rs1420100) Haplotype Cases, n (%) T–C C–Aa

Controls, n (%)

OR (95% CI)

p value

68.8 (23.7) 126.9 (34.3) 0.63 (0.43–0.93) 0.008 114.8 (39.6) 133.9 (36.2) 1.0

0.176

C–C

78.2 (27.0)

87.1 (23.5) 1.05 (0.68–1.61) 0.563

T–A

28.2 (9.7)

22.1 (6.0)

a

1.49 (0.71–3.10) 0.225

Reference haplotype

9 CLINICAL OUTCOME AND RADIOGRAPHIC ASSESSMENT OF LATERAL LUMBAR INTER-BODY FUSION: A MINIMUM 2 YEAR FOLLOW-UP M Pumberger, S Kotwal, I Merino, DR Lebl, C Abjornson, A Hughes, RR Huang, FP Cammisa, AA Sama, FP GIrardi Orthopedic Surgery, Spine and Scoliosis Service, Hospital for Special Surgery, New York, USA Numerous articles have reported Lateral Lumbar Inter-body Fusion for treatment of degenerative intervertebral disc disease and related conditions, but to our knowledge no publication has reported a minimum 2-year clinical outcome with radiological assessment for a large cohort. We performed a retrospective chart and radiographic review of all patients who underwent LLIF between 2006 and 2008 at our institute. All patients were evaluated for demographic data, indications,

Eur Spine J (2011) 20 (Suppl 4):S421–S464 medical co-morbidities, bone density, BMI, pre-operative narcotic use and previous related surgeries. Peri-operative parameters including operative time, blood loss, cage dimension, bone graft substitute were noted. Immediate, early and late complications including neurological deficit, visceral and vascular injury were recorded. Pre-operative, immediate post-operative and minimum 2-year follow-up radiographs were studied for restoration of disc space height, coronal angulation and lumbar lordosis at each level, overall Cobb angle and sagittal balance, end-plate fractures, subsidence, cage overhang, cage location and pseudoarthrosis. The patients were evaluated with patient derived outcome measures (VAS, SF-12 and ODI). There are total of 141 (325 levels) patients with 68 males and 73 females who underwent minimally invasive LLIF at a single institution by 4 surgeons with a minimum 2-year follow-up for each patient. 95.1% levels were confirmed to have secure arthrodesis. Fusion could not be confirmed in 16 levels at final follow-up. There was a mean correction of 2.8 degrees at each instrumented level in the coronal plane and 3.9 degrees of lumbar lordosis. There was significant correction in overall sagittal and coronal plane alignment. 3.2% had motor nerve and 19.3% has sensory deficits which resolved at 6 month follow-up. The most common post-operative complication was anterior thigh pain in 31.5% was self limiting in majority of cases. There was no incidence of vascular or visceral injury. The patient derived clinical outcome scores were improved significantly at the final follow-up. LLIF is an effective technique to achieve interbody arthrodesis while allowing significant correction of coronal and sagittal alignment in axial backpain, spinal stenosis, instability, lateral spondylolisthesis and de novo scoliosis.

10 ADJACENT DISC DISEASE AND REVISION SURGERY AFTER 360-DEGREE CIRCUMFERENTIAL FUSION. A 10-YEARS PROSPECTIVE STUDY ON 73 PATIENTS JP Maruenda, F Garibo, C Barrios, J Burgos, E Hevia Department of Orthopaedic Surgery, Valencia University Clinic Hospital, Valencia, Spain Introduction: Circumferential fusion has shown to be superior to isolated instrumented posterolateral fusion at short-term outcome. However, few reports investigate long-term outcome, specially regarding the appearance of adjacent disc disease and the reintervention rate. Patients and methods: 73 patients underwent lumbar fusion involving one to three levels (from February 1998 to January 2000). Autologous iliac bone graft was used for fusion supplementation in all cases. Patients were evaluated preoperatively, at 2.5 years follow-up and 10 years as after surgery with static and dynamic radiographic studies, CT scan and MRI. Analyzed parameters included the fusion rate, adjacent disc disease (ADD) and the reintervention rate. Patients were also analyzed with the Oswestry-Disability index (ODI), VAs scores, and the patient self-satisfaction questionnaire. Results: At 2.5-year follow-up there was a decrease in pain according to VAs (from 8.4 preop to 4), average ODI score was 30.5, an excellent and good self-satisfaction rate of 82.8%, and a 100% radiologic fusion rate. None of the patients exhibited ADD at this time. There was only a single case of revision surgery because of pedicle screw malposition. At 10-years follow-up, fusion rate remains

S425 in 100%. Clinical outcome according to ODI (65.6) and VAs scores (8) showed a clear worsening. Excellent and good self-satisfaction rate decreases to 41.1%. ADD was detected in 37 cases (50.7%). The clinical worsening of ADD patients conducted to new surgical treatment in 18 of these cases (24.6% of the total series). Conclusion: Circumferential lumbar fusion provides high fusion rate although this factor had no relationship to long-term clinical status. From 2.5- to 10-year follow-up outcome worsened significantly. Due to the high occurrence of ADD, the indications for 360 lumbar fusion should be revised, being stricter when selecting patients. Extreme rigid fusions such as the circumferential fusion technique deserve therefore a certain criticism.

11 BACK PAIN IN ADOLESCENTS WITH IDIOPATHIC SCOLIOSIS T Hirano, K Watanabe, T Izumi, A Sano Department of Orthopedic Surgery, Niigata University Medical and Dental Hospital, Niigata City, Japan Introduction: Although BP is one of the important outcomes in the treatment for idiopathic scoliosis, there have been few studies regarding BP in adolescents with IS compared with proper control group using proper statistical analysis. Patients and methods: Consecutive 147 girls with IS aged from 10 to 15 years were investigated. All had scoliosis more than 20 of Cobb angle (35 on average). Control group consisted of 16,194 school girls with the comparable age without history of scoliosis. Self questionnaire was used to collect the data regarding both high and low BP such as present and past back pain, location of pain (back was divided into 9 regions), duration, recurrence and severity (history of absence from school due to back pain). Mantel–Haenszel test was chosen for statistical analysis to compare IS group and the control group because prevalence of BP increased with age and the age distribution was different between two groups. Results were expressed as odds ratio (OR) and 95% confidence interval (CI). Results: Both point and lifetime prevalence of high BP were significantly higher in IS group (OR 2.9, CI 1.6–5.3, p = 0.001, OR 2.1, CI 1.4–3.3, p = 0.03). Pain was more frequent around right scapula in IS group (OR 4.6, CI 2.2–8.9, p \ 0.0001). Although duration was not significantly different, recurrence of high BP was more frequent (OR 2.8, CI 1.1–4.2, p = 0.03) and it was severer in IS group (OR 3.5, CI 1.1–11.7, p = 0.03). However, there were no significant differences in all analyses regarding low BP between two groups. Discussions: Although there have been a few studies regarding the prevalence of back pain in adolescents with IS, these studies lacked the proper control group and/or the adequate statistical analysis and the results were inconsistent. To our knowledge, this is the first study clarifying the characteristic features of BP in adolescents with IS by comparing to the proper control group using proper statistical analysis. Adolescents with IS had more frequent high BP especially around right scapula, which had more tendency to recur and was severer compared to the control group. This characteristic location of pain may be due to the right rib prominence which is common in idiopathic scoliosis. To the contrary, low BP in IS was not different from the control group. These basic data are important to evaluate BP in IS.

123

S426 12 THE EFFICACY OF COMPLETE CONSERVATIVE TREATMENT FOR ADOLESCENT IDIOPATHIC SCOLIOSIS BASED ON SOSORT MANAGEMENT CRITERIA: RESULTS ACCORDING TO THE SRS CRITERIA FOR BRACING STUDIES F Zaina, S Donzelli, S Negrini PRM, ISICO, Milan, Italy Background context: The SRS criteria give the methodological reference framework for the presentation of bracing results, while the SOSORT criteria give the clinical reference framework for an appropriate bracing treatment. The two have not been combined in a study until now. Purpose: To verify the efficacy of a complete, conservative treatment of Adolescent Idiopathic Scoliosis (AIS) according to the best methodological and management criteria defined in the literature. Study design/setting: Retrospective study in a prospective database Patient sample: We included all AIS patients respecting the SRS inclusion criteria (age 10 years or older; Risser test 0–2; Cobb degrees 25–40; no prior treatment; less than 1 year post-menarchal) who had reached the end of treatment since our institute database start in 2003. Thus we had 44 females and four males, with an age of 12.8 ± 1.6 at the commencement of the study. Outcome measures: SRS criteria fro Bracing studies (unchanged; worsened 6 or more; over 45 at the end of treatment; surgically treated; 2 years’ follow-up); clinical criteria (ATR, Aesthetic Index, plumbline distances); radiographic criteria (Cobb degrees); and ISICO criteria (optimal; minimal). Methods: According to individual needs, two patients have been treated with Risser casts followed by Lyon brace, 40 with Lyon or SPoRT braces (14 for 23 h per day, 23 for 21 h/days, and seven for 18 h/days at start), and two with exercises only (1 male, 1 female): these were excluded from further analysis. Statistics. Paired ANOVA and t-test, Tukey–Kramer and Chi-square test. Results: Median reported compliance during the 4.2 ± 1.4 treatment years was 90% (range 5–106%). No patient progressed beyond 45, nor was any patient fused, and this remained true at the 2-year followup for the 85% that reached it. Only two patients (4%) worsened, both with single thoracic curve, 25–30 Cobb and Risser 0 at the start. We found statistically significant reductions of the scoliosis curvatures (7.1): thoracic (-7.3), thoracolumbar (-8.4) and lumbar (-7.8), but not double major. Statistically significant improvements have also been found for aesthetics and ATR. Conclusions: Respecting also SOSORT management criteria and thus increasing compliance, the results of conservative treatment were much better than what had previously been reported in the literature using SRS criteria only.

13 EVALUATION OF THE EFFECT OF MUSCLE RELAXANTS ON BLOOD LOSS IN CORRECTIVE SURGERY FOR SCOLIOSIS M Spiteri, A Spina, S Paris, J Lehovsky Orthopaedics, Mater Dei Hospital, Sliema, Malta Aim: To evaluate any relationship between blood loss and the regular administration of muscle relaxant drugs during single stage posterior fusion for adolescent idiopathic scoliosis.

123

Eur Spine J (2011) 20 (Suppl 4):S421–S464 Method: A retrospective audit of ten consecutive years, between 2000 and 2009 was performed. Sixty-three patients between the ages of twelve and eighteen underwent a single stage posterior fusion for adolescent idiopathic scoliosis. Ten patients were excluded from the study due to lack of data and due to clotting disorders. All patients were ASA 1. Thirty-one patients were administered regular doses of muscle relaxants during surgery. Twenty-two patients were given a single dose of the drug at induction. The actual blood loss was estimated and compared in both groups. The levels of fusion; duration of surgery; volume and type of intravenous fluids or blood administered intra-operative; and other anaesthetic agents administered were also taken into consideration. Results: The patient’s mean age at the time of surgery was 14.61 years in the group who received regular intra-operative administration of vecuronium or atracorium; and 14.95 years for the other group. The average duration of surgery was 2.6 h (2.7 and 2.6 h respectively). A mean of 10.6 and 11 levels were fused during each operation for either group. The actual blood loss was 1,054 ml in the control group and 604 ml in the group who received regular muscle relaxant drugs (p 0.0028). Discussion: Both groups underwent a single stage posterior fusion for adolescent idiopathic scoliosis. This involved similar levels of fusion, duration of operation and was performed by the same two surgeons. The actual blood loss was significantly different in the two groups, showing statistical significance when compared by student t-test. This implies that regular administration of a muscle relaxant drug during this type of surgery in healthy young individuals would reduce the post-operative blood requirements of these patients.

14 PEDICLE SCREW IMPACTATION IN THE AORTA AFTER DIRECT VERTEBRAL ROTATION LI Alvarez-G-Quesada, R Carrillo, J Carrascoso, I Sampera, M Wagner, C Barrios, J Burgos, E Hevia, P Domenech, J Herrera-Soto Spine Unit-Orthopaedic Surgery, Quiro´n University Hospital, Pozuelo de Alarcon, Spain Objective and introduction: The thoracic pedicle screw implants are accepted as sufficient to firmly secure the correction of spinal deformities, but they are not free of complications. Plowing is defined as the structural failure of the pedicle screw for the shift in the transverse plane during maneuvers direct vertebral rotation (DVR). We present a series of cases with pedicle screws plowing effect that critical approaches to the aorta and required reoperation. Materials and methods: Multicenter retrospective study of 355 pediatric patients undergoing correction of idiopathic and neuromuscular scoliosis between 2004 and 2009 with thoracic pedicle screws and DVR maneuvers. In all cases controlled the position of the screws in the immediate postoperative period by TAC. We selected patients who required removal of screws plowing effect of minimum follow-up series was 2 years. Results: 8 patients (2.2%) had lateral displacement of a screw, provided at the apex and the concavity of the curve, DVR after the maneuver (Table 1): 4 men and 4 women, mean age 14 + 7 years. Intraoperative patients had neurophysiological and radiological confirmation of correct positioning of the screw before placing the bar and perform the derotation maneuver. All of them required reoperation to remove the screw by its proximity to the aorta. No repositioning the screw in either case. No patient underwent vascular or neurological injury or had complications at 2 years of monitoring. A cardiovascular surgeon was present during the reoperations.

Eur Spine J (2011) 20 (Suppl 4):S421–S464

S427

Conclusions: The segmental vertebral derotation maneuver can be cause plowing in appropriately placed screws, breaking the lateral cortex at the maximum correction attempt. The apex of the concavity of the curves is the area of greatest risk for high rotation, dysplasia of the pedicle and the proximity of the aorta. It is difficult to detect this type of problem during surgery, it seems necessary to verify the correct placement of pedicle screws after direct vertebral rotation maneuvers. Table 1 Patient demographics Patient Age at Sex surgery (years + months)

Scoliosis type

Levels of posterior spinal instrumentation

Plowed screw

1

14 + 5

Male

AIS

T2-L3

2 3 4

16 + 3 9+5 11 + 9

Male AIS Female JIS Male NMS

T2-L3 T5-L3 T2-pelvis

5 6 7 8

13 16 12 15

Female Male Female Female

T3-pelvis T2-L1 T2-L3 T2-L3

Left T12 Left T7 Left T9 Right L3 Left T9 Left T9 Left T7 Left T8

+ + + +

11 4 6 8

NMS AIS AIS AIS

AIS adolescent idiopathic scoliosis, JIS juvenile idiopathic scoliosis, NMS neuromuscular scoliosis

and quality of arthrodesis were evaluated. Complications and Spanish SRS-22 version scores were compared. Results: There were no significant differences regarding gender, age, curve type, instrumented levels, or preoperative main curve Cobb angle (LBG 57 ± 13.4 vs. IBG 52.7 ± 11.4). The average surgical time was slightly longer in the iliac bone graft group (4.6 ± 0.6 IBG vs. 4 ± 0.9 h in LBG p = 0.07.) At final follow up the was no significant differences as regards to the main curve Cobb correction however a greater loss of correction was found in the local group (LBG 8.5 ± 6 vs. IBG 4.5 ± 7.3 p = 0.02.) No significant differences were found regarding lumbar lordosis or thoracic kyphosis. Although a greater loss of correction was found in the local bone graft group (less than 10 average), it did not reach high enough values as to be consider clinically significant. Moreover, there were no differences between both groups regarding clinical SRS 22 scores (Global SRS score 4.21 ± 0.4 IBG vs. 4.27 ± 0.38 LBG.) Conclusions: At 5 year follow-up, the local graft is as effective as the iliac crest autograft in order to maintain surgical correction although there was a statistically significant greater loss of radiographic correction at final follow up in the local bone graft group. However clinical differences were not observed as regards to the SRS 22 scores.

Iliac crest graft Age at surgery 15.1* ± 2.1** Preoperative main curve 52.6 ± 11.3 Cobb angle Postoperative main curve Cobb angle

15 IS LOCAL BONE GRAFT ENOUGH TO MAINTAIN SURGICAL CORRECTION IN ADOLESCENT IDIOPATHIC SCOLIOSIS CURVES? A Mardomingo, F Sa´nchez- Mariscal, A Gomez Rice, J Pizones, ´ lvarez, L Zun˜iga, EIzquierdo PA

Final main curve Cobb angle Levels fused

Local bone graft 14.4 ± 1.6 57 ± 13

25 ± 8.5

21 ± 7.7

29.5 ± 9.4

29.6 ± 9.4

0.186 0.464 0.04 0.9

11.6 ± 2.2

10.5 ± 2.8

Surgical time

4.6 ± 0.6

4 ± 0.9

0.07

0.130

Blood units transfused

3.2 ± 1.6

2.6 ± 2.1

0.34

Preoperative thoracic kyphosis T5T12

21.1 ± 11

25.3 ± 11.9 0.77

Spine unit, Hospital Universitario de Getafe, Getafe, Spain

Postoperative thoracic kyphosis T5T12

22.4 ± 10.7

23.58 ± 14.5 0.7

Objective: The purpose of this study was to compare clinical and radiologic results in Adolescent Idiopathic Scoliosis curves using autogenous bone graft from iliac crest versus only local autograft bone. Summary of background data: Harvesting autogenous bone graft has significant morbidity. If only local autogenous bone graft is used, would it be good enough to obtain a similar arthrodesis as that resulting from the use of autogenous iliac crest bone graft? Patients and methods: We carried out a retrospective matched cohort study of 73 patients diagnosed with adolescent idiopathic scoliosis treated by posterior correction and arthrodesis using segmentary instrumentation. Our mean follow was over 8 years. Our minimal postoperative follow-up was 2 years. In the first group, autogenous iliac crest bone graft (IBG) was used while in the other group only local bone (LBG) from the surgical site was used. Radiographic data collection consisted of preoperative, postoperative and final follow-up anteroposterior and lateral full-length radiographs. Loss of correction

Final thoracic kyphosis T5T12

26.2 ± 10.5

26.5 ± 26.5 0.9

Preoperative lumbar lordosis L1S1

57.4 ± 12.5

57.7 ± 13.1 0.263

Postoperative lumbar lordosis L1S1

55.7 ± 10.5

56.4 ± 13.1 0.8

Final lumbar lordosis L1S1

58.3 ± 11

58.3 ± 12.4 0.9

* Mean ** Standard deviation

123

S428 16 ANTERIOR SHORT SPINAL FUSION IN THE TREATMENT OF ADOLESCENT IDIOPATHIC SCOLIOSIS. RETROSPECTIVE REVIEW OF 250 CONSECUTIVE PATIENTS WITH 7 YEARS FOLLOW UP D Zarzycki, T Potaczek, G Makiela, M Tesiorowski Department of Orthopedic Surgery and Rehabilitation, Jagiellonian University, Faculty of Medicine, Zakopane, Poland Summary: The goal of operative treatment of AIS is to obtain a solid fusion with correction of the deformity and restoration of coronal and sagittal plane balance over as few segments as possible. Currently recommended anterior or posterior approaches include ‘‘long segment’’ instrumentation and fusion which includes all vertebrae contained within the Cobb angle of the major. Our described procedure can be several levels shorter. During this study supine stretch films were used for pre-operative planning. Introduction: From 2002 to 2010 we operated 885 patients with AIS using anterior short spinal fusion (‘‘Bone-on-Bone’’ technique).We retrospectively reviewed 250 consecutive patients operated between 2002 and 2005 for single curve at a mean of 7-year follow-up. Methods: The mean age a surgery was 15.8 years (9–48), 86% of the cohort was female and the mean follow-up was 7 years (5.7–8.3).We operated on curves less than 900 by the short segment anterior approach. Results: Surgical correction of the major curve averaged 54.6% over the entire curve, from upper end vertebra to lower end vertebra, correction of the operated segment was 61.1%. The average number of vertebrae fused was 5.4 (4 discs), mean operative time-205 min, blood loss-559 ml and hospitalization time-11.1 days. The compensatory curves spontaneously improved by an average of 34.8%. 89% of the patients’ curves were reduced to below 450, all spines were well balanced in the coronal and sagittal planes. Complications: haemothorax 2 cases, paraplegia 2 cases (epidural haematomaresolved), screw migration 7 cases, chylothorax 4 cases, flat back 8 cases, dural tear 1 case, too short fusion in the upper part-9 cases, too short fusion in the lower part-10 cases. No patient had any pulmonary limitations post-operatively. All the patients were back to an unrestricted lifestyle within 6 months. Conclusion: We report good results following surgical correction of single primary curves with the instrumentation of fewer levels than would have been operated by posterior segmental instrumentation by using our short segment bone-on-bone technique.

17 KYPHOSIS RESTORATION OR MAINTENANCE IN PATIENTS WITH LENKE TYPE I SCOLIOSIS TREATED BY PEDICLE SCREW CONSTRUCT: IS IT REALLY IMPOSSIBLE BY USING 5.5 MM TITANIUM ROD C Ozturk, A Alanay, M Enercan, E Karadeniz, M Baliolu, A Hamzaoglu Orthopedic Surgery, Istanbul Spine Center, Istanbul, Turkey Introduction: Many studies have shown excellent coronal plane correction by using all pedicle screw constructs. However, same papers have shown difficulty in restoration of kyphosis when pedicle screws were used with 5.5 mm titanium rods. The aim of this study is to evaluate the radiographic results in sagittal plane in Lenke type 1 curves treated by pedicle screw construct and 5.5 mm titanium rods.

123

Eur Spine J (2011) 20 (Suppl 4):S421–S464 Materials and methods: One hundred thirty one patients (14 M:117F) with a diagnosis of thoracic idiopathic scoliosis of Lenke type I corrected by polyaxial pedicle screw fixation with 5.5 mm titanium rod were retrospectively analyzed for deformity correction and sagittal plane restoration. Mean age at the time of procedure was 14.9 (10–19) years. All surgeries were carried out by the same surgeon. Pedicle screws were inserted bilaterally at each level fused. Correction of the curve was performed either by cantilever correction, or rod rotation followed by segmental derotation and in situ bending maneuvers. BAVD has not been used as a correction method in any of the patients. Radiographic measurements included coronal thoracic curve Cobb angle, T2-T12 kyphosis, T12-S1 lordosis and CSVL to S1 distance. Proximal junctional kyphosis (PJK) was determined by measuring the kyphosis between upper instrumented vertebrae and the one above. More than 10 degrees kyphosis was accepted as PJK. Results: Average follow-up was 64 (range 24–148) months. Preoperative thoracic kyphosis of 20 and the lumbar lordosis of 32 were improved to 33 and 47, respectively, at the most recent follow-up (p \ 0.05). The preoperative thoracic curve of 50 was corrected to 10 (79% correction, 2% loss of correction) at the most recent followup (p \ 0.05). The noninstrumented lumbar curve of 32 was corrected to 9 (70% correction, 4% loss of correction) at the most recent follow-up. There was no junctional kyphosis at the most recent follow-up. Forty-five percent of patients had preop sagittal plane decompensation (more than 2 cm) preoperatively while 14% had at the final follow-up. Conclusion: Correction of scoliosis by cantilever technique followed by segmental derotation and in situ bending by using 5.5 mm rods provided a significant correction and restoration in thoracic kyphosis. We conclude that the amount of correction in kyphosis depends more on the technique rather than the rod diameter or type.

18 SAGITTAL PLANE ALIGNMENT AFTER POSTERIOR FUSION IN ADOLESCENT IDIOPATHIC SCOLIOSIS: PEDICLE SCREW-ONLY VS. HYBRID CONSTRUCTS M Di Silvestre, G Bakaloudis, F Lolli Spine Surgery Department, Istituto Ortopedico Rizzoli, Bologna, Italy Objectives: To determine whether pedicle screws have any advantages with concern to sagittal plane alignment when compared to hybrid instrumentations in the treatment of thoracic adolescent idiopathic scoliosis (AIS). Summary of background data: Controversy exists regarding loss of thoracic kyphosis seen after thoracic pedicle screws fusion in AIS. Materials and methods: Eighty-eight consecutive AIS patients with a Lenke type 1 curve treated by posterior instrumentation between 1998 and 2003 were analyzed. In 45 patients (Group Hy) thoracic hooks and in 43 patients (Group TPS) thoracic screws alone were used. Preoperative average age (Hy 15.3 vs. TPS 16 years), gender (Hy 38 f/7 m vs. TPS 37 f/6 m), Risser sign (Hy 2.9 vs. 2.9 TPS), main thoracic curve (Hy 64 vs. TPS 65.5) and thoracic kyphosis (Hy 22.6 vs. TPS 21.4) were similar in both groups. Pearson correlation coefficient and univariate analysis of variance were used. Results: At a mean follow-up of 7.3 years (5–10), the TPS group achieved a greater final main thoracic curve correction (Hy 46.4% vs. TPS 58.4%; p \ 0.001), with inferior loss of initial correction (Hy -11.1 vs. TPS -1.3; p \ 0.0005). Absolute final thoracic kyphosis resulted similar (Hy 31.4 vs. TPS 25.4; r = 0.002; p [ 0.05), with both groups showing an equally statistical significant amelioration of the sagittal contour [(Hy 49.2%; p \ 0.001) vs. (TPS

Eur Spine J (2011) 20 (Suppl 4):S421–S464 43.4%; p \ 0.001)] (p [ 0.05). A significant correlation between absolute kyphosis correction (final-preop) and % main thoracic curve initial correction loss was observed in the Hy Group (r = 0.35, p \ 0.001). The SRS-30 assessment showed an improvement in selfimage and satisfaction, without significant differences between groups. Conclusions: According to the present series results, the previously reported loss of kyphosis after a pedicle screws instrumentation in AIS, when compared to hybrid or hooks only implants, should be questioned. At a minimum 5 years follow-up, sagittal contour in the thoracic spine was not found less kyphotic when pedicle screws were compared to hybrid constructs (r = 0.002; p [ 0.05). The higher the loss of initial correction of main thoracic curve, seen in the Hybrid Group, the greater the postoperative absolute kyphosis at T5–T12 level (r = 0.35, p \ 0.001). The clinical relevance of such radiographic differences is still undetectable with current selfassessment tools.

S429 latter to evaluate shoulder balance in Lenke 1 & 2 curves, but both on the clinical anterior and posterior clinical examination.

Lenke

1

2

19 DOES ANTERIOR SHOULDER BALANCE IN AIS CORRELATE WITH POSTERIOR SHOULDER BALANCE CLINICALLY AND RADIOGRAPHICALLY? S Yang, E Feuchtbaum, BC Werner, W Cho, V Reddi, V Arlet

Angle measure

MT

PT

T1 tilt

1st rib angle

CRIA

type R

P

R

P

R

P

R

P

R

P

Posterior inner shoulder

0.572

7.73E-6a

0.239

0.085

0.347

0.011a

0.276

0.045a

0.31

0.024a

Posterior outer shoulder

0.52

6.47E-5a

0.246

0.076

0.214

0.124

0.221

0.113

0.3

0.029a

a

Posterior axillary fold

0.528

4.93E-5

0.228

0.1

0.232

0.095

0.156

0.263

0.282

0.041a

Anterior inner shoulder

0.402

0.003a

0.104

0.458

0.441

0.001a

0.391

0.004a

0.291

0.034a

Anterior outer shoulder

0.258

0.062

0.037

0.791

0.299

0.03a

0.315

0.021a

0.309

0.024a

Anterior axillary fold

0.063

0.655

0.221

0.111

0.212

0.128

0.352

0.01a

0.334

0.015a

Posterior inner shoulder

0.057

0.816

0.164

0.502

0.3

0.21

0.396

0.104

0.224

0.372

Posterior outer shoulder

0.146

0.55

0.072

0.771

0.034

0.893

0.176

0.484

0.143

0.573

Posterior axillary fold

0.11

0.655

0.243

0.317

0.201

0.425

0.013

0.959

0.162

0.522

Anterior inner shoulder

0.387

0.102

0.238

0.327

0.484

0.042a

0.504

0.033a

0.479

0.044a

Anterior outer shoulder

0.375

0.113

0.269

0.266

0.425

0.079

0.466

0.052

0.485

0.041a

Anterior axillary fold

0.264

0.274

0.025

0.918

0.419

0.083

0.504

0.033a

0.44

0.067

a

One of the co-authors received salary/research support from Synthes

Orthopaedic Surgery, University of Virginia, Charlottesville, USA Introduction: AIS patients often present with shoulder imbalance. Shoulder balance (Sh.B) is of importance to self-perception. Studies have correlated posterior Sh.B to X-ray measures. It is important to address Sh.B with respect to an anterior view of the patients’ shoulders as if patients were viewing in a mirror. We thus evaluated the anterior Sh.B and correlated it to posterior Sh.B clinically and radiographically in Lenke type 1&2 curves. Methods: A retrospective review of 74 AIS patients with Lenke 1 (55 patients, Age 15.28 ± 3.35) & 2 (19 patients, Age 15.66 ± 3.72) curves from the Scolisoft database. All patients had PA X-rays, and both anterior & posterior photos. X-ray measures for Sh.B included proximal thoracic (PT) Cobb, main thoracic (MT) Cobb, T1 tilt, 1st rib angle (FRA), and clavicle-rib intersection angle (CRIA). Clinical measures for Sh.B included inner shoulder angle (ISA), outer shoulder angle (OSA), and axillary fold angle (AFA) as viewed from anterior and posterior. Pearson’s correlation was used to analyze relationships. Results: For Lenke 1 curves, there was moderate significant correlation between anterior and posterior Sh.B (ISA R = 0.411, P = 0.002; OSA R = 0.415, P = 0.015; AFA R = 0.346, P = 0.011). There was weak to moderate correlation between X-ray and clinical measures. Ranges of correlation were: MT (R = 0.063–0.572), PT (R = 0.037–0.246), T1 tilt (R = 0.212–0.484), FRA (R = 0.013–0.504), and CRIA (R = 0.143–0.485). For Lenke 2 curves, there was weak to moderate correlation between anterior and posterior Sh.B (ISA R = 0.447, P = 0.055; OSA R = 0.254, P = 0.294; AFA R = 0.429, P = 0.067). There was no statistically significant correlation between any X-ray measures and posterior Sh.B, though moderate significant correlation between X-ray measures and anterior Sh.B (Table 1). Conclusion: There is no strong correlation between anterior and posterior Sh.B, and surgeons should evaluate both sides in planning correction. None of the X-ray measures had strong correlation (R [ 0.8) with anterior or posterior Sh.B. A stronger correlation existed between X-ray measures and anterior Sh.B compared to posterior Sh.B in Lenke 2 curves. Significance: This is the first study that compared anterior and posterior Sh.B. Due to the lack of strong correlation between clinical findings and radiologic parameters surgeons should not rely on the

20 BALANCE OF THE SAGITTAL PROFILE—THE KEY TO AVOID IMPLANT FAILURE IN NEUROGENIC DEFORMITY CORRECTION B Wiedenhoefer, C Carstens, M Akbar Spine Center of the Dept. of Orthopedics, Trauma Surgery & Spinal Cord Injury Unit, University Hospital of Heidelberg, Heidelberg, Germany Introduction: Since the publications of Dubousset, Duval-Beaupe`re and Bridwell about the spino-pelvic balance (SPB) normative values to assess the global sagittal profile (SP) and the spino-pelvic parameters (SPP) sacral slope (SSL), pelvic tilt (PT), pelvic incidence (PI), lumbar lordosis (LL) and thoracic kyphosis (TK) with their interdependence calculated in equations of spino-pelvic harmony are established for idiopathic scoliosis. There is also a strong positive correlation between the correction of the SPP and the Health Related Quality of Life (HRQoL) parameters. Reconstructive surgery for neurogenic deformity has a higher risk for secondary complications. However there is no data about a correlation for wheel chair bound handicapped people and the sagittal profile. Materials and methods: Correction surgery was performed according to the Warner–Fakler-procedure in 32 patients with an average lumbar kyphosis of 131 [90–170]. The correction was achieved by kyphectomy combined with rod stabilization. The average age was 7.3 [1.6 to 16.0] years. The patients had a complete thoracic level paralysis ranging from D6 to D12. None of them was preoperatively able to sit free without the support of their arms or to lie supine. 17 children had marked problems with oral nutrition caused by associated intestinal disorders, 12 had clinically evident dyspnea during physical exercises. 5 patients presented with chronic pressure sores over the apex of the kyphosis. Long term outcome, complications and their relation to the SP and the SPP were analysed. Results: The mean extent of lumbar kyphosis could be corrected from 131 to 42 [15–90] after surgery with an average correction of 89

123

S430 [60–127]. On average, 2.5 [1–4] vertebrae were rejected. The functional outcome was rated very favourable. Long term complications consisted of implant failure (rod penetration and breakage). Implant failure is directly linked to the residual kyphosis postoperatively. There was a significant correlation between implant failure and PT (P \ 0.05) and consequently with the equation PI = PT + SS (P \ 0.05) and LL (P \ 0.05). Conclusion: Correction of thoracolumbar hyperkyphosis can provide excellent correction and gain of functional capacity. Thus surgery always has to be performed with the intention to re-establish the sagittal profile as far as possible in order to reduce the risk of implant failure. The planning and performance is highly demanding but effective.

Eur Spine J (2011) 20 (Suppl 4):S421–S464 than the usage of AICG. Regarding to the complication rate, we see a clear advantage for PEEK-cages. Also the shorter duration of surgery and the tendentially shorter term of postoperative stay lead to the thesis, that VBS with PEEK-cage is superior to fusion with AICG.

21 PEEK-IMPLANT OR AUTOLOGOUS ILIAC CREST GRAFT AS VERTEBRAL BODY SUBSTITUTE AFTER ANTERIOR CERVICAL CORPORECTOMY? C Schulz, S Koschel, U Kunz, U Mauer Neurosurgery, Military Hospital Ulm, Ulm, Germany PEEK and cervical plate implanted after one-level corporectomy Introduction: For cervical stenosis, corporectomy is an established treatment option. Golden standard for vertebral body substitute (VBS) is the autologous iliac crest graft (AICG). To avoid complications and discomfort by harvesting the bone, there is increasingly tried to use synthetic replacements (e.g. PEEK). There is no sufficient information about the advantage of using synthetic placements and therefore no consensus in literature. In this study the quality of osseous consolidation as well as the complication and discomfort rate between the approach with PEEK-cage and AICG are compared. Patients/methods: Non-randomized trial from 01/2002 to 03/2011. The study includes 46 patients with cervical spondylitic myelopathy, who were treated by corporectomy and VBS (27 cases 1 VB, 17 cases 2 VB, 2 cases 3 VB). From these patients 26 (group A; prospective) got a VBS of PEEK (Athlet, Co. Signus) with ventral screw-platefixation (Tosca, Co. Signus; monocortical). 20 Patients (group B; retrospective) were supplied with AICG and a ventral semi-rigid/ dynamic plate (ABC, Co. Aesculap or ACPS, Co. Codman; monocortical). In both groups the distribution of gender, age and amount of replaced VB is uniform. Postoperative radiographs, including functional images and CT scans for assessment of bony consolidation were performed after 19 weeks (group A) resp. 20 weeks (group B). Additionally the following criteria were analyzed: duration of surgery, perioperative complications, term of postoperative stay. Results: The length of the distance, which had to be bridged, was mean 27 mm (group A) and 33 mm (group B). Postoperative fusion rate was 90% (group A) and 80% (group B). In 9% of the cases from group A there were postoperative complications. All of them were typical for the anterior instrumentation. 35% of group B had postoperative complications whereas 60% of them were specific for harvesting AICG. Average duration of surgery was 3:04 h (group A) and 4:16 h (group B), middle term of postoperative stay was 7 days (group A) and 9 days (group B). Discussion: Our analysis shows, that the appropriation of PEEKcages leads to tendentially better, at least equal results in fusion rate

123

Postoperative CT after PEEK cage implantation

Eur Spine J (2011) 20 (Suppl 4):S421–S464

S431

22 COMBINED ANTERIOR AND POSTERIOR FUSION FOR CERVICAL SPONDYLOTIC MYELOPATHY ACCOMPANIED WITH ATHETOID CEREBRAL PALSY—EXPERIENCE IN 25 YEARS H Mihara, S Kondo, S Kato, K Ishida, M Ono, K Itoh Orthopaedic Surgery, Yokohama Minami Kyosai Hospital, Yokohama, Japan Purposes: Adult athetoid cerebral palsy (CP) patients often promote cervical spondylotic myelopathy and/or radiculopathy. We have treated this difficult spinal disorder by combined anterior and posterior fusion while overcoming many difficulties. This study investigates clinical and radiographical outcomes following our surgical treatments in 25 years. Methods: This study reviewed 118 athetoid CP patients (86 men, 32 women) who underwent mid-cervical front back fusion for cervical myelopathy. The mean age at surgery was 46.8 years and the mean follow-up period was 7.5 years. As for the surgical method, all patients underwent posterior fusion using wave-shaped rods followed by anterior interbody fusion on the same day. Regarding anterior reconstruction, 115 cases underwent segmental fusion using iliac bone graft with screw and wire fixation (25 cases), with a constrained type plate (15 cases), and with a semi-constrained type plate (72 cases). In 11 recent cases, box cages were used with a semi-constrained type plate. Neurological improvement (evaluated by JOA score), reoperation rate, instrument failure, correction loss (anterior height reduction) in the fused segments were retrospectively evaluated for each reconstruction method. Results: There were no significant differences in neurological recovery relating to surgical methods. As for radiographic evaluations, bony fusion was achieved in all patients within 6 months postoperatively. Early reoperation was required due to instrument failures in six patients (three after anterior plate fixation and three after screw and wire fixation). Late reoperation was carried out for adjacent segment disease in ten patients and for plate related troubles in three. Regarding correction loss in the fused segments, average anterior height loss was largest with the screw and wire fixation (12.4%) and smallest with box cages reinforced by a semi-constrained plate (3.3%). Conclusions: Regarding surgical treatments for cervical myelopathy in CP patients, we have consistently adopted a wave-shaped rod, which is our original interspinous fixation device, for posterior fixation to avoid the potential risks with screw anchoring techniques. However we needed to modify anterior fixation techniques to achieve successful fusion without instrument failures or large correction losses. Then we consider the most reliable anterior reconstruction method is a combination of intervertebral cage(s) with a semi-constrained plate.

Fig. 1 Recent surgical method for athetoid CP patients

23 MORTALITY AFTER SURGICAL TREATMENT OF CERVICAL MYELOPATHY IN ELDERLY PATIENTS Y Harada, F Suetsuna, T Itabashi Orthopedic surgery, Hachinohe Municipal Hospital, Hachinohe, Japan Introduction: This retrospective analysis of patients with cervical myelopathy is designed to identify the significant risk factors and indices that influence patients mortality.

123

S432 Patients and methods: Patients underwent surgical treatment between 1995 and 2009 and followed until they died or for a minimum of 1 year. The data were gleanded from examination of our hospital records and in some cases conducted phone survey. The information abstracted age, sex, date of operation, approach of surgery, postoperative complications, all existing conditions and past history, American Society of Anesthesiology physical status (ASAPS), preoperative and postoperative severity of cervical myelopathy, preoperative ambulatory ability, date of death and cause of death. Existing conditions and past histories were classified into hypertension, cardiovascular diseases excepted hypertension, respiratory and pulmonary diseases, diabetes mellitus, malignant tumor, renal failure cerebrovascular disease and the others. Severity of cervical myelopathy was evaluated by the Japanese Orthopedic Association score (JOA score). Results: 94 patients were included in this study (65 males and 29 females). Mean age at surgery was 71 years (range 65–81 years). Mean period of follow up was 5.7 years (range 1–15 years). A total of 7 patients (3 males and 4 females) of the 94 patients (7.4%) died within the follow up period. In this study, actual measured mortality rate was 100% (rate, 94 of 94) for 1 year, 96% (rate, 49 of 51) for 5 years, 72% (rate, 18 of 25) for 10 years after surgery. The expected mortality rate for the normal population of the same age and sex was 98% for 1 years, 88% for 5 years, 72% for 10 years. There were significant differences to influence patients mortality with regard to renal failure (p \ 0.0001), ASA-PS (p = 0.0031) and preoperative ambulatory ability (p = 0.0017). There were no significant differences to influence patients mortality with regard to age, sex, approach of surgery, postoperative complications, severity of cervical myelopathy, existing conditions and past histories excepted renal failure. Conclusions: These results suggest that patients underwent surgical treatment with cervical myelopathy may have the same degree of mortality rates compared to normal population. Significant risk factors and indices that influence postoperative mortality were renal failure, ASA-PS and preoperative ambulatory ability.

24 THREE-DIMENSIONAL GAIT ANALYSIS IN CERVICAL SPONDYLOTIC MYELOPATHY: COMPARISON WITH AGE- AND GENDER-MATCHED HEALTHY CONTROLS A McDermott, D Meldrum, C Bolger Physiotherapy, Beaumont Hospital, Dublin, Dublin, Ireland Background: Gait impairment is a primary symptom in cervical spondylotic myelopathy (CSM). Previous studies have shown that people with CSM walk more slowly, with reduced range of motion at the lower limb joints. Information on specific kinetic and kinematic parameters is lacking. Furthermore, gait in CSM has been compared with healthy controls walking at a self-selected, faster speed. Gait speed is a known confounding factor for kinematic and kinetic parameters, therefore comparison at matched speed is necessary to determine true, non-speed dependent differences in CSM. Objectives: 1) To compare gait patterns of people with untreated CSM to those of age and gender-matched healthy controls. 2) To examine the effect of speed on kinematic and kinetic parameters. Methods: Ethical approval was obtained from a local ethics committee. Patients with CSM were recruited consecutively from a neurosurgery clinic. Healthy controls, matched to age (± 5 years) and gender, were recruited for comparison. Patients and controls underwent three-dimensional gait analysis using a Vicon motion analysis system, at self-selected speed over a 10 metre track. Controls were also assessed at the speed of their CSM match.

123

Eur Spine J (2011) 20 (Suppl 4):S421–S464 Results: Sixteen CSM patients and controls were recruited. At selfselected speed, the CSM group walked significantly more slowly, with shorter stride lengths and longer double support duration. They showed significant decreases in several kinematic and kinetic parameters, including sagittal range of motion at the hip and knee, ankle plantarflexion, anteroposterior ground reaction force (GRF) at toe-off, power absorption at the knee in loading response and terminal stance, and power generation at the ankle. At matched speed, the CSM group showed significant decreases in knee flexion during swing, total sagittal knee range of motion, peak ankle plantarflexion, and anteroposterior GRF. Conclusions: and implications: People with CSM were found to have significant gait abnormalities that have not been previously reported. At matched speed, differences persisted in gait parameters at terminal stance, suggesting that there are key differences in the motor strategies used by patients in this phase of gait that cannot be explained by speed alone. Further investigation of these strategies is warranted, using electromyography to evaluate muscle function during gait, and principal components analysis to determine key factors contributing to gait impairment in CSM.

25 THE EFFECTIVENESS OF CERVICAL SPINE ORTHOSES I N RESTRICTING SPINAL MOVEMENT: A 3-D MOTION ANALYSIS STUDY N Evans, G Hooper, R Edwards, G Whatling, V Sparkes, C Holt, S Ahuja School of Engineering, Cardiff University, Winchester, United Kingdom Assessing the efficacy of cervical orthoses in restricting spinal motion has historically proved challenging due to a poor understanding of spinal kinematics and the difficulty in accurately measuring spinal motion. This study is the first to use an 8 camera optoelectronic, passive marker, motion analysis system with a novel marker protocol to compare the effectiveness of the Aspen, Aspen Vista, Philadelphia, Miami-J and Miami-J Advanced collars. Restriction of cervical spine motion was assessed for physiological and functional range of motion (ROM). Nineteen healthy volunteers (12 female, 7 male) were fitted with collars by an approved physiotherapist. ProReflex (Qualisys, Sweden) infra-red cameras were used to track the movement of retro-reflective marker clusters attached to the head and trunk. 3-D kinematic data was collected from uncollared and collared subjects during forward flexion, extension, lateral bending and axial rotation for physiological ROM and during five activities of daily living (ADLs). ROM in the three clinical planes was analysed using the Qualisys Track Manager (Qualisys, Sweden) 6 Degree of Freedom calculation to determine head orientation relative to the trunk. For physiological ROM, the Aspen and Philadelphia were more effective at restricting flexion/extension than the Vista (p \ 0.001), Miami-J (p \ 0.001 and p \ 0.01) and Miami-J Advanced (p \ 0.01 and p \ 0.05). The Aspen was more effective at restricting rotation compared to the Vista (p \ 0.001) and Miami-J (p \ 0.05). The Vista was least effective at restricting lateral bending (p \ 0.001). Through functional ROM, the Vista was less effective than the Aspen (p \ 0.001) and other collars (p \ 0.01) at restricting flexion/extension. The Aspen and Miami-J Advanced were more effective at restricting rotation than the Vista (p \ 0.01 and p \ 0.05) and MiamiJ (p \ 0.05). All the collars were comparable when restricting lateral bending.

Eur Spine J (2011) 20 (Suppl 4):S421–S464

S433

The Aspen is superior to, and the Aspen Vista inferior to, the other collars at restricting cervical spine motion through physiological ROM. Functional ROM observed during ADLs are less than those observed through physiological ROM. The Aspen Vista is inferior to the other collars at restricting motion through functional ROM. The Aspen collar again performs well, particularly at restricting rotation, but is otherwise comparable to the other collars at restricting motion through functional ranges.

27 FIRST EIGHT HOURS CLINICAL SPINAL CORD INJURY DECOMPRESSION CONFIRMED PROMISING EXPERIMENTAL DATA

26 INCREASED INCIDENCE OF ODONTOID FRACTURES IN THE ELDERLY POPULATION—RESULTS FROM THE SWEDISH NATIONAL HOSPITAL DISCHARGE REGISTER DURING 1997 TO 2009

Background: The role of early spinal cord injury decompression is clearly evident from experimental studies. Results of clinical studies are controversial. Pilot study evaluating decompression within first 8 h in a spinal center proved that early surgery is feasible in a low number of patients. Objective: Comparison between injury-to-surgery time of patients that incurred cervical spinal cord injury graded as ASIA A treated in nearest community hospital and spinal center. Methods: All patients included in the study sustained cervical spine fracture dislocation and had total senso-motor deficit upon admission. The Trial group (no = 12) patients were treated in County hospital Pula (Fig. 1). The control group was comprised of 12 out of 25 patients treated in spinal center following transportation from a hospital of first admission. Patients were matched according to age, gender and mechanism of injury. X-ray and CT scan were performed in the trial group and CT myelography or MRI in the control group. For statistical analysis of age, gender, mechanism of injury, encroachment and injury-to-surgery time Mann–Whitney U test was used. All P values below 0.05 were considered significant. For the statistical analysis of neurological improvement signal to noise method was used. Results: Median (range, years) age was 21 (12–45) and 27 (19–36) (P [ 0.05), male to female ratio 10:2 and 8:4 (P [ 0.05), and motorvehicle-accident: head diving ratio 9:3 and 8:4 (P [ 0.05) in the trial and control group, respectively. Median (range, hours) injury-tosurgery time was 5 (2, 5–9) in trial and 19 (17–24) in control group (P \ 0.001). There were no neurological improvements in the control group. Three patients in the trial group recovered neurological function (Fig. 2). Signal to noise rate ratio was six. Conclusion: Our results showed that transportation and diagnostic work-out can delay emergency decompression of injured cervical spinal cord. Three neurological recovered out of 12 patients agreed with laboratory studies results of early decompression benefit. Significant therapeutic effect was confirmed by sound out of noise method and should not be neglected until waiting for results that will be obtained from a longing randomized control trial. Keywords: Secondary injury prevention, spinal cord injury, surgical decompression, traumatic myelopathy

Y Robinson, G Schmeiser, B Sande´n, C Olerud Institute for Surgical Sciences, Uppsala University Hospital, Uppsala, Sweden Objective: Odontoid fractures are a common injury in the elderly population and treatment is often complicated due to patient morbidity. The number of geriatric patients increased during the last decades due to an increase in mean population age. Until now little epidemiological data is available allowing investigating whether the number of odontoid fractures increased during the last decade. Methods: Data for all patients with odontoid fractures admitted to hospital between 1997 and 2009 were abstracted from the Swedish National Hospital Discharge Register (SNHDR). The data in the register are collected prospectively, recording all inpatient admissions throughout Sweden. The SNHDR uses the codes for diagnoses at discharge and surgical procedures according to the Swedish version of the International Classification of Diseases (ICD). Results: A total number of 4444 patients (2072 women, 2372 men) with odontoid fractures were treated as inpatients in Sweden during the years from 1997 to 2009. Of these 1267 were operated. The annual incidence of odontoid fractures showed a linear increase during the years (r = 0.92). This was mainly due to an increase in the geriatric subgroup, while the younger age groups remained unchanged during the observation period. Interestingly the percentage of operated patients rather decreased until 2009 (r = -0.65). Discussion: While the elderly population grows dramatically, the number of hospital admissions due to elderly-specific odontoid fractures increased during the last decade. Possible explanations are greater awareness of fractures, improved diagnostics, and a higher activity level of the patients. Due to the continuous increase of odontoid fractures treatment modalities have to be optimised to reduce fatalities. Obviously there is a trend to more conservative treatment in Sweden. Randomised controlled trials allowing evidence-based recommendations are needed to establish a treatment rationale for odontoid fractures.

KS Duric, M Stancic Department of Neurosurgery, The university of Zagreb School of medicine, Zagreb, Croatia

123

S434

Fig. 1 Patients injured on Istrian peninsula (blue circles and lines) were surgically treated after first admission in the County Hospital Pula—Marine-spital MDCCCLXI. Postoperatively foreigners were transported homeward (white quadrants and lines). Patients that after first admission in a local hospital were re-transported in the level 1 trauma center are presented with black triangles and lines

123

Eur Spine J (2011) 20 (Suppl 4):S421–S464

Fig. 2 Although MRI of the patient that sustained compressive fracture of CVII vertebra and consecutive SCI showed large posttraumatic cavitation. He works his premorbid job

Eur Spine J (2011) 20 (Suppl 4):S421–S464 28 INFLUENCE OF CERVICAL SPONDYLOSIS ON LOW BACK PAIN M Hoshimaru, Y Kawanabe Department of Neurosurgery, Ohtsu Municipal Hospital, Ohtsu, Japan Introduction: Low back pain (LBP) is one of the most common symptoms in humans and is assumed to be due to structural abnormalities in the lumbar spine. However, some authors have reported cervical cord compression may cause LBP. This study was conducted to elucidate the influence of cervical spondylosis on LBP. Materials and method: Between October 2007 and October 2009, 193 consecutive patients underwent cervical laminoplasty using hydroxyapatite implants for cervical spondylotic myelopathy. LBP was assessed using visual analogue scale (VAS) before and after cervical laminoplasty. Fifty seven of 193 patients had a VAS score of 5 or above before surgery and were subjected to this study. This patient group comprised 33 men and 24 women ranging in age from 40 to 89 years (mean 66 years). Radiological studies of the lumbar spine were added in the patient group. The surgical results of cervical laminoplasty were evaluated using the Japanese Orthopaedic Association (JOA) score. Result: The average JOA score was 13.6 ± 1.9 before surgery and 16.1 ± 0.9 after surgery, and the improvement was maintained during the follow-up period. In 27 patients who did not have lumbar structural abnormalities, the VAS score of LBP decreased from 6.1 ± 0.9 before surgery to 0.4 ± 0.5 after surgery and this amelioration lasted to the time of the final follow-up. Remaining 30 patients who showed structural abnormalities of the lumbar spine exhibited modest amelioration of LBP after surgery. In 16 of the 30 patients, the VAS score of LBP decreased from 6.1 ± 1.0 before surgery to 0.5 ± 0.5 after surgery and lumbar decompression surgery was unnecessary. However, in remaining 14 of the 30 patients, VAS score of LBP increased from 5.8 ± 0.9 before surgery to 6.2 ± 1.0 at the time of the final follow-up in spite of transient slight improvement after surgery (an average VAS score of 3.0 ± 1.8), and lumbar laminectomy was performed and resulted in amelioration of LBP. Discussion and conclusion: The above data demonstrate that cervical spondylosis may cause LBP by unknown mechanism and LBP may be alleviated by cervical decompression surgery. Even if patients have lumbar canal stenosis, cervical decompression surgery provide a beneficial effect on LBP although it is limited.

29 MULTIMODALITY TREATMENTS FOR PERIMEDULLARY AV FISTULAS K Hida, T Aoyama, T Asano, K Houkin Department of Neurosurgery, Hokkaido University, Sapporo, Japan Object: Perimedullary arteriovenous fistulas (AVF) lead to myelopathy due to hemorrhage, venous congestion, and mass effect from dilated veins or varix. The goal of treatment is to interrupt the AV shunt by either surgery or embolization. The treatment choice is affected by the involvement of the anterior spinal artery, the location and number of AV shunts, and the presence of associated varix. We analyzed clinical and radiological findings and surgical results of perimedullary AVF. Methods: Since 2000, we have treated 101 cases of spinal AVMs. Among them, 32 cases were perimedullary AVF. They were 18 males and 14 females ranging in age from 3 to 76 years (mean 38 years).

S435 The AV shunts were located in the cervical (n = 8), the thoracic (n = 12), the conus medullaris (n = 10), and filum tereminale (n = 2).. The AVF were fed by the anterior- (n = 26) or posterior spinal artery (n = 6); 26 patients underwent surgery alone, 5 were treated by embolization, and one treated by embolization and surgery. During the surgery, all surgery cases underwent intraoperative DSA. Results: Postoperative angiography revealed complete disappearance of the AV shunt in 25 patients (78%), small fistulas persisted in the other 7 patients (22%); they had large lesions in the conus medullaris that were fed by the anterior spinal artery. Post-treatment, the neurological status was improved in 13 patients, stabilized in 19, and worse in 1 patient. Conclusion: Multimodality treatments including preoperative embolization, intraoperative digital subtraction angiography, motor-evoked potential monitoring, and dye injection are highly useful to treat perimedullary AVF.

30 IN VIVO WEAR PROPERTIES FROM THIRTY EXPLANTED PRODISC-C DEVICES—RETRIEVAL ANALYSIS OF CERVICAL TOTAL DISC REPLACEMENTS DR Lebl, FP Cammisa, FP Girardi, S Lee, T Wright, C Abjornson Spine and Scoliosis Surgery, The Hospital for Special Surgery, New York, USA Introduction: Cervical total disc replacements (CTDRs) have shown promising results compared to conventional anterior cervical discectomy and fusion (ACDF) procedures in several prospective clinical studies. The potential to preserve motion at the operative spinal level by implantation of a CTDR has many potential benefits. To understand the kinematics of these devices in vivo, we performed a prospective analysis of retrieved CTDRs to examine for evidence of wear, surface damage, and bony fixation. Methods: Prodisc-C CTDR devices from revision operations were cleaned and catalogued according to an IRB-approved retrieval program. Polyethylene (PE) and metallic (CoCrMo) components were examined using light stereo-microscopy (6X–31X) for wear, surface damage, and bone ongrowth. Areas of interest were evaluated at higher magnifications using scanning electron microscopy (SEM). Energy dispersive X-ray analysis (EDXA) was employed to identify the source of embedded 3rd body debris. Results: Thirty CTDRs from 29 patients of age 45.1 ± 1.9 years (range 31–57) were studied after a mean implantation time of 372 ± 64 days (range 2–1,295). The operative level was C4–C5 in 20%(6/30), C5–C6 in 47%(14/30), C6–C7 in 20%(6/30), and unknown in 13%(4/30). Indications for revision were axial pain (n = 9), radicular symptoms (n = 6), atraumatic loosening (n = 6), trauma (n = 5), metal allergy (n = 1), myelopathy (n = 1), hypermobility of the spinal segment (n = 1), and unknown (n = 1). Burnishing consistent with impingement was present in 90%(27/30); posteriorly in 63%(19/30), anteriorly in 23%(7/30) and circumferentially in 17%(5/30). There was no evidence of backside wear (n = 16). Third body wear was observed in 23%(7/30). Ongrowth was seen on both components in 43%(13/30), on one component in 40%(12/30), and on neither component in 17%(5/30). Discussion: Endplate impingement of the Prodisc-C occurs in the majority of implants in vivo. Anterior placement of the implant center of rotation may result in posterior impingement in extension during a physiologic range of motion. Backside wear was not present at early time points. 3rd body wear occurs following CTDR. Further longterm follow-up is needed to determine the clinical sequelae of wear patterns on patients outcomes.

123

S436

Eur Spine J (2011) 20 (Suppl 4):S421–S464 31 CFR-PEEK BASED POLYMER-ON-POLYMER ARTICULATIONS A NEW STANDARD FOR CERVICAL DISC PROSTHESIS—AN IN VITRO AND IN VIVO STUDY COMPARING UHMWPE AND CFR-PEEKK Kabir, J Schwiesau, R Pflugmahcer, C Burger, DC Wirtz, TM Grupp Orthopaedics and Trauma Surgery, University Hospital Bonn, Bonn, Germany

Superior Component Prodisc C—Posterior burnishing

Inferior Component Prodisc C—Posterior burnishing

123

Introduction: Total disc arthroplasty (TDA) is designed to preserve motion and to decrease the risk of accelerated degenerative disease at adjacent levels. One aim of designing new implants is to use materials with low wear rate-behaviour, which produces wear debris with low biological activities. Our goal was to compare the biotribological and biological properties of carbon fibre reinforced PEEK (CFR-PEEK) and Ultra High Molecular Weight Polyethylene (UHMWPE) in cervical artificial disc. Methods: In vitro wear simulation was performed with a clinically established cervical artificial disc made of UHMWPE/CoCr29Mo6 in a direct comparison to experimental disc articulations made of CFRPEEK. Each material combination was tested for 10 million cycles with a customised 6 station spinal wear simulator. Gravimetric and geometric wear assessment, an estimation of particle size and morphology were performed. 36 female rabbits were randomly allocated to 3 groups: CFR-PEEK, UHMWPE-particles and sham. The particles were implanted into the epidural space of the cervical region by percutaneous technique (fluoroscopic guidance). Neurobehavioral observations were conducted at pretreatment, on day 1–14 postinjection, then weekly. The rabbits were sacrificed at 3 and 6 months. Histologic sections from the regional lymph nodes, organs, from remote and implantation sites, were analyzed for any abnormalities and inflammation. Results: The cumulative gravimetric wear rate after 10 million cycles of the polyethylene-on-cobalt-chromium was 12.1 ±1.4 mg, compared to 0.5 ±0.23 mg for CFR-PEEK (p \ 0.05). Histopathological examination revealed, that crystalline wear debris was seen in the vertebral canal of examined test injection sites surrounded by inflammatory cells in both particle groups. The inflammation was limited to the epidural space around the particles. CFR-PEEK demonstrated statistical less biological reactivity compare to UHMWPE (p \ 0.05). The time (3 or 6 months) after implantation did not effect the extent of histological response. Conclusion: CFR-PEEK demonstrated an excellent wear behaviour with a wear rate reduction in comparison to UHMWPE. CFR-PEEK particles showed a greater biocompatibility than UHMWPE with reduced inflammatory response in cervical epidural space. Therefore CFR-PEEK based articulations provide an viable alternative to UHMWPE on metal and have a high potential for next generation disc replacements.

Eur Spine J (2011) 20 (Suppl 4):S421–S464 32 SURGICAL TREATMENT OF CERVICAL SPINE TUMOURS: A REVIEW OF 60 PATIENTS M Shousha, H Boehm Spine and Orthopedic surgery, Zentralklinik Bad Berka, Bad Berka, Germany Tumours of the cervical spine are quite rare in comparison to the thoracic and lumbar regions. Between August 2004 and August 2010, 574 patients suffering from tumours of the vertebral column were surgically treated in our institution. The tumour was located in the cervical spine in 60 of them (10.5%) and these represent the material of this work. The mean age at surgery was 60.5 years and 66.7% were males. Five tumours were benign (8.3%). Primary malignant tumours—mostly plasmacytoma—were found in 12 patients (20%) and the remaining 43 patients (71.7%) had cervical metastasis, mostly from the lung followed by the kidney. Plasmacytoma as well as lung and renal metastasis had male predominance, while colon metastasis had female predominance. Neurological impairment was found in 23 patients and pathological fracture in 31. Metastases from the breast, colon and stomach have the highest percentage of pathological fractures (80–100%). The pathology was located in the vertebral body in most of the cases (88.3%) and involved a single segment in 43 patients. On the other hand, 14 patients had bisegmental affection and in three patients, the disease extended to involve three segments. Further metastatic lesions could be detected in 53.3% of the patients. Corpectomy was undertaken in most of the cases (78%). The mean intraoperative blood loss was 1338 ml, the largest amount being with renal metastasis (4883 ml) and the least amount in colon metastasis (310 ml). Neurological improvement was achieved in 56.5% of cases. However, revision surgery was necessary in 6 patients, mostly due to implant failure. The mean period of follow-up was 16 months. 40 patients (66.7%) died due to advanced malignancy. 22 cases (36.7%) died within the first 7 months after surgery, most of them had other metastatic lesions on presentation. More than 50% of patients with colon or lung metastasis died within the first 7 months, while all patients with plasmacytoma survived longer than 7 months. Conclusion: Cervical spine tumours represent 10.5% of all spinal tumours and are mostly malignant (91.6%). The most common pathology is lung metastasis, followed by plasmacytoma and renal metastasis. Neurological deficit occurs in 38.3% of cases and pathological fracture in 51.7%. Intraoperative bleeding is highest with renal metastasis. Although neurological improvement could be achieved in more than one half of the cases, 36.7% die within the first 7 months after surgery. 33 PROSPECTIVE RANDOMIZED COMPARISON BETWEEN SINGLE-LEVEL TRANSFORAMINAL (TLIF) AND POSTERIOR (PLIF) LUMBAR INTERBODY FUSION REGARDING QUALITATIVE AND QUANTITATIVE RADIOLOGIC CHANGES OF THE PARASPINAL MUSCLES AND CLINICAL OUTCOME P Strube, T Hartwig, E Hoff, C Perka, C Gross, M Putzier Klinik fu¨r Orthopa¨die, Centrum fu¨r Muskuloskeletale Chirurgie, Charite´—Universita¨tsmedizin Berlin, Berlin, Germany Aim of the study: Aim of the prospective randomized study was the evaluation of clinical parameters and of quantitative and qualitative radiologic changes of the posterior paraspinal muscles (PPM: multifidus and longissimus muscles) of the lumbar spine in surgically treated segments and superior adjacent segments after single-level posterior (PLIF) and minimally invasive transforaminal (TLIF) lumbar interbody fusion and transpedicular stabilization. Methods: 50 patients with chronic low back pain because of a singlelevel degenerative disc disease Modic II of the segments L4/5 or L5/S1

S437 were randomly divided into 2 groups. 25 patients of group TLIF received a minimally invasive TLIF using a transforaminally placed PEEK-cage and transpedicular fixation employing a paramedian muscle-sparing approach. 25 patients (group PLIF) underwent a PLIF procedure (2 posteriorly placed PEEK-cages and transpedicular fixation) employing a mid-line approach. One week as well as at 12 months postoperatively PPM-volume and fat content as well as qualitative 3Ddistribution of the fatty degeneration were analyzed based on CT. Moreover, clinical parameters (VAS, ODI) were compared. Results: Between 1 week and 12 months postoperatively, radiologic analysis of the treated segments revealed a significant loss of the PPM-volume and in a significant increase of the fat content in both groups. In intergroup comparison no significant difference in PPMvolume and fat content was found at the follow-ups. In group TLIF, fatty degeneration was located in the anterolateral parts of the muscles (longissimus muscle), whereas in group PLIF mostly medial parts (multifidus muscle) were affected. PPM-changes in the superior adjacent segments could not be observed in both groups over time. ODI and VAS improved significantly in both groups without significant differences between the groups at the follow-ups. Conclusion: As reflected by the clinical results, similar amounts of fatty atrophy of the PPM were observed employing a minimally invasive paramedian or a midline approach, but with different distribution. Compared to the midline approach, the paramedian approach does not lead to direct damage of the medial part of the lumbar multifidus, but it results in indirect damage of the PPM, especially the longissimus muscle. This is possibly caused by surgically damaging the rami dorsales. 34 RIGID SPINOUS PROCESS SYSTEM FIXATION VS. PEDICLE SCREW FIXATION IN A POSTEROLATERAL LUMBAR ARTHRODESIS SHEEP CONSTRUCT— COMPARISON OF FUSION RATES MA Hardenbrook, M Kapsokavathis, D Ruehlman, HB Seim, AS Turner Orthopedic Spine Surgery, Boston Spine Group, newton, USA Introduction: Pedicle screws are commonly used in fusion of the lumbar spine. Alternatively, a rigid spinous process fixation device (Aspen; Lanx, Inc) can be used to stabilize the spine with minimal soft tissue dissection or intra-operative fluoroscopy. Biomechanical testing has shown less rigidity in lateral bending/rotation compared to pedicle screws. It is uncertain what effect this decreased stiffness has on fusion rates. Methods: 24 sheep received a PLF at L5–6 with iliac crest autograft placed between the decorticated transverse processes. One group of 12 sheep received a spinous process fixation device, and a second group of 12 sheep received a bilateral pedicle screw and rod construct. Fusions were evaluated with CT scans at 3 and 6 months. 4 sheep from each group were euthanized at 3-months. The remaining 16 sheep were euthanized at the 6 months. CT scans were graded independently by 3 BE/BC spine surgeons. Left and right sides were graded independently. The examiners were blinded to the type of fixation. The CT scans were graded using a published grading system with 2 representing robust bridging fusion mass, 1, incomplete fusion and 0, no bridging bone. Difference among reviewers was evaluated by agreements (Kendall’s tau statistics) between reviewer pairs. Animal level fusion rate were analyzed by using a Fisher’s exact test. Results: There was nearly complete concordance between examiners in grading the fusion mass. The tau statistics between observers were 0.90, 0.91, and 0.83. Evaluating each side independently, 7/8 Aspen devices were graded as 2 versus 3/8 in the pedicle screw construct at 3 mos. At 6 months, 15/16 Aspen devices were graded as 2 versus 8/18 of the pedicle screw group. This was statistically significant

123

S438 (P \ 0.003). In evaluating each sheep specimen using the highest grade of the two sides, all 12 sheep in the Aspen group graded 2, whereas 7/12 graded 2, 2/12, 1 and 3/12, 0 in the pedicle screw construct. This was statistically significant (P \ 0.037). Conclusion: In this sheep model, the decreased rigidity of the spinous process fixation device enhanced the bone formation and fusion rates when compared to the more rigid pedicle screw fixation construct.

Eur Spine J (2011) 20 (Suppl 4):S421–S464 35 IMPACT OF BIOMEDICAL AND BIOPSYCHOSOCIAL EDUCATIVE MODULES ON THE ATTITUDES, BELIEFS AND RECOMMENDATIONS OF HEALTH CARE PROVIDERS ABOUT LOW BACK PAIN: A RANDOMISED CLINICAL TRIAL J Domenech, D Sa´nchez-Zuriaga, E Segura-Orti, B Espejo, JF Lison Physiotherapy Department, University CEU-Cardenal Herrera. Moncada (Valencia), Spain., Moncada, Valencia, Spain

Grade 2 fusion

Grade 0 fusion

123

Background: The attitudes and beliefs that patients hold towards low back pain (LBP) are strong predictors of clinical outcome. Also, the orientations that Health Care Providers (HCP) hold may have an impact on the patients by reinforcing maladaptative beliefs and behaviours and by limiting the clinicians adherence to Clinical Practice Guidelines (CPG). Little is known about the impact that medical formation has to shape the HCP beliefs and attitudes towards LBP. Objective: to determine the effect of two educational modules with different orientations (biomedical or biopsychosocial) on changing HCP fear avoidance and pain and impairment beliefs and the recommendations given to patients. Methods: 184 physiotherapy students were randomly assigned to receive two different teaching modules of identical duration in addition to their ordinary academic clinical program. The intervention module consisted in the presentation of the biopsychosocial model. The control intervention (biomedical module) consisted on classes of anatomy and biomechanics of lumbar spine without instructions about treatment. The fear avoidance beliefs were evaluated with the FABQ questionnaire and the beliefs that pain justifies disability by the HC PAIRS questionnaire. Their recommendation for work and activity were assessed in three clinical cases vignettes. Results: The recommendations for work and activity significantly correlated with the fear avoidance and pain-impairment beliefs in both groups. After the interventions those who received the biopsychosocial module reduced significantly their fear avoidance and pain impairment beliefs, which was strongly correlated with an improvement in activity and work recommendations. Although the biomedical module was designed as a placebo intervention without advising treatment recommendations those who attended at it worsened the FABQ score and were more restrictive in the recommendations for work and activity. Previous personal episodes of back pain did not influence this changes. Conclusion: Our results confirm for the first time the previous suspicion that a strictly biomedical education worsens the maladaptive beliefs of HCP and consequently makes their recommendations for work and activity not in agreement with CPG. The knowledge of the biopsychosocial model may improve adherence to CPG by improving beliefs and attitudes towards LBP. The implications of this study seem important in the design of the contents of continuous medical education.

Eur Spine J (2011) 20 (Suppl 4):S421–S464 36 COMPLIANT ARTIFICIAL DISC REPLACEMENT: IN VIVO IMPLANT STABILITY AND OSSEOINTEGRATION IN AN ANIMAL MODEL T Steffen, B Freeman, M Aebi Director Orthopaedic Research Laboratory, McGill University, Montreal, Biomedical Engineering Dept, Montreal, QC, Canada Long term, secondary implant fixation of Total Disc Replacements (TDR) can be enhanced by hydroxyapatite or similar osseo-conductive coatings, historically on metal substrates. The objective of this in vivo study was to investigate the early stability and subsequent bone response adjacent to an all polymer TDR implant over a period of 6 months. A scaled down version of the calcium phosphate coated polymeric TDR device was implanted into the disc space one level above the lumbo-sacral junction in 6 Baboons (Papio annubis). Anterio-posterior and lateral X-ray images of the full lumbar spine were captured prior to surgery, and at intervals up to 6 months post-operatively. Systematic review of these serial images was conducted for subsidence and migration by an independent radiographer. Fluorochrome markers (4), which contain molecules that bind to mineralization fronts, were injected at intervals (7, 13, 19 and 25 weeks post-op) in order to investigate bone remodeling over time. Animals were humanely euthanized 6 months after index surgery. Test and control specimens were retrieved, fixed and subjected to histological assessment of the bone-implant-bone interface. Fluorescence microscopy and confocal scanning laser microscopy were utilized with BioQuant image analysis to determine the bone mineral apposition rates and gross morphology. Radiographic evaluation revealed no loss of disc height at the operative level or adjacent levels. No evidence of subsidence or significant migration of the implant up to 6 months. Heterotopic ossification was observed to varying degrees at the operated level. Histology revealed the implant primary fixation features embedded within the adjacent vertebral endplates. Fluorochrome distribution revealed active bone remodeling adjacent to the polymeric end-plate with no evidence of adverse biological responses. Mineral apposition rates of 0.7–1.7 microns/day are in keeping with literature values for hydroxyapatite coated implants in cancellous sites of various species. Radiographic assessment demonstrates that the calcium phosphate coated polymeric implant remains stable in vivo with no evidence of subsidence or significant migration. Histological analysis suggests the primary fixation features are engaged, and in close apposition with the adjacent vertebral bone. Fluorochrome markers provide evidence of a positive bone remodelling response in the presence of the implant.

37 LUMBAR SPINAL STENOSIS: A SIMPLE, PRACTICAL AND RELIABLE CLASSIFICATION M Pietrek, C Spitzer, L Papavero Clinic for Spinal Surgery, Scho¨n Klinik Hamburg Eilbek, Hamburg, Germany Background context: The incidence of lumbar spinal stenosis (LSS) shows an increase due to the increase in life expectancy, and averages 10% in the population of 65 years and older. Decompression for LSS is already the most frequent procedure in spine surgery, both in Europe and in the US. So far there is no distinct classification of LSS,

S439 terms such as ‘‘relative’’ or ‘‘absolute’’ are not defined. Therefore, a modified classification is presented for better comparison of different degrees of LSS. This MRI-based classification divides LSS into four grades (A-D). It includes the nerve root sedimentation sign, the level of cerebrospinal fluid, and the distribution of epidural fat (Fig. 1). Purpose: To determine inter-rater and intra-rater reliability of the presented AD-classification of LSS. Methods: Four experienced spine surgeons (senior surgeons) and three residents (junior surgeons) graded 100 axial T2 MRI-scans of LSS of different degrees as well as normal findings according to the AD-classification (normal finding, grade A, B, C or D). This grading was repeated after 2 weeks under similar conditions, but in a different sequence. Inter-rater reliability was determined using Fleiss’ kappa coefficient, both overall and for each of the two groups. Intra-rater reliability was measured using Cohen’s kappa. Results: Inter-rater reliability shows a good level of agreement overall (= 0.70, standard error (SE) = 0.01), as well as within the senior surgeons (= 0.77, SE = 0.02) and within the junior surgeons (= 0.74, SE = 0.03). Intra-rater reliability for the senior surgeons ranges from = 0.73 (SE = 0.06) to = 0.89 (SE = 0.06), and for the junior surgeons from = 0.71 (SE = 0.06) to = 0.86 (SE = 0.06). Conclusion: The AD-classification of LSS presented shows a good inter-rater as well as intra-rater reliability, in both experienced and less experienced clinicians. Therefore, it can be easily used to communicate and compare the degree of LSS on T2 axial MRI-scans in clinical practice.

Fig. 1 AD-classification of lumbar spinal stenosis

38 SIX SENIOR SPINE SURGEONS ARE NOT ABLE TO RELIABLY GRADE FACET DEGENERATION AND MULTIFIDUS DEGENERATION! J Franke, JC LeHuec, K Scheufler, U Liljenqvist, R Hedlund, H Chhabbra, F Awiszus Department of orthopedics, University of Magdeburg, Magdeburg, Germany Introduction: As spinal surgery relies more and more on dynamic solutions the assessment of the posterior structures of a lumbar segment are of paramount importance. To assess the degeneration of the facet joint Weishaupt and Fujiwara concluded that the grading could

123

S440 be validly done by MRI. We want to test this with 6 experienced spine surgeons as investigators in an everyday setting. Material and method: Out of the pool of available MRI‘s of the lumbar spine 24 MRI0 s of 24 patients were chosen by two (jf and jcl) of the authors. Inclusion criterion was just an MRI investigation with sagittal images in t1 and t2 and axial images of L3/4 to L5/S1 with a variety of degenerative status. For investigation one sagittal view for t1 and t2 and two optimal axial images of both t1 and t2 for the lumbar segments L3/4 to L5/S1 was taken for each patient. The grading was done on two separated occasions. All MRI0 s were assessed on t1 and t2 images. Additionally all investigators were asked to grade the muscle fatty degeneration left and right for each level with just minus (grade 1) or plus (grade 2) 50%. The statistical analysis comprised the overall kappa values for 6 investigators and 4 grades and weighted kappa values in between the 6 investigators. Results: The generalised Kappa for all 6 investigators and 4 grades was for t1 Kappa = 0.20169 (95% CI 0.17485–0.22854) and resembles only a moderate agreement. The generalised Kappa coefficient for t2 was less as for t1 with Kappa = 0.15437 (95% CI 0.12819–0.18055). From the 4 different grades we found a non significant agreement for the indicative and clinically relevant grade 2 for t1 Kappa = 0.08330 (95% CI -0.05395 to 0.22056) as well as t2 Kappa = 0.06095 (95% CI -0.07562 to 0.19753). The best agreement was found for grade 1 and 4 (Kappa von 0.21 to 0.47; 95% CI 0.16–0.56). The weighted kappa values in between two investigators were found to be in a wide range from 0.277 to 0.724. The generalised Kappa fort he muscle score for T1 was for 6 investigators and 2 grades = 0.39951 (95% CI 0.35733–0.44168) and for T2 = 0.38559 (95% CI 0.34342–0.42776). Conclusion: The basic conclusion is that facet grading using MRI investigation in the currently used classification by Fujiwara needs to be questioned for the everyday use. Even for the multifidus there is a higher agreement but still far from prefect. The authors want to stress the need for further improvement of current grading systems.

123

Eur Spine J (2011) 20 (Suppl 4):S421–S464 39 REDUCTION OF MUSCULAR INSUFFICIENCY BY MINIMAL INVASIVE MUSCLE PRESERVING APPROACH FOR SPONDYLODESIS IN PATIENTS WITH DEGENERATIVE LUMBAR INSTABILITY AND OBESITY (BMI [ 35) TH Pfandlsteiner, C Wimmer Center for Spine surgery with Scoliosis center, Scho¨n-Klinik Vogtareuth, Vogtareuth, Germany Background: Obesity (BMI [ 35) and lumbar de novo scoliosis is often combined with muscular insufficiency of the M. erector spinae. Aim of the percutaneous minimal invasive approach with or without decompression, is reduction of the muscle damage. Quicker rehabilitation and lower adjacent segment problems are seen. Outcome of open conventional spondylodesis (group I) and percutaneous stabilization combined with spondylodesis (group II) in obese patients degenerative instability have bee compared. Method: Retrospective, monocenter study, 06/2006–12/2010. T-score \-2.3, group I: 65 and group II: 58 patients. Age at operation was 55a (48–78), BMI 39 (35- 43). Bi-, tri- and polysegmental instrumentation, VAS and patient satisfaction score have been compared. Indication: Degenerative lumbar scoliosis, osteochondrosis, spondylolisthesis and FBSS. To control fusion X-ray or CT-scan of the lumbar spine after 6, 12, 24 months was done. Muscle insufficiency was measured by pain threshold measurement with a digital dolorimeter at defined triggerpoints before and 6, 12 and 24 months after operation. Results: Fusion rate in the dorsal group was 85% (group I and II) and in the dorsoventral group was 92% in both groups. Follow up 28 (22–36) months, lost to follow up 1/65 and 0/58. Screw loosening combined with pseudo-arthrosis in group I was found in 7 patients (19 screws) and in group II in 2 patients (3 screws), adjacent disc degeneration in group I 6/65 and in group II 1/58, screw breakage in group I 1/65 and in group II 0. The length of walking distance improved in the MIS group 3 months earlier, the load-carrying capacity and ability was reached 1.5 months earlier. Demand for rehabilitation after fusion was significant lower in group II (92%/ 83%). Triggerpoint measurement showed a increase from pre OP 1.8 (0.7–2.2) kg/cm to 2.3 (1.6–3.8) kg/cm 6 months post OP and 4.6 (3.0–6.7) kg/cm 12 months post OP. At least control 24 months post OP they were 5.5 (4.6–8.9) kg/cm. The patient satisfaction score was much better in the MIS group, and the VAS score decreased significant earlier. Conclusion: Rate of adjacent disc degeneration and screw loosening is significant lower in the MIS group. The preoperative existing muscular insufficiency in obese patients is not that much increased in the MIS group than in the conventional group after operation. The minimal invasive, percutaneous instrumentation shows advantages especially in obese patients.

Eur Spine J (2011) 20 (Suppl 4):S421–S464 40 SENSORY AND MOTOR DEFICIT FOLLOWING LATERAL LUMBAR INTERBODY FUSION M Pumberger, AP Hughes, RR Huang, AA Sama, FP Cammisa, FP Girardi Orthopedic Surgery, Spine and Scoliosis Service, Hospital for Special Surgery, New York, USA Lateral lumbar interbody fusion (LLIF) as a minimally-invasive technique has gained growing interest in recent years. However, one of the procedure’s limitations remains the unknown incidence of sensory and/or motor deficit following a trans-psoatic approach. We seek to identify the incidence and nature of neurological events following LLIF. We performed a retrospective of 247 patients charts undergoing LLIF between 2006 and 2009. We identified the new onset of anterior thigh pain, sensory and motor deficit. Each sensory deficit was reported to the according dermatome. Motor deficits were divided into the muscle movement. Patient charts were reviewed for demographics, medical co-morbidities, subjective neurological complaints, and physical exam findings at the following time points: pre-, peri- and postoperatively at 6 weeks, 3, 6 and 12 months follow-up. The specific deficits were correlated to the side and level of approach, numbers of levels fused, length of surgery, and bone graft material employed. A multivariate logistic regression models were created to evaluate the independent associations of each potential explanatory variable. A total of 237 patients (139 Female, 98 Male) underwent LLIF surgery at our institution with a total of 463 levels fused. Average age 61.5 years (range 31–88) and mean BMI was 28.3 (range 17.4–60.3). At 6 weeks, 41.4%, 16.0% at 12 weeks, 3.7% at 6 months and 0.8% at 12 months experienced thigh pain after surgery. We found a significant correlation between the involvement of L 4–5 and the occurrence of motor deficits at 6 weeks (p = 0.009), 12 weeks (p = 0.009) and 12 months (p = 0.027). However, no correlation between a sensory deficit and the involvement of L1–2, L2–3, L3–4 was found (6 weeks (p = 0.347), 12 weeks (p = 0.999) and 6 months (p = 0.999) 12 months (p = 0.817). There is an association between the numbers of levels fused and the persistent nerve injury (12 months follow up p \ 0.001). Bone graft material do not have an influence on sensory or motor deficit at any time point. The multivariate logistic regression model showed an independent association of the following risk factor for a motor deficit at 6 weeks: gender (female), duration of surgery, BMI ([ 30) and involvement of L 4–5. LLIF remains a valuable tool for achieving fusion through a minimally invasive approach with little risk to neurovascular structures. Approach related iliopsoas weakness remains a common postoperative deficiency.

41 THE CORE OUTCOME MEASURES INDEX (COMI) FOR THE ROUTINE DOCUMENTATION OF OUTCOME IN PATIENTS UNDERGOING SPINE SURGERY: STILL PERFORMING WELL FIVE YEARS ON? AF Mannion, FS Kleinstu¨ck, UM Mutter, F Porchet, TF Fekete, F Lattig, D Jeszenszky, D Grob Spine Center, Schulthess Klinik, Zu¨rich, Switzerland Introduction: In 1998, Deyo et al. proposed the use of 6 single questions for the routine examination of outcome in patients with back problems. The questions have since been put together to form an

S441 outcome instrument (the Core Outcome Measures Index) and adapted for use in many different languages. After more than 5 years’ experience using the COMI in routine practice, we sought to evaluate its overall performance in the systematic patient-centred evaluation of surgical outcome. Methods: The COMI has one question each on back (or neck) pain intensity, leg/buttock (or arm/shoulder) pain intensity, function, symptom-specific well-being, general quality-of-life, work-disability and social-disability, and is scored as a 0–10 index. At follow-up (FU), patients rate the global treatment outcome and their satisfaction, on a 5-point Likert scale. Results: From 1.1.2005 to 31.12.2010 the questionnaire was completed by 6644/7344 (90%) spine patients before surgery, 6423/6900 (93%) at 3 mo FU, 5480/5987 (92%) at 12 mo FU, 4170/4632 (90%) at 24 mo FU, and 940/1106 (85%) at 5 y FU. In a sub-group of patients operated for lumbar degenerative disease (N = 452, 56% female, age 60 ± 15 years), with pain relief as one of the goals of surgery and with completed COMI questionnaires at all follow-ups (up to 5 years), mean (± SD) COMI scores of 7.8 ± 1.7, 4.2 ± 2.7, 3.9 ± 2.9, 3.8 ± 2.9 and 3.7 ± 2.9 were recorded pre-op and at 3, 12, 24, 60 mo FU, respectively. The large and significant reduction in COMI seen at the first FU (effect size, 1.21) was retained with slight improvement up to 5 years later. Consistent with the COMI scores, a ‘‘good global outcome’’ (operation helped/helped a lot) was recorded by 79%, 76%, 75% and 81% at each time point, respectively. Conclusion: The questionnaire has proven highly feasible to implement on a prospective basis within the routine of a busy Spine unit ([ 1200 surgeries/year, [ 10 surgeons). The responsiveness (sensitivity to change) of the COMI score was as good as or better than that of many longer symptom-specific outcome questionnaires. The shortness of the questionnaire and its multidimensional nature make it an attractive option to comprehensively assess almost every patient undergoing surgery. We therefore continue to recommend its widespread and consistent use, to allow standardisation of outcome measurements in future trials, multicentre studies, in-house quality management systems and surgical registries.

42 OUTCOMES OF FUSION SURGERY FOR OSSIFICATION OF THE POSTERIOR LONGITUDINAL LIGAMENT OF THE THORACIC SPINE: A MULTI-CENTER RETROSPECTIVE STUDY M Matsumoto, Y Toyama, K Abumi, Y Nohara, H Chikuda, N Kawahara, M Yamazaki, T Kato, S Imagama, Y Matsuyama Dept. of Orthopaedic Surgery, Keio University, Tokyo, Japan Object: To evaluate the outcomes of fusion surgery for patients with ossification of the posterior longitudinal ligament in the thoracic spine (T-OPLL), and to identify factors significantly related to surgical outcomes. Methods: This study included 76 patients (34 men and 42 women; mean age, 56.3 years) who underwent fusion surgery for T-OPLL at seven spine centers during the five-year period from 2003 to 2007. The study examined the patients’ demographic data, underlying disease, preoperative comorbidities, past history of spinal surgery, radiological findings, surgical methods, surgical outcomes, and complications. Surgical outcomes were assessed using the JOA scores for thoracic myelopathy (11 points) and the recovery rate. Results: The mean JOA score was 4.6 2.1 points preoperatively, and 7.7 2.5 points at the time of the final follow-up examination, yielding a mean recovery rate of 45.4 39.1%. The recovery rates by surgical

123

S442 method were 38.5 ± 37.8% for posterior decompression and fusion, 65.0 ± 35.6% for anterior decompression and fusion via an anterior approach, 28.8 ± 41.2% for anterior decompression via a posterior approach, and 57.5 ± 41.1% for circumferential decompression and fusion. The recovery rate was significantly higher for patients treated using an anterior approach than for those treated using a posterioronly approach, and for patients without diabetes mellitus (DM) compared to those with DM. One or more complications were experienced by 31 patients (40.8%), including 20 patients with postoperative neurological deterioration, 7 with dural tears, 5 with epidural hematoma, and 4 with respiratory complications. Conclusions: The outcomes of fusion surgery for T-OPLL were favorable. An anterior approach and the absence of DM correlated with better outcomes. However, a high rate of complications was associated with the surgery.

43 RECRUITMENT OF HUMAN ANNULUS FIBROSUS CELLS BY THE CHEMOKINES CXCL10, XCL1 AND HUMAN SERUM AA Hegewald, K Neumann, G Kalwitz, U Freymann, M Endres, K Schmieder, C Kaps, C Thome´ Department of Neurosurgery, Innsbruck Medical University, Innsbruck, Austria Biological attempts to close the ruptured annulus fibrosus with repair tissue occurs spontaneously after a disc herniation incident. Although chemokines are suggested to play a role in resorption of herniated disc tissue, the role of chemokines in annulus fibrosus homeostasis and repair remains unclear. The objective of this study was to investigate spontaneous repair mechanisms and underlying cell recruitment in degenerative annulus defects. Therefore, human annulus fibrosus tissue and cells were analyzed for the presence of chemokine receptors and the migratory effect of selected chemokines on them. Cells were isolated from annulus fibrosus tissue and expanded in the presence of human serum. Multiwell-chemotaxis assays were used to analyze the migratory effect of human serum and 0 to 1,000 nM of the chemokines CXCL7, CXCL10, CXCL12, CCL25 and XCL1 on annulus fibrosus cells (n = 9 per chemokine and dose). Presence of corresponding chemokine receptors on annulus fibrosus cells was determined by real-time PCR analysis and immuno-histochemistry. Serum (0.1 to 10%) significantly (p \ 0.01) stimulates the migration of annulus fibrosus cells. Compared to untreated cells, the migration of cells was significantly (p \ 0.01) enhanced upon stimulation with 100–1,000 nM CXCL10 and 1,000 nM XCL1. Chemokine receptors showed low expression levels in expanded annulus fibrosus cells as assessed by PCR. Immuno-histochemical staining of the CXCL10 receptor CXCR3 and the XCL1 receptor XCR1 showed presence of the particular receptors in annulus fibrosus cells expanded under conventional cell culture conditions. In native annulus fibrosus tissue, CXCR3 could be detected, while XCR1 was not evident. The findings suggest that chemokines, in particular CXCL10, effectively recruit isolated cells from the annulus fibrosus. This suggests that chemokines have an important role in annulus fibrosus homeostasis and possibly in spontaneous annulus repair attempts. This might have important implications for biological annulus sealing strategies.

123

Eur Spine J (2011) 20 (Suppl 4):S421–S464 44 LONG-TERM CULTURE OF BOVINE NUCLEUS PULPOSUS EXPLANTS IN A NATIVE ENVIRONMENT BGM van Dijk, E Potier, K Ito Biomedical Engineering, Eindhoven University of Technology, Eindhoven, Netherlands Regenerative therapies are promising treatments for intervertebral disc degeneration, and to screen for their efficacy, a long-term in vitro model would be valuable. As some of the most substantial degenerative changes which must be reversed occur in the nucleus pulposus (NP), NP explant culture may constitute such a model. This model should include hypoxia, low glucose, low pH as well as high osmolarity to prevent swelling, proteoglycan loss and maintain a native cell environment. To prevent swelling, the NP tissue osmolarity can be balanced by increasing medium osmolarity or by using a physical constraint. In a previous study we showed that NP explants were successfully cultured in hyper-tonic medium for 21 days. However, the other factors of the native disc environment were not included. Moreover to investigate efficacy of treatments on tissue biochemistry, even longer culture durations are desirable. Therefore in this study, we investigated the effect of physiological oxygen (O2) and glucose levels on bovine NP explants, cultured either in a physical constraint or in medium with raised osmolarity for 42 days. Osmolarity was raised to iso-tonic and hyper-tonic levels compared to native tissue using polyethyleneglycol (PEG). O2 and glucose were adjusted to standard culture (21% O2 and 4.5 g/l glucose) or physiological (5% O2 and 1 g/l glucose) levels. Explants were analyzed for water, and biochemical (DNA, collagen, glycosaminoglycan (GAG) and fixed charge density) content, cell viability, gene expression and tissue histology after 42 days and compared to day 0 samples. After culture, there were no significant changes in GAG and DNA content for all conditions (Fig. 1B and C). In the constrained condition with standard O2 and glucose levels and in both iso-tonic PEG conditions, water content was increased (Fig. 1A). In the constrained condition with physiological O2 and glucose levels and in both hypertonic PEG conditions, all biochemical conditions were not significantly different from day 0. Safranin-O/Fast Green staining showed that the tissue composition of both approaches resembled day 0 in physiological conditions (Fig. 2). In conclusion, physiological O2 and glucose levels have a beneficial effect only in the constrained condition. Both approaches can maintain matrix composition specific to NP and important for cell behavior and thus constitute promising long-term models to test regenerative therapies. This research is part of the BMM project IDiDAS.

Fig. 1 Water content (A), sGAG content expressed per dry weight (B) and DNA content expressed per dry weight (C). Values are mean ± SD, n = 5, *p \ 0.05 compared to day 0

Eur Spine J (2011) 20 (Suppl 4):S421–S464

Fig. 2 Safranin-O/Fast Green stained sections of fresh and 42-days cultured NP explants. Day 0 (A); hyper-tonic PEG (B) and physical constrained (C) with physiological O2 and glucose levels

45 COMBINED HYPERBARIC OXYGEN AND HYPOTHERMIA TREATMENT ON OXIDATIVE STRESS PARAMETERS AFTER SPINAL CORD INJURY: AN EXPERIMENTAL STUDY K Topuz, A Colak, B Cemil, M Kutlay, M Demircan, A Eroglu, O Ipcioglu, Z Kucukodaci, G Uzun Neurosurgery, GATA Haydarpaa Training Hospital, Istanbul, Turkey Background and aims: This study was aimed to investigate the possible beneficial effects of combined hypothermia and hyperbaric oxygen (HBO) treatment in comparison with methylprednisolone in experimental spinal cord injury (SCI). Methods: Forty eight male Wistar albino rats (200 g to 250 g) were randomized into six groups; A (Normothermic control group; only laminectomy), B (Normothermic trauma group; laminectomy + spinal trauma), C (Normothermic methylprednisolone group; laminectomy + spinal trauma + methylprednisolone treated), D (Hypothermia group; laminectomy + spinal trauma + hypothermia treated); E (HBO group; laminectomy + spinal trauma + HBO therapy), F (Hypothermia and HBO group; laminectomy + spinal trauma + hypothermia and HBO treated) each of contains 8 rats. Neurological assessments were performed 24 h after trauma and spinal cord tissue samples had been harvested for both biochemical and histopathological evaluation. Results: After SCI, tissue malondialdehyde (MDA) level of the control group was measured increased, and superoxide dismutase (SOD), glutathione peroxidase (GSH-Px), and catalase (CAT) enzyme activities were measured decreased. In group F, it was also shown that MDA level elevation had been prevented, and group F has increased the antioxidant enzyme activities than the other experimental groups C, D, E (p 0.05). Conclusions: We concluded that the use of combined hypothermia and HBO treatment might have potential benefits in spinal cord tissue on secondary damage.

S443 patients has been reported to be twofold higher than in healthy people, which is higher than the risk expected using the bone mineral density (BMD). Therefore, it is very important to investigate the bone strength in order to discuss the fracture risk in patients with RA. Purpose: The purpose of this study was to evaluate the risk of lumbar spine fracture using CT-based finite element methods and FRAX. Patient sample: Forty-five patients with RA and 18 patients with postmenopausal osteoporosis (PO) were included in this study after obtaining fully informed written consent. All subjects were postmenopausal females. Methods: We recorded the age, body height, body weight, and body mass index. BMD values of the lumbar spine and right femoral neck were evaluated using dual-energy X-ray absorptiometry. All patients were assessed by CT to evaluate the vertebral bone strength in the lumbar spine (L2). When a compression force was gradually loaded onto the upper endplate of L2 using Mechanical Finder software, the compression load at the first solid collapse was defined as the fracture load. FRAX values were calculated for all subjects using FRAX software, proposed by the WHO. Mann–Whitney’s U test was used for statistical analysis between the two groups. Spearman’s rank correlation coefficient was employed for analysis between the fracture load and other factors. Results: There was no significant difference in the body height, body weight, body mass index, and BMD at the femoral neck between the two groups. The average age of the RA group was lower than that of the PO group. The RA group had a higher BMD of the lumbar spine than the PO group. The average fracture load of L2 was 4,399.1 N and 3,646.7 N in RA and PO groups, respectively. The average FRAX for vertebral fracture was 17.2 and 15.6% in RA and PO groups, respectively. There was no significant difference in the fracture load and FRAX between the two groups (Table 1). A positive correlation was observed between the fracture load and BMD of the lumbar spine. Negative correlations were observed between the fracture load and age, and between the fracture load and FRAX (Table 2). Conclusions: Our data showed that RA patients exhibited poorer vertebral strength and more fracture risks than age-matched PO patients, which were not affected by glucocorticoid therapy. Table 1 Demographic data and results RA group

PO group

p-value

N

45

18

–

Age (years)

62.3 (36–77)

72.7 (59–84)

0.001

26.3 (15–42)

0.001

Duration of postmenopausal 15.7 (3–32) period (years) Body height (cm2)

153.3 (141.6–165.5) 149.2 (140.2–155.4)

0.399

Body weight (kg)

52.9 (38.5–71.6)

48.6 (39.6–58.0)

0.339

Body mass index

22.6 (17.2–33.8)

21.9 (16.9–25.0)

0.666

BMD

46 LOW BONE STRENGTH IN LUMBAR SPINE OF PATIENTS WITH RHEUMATOID ARTHRITIS T Dokai, H Hagino, H Nagashima, Y Nanjo, A Tanida, R Teshima Department of Orthopedic Surgery, Faculty of Medicine, Tottori University, Yonago, Japan Background: A decrease of bone minerals is often observed in patients with rheumatoid arthritis (RA). The fracture risk in RA

Lumbar spine (g/cm2)

0.833 (0.618–1.132) 0.701 (0.526–0.867)

0.030

Right femur (g/cm2)

0.676 (0.481–0.909) 0.644 (0.486–0.841)

0.335

Fracture load of L2 (N)

4,399.1 (400–9,400) 3,646.7 (1,700–5,292) 0.380

FRAX Vertebral fracture (%)

17.2 (5.6–35.0)

15.6 (6.8–27.0)

0.798

Femoral neck fracture (%)

5.6 (0.6–16.0)

5.2 (1.2–12.0)

0.648

Values were given as the mean (range) A Mann–Whitney’s U-test was employed for statistical analysis Statistical significance was set at p \ 0.05 BMD bone mineral density, RA rheumatoid arthritis, PO postmenopausal osteoporosis

123

S444

Eur Spine J (2011) 20 (Suppl 4):S421–S464

Table 2 Relationships between the fracture load and other parameters

Fx load

Fx load

Age

BMD (L)

BMD (F) FRAX (V)

FRAX (F)

–

-0.361 (0.015)

0.346 (0.025)

0.244 (0.123)

-0.294 (0.070)

-0.231 (0.158)

–

-0.419 (0.006)

-0.174 (0.277)

0.510 (0.001)

0.412 (0.009)

–

0.668 (0.000)

-0.402 (0.014)

-0.487 (0.002)

–

-0.527 (0.001)

-0.696 (0.000)

–

0.935 (0.000)

Age BMD (L) BMD (F) FRAX (V)

non-Powder sign (n = 47) included 14 (30%), and Powder sign included pseudarthrosis significantly less than non-Powder sign (p = 0.01). Black line (n = 24) included 11 (46%) pseudarthrosis, non-Black line (n = 39) included 4 (10%), and Black line included pseudarthrosis significantly more than non-Black line (p = 0.004). Kyphosis progression rate of Powder sign was 11% (± 21%) and that of non-Powder sign was 23% (± 25%), and there was no significant difference (p = 0.08). Kyphosis progression rate of Black line was 32% (± 22%) and that of non-Black line was 13% (± 22%), and there was a significant difference (p = 0.003). Conclusions: On MRI STIR, Powder sign can predict bone union, and Black line was the risk factor of pseudarthrosis, kyphosis, and back pain on osteoporotic vertebral fractures.

Values were given as the correlation coefficient (p-value) Spearman’s rank correlation coefficient was used to investigate associations Statistical significance was set at p \ 0.05 Fx fracture, BMD bone mineral density, L lumbar spine, F femur, V vertebrae

47 CAN MRI STIR PREDICT PSEUDARTHROSIS AND KYPHOSIS AND BACK PAIN OF OSTEOPOROTIC THORACOLUMBAR VERTEBRAL FRACTURES? H Omi, T Yokoyama, A Ono, T Numasawa, K Wada, Y Yamasaki, H Jin, T Naraoka, Y Fujisawa, S Toh

Fig. 1 Methods

Orthopaedic department, Odate Municipal General Hospital, Odate, Japan Objective: There were few reports about the relationship between MRI short-TI inversion-recovery (STIR) findings and prognoses of injured vertebrae. The purpose of this study was to assess STIR for predicting pseudarthrosis and kyphosis and back pain with osteoporotic vertebral fractures. Study design: prospective cohort study. Patient sample: From November 2008 to June 2010, sixty-three vertebrae in 56 patients (48 women and 8 men) were enrolled in this study. The average age was 77.5 years old (60–92 years old). Inclusion criteria were the presence of back pain after injury by minor trauma, age over 60 years old, MRI (1.5 Tesla) performed within 30 days after injury, fractured vertebra level from Th10 to L2, and treatment with thoracolumbar hard orthosis over 3 months. Exclusion criteria were bone tumor, bone marrow disease, inflammatory disease, infection, and dialysis treatment. Methods: On STIR, the midsagittal slice and adjacent slices (total 3 slices, 3 mm interval) of all injured vertebrae were evaluated. A homogeneous high intensity vertebra was defined as Powder sign. All injured vertebrae were divided into Powder sign or non-Powder sign. A low intensity line in a high intensity area was defined as Black line, and all injured vertebrae were divided into Black line or non-Black line. Pseudarthrosis, Kyphosis rate and Kyphosis progression rate were calculated on lateral view at 6 months X-p (Figure 1). Visual analogue scale for back pain (VAS) was evaluated at 6 months examination. Results: Fourteen of 63 vertebrae (22%) resulted in pseudarthrosis (Figure 2). Powder sign (n = 16) included 0 (0%) pseudarthrosis, and

123

Fig. 2 Results

Eur Spine J (2011) 20 (Suppl 4):S421–S464

S445

48 MRI INTERPRETATION OF POSTERIOR LIGAMENTOUS COMPLEX (PLC) SEQUENTIAL DAMAGE AFTER TRAUMATIC FRACTURES. PROSPECTIVE EVALUATION OF 74 SPINAL FRACTURES J Pizones, E Izquierdo, F Sa´nchez-Mariscal, L Zu´n˜iga, ´ lvarez-Gonza´lez, A Go´mez-Rice PA Orthopaedics, Spine Unit, Hospital Universitario de Getafe, Madrid, Spain Introduction: PLC plays an important role in vertebral stability. However, the sequence in which the different components of the complex tear as damage forces increase, has not been yet investigated. Methods: Prospective study of 74 consecutive acute vertebral traumatic fractures, analyzed using X-rays and MRI scan (FS-T2-w/STIR sequences). Fracture morphology was classified according to AO (Magerl et al.). Integrity of each PLC component-facets, interspinous ligaments (IS), supraespinous ligaments (SS) and ligamentum flavum (LF)- was assessed and classified as intact, edema or disruption. IS edema was further subdivided depending on the extension ([/\50%). We analyzed the association between MRI signal and the progressive scale of morphological damage AO (Chi2). Results: AO type A1/A2 fractures (compression forces) associated with only facet distraction. A3 fractures (axial forces) showed additional IS edema, usually\50%, with neither SS nor LF disruption. B1 fractures (flexion-distraction forces) associated with facet distraction, IS edema or disruption, and low rate of SS/LF disruptions; B2 fractures increased SS/LF disruption rates. C fractures (rotation forces) associated with facet distraction or disruption, and IS, SS or LF complete rupture. We found high association (p \ 0.001) between AO-MRI (Table 1). MR images analysis showed: •

•

•

Posterior distraction forces could begin in the facets, and extend throughout the IS ligament starting from the postero-superior margin (finally deattaching the SS superiorly) and heading the antero-inferior border, finally tearing the LF. The SS ligament can be found disrupted with only partial IS disruptions. LF distraction ruptures always appear after IS complete ruptures. IS ligaments can show any of the three signal patterns (intact, edema or disruption), SS and LF are found either intact or disrupted.

Conclusion: MR images associate with AO progressive scale of morphological damage, showing what could be a common progressive sequence of rupture among the different components of the PLC as traumatic forces increase. Table 1 Data of the 74 fractures showing the distribution of MRI findings (FS T2-w/STIR sequences) among the different fracture patterns Facet alterations IS edema

IS rupture

SS rupture

LF rupture

N

BME 12.5% AO 25% A1

0 (2) 8%

0 0

0 0

0 0

8 24

60% AO A2

0

0

0

0

5

84% AO A3

(6) 46%

0

0

0

13

Table 1 continued Facet alterations IS edema

IS rupture

SS rupture

LF rupture

N

AO 79% B1

(10) 71%

21%

36%

29%

14

AO 67% B2

(1) 16%

83%

67%

67%

6

AO – B3

–

–

–

–

0

AO 100% C1

0

100%

100%

100%

1

AO 100% C2

(1) 33%

66%

100%

67%

3

AO – C3

–

–

–

–

0

P

0.000

0.000

0.000

0.000

0.000

Chi2 AO classification (Magerl et al.); BME bone marrow edema/vertebral contusion, IS interspinous ligament, SS supraspinous ligament, LF ligamentum flavum, N number

49 DEVELOPMENT AND VALIDATION OF THE NEW AOSPINE THORACOLUMBAR INJURY CLASSIFICATION SYSTEM KJ Schnake, C Oner, C Bellabarba, M Reinhold, LY Dai, L Audige´ Center for Spinal Surgery and Neurotraumatology, BG Trauma Clinic Frankfurt, Frankfurt, Germany Introduction: Spinal fracture classifications like the Denis and Magerl comprehensive classification are widely in use. However, they were based on retrospective studies of thoracolumbar injuries without a well defined methodological concept or scientific validation. The need for a comprehensive, valid and universally accepted spinal injury classification system motivated the AOSpine International to establish a classification group with the aim to develop such system. We present the driving methodological and clinical issues, as well as the current development stage. Methods: Based on the concepts of the Magerl classification, a group of clinical experts coming from different countries developed a scheme for thoracolumbar injuries which distinguishes between A (compression), B (tension band) and C (displacement) type injuries. Type A injuries include 4 subtypes (wedge-impaction/split-pincer/ incomplete burst/complete burst); B type injuries include 2 subtypes (pure osseous and (osseo-)ligamentous disruption); C type injuries were further categorized in 3 subtypes (hyperextension/translation/ separation). The validation process (plain X-rays and CT-scans) involved 4 agreement studies so far. Reliability and accuracy parameters were assessed by kappa statistics and latent class analysis, respectively. Results: 116 fractures were included in the last validation process. The reliability of identifying cases with A type injuries only was good (kappa 0.77). Complete burst fractures (A4) could be well differentiated from the other type A fractures (n = 86; kappa 0.84). Latent class analysis helped identify patterns of coding by participating surgeons, and thus support the revision and specification of definitions. Discussion: This development process represents a major scientifically valid effort to provide spine surgeons with a comprehensive

123

S446 classification of spinal trauma. After 4 evaluation sessions, results are very supportive and helped consolidating the current proposal. At this stage of the development the predictive clinical value of the proposed system is at best suspected by expert opinion, and its quantification will require follow-up evaluation sessions, which will be performed by a considerable number of surgeons from different countries.

Eur Spine J (2011) 20 (Suppl 4):S421–S464

TLICS 10

*

9

*

8

*

7

50 QUANTITATIVE TREATMENT DECISION MAKING FOR THORACOLUMBAR BURST FRACTURE

6 5

*****

Department of Orthopaedic Surgery, Kobe Red Cross Hospital, Kobe City, Japan Purpose: Classification and therapeutic strategy of thoracolumbar burst fracture is controversial. Load Sharing Classification Score (LSCS; McCormack T, et al. 1994) and Thoracolumbar Injury Classification and Severity Score (TLICS; Vaccaro AR, et al. 2005) are both quantitative evaluation system of thoracolumbar burst fracture. Theoretically, combination of these scoring systems is helpful not only for surgical indication but also surgical approach (i.e. necessity of anterior reconstruction). However, relationship of these scoring systems has not been investigated. The purpose of this study was to clarify the relationship of LSCS and TLICS, and to investigate the clinical usefulness of the combination of these systems. Methods: Total 102 cases treated in one institution from December, 2003 to June, 2010 were evaluated using LSCS and TLICS retrospectively. Results: Mean LSCS point with TLICS point of 5 or more (recommendation of surgical treatment) was 7.2 ± 1.2, and mean LSCS point with TLICS point of 3 or less (recommendation of conservative treatment) was 5.8 ± 1.4 (p \ 0.0001). On the other hand, mean TLICS point with LSCS point of 7 or more (recommendation of additional anterior reconstruction surgery) was 6.9 ± 2.3, and mean TLICS point with LSCS point of 6 or less (expectation of good clinical results with posterior short fusion) was 4.6 ± 2.7 (p \ 0.0001). Discussion: TLICS point and LSCS point were well correlated. The combination of these systems extracted conflicting cases with less than 4 TLICS points (indication for conservative treatment) and 7 or more LSCS points (indication for additional anterior reconstruction). Fractures with highly destructed vertebral body without neurological deficit or posterior ligamentous complex injury belong to this contradictory category. Further multidirectional analysis is required for these cases. Fundamental components of surgical indication are injury mechanism including posterior ligamentous complex injury and neurological deficit, which are comprehended in TLICS system. We believe that rigid spinal stability is not acquired by posterior short segment fixation alone when the anterior and middle column is severely injured. LSCS system can quantitatively evaluate the severity of vertebrae. As a conclusion, the combination of TLICS and LSCS is useful for treatment decision making for thoracolumbar burst fracture.

123

*

3 2

****

*

4

T Takigawa, K Koshimune, T Morita, S Mizuno, H Ohashi, Y Ito

**

****

****

****

******

***

***

*

**

****** *****

****

** **

*

**

**

***

******

***

*

*

* *

***

*****

*****

*****

**

3

4

5

6

7

8

1 9

LSCS

Distribution of LSCS point and TLICS point

51 DOSE RESPONSE AND STRUCTURAL INJURY IN THE DISABILITY OF SPINAL INJURY MS Patel, P Sell Spine and Orthopaedic Surgery, University Hospitals of Leicester NHS Trust, Leicester, United Kingdom Introduction: In almost all traumatic injury there is a clear relationship between the structural tissue damage and resultant disability after recovery. There are no publications that compare significant osseous injury to soft tissue injury. Aim: To compare spinal outcome measures between patients reviewed for medico-legal compensation claims relating to perceived injury in the work place to those having sustained serious structural injury in the form of unstable thoraco-lumbar fractures requiring surgical stabilisation. Method: Two consecutive cohorts consisting of 23 patients with healed spinal fractures and 21 patients with a perception of work related soft tissue injury were compared. Patient demographics and a range of outcome measures including the Oswestry Disability Index (ODI), Low Back Outcome score (LBOS), Modified Somatic Perception (MSP) and Modified Zung Depression (MZD) indices were measured. Results: 23 patients (8 females; 15 males) with spinal fractures (group 1) of average age 42 years (range 22–66) were followed up for a mean of 41 months (range 14–89, SD 23.3) post trauma and compared to 21 patients (6 females; 15 males) with self reported back pain pursuing compensation claims, (group 2) of average age 47 years (range 37–63) of mean time since perceived injury of 42 months (range 12–62, SD 14.5). Both groups were comparable in terms of age and sex with P values of 0.254 and 0.752 respectively.

Eur Spine J (2011) 20 (Suppl 4):S421–S464 The average ODI was 28% (SD 18.5) compared to 52% (SD 17.1) (P: 0.0003); LBOS 39.7 versus 20.3 (P: 0.0003); MSP 4.3 versus 9.65 (P: 0.03); and MZD 20.2 versus 35.9 (P: 0.001) in groups 1 and 2 respectively. Conclusion: Despite high energy trauma and significant structural damage to the spine, post-traumatic patients had better spinal outcome scores in all measures (ODI, LBO, MSP, MZD). This thereby defies the majority of the Bradford Hill criteria of medical causality. The reasons for such differences are primarily psychosocial. Addressing obstacles to recovery may improve outcomes. There is no ‘dose–response’ curve to functional outcomes. In fact, uniquely the disability seems greater in the lower energy injury which is unique in trauma care.

52 COMPARISON OF CLINICAL AND RADIOLOGIC RESULTS OF OPEN AND PERCUTANEOUS SHORT SEGMENT PEDICLE SCREW FIXATION IN THE TREATMENT OF THORACOLUMBAR BURST FRACTURE J Lee, J Jang, B Seo, S Kim Neurosurgery, Chonnam National University Hospital, Gwangju, Korea (ROK/South Korea) Objective: The optimal management of unstable thoracolumbar burst fractures continues to be a matter of controversy. In many studies, the authors advocated that surgical interventions were superior to nonsurgical conservative treatment. Among many surgical methods for thoracolumbar burst fractures, short-segment pedicle screw fixation was widely adopted. Recently, several percutaneous pedicle screw fixation system has been introduced. The purpose of this study is to compare clinical and radiologic outcomes of open and percutaneous short segment pedicle screw fixation in the treatment of thoracolumbar burst fracture. Methods: This study involved fifty patients with thoracolumbar burst fracture underwent either open (n = 28) or percutaneous short-segment pedicle screw fixation (n = 22). Inclusion criteria required the following: single-level closed burst fracture involving T11-L2, no fracture dislocations or pedicle fractures, no osteoporosis, and no other major organ system or musculoskeletal injuries. Radiographs were obtained before surgery, immediately after surgery, and at final follow-up for accessing the restoration of spinal column. For patient’s pain and functional assessment, visual analogue scale (VAS) and Greenough Low Back Outcome Score (LBOS) were measured. Operation time, intraoperative bleeding loss, and the day of hospital stay were also evaluated. Results: The average follow up period was 27 months and 11.8 months in open and percutaneous surgery group respectively. There were no statistically differences between open and percutaneous surgery group concerning age, sex, fracture site, cause of injury, and preoperative regional kyphosis. In both groups, regional kyphosis after surgery and final follow-up was statistically improved (P \ 0.05). Postoperative correction loss of kyphosis showed that there was no significant difference between the both groups at final follow-up (P [ 0.05). Significant differences were observed between the two groups in intraoperative blood loss, operation time, and the day of hospital stay (P \ 0.05). In clinical results, the percutaneous surgery group had less VAS score and a better LBOS up to 3 months after surgery (P \ 0.05), but the outcomes were similar in final follow-up. Conclusions: Percutaneous short segment pedicle fixation in the treatment of thoracolumbar burst fracture provided earlier pain relief and functional improvement through less surgical related trauma

S447 compared with open surgery group. Therefore, it is expected that the patients underwent percutaneous surgery may get back to his social activities quickly, although the both surgery groups provided favorable clinical and radiologic outcomes at final follow-up.

53 FAILURE OF PERCUTANEOUS VERTEBROPLASTY TO RECTIFY VERTEBRAL COMPRESSION FRACTURES WITH AVASCULAR OSTEONECROSIS DH Heo Neurosurgery, Chuncheon Sacred Heart Hospital, College of Medicine, Hallym University, Kangwon-do, Korea (ROK/South Korea) Objectives: To assess for at least 2 years the radiologic and clinical outcomes of patients who underwent polymethylmethacrylate (PMMA) vertebroplasty to treat osteoporotic vertebral compression fractures with avascular necrosis. Summary of background data: Recently, osteoporotic vertebral compression fractures with avascular osteonecrosis have been treated with percutaneous vertebroplasty. However, there have been no previous multi-year, clinical and radiological studies of the results of vertebroplasty in the vertebral body with non-infected avascular osteonecrosis. Methods: Thirty patients were followed for at least 2 years after vertebroplasty. We retrospectively reviewed several parameters, including visual analog scale (VAS) score, compression ratio, kyphotic angle, injection pattern of PMMA (interdigitation and solid mass), and morphological changes of the PMMA cemented vertebral bodies. Results: The vertebral height and kyphotic angle were significantly corrected after vertebroplasty. However, the restored vertebral height re-collapsed (P \ 0.05), and the kyphotic angle became aggravated (P \ 0.05) during the 2 years or longer of postoperative follow-up. There were four kinds of morphological changes of the injected PMMA cemented vertebral body, including heterotopic ossification, fusion with the adjacent vertebral body, bone cement fragmentation and migration, and development of a radiolucent line around the PMMA mass in the vertebral body. Conclusion: After vertebroplasty, the compression and kyphosis of avascular necrotic vertebral bodies progressed continuously for 2 years or longer. Vertebroplasty may not provide sufficient stability. Therefore, we strongly recommended that strict observation and follow-up be used after vertebroplasty.

Fig. 1 6 months after the vertebroplasty, recollapse has occurred, and a radiolucent line has developed around the solid PMMA mass (arrow). 25 months after the vertebroplasty, a lateral plain X-ray shows that the augmented vertebrae has experienced further recollapse

123

S448

Eur Spine J (2011) 20 (Suppl 4):S421–S464 fractures. The typical shortcomings of conventional instrumentation (implant loosening; cut-out of screws) can be avoided. With the fixator providing sufficient axial stability for the posterior spinal wall, this technique allows for far anterior placement of the cement during kyphoplasty, thus adding to its safetyness. It is performed percutaneously, additionally fitting elderly patients needs. However, verification of disc-integrity is necessary, as this technique does not address the disc space.

Fig. 2 Radiologic studies of a 71-year-old man with an L1 compression fracture. 2 months after the vertebroplasty, the patient presented with weakness in both legs. An MRI shows that the recollapsed cemented L1 vertebral body has compressed the thecal sac

54 IN SITU CEMENT AUGMENTATION OF PERCUTANEOUS PEDICLE SCREWS—A NOVEL SHORT SEGMENT INSTRUMENTATION SYSTEM FOR THE OSTEOPOROTIC SPINE TR Blattert, C Josten

Screw and percutaneous cement injection

Spine Surgery and Traumatology, Orthopaedische Fachklinik Schwarzach, Schwarzach, Germany Introduction: Stabilization of osteoporotic burst fractures comprises a major challenge. Balloon-kyphoplasty as a single procedure does not address the posterior wall fragment and thus cannot restore axial stability. Classic-type posterior instrumentation tends to fail due to implant loosening. We therefore prefer combined vertebral stabilization by means of cement-augmented short segment posterior instrumentation and single-level kyphoplasty. We introduce a novel percutaneous system with in situ cement augmentation of pedicle screws. Methods: Inclusion criteria for this prospective trial: A3-fractures of Th11-L5; integrity of adjacent discs (MRI); t-score -2.5 (DXA). Initial reduction and cement-augmentation was performed by percutaneous Balloon-kyphoplasty (PMMA). Final reduction was achieved by percutaneous instrumentation of the adjacent vertebrae with in situ PMMA screw-augmentation (FNS; Medtronic). Following data were acquired: subjective pain rating (Visual Analogue Scale-VAS); bisegmental endplate-angle (plain X-rays). Patients were subject to full weight-bearing on day 1. Follow-up was performed on day 1; week 6; and months 3, 6, and 12. Results: 52 patients with 208 augmented pedicle screws were included. Average patient age was 74 (60 to 92). Average t-score was -2.7. (-3.1 to -2.5). In 41/208 pedicle screws, leakage of cement was noted. Direction of leakage was anterior or lateral for 40, and epidural for 1 case. All 52 patients experienced marked pain-relief as expressed on the VAS (p \ 0.005). Average correction of bisegmental endplate-angle was 8.7 (p \ 0.005). 47 patients concluded FU. No significant loss of correction was noted with the exception of 2 patients in which there was cut-out of the cement-augmented cranial pairs of screws. Except those 2 cases, there was no case of implant loosening or cut-out of pedicle screws. Conclusions: Combined cement-augmented instrumentation and kyphoplasty is efficient for stabilization of osteoporotic burst

123

In situ cement augmentation of percutaneous screws-intraoperative fluoroscopy

55 NEUROLOGICAL LATERALITY IN CERVICAL COMPRESSION MYELOPATHY H Mihara, S Kondo, S Kato, K Ishida, M Ono, M Hachiya Orthopaedics, Yokohama Minami Kyosai Hospital, Yokohama, Japan Introduction: According to Crandall’s classification, transverse lesion syndrome manifests neurological deterioration almost equally

Eur Spine J (2011) 20 (Suppl 4):S421–S464 on both sides. Although most patients with compression cervical myelopathy are classified as this type, they exhibit substantial laterality in their neurological symptoms. As very few studies have examined such laterality of neurological deterioration, we analyzed the laterality of motor and sensory disturbance in cervical myelopathy patients and investigated its improvement following surgical treatments. Methods: Subjects were 141 cervical myelopathy patients (cervical spondylosis in 93 patients and ossification of the posterior longitudinal ligament in 48) who underwent surgical treatments in our hospital and were followed more than 1 year. The neurological status of each subject was evaluated by Japan Orthopaedic Association (JOA) score and also by performance tests including finger grip and release test (GRT) and triangle step test (TST). Simultaneously sensory scores were recorded right-left separately by the pinprick test. Laterality of motor dysfunction was defined positive if a patient showed 4 times or more laterality on the performance tests (GRT and TST). Concerning the assessment of surgical results, recovery rates of every scores were calculated. Results: According to the performance tests, 4 times or more laterality was recorded in 36 patients (26%) on the GRT and 30 patients (21%) on the TST. Then 15 patients (11%) demonstrated obvious laterality on both tests (asymmetric group). The sensory scores did not correlate with the laterality of motor dysfunction in total. In the asymmetric group, surgical treatments improved the JOA score by 50.2% as well as the GRT score by 50.0%, the TST score by 67.9% and the sensory score by 55.7% on average. Among 90 patients who did not show laterality on the performance tests, 24 patients with significantly lower scores (GRT, TST 15) were classified as symmetric severe group and 12 patients with mild deterioration (GRT, TST [ 20) were classified as symmetric mild group. As to postoperative results, the asymmetric group showed better recovery than the symmetric severe group but poorer improvement than the symmetric mild group (Figure 1). Conclusions: According to the results in this study, patients showing asymmetric motor dysfunction should consider surgical decompression before the symptoms advance to symmetric severe conditions.

S449 56 COMPARISON OF THE MOBI CÒ CERVICAL ARTIFICIAL DISC TO ANTERIOR CERVICAL DISCECTOMY AND FUSION IN THE TREATMENT OF SYMPTOMATIC CERVICAL DEGENERATIVE DISC DISEASE AT 2 LEVELS RJ Davis, HW Bae, S Gaede, MS Hisey, G Hoffman, K Kim, PD Nunley, D Peterson, R Raushbaum, J Stokes Neurosurgery, Greater Baltimore Medical Center, Towson, USA Background: Anterior discectomy and fusion (ACDF) is a widely accepted treatment for radiculopathy and myeloradiculopathy associated with degenerative disc disease (DDD) of the cervical spine at both one and multiple vertebral levels. Recently, several FDA Investigational Device Exemption trials (IDE) have evaluated the success of total disc replacement (TDR) compared to ACDF at a single vertebral level. Study success in these studies was based on many factors including absence of radiographic failure. However, the relevance of radiographic failure to overall clinical outcome of a patient has been questioned, particularly for ACDF treatment groups. Purpose: Evaluate the contribution of radiographic failures towards overall study success rates of an FDA IDE trial that compared TDR to ACDF at 2 consecutive levels of the cervical spine. Outcome measures: The primary outcome measure was overall success rate at 24 months. Study success requires: improvement in NDI score, absence of subsequent surgeries, and absence of major complications due to neurologic deterioration, radiographic failure, and adverse events. Radiographic failure was defined as more than 2 segmental motion and/or radiolucent lines for ACDF and less than 2 of segmental motion and/or presence of heterotopic ossification, development of bridging bone, or osteophytes for TDR. Methods: 330 patients with two-level symptomatic cervical DDD were randomized to either the TDR group (Mobi-C) or ACDF group (allograft with anterior plate) in a 2:1 ratio (225:105). Patients were evaluated pre-operatively and post-operatively at 6 weeks, 3, 6, 12, 18, and 24 months. Results: At 24 months, TDR group success rate was 70.59% compared to 37.37% for ACDF group which establishes statistical superiority for TDR over ACDF (p \ 0.0001). At this primary study endpoint 9.2% of overall study failures in the TDR group and 19.4% in the ACDF group were due to radiographic failure alone. Overall study success rates without these study failures were 73.30% for TDR and 49.49% for ACDF which also establishes statistical superiority for TDR over ACDF (p \ 0.0001). Furthermore, this statistical superiority is shown at each time point under both conditions for study success (p \ 0.01). Conclusions: The results of this analysis validate the robustness of the superiority of TDR with Mobi-C over ACDF at 2 levels and continue to support the conclusion that TDR may be a superior treatment to ACDF at 2 levels.

Fig. 1 Neurological laterality and postoperative recovery

123

S450 57 DOSE- AND TIME-DEPENDENT EFFECT OF HIGH GLUCOSE CONCENTRATION ON PROLIFERATION AND APOPTOSIS OF NOTOCHORDAL CELLS JB Park, YY Kim, CG Kong Orthopaedic Surgery, The Catholic University of Korea, Uijeongbu, Korea (ROK/South Korea) Introduction: Notochordal cells play an important role in the formation and maintenance of the nucleus pulposus (NP) during embryonic development. After birth, notochordal cells gradually disappear via Fas-mediated apoptosis from NP. Disappearance of notochordal cells from NP is thought to be the starting point of disc degeneration. Diabetes mellitus (DM) is thought to be associated with premature, excessive apoptosis of NP notochordal cells, which leads to early disc degeneration. It remains to be determined if these premature changes are due to hyperglycemia or some other factors associated with DM. The purpose of current study was to investigate the effect of high glucose on apoptotic pathway, viability, and expression of matrix metalloproteinases (MMPs) and their tissue inhibitors of metalloproteinase (TIMPs) of rat notochordal cells. Methods: Rat notochordal cells were isolated, cultured, and placed in either 10% (normal control) fetal bovine serum (FBS) or 10% FBS plus 3 different glucose concentrations (100 nM, 200 nM, and 400 nM) for 1, 3, 5, and 7 days, respectively. To determine the high glucose-induced apoptotic pathways, activation of caspases (-3, -8, and -9) as well as cleavages of Bid and cytochrome-c were evaluated with Western blotting. To investigate the effect of high glucose on viability of notochordal cells, degree of proliferation and apoptosis of notochordal cells was quantified. Finally, the effect of high glucose on expression of MMPs and TIMPs of notochordal cells was evaluated with Western blotting. Results: High glucose decreased proliferation of rat notochordal cells from culture day one to seven with dose-dependent manner. High glucose increased apoptosis of rat notochordal cells from culture day one to seven with dose-dependent manner and led to activations of caspases-9 and caspases-3. It also caused cleavages of Bid and cytochrome-c. Finally, high glucose significantly increased expression of MMPs (-1, -2, -3, -7, -9, and -13) and TIMPs (-1 and -2) of rat notochordal cells, respectively. Conclusions: The current findings demonstrate that high glucose decreases proliferation and increases apoptosis of rat notochordal cells via Fas-mediated type 2 pathway with dose- and time-dependent manner, which enhances expression of MMPs and TIMPs. Therefore, aggressive glucose control from early stage of DM is recommended to prevent or slow intervertebral disc degeneration in diabetic patients.

58 MRI STUDY OF POSTERIOR LIGAMENTOUS COMPLEX (PLC) INSTABILITY AFTER TRAUMATIC FRACTURES. THE IMPORTANCE OF THE SUPRASPINOUS LIGAMENT. J Pizones, E Izquierdo, L Zu´n˜iga, F Sa´nchez-Mariscal, ´ lvarez-Gonza´lez, A Go´mez-Rice PA Orthopaedics, Spine Unit, Hospital Universitario de Getafe, Madrid, Spain Introduction: We still do not know how many of the posterior structures have to be damaged to consider the PLC as disrupted, and how to recognize the posterior complex rupture with imaging techniques.

123

Eur Spine J (2011) 20 (Suppl 4):S421–S464 Materials and methods: Prospective study of 74 consecutive vertebral acute traumatic fractures, analyzed using X-rays and MRI scan (FS-T2-w/STIR sequences). We analyzed the association between MRI signal (intact, edema, disruption) of each PLC component-facets, interspinous ligaments (IS), supraespinous ligaments (SS) and ligamentum flavum (LF)- and the variables: AO classification, TLICS classification (fracture morphology, neurologic status, PLC consideration), treatment, surgical findings, interspinous diastasis ratio (SDR), local kyphosis (LK) and ISS (TLICS) score. Chi2 test was used for categorical variables and U Mann–Whitney for continuous ones. Results: MR images of IS, SS or LF complete disruptions showed in all cases ruptures during surgical examination. Images of IS edema correlated surgically with intact ligaments or laxity. MR images of IS edema associated with 87.5% of facet distraction, 20% of SS ruptures and 0% of LF ruptures, LK:11.6, SDR:1.2. MR images of IS complete disruption were always associated with SS or LF rupture, and LK:16, SDR:2. MR images of SS complete disruption associated LK:14.5 and SDR:1.8; while LF complete disruption showed LK:18 and SDR:1.9. Therefore, we compared two groups: we called the first group ‘‘Stable PLC’’ (images of facet distraction and IS edema) and the second, ‘‘Unstable PLC’’ (images of IS, SS, or LF disruption). ‘‘Unstable PLC’’ showed severer scores in every variable: worst morphologic damage in the AO classification, higher punctuations in TLICS, surgical treatment was more predominant, higher SDR, greater LK, and worst ISS. All showed p \ 0.001, except neurological damage (Table 1). Conclusion: IS edema seen by MRI scan seems not to be enough to create a complete instability, it could be the rupture of the SS or LF what gives the key to PLC instability. Two different patterns of PLC could be define (stable/unstable), which would allow a better understanding of the category ‘‘PLC consideration’’ of the TLICS classification clearing the term ‘‘indeterminate’’. Table 1 Comparison between stable and unstable PLC considering MRI findings (FS T2-w/STIR sequences). Chi2 and U-Mann– Whitney

Number Localization

AO classification type

Stable (non disrupted)

Unstable (disrupted)

58 45% T 48% TL 7% L 55% A1; 9% A2; 22% A3 12% B1; 2% B2

16 12.5% T 75% TL 12.5% L 0% A

0% C Fracture morphology Compression Burst Translation Flexiondistraction Neurologic status Intact Damaged

P

0.06

0.000

44% B1, 31% B2 6% C1; 19% C2 0.000

65% 28% 0% 7%

0% 12% 19% 69%

91.4% 8.6%

75% 25%

0.07

Eur Spine J (2011) 20 (Suppl 4):S421–S464

S451

Table 1 continued Stable Unstable P (non disrupted) (disrupted) PLC consideration Intact Indeterminate Disrupted Treatment Conservative Surgical Surgical findings IS Intact Laxity Rupture SS Intact Rupture LF Intact Rupture Spinous diastasis ratio Local kyphosis ISS (TLICS score)

0.000 72% 26% 2%

0% 6% 94%

60% 40%

0% 100%

62% 38% 0%

0% 19% 81%

100% 0%

12.5% 87.5%

0.000

100% 0% 1.1 8.6 ± 5.2 2.1 ± 1.2

31% 69% 1.8 15.9 ± 6.9 7.2 ± 1.5

0.000

0.000

newly-developed DS. LDK was diagnosed by the angle of L1S1 \ 15 among all subjects. Results: Subjects’ baseline age was 57.5 and mean follow-up period was 12.0 years. DS was diagnosed in 12.7%, and PI was significantly bigger in DS patients (DS 62.4 vs. normal 54.7, p = 0.0056). LDK was diagnosed in 13.7%, and PI was significantly smaller in LDK patients (LDK 48.2 vs. normal 56.3, p = 0.0021; Student t-test). Conclusion: This is the first to study the relationship between PI and spinal pathologies in a mean 12-year follow-up. PI determines the capacity of pelvic retroversion, and smaller PI led to the development of kyphotic deformity, while bigger PI led to the development of spondylolisthesis.

0.000

60 FUSION DOES NOT IMPROVE OUTCOME IN DECOMPRESSIVE SURGERY FOR LUMBAR SPINAL STENOSIS. A STUDY OF 5390 PATIENTS FROM SWESPINE P Forsth, B Sande´n, K Michaelsson Department of Orthopedics, Spine Surgery Unit, Uppsala, Sweden

0.001 0.000 0.000

Non disrupted: facet distraction + IS edema; Disrupted: IS, SS or LA disruption. T: thoracic; TL: thoracolumbar; L: lumbar; IS: interspinous ligament; SS: supraespinous ligament; LF: ligamentum flavum

59 SAGITTAL SPINOPELVIC ALIGNMENT AND ITS RELATIONSHIP WITH DEGENERATIVE DEFORMITY IN A PROSPECTIVE COHORT

Introduction: The role of fusion in surgery for lumbar spinal stenosis is a major controverse in spine surgery. The aim of this study was to examine if adding fusion improves results of decompression for lumbar stenosis with or without degenerative olisthesis. Methods: Swespine, the Swedish Spine Register, was used for the study. Data were obtained for surgical procedures for 1- or 2-level central lumbar stenosis L2–L5 for patients older than 50 years at the time of surgery. The register contains data on the presence of any preoperative olisthesis [3 mm in the operated segments. 5390 patients fulfilled the inclusion criteria and had completed the 2-year follow-up. The results of decompression with and without concomitant fusion were assessed in a multivariate analysis and adjusted for gender, age, smoking habits and differences at baseline for the studied outcome measures (Back pain, Leg pain, EuroQol, ODI and Global Assessment). Results: At 2 years follow-up, adjusted means with 95% CI.

T Kobayashi, K Aono, S Jimbo, I Senoo, Y Atsuta, T Matsuno Orthopaedic Surgery, Asahikawa Medical University, Asahikawa, Japan Introduction: There is a growing concern about the relationship between sagittal spinopelvic alignment and spinal pathologies. Pelvic incidence (PI), a morphological parameter not affected by the posture or the pelvis position, has been suggested as a unique predictor of spinal degenerative deformities such as spondylolisthesis (DS) and kyphosis (LDK). The purpose of this study was to investigate the influence of PI on the development of DS and LDK in a prospective cohort. Methods: A final total of 227 community-based female volunteers, aged 40 + years at baseline and without severe spinal history, were followed for more than 8 years. Standardized upright entire spine radiographs were taken at each visit and used for measuring sagittal spinal alignment including PI. Magnitude of spondylolisthesis was also evaluated in a subgroup of 142 women without DS at baseline radiograph, and development of more than 5% slip was diagnosed as

Decompression + fusion

Decompression only

Preop. olisthesis n = 689

n = 617

Back pain (VAS)

32.2 (29.8–34.5)

34.8 (32.5–37.2) p = 0.12

Leg pain

32.2 (29.7–34.8)

34.8 (32.5–37.4) p = 0.17

EuroQoL

0.62 (0.59–0.64)

0.63 (0.61–0.66) p = 0.34

No olisthesis

n = 488

n = 3596

Back pain

(VAS) 34.2 (31.5–36.9)

35.8 (34.7–36.8) p = 0.30

Leg pain

36.7 (33.8–39.7)

36.1 (35.0–37.2) p = 0.69

EuroQoL

0.60 (0.57–0.64)

0.61 (0.60–0.63) p = 0.56

123

S452

Eur Spine J (2011) 20 (Suppl 4):S421–S464

No differences were found regarding ODI and global assessment. In summery no significant differences in outcome were found between patients operated with decompression only and those who had a concomitant fusion. The frequency of patients in the whole register who had a second operation for lumbar spinal stenosis was 6.0% in the decompression group and 7.7% among those with a fusion. For those with preop olisthesis these numbers were 7.2% after decompression and 5.7% with a fusion. Conclusions: To our knowledge this is the largest cohort used for studying this issue. The results suggest that adding fusion to decompressive surgery does not favor the outcome in lumbar spinal stenosis, regardless of the presence of any preoperative degenerative olisthesis.

61 A TISSUE-ENGINEERED TOTAL DISC REPLACEMENT. 8 MONTH IN OUTCOMES OF AN IN VIVO STUDY

TE-TDR Disc assessment modalities

AR James, R Ha¨rtl, RD Bowles, H Gebhard, LD Bonassar Neurological Surgery, Weill Cornell Brain and Spine Center, New York, USA Introduction: Degenerative disc disease is a commonly associated with pain, disability and cost. A tissue-engineered total disc replacement (TE-TDR) has the potential to restore motion and mechanical damping, integrate with the native tissue and produce a disc like extracellular matrix. Final in-vivo results are presented. Methods: The rat caudal spine is easily approached, surgically reproducible and the disc experiences loading and high strain. Athymic rats allow implantation of xenograft cells. Anatomically sized cell seeded TE-TDR were created with an annulus fibrosus with circumferentially aligned collagen fibrils contracted around an alginate nucleus pulposus. Ethical approval was obtained. Control groups were discectomy alone (n = 6) and discectomy with immediate reimplantation of the excised disc (n = 6). The intervention group (n = 24) was studied up to 8 months. Caudal and cranial discs were internal controls. Sequential high-resolution MRI (7T) allowed non-invasive assessment, whilst quantitative MRI (T1rho) of collagen and proteoglycan. MicroCT and histology with staining for proteoglycans and collagen was undertaken. Biochemical and mechanical testing was performed. Results: Disc space height (DSH) of TE-TDR implants was maintained at 70% of normal controls, with no significant deterioration to 8 months. DSH was not significantly different from reimplanted discs, whilst discectomy led to a rapid collapse in DSH. MRI T1rho assessment of the nucleus indicated the presence of proteoglycan with reduced water content. The annular component was thickened, the signal consistent with the presence of high collagen concentrations. Histological analysis showed a sustained production of collagen and proteoglycan with homogeneity, and tissue integration. MicroCT demonstrated no bony bridging. Biomechanical testing was similar with a dynamic modulus only 30% stiffer than normal, at 5 months. Conclusions: The TE-TDR maintains intervertebral DSH, biomechanical function and morphology, and tissue integration at 8 months, which is the life expectancy of these animals.

123

Graph of Disc Space Height as a Percentage of the Control

62 DEGENERATIVE FACET JOINT ALTERATIONS SECONDARY TO PERCUTANEOUS PEDICLE SCREW FIXATION: RADIOGRAPHIC ANALYSIS OF 30 PATIENTS L Proietti, L Scaramuzzo, GR Schiro`, S Sessa, CA Logroscino Orthopaedic and Traumatology, Catholic University, Rome, Italy Introduction: Aim of the study was to evaluate the appearance of degenerative lumbar facet joints alterations secondary to percutaneous pedicle screw fixation for the treatment of lumbar fractures without neurological involvement. Materials and methods: 30 patients affected by fractures of the lumbar spine underwent percutaneous pedicle screw fixation. In all patients a short fixation (4 screws) without fusion was performed.

Eur Spine J (2011) 20 (Suppl 4):S421–S464 Radiographic analysis was performed by a thin cut CT scan in the preoperative and post-operative time at 4 months in 10 patients, 8 months in 7 patients and 12 months in 13 patients. Facet joints degenerative alterations were classified according to Pathria criteria. In every patient the 4 facet joints neighbouring the screws and these at fracture’s level were evaluated. Screw placement respect to facet joint surface was also evaluated. Results: Eleven patients presented an A3.1, 4 an A3.2, 15 an A.3 fracture. Patients were divided into 3 groups considering CT scan execution time: group A 4 months, group B 8 months, group C 12 months. Patients in the group A showed no statistically significant differences between pre and post-operative facet joints (p [ 0.05). Patients in the group B showed a statistically significant worsening in the 28% of the facet joints at fracture’s level (p = 0.03). Patients in the group C showed a plain degeneration in the 100% of the cases (p = 0.001). Considering screws placement there was an impingement with the articular surface in 6 cases, responsible of an early degeneration of the related facet joint. Seven patients were older than 65; 3 of them showed a complete degeneration of the facet joints at fracture’s level comparable to a fusion yet to 8 months. Discussion: Results’ analysis showed the presence of significant facet joints degeneration yet to 8 months after a percutaneous screw fixation. Facet joints fusion was observed only in patients older than 65 with a high pre-operative degeneration grade. Conclusion: In the authors’ opinion lumbar fractures treatment with percutaneous screw fixation without fusion causes an early facet joints degeneration and has so not to be considered an harmless procedure. This study encourages to remove percutaneous pedicle screw instrumentation no later than 1 year to avoid severe facet joints degenerative changes.

63 WHICH FACTORS INFLUENCE SPONTANEOUS LUMBAR CURVE CORRECTION AFTER SELECTIVE THORACIC FUSION IN LENKE TYPE IB AND C CURVES? A Hamzaoglu, C Ozturk, A Alanay, M Enercan, E Karadeniz, S Karaca Orthopedic Surgery, Istanbul Spine Center, Istanbul, Turkey Introduction: The aim of this retrospective study was to analyse Lenke type 1B and C curves treated by selective fusion by using PS and to determine the critical factors which may have influenced spontaneous correction in lumbar spine. Materials and methods: Between 1999 and 2009, 111 (100 female and 11 male) consecutive patients with Lenke type 1 B and C curves (44 B and 67 C modifier) who underwent selective thoracic fusion by using PS constructs were included. All surgeries were done by a single surgeon. The rule for a safe amount of correction in main thoracic curve was not to correct more than the lumbar curve magnitude in the preoperative side bending x-ray. All patients had intraoperative under-table 35 inch x-ray to apply the rule. Preop, postop and follow-up x-ray images were evaluated in terms of curve magnitudes and flexibility, sacral tilt (more than 2), lower instrumented vertebra (LIV), apical vertebra translation of lumbar curve and maturity. Results: Average age at the time of surgery was 15 years (range 11–19) and follow-up was 64 (range 24–148) months. The fusion stopped at stable vertebra in 71 patients, at neutral vertebra in 29 patients and stable and neutral vertebra which were the same in 11 patients. Average correction rates at final follow-up for thoracic curve was 81% while it was 68% for lumbar curve. Spontaenous lumbar curve correction ratio was 75 versus 54% for lumbar curves with

S453 more or less than 60% flexibility; 76 versus 66% whether LIV being both stable and neutral or not; 72 versus 66% for curves with or without sacral tilt; 71 versus 66% for patients younger or older than 14 years of age; 70 versus 66% for risser grade less or more than 2, 68 versus 53% for lumber curves lesser or greater than 45 degrees and 71 versus 63% for apical vertebra translation of lumbar curve less or more than 2 cm. There was no coronal plane decompensation and imbalance in any of the patients and no reoperation. Conclusion: Patients younger than age 14, with LIV being both stable and neutral, with lumbar curve flexibility more than 60%, with no sacral tilt, risser sign 2, preoperative lumbar curve magnitude of less than 45 degrees and apical vertebra translation of lumbar curve less than 2 cm. had more spontaneous lumbar curve correction rates after selective thoracic fusion (p \ 0.05). 64 THE NATURAL HISTORY OF SCHEUERMANN’S KYPHOSIS—A COMPARATIVE STUDY AFTER 37-YEAR FOLLOW-UP L Ristolainen, JA Kettunen, M Helio¨vaara, UM Kujala, A Heinonen, D Schlenzka Research, ORTON Orthopaedic Hospital and Research Institute, Helsinki, Finland Present knowledge of the natural course of Scheuermann’s kyphosis is mainly based on a study by Murray et al (1993).They stated that ‘‘although patients who have Scheuermann’s kyphosis may indeed have some functional limitations, they do not have major interference with their lives’’. Their control group consisted of individuals who were ‘‘friends or relatives who had accompanied patients who had come to the orthopaedic clinics’’. The purpose of this study was to investigate back pain and disability, their relationship to vertebral changes in patients with untreated Scheuermann’s, and to compare the results to a representative sample of the population. The clinical records, radiographs, and addresses of 136 patients who had attended the outpatient clinics between 1950 and 1990 for suspected Scheuermann’s were available.Eighty responded to a questionnaire, 49 of them (12 females, 37 males) fulfilled the radiographic criteria for ‘‘classic’’ Scheuermann’s.There was no difference in the baseline data between responders and non-responders. Th-kyphosis, l- lordosis, and scoliosis were measured from radiographs. The number of affected vertebrae and the degree of wedging were registered. Anthropometric data, occurrence of back pain, disability scores, and employment status were compared to a representative sample (n = 3835) of the normal population. After mean follow-up of 37 (SD 6.5; range 25.9–53.7) years, the average age of Scheuermann’s patients was 58.8 (8.2; 44.4–79.3) years. Male patients were significantly taller than the control subjects at age 20 and at follow-up. Female patients were on average 6 kg heavier (P = 0.02) and their mean BMI was higher (23.9 vs. 20.8 kg/m2, P = 0.001) at age 20 than in the controls. Females had a greater mean kyphosis than males (51.7 vs. 43.2, P = 0.11). There was no correlation between the degree of thoracic kyphosis and disability. Patients who had difficulties in walking upstairs had more affected vertebral bodies than patients who did not (6.9 vs.5.7, P = 0.04). Compared to the refrence group, Scheuermann’s patients had an increased risk for constant back pain (P = 0.003), a 2.6-fold risk for disability because of back pain during the past 5 years (P = 0.002), a 3.7-fold risk for back pain during the past 30 days (P \ 0.001), and a 2.3-fold risk for sciatic pain (P = 0.005). They reported a poorer quality of life (P \ 0.001) and general health (P \ 0.001).

123

S454 There was no difference in working ability and employment status between patients and controls.

65 RIB-VERTEBRAL ANGLE DIFFERENCE AND CONVEX RIBVERTEBRAL ANGLE AS RISK FACTORS FOR CURVE PROGRESSION DURING BRACING IN YOUNG CHILDREN WITH IDIOPATHIC SCOLIOSIS Q Ding, X Sun, X Zheng, M Ji, S Mao, F Lv, X Zhang Spine Surgery, Drum Tower Hospital, Nanjing University Medical School, Nanjing, China Introduction: RVA and RVAD has been considered as prognostic factors for the natural history of infantile or juvenile scoliosis patients. However, few studies investigated the prognostic value of RVA and RVAD in young children with bracing treatment. Methods: IS patients who were less than 12 years old and had Risser sign 0 at the beginning of bracing were selected for this study. All patients had a regular follow-up over 2 years or were followed-up till the patients were indicated for surgical correction due to curve progression. Totally 28 patients were recruited, including 1 boy and 27 girls. They were divided into two groups according to the outcome of bracing treatment at latest follow-up: Group A, patients with the curve improved or stable and Group B, patients with curve worsened or indicated for surgery. Results: Fifteen patients were included in Group A and 13 patients in Group B, respectively. Till the last follow-up, curve progression was found in 46% (13/28) patients. No difference of the initial curve was observed between Group A and Group B (P = 0.194). At last followup, the average primary curve Cobb’s angle was 25.4 ± 5.3 in Group A and 38.4 ± 6.7 in Group B, with a significant difference between the two groups (P \ 0.001). In addition, significant difference was observed in the initial RVAD (9.8 ± 5.3 vs 19.0 ± 10.9, P = 0.013), initial convex RVA (74.8 ± 7.8 vs 67.7 ± 5.9, P = 0.013), final RVAD (9.3 ± 6.1 vs 26.2 ± 12.3, P = 0.001) and final convex RVA (72.2 ± 10.5 vs 61.2 ± 10.1, P = 0.009) between Group A and Group B. Association analysis showing that RVAD is associated with curve progression (P = 0.029). Although no association of curve progression with convex RVA (P = 0.228) was observed, the progression rate of IS with convex RVA68(66.7%) was higher than that of IS with convex RVA [68(36.8%). Conclusion: Bracing treatment could effectively prevent curve progression in half of IS patients with Risser grade 0. RVAD20 and convex RVA68 are two risk factors in predicting curve progression during bracing in young children with idiopathic scoliosis. Significance: RVAD and convex RVA at brace initiation can help to predict the effectiveness of bracing in skeletally immature IS patients. 66 USE OF VERTICAL EXPANDABLE PROSTHETIC TITANIUM RIB (VEPTR) FOR TREATING CONGENITAL KYPHOSIS IN THORACIC MENINGOMYELOCELE PATIENTS GR Zuiani, PTM Cavali, MA Lehoczki, AJ Rossato, E Landim, MI Risso-Neto, IG Veiga, W Pasqualini Ortop e Trauma, Grupo de Escoliose, AACD, Associac¸a˜o de Assisteˆncia a` Crianc¸a Deficiente, Campinas, Brazil Objective: to evaluate clinical and radiographic postoperative results of congenital kyphosis correction in thoracic meningomyelocele patients using Vertical Expandable Prosthetic Titanium Rib (VEPTR).

123

Eur Spine J (2011) 20 (Suppl 4):S421–S464 Methods: A retrospective study of 19 thoracic meningomyelocele and congenital kyphosis patients that were subjected to the VEPTR treatment between October 2005 and October 2008, with radiographic evaluation and immediate post and pre-operative clinical practice. Also, the duration of surgical procedure, the need for blood transfusion and postoperative complications were assessed. Results: The patients’ average age was 70 months (from 32 to 130 months). The average follow-up from patients was 13.5 months (from 2 to 26 months). The average duration of the procedure was 117 min (variation between 70 and 195 min). All children reached trunk balance, 13 of whom had not showed it in the postoperative period. The average of pre-operation kyphosis was 115 (from 80 to 150) and 77 (from 50 to 104) for postoperative, with an average correction percentage of 31.2% (from 1.1 to 61.5%). The previous pre-operative imbalance of trunk was an average 7.9 cm (from 1.0 to 15.5 cm) and 3.4 cm (from 0 to 8 cm) for post-operative. The average correction of this imbalance was of 50.4% (from 0 to 100%). Regarding weight, in pre-operative the average was 15.4 kg (from 8 to 30 kg) and 20.6 kg (from 8.5 to 35 kg) for postoperative. The average gain of weight was of 36.6% (from 9.8 to 100%). Five of the 19 patients (26.3%) presented postoperative complications. No patient needed blood transfusion. Conclusion: The use of VEPTR in thoracic meningomyelocele and congenital kyphosis patients has proven to be an effective and promising alternative for the control of physical deformity in patients with a potential for growth.

67 VEPTR 7 YEARS FOLLOW UP IN THE TREATMENT OF SEVERE SPINAL DEFORMITIES. C Wimmer, T Pfandlsteiner, P walno¨ffer Spine center with center of deformities, Scho¨n Klinik Vogtareuth, Vogtareuth, Germany Indroduction: Since 2002 the treatment with VEPTR is established in Europe. From 2005 to 2011 65 patients were treated with VEPTR. Diagnosis were congenital-, neuropathic-, infantile scoliosis. Material and method: There were 27 congenital,23 neuropathic, and 15 infantile scoliosis. The average age of the 43 female and 22 male patients were 7,5 (3–13 years). Correction of the primary curve and correction after surgery were measured according to. Complications were noted, a patient satisfaction score and growing of length was measured. 9 of the 65patients had had previous surgery. Results: The primary curve measured 65 (45–130) and improved to 32 (25–75) at index operation. No complications during surgery were noted. The time at operation was 95 min. (65 to 175). In 29 cases a rib to lumbar spine hybrid was used, and in 25 cases a combination from rib to rib and in 11 cases a rib to pelvis construct. The blood loos during surgery was in mean 125 ml (65 to 180). One patient showed a prolonged wound healing, another patient had had a pneumonia, which resolved with adequate therapy. In five cases the rib devices pull out after the lengthening procedure. In four cases a removal of the metal was necessary in cause of skin breakage. No infection or neurological deficit occurred in the 1209 surgery’s of VEPTR. In all cases a spinal monitoring with SEP and MEP was used. The average stay in the hospital was 18 days (14 to 31). 55 out of 65 patients had 1 to 7 lengthening procedures. The average correction of the lengthening procedures was 15,7 (19.8%). In 19 of the patients the second curve must be instrumented. All patients and parents were satisfied with this procedure and would to this procedure again.

Eur Spine J (2011) 20 (Suppl 4):S421–S464 Conclusion: The first results of the VEPTR instrumentation are encouraging. Remarkable is the low complication rate and the high patients satisfaction.

S455 69 CLINICAL EVALUATION OF THE PRE- AND POSTOPERATIVE ANTERIOR CHEST WALL CONTOUR IN THORACIC ADOLESCENT IDIOPATHIC SCOLIOSIS S Mao, F Zhu, B Qian, Z Zhu, Z Liu, B Wang, Y Qiu

68 DYNAMIC RIB REGENERATION AFTER THORACOPLASTY IN IDIOPATHIC SCOLIOSIS PATIENTS X Sun, C Xia, W Ma, F Zhu, B Qian, Y Yu, Z Zhu, B Wang, Y Qiu Spine Surgery, Drum Tower Hospital, Nanjing University Medical School, Nanjing, China Summary: This study evaluated the rib regeneration in adolescent idiopathic scoliosis (AIS) patients after thoracoplasty. Most of the excised ribs regenerated up to Grade 5 within 6 months postoperatively (rib formation phase), while almost all ribs regenerated beyond Grade 5 at 2-year follow-up (rib remodelling phase). Introduction: In AIS with thoracic curves, posterior correction is most commonly used, and additional thoracoplasty is sometimes performed for obvious rib hump deformity. However, few studies investigated the regeneration of the resected ribs. This study was to evaluate the rib regeneration after thoracoplasty in AIS patients. Methods: A retrospective review was performed on 66 AIS patients with Lenke type 1 who underwent posterior correction and thoracoplasty from 1999 to 2004. There were 43 girls and 23 boys, with a mean age of 16.0 years and a mean thoracic curve of 59.6. During the post-operative follow-ups, standing postero-anterior and lateral radiographs of spine were taken. Rib regeneration was analyzed according to the classification established by Philips. And rib formation phase referred to Grade 1 to 5, while rib remodelling phase referred to Grade 6 to 7. Results: Totally 248 ribs which maximally contributed to the rib hump deformity were excised during the surgical procedures. At 3 months postoperatively, regeneration of Grade 4 and Grade 5 were found in 31.6% (71 ribs) and 52.9% (119 ribs), respectively, of 225 ribs in 60 patients. At 6 months, rib regeneration was observed in 18.7% (20 ribs) at Grade 4, 71.0% (76 ribs) at Grade 5 and 8.4% (9 ribs) at Grade 6 respectively, of 107 ribs in 29 patients. At 1 year, rib regeneration was found in 90.1% (109 ribs) at Grades 5–6, of 121 ribs in 32 patients, but none of these ribs reached Grade 7. At 2-year follow-up, rib regeneration beyond Grade 5 was observed in 98.7% (78/79 ribs, in 22 patients). Additionally, 5.1% (4 ribs) reached Grade 7 and none was less than Grade 4. No significant difference in rib regeneration between genders was found. Conclusion: After thoracoplasty, the formation of the resected rib mainly completes within 6 months after thoracoplasty, while the remodeling procedure begins nearly 1 year postoperatively and takes a rather longer duration. Significance: This observation evidences the dynamic reconstruction of the integrity of thorax cage after thoracoplasty, and thus gives answer to the finding that pulmonary function decreases shortly after surgery but returns to preoperative baseline at a long-term follow-up in most AIS patients. Keywords: Rib, regeneration, thoracoplasty, idiopathic scoliosis, posterior correction

Spine Surgery, Drum Tower Hospital, Nanjing University Medical School, Nanjing, China Summary: Anterior chest wall deformity is one of the major cosmetic concerns in thoracic adolescent idiopathic scoliosis (AIS). This study demonstrated significant correlation between variations of the anterior chest wall contour and Cobb angle, but not apical vertebral rotation. Moreover, surgery did not reliablely result in improved anterior surface shape though it may excellently correct Cobb angle. Large preoperative chest wall angle (CWA) and T9 as apex were identified to be associated with lower risk of chest wall shape aggravation postoperatively. Introduction: Similar spinal deformities in AIS usually accompany different anterior chest wall appearance. The surgical correction of the deformed chest wall is quite elusive despite excellent Cobb angle reduction. We wonder if the anterior chest wall deformity is independent in the severity on the co-development of spinal curvature, trunk torsion and vertebral rotation. This study was to assess the anterior chest wall deformity and define its relationship with other deformed components in AIS. Risk factors leading to aggravated chest wall shape postoperatively were also analyzed. Methods: Pre and post operative 3-dimensional anterior chest wall reconstructions were performed in 110 AIS patients. Deformity parameters to be measured included CWA, Sternum-Rib Ratio (S-R Ratio), the angle of the sternum relative to the apical vertebral body, Cobb angle, rib hump (RH), apical vertebral rotation (AVR) and angle of trunk rotation (ATR). Correlation analysis and linear regression analysis were performed. Surgical improvements were revealed with paired-sample t test. Results: The Cobb angle demonstrated good correlation with CWA, RH,, S-R Ratio, AVR and ATR (r = 0.377, 0.604, -0.401, 0.514, 0.530 and 0.517, respectively, p \ 0.001). The rib hump demonstrated good correlation with AVR (r = 0.546, p \ 0.001). No significant relationship between CWA and rib hump, AVR, ATR were detected (r = 0.129, 0.043, -0.039, respectively). Averaged CWA with different curve apex showed a normal distribution shape, with the highest value at T9 level. The Cobb angle, S-R Ratio, AVR and ATR improved significantly postoperatively (p \ 0.05), while the CWA andshowed no difference. Postoperative CWA aggravated in 52.8% of the patients, which showed a significantly lower average preoperative CWA (3.0 ± 2.8) compared with those with decreased CWA (7.9 ± 3.1). Conclusions: The variation of the anterior chest wall contour is partially attributable to Cobb angle, but not directly correlated with vertebral rotation, indicating that other intrinsic factors (Rib deformity, thoracic torsion, etc) may contribute to the anterior chest wall deformity. The deformed surface shape can not be reliably improved with posterior pedicle screw instrumentation. The patients with large preoperative CWA and T9 as apical vertebra showed lower risk of chest wall shape aggravation postoperatively. Significance: Post-op improvement of the deformed anterior chest wall contour is full of uncertainties, and AIS patients should be informed of the risk of chest wall shape aggravation postoperatively.

123

S456 70 A PROSPECTIVE STUDY OF ONE-STAGE SURGERY OF CONGENITAL SCOLIOSIS WITH SPLIT CORD MALFORMATION Q Guixing, L Jiaming Department of Orthopaedics, Peking Union Medical College Hospital, Beijing, China Background: Split Cord Malformation (SCM) is a rare congenital disorder. The nature history of SCM in CS has not been well documented. Many spine surgeons suggested to remove spur prior previously before scoliosis correction. However, the two-stage surgery of removing spur prior to scoliosis correction caused neurologic complication and made subsequent correction more difficult. Objective: The aim of this study was to investigate whether one-stage scoliosis correction without previously removing bone spicule or fiber band was safe and effective. Methods: A Prospective clinical study to evaluate the safety and effectiveness of one-stage correction of congenital scoliosis (CS) with split cord malformation (SCM) during February 2000 and November 2010 in one spinal center. Among 500 congenital scoliosis patients, 95 patients with SCM (66 girls and 29 boys), with type I SCM 35 patients and type II SCM 60 patients underwent one-stage scoliosis correction without previously removing intraspinal bone spicule or fiber band. The mean age was 13.5 years in type I SCM and 14.4 years in type II SCM. The mean Cobb angle was 74.1 in type I SCM and 65.2 in type II SCM. Clinical symptoms and signs, postoperative and follow-up Cobb angle and postoperative Complication were evaluated preoperatively and postoperatively at least 2 years. Results: 74 patients had been followed for minimal 2 years and the mean follow up was 5.3 years. There were 12 patients with lower limbs abnormal, 13 patients with reflex abnormal, 10 patients with diminished sensation and 6 patients with decreased muscle strength. All the patients underwent scoliosis correction surgery. None of bone spicule in SCM I or fiber band in SCM II were removed. Among these patients, 5 underwent anterior approach and 90 underwent posterior approach. The mean postoperative Cobb angle was 38.3 in type I SCM and 30.2in type II SCM. At last follow up, the mean Cobb angle was 41.2 in type I SCM and 32.9in type II SCM. No neurologic deficit occurred perioperartively. And neurologic symptoms and signs were stable on follow up. Conclusions: If neurological sign was stable, previously removing of bone spicule or fiber band prior to scoliosis correction may be not necessary.

71 TYPE 3 HEMIVERTEBRA RESECTION VIA POSTERIOR APPROACH IN YOUNG CHILDREN M Enercan, A Alanay, C Ozturk, S Karadereler, M Sarier, A Hamzaoglu Orthopedic Surgery, Istanbul Spine Center, Istanbul, Turkey Introduction: Hemivertebrectomy via posterior approach can be done by removing only osseous hemivertebrae (HV) with the eggshell technique (type 1), both osseous HV and endplates leaving the adjacent discs intact (type 2) and whole structures between two vertebral bodies adjacent to HV and endplates of adjacent vertebrae (type 3). Type 1 is less complex to perform via posterior approach however with a risk of regrowth of HV. Type 2 creates a gap filled with fibrous scar and stability is doubtful. Type 3 is a technically challenging

123

Eur Spine J (2011) 20 (Suppl 4):S421–S464 procedure with no risk of regrowth. The aim of this study is to analyse the results of type 3 hemivertebrectomy. Materials and methods: 33 patients having 39 posterior hemivertebrectomy and transpedicular fixation were reviewed. Radiological and clinical charts were retrospectively evaluated. Results: Average age (15 male and 18 female) was 4.2 years (2–10). 17 patients had scoliosis with 34 (18–52), 3 had kyphosis with 53.3 (43–68) and 13 had kyphoscoliosis [mean scoliosis 38 (20– 62)], kyphosis 32(11–78)]. Twenty HV were located in thoracic spine (T3-T11), 9 in thoracolumbar spine (T12-L1) and 10 in lumbar spine (L2–L5). In 6 patients, two-level hemivertebrectomy was done. The mean level of instrumentation was 2.2, operation time was 4.2 h and blood loss was 412 ml. Mesh cages were placed at resected area in all patients Mean follow-up was 48 m (24-120). Coronal plane deformity improved to 6 (82%) and was 6.2 at final follow-up. Sagittal plane deformity improved to 3 (94%) and it was 4at final follow-up. None of the patients exhibited neurological problems associated with surgery. One patient had dural tear and 2 had superficial infection. Pseudoarthrosis and early or late implant failure was not detected. Conclusion: Type 3 hemivertebrectomy and strut grafting via posterior approach and transpedicular instrumentation is safe and effective in young children. Several advantages are; immediate correction with no reliance on concave growth, no risk of regrowth of HV, immediate stability and reconstruction of anterior column with successful restoration of sagittal plane alignment.

72 THE RESOLUTION OF SYRINGES AFTER POSTERIOR FOSSA DECOMPRESSION IN CHIARI MALFORMATION PATIENTS ASSOCIATED WITH SCOLIOSIS T Wu, X Sun, F Zhu, B Qian, Y Yu, Z Zhu, B Wang, Y Qiu Spine Surgery, Drum Tower Hospital, Nanjing University Medical School, Nanjing, China Summary: This study investigated the resolution of syringes after posterior fossa decompression (PFD) in Chiari malformation Type I (CMI) patients associated with scoliosis through measurements of the maximal S/C ratio and length of syringes. Significant decrease in syrinx size was observed in nearly all patients. Patients with more severe tonsil descent are likely to have a better resolution of syringes. Induction: The behavior of syrinx resolution after PFD in CMI patients associated with scoliosis was rarely reported in literatures. This study was to investigate the outcome of syringes after PFD in CMI patients associated with scoliosis and to find out the possible factors in predicting better outcomes. Methods: CMI patients associated with scoliosis receiving PFD procedures between 2000 and 2009 were included after meeting the following criteria: (1) Age \20 years; (2) Diagnosed as syringomyelia associated with CMI; (3) Scoliosis as a main complaint in each patient; (4) At least MRI scans of spine at pre-op and 3 to 6 months post-op; (5) No prior surgical treatment. Cases with acquired CMI anomalies or receiving syringosubarachnoid shunting were excluded. The maximal S/C ratio and length of syringes were measured to evaluate the resolution of syringes. And 20% decrement in S/C or length at latest visit was defined as a significant radiographic improvement. Result: Totally 46 patients were included. Of all patients, 38 were identified with one follow-up MRI scan, 7 with two follow-up MRI scans and 1 with three follow-up MRI scans. Till the latest follow-up, the mean improvement ratios of S/C and of length of syringes after PFD were 53and 57%, respectively. And 97.8% (45 of 46) of patients had a significant radiographic improvement of syringes

Eur Spine J (2011) 20 (Suppl 4):S421–S464 on MRI. The degree of cerebellar tonsil descending correlated with the surgical outcome (r = 0.100, p = 0.006). In 8 patients with more than one follow-up MRI scans, significant improvement of syringes can be observed at the first follow up, but no significant difference of the size of syringes was found between at the first follow up and at the latest one. Conclusion: Syringes presented a significant resolution after PFD in most CMI patients associated with scoliosis. Severe tonsil descending might be a predictor of better surgical outcomes. The main resolution of syringes took place within 3 to 6 months after PFD. And no obvious changes in the size of syringes were observed after the acute improvement. Keywords: Resolution, Syrinx, Chiari malformation, Scoliosis, Posterior fossa decompression.

73 COMPARISON OF SOMATOSENSORY EVOKED POTENTIALS BETWEEN ADOLESCENT IDIOPATHIC SCOLIOSIS AND CONGENITAL SCOLIOSIS WITHOUT NEURAL AXIS ABNORMALITIES Z Chen, Y Qiu, W Ma Spine Surgery, Drum Tower Hospital, Nanjing University Medical School, Nanjing, China Summary: This study evaluated somatosensory evoked potentials (SEPs) in adolescent idiopathic scoliosis (AIS) and congenital scoliosis (CS) patients with similar curve pattern and severity both in coronal and sagittal planes. It showed that the rate of abnormal SEPs was higher in AIS group than CS group. It reveals somatosensory pathway dysfunction exist in a subgroup of AIS patients, and it might be not secondary to scoliosis curve. Introduction: Abnormal somatosensory evoked potentials (SEPs) have been documented in adolescent idiopathic scoliosis (AIS) patients with different cure severity. However, few studies investigated whether abnormal SEPs were the cause or effect of idiopathic scoliosis. This study was to investigate the significance of abnormal SEPs in AIS patients, and to explore its effect on the etiopathogenesis of AIS. Methods: Posterior tibial nerve SEPs (PTN-SEPs) were performed on AIS and CS female patients. The inclusion criteria were: AIS patients with a Lenke type 1 curve; CS patients with right thoracic curve (apex between T5 and T12) and normal sagittal profile (kyphosis less than 50 degree measured from T2 to T12). All patients were evaluated with a total spine magnetic resonance imaging, and those with neural axis abnormalities were excluded. The patients with neurological deficits on detailed physical examination were also excluded. Absence of SEPs waveforms, prolongation of peek latency or asymmetrical peek latency were defined as pathological change. The incidence of pathological SEPs and clinical characteristics were compared between AIS and CS patients. Results: Forty-six AIS and thirty-three CS patients were included in this study. There were no significant difference in coronal and sagittal Cobb angle between the two groups. The rate of abnormal SEPs was 32.6% (15/46) and 15.2% (5/33) in AIS and CS groups, respectively, and the difference was statistically significant (P \ 0.05). Conclusion: Somatosensory pathway dysfunction exist in a subgroup of AIS patients, and it might be not secondary to scoliosis curve. Significance: –

S457 Keywords: idiopathic scoliosis, somatosensory evoke potentials, etiopathogenesis

74 IMPACT OF MULTIMODAL INTRAOPERATIVE MONITORING DURING SURGERY FOR SPINE DEFORMITY AND POTENTIAL RISK FACTORS FOR NEUROLOGIC MONITORING CHANGES B Feng, G Qiu, J Shen, J Zhang Department of Orthopaedic Surgery, Chinese Academy of Medical Science and Peking Union Medical College, Beijing, China Objectives: The aim of this study was to evaluate efficacy of multimodal intraoperative neuromonitoring for predicting iatrogenic neurologic injury during surgical correction of spine deformity and evaluate the potential risk factors for neurologic monitoring changes. Methods: The records of 176 consecutive patients who underwent surgery for the treatment of spinal deformity were reviewed. The patients were monitored using transcranial electric motor evoked potential (tceMEP) and/or somatosensory-evoked potential (SEP). Alterations with MEP wave amplitude decreasing more than 75% and SEP amplitude decreasing more than 50%, as compared to the baseline, were diagnosed as positive changes. Risk factors related to neurologic monitoring changes (NMCs) were evaluated, in light of preoperative neurolgocial deficit, comorbidity of spinal cord deformity, procedure of osteotomy, main curve cobb’s angle and diagnosis of kyphosis. Results: Combined MEP/SEP monitoring were successfully achieved in 175 of 176 cases. Eleven cases presented with ture NMCs according to MEPs. One patients had irreversible neurological deficit and 4 patients had transient neurological deficit after waking-up from operation. All the five cases were detected by MEP monitorings. SEP lagged MEP for average 15 min when both presented with positive changes. SEP detected 2 cases with dorsal impingement of spinal cord. The sensitivity and specificity of MEP were 91.7 and 98.8% respectively. Solo SEP were 50 and 95.2%. Combined MEP and SEP were 92.9 and 99.4%. Procedure of osteotomy, curve Cobb’s angle more than 90 and preoperative kyphosis were correlated with higher incidence of NMCs. Conclusion: MIOM provides higher sensitivity for monitoring during spine deformity surgery and can predict events of neurological injury. Detection of NMCs and adjustment of surgical strategy may prevent irreversible neurologic deficit. The possible risk factors for NMCs during spine deformity surgery included osteotomy procedure, kyphosis correction and preoperative Cobb’s angle more than 90

Monitoring result during a VCR procedure for 107 kyphosis

123

S458

Eur Spine J (2011) 20 (Suppl 4):S421–S464

Risk factors for neurologic monitoring changes (NMCs) in light of the osteotomy procdure, diagnosis of kyphosis, Cobb’s angle of main curve, presence of preoperative neurologic deficit and spinal cord deformity Title observed Procedure

Item

Osteotomy

Number Number Incidence Chiof cases of (%) square NMCs tests (P) 44

7

16

131

4

3

With kyphosis

15

3

20

Without kyphosis

160

8

5 17

No

\0.01

osteotomy Kyphosis

Cobb’s angles (major curve)

\90

29

5

C90

132

6

5

Preoperative neurologic deficit

Presence

15

2

13

Absence

160

9

7

Spinal cord deformity

Presence

22

1

5

Absence

153

10

7

0.02

0.01

beforehand, was tightened over the graft to prevent anterior migration. Collar orthosis (Vista) was applied, and started walking in two days postoperatively. The collar was continued for three months. Results: There was no case of infection, deterioration of neurological status, or postoperative graft failure. All the cases showed bone union within three months. Discussion: Multilevel corpectomy extremely increases segmental instability, which was thought to be a cause of failure in anterior reconstruction. The study showed that stabilizing the middle segment with pedicle screws reduced segmental instability, because it work as a strut to support the cervical spine. Upper and lower interface between the graft and host vertebral bodies were thought to be mechanically weak with this technique. But as shown here, fixation with a screw and a small plate proved to be sufficient. Anterior pedicle screw technique showed excellent stability with safety after multilevel corpectomy.

[0.05

[0.05

75 ANTERIOR CERVICAL RECONSTRUCTION WITH PEDICLE SCREWS AFTER FOUR LEVEL CORPECTOMY M Ikenaga, M Mukaida, R Nagahara, N Tsubouchi, Y Nakagawa Orthopedics, Kyoto Medical Center, Kyoto, Japan Introduction: There were many previous reports of complications concerning cervical reconstruction after multilevel corpectomy. Lack of stability seems to be a main cause of this complication. On the other hand, posterior reconstruction with cervical cervical pedicle screws provided excellent stability as Abumi et al reported. We then introduced anterior pedicle screw fixation technique after multilevel corpectomy. The paper reports the operative technique, and short-time results of four cases who received four level corpectomy. Materials and methods: Four patients with cervical myelopathy who received four corpectomy were included in the study. Two CSM and two OPLL. Fusion level was C2-7. Preoperative, and postoperative clinical status was evaluated with JOA (Japanese Orthopedic Association) scores for cervical myelopathy. Operative technique: After corpectomy from C3 to C6 through left anterior approach, their right pedicles were exposed under microscope. The pedicles were reamed with probe, and poliaxial screws (Oasys, medial biased angle screw, Stryker Spine) 3.5 mm in diameter, 20 mm in length, were inserted. Rod was placed and locked. Fibula in an appropriate length was harvested, shaped to fit the cavity, and tapped into left side of the instrument. Caudal end of the graft was fixed with a diagonal screw into C7. Cephalic end was fixed with small cervical plate (Reflex Hybrid, Stryker Spine). Circular wiring with polyethylene cable (Nespron Cable, Nesco), passed under the rod

123

54-year-old male. Postoperative AP view. The patient had massive OPLL from C3 to C6. After corpectomy from C3 to C6, pedicle screws were put into the right pedicles, and fibula strut grafting was performed with screws and plate

Eur Spine J (2011) 20 (Suppl 4):S421–S464

S459 (45–69 years). A surrogate head was attached to the occipital mount. The spine was stabilized using a pneumatic muscle force simulation. The specimens were incrementally accelerated from 2 to 6 g using a custom built sled system. The experiment was stopped as soon as any structural injury became macroscopically visible. Shear force (Fs) in the direction of acceleration as well as the moment (My) around the axis perpendicular to the direction of acceleration were measured using a 6-component load cell. Horizontal head (axhead) and T1 (axT1) acceleration were measured using accelerometers. Head rotation was measured using an optical motion capturing system. Statistical analyses were performed using a Wilcoxon’s rank sum test to a significance level of p = 0.05. Results: axhead and axT1 were greater in RI than in FI and SI and were almost equal in SI and FI. Contrary, head rotation in RI was significantly lower than in SI and FI. No significant difference was found between SI and FI. My was greater in SI and FI than in RI. Fs was significantly different between RI and SI. Greatest Fs was measured in RI. Structural injury occurred at an acceleration level of C3 g in SI and FI and at C4 g in RI. RI resulted in injuries of the anterior disc at levels C56 and C67. FI resulted in injuries of the posterior ligaments and facet joint at levels C45, C56, C67. SI resulted in injuries of the facet joints at levels C34, C45, C56, C67. Conclusion: SI and FI seem to have a higher injury risk. axhead and axT1 do not seem to be appropriate predictors of injuries, while head rotation and My could possibly be considered as an appropriate injury criterion.

77 VALIDITY AND RESPONSIVENESS OF THE CORE OUTCOME MEASURES INDEX (COMI) FOR THE NECK CD Fankhauser, UM Mutter, E Aghayev, AF Mannion Spine Center, Schulthess Klinik, Zu¨rich, Switzerland

Postoperative lateral view

76 INJURY RISK DURING REAR, FRONTAL AND SIDE IMPACTS KL Ha¨ußler, HJ Wilke, M Zieringer, R Wagner, A Rakebrand, A Kienle Institute of Orthopaedic Research and Biomechnics, Centre for Muskuloskeletal Research, Ulm University, Ulm, Germany Introduction: Whiplash injuries of the cervical spine belong to the most common but poorly understood injuries in traffic accidents. A deeper understanding of the injury mechanisms could lead to enhanced diagnosis and treatment of these injuries. Several in vitro studies investigated rear (RI), frontal (FI) and side (SI) impacts. However, differences between the impact directions concerning the injury risk to the cervical spine remain unclear. Therefore, the aim of this study was to compare kinetic and kinematic parameters between the three impact directions. Materials and methods: 20 fresh-frozen human cervical spine specimens (C0-T1) were used in this study. 7 specimens were each tested in FI (44–82 years) and SI (51–90 years) and 6 in RI

Patient-orientated outcome questionnaires are essential to evaluate treatment success. To compare different treatments, hospitals, and surgeons, standardized questionnaires are required. The present study examined the validity and responsiveness of the Core Outcome Measurement Index for neck pain (COMI-neck), a short, multidimensional outcome instrument. Questionnaires were completed by patients with degenerative problems of the cervical spine undergoing cervical disc arthroplasty before (N = 89) and 3 months after (N = 75) surgery. The questionnaires comprised the EuroQol-Five Dimension (EQ-5D), the North American Spine Society Cervical Spine Outcome Assessment Instrument (NASS-cervical) and the COMI-neck. The COMI and NASS-cervical scores displayed no notable floor or ceiling effects at any timepoint whereas for the EQ-5D, the highest or lowest values were reached in around 32.5% of patients at follow-up. With one exception (symptom-specific well-being), the individual COMI items and the COMI summary score correlated to the expected extent (R = 0.4–0.8) with the scores of the chosen reference questionnaires. The area under the curve (AUC) generated by ROC analysis was significantly higher for the COMI (0.96) than for any other instrument/sub-scale when self reported treatment outcome was used as the external criterion, dichotomised as ‘‘good’’ (operation helped a lot/helped) versus ‘‘poor’’ (operation helped only little/did not help/made things worse). The COMI had a high effect size (standardised response mean; SRM) (2.34) for the good global outcome group and a low SRM for the poor outcome group (0.34). The EQ-5D and the NASS-cervical lacked this ability to differentiate between the two groups, showing less distinct SRMs for good and poor outcome groups. This study provides evidence that the

123

S460

Eur Spine J (2011) 20 (Suppl 4):S421–S464

COMI-neck is a valid and responsive questionnaire in the population of patients examined. Further investigations should examine its applicability in other patient groups with less severe neck pain or undergoing other treatment modalities.

78 LONG TERM FOLLOW-UP OF RECONSTRUCTED PEDIATRIC CERVICAL SPINE USING PEDICLE SCREW FIXATION SYSTEM: MORPHOLOGIC CHANGE OF FIXED CERVICAL SPINAL COLUMN BY STRESS SHIELDING Y Abe, K Abumi, M Ito, Y Kotani, H Sudo, K Nagahama, A Iwata, A Minami

Fig. 1 A case in Reduced-group. The screw tip penetrated the anterior cortex of the vertebral body

Department of Orthopaedic Surgery, Hokkaido University Graduate School of Medicine, Sapporo, Japan Introduction: Posterior fixation using the constrained type pedicle screw system in the cervical spine reduces the axial load of the anterior column. However, this mechanical alternation of the reconstructed spinal column by changing of anterior/posterior axial load share is not well known. In this study, we investigated the morphologic change of cervical column in pediatric patients fixed using pedicle screw system. Methods: Nine pediatric patients (age at surgery: 9–18 years, 7 boy and 2 girl) who underwent cervical fixation using the pedicle screw system between 1994 and 2004 were included in this study. Five patients had cervical kyphosis preoperatively; postlaminectomy kyphosis in three, post-atlantoaxial fixation in one and posttraumatic in one. Remaining 4 patients had posttraumatic cervical instability. Five of 9 patients underwent posterior reconstructive fusion alone, and 4 patients underwent combined anterior fusion with autologous bone grafting. Antero-postero diameter of vertebral body (D.Vb), and fusion level lordosis angle were evaluated with lateral X-ray images. Results: The mean follow-up term was 8.9 years and all patients obtained bony union. Four patients showed the reduction in D.Vb at the final follow up (Reduced-group) and 5 showed the increase in D.Vb with growth (Non-reduced-group). All 4 in Reduced-group underwent combined anterior fusion. Fusion level lordosis at final follow up was 4.0lordosis in Reduced-group and 2.2kyphosis in NonReduced-group. Statistic analysis showed that combined anterior fusion (p = 0.008) and postoperative fusion level lordosis (p = 0.048) were significantly related to the reduction of D.Vb. The anteroposterior diameter tended to reduce especially at the middle of fusion levels, and this trend was remarkable at the level of combined anterior bone grafting. In the patient of 4-level fusion, the D.Vb reduced to 45% of the original at the postoperative 5 years follow-up (Fig. 1). Conclusion: The results of present study suggest that postoperative lordosis achieved by strong correction force of pedicle screw system and destruction of the endplate by combined anterior bone graft are the risk factors for the atrophic change of vertebral body in pediatric patients. Destruction of endplate and instrumentation-induced stress shielding could alter the remodeling condition and cause the resultant atrophic change. Early removal of the instruments before reduction of vertebral body diameter might be recommended.

123

79 OPTIMAL ANTERIOR APPROACH FOR THE CERVICOTHORACIC JUNCTION LESIONS SB Im, KW Park, BT Kim, DS Shin, WH Shin, SC Hwang Neurosurgery, Soonchunhyang Univerisity, Bucheon, Korea (ROK/South Korea) Objectives: We would like to report the advantage and technical feasibility of partial transmanubrial approach for the cervico-thoracic junction (CTJ) lesions and emphasize its appropriateness. Background: The anterior approach for the CTJ is generally considered to be treacherous because the reversal of lordosis to kyphosis makes visualization deeper and is hindered by the several vital structures. Anterior approach is selected because many lesions on cervico-thoracic junction develop on the anterior column, and it offers direct decompression and reconstruction. Although many types of approach were developed, each method has a limit. We performed transmanubrial approach, which bisect upper part of the manubrium down to 2nd intercostal space with inverted T-shape. Materials and methods: Between 2005 * 2010, 10 patients were undergoing the transmanubrial approach for the CTJ junction lesion. The etiology was consisted with five metastases, two primary bone tumors, two tuberculosis epidural abscesses and one trauma. We analyzed the preoperative decision method for manubriotomy, surgical feasibility and difficulty, postoperative result and complications. The mean follow-up duration was 2.5 years. Results: The supramanubrial parallel line was critical for the decision of manubriotomy. Parallel line to the upper margin of the lesion was more important than lower one. Because the main obstacle of downward dissection was left subclavian vein and aorta, the area caudal to the T3 cannot be extended even we do full sternotomy. Thus, full sternotomy seems like to have a part of unnecessary exposure. The anterior decompression and spinal column reconstruction can be performed with Harms cage. Thoracic duct injury and chylothorac was occurred in one case, and it was managed with maintaining chest tube drain and lipid free diet for 2 weeks. Preoperative impending neurological symptoms were resolved in all cases. Conclusions: Transmanubrial approach provides effective direct decompression and reconstruction method for the lesion on CTJ.

Eur Spine J (2011) 20 (Suppl 4):S421–S464

S461

Partial splitting manubriotomy approach might be optimal surgical option for the various pathology on anterior CTJ and can be performed with low morbidity. 80 DOES ATLANTOAXIAL DISLOCATION INFLUENCE THE SUBAXIAL CERVICAL ALIGNMENT AND DISC DEGENERATION S Wang, C Wang Orthopaedic, Peking University Third Hospital, Beijing, China Introduction: Atlantoaxial dislocations (AAD), although uncommon, are associated with complex deformities of the cervical spine, significant risk of neurological deterioration, and typically require surgical treatment. We noticed that severe AAD were always associated with kyphosis of the occipital-cervical junction. In this setting, questions arise during the occipital-vertebral surgery: Does atlantoaxial dislocation influence the subaxial cervical alignment and disc degeneration? Is anatomic reduction necessary for atlantoaxial dislocation? Despite this, very little information is available concerning cervical alignment in individuals living with this pathologic condition. In addition, whether or not these alignment abnormalities accelerate cervical disc degeneration remains uncertain. Previous studies have reported the alignments of the occipital-cervical and subaxial spine was closely interrelated in asymptomatic individuals. To our knowledge there is no prior study to examine these relationships in such a population. Methods: From 2007 to 2010, 197 patients with atlantoaxial dislocation and atlas occipitalization were studied. Angles formed between Occiput-C2 and C2-C7 were measured (Fig 1). The relationship between the alignment of the occipital-cervical junction and the subaxial cervical spine was evaluated, as well as those between cervical alignment and cervical disc degeneration. Results: The range of values for the angles measured were as followed: The Occiput-C2 angles were -32.1 to 44.6o; and the C2-C7 angles were -17.4 to 77.8o. Statistically significant negative correlations were observed between the Occiput-C2 and C2-C7 angles (Fig 2). Among patients in their 4th, 5th, and 6th decades of life, C5-C6 disc degeneration had a significantly negative correlation with the Occiput-C2 angles. Conclusion: The alignment of the subaxial cervical spine was found to have a significant correlation with that of the occipital-cervical junction in this population. Anterior dislocations of the atlas are associated with diminished lordosis or even kyphosis of the occipital-cervical junction, and result in compensatory hyperlordosis of the subaxial cervical spine, collectively presenting as a ‘‘swan neck’’ deformity. Atlantoaxial dislocation may influence the global cervical alignment and may subsequently accelerate cervical disc degeneration.

Fig. 2 Correlation coefficient of angulation between the Occiput-C2 and C2–C7 (Y = 26.33–0.79X, R2 = 0.73, P = 0.001)

81 DEVELOPMENT, TREATMENT AND OUTCOME OF NEUROPATHIC (CHARCOT) ARTHROPATHY OF THE SPINE IN PATIENTS WITH SPINAL CORD INJURY J Krebs, N Aebli, T Po¨tzel Orthopaedics & Spine Surgery, Swiss Paraplegic Center, Nottwil, Switzerland Introduction: Neuropathic (Charcot) arthropathy of the spine is a rare, but severe, progressive, degenerative disease. It is characterized by the destruction of the intervertebral disc and vertebrae, hypertrophic ossification, ossification of soft tissue, hyperkyphosis, instability and in some advanced cases pseudarthrosis. Neuropathic arthropathy develops in the absence of deep sensation in a joint, which is subjected to repetitive overload. There is a lack of data concerning the early signs and risk factors of Charcot’s disease of the spine and the complications and outcome of surgical treatment. Methods: The case histories of patients suffering from spinal cord injury who were admitted to the Swiss Paraplegic Centre from January 1, 1999 to December 31, 2010 and who were diagnosed with Charcot’s disease of the spine were investigated. A total of 25 patients (18 male, 7 female) with 31 Charcot joints of the spine were identified. Results: The majority (n = 22) of affected patients were paraplegics with an ASIA A impairment score (n = 19). Spinal cord injury (SCI) had mainly resulted from trauma (n = 16) or infection (n = 5). The most common symptoms were back pain, sitting imbalance, kyphosis and pressure sores. High intensity sport or vocational activities and laminectomy were identified as potential risk factors. Charcot joints were observed in the lumbo-sacral (n = 14), thoraco-lumbar (n = 9) and lumbar spine (n = 8). All Charcot joints occurred below the level of SCI and in the trauma patients below the initial fracture site. Charcot joints were located within or below an instrumented area of the spine. The time from SCI to diagnosis of a Charcot joint was on average 22.4 ± 12.1 years (4.9–39.9 years). Three patients were treated conservatively. The other underwent instrumented, multilevel, postero-lateral spondylodesis with additional anterior spondylodesis in four patients. The mean follow-up was 6.4 ± 4.1 years (1.0–16.0 years). Implant loosening in sacral (n = 6) and lumbar (n = 1) vertebrae, formation of a second Charcot joint (n = 2), infection (n = 1) and increased spasticity (n = 1) were identified as

Fig. 1 Measurements of the angels on the neutral lateral X-rays

123

S462

Eur Spine J (2011) 20 (Suppl 4):S421–S464

complications. In one patient, pain relief after surgery was not satisfying. Conclusions: Surgical treatment achieved satisfying pain relief and sitting balance in the majority of patients. Spondylodesis including the sacrum showed a high complication rate (60%).

83 IMPLANT SURVIVAL AFTER DEEP WOUND INFECTION FOLLOWING INSTRUMENTED SPINE SURGERY

82 PROPHYLACTIC ANTIBIOTIC TREATMENT AND RISK FACTORS FOR POSTOPERATIVE WOUND INFECTIONS IN SURGERY FOR LUMBAR DISC HERNIATION. A COHORT STUDY FROM A NATIONAL REGISTRY FOR SPINE SURGERY.

Department of Surgery, Spine Unit, 1) Universitat Auto`noma de Barcelona, 2) German Scoliosis Centre Bad Wildungen, Barcelona, Spain

S Habiba, A Sørlie, T Solberg Neurosurgery, University of North Norway, Tromsø, Norway Objective: There is no consensus as to whether prophylactic antibiotics treatment (PAT) should be given routinely or to selected groups of patients operated for lumbar disc herniation (LDH). The aims of the study were to assess the rate and effectiveness of PAT, and evaluate risk factors for postoperative wound infections (PWI). Methods: In 2006 a National Registry For Spine Surgery was established to monitor the quality and safety of spine surgery for degenerative disorders of the lumbar spine. This is a prospective study of the first 2958 consecutive cases operated for LDH from Oct.06 to Dec.10. Primary outcome measure was occurrence of PWI, registered at first postoperative follow up (mean:2.8 months). Risk factors were evaluated in multivariate analyses with PWI (Yes/No) as dependent variable and PAT (Yes/No) as the exposition variable. We adjusted for possible confounders in a logistic regression model. Results: PWI was reported in 74 (2.5%) cases. Among them 53 (71.6%) were superficial, 13 (17.6%) were deep and 8 (10.8%) cases could not be classified. PAT was given in 2289 (77.4%) of the cases. Of the population with PWI 48 (64.9%) had received PAT and 26 (35.1%) had not (p = 0.009). However, of those who received PAT, more patients had been operated previously. They were also older and more obese. The duration of surgery was longer among patients with PWI (p = 0.02). When adjusting for these possible confounders, PAT protected against PWI (OR 2.4, 95% CI 1.4–4.08, p = 0.001). In addition, duration of surgery above median time ([ 60 min) was a independent risk factor for PWI (OR 1.8, 95% CI 1.04–3.08, p = 0.034). Conclusion: This study shows that PAT is frequently used, and is given to a proportion of the population who are at higher risk for PWI. PAT seems to be effective in preventing PWI. Long duration of surgery is also an independent risk factor for developing PWI. We propose that PAT should be administered routinely, especially if surgery is expected to last more than 60 min.

123

S Nu´n˜ez Pereira, D Rodrı´guez-Pardo, C Pigrau, J Bago´, C Villanueva, E Ca´ceres, F Pellise´

Background: Deep surgical site infection (DSSI) after instrumented thoracolumbar spine surgery is a major complication that can considerably affect the outcome of surgery. DSSI treatment efficacy and its final consequences are under-reported. PURPOSE: Our aim was to evaluate long-term implant survival after DSSI, and associated risk factors. Materials and methods: Forty-four patients (mean age 51.4 y, SD 20. 27; 61.3% women) operated at the same centre between January 2006 and December 2008, consecutively developed acute postoperative DSSI. Preoperative diagnoses were: 15 degenerative lumbar disease, 13 scoliosis, 10 fractures, 6 tumours. All patients were properly treated under surveillance of the infectious diseases department with surgical debridement and antibiotic therapy for a minimum of 8 weeks, based on culture and antibiogram. Kaplan–Meier survivorship analysis and Cox proportional risks analysis were performed. Terminal events were defined as implant removal or death related to DSSI. The mean follow-up was 26 (± 13.9) months. Results: Eleven patients (25%) presented a terminal event. In 10 patients implants were removed: 7 cases because of persistent DSSI activity despite multiple debridements and antibiotics, and 3 others because of DSSI recurrence after a symptom-free period. One patient died due to sepsis. Survival rates after the first debridement were 85% at 6 months, 83% at 1 year, 73.4% at 2 years, and 67% at 3 and 4 years. In 4 of the 11 patients, new implants were placed to avoid progression of the deformity. In 2 of these patients, DSSI recurred. Patients who developed septic shock (HR 15.07; 95% CI 3.6–62.2), those who had more than one debridement (HR 3.78 95% CI 1.5–9.4), and those with [3 fused segments (HR 5.8 95% CI 1.6–27.1) had a higher risk of developing a terminal event (p \ 0.05). Patients with polymicrobial infection (HR 1.7 95% CI 0.5–5.6) and those with higher ASA scores (HR 1.8 95% CI 0.7–4.9) tended to have a higher risk of implant removal. Conclusions and discussion: Implant survival is seriously compromised after properly treated DSSI, and decreases progressively over the first 24 months. Almost one third of implants will not survive at 3 or 4 years. The fusion length and ‘‘aggressiveness’’ of the infection predict the final survival rate.

Eur Spine J (2011) 20 (Suppl 4):S421–S464 84 IS POSTOPERATIVE ANTIMICROBIAL PROPHYLAXIS NEEDED FOR THE MANAGEMENT OF SURGICAL SITE INFECTION AFTER SPINAL INSTRUMENTATION SURGERY? T Numasawa, A Ono, K Wada, Y Yamasaki, G Kumagai, T Yokoyama, S Toh Orthopaedic Surgery, Hirosaki University Graduate School of Medicine, Hirosaki, Japan Background: It is widely accepted that postoperative antimicrobial prophylaxis (AMP) is effective in reducing the risk of surgical site infections (SSI) following spinal surgery. After publication of the Guideline for Prevention of Surgical Site Infection by the Centers for Disease Control and Prevention in 1999, a large number of studies confirmed the effectiveness of AMP. Due to the possible emergence of AMP resistant bacteria or appearance of side-effects, we have treated and managed patients who underwent spinal surgery without postoperative antimicrobial agents since 2003. Purpose: To investigate the incidence of SSI in patients without administration of antibiotics after spinal instrumentation surgery. Subjects: A consecutive 468 patients (230 males and 238 females) were adopted in this study from November 2003 to June 2010. Mean age at the time of operation was 57.1 years. We defined this group as the non-postoperative dose group. There were 121 patients (25.9%) who underwent instrumentation surgery. On the other hand, we defined patients who were administered postoperative multiple doses of AMP between January 2000 and October 2003 as the postoperative dose group. There were 340 cases, consisting of 198 males and 142 females in this group. Average age at the time of operation was 51.3 years. There were 147 patients (43.2%) who underwent spinal instrumentation surgery. Methods: All patients were administered 1 g of cefazolin within 30 min of skin incision, and the same dose of antimicrobial agent was added every 4 h during surgery in the non-postoperative dose group. We administered AMP before and for 7 days after surgery in the postoperative dose group. Results: The postoperative infection rate was only 1.50% (7 cases), of which 6 cases were superficial infections and 1 case was a deep infection in the non-postoperative dose group. In the postoperative dose group, there were 9 confirmed ostoperative wound infections in the 340 patients for an overall SSI rate of 2.65%. There was no significant difference between the two groups. The incidence of SSI in patients who underwent spinal instrumentation surgery was 0.83% (one of 121 patients) in the non-postoperative dose group and 2.04% (three of 147 patients) in the postoperative dose group. There was no significant difference between two groups even with the use of spinal implants. Conclusions: The duration of antimicrobial prophylaxis was not related to the SSI rate at our institution. Postoperative administration

S463 of antibiotics appears to be unnecessary for spinal surgery even with spinal implants when perioperative management was achieved for the patient condition and surroundings as recommended in the CDC guidelines.

85 DOES POST-OPERATIVE SURGICAL SITE INFECTION INFLUENCE THE OUTCOME FOR PATIENTS UNDERGOING SURGERY FOR METASTATIC SPINAL CORD COMPRESSION? Q Nasir, R Trichy Centre for Spinal Studies and Surgery, Queeens Medical Centre, Nottingham, United Kingdom Background: The incidence of infection in surgery for metastatic spinal cord compression (MSCC) has been reported to occur between 12–20%. Most studies in the literature report on the incidence of surgical site infection but do not report on the influence of infection on the neurological outcome and survival of these patients. The objective of this study was to establish if the surgical site infection adversely affected the neurological outcome and survival in these patients. Patients and methods: We analysed a consecutive cohort of 154 patients who underwent surgery for MSCC between October 2003 and March 2010. The patients were divided into infected and noninfected groups and the outcome measures were analysed. The outcome measures studied included the morbidity (length of in-patient stay and number of surgical procedures), neurological outcome (Frankel grade) and survival (one-year survival rate and mean survival in days). Results: The incidence of infection was 13% in our study (20/154). The mean length of stay was 28.4 ± 22 days in the infected group and 18.2 days ± 22 in the non-infected group (p = 0.05). Most patients with an infected wound required additional procedures (16/20, 80%) with wound debridements/washout. There was no difference in the neurological outcome between the groups, with most patients remaining the same or improving by one Frankel grade (16/20, 80% in Infected group versus 117/134, 86% non-infected group, p = 0.45). The median survival was 151 days (mean 356 ± 437) in the infected group and 227 days (mean 545 ± 720) in the non-infected group but this difference was not statistically significant (p = 0.10). Discussion and conclusions: The occurrence of surgical site infection increases the morbidity following surgery for MSCC with significantly longer in-patient stay in these patients. However, the neurological outcome and survival are not adversely affected.

123

S464 86 THE REVISED TOKUHASHI SCORE AND SURVIVAL IN PATIENTS WITH METASTATIC SPINAL CORD COMPRESSION (MSCC) S Elsayed, V Dvorak, N A Quraishi Centre of Spine Studies & Surgery, Queen’s Medical Centre, Nottingham, United Kingdom Objectives: The revised Tokuhashi score has been widely used to evaluate indications for surgery and predict survival in patients with metastatic spinal disease. Our objective was to determine whether the score accurately predicted survival in those with MSCC. Study design: Retrospective analysis Subjects: All patients with MSCC presenting to our unit were included in this study from October 2004 to December 2009. Patients were divided into three groups—The Revised Tokuhashi score 0–8, 9–11 and 12–15.

123

Eur Spine J (2011) 20 (Suppl 4):S421–S464 Outcome measures: Neurological outcome and survival Results: A total of 109 patients with MSCC were managed in our unit during this time. Mean age of patients was 61 years (range 7–86). Mean and median survival was 350 (5–2256) and 93 days in the 0–8 group, 439 (8–1902) and 229 days in the 9–11 group, and 922 (6–222) and 875 days in the 12–15 group; p = 0.01. All patients underwent decompression and stabilisation surgery. The rate of consistency between the prognostic score and actual survival was 64% (0–8), 64% (9–11) and 69% (12–15). Overall the consistency was 66%. There was no difference in neurological outcome between the 3 groups. Conclusions: There was a significant difference in the mean survival between groups. There was a moderate consistency between predicted and actual survival in this group of patients (66%) who all had cord compression. All patients had undergone some form of decompression and stabilisation surgery regardless of the overall revised Tokuhashi score.