Eur Spine J (2015) 24 (Suppl 6):S711–S742 DOI 10.1007/s00586-015-4130-8

ABSTRACTS

QUICK FIRES (Concurrent Short Communications) EUROSPINE 2015 BASIC SCIENCE

QF1 MODELLING THE FAILURE MECHANISM OF THE ANNULUS FIBROSUS Marle`ne Mengoni, Alison Jones, Nagitha Wijayathunga, Ruth Wilcox Institute of Medical and Biological Engineering - University of Leeds, Leeds, UK Introduction: A number of surgical techniques in the intervertebral disc, such as needle puncture during nucleus replacement, discography, or simple annulus suturing, can cause direct or progressive damage leading to degeneration. The aim of this study was to assess the damage and failure strength of the anterior annulus fibrosus (AF) under radial tension with the goal of including damage in in-silico models of interventions. Methods: Twelve discs were extracted from frozen mature ovine thoracolumbar spines. Cuboid AF specimens were dissected and divided into three groups: outer AF (N = 6), inner AF (N = 8) and specimens across the AF thickness (N = 12). Tensile radial tests were performed at 1mm/min and load was recorded. The mechanical behaviour of the specimens was defined by the linear modulus, stress and strain values at occurrence of damage, initial local failure and macroscopic failure. A 1D variable-stiffness spring model was designed to replicate the damage part of the experimental data. Springs were assembled in series representing the lamellae or the inter-lamellar connections. Each spring was given an initial stiffness value derived from a linear model [Mengoni et al., WCB2014]. The damage variation was computed for each group of specimens with an optimization process to minimize the difference between in-vitro and in-silico stiffness. Results: Significant differences were found between the outer annulus group and the other groups for the stress at which apparent damage or failure occurred. No significant difference was observed within each group between the stress at the occurrence of first failure and at macroscopic failure. The optimization converged to damage values reproducing the sample stiffness within 10 % of the experimental value. At a given strain level, the predicted damage of the lamellae was higher than that of the inter-lamellar connections. Discussion: The experimental data suggests that, in radial loading, the annulus has an almost linear behaviour up to about 20 % strain

followed by a modulus decrease until the first instance of local failure. Several local failures follow at similar stress levels before macroscopic failure. The in-silico results show the lamellar damage is predominant in the damage mechanism of the samples. In conclusion, this study demonstrated that damage can be adequately modelled using 1D variable-stiffness springs, and this approach can now be applied to models of the whole annulus to examine clinical interventions. Disclosures: author 1: none; author 2: none; author 3: none; author 4: grants/research support; Company = Engineering and Physical Sciences Research Council.

QF2 TOTAL DISC REPLACEMENT USING TISSUEENGINEERED INTERVERTEBRAL DISCS IN A CANINE MODEL Yu Moriguchi, Jorge Mojica Santiago, Peter Grunert, Rodrigo Navarro, Thamina Khair, Lawrence Bonassar, Roger Ha¨rtl Weill Cornell Medical College, New York, USA Introduction: The most commonly performed treatments for degenerative disc disease, anterior cervical decompression and fusion and prosthetic total disc replacement devices (TDR), despite being effective still pose risks of pseudoarthrosis, implant dislodgement, and adjacent segment disease. Tissue engineered intervertebral disc (TE-IVD), an alternative treatment option, has been previously developed by our group as a biological TDR device and tested in a rat tail model [Bowles 20011]. Presently, we evaluate the in-vivo efficacy of our TE-IVD in the canine cervical spine. Methods: TE-IVD components were constructed in vitro using either AF or NP cultivated canine disc cells; the collagen gel based composite AF enclosed an alginate gel based composite NP. 12 skeletally mature beagles underwent discectomy with whole IVD resection and were divided into two groups: solely discectomized control and TEIVD implanted group. Adjacent proximal segments served as internal healthy control. All beagles were imaged post-operatively at 2 and 4 weeks. Quantitative analysis using T2 intensity measured NP size and hydration of implanted TE-IVD, while X-rays measured disc height indices of treated segments. Qualitative histological analysis evaluated implant engraftment and ingrowth over time plus secondary degeneration post discectomy. Results: 2-week MRIs of implanted TE-IVDs revealed T2 high intensity with acute outer inflammation due to surgical breach, which faded by 4-weeks. At 4 weeks, TE-IVD sustained position in disc space with relatively increased T2 intensity, resulting in over 70 % of NP hydration as that of healthy discs. Conversely, discectomized disc segments were void of hydration. Disc height indices of TE-IVDs and discectomized discs were 71 % and 49 %, respectively, of that of

123

S712 healthy control discs. Further, TE-IVD histological sections exhibited chondrocyte-like cell viability, abundant proteoglycan content in the extracellular matrices, and substantial integration into the host tissues without signs of immune reaction. Conclusions: Despite significant biomechanical demands of the beagle cervical milieu, our in vivo TE-IVDs maintained structure and hydration in addition to disc height of the treated segment at up to 4-weeks. Discs displayed dynamic adaptation to the host environment, with extracellular matrix production, and cell proliferation. Further long-term experiments will elucidate the clinical applicability and efficacy of the presented innovation. Disclosures: author 1: none; author 2: none; author 3: grants/research support; Company = AO-Spine; author 4: none; author 5: none; author 6: stock/shareholder; Company = 3D BioCorp; author 7: grants/research support; Company = Depuy synthes, consultant; Company = brainlab.

QF3 DECREASED SCIATIC NERVE MOVEMENT IN PEOPLE WITH POSTOPERATIVE RESIDUAL LEG SYMPTOMS FOLLOWING DISCECTOMY FOR LUMBAR INTERVERTEBRAL DISC DISORDER Gary Shum, Sally Cinnamond, Andrew Clarke, Daniel Chan, Mike Hutton, Rohan Chauhan, Jon Marsden Plymouth University, Plymouth, UK Background: Lumbar intervertebral disc disorder is one of the most commonly diagnosed pathologic conditions affecting the spine. Discectomy is cost effective and largely successful in the management of lumbar intervertebral disc disorders. However, postoperative residual leg pain can occur in up to 32 % people with lumbar intervertebral disc disorder following discectomy. This results in continued pain and functional limitations that impact on the ability to work and quality of life despite after a successful surgery. It is hypothesised that postoperative residual leg symptoms could due to persistent shortening or adhesions of the sciatic nerve, leading to decreased sciatic nerve movement. Methods: Fifteen participants with postoperative residual leg symptoms following discectomy for lumbar intervertebral disc disorder were recruited. An innovative ultrasound imaging technique has been developed to measure the sciatic nerve movement at the knee level during forward bending movement. Off-line cross-correlation analysis of the ultrasound imaging was used to track the movement of the sciatic nerve at the popliteal fossa of both the pain and non-painful side. The range of motion of the lumbar spine and hip were also measured simultaneously by accelerometers, which were placed on the L1, S2 and bilateral thighs. Paired t-test was used to compare the sciatic movement between the painful and non-painful sides in patients with postoperative residual leg symptoms following discectomy for lumbar intervertebral disc disorder. Results: In the limited lumbar spine flexion during forward bending (30.92 ± 8.57°), there was a significant decrease (p \ 0.05) in the sciatic nerve movement in the painful side (2.19 ± 2.01mm) when compared with the non-painful side (4.19 ± 2.32mm). Discussion and Conclusion: The results suggested that decreased sciatic nerve movement may contribute to the continued postoperative residual radiating pain even when the spinal disorders have been treated successfully by surgical intervention. The observed limited sciatic nerve movement could be a consequence of prolonged immobilisation prior to spinal surgery, leading to the shortening, adhesion and thus decreased movement of the sciatic nerve. Intervention should be developed to mobilise and stretch the

123

Eur Spine J (2015) 24 (Suppl 6):S711–S742 sciatic nerve in patients with residual postoperative leg pain following spinal surgery. Disclosures: author 1: grants/research support; Company = Royal Devon and Exeter Hospital, United Kingdom, author 2: employee; Company = Plymouth University, author 3: grants/research support; Company = Royal Devon and Exeter Hospital, author 4: none; author 5: no indication, author 6: grants/research support; Company = RDEFT NHS Foundation Trust, employee; Company = RDEFT NHS Foundation Trust, author 7: grants/research support; Company = william Scholl Podaitric endowment Fund.

QF4 IS THERE ANY CORRELATION BETWEEN PATHOLOGICAL PROFILE OF FACET JOINTS AND CLINICAL FEATURE IN PATIENTS WITH THORACOLUMBAR KYPHOSIS SECONDARY TO ANKYLOSING SPONDYLITIS? AN IMMUNOHISTOCHEMICAL INVESTIGATION Bangping Qian, Mingliang Ji, Jun Hu, Yong Qiu, Saihu Mao Spine Surgery, the Affiliated Drum Tower Hospital of Nanjing University Medical School, Nanjing, China Objective: To systematically and extensively evaluate the immunopathology of the facet joints in patients with thoracolumbar kyphosis secondary to ankylosing spondylitis (AS). Methods: The facet joints were obtained from 30 AS patients and 23 age- and gender-matched controls (patients with fresh thoracolumbar fracture). The facet joints were assessed immunohistochemically by analysing the number of infiltrating T cells (CD3, CD4, CD8), B cells (CD20), microvessel density (CD34), osteoblasts (CD56), bone marrow macrophages (CD68), and osteoclasts (CD68) per high-power field (hpf). According to the presence or absence of persistent inflammation, AS patients were divided into 2 groups: A (patients with persistent inflammation) and B (patients without persistent inflammation). Lumbar spinal mobility was assessed using the modified Schober index (MSI). Results: Two or more CD3+ T cell aggregates were found in the facet joints from 18 of 30 AS patients, whereas one CD3+ T cell aggregate was noted in 5 of 23 patients with thoracolumbar fracture. The levels of T cells (CD4+ and CD8+), CD20+B cells, CD56+ osteoblasts and CD34+ microvessel density were significantly higher in AS patients compared with the controls (all P \ 0.01). Notably, the MSI score in group A was significantly higher than that in group B (P \ 0.01). Conclusion: Active spinal inflammation is frequently observed in AS patients with thoracolumbar kyphosis. In addition, persistent inflammation in facet joints may further contribute to the loss of spinal mobility in the later stages of AS. These findings indicate that careful monitoring of disease activity is mandatory for AS patients in its advaced stage. Disclosures: author 1: none; author 2: none; author 3: none; author 4: none.

QF5 STUDY OF NEURAL CONNECTIVITY USING DTITRACTOGRAPHY OF THE THE SENSORIMOTOR NETWORK IN IDIOPATHIC SCOLIOSIS Julio Domenech, Angel Alberich-Bayarri, Luis Martı´-Bonmatı´, Gracia´n Garcia-Martı´, Marı´a De la Iglesia-Vaya´, Jose Marı´a Tormos, ´ lvaro Pascual-Leone A

Eur Spine J (2015) 24 (Suppl 6):S711–S742

S713

Hospital Arnau de Vilanova. University Cardenal Herrera CEU. Valencia. Spain

Emilia-Romagna Region Health Authority - Department of Hospital Services, Bologna, Italy

The aetiology of idiopathic scoliosis (IS) remains largely unknown, but growing evidence suggests an underlying CNS disorder. IS patients have an asymmetric interhemispheric motor cortex hyperexcitability evidenced with paired pulse transcranial magnetic stimulation, and an increased activation of supplementary motor cortex as revealed by functional MRI. Such findings are consistent with a subclinical sensorimotor integration disorder, reminiscent of idiopathic dystonia in which a sensorimotor integration disorder may be a causative factor. Diffusion tensor imaging (DTI) allows identification of white matter fibres, which appear grouped in compact packages and thus limit the movement of water molecules primarily to a direction parallel to the main vector of axons structure. This characteristic allows the processing of a synthetic reconstruction of the brain nervous routes (tractography) using probabilistic approaches. Objective: To investigate the microarchitecture of white matter pathways in IS compared to healthy controls. Methods: Twenty Patients with IS and 20 healthy controls, matched for age (15, 1 SD 1.2 respectively) were studied using 3T-MRI probabilistic diffusion and tractography. IS Patients had an average deformity of 418 Cobb (range 30° to 64°). MR acquisition parameters for DTI imaging included a spin-echo echo-planar imaging (SE-EPI) single-shot pulse sequence with a b-value of 1300 s/mm2; 64 different space directions; TR: 6200 ms; TR: 67 ms; voxel size 2 9 2 9 2 mm; 60 axial slices. Results: There was a significant reduction in maximal diffusivity of the cerebello-thalamo-cortical tract and body of corpus callosum in IS patients compared with healthy controls. No differences were found in the corticospinal tract neither as compared to controls nor between hemispheres within group. Conclusion: This study shows an alteration in the axonal architecture of the connectivity pathways in the sensory-motor network in patients with IS. These abnormalities may be the substrate of the neurophysiological and subclinical abnormalities found in IS. They may also support the hypothesis of a defective motor control at the central level, involving transcallosal and cortico-subcortical-cerebellar circuitries as a causative factor of the spinal deformity. This investigation line may help to deepen in the understanding of IS aetiology and postulate new treatment strategies such as neuromodulation of the sensoriomotor circuitry in early stages of the deformity. Disclosures: author 1: none; author 2: stock/shareholder; Company = QUIBIM SL, none; Company = none, author 3: none; author 4: none; author 5: grants/research support; Company = author 6: none; author 7: none.

To our knowledge, there are not currently available biochemical markers to predict the severity of functional outcome after spinal cord injury (SCI). These factors would allow to complete the patient’s outline with further clinical information, thus enhancing the reliability of early prognosis. The aim of the study is to identify new possible biomarkers for SCI clinical outcome through the analysis of the cerebrospinal fluid (CSF) in the context of primary care of injured patients. CSF was obtained by lumbar puncture performed within the first 24 hours from the injury, before surgical intervention. The study design will include 30 patients matching the inclusion/exclusion criteria detailed in the study plan. Eight patients were recruited up to now. The CSF was then processed using high-throughput technology for multiparametric protein quantification (Luminex platform). Three magnetic bead panels were used for biomarker assays: Cytokine/ Chemokine 21-plex (Bio-Rad), Neurological Disorders 4-plex and Neurodegenerative Diseases 10-plex (Millipore), so that 35 protein markers were investigated. Samples of CSF from healthy volunteers were purchased (PrecisionMed) and used as controls. Among cytokines/chemokines, we found an increase in GRO-A (Chemokine C-X-C Motif Ligand 1) and MIF (Macrophage/microglia migration inhibitory factor) protein levels, while HGF, SDF1a, SCGF-B, IL16, MIG and CTAK were unchanged. P-tau (but not T-tau) also increases, such as RANTES (Chemokine C-C motif ligand 5), brain-derived neurotrophic factor, platelet-derived growth factor, siCAM (soluble Intercellular Adhesion Molecule 1) and PAI-1 (Plasminogen activator inhibitor-1). IL2Ra, INFa2, SDF1a, MCP3, IL12, LIF, TNFb, IL18, IL1a, IL3, SCF, TRAIL, M-CSF resulted undetectable in both control and lesioned patients. In few subjects (N = 3) we measured the same markers also 3 months after lesion. We observed that Stem Cell Growth Factor beta is higher than in controls immediately after lesion, thus decreasing below the control values at three months after lesion. Although these results are still preliminary and the data have yet to be related to the functional outcomes of each patient (under collection), the use of high-throughput technologies seems to be a promising approach for the discovery of prognostic markers of clinical outcome in SCI patients. (Ethical committee approval: n. 281, 14/02/2013; Dr. F. DeIure and Dr. F. Boni, Maggiore Hospital, are gratefully acknowledged). Disclosures: author 1: none; author 2: none; author 3: none; author 4: none; author 5: none; author 6: none; author 7: none; author 8: none; author 9: none; author 10: none.

QF6 INVESTIGATION ON THE PROGNOSTIC VALUE OF BIOCHEMICAL MARKERS IN THE CEREBROSPINAL FLUID FOR THE FUNCTIONAL OUTCOME OF THE SPINAL CORD INJURED PATIENTS. Laura Calza`, Mercedes Fernandez, Rita Capirossi, Roberto Montevecchi, Nozha Borjini, Jacopo Bonavita, Tiziana Giovannini, Gian Piero Belloni, Salvatore Ferro, Giovanni Gordini Health Sciences and Technologies – Interdepartmental Center for Industrial Research, University of Bologna; Bologna Local Health Authority - Intensive Care Unit, EMS and Trauma Centre, Maggiore Hospital, Bologna; Montecatone Rehabilitation Institute SpA, Imola;

QF7 THE MECHANISM OF LIGAMENTUM FLAVUM HYPERTROPHY: INTRODUCING ANGIOGENESIS AS A CRITICAL LINK THAT COUPLES MECHANICAL STRESS AND HYPERTROPHY Hong Joo Moon, Junseok Hur, Joohan Kim, Youn-Kwan Park Korea University Guro Hospital, Seoul, Korea Background: Biochemical alterations associated with mechanical stress have been explored as an initiating step in the pathological progression of ligamentum flavum hypertrophy (LFH); however, this mechanism remains poorly understood. Recently, the inflammation induced after mechanical stress and the subsequent response of ligamentum flavum (LF) cells have been implicated in LFH pathology.

123

S714

Eur Spine J (2015) 24 (Suppl 6):S711–S742

Objective: This study investigated the hypothesis that angiogenesis may be a critical link between hypertrophy and a series of stimulating events, including mechanical stress. Methods: LF from 20 lumbar spinal canal stenosis (LSCS) patients and 16 non-LSCS patients (control group) were collected during surgery. Patients’ demographic and radiographic data were obtained. The levels of angiogenic factors (VEGF, angiopoietin-1, VCAM, and bFGF) in the LF were investigated using an enzyme-linked immunosorbent assay (ELISA). Angiogenesis was also quantified by immunohistochemical detection of CD34-positive capillaries. The correlations among clinical factors, including radiographic factors, angiogenic factors and angiogenesis, were statistically analyzed. Results: The LSCS group was older and exhibited a longer symptom duration, wider segmental motion, and thicker LF than the control group. The LSCS group showed significantly higher tissue concentrations of VEGF (p \ 0.001) that positively correlated with LF thickness (r = 0.557, p \ 0.001) and segmental motion (r = 0.586, p \ 0.001). The LSCS group showed significantly more CD34-positive capillaries than the control group (p = 0.004). Conclusions: The LSCS group showed greater segmental motion, higher VEGF concentrations and more CD34-positive capillaries than the control group. These data indicate that VEGF-mediated angiogenesis following mechanical stress may be a critical step within the series of pathological events in LFH. Disclosures: author 1: none; author 2: none; author 3: none; author 4: none.

similar across segments (L23:19 %; L34:27 %; L45:17 %; L5S1:14 %) compared to A-P. A-P ICR coordinates, averaged to provide a single centre of rotation (COR) value, converged slightly posterior to the mid-coronal plane (L23: -5 %; L34: -3 %; L45: -6 %; L5S1: -11 %) but tended to mask differences in ICR migration patterns across segments. Averaged S-I COR coordinates, however, provide a simple metric to highlight differences in segmentspecific ICR patterns. L23 and L34 CORs were located midway between mid-transversal plane and superior endplate of inferior vertebra (27 % and 30 % respectively). Comparatively, L45 and L5S1 ICRs were located significantly higher- at the superior endplate of the respective inferior vertebrae (48 % and 53 % respectively). Plots of continuous segmental ICR/FHA migration patterns will also be presented. One of the potential implications of level-specific ICR pattern differences lies in lumbar disc arthroplasty device design. Disclosures: author 1: grants/research support; Company = CDC/ NIOSH R21 OH00996; Marie Curie COFUND Postdoctoral Fellowship; Swiss National Science Foundation Ambizione Research Grant, author 2: none; author 3: grants/research support; Company = NIH, author 4: none; author 5: grants/research support; Company = University of Pittsburgh.

QF8 LEVEL-SPECIFIC DIFFERENCES IN MIGRATION OF INSTANTANEOUS CENTRES OF ROTATION (ICR) OF LUMBAR INTERVERTEBRAL JOINTS DURING LIFTING

Werner Schmoelz, Alexander Keiler, Marko Konschake, Nicola Marotta, Alessandro Gasbarrini

Ameet Aiyangar, Liying Zheng, William Anderst, Bernhard Weisse, Xudong Zhang

Introduction: In patients suffering from osteoporosis in situ augmentation of pedicle screws with PMMA is currently the common method to improve the anchorage of pedicle screws However, PMMA augmentation of pedicle screws bears the risk of cement leakage, PMMA may have toxic effects and can make a revisions difficult. To overcome these drawbacks, a modified technique applying a selfcuring elastomeric material into a balloon created cavity prior to screw insertion was developed. The aim of the present study was to compare screw anchorage of the two augmentation techniques. Methods: Nine lumbar vertebral bodies were used for testing. Right pedicles were instrumented with cannulated and fenestrated pedicle screws and augmented in situ with 2ml PMMA. Left pedicles were instrumented with cannulated pedicle screws. Prior to left screw insertion a cavity was created with a balloon and was filled with 3ml of self-curing elastomer (VK100, BONWRX, USA). Each screw was subjected to a cranio-caudal cyclic load starting from -50 to 50N while the upper load was increased by 5N every 100 load cycles until failure or 11000 cycles (600 N). The relative motion of the screw during cyclic loading was recorded using a 3D motion analysis system mounted to the screw head. After the cyclic loading a pull out test of the screws was conducted. Results: All in situ PMMA augmented screws failed by screw cut out, while in the elastomer group with the balloon cavity technique 4 screws cut out and 5 screws did not fail during cyclic loading. In 7 of 9 vertebrae the cavity creation augmentation technique with the elastomer outperformed the in situ PMMA augmentation. The mean cycles to failure were 9824 (±1982) and 7401 (±1644) for the elastomer and PMMA group, respectively (p = 0.012). The pullout test showed significant differences in pullout force (p = 0.003) and displacement (p = 0.002); with 671.3 N ±332.1 and 5.9 mm ±4.6 for elastomer group and 1188.6 N ±288.1 and 12.3 mm ±2.2 for the

EMPA, Mechanical Systems Division, Duebendorf, Switzerland; University of Pittsburgh, Pittsburgh, USA Although significant migration of the ICR—a popular parameter for characterizing translation of lumbar intervertebral segments (LIS—; can occur during motion, in vivo ICR data during dynamic, functional tasks are sparse. This study aimed to map ICRs of LIS during a lifting task and investigate how these might differ across the lumbar segments. 11 healthy subjects lifted 3 known weights (4.5, 9, 13.5 kg) from a trunk-flexed position (*75°) to an upright position. A Dynamic Stereo X-ray system (DSX) recorded dynamic images of lumbar (L2S1) motion. A custom, markerless model-based tracking software was used to process the dynamic, biplane X-ray data into highly accurate (B 0.268, 0.2 mm) 6DOF bone kinematics. LIS motion was expressed as motion of cranial- relative to the caudal vertebrae for L23, L34, L45 and L5S1. Finite Helical axes (FHA) of rotation were computed at 1° increments of respective LIS extensions and plotted against overall L2-S1 ROM progression to allow comparison across segments. Anterior-posterior (A-P) and superior-inferior (S-I) coordinates of the ICRs were then normalized to the depth and height respectively of the inferior vertebral body for each LIS (0 % = vertebral body centre). Data from the 3 weights were aggregated since, on initial, preliminary analysis, inter-weight differences seemed relatively small. Average A-P migration range (MR) of ICRs was 47 % and 32 % of bone depth for L23 and L34 respectively. By contrast, L45 (11 %) and L5S1 (20 %) ranges were much lower. S-I MRs were more

123

QF9 AUGMENTATION OF PEDICLE SCREWS WITH A SELF-CURING ELASTOMERIC MATERIAL

Medical University Innsbruck, Austria

Eur Spine J (2015) 24 (Suppl 6):S711–S742 PMMA group. In the elastomer group the main failure mode was screw pullout, while in the PMMA the main failure mode was pedicle fracture. Conclusion: Compared to in situ PMMA augmented pedicle screws the modified pedicle screw augmentation technique with balloon cavity creation and self-curing elastomeric silicone resulted in an improvement in pedicle screw anchorage under cyclic cranio-caudal loading. The pullout force after cyclic loading for the PMMA augmented screws was higher with a different failure mode. Disclosures: author 1: grants/research support; Company = B.Braun Aesculap Italy; author 2: none; author 3: none; author 4: none; author 5: none.

QF10 DOES THE TIMING OF CEMENT APPLICATION AND REPOSITION EFFECT PEDICLE SCREW ANCHORAGE? Luis Alvarez, Christian Hainrich, Werner Schmoelz, Felix Tome, Angel R. Pin˜era, Marta Martin-Fernandez, Javier M. Duart Fundacio´n Jimenez Diaz Introduction: Cement augmentation is an establish method to increase the pedicle screw (PS) anchorage in osteoporotic vertebral bodies.If a reposition maneuver of the instrumented vertebra is part of the surgical procedure, the ideal time for augmentation is still controversially discussed. Augmentation of the PS prior to the reposition maneuver will likely increase the force which can be applied during the maneuver. However, it bears the risk to impair the anchorage of PS already during the maneuver. Augmentation of the PS after the reposition maneuver might result in trabecular bone fractures during the maneuver. Possibly impaired screw anchorage due to the reposition maneuver might be restored by the subsequent augmentation and early screw loosening might be prevented. The purpose of the present study was to evaluate effect of cement application timing on PS anchorage in the osteoporotic vertebral body. Methods: Ten lumbar vertebrae (L1-L5) were used for testing. The left and right pedicles of each vertebra were instrumented with same PS size and used for pairwise comparison of the two timing points for augmentation. For the reposition maneuver the left PS was loaded axially under displacement control (2 9 ±2mm, 3 9 ±6mm, 3 9 ±10mm) in order to simulate a reposition maneuver. Subsequent both PS were augmented with 2ml PMMA cement. The same force as measured during the left PS maneuver was applied to the previously augmented right hand side PS (2 9 F(±2mm), 3 9 F(±6mm), 3 9 F(±10mm)). Both PS were cyclic loaded with initial forces of +50 and -50N while the lower force was increased by 5 N every 100 cycles until total failure of the PS. The PS motion was measured with a 3D motion analysis system. After cyclic loading stress X-Rays were taken to document the PS loosening and to identify the PS loosening mechanism. Results: In comparison with PS augmented prior to the reposition maneuver, PS augmented after the reposition maneuver showed a significant (p \ 0.05) higher number of load cycles until failure (5930 ± 1899 vs 3830 ± 1706). The loosening mechanism for PS augmented after the reposition maneuver was PS toggling with the attached cement cloud with the trabecular bone. While PS augmented prior to the reposition maneuver showed a motion of the screw within the cement cloud. Conclusion: The time of cement application has an effect on PS anchorage in the osteoporotic vertebral body.PS augmented after the reposition maneuver showed a significant higher number of load cycles until screw loosening.

S715 Disclosures: author 1: grants/research support; Company = Biomet, consultant; Company = Spineart; author 2: none; author 3: grants/ research support; Company = Biomet; author 4: none; author 5: none; author 6: none; author 7: none.

CERVICAL SPINE

QF11 NOVEL CERVICAL ANGULAR MEASURES ACCOUNT FOR BOTH UPPER CERVICAL COMPENSATION AND SAGITTAL ALIGNMENT Christopher Ames, Themistocles Protopsaltis, Renaud Lafage, Daniel Sciubba, Robert Hart, Justin Smith, Christopher Shaffrey, Frank Schwab International Spine Study Group Hypothesis: Cervical measures can account for upper cervical compensation and cervical sagittal deformity. Design: Prospective cohort and retrospective review of prospective database Introduction: Current descriptions of CD like C2-C7 plumbline (cSVA) do not account for upper cervical compensation. Thoracolumbar deformity (TLD) angular measures like the T1 Pelvic Angle (TPA), can account for both global and pelvic tilt and are less affected by lower extremity compensation. Such advantages are lacking in established cervical measures. The craniocervical angle (CCA) combines the slope of McGregor’s line and the inclination from C7 to the hard palate, thus it accounts for cervical alignment and upper cervical compensation (C0-2A). The C2-Pelvic Tilt (CPT) is an angle that combines C2 tilt and pelvic tilt, thus, like TPA, it is less affected by lower extremity and pelvic compensation. Methods: Novel and existing CD measures were correlated in 781 pts from a TLD database and 61 pts from a prospective CD database. CD pts were subanalyzed by region of deformity driver: cervical (C), and cervico-thoracic junction (CT). TLD pts were grouped if they had cervical deformity (cSVA[4 cm or TS-CL[20) or not. Results: TLD cohort: Mean cSVA was 31.7° ± 17.8 mm. In pts with cervical deformity, mean CCP = 56.0° ± 7.4 and CPT = 33.6° ± 15.8 were significantly different than nonCD pts (p\0.001). CCA and CPT correlated with cSVA (r = 0.49/r = 0.42, p \ 0.001) and C0-2A (r = 0.63/r = 0.29, p \ 0.001). CD cohort: mean cSVA was 47.3° ± 32.2 mm. CCA and CPT correlated with cSVA (r = 0.71/r = 0.66, p \ 0.001) and C0-2A (r = 0.66/r = 0.61, p \ 0.001). Correlation of cSVA and C0-2A was weaker (CT pts were significantly more deformed by cSVA (71.3mm vs 24.0, p \ 0.001), CCA (47.1° vs 59.1°, p \ 0.001), and CPT (63.3° vs 43.8°, p = 0.002). Using linear regression analysis, cSVA of 4cm corresponded to CCA of 53.2° (r2 = 0.5) and CPT of 48.5° (r2 = 0.4). Conclusion: CCA and CPT account for both cervical sagittal alignment and upper cervical compensation. These novel parameters can be utilized in pre and postoperative assessments of cervical sagittal alignment. Future studies should gauge their clinical relevance with health measures relevant to cervical deformity. Disclosures: author 1: grants/research support; Company = Depuy Synthes Spine, consultant; Company = Depuy, Medtronic, Stryker, stock/shareholder; Company = Baxano Surgical, Doctor’s Research Group, royalties; Company = Aesculap, Biomet Spine, employee; Company = University California San Francisco; author 2: grants/ research support; Company = Zimmer, consultant; Company =

123

S716 Medicrea, Biomet, AlphaSpine; author 3: none; author 4: consultant; Company = Medtronic, Depuy-Synthes, Styker, Orthofixauthor 5: grants/research support; Company = Medtronic, ISSGF, consultant; Company = DepuySynthes, Globus, Medtronic, stock/shareholder; Company = Spine Connect, royalties; Company = Seaspine, DepuySynthes, other financial report; Company = DepuySynthes; author 6: grants/research support; Company = Depuy-Synthes, ISSGF, consultant; Company = Biomet, Nuvasive, Cerapedics, Medtronic, royalties; Company = Biomet; author 7: grants/research support; Company = Depuy-Synthes, consultant; Company = Biomet, Medtronic, Nuvasive, Stryker, stock/shareholder; Company = Nuvasive, royalties; Company = Biomet, Medtronic, Nuvasive; author 8: grants/research support; Company = DePuy, MSD, AO, consultant; Company = Medicrea, MSD, K2M, stock/ shareholder; Company = Nemaris INC, royalties; Company = MSD, K2M, other financial report; Company = Biomet, NuVasive.

QF12 CERVICAL SPINE BALANCE IN SCHEUERMANN DISEASE Piotr Janusz, Marcin Tyrakowski, Tomasz Kotwicki, Kris Siemionow 1

Department of Orthopeadic Surgery University of Illinois at Chicago, Chicago, IL, 2Department of Orthopedics, Pediatric Orthopedics and Traumatology, The Centre of Postgraduate Medical Education in Warsaw, Poland and 3Spine Disorders Unit, Department of Pediatric Orthopaedics and Traumatology, University of Medical Sciences, Poznan, Poland Introduction: The structural hyperkyphosis characterizing Scheuermann disease (SD) may be localized in the thoracic (SDT) or thoracolumbar (SDTL) spine segments. This may effect sagittal cervical balance. Purpose: The aim of this study was to assess sagittal cervical balance in patients with the two types SD. Methods: Seventy-one patients (41 males and 30 females), aged: 16.3 ± 3.8 (11.6 - 31.0) years with SD were enrolled into the study. None of the patients had symptoms attributable to the cervical spine. On standing lateral long cassette radiographs the following measurements were made: 1) cervical spine parameters: C0-2 angle, C2-7 angle (CL), C1-C2 angle, relative rotation angle (RRA) - measured at levels from C2 to C7, cervical tilt (CT), cervical sagittal axis (cSVA); 2) thoracic inlet parameters: thoracic inlet angle (TIA), T1 slope, neck tilt (NT); 3) cranial parameters: C0 angle, cranial offset (CO), and cranial tilt (CRT). Comparison with t test was performed with P value significance level \0.05. Results: There were 38 patients with SDT and 33 with SDTL. In SDT T1 slope was significantly greater than in SDTL (38.1° vs. 28.9°; p = 0.0002), and consequently CL (-19.8° vs. -8.9°; p = 0.0160), CT (29.8° vs. 24.3°; p = 0.0190) and TIA (81.9° vs. 71.1°; p = 0.0022) in SDT were significantly greater as well. The difference in CL was expressed mainly in C6-C5 (-4.8° vs. -1.4°; p = 0.0146) and C5-C4 (-4.4°; vs. -1.5° p = 0.0464) segments. There were no significant difference in proximal cervical lordosis : C0-2 angle (-21.6° vs. -20.8°; p = 0.7597), C1-C2 angle (-30.8°vs. -27.5°; p = 0.1746), C3-C2 (-5.4° vs. -5.1°; p = 0.7976) and C4-C3 (3.5°-1.5°; p = 0.1464) segments. There was no significant difference between the SDT and SDTL patients in cranial parameters C0 angle(8.2° vs. 7.4°; p = 0.6942), angle CRT (8.1° vs. 5.3°; p = 0.1020), CO (-13.9mm vs. -14.9 mm; p = 0.9159).

123

Eur Spine J (2015) 24 (Suppl 6):S711–S742 The values of the NT (47.8° vs. 46.6° p = 0.6768), cSVA (18.5 mm vs. 14.4 mm, p = 0.2880) and C7-C6 (-7.8° vs. -5.1°; p = 0.0921) were comparable in both groups. Conclusion: Scheuermann disease type has an influence on cervical sagittal balance. Localization of structural kyphosis affects T1 slope as well as C2-C7 lordosis, cervical tilt and thoracic inlet angle. Significant difference in C2-7 lordosis in comparison to similar C0-2 lordosis demonstrate that compensation is present in subaxial cervical spine. Position of the head center of gravity is not dependent on SD type. Disclosures: author 1: employee; Company = Department of Orthoaedics, University of Illinois at Chicago; author 2: none; author 3: none; author 4: grants/research support; Company = dePuy, consultant; Company = dePuy, captureproof.com, Providence Medical Technologies, stock/shareholder; Company = providence Medical Technologies, royalties; Company = Amedica, other financial report; Company = Globus, Providence Medical Technologies.

QF13 REGIONAL THORACIC AND LUMBAR SAGITTAL COBB ANGLE CHANGES AND UIV DETERMINE EVOLUTION OF CERVICAL ALIGNMENT AFTER ASD SURGERY: SERIES OF 171 PATIENTS WITH 2 YEAR FOLLOW UP Christopher Ames, Amit Jain, Daniel Sciubba, Han Jo Kim, D. Kojo Hamilton, Justin Scheer, Brian Neuman International Spine Study Group Hypothesis: Operative changes in thoracic kyphosis, lumbar lordosis, and the selection of the upper instrumented vertebrae (UIV) significantly influences changes in cervical alignment over a 2-yr follow-up (F/U) in adult spinal deformity (ASD) pts. Design: Retrospective review of prospectively collected multicenter database Introduction: The aim of our study was to assess the influence of operative changes in spinopelvic parameters on cervical alignment in ASD pts. Methods: 171 ASD pts C18yrs were assessed for changes from baseline to the 2-yr F/U (base-2yr) in the: C2-C7 sagittal vertical axis (C2-C7 SVA), T1-slope (T1S), and C2-C7 lordosis (C2-C7Lord). Multivariate models were constructed to analyze the influence of: UIV selection (T9 and below vs. above T9), and operative changes from baseline to 6-weeks (base-6wk) in the following spinopelvic parameters: thoracic kyphosis (TK), lumbar lordosis (LL), C7-S1 SVA, pelvic incidence, pelvic tilt and sacral slope. Results: The base-2yr changes in C2-C7 SVA and in T1S were both significantly associated with the surgical changes from base-6wk in TK, LL and with the UIV selection (Figure 1). Interestingly, the operative correction of C7-S1. SVA from base-6wk was not significantly associated with either changes in C2-C7 SVA or T1S over the 2-yr F/U. Multivariate model revealed that changes from base-2yr in the C2-C7Lord were associated with the base-6wk changes in the C7-S1 SVA (P = 0.004). The majority of changes in the C2-C7 SVA over the 2-yr F/U occurred in the first 6 weeks after surgery (base-2yr 95 % CI: -0.1mm to +4.6mm, and base-6wk 95 % CI: +0.7mm to +4.7mm). Over the 2-yr F/U, on average, there was loss of C2-C7Lord, majority of which was lost in the first 6 weeks after surgery (base-2yr 95 % CI: -3.2 to +0.5deg, and base-6wk 95 % CI: -4.8 to -1.2deg). Conclusion: Reciprocal changes in cervical alignment occur in response to operative changes in TK, LL and C7-S1 SVA. Cervical alignment is also influenced by UIV selection. Majority of changes occur in the first 6 weeks and persist over.

Eur Spine J (2015) 24 (Suppl 6):S711–S742 Disclosures:: author 1: grants/research support; Company = Depuy Synthes Spine, consultant; Company = Depuy, Medtronic, Stryker, stock/shareholder; Company = Baxano Surgical, Doctor’s Research Group, royalties; Company = Aesculap, Biomet Spine, employee; Company = University California San Francisco; author 2: none; author 3: consultant; Company = Medtronic, Depuy-Synthes, Styker, Orthofix; author 4: consultant; Company = K2M, Medtronic, Biomet, other financial report; Company = Depuy, Stryker; author 5: none; author 6: none; author 7: grants/research support; Company = Depuy Synthes.

QF14 KYPHOPLASTY C2: INDICATIONS AND SURGICAL TECHNIQUE Ahmed Shawky, Ali Ezzati Helios Klinikum Erfurt - Germany; Assiut University Hospitals, Egypt Introduction: Osteolytic lesions of C2 are rare. Pathological fractures C2 have devastating consequences regarding C1/2 stability. Kyphoplasty C2 provides primary stability of C2, prevents possible C2 fractures and subsequent instability with preservation of C1/2 mobility. Only few reports about the results of Kyphoplasty C2 either through trans-oral or anterolateral approach are available. Patients and Methods: Two patients with osteolytic lesions C2 were included. Under general anesthesia antero-lateral approach at the level of C4/5 with deep dissection to the Level C2/3 was used. Under x-ray guidance in anteroposterior and lateral plains Jamshidi-needle was inserted in the midline at the lower anterior point of C2. The needle was then replaced over a guide wire with the working cannula. After obtaining biopsy for histo-pathology, a single balloon is guided through the working cannula and carefully inflated. The balloon is then deflated and cement is applied to fill in the cavity. Both patients were followed up for 12 months. Results: Biopsy confirmed the diagnosis of multiple myeloma in one patient and solitary plasmacytoma in the other. There were no intraoperative or postoperative complications. No cement leak. Both patients showed normal range of motion of the cervical spine postoperatively and at the final follow up. Functional views of the cervical spine at the final follow up did not show C1/2 instability. Discussion and Conclusion: Balloon kyphoplasty of C2 is indicated in osteolytic lesions C2 with different underlying pathologies. Infection is still a contraindication. It showed good clinical and radiographic results. It prevented pathological fractures of C2 with its subsequent morbidities and preserved motion of the upper cervical spine. Compared to the trans-oral approach, the antero-lateral one provides a better access to the whole length of C2 in line with its longitudinal axis. Disclosures: author 1: none; author 2: none.

S717 Introduction: Some authors have reported using allograft bone instead of autograft bone to avoid donor-site complications in posterior atlantoaxial fusion and getting good results. However, the assessment of fusion is different in different literatures. Most of the authors confirmed fusion based on no movement on dynamic radiographs, which is not reliable. There is no prospective comparative study to compare the effectiveness of allograft and autograft in atlantoaxial fusion routinely using computed tomography (CT) image yet. Purpose: To compare the effectiveness of allograft and autograft in atlantoaxial fusion, and to evaluate the feasibility of using allograft for atlantoaxial fusion. Methods: From January 2012 to December 2012, forty-one consecutive patients underwent posterior atlantoaxial fusion with screw-rod system fixation in our spine centre. The choice to use allograft bone or iliac crest autograft bone was made by the patient himself or herself after being told about the advantages and disadvantages of the both methods. In the allograft group, we used the mixed material of morcellized allograft and local autograft for posterior atlantoaxial fusion. In the autograft group, we used the morcellized iliac crest autograft for fusion. Patients were followed-up regularly. All the patients underwent CT scans and dynamic radiographs 6 months after surgery, and every 6 months thereafter till the end of the study or confirmation of fusion. Results: Twenty-four patients underwent posterior atlantoaxial fusion with allograft, and seventeen underwent fusion with autograft. All the patients have been followed-up at least 18 months. At the final followup, only two (8.3 %) patients were confirmed bony fusion based on CT image in the allograft group; while fifteen (88.2 %) patients were confirmed bony fusion in the autograft group. The fusion rate of allograft group is significantly lower than that of autograft group (P\ 0.01). All the forty-one (100 %) patients showed no movement on dynamic radiographs. Conclusions: Allograft is not reliable for posterior atlantoaxial fusion even with the rigid internal fixation of modern constructs. Autograft is still the golden standard for atlantoaxial fusion despite of the donor-site morbidity. The assessment of fusion based on no movement on dynamic radiographs is not reliable. The confirmation of fusion should be based on the presence of bridging bone on CT image. Disclosures: author 1: none; author 2: none; author 3: none; author 4: none; author 5: none; author 6: none; author 7: none; author 8: none.

QF16 CLINICAL APPLICATION OF C1 PEDICLE SCREW AND LATERAL MASS SCREW FOR ATLANTOAXIAL INSTABILITY PATIENTS WITH A NORMAL C1 POSTERIOR ARCH: A PROSPECTIVE, DOUBLE-BLIND, RANDOMIZED CONTROLLED TRIAL Liang Yan, Baorong He, Dingjun Hao

QF15 ALLOGRAFT VERSUS AUTOGRAFT FOR POSTERIOR ATLANTOAXIAL FUSION WITH SCREW-ROD SYSTEM: A PROSPECTIVE COMPARATIVE STUDY Da-Geng Huang, Xin-Liang Zhang, Xiao-Dong Wang, Ding-Jun Hao, Bao-Rong He, Tuan-Jiang Liu, Qi-Ning Wu, Hua Guo Department of Spine Surgery, Honghui Hospital, Xi’an Jiaotong University Health Science Center Xi’an, China

Hong Hui Hospital, Xi’an Jiaotong University College of Medicine, Xi’an, China Purpose: To compare the feasibility and clinical outcomes of C1 pedicle screw and lateral mass screw for atlantoaxial instability patients with C1 posterior arches measuring [4 mm. Methods: A total of 140 patients with atlantoaxial instability (C1 posterior arches measuring [4 mm) were enrolled in a single centre double-blind trial and randomized to receive either C1 pedicle screw fixation or lateral mass screw fixation. Patients were assessed preoperatively, at six weeks, six months, one year and three years after

123

S718 surgery. The operation time, blood loss, intraoperative complications, JOA Score, VAS Score, and bone fusion rates were recorded. Results: The operation was successful in all 140 cases, with all patients showing improvement in clinical symptoms. There were significant differences in operation time and blood loss between the 2 groups (P \ 0.001). The mean follow-up time was 42 months. At the final follow-up, the JOA score was significantly better than the preoperative score (mean, 14.2; P \ 0.05). The mean postoperative improvement rate was 88.7 % and the mean VAS score was 1.7; both results were significant as compared with preoperative results (P \ 0.05). Bone fusion was achieved within 6 months after operation. No screw loosening, shifting, breakage, or atlantoaxial instability was observed. For the C1 lateral mass screw fixation group, nine patients had burst bleeding of C1-2 venous plexus during surgery, and six patients had immediate pain and numbness at the occipitocervical region. Conclusions: C1 pedicle screw fixation is less invasive, simple, less complications, and satisfied clinical results for atlantoaxial instability as compared with C1 lateral mass screw fixation. Disclosures: author 1: none; author 2: none; author 3: none.

QF17 SAFETY AND ACCURACY OF CT-REGION MATCHING NAVIGATED PEDICLE SCREW INSTRUMENTATION OF THE SUBAXIAL CERVICAL AND CERVICOTHORACIC SPINE Yu-Mi Ryang, Niels Buchmann, Jens Lehmberg, Jens Gempt, Bernhard Meyer Department of Neurosurgery, Klinikum rechts der Isar, Technical University Munich, Germany Objective: Small pedicle diameters and the vicinity to the vertebral artery, nerve roots and spinal cord make pedicle screw (PS) placement in the cervical spine technically demanding and especially challenging for the spine surgeon. Because of the aforementioned reasons the use of pedicle screws in the cervical spine is relatively uncommon compared to the thoracolumbar spine and the indication for cervical pedicle screw placement should be meditated with care. Navigation might reduce the risk for pedicle breaching and neurovascular injuries. This prospective study assessed the accuracy and safety of CTnavigated (BrainLABÒ region matching) PS of the subaxial cervical spine (C3-7) and cervicothoracic junction (C6-TH3). Methods: Between March 2014 and February 2015 a prospective cohort of 41 patients received PS instrumentation of the subaxial cervical spine or cervicothoracic junction with the aid of CT-navigated region matching (BrainLABÒ CurveTM Image Guided Surgery). A total of 194 PS (90 cervical; 104 thoracic) were implanted. Indications were metastatic, infectious, traumatic and degenerative spine disease. Pedicle screw accuracy was assessed on routinely performed postoperative CT- or 3D-fluoroscopy scans. Accuracy was graded according to the Gertzbein and Robbins classification (grade A: no pedicle breach, grade B: 0.1-2 mm, grade C: 2.1-4 mm, grade D: 4.1-6 mm, grade E: [6 mm pedicle breach). Pedicle screws graded A or B were classified accurate. Results: Overall accuracy (grades A and B) was 96 %. Four percent of PS were graded as C (7 PS) and D (1PS). There was no difference in accuracy between cervical and thoracic PS (96 %). No neurovascular lesion was experienced and no revision surgery for misplaced pedicle screws was necessary.

123

Eur Spine J (2015) 24 (Suppl 6):S711–S742 Conclusion: CT-region matching navigated pedicle screw instrumentation of the subaxial cervical and cervicothoracic spine is a safe procedure with excellent pedicle screw accuracy. In our study no neurovascular lesion occurred and no revision surgery was necessary. Pedicle screw instrumentation is technically demanding even to the experienced spine surgeon. Indications should be well-considered and performed with the support of CT-navigation to enhance safety and accuracy of this procedure. Disclosures: author 1: consultant; Company = BrainLAB, AOSpine; author 2: none; author 3: none; author 4: none; author 5: grants/ research support; Company = medtronic, Depuy Synthes, Ulrich medical, Brainlab, consultant; Company = medtronic, Depuy Synthes, Ulrich medical, Brainlab, royalties; Company = ulrich medical, spine art.

QF18 SURGICAL TREATMENT OF LOWER CERVICAL FRACTURE-DISLOCATIONS WITH SPINAL CORD INJURIES BY ANTERIOR APPROACH. FIVE-TO 15-YEARS FOLLOW-UP Hua Guo, Biao Wang Honghui Hospital, Xi’an Jiaotong University College of Medicine, Xi’an, China Purpose: Lower cervical fracture-dislocations are often caused by flexion-stretch injuries, and frequently combined with spinal cord injuries, which can cause seriously damage. Up to present, there is no consensus on treatment denominator for lower cervical fracturedislocations. In recent years, anterior approach surgery with directly decompression and reduction has been widely accepted. However, large sample size, long-term follow-up study to assess the clinical efficacy of anterior approach is rarely seen in the literature. Through this retrospective cohort study we assessed the clinical outcomes of anterior approach surgury. Methods: From January 2000 to January 2010, 312 patients with lower cervical spine fracture-dislocations with spinal cord injuries treated by anterior approach were retrospectively analyzed. 218 cases (70 %) were data integrity and obtained follow-up. The average age of 218 patients was 41.5 years (ranged 21-72), including 121 males and 97 females. Classified by the degree of dislocation, we had grade I: 92 cases, grade II: 65 cases, grade III:50 cases and grade IV: 11 cases. All cases underwent skull traction (3-6 kg) for cervical immobilization and avoiding the secondary spinal cord injury before surgery, then anterior discectomy and reduction were performed. If the reduction failed, corpectomy was performed for further reduction. Results: The follow-up time was 8.3 years in average, range from 5 to 15 years. Complete reduction was got in 178 cases (81.7 %), and 40 cases (18.3 %) obtained more than 90 % reduction. The postoperative radiologic indexes of all patients were higher than the preoperative ones (P \ 0.05), but there was no statistical difference between postoperation and the final follow-up (P [ 0.05). The cervical spine normal intervertebral height and physiological curvature were maintained, and there were no plates or screws associated complications observed during the follow-up. 163 cases (74.8 %) presented with neurological functional recovery, and the remaining 55 patients (25.2 %) had no significant changes of neurological function. Conclusions: For lower cervical fracture-dislocations with spinal cord injuries, satisfied clinical outcomes can be obtained by choosing anterior surgery approach. By restoring the normal structure of

Eur Spine J (2015) 24 (Suppl 6):S711–S742 cervical spine and promoting neurological functional recovery, anterior approach achieved good long-term curative effect. Keywords: cervical spine; spinal fracture-dislocations; anterior approach. Disclosures: author 1: none; author 2: none.

QF19 TITANIUM ION LEVELS IN PATIENTS WITH METAL-ONMETAL CERVICAL DISC ARTHROPLASTY: A PROSPECTIVE STUDY UP TO 84 MONTHS Matthew F. Gornet, Vaneet Singh, Francine W. Schranck, Anastasia K. Skipor, J.J. Jacobs The Orthopedic Center of St. Louis - Spine Research Center, St. Louis, USA Introduction: Cervical disc arthroplasty (CDA) has emerged as a motion-sparing alternative to rigid fusion to treat degenerative cervical disc disease. CDA preserves motion by allowing implant components to articulate against each other. However, as observed in other articulating joint replacement implants, bearing surface contact may generate wear debris and, in the case of hard bearing surfaces, may release ions. This study measured serum titanium (Ti) ion levels for patients implanted with a cervical disc through 84-months post implantation. This metal-on-metal disc with ball-in-trough articulation is made of titanium alloy/titanium carbide composite (Ti-6Al-4V/ TiC). Methods: The study included 30 patients following strict exclusion criteria (including no previous permanent metal implants, no professional exposure to metal particles). The serum Ti levels were assayed using high-resolution inductively-coupled plasma-mass spectrometry at pre-op and 3, 6, 12, 24, 36, 60, and 84 months postoperatively. The detection limit for Ti was 0.2 ng/mL, and any value below this limit was assigned a value of half of the detection limit (0.1 ng/mL). The data was statistically analyzed using the Friedman test for longitudinal comparisons. Results: Median serum Ti levels were determined to be 0.10, 1.22, 1.15, 1.27, 1.21, 1.46, 1.34, and 1.42 ng/mL at pre-op, 3, 6, 12, 24, 36, 60, and 84 months, respectively. Serum ion levels at all postoperative time points were significantly higher than preoperative levels (P-value \0.01). Conclusion: Long-term serum Ti ion levels were reported for patients implanted with a metal-on-metalcervical disc at a single level. The results show that the postoperative ion levels were significantly higher than the preoperative levels. The observed Ti ion levels in this study are lower than the reported Ti ion levels in subjects undergoing bilateral total hip joint replacement with a modular stem made out of Ti alloy and Ti based posterior instrumentation used for lumbar arthrodesis [1, 2]. References: 1. Omlor GW, Kretzer JP, Reinders J, Streit MR, Bruckner T, Gotterbarm T, Aldinger PR, Merle C (2013) In vivo serum titanium ion levels following modular neck total hip arthroplasty–10 year results in 67 patients. Acta biomaterialia 9:6278-6282. 2. Richardson TD, Pineda SJ, Strenge KB, Van Fleet TA, MacGregor M, Milbrandt JC, Espinosa JA, Freitag P (2008) Serum titanium levels after instrumented spinal arthrodesis. Spine (Phila Pa 1976) 33:792-796. Disclosures: author 1: consultant; Company = K2M; Medtronic, stock/shareholder; Company = Bonovo; International Spine & Orthopedic Institute, LLC; Nocimed; OuroBoros; Viscogliosi Bros

S719 Venture Partners LLC, royalties; Company = Medtronic; author 2: employee; Company = Medtronic plc; author 3: none; author 4: grants/research support; Company = Medtronic-Spinal and Biologics; author 5: grants/research support; Company = Medtronics, Zimmer, Nuvasive, NIH, stock/shareholder; Company = Implant Protection.

QF20 QUALITY OF LIFE AFTER CERVICAL DISC ARTHROPLASTY TO TREAT DEGENERATIVE DISC DISEASE: TWO YEAR RESULTS OF AN INTERNATIONAL PROSPECTIVE, MULTICENTER, OBSERVATIONAL STUDY(NTC00875810) Saleh Baeesa, Ronai Marton, Jan Stulik Faculty of Medicine King Abdulaziz University, Jeddah, Saudi Arabia Study Design/Setting: Multicenter 2-year prospective observational study. Objective: To describe routine surgical practice and patient reported outcomes (PROs) when treating cervical degenerative disc disease (CDDD) with Cervical Disc Arthroplasty (CDA). Summary of Background Data: CDA is an alternative treatment to Anterior Cervical Decompression and Fusion (ACDF), however it brings the same benefits as the latter in addition to maintaining segmental motion, keeping the kinematics of the spine and protecting the adjacent levels. Methods: A total of 194 patients were enrolled (190 patients implanted) by 15 sites across 7 countries. Patient demographics and intra-operative data were collected; EQ-5D, EQ-VAS and Neck Disability Index (NDI) (PROs), intervertebral disc space and adverse events (AEs) intervertebral disc space and AEs were assessed preoperatively and at 3, 6, 12 and 24 months (mo) follow-ups (FUs). Results: Sixty seven percent of the patients were female, mean age 44.0 years, mean BMI 25.6. Compliance to follow up was: 92 % (175/ 190) at 3 mo, 42 % (80/190) at 6 mo, 59 % (112/190) at 12 mo and 78 % (148/190) at 24 mo. Disc herniation was the most frequent primary indication for CDA (80.5 %). Thirty seven percent of patients already experienced pain for more than 1 year prior to baseline assessment. Mean procedure duration was 87.1 min and mean blood loss was 43.8 ml. The majority (71.0 %) of the artificial cervical discs was implanted at level C5-C6 while 16.3 % of patients received implants at 2 levels. There was a significant improvement from baseline to 3, 6, 12 and 24 months follow ups (FUs) in all 3 PROs: EQ-5D index (baseline 0.59 to 0.78, 0.77, 0.8, 0.76 FUs, p \ 0.001), EQ-VAS (baseline 51.8 to 73.0, 72.0, 74.3, 71.5 FUs, p \ 0.001) and NDI (baseline 24.1 to 13.2, 12.6, 12.3, 13.4 FUs, p \ 0.001). After implantation, the mean disc height at the affected level increased by 0.19mm from baseline (0.22 mm) to the 3 mo FU (0.41 mm) remaining constant thereafter (0.42, 0.40, 0.39). Mean disc height of levels above and below the implant remained comparable among baseline and all FUs. A total number of 63 AEs were recorded in 44 patients, of which 8 serious AEs and 2 serious adverse device effects. Conclusions: Our study shows a significant improvement in patient reported outcomes in the first three months after the CDA which is maintained throughout the 24 months follow-up. Disclosures: author 1: consultant; Company = Medtronic; author 2: consultant; Company = Medtronic and Sanatmetal; author 3: none.

123

S720 QF21 BRYAN DISC ARTHROPLASTY VERSUS ANTERIOR CERVICAL DISCECTOMY AND FUSION FOR THE TREATMENT OF CERVICAL DISC DISEASE. A REVIEW OF LEVEL I-II RANDOMIZED CLINICAL TRIALS INCLUDING CLINICAL OUTCOMES Eduardo Hevia, Marı´a Aragone´s, Alberto Caballero, Carlos Barrios Spine Surgery Unit, Hospital La Fraternidad-Muprespa, Madrid, Spain; Institute for Research on Musculoskeletal Disorders, Valencia Catholic University, Valencia, Spain Introduction: To date, a compilation of the clinical and radiologic outcomes and adverse events of anterior cervical discectomy and fusion (ACDF) compared with a single cervical disc arthroplasty (CDA) design, the Bryan disc has partially accomplished. The controversy concerning the benefits of unisegmental CDA over ACDF is still open because (RCTs) comparing ACDF with cervical arthroplasty has been highly inconclusive. Most of these studies mixed disc prosthesis with dissimilar kinematic characteristics. Methods: Systematic review of randomized clinical trials with evidence level I-II. Only RCTs reporting clinical outcomes were included in this review. After a search on different databases including PubMed, Cochrane Central Register of Controlled Trials, and Ovid MEDLINE, a total of 10 RCTs out of total 51 studies were entered in the study. RTC’s were searched from the earliest available records in 2005 to December 2014. Results: Five studies were Level I, and five were Level II. Out of a total of 1101 patients, 562 patients were randomly assigned into the Bryan arthroplasty group and the remaining 539 patients into the ACDF group. The mean follow-up was 30.9 months. Patients undergoing CDA had lower Neck Disability Index, and better SF-36 Physical component scores than ACDF patients. Patients with Bryan CDA had also less radiological degenerative changes at the upper adjacent level. Overall adverse events were twice more frequent in patients with ACDF. The rate of revision surgery including both adjacent and index level were slightly higher in patients with ACDF, showing no statistically significant difference. Conclusions: This review of evidence level I-II RCTs comparing clinical and radiological outcomes of patient undergoing Bryan arthroplasty or ACDF indicated a global superiority of the Bryan disc. The impact of both surgical techniques on the cervical spine (radiological spine deterioration and/or complications) was more severe in patients undergoing ACDF. However, the rate of revision surgeries at any cervical level was equivalent for ACDF and Bryan arthroplasty. These data suggest that even though the loss of motion has a determinant influence in the development of degenerative changes in ACDF cases, these kinematic factors do not imply a higher rate of symptomatic adjacent segment degeneration requiring surgery. Disclosures: author 1: none; author 2: none; author 3: none; author 4: none.

QF22 DOES PREOPERATIVE DISC HEIGHT AFFECT POSTOPERATIVE MOTION AFTER CERVICAL TOTAL DISC REPLACEMENT? ANALYSIS OF PROSPECTIVE CLINICAL TRIAL WITH 2-YEAR FOLLOW-UP

Eur Spine J (2015) 24 (Suppl 6):S711–S742 Loyola University Chicago, Chicago, Illinois, USA; Edward Hines Jr. VA Hospital, Nines, Illinois, USA Background: There is limited clinical data on the relationship between preoperative disc height and quantity and quality of postoperative motion after cervical total disc replacement (TDR). We investigated this relationship by analyzing the radiographic and clinical results of a prospective, FDA-regulated feasibility study of compressible cervical disc prosthesis. Methods: The study included 30 patients: 12 single-level and 18 twolevel implantations (C4-C5:7; C5-C6:27, C6-C7:14). All patients received a 6mm-height prosthesis (M6-C). An independent core facility performed measurements on preoperative and 2-year postoperative radiographs. Anterior, posterior, and average disc heights were measured at the operated (index) and adjacent segments. Segmental and total (C2-C7) range of motion (ROM) was measured on flexion-extension films. We analyzed the influence of preoperative disc height on the postoperative ROM, location of flexion-extension center of rotation (COR), and clinical outcomes (VAS neck and arm pain, NDI) two-years following TDR. Results: The preoperative disc height at the TDR level was 3.7±0.8 mm (median: 3.7). Group #1 with disc height below the median height had significantly narrower discs than Group #2 with above median disc heights (3.0 vs. 4.4 mm, p \ 0.05). Postoperatively the disc height increased to 5.8±1.0 mm at 2 years, with no significant group difference. Narrower (Group #1) discs were significantly less mobile preoperatively than taller (Group #2) discs (6.7 vs. 10.5 degrees, p \ 0.05). Both groups achieved nearly the same motion postoperatively (5.6 vs. 5.0 degrees, p[0.05); thus, narrower discs experienced greater improvement in motion than the taller discs (p \ 0.05). Postoperatively, the index level COR was maintained posterior to disc midline, although it was located more cephalad than preoperatively (p \ 0.05), closer to the inferior endplate of the implanted disc space. The COR location was unaffected at adjacent levels. The VAS neck and arm pain scores and NDI scores all significantly improved after TDR (p \ 0.05). The preoperative disc height did not influence the postoperative index level COR location, pain scores, or NDI scores (p [ 0.05). Conclusions: The results suggest that disc-space distraction up to twice the preoperative height in a collapsed segment may not degrade the postoperative motion or clinical outcomes after TDR with compressible disc prosthesis; and thus, collapsed discs may be amenable to disc arthroplasty. Disclosures: author 1: grants/research support; Company = Spinal Kinetics, Inc., USA, stock/shareholder; Company = Spinal Kinetics, Inc., USA; author 2: none; author 3: none; author 4: none; author 5: stock/ shareholder; Company = Paradigm Spine, Spinal Motion, Amedica, Spinal Kinetics, royalties; Company = Medtronic, Depuy, Amedica; author 6: consultant; Company = Spine Wave/Globus/Medtronic, stock/ shareholder; Company = Spine Wave, royalties; Company = Spine Wave; author 7: consultant; Company = orthofix, stock/shareholder; Company = Spinal Kinetics; author 8: grants/research support; Company = Orthofix, consultant; Company = Amendia.

QF23 CHARACTERISTICS OF PICA END VERTEBRAL ARTERY: AS ANATOMICAL VARIATION, AND POTENTIAL RISK FOR CERVICAL SPINE SURGERY Takeshi Aoyama, Naoshi Obara

Avinash Patwardhan, Gerard Carandang, Leonard Voronov, Robert Havey, Carl Lauryssen, Domagoj Coric, Thomas Dimmig, David Musante

123

Spine Center, Department of Orthopaedic Surgery, Teine Keijinkai Hospital, Sapporo, Japan

Eur Spine J (2015) 24 (Suppl 6):S711–S742 Introduction: Since asymmetry of vertebral artery(VA) is common, it is advocated that dominant side of VA should be taken care during cervical spine surgery. Although the importance of non-dominant VA is not discussed, there is anatomical variation which terminates as poserior inferior cerebellar artery(PICA): PICA end VA. The injury of PICA end VA possibly leads to cerebellar infarction. Features of PICA end VA are not known well. Frequency, characteristics of PICA end VA, and whether this variation is congenital or acquired change are described in this study. Materials and Methods: Study1; Consecutive 358 cases of head and neck MR angiogram(MRA) in 2 months were checked. MR equipment was 1.5Tesla. The purposes of taking MRA were for acute or chronic strokes, headaches, medical check-up, etc. Diameter of VA was mesured at V2 portion. If non-domonant side VA was less than 3/4 to dominant side, it was determined as asymmetric. Relation between frequency of PICA end VA and asymmetry, diameter of VA was analysed. Study2; MRA of consecutive 62 cases younger than 39yrs in 14months were also measured as same method. Frequency of PICA end VA of this young age group and 324 of the 358 cases, who were over 50yrs, was compared. Results: Of 358 cases male:female were 207:151, age were 10-94 (67.8±13.8). As typical, both VA configure BA in 296 cases (82.7 %), one side were abscent in 18 cases (5.0 %). PICA end VA were in 44 cases (12.3 %). VA were asymmetric in 164 cases (45.8 %). Diameter of all VA were 3.2±0.76 mm. Diameter of PICA end VA were 2.0±0.55 mm, which were significantly smaller than non-dominant but not PICA end VA, 2.8±0.59 mm (p \ 0.001). 38/44 cases (86.4 %) were less than 2.5 mm. Among VA less than 2.0mm, 26/56 cases (46.4 %) were PICA end VA. 40 PICA end VA were in asymmetric VA. And if dominant side is more double than the other, 55.9 % were PICA end VA. PICA end VA were 1/62 cases (1.6 %) and 43/324 cases (13.3 %) in B39yrs and C50yrs, respectively. It is significantly fewer in young population (p \ 0.001). Conclusions: PICA end VA is not rare, especially in cases of VA asymmetry, thin VA, and in the elderly; it is clarified PICA end VA is mainly by aging process. It is not obvious that whether PICA end VA injury result in cerebellar infarction. Such case hasn’t been reported. But PICA end VA also shoud be preserved, since potential risk exists. If VA is thin, has large difference in both side, special care should be taken during cervical spine surgery. Disclosures: author 1: none; author 2: none.

QF24 COMPARISON OF CAGE WITH LOCAL BONE GRAFT AND ILIAC BONE GRAFT IN ANTERIOR CERVICAL DISCECTOMY AND FUSION(ACDF): A PROSPECTIVE STUDY Zhili Liu, Xu Xiong, Jiaming Liu Department of Orthopedics Surgery, the First Affiliated Hospital of Nanchang University, Nanchang, 330006, China Objective: A prospective study was performed to compare the clinical and radiological results between cage filled with local bone graft and iliac bone graft in anterior cervical discectomy and fusion. Methods: A total of 60 consecutive patients who underwent ACDF were evaluated from January 2010 to January 2013. 29 patients received ACDF with a PEEK cage filled with local bone graft (local bone group) and 31patients received ACDF with iliac bone graft (iliac bone graft).The clinical data and perioperative complications of the patients in the two groups were recorded. Preoperative and postoperative radiographs were taken to calculate the ratio of interbody height to disc height, and the interbody bone fusion rate. The Japanese

S721 Orthopedic Association (JOA) score and visual analogue scale (VAS) were used to estimate the postoperative functional outcome. The results between the two groups were compared. Results: The mean follow-up time was 25.0 ± 3.8 months in local bone group and 24.4 ± 3.4 months in iliac bone group (P = 0.56). Although there was no significant difference for the blood loss between the two groups during surgery (P = 0.17), the operation time was significantly less in local bone group than that of iliac bone group (P = 0.01). According to the radiographs measurement, the DSH and SIA at the surgical segment was significantly greater in local bone group than those of iliac bone group after the surgery (P = 0.00 and P = 0.00). In each group, there was significant difference between preoperative and postoperative values for VAS and JOA scores (P = 0.00 and P = 0.00). However, no significant differences were found between the two groups (P = 0.74 and P = 0.69). The incidences of perioperative complications were 6.8 % in local bone group and 29 % in group iliac bone group, respectively (P = 0.04). The fusion rates at the 24th week after surgery were 93.1 % (27/29) in local bone group and 90.3 % (28/31) in iliac bone group (P = 0.70). Conclusion: Compared with iliac bone graft, the cage with local bone graft has advantages of shorter operation time, better radiological results and lower surgical complications. We believe that it is an ideal alternative bone graft for ACDF. Disclosures: author 1: none; author 2: no indicationl; author 3: none.

QF25 THE DURATION OF SYMPTOMS AND CLINICAL OUTCOMES IN PATIENTS UNDERGOING ANTERIOR CERVICAL DISCECTOMY AND FUSION FOR DEGENERATIVE DISC DISEASE AND RADICULOPATHY Alan Villavicencio, Sigita Burneikiene, Lee Nelson, Alexander Mason, Sharad Rajpal Boulder Neurosurgical Associates, Justin Parker Neurologic Institute, Boulder, CO, USA Introduction: There have been controversial reports published in the literature on the duration of symptoms (DOS) and clinical outcome correlation in patients undergoing ACDF for painful degenerative disc disease and radiculopathy. The primary purpose of this study was to analyze if the DOS has any effect on clinical outcomes. Methods: A post hoc analysis was performed on an original prospective clinical study analyzing clinical outcomes and cervical sagittal alignment correlations. Fifty-eight patients undergoing oneor two-level ACDF surgeries for cervical degenerative radiculopathy were analyzed. Standardized questionnaires were used to evaluate clinical outcomes. Neck and arm pain was evaluated using VAS (Visual Analog Scale). Two scales of Health-Related Quality of Life Questionnaire (SF-36) were utilized for this study: the Physical Component Summary (PCS) and Mental Component Summary (MCS). Neck Disability Index (NDI) was used to evaluate chronic disability in activities of daily living. The patients completed a selfreported Patient Satisfaction with Results Survey. Patients who had previous or redo surgeries, were diagnosed with myelopathy or had more than two-level ACDF surgeries were excluded, leaving a total of 58 patients. The mean follow-up was 37.2 months (range, 12-54). Patients were divided into two groups for clinical outcome analyses according to the DOS: patients who had surgery within 6 months (n = 29) or more than 6 months (n = 29) after becoming symptomatic. Results: There were no statistically significant differences in any demographic or clinical parameters among the patient groups. Controlling for preoperative scores, the patients who had surgery within 6

123

S722 months reported significantly higher reduction (p = 0.04) in arm pain scores compared to the patients who waited longer than 6 months. No significant differences were detected in postoperative neck pain VAS (p = 0.3), NDI (p = 0.06), SF-36 PCS (p = 0.08) and MCS (p = 0.8) scores. Conclusions: Neck and upper extremity pain can be successfully treated conservatively. In those cases when surgical intervention is pursued, patients with shorter DOS have better improvement in radiculopathy symptoms that is statistically significant. Disclosures: author 1: none; author 2: none; author 5: consultant; Company = Medtronic, Biomet, LDR.

Eur Spine J (2015) 24 (Suppl 6):S711–S742 Disclosures: author 1: none; author 2: none; author 3: none; author 4: none.

QF27 DOES THE ADDITION OF CLONIDINE TO LUMBAR NERVE ROOT BLOCKS IMPROVE OUTCOME. A RANDOMISED, PROSPECTIVE, SINGLE BLINDED CONTROLLED PILOT STUDY Patrick Hourigan, Helen Challinor, Sarah Whitehouse, Andrew Clarke

THORACO-LUMBAR SPINE, DEGENERATIVE Royal Devon and Exeter Hospital, Exeter, UK QF26 SHORT FORM OF THE DISTRESS AND RISK ASSESSMENT METHOD: DEVELOPMENT AND VALIDATION Aron Lazary, Julia Szita, Zsolt Szo¨ve´rfi, Peter Pal Varga National Center for Spinal Disorders, Budapest, Hungary Background: Distress and Risk Assessment Method (DRAM) is a tool for measuring the psychological risk factors in low back pain patients. In our previous study, DRAM scale, its categories as well as its components (Zung Depression Scale (ZDS) and Modified Somatic Perception Questionnaire (MSPQ)) proved to be significant predictors for global treatment outcome after routine lumbar degenerative surgeries. On the other hand, the original DRAM has to be calculated from 45 items making that a long and clinically impractical tool. Aim: The aim of the study was to develop a short form of the DRAM (sDRAM) and to determine the prognostic power of the new version in terms of the global treatment outcome. Methods: Classic Test Theory were applied for item reduction after the factor analysis of ZDS and MSPQ on the data of a large (N = 1837) cohort of surgically treated low back pain patients. Internal consistency and the correlation of the shortened DRAM (sDRAM) with the original scale were analyzed. Two-step cluster analysis of sDRAM was applied to determine the cut-off scores. The prognostic value of sDRAM was tested on a prospective cohort of routine lumbar degenerative surgeries (N = 445). Results: ZDS had a five-factor structure while MSPQ showed three factors. In each factor, the most important item has been determined by item-score correlation and reliability testing. The sDRAM consisted of five ZDS items and three MSPQ items. sDRAM scale was highly correlated with the DRAM scale (r = 0.88). Cluster analysis resulted in a four-cluster structure of the cohort similarly to the original questionnaire. The sDRAM scale showed a significant association with the global treatment outcome in a logistic regression model (p \ 0.001, OR = 1.3) and in ROC analysis (AUC = 0.744, p \ 0.001) and it was similar to the original DRAM (p \ 0.001, OR = 1.07 as well as AUC = 0.72, p \ 0.001). Risk for poor outcome was significantly associated with the categories of sDRAM (6 %, 20 %, 39 % of poor outcome in ‘‘Normal’’, ‘‘At Risk’’ and ‘‘Distressed’’ subgroups, p \ 0.001). Conclusions: A shorter, clinically applicable method for the assessment of distress has been developed. The sDRAM consists of 8 items and correlates with the original questionnaire. sDRAM and its categories proved to be significantly associated with global treatment outcome. The tool can be used for preoperative risk assessment and reduction strategies, however, further validation is needed to show its utility in different, wider patient-groups.

123

Objectives/Purpose: To evaluate if adding clonidine to a standard nerve root block containing local anaesthetic and steroid improved the outcome for patients with severe lumbar nerve root pain secondary to MRI proven lumbar disc prolapse. Methods: A single blind, prospective, randomised controlled trial was undertaken at our institution between July 2011 and July 2014. 100 consecutive patients undergoing trans-foraminal epidural steroid injection for nerve root pain secondary to lumbar disc prolapse were randomised to receive the injection either with or without the addition of clonidine. 50 patients were allocated to each arm of the study by block randomisation. The primary outcome measure used was the avoidance of the need for a second procedure- either a repeat injection or micro-discectomy surgery. Secondary outcome measures were also studied: pain scores for leg and back pain using a visual analogue scale (VAS), the Roland Morris Disability Questionnaire (RMDQ) and the Measure Your Own Medical Outcome Profile (MYMOP). Follow up was carried out at 6 weeks, 6 months and 1 year. Results: No patient was lost to follow up. No serious complications occurred. Female to male ratio was 57: 43 with a mean age of 42 (range 21-64). The groups were comparable for age, sex, and median duration of symptoms. 55 were right sided injections, and 45 left. Of the 50 patients receiving the addition of clonidine, 56 % were classified as successful injections, with no further intervention required. Of the 50 cases receiving a standard injection, only 40 % did not need to proceed to further intervention. This difference did not reach statistical significance (p = 0.109, chi-squared test). All other secondary measures showed no statistically significant differences between the groups. Conclusions: This pilot study has shown a 16 % treatment effect with adding clonidine to lumbar nerve root blocks. In addition, it has shown that clonidine is a safe injectate for lumbar nerve root blocks. Disclosures: author 1: grants/research support; Company = Royal Devon and Exeter Hospital; author 2: grants/research support; Company = Royal Devon & Exeter Hopsital; author 3: consultant; Company = Spinal Research Foundation, RD&E, Exeter; author 4: grants/research support; Company = Royal Devon and Exeter Hospital.

QF28 PREOPERATIVE FIBRINOGEN PLASMA CONCENTRATION BUT NOT FACTOR XIII ACTIVITY PREDICTS BLEEDING IN LUMBAR SPINAL FUSION PATIENTS Malin Carling, Anders Jeppsson, Bengt Eriksson, Helena Brisby Dept. of orthopaedics, Institute of clinical sciences, Sahlgrenska Academy, University of Gothenburg, Sweden

Eur Spine J (2015) 24 (Suppl 6):S711–S742 Background: Spine fusion surgery involves a calculated risk of bleeding and blood transfusions. In order to minimize perioperative bleeding and subsequent transfusions it is important to understand and optimize the perioperative haemostasis. Fibrinogen and factor XIII are two coagulation factors that have gained interest in the field to predict bleeding complications. In adolescent scoliosis patients preoperative fibrinogen plasma concentration has been demonstrated to be a relatively good predictor of excessive bleeding and transfusion requirements. Aim: To investigate the association between preoperative fibrinogen plasma concentration and factor XIII activity on bleeding and transfusions in adult patients undergoing spinal fusion procedures. Patients and Methods: Fifty-three patients undergoing spine fusion surgery of two or more levels were included. Blood samples were preoperatively and analyzed for fibrinogen concentration and FXIII activity. Perioperative bleeding, allogeneic transfusions and autologous transfusion volumes of cell saver-processed blood were recorded. Large transfusion was defined as [2 units of red blood cell concentrate (RBC) or [500 ml of autologous transfusion. The associations between preoperative fibrinogen concentration and factor XIII activity, and intraoperative and postoperative bleeding as well as transfusion requirements, were assessed. Results: Mean fibrinogen concentration was 3.0 ± 0.8 (1.7-4.9) g/L and mean FXIII activity 1.06 ±0.23 (0.65-1.52) kIU/L. Median bleeding volume was 1690 (25th and 75th percentile 1715-2530) mL. Thirty-five patients (67 %) received RBC transfusion and 33 (64 %) autologous transfusion. Twenty-five patients (48 %) were excessively transfused. Patients with a preoperative fibrinogen \2.5 g/L (n = 15) had significantly larger median perioperative bleeding than patients with [2.5 g/L (2430 ml (1910-3150) vs 1390 ml (940-2100), p = 0.029). The proportion of patients with perioperative bleeding [2000 ml was higher in patients with fibrinogen concentration B2.5 g/L (61 vs 32 %, p = 0.046). No associations between factor XIII activity and bleeding, or between fibrinogen concentration or FXIII activity and transfusion rate, were observed. Conclusion: Fibrinogen plasma concentration of \2.5 g/L is associated with increased postoperative bleeding after spinal fusion surgery. Assessment of fibrinogen concentration preoperatively can be used as a tool to identify patient at increased risk for bleeding. Disclosures: author 1: none; author 2: grants/research support; Company = CSL Behring, consultant; Company = CSL Behring; author 3: none; author 4: none.

QF29 TIMING OF TREATMENT OF CAUDA EQUINA SYNDROME AT A NATIONAL TREATMENT CENTRE Daryl Blades, Gavin Heyes, Katie Robinson, Niall Eames Royal Victoria Hospital Belfast Cauda Equina Syndrome (CES) is an orthopaedic emergency not to be missed. However, the timing of surgery is still up for debate. It may in part be due to inconsistent reporting of both cauda equina syndrome incomplete (CESI) and those presenting with urinary symptoms (CESU). This study evaluates the patient demographics and outcomes for patients referred with possible CES treated in a tertiary referral unit. Particular attention will be paid to the influence of surgical timing. Materials and Methods: This study was performed at Royal Victoria Hospital Belfast, a single unit providing emergency spinal surgery for Northern Ireland. All referrals made to the unit between 2008 and 2014 were reviewed. Data from this study was collected using a

S723 prospective outcomes database and by reviewing all patient outpatient notes. Data recorded included age, gender, presenting complaint, radiological confirmation of cauda equina compression, timing of surgery and resolution of symptoms. Results: 344 patients were referred with possible CES. Of the 344, 205 had no CE compression on imaging, 3 had metastatic cord compression and 137 had CE compression. Presenting complaints more commonly found in those with CES were; Unilateral leg symptoms, pain, weakness, urinary and bowel symptoms. Presenting complaints more commonly found in those without radiological CES were; bilateral leg symptoms, leg paraesthesia and saddle paraesthesia. Combined results for those with CES and operated on \48 hours included significant improvement in resolution of paraesthesia and chronic back pain. Surgery on those with CESI had a significant improvement in chronic back and leg pain compared to CESU (p0.03). CESI groups demonstrated a significant improvement in paraesthesia (p0.008) and chronic back pain (p0.004) if operated on \48 hours and in CESU group, paraesthesia rates were improved if operated on \48 hours (p0.007). Most importantly urinary (p0.37) and or faecal symptoms (p0.55) did not improve significantly with surgery within \48 hours compared to surgery [48 hours in CESU group. Conclusions: This study shows that paraesthesia is improved significantly if operated on early in both groups and back and leg pain is improved if operated on\48 hrs CESI. Importantly when urinary and or bowel symptoms are present, time to surgery does not appear to significantly influence outcome. However one should not wait for a CESI to become a CESU, CESI should be treated promptly. Disclosures: author 1: none; author 2: none; author 3: none; author 4: none.

QF30 TWO YEAR OUTCOMES OF PROSPECTIVE RANDOMIZED TRIAL COMPARING LUMBAR DECOMPRESSION WITH OR WITHOUT INTERLAMINAR STABILIZATION Michael Rauschmann, Matteo Bonsanto, Joerg Franke, Steffen Sola Department of Spine Surgery, Frankfurt, Germany Background: Management of lumbar spinal stenosis (LSS) is characterized by variability in treatments. Use of a modality is conditional on disease severity and the potential for under or over treatment. Purpose: Compare clinical and functional outcomes in patients with moderate to severe lumbar spinal stenosis (LSS) treated with decompression plus interlaminar stabilization (ILS) or decompression alone (DA). Methods: A clinical trial comparing ILS to DA (1:1) was conducted at 6 sites in Germany. At inclusion, subjects had MRI-confirmed LSS at \ = 2 contiguous levels between L1-L5 and a VAS Back Pain of C50mm. This administrative interim analysis includes 184 patients (87 ILS and 97 DA) due for 2-yr follow-up as of Dec 22, 2014. Responder analyses based on composite clinical success (CCS) at m24 were used to compare groups. CCS required no reoperation or steroid injection events. In patients with no events, a meaningful improvement was required, evaluated using either ODI, ZCQ, a treadmill test, or VAS pain. Treadmill success required improvement in walking time [ = 8 min or for 15 min. ZCQ required success in [ = 2 of 3 ZCQ components. VAS back and leg pain success required [ = 20 mm improvement. Analyses included using alternative CCS definitions by replacing or incrementally adding components including narcotic use at m24.

123

S724 Results: At m24, 17 (19.5 %) ILS and 31 (32.0 %) DA had either an event (reoperation or injection, p = 0.056). Among event-free subjects, 47/57 (82.5 %) ILS and 42/48 (87.5) DA achieved ZCQ success. This resulted in 63.5 % (47/74) ILS and 53.2 % (42/79) DA achieving m24 CCS (p = 0.078). ILS had less m24 narcotic use (p = 0.075). When no m24 narcotic use is required to achieve CCS, responder rates become 58.1 % and 43.0 % for ILS and DA respectively (p = 0.06). Treadmill success was 73.8 % (31/42) ILS vs 70.0 % (28/40) DA in event-free subjects. CCS using treadmill test rather than ZCQ showed success rates of 52.5 % (31/59) and 39.4 % (28/71) for ILS and DA. With no m24 narcotic use added, 47.5 % ILS vs 32.4 % DA (p = 0.078) had success. Comparable results were observed for CCS based on VAS pain and no m24 narcotic use; rates were 63.5 % vs 53.2 % using back pain and 60.8 % vs 39.2 % for improvements in either leg (p = 0.008). Safety was comparable. Conclusion: Decompression with ILS resulted in significantly better functional outcomes, lower failure rates, and less use of oral narcotics pain compared with decompression alone using the responder analysis based on CCS. Disclosures: author 1: none; author 2: none; author 4: grants/research support; Company = Baxter, Medtronic, relievant, zimmer, consultant; Company = Medacta, silony, Medtronic, exp. orthopedics, royalties; Company = Ohst; author 5: none; author 6: grants/research support; Company = Paradigm Spine, consultant; Company = Paradigm Spine, Depuy-Synthes, Aesculap, Signus, Spontech, Biomet, Medtronic.

Eur Spine J (2015) 24 (Suppl 6):S711–S742 identify the predisposing factors for IEI. Results: 21 levels (10.4 %) of 17 patients had a sign of IEI. This injury group had a significantly higher rate of females (95.2 % vs 61.7 %, p = 0.002), PEEK cage (100 % vs 82.2 %, p = 0.03), lower T-score of BMD (-2.14 ± 0.16 vs -1.2 ± 0.15, p = 0.02) and taller cage height (10.3 ± 0.33mm vs 9.7 ± 0.1mm, p = 0.04) compared with the no injury group. Multivariate analysis revealed BMD as a negative (OR: 0.5, 95 %CI: 0.27-0.93, p = 0.03), and cage height as a positive (OR: 1.84, 95 %CI: 1.07-3.17, p = 0.03) predisposing factor for IEI. Conclusion: Reduced BMD and taller cage height are significant predisposing factors for IEI of MIS-LIF. Preoperative evaluation of bone quality, intensive treatment for osteoporotic cases, and cautious choice of cage height are essential to prevent IEI. Disclosures: author 1: none; author 2: consultant; Company = Madtronic, DePuy Synthes, AOSpine, NuVasive; author 3: none; author 4: none.

QF32 RECIPROCAL CHANGES ASSESSMENT IN ADJACENT MOBILE LEVEL FOLLOWING MONOSEGMENTAR INTERBODY FUSION Enrique Vargas, Rodrigo Amaral, Luis Marchi, Leonardo Oliveira, Fernanda Fortti, Etevaldo Coutinho Instituto de Patologia da Coluna, Sa˜o Paulo, Brazil

QF31 PREDISPOSING FACTORS FOR INTRAOPERATIVE ENDPLATE INJURY OF MINIMALLY INVASIVE LATERAL INTERBODY FUSION Kotaro Satake, Tokumi Kanemura, Hidetoshi Yamaguchi, Naoki Segi Konan Kosei Hospital, Konan, Japan Introduction: Unintended Intraoperative endplate injury (IEI) and subsequent cage subsidence is one of the serious complications of minimally invasive lateral interbody fusion (MIS-LIF) for compromising both of indirect decompression of neural element and restoration of segmental lordosis. It is unknown whether it is derived from inexperienced surgical technique or a patient’s inherent problem. This study was conducted to compare the injury levels and no injury levels in a MIS-LIF series and to identify the predisposing factors for IEI. Methods: 102 consecutive patients (41 males and 61 females, age at surgery: 69.0±0.8 years, total 201 levels) underwent MIS-LIF and were supplemented with bilateral pedicle screw fixation at a single institute. The applied cages were all 18mm width. Preoperative and immediately postoperative x-rays were compared at each level and all IEIs (cage subsidence into an endplate C1mm) were identified. They were classified as the injury group. The following parameters were compared between the injury- and the no injury group; age, sex, T score for bone mineral density (BMD), surgeon’s experience (high, middle, and low volume), surgical level (thoracic, upper lumbar, or lower lumbar), cage material (titanium or PEEK), cage height and lordosis, preoperative disc height and segmental lordosis, distracted disc height (cage height-preoperative disc height), gained segmental lordosis (cage lordosis-preoperative segmental lordosis), preoperative coronal segmental angle and T12-L4 Cobb angle, and postoperative cage position in sagittal view of computed tomography multi-planar reconstruction. Uni- and multivariate analyses were performed to

123

Introduction: Lumbar spine degenerative changes tend to lead to disc space loss and lordosis reduction. Compensatory mechanisms act to maintain good sagittal alignment, but they correlate with pain, disability and the degenerative cascade. Hyperextension (increased lordosis) at adjacent levels and has been associated with adjacent level disease. The spontaneous correction of this compensatory mechanism is not well described in the literature. The aim of this study was to verify if there is a lordosis reduction in the adjacent level following a monosegmentar anterior lumbar interbody fusion (ALIF) in L5S1. Methods: Prospective enrollment in a single-center study. Inclusion criteria: stenosis, DDD, instability and/or low-grade spondylolisthesis diagnosis; stand-alone ALIF with self-locking lordotic interbody cage. Exclusion criteria: previous lumbar fusion. All analyses were done with preoperative and 3-month data. Thirty-five eligible patients versus 47 enrolled. The group had 25 females; average age of 47±15 years old. Primary outcome measures: L4L5 and L5S1 lordosis. Secondary outcomes: lumbar lordosis (LL), L5 takeoff angle (L5ta) and sacral slope (SS). Analysis utilized preoperative and 3-month lateral radiographs in orthostatic neutral position. Two authors reviewed the images. Alpha was set to 0.05. Results: The change of LL was not significant: 52° versus 53° (p = 0.461). It was observed correction of L5S1 lordosis, average of 19° in preoperative versus 28° in 3 months (p \ 0.001), a decrease in L4L5 angle, 17° versus 14° (p \ 0.001), reduction in L5ta, 17° to 13 ° (p = 0.007), and increase in SS, 37° versus to 40° (p = 0.002). Before ALIF, the L5S1 contribution to LL was average 37 %, and after 54 % (p \ 0.001), and the angle in L4L5 contributed 34 % and decreased to 24 % (p \ 0.001). Conclusions: Following lordosis restoration with ALIF in L5S1 it was observed a reciprocal change in the cranial level as a slight decrease in lordosis. Disclosures: author 1: none; author 2: consultant; Company = NuVasive, Inc; author 3: none; author 4: none; author 5: none; author 6: none.

Eur Spine J (2015) 24 (Suppl 6):S711–S742 QF33 RHBMP-2-INDUCED RADICULITIS IN TRANSFORAMINAL LUMBAR FUSION PATIENTS: RELATIONSHIP TO DOSE Alan Villavicencio, Sigita Burneikiene Boulder Neurosurgical Associates, Justin Parker Neurologic Institute, Boulder, CO, USA Introduction: The off-label use of rhBMP-2 is exponentially increasing in spine surgery procedures along with the reports on complications. The early inflammation reaction to rhBMP-2 manifesting with radicular symptoms has been previously reported in patients undergoing TLIF. There is a disagreement in regards to the factors affecting its occurrence and whether such symptoms are dosedependent. The purpose of this study was to determine the incidence of rhBMP-2 induced radiculitis and its relationship to dose. Methods: A retrospective cohort analysis was performed of prospectively collected data. A total of 204 consecutive patients who underwent one- or two-level TLIF and instrumented posterolateral fusion with an off-label rhBMP-2 use were included in this analysis. The patients that developed radiculitis in the immediate postoperative period were believed to have rhBMP-2-induced radiculitis in the absence of any other morphological abnormalities. Primary outcome measure was radicular symptoms identified in the immediate postoperative period. Correlations between the total rhBMP-2 dose, dose per spinal level and incidence of radiculitis evaluated. Results: The incidence of postoperative radiculitis was 11.3 % (23 out of 204) and the average rhBMP-2 total dose was 4.9 mg (range 2.1-12) and 3.8 mg (range 1.05-12) per spinal level. Logistic regression analysis did not identify any significant correlations between the rhBMP-2 doses and the incidence of radiculitis (p = 0.6). Conclusions: The incidence of rhBMP-2-induced radiculitis in patients undergoing TLIF is quite high, but there were no dose-related correlations found. Disclosures: author 1: none; author 2: none.

QF34 AN OBSERVATIONAL STUDY ON THE OUTCOME AFTER SURGERY FOR LUMBAR DISC HERNIATION IN ADOLESCENTS COMPARED TO ADULTS BASED ON THE SWEDISH SPINE REGISTER. Tobias Lagerba¨ck, Peter Elkan, Hans Mo¨ller, Anna Grauers, Elias Diarbakerli, Paul Gerdhem Karolinska Institutet and Karolinska University Hospital, Stockholm Sweden Introduction: Disc-related sciatica has a prevalence of about 2 % in adults, but is rare in adolescents. If conservative treatment is unsuccessful, surgery is an option. The aim was to compare the outcomes of surgery for lumbar disc herniation in adolescents with adults in the SweSpine Register. Methods: This is a prospective observational study in a National Quality Register. This study included 151 patients, aged 18 years or younger, 4, 386 patients aged 19-39 years and 6, 078 patients aged 40 years or older, followed for 1-2 years after surgery. The primary outcomes were patient satisfaction and global assessment of back and leg pain. Secondary outcomes were visual analog scale (VAS) back pain, VAS leg pain, Oswestry Disability Index (ODI), and

S725 EuroQol 5-Dimensions (EQ-5D). Statistical analyses were performed with the Welch F-test, the Chi-square test, and the Wilcoxon test. Results: At follow-up, 86 % of the adolescents were satisfied compared to 78 % in the younger adults and 76 % in the older adults group (p \ 0.001). According to the global assessment, significantly decreased leg pain was experienced by 87 % of the adolescents, 78 % of the younger adults and 71 % of the older adults (p \ 0.001). Corresponding figures for back pain were 88 %, 73 % and 70 %, respectively (p \ 0.001). All groups experienced significant postoperative improvement of VAS leg pain, VAS back pain, ODI and EQ-5D (all p \ 0.001). Conclusions: The adolescent age group was more satisfied with the treatment than the adult groups. There was a significant improvement in all age groups after surgery. Disclosures: author 1: none; author 2: none; author 3: none; author 4: none; author 5: none; author 6: none.

QF35 CLASSIFICATION OF THE SAGITTAL PLANE BY INTEGRATING ALIGNMENT, BALANCE AND COMPENSATION Dominique A Rothenfluh, Jeremy Reynolds, Dennis Dominguez Oxford University Hospitals NHS Trust, University of Oxford, United Kingdom Introduction: We present a novel classification of the sagittal plane based on regression models of measurements of lumbosacral alignment and pelvic retroversion against sagittal balance.While sagittal imbalance is a well-recognised sagittal plane deformity, malalignment and retroversion are found in the degenerative spine also in the absence of sagittal imbalance which need to be recognised. The SRSSchwab classification of adult deformity uses sagittal modifiers to describe the sagittal plane component of the deformity whereas the Roussouly types focuses on the shape of the spine. Here, we present a novel classification of the sagittal plane which places itself between the two established systems. Methods: 161 whole spine x-rays were analysed of patients with manifest adult deformity including sagittal imbalance and (n = 87) and of patients with degenerative lumbar spines without scoliosis [10deg. and no sagittal imbalance (n = 74). For classification, logistic regression and ROC analysis against dichotomous nominal dependants was carried out. Sagittal imbalance was defined at a cutoff of 47mm based on earlier reports linking it to an ODI[40 %. Results: Logistic regression and ROC analysis revealed a cut-off of 10.9° for pelvic incidence-lumbar lordosis (PILL) mismatch predicting sagittal imbalance (AUC 0.83, sens 0.92, spec 0.59, OR 15.8). The balanced group was further classified in a compensated vs. noncompensated group by calculating a cut-off for pelvic tilt (PT) of 25.1° (AUC 0.77, sens 0.88, spec 0.71). The balanced and compensated group was classified according to the PILL criterion into an aligned and malaligned group. Limits of agreement of measurements and reliability measures were calculated. Conclusion: The above calculations support the classification of the sagittal plane and connect the degenerative spine to sagittal imbalance. While type A describes a balanced and aligned spine, type B1 exhibits lumbo-sacral malalignment (PILL [10°). Type B2 is a malaligned (PILL[10°) spine and compensating (PT[25°) to maintain the balance. Type C is the sagittally imbalanced spine (SVA [47mm). This classification is useful to understand the relationship of the intrinsic to the global deformity and can guide surgical

123

S726 decision-making by identifying the key aspects of the sagittal deformity which have been shown to have biomechanical consequences. Disclosures: author 1: none; author 2: none; author 3: none.

QF36 THE RELATIONSHIPS BETWEEN THE SRS-SCHWAB CLASSIFICATION MODIFIERS AND THE ODI SCORE ARE ROBUST TO CONFOUNDING Dennis Hallager Nielsen, Lars Valentin Hansen, Casper Rokkjær Dragsted, Nina Caroline Peytz, Martin Gerhchen, Benny Dahl Spine Unit, Department of Orthopaedic Surgery, Rigshospitalet, University of Copenhagen, Copenhagen, Denmark Introduction: The Scoliosis Research Society (SRS)-Schwab Adult Spinal Deformity Classification includes sagittal modifiers considered important for health related quality of life (HRQOL): Pelvic Tilt (PT), Pelvic Incidence minus Lumbar Lordosis (PI-LL) and Sagittal Vertical Axis (SVA). Each modifier is graded 0, + or ++ with increasing abnormality. The clinical impact of the classification has been evaluated in patients drawn from the American International Spine Study Group database, however confounding has not been evaluated, although age and aetiology of spinal deformity have been suggested as clinical relevant. Objective: To evaluate confounding effects of age, aetiology of spinal deformity and history of spine surgery on the relationship between the SRS-Schwab sagittal modifiers and the ODI score. Methods: Between March 2013 and May 2014 all patients at least 18 years of age with sufficient radiographic images taken at one outpatient clinic to evaluate a spinal deformity were prospectively enrolled. Exclusion criteria were deformity surgery within 6 months or missing ODI questionnaires. History of spine surgery was recorded. Patients were classified as young (18-44), middle aged (45-64) or old (65+). Aetiology of spinal deformity included degenerative, AIS, Ankylosing Spondylitis (AS), congenital, neuromuscular, Scheuermann’s disease and secondary causes. These potential confounding effects were evaluated with univariate and three multivariate proportional odds regressions, one for each of the modifiers (PT, SVA and PI-LL). Outcome was ODI scores aggregated into 5 equidistant levels. Results: 286 of 460 eligible patients were included. 30 had deformity surgery within 6 months and 144 did not return the questionnaire. From three multivariate analyses we found increased proportional odds for worse ODI scores with each step increase in SRS-Schwab modifier grades (Odds Ratio (OR) range: 1.4-1.7), increased odds for middle aged (OR 2.7-2.9) or old (OR 2.3-2.5) vs. young patients, and lower odds for AS (OR 0.1) or congenital (OR 0.2) vs. degenerative deformities. All effects were statistically significant (P B 0.032). Effects of previous spine surgery and AIS vs. degenerative deformity were significant in univariate analyses. Conclusion: The SRS-Schwab modifiers have robust associations with the ODI score when adjusted for potential confounding variables. However, patient age and aetiology of spinal deformity have significant associations as well. Disclosures: author 1: grants/research support; Company = Globus Medical Inc. author 2: none; author 3: none; author 4: none; author 5: grants/research support; Company = Medtronic, Globus Medical, K2M, consultant; Company = Medtronic, Globus Medical, K2M; author 6: grants/research support; Company = Medtronic, Globus Medical, K2M.

Eur Spine J (2015) 24 (Suppl 6):S711–S742 QF37 SPINAL FUSION TERMINATING AT L5 IN ADULT SCOLIOSIS: FACTORS ASSOCIATED WITH SUBSEQUENT EXTENSION OF THE FUSION TO THE SACRUM Santo Creaco, Anne F. Mannion, Dezso¨ Jeszenszky, Frank S. Kleinstu¨ck, Daniel Haschtmann, Tamas F. Fekete Spine Center, Schulthess Klinik, Zu¨rich, Switzerland, School of Specialization in Orthopedic and Traumatology Surgery – University of Messina – Messina, Italy Introduction: When planning surgery for adult deformity, it is often difficult to decide whether to fuse a lumbar construct to the sacrum. We analyzed the outcomes of patients who underwent fusion stopping at L5 for adult scoliosis (defined as C 20° coronal Cobb angle of the lumbar spine), and evaluated risk factors associated with subsequent caudal extension of the fusion. Methods: We searched our local Spine Outcomes database of prospectively collected data to identify patients who had undergone fusion terminating at L5. We identified 56 patients (43 F, 13 M; mean age, 64.9 (23-84) y) who were at least 36 mo postoperative (ave 78 (38-117) mo). Radiographs taken before and after surgery and at the latest follow up were evaluated. Risk factors for subsequent extension of the fusion (i.e., surgical procedures and radiographic/sagittal balance parameters) were evaluated using multivariable analyses. Patients had completed the Core Outcome Measures Index (COMI) preop and at 3, 12 and 24 months postop. Results: 45 patients underwent fusion without osteotomy, while 11 had different types of osteotomy. There were 22 short fusions (3 segments); 21, medium (4-6 segments); and 13, long (C7 segments). 43 patients had circumferential fusion at L4/L5. 20/56 patients underwent subsequent surgery, with 15 of these involving extension to the sacrum (EXT) (after a mean 36.1 mo). Proportional Hazards multiple logistic regression showed that the odds of undergoing EXT were significantly greater for long fusion (p = 0.007), lower preop lordosis (p = 0.012) and no osteotomy (p = 0.038) and showed a tendency to be greater with interbody fusion at L4/5 (p = 0.107). Although no other baseline radiographic parameters (Cobb, SVA, SS, PI, PT, L5/S1-coronal tilt) were associated with subsequent EXT, the LL was significantly lower and L5/S1 coronal tilt significantly greater at 3mo postop in the EXT group (p \ 0.05). There was a trend (p = 0.19) towards a different pattern of change in COMI scores over the course of follow-up: both groups reduced to a similar extent up to 1y FU, but the EXT group worsened between 1 and 2 yr FU while the non-revised group remained stable. Conclusion: Long fusion, interbody fusion at L4/5, no osteotomy procedures and lower preoperative lordosis all appeared to be risk factors for revision extending to the sacrum. These factors may lead to a greater biomechanical stress and subsequent degeneration of the L5/S1 segment, necessitating later revision. Disclosures:: author 1: none; author 2: none; author 3: consultant; Company = DePuySynthes, royalties; Company = DePuySynthes; author 4: none; author 5: none; author 6: none.

QF38 RISK FACTORS ASSOCIATED WITH ROD FRACTURE AFTER OSTEOTOMY FOR ADULT SPINAL DEFORMITY Evalina Burger, Cameron Barton, Andriy Noshchenko, Vikas Patel, Christopher Cain, Christopher Kleck University of Colorado, Anschutz Medical Campus, Aurora, USA

123

Eur Spine J (2015) 24 (Suppl 6):S711–S742 Introduction: Osteotomies including pedicle subtraction (PSO) and Smith-Peterson (SPO) are widely used to facilitate correction in the treatment of adult spinal deformity (ASD), but are associated with complications including instrumentation failure and rod fracture (RF). No study to date has analyzed a comprehensive list of variables to determine risk factors for RF in a population of ASD patients treated with osteotomy. Purpose: The purpose of this study is to determine incidence and risk factors for rod fracture after osteotomy for adult spinal deformity. Design: Retrospective review of adult deformity database (COMIRB #14-1258). Methods: Eighty consecutive ASD patients (54F, average age 59) treated with osteotomy met strict inclusion/exclusion criteria and follow-up of at least 1 year. Data was extracted including patient variables (e.g. age, gender), surgical variables (e.g. levels fused, osteotomy type and location), instrumentation variables (e.g. rod and screw characteristics), and postoperative variables (e.g. spinopelvic parameters, rod fracture characteristics). Patients were divided into two groups (rod fracture or non-rod fracture) and odds ratios were calculated to assess risk factors for RF. Results: Incidence of RF was 6.3 % (5/80) in total population, 9.3 % (4/43) for PSO group vs 2.7 % (1/37) for SPO only group (OR = 3.44, P = .49). Risk factors for RF included sagittal rod contour[60 degrees (OR = 10.67, P = .05), presence of pelvic fixation (OR = 14.85, P = .02), presence of dominos/cross connectors at time of fracture (OR = 9.17, P = .04), and pseudarthrosis in follow-up period (OR = 84.00, P = .0005). Individual rod fractures showed evidence of repeated rod bending and fractures at current or prior connectors. Conclusions: Current study shows total RF incidence of 6.3 % and trends higher in PSO vs. SPO. Risk factors for RF meeting statistical significance included pseudarthrosis, sagittal rod contour [60 degrees, presence of dominos/cross connections at time of fracture, and presence of pelvic fixation. Disclosures: author 1: grants/research support; Company = Aesculap, consultant; Company = Medicrea, Paradigm, Signus; author 2: grants/research support; Company = Medicrea; author 3: employee; Company = Colodado University Denver; author 4: grants/research support; Company = Medtronic, MTF, OMeGA, SI-Bone, Aesculap, Vertiflex, consultant; Company = Aesculap, Allosource, royalties; Company = Biomet, Aesculap, Springer, SLACK; author 5: grants/ research support; Company = Medicrea, Medtronic, SI Bone, Vertiflex, Aesculap, consultant; Company = DePuy Synthes, royalties; Company = DePuy Synthes; author 6: grants/research support; Company = Medtronic Sofamor-Danek, Integra, Orthofix, Aesculap, Medicrea, Vertiflex.

QF39 PROSPECTIVE MULTICENTER ASSESSMENT OF INTRAOPERATIVE AND PERIOPERATIVE COMPLICATION RATES ASSOCIATED WITH ADULT SPINAL DEFORMITY (ASD) SURGERY IN 558 PATIENTS Justin Smith, Eric Klineberg, Frank Schwab, Christopher Shaffrey, Shay Bess, Gregory Mundis, Han Jo Kim, Justin Scheer, Christopher Ames International Spine Study Group Hypothesis: Early complication rates associated with ASD surgery are more accurately assessed with rigorous prospective documentation. Design: Prospective multicenter cohort study of consecutive ASD patients treated surgically. Introduction: Few previous studies have focused on early (intraop and periop) complication rates for ASD surgery, with reported overall

S727 rates ranging from *10 % to 75 %. However, most available studies are relatively small single- center series and likely underestimate complication rates due to retrospective design and lack of rigorous collection of complications. Accurately defining the early complication rates is particularly important for patient counseling with regard to the risks and benefits of surgical treatment. Methods: As part of a prospective ASD database, standardized collection forms, on-site coordinators and auditing helped ensure complete capture of complications. Intraop and periop (\6 wks) complications were collected for all patients with these available data, regardless of subsequent length of available follow-up. Results: 558 patients underwent surgical treatment for ASD and had a mean age of 57 yrs, mean Charlson Comorbidity Index of 1.5 and previous surgery in 48 %. The majority (98 %) of patients had treatment including a posterior instrumented approach and included a 3-column osteotomy in 22 % of patients. A total of 171 intraop complications (69 major, 102 minor) affected 127 (22.8 %) patients, and a total of 272 periop complications (133 major, 139 minor) affected 185 (33.2 %) patients. Collectively, 443 early complications (202 major, 241 minor) were reported, with 264 (47.3 %) patients experiencing one or more complication. The overall mean numbers of intraop, periop and total complications per patient were 0.31, 0.49 and 0.79, respectively. Conclusion: ASD surgery is associated with high rates of early complications, with 47.3 % of patients having at least one complication. These rates are substantially higher than previously reported. These finding may prove useful in treatment planning and patient counseling. Disclosures: author 1: grants/research support; Company = DepuySynthes, ISSGF, consultant; Company = Biomet, Nuvasive, Cerapedics, Medtronic, royalties; Company = Biomet; author 2: grants/ research support; Company = AO Spine, Depuy, OREF, other financial report; Company = Speakers Fees Depuy AO Spine; author 3: grants/research support; Company = DePuy, AO, MSD, consultant; Company = K2M, Medicrea, MSD, stock/shareholder; Company = Nemaris INC, royalties; Company = MSD, K2M, other financial report; Company = Biomet, Nuvasive; author 4: grants/research support; Company = Depuy-Synthes, consultant; Company = Biomet, Medtronic, Nuvasive, Stryker, stock/shareholder; Company = Nuvasive, royalties; Company = Biomet, Medtronic, Nuvasive; author 5: grants/research support; Company = depuy, medtronic, k2, innovasis, stryker, consultant; Company = k2, allosource, nuvasive, royalties; Company = pioneer; author 6: grants/research support; Company = dePuy Synthes, consultant; Company = nuvasive, K2M, Medicrea, Misonix, royalties; Company = nuvasive, K2M; author 7: consultant; Company = Biomet, K2M, Medtronic, other financial report; Company = Stryker, Depuy; author 8: none; author 9: grants/research support; Company = DePuy Synthes Spine, consultant; Company = DePuy, Medtronic, Stryker, stock/shareholder; Company = Baxano Surgical, Doctor’s Research Group, royalties; Company = Aesculap, Biomet Spine, employee; Company = University California San Francisco.

QF40 EFFECT OF TREATMENT COMPLICATIONS ON THE OUTCOMES IN ADULT SPINAL DEFORMITY (ASD): A DECISION ANALYSIS (DA) APPROACH Emre Acaroglu, Umit Ozgur Guler, Aysun Cetinyurek-Yavuz, Selcen Pehlivan, Yasemin Yavuz, Montse Domingo-Sabat, Ferran Pellise, Ahmet Alanay, Francesco Sanchez Perez Grueso, Frank Kleinstu¨ck, Ibrahim Obeid, ESSG 1

Ankara Spine Center, Ankara TR; 2Clinistats EU, Zurich, CH; Hospital Universitari Vall d’Hebron, Barcelona, E; 4Acibadem

3

123

S728 Maslak Hospital, Istanbul, TR; 5Hospital Universitari La Paz, Madrid, E; 6Schultess Clinic, Zurich, CH; 7Bordeaux University Hospital, Bordeaux, F Background: Treatment of ASD is known to be associated with a fairly high rate of complications where as the impact of these complications on treatment outcomes is less well known. Aim: To analyse the impact of treatment complications on outcomes in ASD using a decision analysis model. Methods: From an international multicentre database of ASD patients (968 pts), 535 who had completed 1 year follow-up (371 non-surgical -NS), 164 surgical -S), constitute the population of this study. DA was structured in two main steps of: 1) Baseline analysis (Assessing the probabilities of outcomes, Assessing the values of preference -utilities-, Combining information on probability and utility and assigning the quality adjusted life expectancy (QALE) for each treatment) and 2) Sensitivity analysis. Complications were analysed as life threatening (LT) and non-life threatening (NLT) and their probabilities were calculated from the database as well as a thorough literature review. Outcomes were analysed as improvement (decrease in ODI[8pts), no change and deterioration (increase in ODI[8pts). Death/complete paralysis was considered as a separate category. Results: All 535 patients (371 NS, 164 S) could be analysed in regard to complications. Overall, there were 78 NLT and 12 LT complications and 3 death/paralysis. Surgical treatment was significantly more prone to complications (31.7 % vs. 11.1 %, p \ 0.001). On the other hand, presence of complications did not necessarily decrease the chances of improvement and surgical patients tended to rate better despite higher complications (improvement rates for S group: 37.5 % for LT, 44.4 % for NLT and 58.8 % for none; for NS group: N/A for LT, 17.9 % for NLT and 13.8 % for none). Likewise, QALE was not particularly affected by the presence or absence of complications regardless of the type of treatment. QALE ranged between 61 to 63 in the S group vs. 61 to 62 in the NS in patients with favorable outcomes; 49 to 65 in the S group vs. 68 to 70 in the NS in patients with intermediate outcomes; and 38 to 58 in the S group and 40 to 75 in the NS in patients with unfavourable outcomes. Discussion and Conclusions: This study has demonstrated that surgical treatment of ASD is more likely to cause complications compared to non-surgical treatment. On the other hand, presence of complications neither has a significantly negative impact on the likelihood of clinical improvement nor affects the QALE at the 1st year detrimentally. Disclosures: author 1: grants/research support; Company = Medtronic, Depuy Synthes, stock/shareholder; Company = IncredX; author 2: grants/research support; Company = Eurospine; author 3: none; author 4: consultant; Company = Yildirim Beyazit University; author 5: none; author 6: grants/research support; Company = Depuy-synthes; author 7: grants/research support; Company = DePuy Synthes, K2M, consultant; Company = DePuy Synthes, Biomet; author 8: grants/research support; Company = Depuy Synthes, consultant; Company = Stryker; author 9: grants/research support; Company = DePuy Synthes, consultant; Company = DePuy syntes; author 10: grants/research support; Company = DePuy-synthes.

GROWING SPINE

QF41 PROVIDENCE NIGHTTIME BRACING, IN TREATMENT OF ADOLESCENT IDIOPATHIC SCOLIOSIS. Ane Simony, Inge Beuschau, Lena Quisth, Mikkel O. Andersen

123

Eur Spine J (2015) 24 (Suppl 6):S711–S742 Middelfart Hospital, Middelfart, Denmark Introduction: Since 2008 the non-surgical treatment of adolescent idiopathic scoliosis (AIS) in the southern part of Denmark, went from full-time bracing with Boston brace, to Providence night-time bracing. Methods: Since 2008, skeletally immature patients diagnosed with AIS and a primary curve with apex at T7 or below, were treated with the Providence night-time brace. The patients were evaluated every 6 months, with standing x-rays. Compliance with the treatment was recorded during the study. Bracing was continued up to two years post menarche or, for males upon reaching the expected adult height. Cobb angle of the major curve pre-treatment, in-brace correction and at last follow-up were determined. Curve progression of more than 5 degrees and need for surgical intervention were considered brace failures. The patients received the SRS22r questionnaire at least 8 months after termination of brace treatment. Results: A total of 62 patients, 10 males and 52 females, mean age of 14 years with a mean Cobb of 31.3° degrees were included. There were 31 primary thoracic curves, 19 thoraco-lumbar curves, 11 lumbar curves and 1 double curve. The mean length of brace treatment was 17.8 months (6-59 months). The mean in-brace correction was 81 % (24-100 %). After the end of bracing, the mean Cobb angle was 29.1° (7-50), an average of no progression. There were 11 (18 %) brace failures, but only 5 (10 %) patients had surgery. All patients returned the SRS22 r questionnaire, and the follow up was 12 months (8-37m). The SRS22r Function score was 4, 32, Mental Health 3, 81, SRS22r total 3, 9 and satisfaction with management 3, 90. The Pain domain was 3, 8. Conclusion: This study shows a good curve control with the Providence Night time brace and an acceptable 20 % curve progression rate. The night time brace is an excellent alternative to standard conservative treatment. The patients tolerate the night-time brace treatment well, and compliance is high. Function, Self-image, Mental Health, and SRS 22r total are similar to surgically treated patients and health controls, but it seems like patients treated with Providence braces are more likely to experience back pain than healthy adolescents and surgically treated scoliosis patients. Disclosures: author 1: none; author 2: none; author 3: none; author 4: none.

QF42 THE PREVALENCE OF INTRASPINAL ANOMALIES IN INFANTILE AND JUVENILE PATIENTS WITH PRESUMED IDIOPATHIC SCOLIOSIS: A MRI-BASED ANALYSIS OF 504 PATIENTS Zezhang Zhu, Wen Zhang, Zhen Liu, Xu Sun, Yong Qiu Spine Surgery, the Affiliated Drum Tower Hospital of Nanjing University Medical School, Nanjing, 210008, China Background: Though several published studies have reported the incidence of intraspinal neural axis abnormalities in infantile and juvenile ‘‘presumed idiopathic’’ scoliosis, there has been a varying prevalence ranging from 11.1 % to 26.0 % based on a limited sample size. Purpose: To investigate the prevalence and distribution of intraspinal anomalies in the infantile and juvenile patients with ‘‘presumed idiopathic’’ scoliosis and to explore the radiographic and clinical indicators. Methods: A total of 504 infantile and juvenile patients diagnosed with ‘‘presumed idiopathic’’ scoliosis were examined for potentially-

Eur Spine J (2015) 24 (Suppl 6):S711–S742 existing neural axis abnormalities by MRI. Patients were grouped into two cohorts according to the presence of neural axis abnormalities. Radiographic parameters including curve magnitude, curve pattern, location of apex, degree of thoracic kyphosis, and span of curve were recorded and compared between the two groups. The prevalence of the neural abnormalities between the infantile-age group and juvenileage group was also compared. Results: 94 patients (18.7 %) were found to have a neural abnormality involving spinal cord: Arnold-Chiari malformation alone in 43 patients, Arnold-Chiari malformation combined with syringomyelia in 18 patients, isolated syringomyelia in 13 patients, diastematomyelia in 6 patients, tethered cord combined with diastematomyelia in 6 patients, tethered cord alone in 4 patients, and other uncommon intraspinal abnormalities in the remaining 4 patients. In the infantileage group, 5 patients with isolated syringomyelia, 2 patients with Arnold-Chiari malformation, 2 patients with diastematomyelia, 1 patient with tethered cord and 1 patient with Arnold-Chiari malformation and syringomyelia were observed among 71 patients. Totally Arnold-Chiari malformation with or without syringomyelia accounted for 64.8 % (61/94) among all these abnormalities. Then male gender, left thoracic curve and right lumbar curve were found to be significantly associated with the presence of neural axis abnormalities on MRI. Conclusions: The incidence of neural axis abnormalities in the presumed IIS and JIS was 18.7 %. Thus a routine MRI evaluation appears warranted for those ‘‘presumed idiopathic’’ scoliosis patients if aged less than 10 years, being male or having left thoracic or right lumbar curve. Disclosures: author 1: none; author 2: none; author 3: none; author 4: none; author 5: none.

QF43 DOES PEDICULE SCREW FIXATION UNDER AGE FIVE CAUSE SPINAL CANAL NARROWING? A CT STUDY WITH MINIMUM 5 YEARS FOLLOW-UP Sinan Kahraman, Meric Enercan, Mutlu Cobanoglu, Levent Ulusoy, Ayhan Mutlu, Tunay Sanli, Bahadir Gokcen, Erden Erturer, Cagatay Ozturk, Azmi Hamzaoglu Istanbul Spine Center, Turkey Summary: Pedicle screw instrumentation has no adverse effect on further spinal growth and does not result in iatrogenic spinal stenosis. Design: Retrospective. Introduction:The influence of pedicle screw fixation on canal diameter below age 5 is controversial. Animal studies demonstrated development of canal stenosis after pedicle screw fixation. In contrast to this results, clinical studies demostrating no spinal stenosis after pedicle screw fixation has been published. The aim of this retrospective study was to evaluate the changes in the canal area in a group of patients who had pedicle screw fixation under age 5 for the treatment of spinal deformity at least 5 year follow-up. Methods:11 patients who had been operated due to spinal deformity under age 5 with who had a CT examination due to several reasons at least 5 years after the initial spinal operation were included in the study. All patients had congenital scoliosis and underwent hemivertebrectomy and transpedicular fixation procedures at an average age of 3.18 (range; 2 to 5). All had preoperative CT to evaluate the congenital deformities. Measurements were done at the instrumented vertebrae as well as the uninstrumented ones above and below them to evaluate; anterior vertebral body height (AVBH), posterior vertebral body height (PVBH), cranial end plate length (CrEPL), caudal end

S729 plate length (CaEPL), spinal canal area (SCA), anteroposterior diameter of vertebral body (APD) and lateral diameter of vertebral body(LD) of upper instrumented vertebra (UIV), lower instrumented vertebra (LIV), upper adjacent uninstrumented vertebra(UAV) and lower adjacent uninstrumented vertebra (LAV). Results: The average follow-up was 7, 2 (range; 5 to 12) years. 6 of the patients were over age 10 during the final CT examination while 5 were at age 7. Female to male ratio was 7 to 4. Measurement of all the parameters in 22 instrumented and 22 non-instrumented segments showed a proportional increase rather than a decrease at each segment (Figure 1). The percentage of canal area growth at UIV and LIV were 21 % and 17.5 % respectively. Conclusion: Pedicle screw instrumentation has no adverse effect on further spinal growth and does not result in iatrogenic spinal canal stenosis. Disclosures: author 1: none; author 2: none; author 3: none; author 4: none; author 5: none; author 6: none; author 7: none; author 8: none; author 9: none; author 10: consultant; Company = Medtronic.

QF44 A SIMPLE METHOD FOR ASSESSING ROTATIONAL FLEXIBILITY IN ADOLESCENT IDIOPATHIC SCOLIOSIS: MODIFIED ADAM’S FORWARD BENDING TEST Alpaslan Senkoylu, Mustafa Ilhan, Necdet Altun, Dino Samartzis, Keith Luk Gazi University, Orthopaedics and Traumatology, Ankara, Turkey Introduction: Determination of curve flexibility is a critical parameter for the treatment plan of AIS. Numerous published reports exist that define curve flexibility in the sagittal and coronal planes. However, robust assessment of rotational flexibility by a physical examination method remains inconclusive. As such, this study attempted to define the coupling phenomenon of axial rotation with reference to the side-bending movement, further facilitated by the introduction of our curve flexibility index (CFI). Method: Forty AIS patients were assessed prospectively. In the mAFBT, without coming up to an erect position, subjects were asked for bending to the convex side of the curve in the axial plane. Evaluation of the rib/loin hump flexibility was performed during this maneuver. Scoliometric measurements were done during the AFBT and mAFBT. Utilizing plain radiographs, Cobb angles were measured and curve flexibility indices were calculated. The fulcrum and sidebending were performed to assess radiographic curve flexibility. Pearson’s correlation and ROC curves were used for statistical analysis. The CFI was used, defined as follows: CFI = [(AFBTmAFBT)/AFBT]x100. Results: Significant correlations were noted between the Cobb angle and AFBT (p = 0.005), fulcrum bending and the mAFBT (p = 0.0001), side-bending and mAFBT (p = 0, 0001), and the postoperative Cobb angles and AFBT (p = 0.003). There were significant positive correlations between curve flexibility as based on the fulcrum bend to that of the CFI (r = 0.347, p = 0.036) and side-bending technique (r = 0.416, p = 0.008). Based on ROC analyses (AUC range = 0.71-0.78), the mAFBT demonstrated high specificity and sensitivity rates for flexible and rigid curves, respectively. Conclusion: The mAFBT was found to be a reliable test for clinical assessment of rotational flexibility in AIS patients. Disclosures: author 1: none; author 2: none; author 3: none; author 4: none; author 5: grants/research support; Company = university of Hong Kong, employee; Company = university of Hong Kong.

123

S730 QF45 SLIDING-GROWING ROD TECHNIQUE (SGRT) IN THE TREATMENT OF EARLY ONSET SCOLIOSIS - MORE THAN 2 YEARS OF FOLLOW-UP Meric Enercan, Sinan Kahraman, Bahadir Gokcen, Tunay Sanli, Sinan Yilar, Erden Erturer, Cagatay Ozturk, Azmi Hamzaoglu Istanbul Spine Center, istanbul, Turkey Summary: Sliding-growing rod technique provides a dynamic fixation which obtain and maintain satisfactory curve corrections in EOS and allows self growing of the spine with a rate of 1.28 mm growth per month, decreases number of lengthening procedures and demonstrated low complication rates with improved pulmonary functions. Design: Retrospective. Introduction: The main goal of treatment in EOS is to obtain and maintain curve correction while simultaneously preserving spinal, trunk, and lung growth. This study introduces a new surgical strategy, called ‘‘sliding-growing rod technique’’ (SGRT) developed to decrease the number of lengthening procedures. The aim of this study is to assess whether self growing system works or not, determine complication rates and effects on pulmonary functions in patients who had more than 2 yrs f/up. Methods: 15 (9F/6M) pts, mean age 6.8 (5-10) were evaluated. Surgical technique included placement of pedicle screws with a muscle-sparing technique. Following rod placement and correction, the most proximal and most distal two segments were fixed and fused; the rest of the screws were left with unlocked set screws to allow vertical growth. Proximal and distal rods are connected with side to side connectors (domino) mostly at distal level. Distal rod was fixed to domino connector whereas proximal rod kept loose to allow self growing (Figure). Preop, f/up, final x-rays and pre/postop pulmonary function tests (PFT) were evaluated. Results: Mean f/up was 24.8 months (24-32). Ave preop MT curve of 61.1° was corrected to 23, 3° with a correction rate of %62.6 . Ave TL/ L curve of 43, 2° was corrected to 15, 5° with a correction rate of %68.7. Preop TK of 35, 1° and LL of 55, 3° was maintained at 29, 4° and 55, 7° respectively. Mean increase in T1-T12 length was 1.14 mm/month and 1.28 mm/month in T1-S1 height. No patient had neurological impairments. There was no rod breakages or other implant failure. This modification prevented 42 planned lengthening procedures. Mean preop %predicted FVC of 68.76 improved to 72.43 and mean preop %predicted FEV1 of 67.43 improved to 71.28 at the latest f/up. Conclusion: In contrast to traditional growing rod systems, SGRT provides a dynamic fixation which allows self growing of spine with a rate of 1.28 mm per month. SGRT demonstrated low complication rates and improved pulmonary functions at the end of 2 yrs f/up. Disclosures: author 1: none; author 2: none; author 3: none; author 4: none; author 5: none; author 6: none; author 7: none; author 8: consultant; Company = Medtronic.

QF46 MAGNETICALLY CONTROLLED EXTENSION DEVICES: DOES IT REALLY DO EXACTLY WHAT IT SAYS? Daniel Winson, Kar Teoh, Narendra Rath, Sashin Ahuja University Hospital of Wales, Cardiff, UK A discrepancy has been noted clinically between the desired extension length and the actual extension length in growth rods that use a

123

Eur Spine J (2015) 24 (Suppl 6):S711–S742 magnetically controlled device. This prospective study was designed to identify whether this perceived discrepancy exists. Patients with magnetically controlled growing rods in place who have had extensions performed in the outpatient clinic using the External Remote Controller (ERC) were monitored and outcome was measured in terms of actually extension achieved compared to the planned extension using the ultrasound measurements. Ultrasound scan was performed on the magnetic rods both pre and post extension. The distraction achieved was compared to the reading recorded on the ERC. In total 14 extensions were undertaken on 6 patients with double rod constructs. This provided 28 extension readings. The desired distraction was recorded on the ERC and this was compared to the distraction achieved on US examination. 3 separate US measurements performed and the final reading was logged as the average of the 3 readings as per the best view on US to assess the distraction of the magnet. All the recordings were normalised and the ERC reading was taken as 100 % and the actual distraction of the magnet was assessed as a percentage of the ERC reading. The range of extensions was from 0-100 % with an average of 47.16 %. Only 2 out of 28 readings (7 %) were accurate. The rest of readings on ERC overestimates the actual distraction obtained. There is no significant difference between the amount of actual distraction obtained for left and right rods (p = 0.93, Wilcoxon Signed-Rank Test). This study shows that the recordings shown on the ERC device used in clinic used for the extension of magnetically controlled growing rods is consistently misrepresenting the actual amount of distraction as per ultrasound measurements. Disclosures: author 1: none; author 2: none; author 3: none; author 4: other financial report; Company = Teaching/Education-Globus Medical.

QF47 THE EFFECT OF MAGNETICALLY CONTROLLED GROWING ROD ON THE SAGITTAL PROFILE IN EARLY-ONSET SCOLIOSIS PATIENTS Caglar Yilgor, Kenneth Cheung, Kenny Kwan, Dino Samartzis, John Ferguson, Colin Nnadi, Ilkka Helenius, Muharrem Yazici, Gokhan Demirkiran, Ahmet Alanay, Behrooz Akbarnia Acibadem University School of Medicine; The University of Hong Kong, Starship Children’s Hospital/Auckland Bone and Joint Surgery; Oxford University Hospitals; Turku University Central Hospital, Department of Paediatric Orthopaedic Surgery; Hacettepe Universit; Orthopaedic Surgery, University of California, San Diego Summary: The is a retrospective review of prospectively collected data from a multicentre study of early-onset scoliosis treated by magnetically-controlled growing rod with a minimum of 2-year follow-up. Thoracic kyphosis was reduced in patients with pre-operative [40° and the overall sagittal balance improved or returned to neutral in 60 % of cases. Hypothesis: MCGR lengthening has an effect on the regional and global sagittal profile. Design: Retrospective review of prospectively collected data. Introduction: Magnetically controlled growing rod (MCGR) has a straight central housing portion that cannot be bent. The proximal and distal portions are contoured intra-operatively according to the desired kyphosis and lordosis. The effects of gradual lengthening on the regional and overall sagittal profile in early onset scoliosis (EOS) are not been well-documented. This study aimed to report on the changes of the sagittal profile after MCGR implantation.

Eur Spine J (2015) 24 (Suppl 6):S711–S742 Methods: A retrospective review of prospectively collected data from consecutive patients undergoing MCGR treatment with a minimum of 2-year follow-up from 6 centres was carried out. Clinical data and complications were noted. Radiographic measurements including thoracic kyphosis (TK), lumbar lordosis (LL) and sagittal vertical axis (SVA) were analysed. Results: Thirty patients were reviewed and twenty-three patients had full radiographic data for analysis. The mean age at the time of surgery was 7.3 years (range: 4-14 years) and the mean follow-up period was 39.2 months (range: 24- 61 months). Patients were divided into 3 groups according to their pre-operative TK: group 1 (TK \ 20°), group 2 (TK 20°-40°) and group 3 (TK[40°). Mean TK did not change in group 1 or 2 during MCGR lengthening but decreased in group 3, and mean LL remained the same in all 3 groups. At final follow-up, global sagittal balance (SB) improved or returned to neutral alignment in 60 % of cases, and did not change in 27 %, and worsened in 13 %. Conclusion: Growth sparing techniques allow coronal curve correction in EOS but its effect on the sagittal profile is not well understood. This study showed that MCGR reduced TK in those with pre-existing TK [40° and had no effect on other regional sagittal parameters. It had a tendency to improve the global sagittal balance. Further studies are required to evaluate fully the effect of MCGR on the sagittal profile. Disclosures: author 1: none; author 2: grants/research support; Company = Ellipse Technologies; author 3: none; author 4: none; author 5: consultant; Company = K2M, Ellipse, stock/shareholder; Company = Nuvasive, K2M; author 6: royalties; Company = Ellipse Technologies; author 7: grants/research support; Company = Medtronic, Baxter, consultant; Company = Medtronic, Depuy Synthes; author 8: consultant; Company = DePuy Synthes; author 9: none; author 10: grants/research support; Company = Depuy Synthes, consultant; Company = Stryker.

QF48 ASYMMETRIC APPEARANCE IN THE OSSIFICATION CENTER OF RING APOPHYSIS IN PATIENTS WITH ADOLESCENT IDIOPATHIC SCOLIOSIS (LENKE TYPE 1) Takahiro Makino, Takashi Kaito, Tsuyoshi Sugiura, Masafumi Kashii, Hideki Yoshikawa Department of Orthopaedic Surgery, Osaka University Graduate School of Medicine, Osaka, Japan Purpose: Little has been known about the effect of the asymmetric loading to vertebral bodies on the ossification center of ring apophysis in patients with adolescent idiopathic scoliosis (AIS). The aim of this study was to clarify the asymmetric appearance in the ossification center of ring apophysis between the convex and concave side in patients with AIS using multiplanar reconstruction computed tomography (CT). Methods: Sixteen patients (all female; mean age, 14.5±2.0 years) who underwent corrective surgery for AIS (Lenke type 1) between 2012 and 2013 were included. The mean Cobb angle was 55.5±7.1° at major thoracic (MT) curve and 39.3±10.3° at thoracolumbar/ lumbar (TL/L) curve, and mean Risser sign was 4.3 (range, 3-5). The coronal sections (anterior one quarter of vertebral body, anterior; middle of vertebral body, middle; posterior one quarter of vertebral body, posterior) parallel to posterior wall of vertebral body were reconstructed from preoperative CT scans (320-row detector; slice thickness, 0.5mm), and the appearance of the ossification center was evaluated at the four corners of each sections.

S731 Results: The appearance rates of the ossification center (convex / concave) within MT curve excluding end vertebrae (86 vertebral bodies) were 99.4 % / 85.5 % (anterior), 95.9 % / 47.7 % (middle), 93.6 % / 42.4 % (posterior); those within TL/L curve excluding end vertebrae (56 vertebral bodies) were 95.5 % / 70.5 % (anterior), 77.7 % / 51.8 % (middle), 79.5 % / 41.1 %(posterior) respectively. The appearance rates of the ossification center at both MT and TL/L curve were significantly lower at the concave side than convex side at every section. In the end vertebrae (48 vertebral bodies), the appearance rate were 89.6 % / 89.6 % (anterior), 77.1 % / 82.3 % (middle), 65.6 % / 76.0 % (posterior) respectively, and all these appearance rate did not differ between the concave and convex sides. Conclusion: The appearance rates of the ossification center of ring apophysis were lower at the concave side than convex side except end vertebrae. The result suggests that the asymmetric bony growth of vertebral body come into existence at both structural and non-structural curves. The evaluation of the asymmetric appearance in the ossification center of ring apophysis have possibility to be one of the indices to predict the progression of uninstrumented curve after corrective surgeries or the effectiveness of brace treatments. Disclosures: author 1: none; author 2: none; author 3: none; author 4: none; author 5: none.

QF49 SERIAL ELONGATION, DEROTATION AND FLEXION (EDF) CASTING UNDER GENERAL ANESTHESIA AND NEURO-MUSCULAR BLOCKING DRUGS IMPROVE OUTCOME IN PATIENTS WITH JUVENILE SCOLIOSIS: PRELIMINARY RESULTS Federico Canavese University Hospital Estaing, Dept. of Pediatric Surgery, Clermont Ferrand, France Purpose: Juvenile scoliosis, among different types of spinal deformity, remains still a challenge for orthopedic surgeons. Elongation, Derotation and Flexion (EDF) casting technique is a custom-made thoracolumbar cast based on a three dimensional correction concept. The primary objective of the present study was to measure changes on plain radiographs of patients with Juvenile Scoliosis treated with EDF plaster technique. The second aim was to evaluate the effectiveness of the EDF plaster technique realized under general anesthesia (GA) and neuromuscular blocking drugs, i.e. curare, on the radiological curve correction. Methods: A retrospective comparative case series study was performed in which were included forty four skeletally immature patients. Three patient groups were selected. Group 1: EDF cast applied with patients awaken and no anesthesia; Group 2: EDF cast applied under GA without neuromuscular blocking drugs; Group 3: EDF cast applied under GA with neuromuscular blocking drugs. All the patients were treated with two serial EDF casts by two months and a half each. All measurements were taken from the radiographic exams. Cobb’s angle; Metha’s angle and Nash and Moe grade of rotation were assessed before and after applying the cast. Thirty-four (77.3 %) patients were followed-up at least 24 months after removal of last EDF cast. Results Eighteen patients (3 males, 15 females) were included in Group 1, 12 (2 males, 10 females) in Group 2 and 14 (5 males, 9 females) in Group 3. Serial EDF casting was more effective at initial curve reduction and in preventing curve progression when applied under GA with neuromuscular blocking drugs, i.e. curare. Metha’s anle and Nash and Moe score improved significantly in all groups of patients treated according to principles of EDF technique. During follow-up period, six patients required surgery in Group 1 (6/

123

S732 18; 33.3 %), 3 patients required surgery in Group 2 (3/12; 25 %) and 2 patients underwent surgery in Group 3 (2/14; 15 %). Conclusions: Preliminary results support the hypothesis EDF casting is effective in controlling the curve. It can control the natural evolution of juvenile scoliosis by reducing and slowing curve progression in both frontal (Cobb’s angle) and transverse plane (rib vertebral angle difference and apical vertebral rotation degree). Keywords: juvenile scoliosis; EDF casting; conservative treatment; neuromuscular blocking drugs. Disclosures: author 1: none.

TRAUMA, TUMOR

QF51 AGE-RELATED CHANGES IN THORACIC SPINE KINEMATICS Dominika Ignasiak, Andrea Ru¨eger, Stephen J. Ferguson Institute for Biomechanics, ETH Zurich, Switzerland Osteoporotic vertebral fractures lead to serious clinical consequences such as pain, immobility, posture loss and increased risk of future fractures [1]. The fractures typically develop after the age of 50 years in the vertebrae around the thoracic apex and thoracolumbar junction [2], even under the load of normal daily living activities. An association of motion and pathologies has been previously reported for the lumbar spine, but not for the thoracic spine. Therefore, the aim of this study was to investigate thoracic spine kinematics and their change with age, which could be linked to age-related changes in spinal loading and fracture risk. Forty-two healthy volunteers with no prior spinal surgery or current back problems were recruited in two age groups: young (N = 21, age 27.0±4.0) and elderly (N = 21, age 70.1±3.9). The volunteers performed a forward flexion maneuver, during which positions of reflective skin markers were tracked using a motion-capture system (Vicon). The temporal change of spinal curvature was analyzed by fitting a cubic function to the spine markers positions in the sagittal plane. The correlation of spine kinematics characteristics with age group was tested using ANOVA and Chi-square tests. A sequential flexion maneuver, beginning with flexion of lower spine followed by thoracic flexion, was a characteristic most often observed in the young, but never in the elderly. The temporal sequence of the lower thoracic spine (T7T9, T9T11 and T11L1) was described as ‘‘from bottom to top’’ in most of the young and as ‘‘simultaneous’’ in most of the elderly. In the majority of the young, T11L1 flexed the most of the lower thoracic levels; on the contrary, it flexed the least in most of the elderly. On average, the elderly showed an increased range of T5T7 flexion and a reduced range of T9T11 and T11L1 flexion. Also, mean segmental velocity (normalized to flexion velocity) was higher for T5T7 and lower for T11L1 segments in the elderly. This study provides a novel description of thoracic spine kinematics in healthy young and elderly subjects. Significant differences were found between the age groups in the motion characteristics of the lower and middle thoracic levels, which are the most common sites of vertebral fractures. Further investigation is needed to explore the biomechanical consequences of these age-related changes in spine kinematics.

123

Eur Spine J (2015) 24 (Suppl 6):S711–S742 References [1] Ross, Am J Med 103(2), 1997. [2] Delmas et al., JBMR 20(4), 2005. Disclosures: author 1: grants/research support; Company = AOSpine; author 2: none; author 3: grants/research support; Company = AOSpine.

QF52 DIFFUSION-WEIGHTED MRI ASSESSMENT OF ADJACENT DISC DEGENERATION AFTER THORACOLUMBAR VERTEBRAL FRACTURES David Noriega, Francisco Ardura, Ruben Hernandez Ramajo University of Valladolid, Spain Objective: The purpose of our study was to determine the water diffusion in the thoracolumbar discs adjacent to a previous vertebral fracture. By using the mean apparent diffusion coefficient (ADC), the aim of this work was to analyze if a relationship exists between disc ADC and MR findings of adjacent disc degeneration after thoracolumbar fractures treated by anatomic reduction using vertebroplasty. Material and Methods: A series of 20 non-consecutive voluntary patients treated because of vertebral fractures (mean age, 50.7 years; range, 45-56) were included in the study. There were 10 A3.1 and 10 A1.2 fractures (AO classification). Surgical treatment was applied in 14 cases, and conservative in 6. The intention of surgery was the anatomic restoration of the vertebral endplates by placement of expandable implants into the vertebral body through a minimally invasive transpedicular approach. MRI T2-weighted images and mapping of apparent diffusion coefficient (ADC) of the intervertebral disc adjacent to the fracture segment were performed after a mean follow-up of 32 months. A total of 60 discs, 3 per patient, were analyzed: infra-adjacent, supra adjacent and a control disc one level above the supra-adjacent. As compared to surgically treated patients, discs at the supra-adjacent fracture level showed statistically significant lower values in cases treated orthopedically (p \ 0.001). Results: There were no differences between patients surgically treated and those following a conservative protocol regarding the average ADC values obtained in the 20 control discs analyzed. Taken all cases together, the average ADC in the supra-adjacent level was lower than in the infra-adjacent (1.53 ± 0.06 versus 1.35 ± 0.12; p \ 0.001). Average ADC values of the disc used as a control were similar than those of the infra-adjacent level (1.54 ± 0.06). as compared to surgically treated patients, discs at the supra-adjacent fracture level showed statistically significant lower values in cases treated orthopedically (p \ 0.001). The variation in the delay of surgery had no influence on the average values of ADC at any of the measured levels. Conclusions: ADC measurements of the supra-adjacent disc after a mean follow-up of 32 months following thoracolumbar fractures showed that restoration of the vertebral collapse by minimally invasive vertebral reduction and augmentation prevent posttraumatic disc degeneration. Disclosures: author 1: consultant; Company =; author 2: consultant; Company = Vexim; author 3: consultant; Company = Sacyl.

Eur Spine J (2015) 24 (Suppl 6):S711–S742 QF53 CLINICAL OUTCOMES OF INTRAOPERATIVE CONEBEAM CT NAVIGATION FOR PEDICLE SCREW INSTRUMENTATION IN EMERGENT SPINE SURGERY John Street, Daniel Mendelsohn, Jason Strelzow, Juliet Batke, Nicolas Dea University of British Columbia, Vancouver, Canada Background: Computer-assisted spine navigation with intraoperative cone-beam CT is being increasingly utilized for elective degenerative and major deformity spine surgery. ‘After hours’ use of intraoperative CT for unscheduled emergency and oncology operations is constrained by personnel and technical support availability. The purpose of this study was to investigate the clinical outcomes of patients requiring emergent spine surgery comparing navigated and non navigated techniques. Methods: We identified patients who underwent emergent spine trauma and oncology surgery using intraoperative cone-beam CT navigation (NAV). A historical control cohort was matched from patients undergoing emergent surgery using conventional fluoroscopy (nonNAV). Operative time, estimated blood loss, readmissions and reoperations, postoperative surgical-site infections, and screw positioning accuracy were recorded. Results: 55 patients underwent surgery for trauma and 14 for metastatic tumors in each cohort. At total of 800 NAV and 912 nonNAV pedicle screws were inserted. The majority of patients (83.3 %) were admitted through the emergency department or were transferred directly from a different hospital. The use of navigation significantly increased operative time (291.5 min vs. 246.4 min). Estimated blood loss was not significantly different between the cohorts (1150 vs. 1350 mL). The number of postoperative infections requiring readmission (7.2 % vs. 2.9 %) and reoperation (10.1 % vs. 5.8 %) were higher in the navigated cohort but not significantly different. Two nonNAV patients underwent reoperation for screw malposition compared to none in the NAV cohort; the difference was not statistically significant. Significantly more grade I (NAV: n = 21, 2.6 %, nonNAV: n = 34, 3.7 %) and grade II (NAV: n = 18, 2.3 %, nonNAV: n = 47, 5.2 %) pedicle breeches in the nonNAV cohort compared to the NAV cohort. In terms of the direction of breeches, significantly more nonNAV breeches were medial (NAV: 1.5 %, nonNAV: 3.5 %). Conclusions: The use of intraoperative cone-beam CT navigation increased operative time by 97 minutes for metastatic tumors and 32 minutes for trauma cases, resulted in no change in operative blood loss or postoperative infection rate, and decreased Grade 1 and 2 and medial screw malpositions. Our findings suggest that intraoperative cone-beam CT navigation is safe and feasible for operative management of emergent and traumatic spinal cases. Disclosures: author 1: grants/research support; Company = Medtronic; author 2: none; author 3: grants/research support; Company = Medtronic; author 4: none; author 5: none.

S733 Introduction: Dorso-ventral stabilization is a frequently used procedure for the treatment of traumatic vertebral body fractures [1]. Many different implants are offered for vertebral body replacement. The aim of the current study was the analysis of radiological images and clinical outcome 3 years after implantation of a vertebral body replacement that is expanded by water pressure. Methods: The current study includes all patients of our trauma center which suffered a singular traumatic fracture of a vertebral body (Th 5 - L 5) in a 13 months’ period, that (i) underwent a dorsal instrumentation (VIPER 2Ò, DePuy Synthes, Tuttlingen, Germany) and (ii) afterwards underwent the implantation of a hydraulic expansible vertebral body replacement (Hydrolift, Aesculap AG, Tuttlingen, Germany) via mini-open lateral thoracotomy. The radiological findings of all patients were evaluated (pre- and postoperatively and 3 years after implantation): regional sagittal angle was determined; the implants’ sintering and current position were analyzed. Moreover, the clinical outcome 3 years after implantation was analyzed via VAS spine score [2]. Statistical Analysis was performed with the software SAS 2.9. Results: n = 53 patients could be included in the current study. The complete follow-up was successful for n = 47 patients (follow-up rate: 89 %). The analysis of the radiological data showed an average sintering of the implants of 1.1 ± 1.2 mm. After the initial operation, the local sagittal angle remained stable in the 3-years-follow up at thoracic spine (5.4 ± 4.8 ° vs. 5.2 ± 4.8 °; p = 0.451) and lumbar spine (-7.2 ± 6.0 ° vs. -7.0 ± 6.0 °; p = 0.451). Furthermore, no change in the implants’ position could be observed. Most of the patients (n = 40) were satisfied with their outcome. The mean rating of the VAS spine score was 65.2 ± 23.1. Discussion: This study shows that the implantation of a hydraulic expansible vertebral body replacement allows a permanent stable fixation after traumatic fractures of the thoracic and lumbar spine. The local sagittal angle is fixed after 3 years. Secondary dislocation and serious sintering of the implant were not observed. The clinical outcome after implantation of a hydraulic expansible vertebral body replacement is comparable to the outcome after the ventral stabilization with an iliac crest bone graft [2]. References: [1] Reinhold M et al. (2010) Eur Spine J, 19: 1657 [2] Knop C et al. (2001) Unfallchirurg, 104: 488 Disclosures: author 1: grants/research support; Company = Aesculap AG, Tuttlingen, Germany; author 2: none; author 3: none; author 4: none.

QF55 SPINAL TRAUMA IN PATIENTS WITH ANKYLOSING SPINAL DISORDERS - A MULTICENTER STUDY, NAGOYA SPINE GROUP Shiro Imagama, Zenya Ito, Kei Ando, Mikito Tsushima

QF54 RADIOLOGICAL AND CLINICAL RESULTS AFTER IMPLANTATION OF A HYDRAULIC EXPANSIBLE VERTEBRAL BODY REPLACEMENT FOLLOWING TRAUMATIC VERTEBRAL FRACTURES IN THE THORAKOLUMBAR SPINE - A THREE-YEARS-FOLLOW-UP Michael Kreinest, Dorothee Schmahl, Paul A. Gru¨tzner, Stefan Matschke BG Trauma Center Ludwigshafen, Department of Trauma and Orthopedic Surgery, Ludwigshafen, Germany

Nagoya University Graduate School of Medicine, Department of Orthopedic Surgery, Nagoya, Japan Background: In Japan, the number of patients with ankylosing spinal disorders (ASD), such as ankylosing spondylolisis (AS) and diffused idiopathic skeletal hyperostosis (DISH), is increasing. ASD patients are prone to unstable spinal fractures, even when the trauma is trivial. It is often difficult to be noted unstable fractures. Because such unstable fractures cannot be managed using conservative treatments, surgical treatments are employed. This is a retrospective study of a consecutive series of operatively managed ASD patients with spinal fractures.

123

S734 Methods: From January 2003 to December 2012, 79 with ASD fractures (AS 11 and DISH 68, 55 men and 24 women) were operated in 5 institutions, belonging to the Nagoya Spine Group. Patient demographics, injury characteristics, outcomes and complications were obtained from medical records and radiographs, and subsequently used for the study. Exclusion criteria were reoperation and not spinal column fractures, such as transverse fracture. Results: The mean age was 73.8 years old. 52 cases were the results of low impact trauma, such as falls from ground level. The majority of fractures were located in the thoracic spine 41, and in particular 37 cases (47 %) were affected in thoracolumbar transition (T11-L1). All fractures were damaged from anterior to posterior columns. 26 cases (33 %) had more than 10 days delay in the diagnosis of their fractures. 44 patients (56 %) had pre-operative neurologic paraparesis. During surgery, multilevel posterior instrumentations, 2 above and 2 below the injury at a minimum were performed. Patients with neurological deficits had additional spinal decompression performed. Based on the Frankel Classification system, 25 patients (32 %) had post-operative improvement of more than one grade. 7 patients had surgical complications whereby the surgical site was infected, and 1 had newer neurological deficits post-operatively. In 54 cases (68 %) that were followed up for more than 6 months, there was confirmation of bone union at the site of injury. Conclusion: In the surgical cases of spinal trauma with a background of ASD, there are cases where the diagnosis of 3-column damage is not made, and thus treated conservatively. Especially in the cervical spine and thoracolumbar transition, delay in diagnosis is highly associated with paralysis. Thus, the stability of the injured spine should be ascertained as soon as possible by MRI and CT, and surgical intervention should be considered. Disclosures: author 1: none; author 2: none; author 3: none.

QF56 PREDICTING HIGH SURGICAL TREATMENT COSTS AT PRIMARY SPINAL TUMOR PATIENTS Zsolt Szo¨ve´rfi, Aron Lazary, Izabella Ko´szo´, Korne´l Papik, Pe´ter Pa´l Varga National Center for Spinal Disorders, Budapest, Hungary Background: Surgical treatment cost of primary spinal tumors -due to the complexity and magnitude of the surgery- can be high. From a healthcare provider perspective being aware of those factors that influence treatment cost is essential for the best hospital resource management. However the impact of the individual patient characteristics on direct medical costs is unknown. Objective: Our objective was to identify those preoperative factors that predict a high direct cost at patients with surgically treated primary spinal tumors. Methods: A retrospective micro-costing study was conducted from a healthcare provider perspective. Total cost per patient resulted from the aggregation of inpatient-days costs, ICU days costs, OR time costs (surgery and anesthesiology), medication costs, blood transfusion costs and other material costs, including appropriate allocation of overheads. Patients with primary spinal tumors operated in the National Center for Spinal Disorders between 2011 and 2014 were selected. The prognostic value of ten preoperative factors (dignity, previous surgery, pathological fracture, neurological dysfunction, age, spinal level, ASA score, Charlson Comorbidity Index, BMI, tumor volume) was investigated in linear regression modelling. Results: 72 surgically treated primary spinal tumor patients were included in the study. Median length of hospital stay was 12 days

123

Eur Spine J (2015) 24 (Suppl 6):S711–S742 (3-115), the median ICU treatment was 1.4 days (0-21.7), the median OR time was 2.8 hour (0.4-20.4). Median cost per patient was 3, 088.5 € (537.7 €-21, 860.6 €). In the univariate analysis only ASA score, Charlson Comorbidity Index and pathological fracture were not significantly associated with higher cost. The seven significant variables were further assessed in a multivariate model. Dignity, previous surgery, spinal level and tumor volume were associated with high cost and preoperative neurologic dysfunction showed only a trend towards significance (whole model: R = 0, 69, R2 = 0, 44, df = 5, F = 12, 34 p \ 0, 001). Conclusion: The present study identifies four predictive variables for high hospital cost related to primary spinal tumor surgery. Malignant tumors, previous surgery, sacral involvement and large tumor volume significantly increase the cost of treatment. This method of cost analysis provides useful insights for resource management in the surgical treatment of spinal tumors. Disclosures: author 1: none; author 2: none; author 3: none; author 4: none.

QF57 EVALUATION OF SCORING SYSTEM AND PROGNOSTIC FACTORS IN PATIENTS WITH SPINAL METASTASES FROM LUNG CANCER Kimberly-Anne Tan, Jonathan J.H. Tan, Sandar Zaw Aye, Naresh Kumar National University Hospital, Department of Orthopaedics, Singapore Introduction: The decision for operative treatment of patients with spinal metastases is dependent on the patient’s predicted survival. Current prognostic systems include the Tokuhashi, Tomita, Bauer and Oswestry systems. However, the best system for predicting survival of patients with spinal metastases specifically from lung cancer has yet to be found. The high incidence of spinal metastases from lung cancer as well as the improved survival of patients treated with tyrosine kinase inhibitor therapy warrants investigation of the accuracy of these scoring systems. Methods: 180 patients with spinal metastases from lung cancer treated at our institution between May 2001 and August 2012 were studied. 51 of these patients underwent operations for their spinal metastases while 129 received other non-operative modalities of treatment. The primary outcome measure was survival time from the time of diagnosis of spinal metastases. Survival data was obtained from the National Registry of Diseases Office, with survival data being complete at the time of conclusion of the study. Potential prognostic factors of survival included within the various scoring systems, amongst others, were examined. Predicted survival according to the four scoring systems was compared with actual survival. Potential prognostic factors were investigated using Cox regression analyses. Kaplan-Meier survival estimates and Log-rank tests were conducted for all scoring systems, and their predictive values were determined using post-estimation. Results: Cancer type (NSCLC) (p \ 0.001), treatment with tyrosine kinase inhibitor therapy (p = 0.01), Karnofsky Performance Status (pB0.015), absence of palsy (pB0.001) and absence of visceral metastases (p = 0.007) are significant predictors of survival. The absolute score of all four scoring systems was significantly associated with actual survival, although this significance did not extend to their different prognostic subgroups, except the Oswestry Spinal Metastasis Risk Index. Predictive values were 0.70, 0.69, 0.65 and 0.60 for Tokuhashi, Tomita, Bauer and Oswestry scores respectively. Conclusion: Current scoring systems used to predict survival of patients with spinal metastases from lung cancer are inadequate. As

Eur Spine J (2015) 24 (Suppl 6):S711–S742 lung cancer histology appears prognostic, a new scoring system incorporating these factors should be considered, to account for increased survival gained from the use of more targeted lung cancer therapy. Disclosures: author 1: none; author 2: none; author 3: none; author 4: none.

S735 QF59 METASTATIC SPINE TUMOUR SURGERY: MINIMALLY INVASIVE APPROACH VERSUS OPEN APPROACH Naresh Kumar, Pang Hung Wu, Aye Sandar Zaw, Rishi Malhotra National University Health System, Singapore

QF58 A MODIFICATION OF THE TOKUHASHI REVISED SCORE IMPROVES PROGNOSTIC ABILITY IN PATIENTS WITH METASTATIC SPINAL CORD COMPRESSION. Søren Schmidt Morgen, Martin Gehrchen, Sebastian Bjørck, Claus Falck Larsen, Svend Aage Engelholm, Benny Dahl 1

Spine Unit, Department of Orthopedic Surgery, Copenhagen University Hospital, Rigshospitalet; 2Trauma Center, Copenhagen University Hospital, Rigshospitalet; 3Department of Radiation and Oncology, Copenhagen University Hospital, Rigshospitalet, Copenhagen, Denmark

A Modification of the Tokuhashi Revised Score Improves Prognostic Ability in Patients With Metastatic Spinal Cord Compression. Summary: Patients suffering from metastatic spinal cord compression (MSCC) are routinely evaluated with pre-operative scoring systems. This study on 1143 MSCC patients modify the most recommended prognostic scoring system and show that it improves the prognostic value for patients with MSCC. Hypothesis: A modification of the Tokuhashi Revised score (TR) can improve its prognostic precision. Design: Single-center, prospective, cohort study. Introduction: A modification of the TR has been proposed in recent studies. This study proposes a modification of the TR in order improve the prognostic value. Methods: In 2011 and 2012 a total of 1143 consecutive patients admitted with MSCC were prospectively included and variables for the TR score were collected. In the 2011 cohort (544 patients) each component of the TR including primary cancer diagnosis was analyzed regarding prognostic ability by Cox regression analysis. Based on these findings a modified TR score (MTR) was defined. New prognostic groups were determined using Receiver Operating Characteristics (ROC) analysis for survival \ 6 months, [ = 6 months, and [ = 12 months. These prognostic groups were used to compare the TR and the MTR in the 2012 cohort (599 patients) based on Kaplan-Meier (KM) survival curves and ROC analysis. Results: There were no significant differences between the 2011 and the 2012 cohort regarding age and gender distribution. The mean age was 66 years (range 20-97) and the most common primary tumor sites were prostate (21 %) and lung (20 %). In the 2012 cohort the KM curves showed that the actual survival for each prognostic group was separated more accurate with the MTR compared to the TR. The areas under the ROC curves were significant larger for the MTR compared with the TR. The ROC area were for \ 6 months survival, MTR = 0.71 and TR = 0.65; p = 0.003, for [ = 6 months survival, MTR = 0.71 and TR = 0.65; p = 0.003, for [ = 12 months survival MTR = 0.72 and TR = 0.67; p = 0.0015. Conclusion: A modification of the TR can improve the precision in the estimation of survival among patients with MSCC. Disclosures: author 1: none; author 2: grants/research support; Company = Medtronic, Globus Medical, K2M, consultant; Company = Medtronic, Globus Medical, K2M; author 6: grants/research support; Company = Medtronic, Globus Medical, K2M.

Background: Minimally invasive surgery (MIS) approaches have shown for early wound healing and reduced intra-operative blood loss in degenerative spine diseases. Such surgical strategies in metastatic spine disease (MSD) are highly desirable in reducing surgical morbidity and enabling earlier initiation of oncological treatment. However, previous studies were case series; none comparing the outcomes between MIS and open approaches. We aimed to compare the outcomes of MIS and open approach in patients with symptomatic MSD. Materials & Methods: We included patients undergoing surgery for MSD in our institution between 2011 and 2014: 22 patients who underwent posterior surgery using MIS and other 22 using open approach. Preoperative, intraoperative and postoperative data were collected for direct comparison of the two approaches. Generalized Linear Model was exploited to estimate the effect of MIS on outcomes, adjusting potential confounders. All patients were followed up for a minimum period of three months from time of surgery until the end of study period or till their demise. Results: All patients showed improvement in pain and neurological status. Full normal function (Frankel score E) was achieved or maintained in 82 % in MIS group compared to 54 % in open group post-operation. Kruskal-Wallis analysis showed a significant difference in Frankel score between pre-operation and post-operation in MIS group (P \ 0.01). Independent ambulation was observed within 3 months of surgery in 88 % in MIS as compared to 64 % in open group. This difference, however, was not significant. In multivariate analysis, the amount of blood loss was significantly lower (537ml less) in MIS group than open group. A significant difference in time to start radiotherapy from index surgery was observed between MIS and open group with initiation of radiotherapy 7 days earlier in patients in MIS group compared to open group. Operative time, duration of hospital stay and time to initiate chemotherapy were also favourable in MIS group though the difference was not statistically significant. Conclusion: MIS in MSD have shown promising results for patients suffering from neurological deficit, instability or back pain. MIS approaches are there to stay and evolve with time for the treatment of MSD due to less peri-operative morbidity and allowing earlier radiotherapy. Disclosures: author 1: none; author 2: none; author 3: none; author 4: none.

QF60 POSTERIOR THORACIC CORPECTOMIES WITH CAGE RECONSTRUCTION FOR METASTATIC SPINAL TUMORS: COMPARING THE MINI-OPEN APPROACH TO THE OPEN APPROACH Dean Chou, Darryl Lau Department of Neurological Surgery, University of California San Francisco Object: There has been a shift to utilize posterior-only approaches to performing thoracolumbar corpectomies for spinal metastasis, but most are still done through an open approach. This study compares

123

S736 outcomes of patients who undergo mini-open vs. traditional open transpedicular corpectomy for thoracic spinal metastasis. Methods: A consecutive cohort from 2006 to 2013 of 49 adult patients who underwent transpedicular corpectomies in the thoracic spine for spinal metastases was retrospectively identified. Patients were categorized into open surgery and mini-open surgery. The mini-open transpedicular corpectomy was performed with a midline facial incision only over the corpectomy level of interest and percutaneous instrumentation above and below that level. The open procedure consisted of a traditional posterior transpedicular corpectomy. Chi-square test and two-tailed t-test were employed to compare perioperative and follow-up outcomes between the two groups. Results: In the analysis, 21 patients had mini-open surgery, and 28 patients had open surgery. The mean age was 57.9 years, and 59.2 % were male. There were no significant differences in demographics, comorbidities, tumor type, neurological status (ASIA score), number of corpectomies performed, and number of levels instrumented between the two groups. The open group had a mean operative time of 413.6 minutes, and the mini-open group had a mean operative time of 452.4 minutes (p = 0.329). Compared to the open group, the miniopen group had significantly less blood loss (917.7 cc vs. 1697.3 cc, p = 0.019) and a significantly shorter hospital stay (11.4 days vs. 7.4 days, p = 0.001). There was a trend towards a lower perioperative complication rate in the mini-open group (9.5 %) compared to the open group (21.4 %), but this was not statistically significant (p = 0.265). At follow-up, there were no significant differences in ASIA score (p = 0.342), complication rate after the 30-day postoperative period (p = 0.999), or need for surgical revision (p = 0.803). The open approach had a higher overall infection rate of 17.9 % compared to the mini-open approach (9.5 %), but this was not statistically significant (p = 0.409). Conclusion: The mini-open transpedicular corpectomy is associated with less blood loss and shorter hospital stay compared to open transpedicular corpectomy. The mini-open corpectomy also trended towards a lower infection and complication rates, but these did not reach statistical significance. Disclosures: author 1: consultant; Company = Orthofix; author 2: none.

NEW TECHNIQUES, IMAGING, PATIENT SAFETY, INFECTION, COMPLICATIONS

QF61 DEGENERATIVE CHANGES IN T1q MRI OF THE LUMBAR SPINE IN ASYMPTOMATIC SUBJECTS Tobias Schulte, Volker Vieth, Georg Gosheger, Christoph Stehling, Walter Heindel, Joachim Gerß, Tobias Lange, Raphael Gu¨bitz Department of Orthopedics and Tumor Orthopedics, University Hospital Mu¨nster, Germany; Department of Clinical Radiology, University Hospital Mu¨nster, Germany; Institute of Biostatistics and Clinical Research, University of Mu¨nster, Germany; Fraunhofer MEVIS, Institute for Medical Image Computing, Bremen. The clinical relevance of degenerative changes in lumbar discs on MRI scans are a matter of discussion in everyday practice as well in symptomatic as in asymptomatic patients. In 1990, Boden et al. published a study on pathologic changes in asymptomatic subjects in qualitative MRI. In the present study lumbar discs of asymptomatic subjects were investigated using a modern quantitative MRI T1q mapping technique. T1q relaxation times positively correlate with the

123

Eur Spine J (2015) 24 (Suppl 6):S711–S742 proteoglycan content of discs. The study was supported by a Research Grant of the German Spine Foundation. The study included 81 volunteers without any history of back pain divided into 3 age groups (A: 20-39 y, n = 24; B: 40-59 y, n = 30; C: 60-80 y, n = 27). They underwent MRI at 1.5 T with acquisition of sagittal T1q-mapping sequences. Segmentation was done focusing on the entire disc as well as on the nucleus pulposus and the anulus fibrosus. Analysis of the influence of age, BMI and lumbar level on the T1q relaxation time was done using the pearson correlation test and a linear mixed model. Mean T1q relaxation times (in ms) were: Total disc: group A: 68 ms; group B: 56 ms; group C: 45 ms. Anulus fibrosus: group A: 51 ms; group B: 47 ms; group C: 41 ms. Nucleus pulposus: group A: 95 ms; group B: 71 ms; group C: 51 ms. There was a significant influence of age on T1q relaxation times with decreasing times with increasing age. Furthermore, there was a significant difference between age groups A vs. C as well as B vs. C. BMI showed a significant impact on T1q relaxation times with high BMI correlating with low T1q times. Subgroup analysis showed this to be especially true for subjects of group C. Considering the lumbar level there was a significant decrease from L1/2 to L5/S1. Analysis of the disc parts showed significant differences with the highest T1q relaxation times for the nucleus pulposus and the lowest for the anulus fibrosus. In conclusion, this study shows that in asymptomatic subjects increasing age and increasing BMI correlate with increasing disc degeneration measured by T1q MRI. The lower discs showed more degeneration than the upper. As our subjects were all asymptomatic, it needs to be stressed that correlating symptoms and images is crucial in everyday practice. This study using quantitative imaging supports and extends the qualitative findings by Boden. T1q MRI is a useful non-invasive complementary tool to quantify disc degeneration as well in clinical situations as in research. Disclosures: author 1: grants/research support; Company = Research Grant of the German Spine Foundation; author 2: none; author 3: none; author 4: none; author 5: none; author 6: none; author 7: none; author 8: none.

QF62 SANDWICH VERTEBRAL FRACTURE IN THE STUDY OF ADJACENT-LEVEL FRACTURE AFTER CEMENTED VERTEBRAL AUGMENTATION Fe´lix Tome´-Bermejo, Angel R. Pin˜era, Javier Melchor Duart Clemente, Silvia Pe´rez-Esteban, Luis Alvarez Spine Department. Fundacio´n Jime´nez Dı´az University Hospital, Madrid, Spain Objectives: Cemented vertebral augmentation procedures provide fast pain relief and quality of life improvement in patients with osteoporotic vertebral fractures (OVF). However, cement injection could alter the mechanisms of load transmission on the adjacent vertebral bodies in such a way that it could predispose to the occurrence of future adjacent vertebral fractures (AVF). The incidence of a second OVF in the following year in the untreated population is 19.2 %. Literature does not provide conclusive results about whether the AVF are related to the natural progression of osteoporosis or if on the contrary they are a result of the cement injection. Prophylactic vertebroplasty of adjacent levels was proposed to prevent the occurrence of future AVF. However, prophylactic vertebroplasty is not without risk, so a definitive justification for cementing an intact vertebra is needed. The purpose of this study is to investigate if there is a real increase in the risk of AVF in patients undergoing cemented

Eur Spine J (2015) 24 (Suppl 6):S711–S742 vertebral augmentation procedures to justify prophylactic adjacent cementation. Material and Methods: A sandwich vertebra is an intact vertebral body located between 2 previously cemented vertebrae, and is probably the maximum risk scenario for the occurrence of an AVF. We reviewed 358 patients that underwent percutaneous vertebroplasty for painful osteoporotic vertebral fracture with a mean follow-up of 9.35 years (range 5.1 to 21.1 years). Results: After the initial treatment, a total of 39 sandwich situations (10.8 %) were identified, of which 34 patients were available for follow-up. Six patients (17.6 %) developed a sandwich vertebra fracture, of which four patients were treated with a new vertebroplasty. Eight patients (20.5 %) developed one or more OVF at different levels to the sandwich vertebra. The incidence of sandwich vertebra fracture observed was not higher than at distant levels (p[0.05). The time interval of recompression after vertebroplasty was 24.8 months on average. In addition, gender, location of original fracture segment, the amount of cement, cement leakage, and initial number of OVFs were documented, but these were not influencing factors in this study (p[0.05). Conclusions: The presence of sandwich situation is not associated with an increased secondary fracture rate compared with the incidence of distant fractures, or the population with no history of cementation. Given the progressive nature of osteoporosis, the occurrence of AVF after vertebral augmentation procedures could be result of the natural course of osteoporosis and prophylactic augmentation for sandwich situation is not necessary. Disclosures: author 1: none; author 2: none; author 3: none; author 4: none; author 5: grants/research support; Company = Biomet, consultant; Company = Spineart.

QF63 NEW SURGICAL APPROACH AND DECOMPRESSION PRESERVING MUSCLE-LIGAMENTOUS COMPLEX FOR LUMBAR CANAL STENOSIS Futoshi Suetsuna Hachinohe Municipal Hospital, Hachinohe, Japan Background: Conventional fenestration for lumbar canal stenosis (LCS) makes us often difficult to leave 1/2 of facet joint in cases with steep facet angles or large spinous process, and causes muscle atrophy due to wide spread of paraspinal muscles. Purpose: To evaluate our fenestration method using new approach of bilateral oblique cutting of spinous process (BOCS). Design: A retrospective study of patients treated with BOCS approach. Patient sample: 20 patients with an average age of 71.7 years. All patients underwent fenestration using BOCS approach for LCS. Outcome measure: Radiographic parameters included bony fusion of spinous process temporarily disconnected to the left and right, atrophy rate of the paraspinal muscles at 1 year after surgery using MRI and cross-section area (mm2) of the spinal canal using CT at L4/5 level. Clinical parameters included operation time and blood loss per onelevel, recovery rate using JOA score and Visual Analogue Scale (VAS, maximum 100mm) of low back pain on 3rd day after surgery. Methods: Radiographic and clinical outcomes of the patients who underwent our method in 2011-2013 were reviewed with an average follow-up period of 20 months ([1 year). Surgical procedure of L4/5 fenestration is as follows. Skin incision is about 3-4 cm on L4 and 5 spinous processes. Posterior ligament is divided into right and left,

S737 and we drill the lower one-third part of L4 spinous process with an air drill, leaving the periosteum. Subsequently, we obliquely cut the L4 spinous process left and right with micro bone saw, and divide the detached bone fragment with the ligaments to the left and right leaving the L4 spinous process, and expose the L4 lamina. After fenestration, we suture the temporarily detached bone fragments to the remained L4 spinous process. Results: The average operation time and blood loss were 59 minutes and 83 ml. The average preoperative and postoperative JOA scores were 16.7 points and 24.6 points. Recover rate was 79.6 %. The average VAS score was 9 mm. CT showed 90 % bony fusion of spinous process. The average muscle atrophy rate was 10.0 %. The average preoperative and postoperative cross-section areas of spinal canal were 204.9 mm2 and 318.3 mm2. Conclusion: Our method using BOCS approach preserving posterior elements produced satisfactory outcomes, safe decompression and adequate prevention of paraspinal muscle atrophy. It is indicated for not only standard LCS but also LCS with steep facet angle. Disclosures: author 1: none.

QF64 OCCULT INFECTIONS IN PATIENTS UNDERGOING REVISION SURGERY Celeste Abjornson, Tucker Callanan, Victor Yoon, Frank Cammisa Hospital for Special Surgery, New York, USA Often in revision spinal surgery, chronic pain and/or pseudoarthrosis are the main indications, and normal screening rules out the presence of infection. Recent findings from total joint arthroplasty are that over 10 % of presumed aseptic or culture-negative implants have been removed to only later be found to be positive for bacteria. If insidious organisms exist in low concentrations in a biofilm that surrounds the majority of implant surfaces and may be identifiable only after non-routine longer culture times, these organisms can disrupt bone formation leading to micro-motion (pain generation) around the implants or failure of a solid fusion. Occult spinal infections provide a differential diagnoses to be considered in patients experiencing chronic pain when surgical parameters such as fusion status and hardware placement are normal. The purpose of this investigation was to ascertain the incidence of occult pathogens in revision spinal surgery, as well as identify properties in the spinal implants that may have predisposed them to infections. Under IRB approval, implant retrievals of a prospective, consecutive series of twenty adult patients undergoing a posterior revision in either the cervical, thoracic, or lumbar spine requiring hardware removal were studied. Peri-operative antibiotics were held until hardware removal and swabs had been taken. Swabs were taken in each screw hole and around each screw head for long cultures (10 days). The hardware was analyzed for wear by light microscopy and scanning electron microscopy. Occcult adherent bacteria were found in 35 % of the tissue surrounding the implants of patients undergoing spine revision surgery. The most prevalent bacteria found were P. acnes, in 71 % of the infected patients. Other cultures grew Klebsiela pneumoniae, Coccobacilli, Trueperella bernardia, as well as Staphylococcus simulans, saccharolyticus, epidermidis, and capitis. Etching and pitting consistent with normal wear and removal was visible on each implant analyzed via SEM, but not abnormal to revision surgeries collected for known pathologies.

123

S738 The findings of this study suggest that occult infections are far more prevalent in patients requiring revision surgery than previously identified. Identifying the factors that contribute to failed surgery and/ or chronic pain are critical. The results of this study demonstrate that occult infections could be a contributing factor to this issue. Disclosures: author 1: grants/research support; Company = Seaver Foundation; author 2: grants/research support; Company = Seaver Foundationl; author 3: none; author 4: grants/research support; Company = Seaver Foundation.

QF65 AN ALGORITHMIC APPROACH FOR MANAGING ORTHOPAEDIC SURGICAL WOUNDS OF THE CICATRICIAL SCOLIOSIS: NEW TECHNOLOGIES En Xie, Dingjun Hao Department of Spine Surgery, Hong Hui Hospital, Xi’an Jiaotong University College of Medicine, Xi’an, China Background: Wound breakdown after Cicatricial scoliosis surgery may necessitate secondary soft tissue coverage. The Cicatricial scoliosis region is difficult to reconstruct for orthopaedic and plastic surgeons owing to its complex bony anatomy and unique functional demands. Purposes: We evaluated our current algorithmic approach for managing Cicatricial scoliosis surgical wounds of the Cicatricial scoliosis with respect to whether (1) prophylactic or simultaneous soft tissue coverage affected wound healing complications (secondary plastic surgery, orthopaedic hardware removal, malunion, further orthopaedic surgery, ultimate failure) and (2) postoperative referral for soft tissue management was associated with wound location, size, and Cicatricial scoliosis procedure. Methods: We retrospectively reviewed 37 patients who underwent elective Cicatricial scoliosis surgery and required concomitant plastic surgery at our institution. Study end points included secondary plastic surgery procedures, hardware removal for infection, Cicatricial scoliosis malunion, further orthopaedic surgery, and wound-healing failure as defined by a chronic non healing wound. Minimum followup was 12 months (mean, 27.2 months; range, 12-36 months). 3 patients were dropped to complete followup. We developed an algorithm that centers on two critical points of care: preoperative evaluation by the orthopaedic surgeon and evaluation and treatment by the plastic surgeon after referral. Results: Compared with postoperative intervention, prophylactic or simultaneous soft tissue coverage did not lead to differences in frequency of secondary plastic surgery procedures (p = 0.55), hardware removal procedures (p = 0.13), malunions (p = 0.47), further orthopaedic surgery (p = 0.48), and ultimate failure (p = 0.27). Patients referred postoperatively for soft tissue management most frequently had Cicatricial scoliosis wounds (p \ 0.001) of smaller size (p = 0.03). Most commonly associated with total Cicatricial scoliosis surgery (p = 0.004). Conclusions: Using our algorithmic approach, prophylactic or simultaneous soft tissue coverage did not increase the study end points. Our algorithm facilitates early identification of skin instability and enables prompt soft tissue coverage before or concurrently with orthopaedic procedures. The effect of prophylactic or simultaneous soft tissue coverage on postoperative wound-healing requires additional investigation. Disclosures: author 1: none; author 2: none.

123

Eur Spine J (2015) 24 (Suppl 6):S711–S742 QF66 IS PREOPERATIVE FIBRINOGEN PLASMA CONCENTRATION A GOOD BLEEDING PREDICTOR IN ADULT SPINAL DEFORMITY SURGERY? Lidia Mora, Alba Vila-Casademunt, Ferran Pellise´, M Jose´ Colomina, Montse Domingo-Sa`bat, Francisco Javier S Pe´rez-Grueso, Ahmet Alanay, Emre Acaroglu, Joan Bago´, ESSG European Spine Study Group Hospital Vall Hebron, Barcelona; Vall Hebron Institut de Recerca, Barcelona; Hospital De La Paz, Madrid; Acibadem University School of Medicine, Istanbul; Ankara Spine Center, Ankar, Turkey Introduction: Since bleeding and transfusions are associated with increased postoperative complications, a preoperative variable to predict excessive bleeding would be very valuable. Fibrinogen was previously determined to be a good bleeding predictor in cardiac surgery, and recently was proposed for Adolescent Idiopathic Scoliosis surgery as well (Spine 2011, 1; 36(7). The main objective of this study was to investigate the potential correlation between preoperative fibrinogen and bleeding and transfusion requirements in Adult Spinal Deformity surgery. Methods: All patients older than 18 years with coronal Cobb [208 or Sagittal Vertical Axis [5cm or Pelvic Tilt [258 or Thoracic Kyphosis [608 who consecutively underwent surgery in the same center were included in the study. Preop-lab variables, intraoperative data, transfusions and complications were analyzed. Preoperative fibrinogen, intraoperative Estimated Blood Loss, (EBL, % of estimated blood volume: % of estimated blood volume: men 75ml/kg, women 65ml/kg) and transfusion requirements were stratified according to the sample distribution. The correlation between fibrinogen and EBL was calculated using the Spearman’s rho. Fibrinogen and transfusions were compared between patients with limited and extensive transfusion (0-3 vs. C4 units) using the Student’s t test. Results: 75 patients, 57 females, mean age 50.1y (SD19.5), who underwent 88 surgeries, met inclusion criteria. Mean surgical time was 405.5min (SD141.8); mean number of fused levels was 10; intraoperative blood loss was 1988.5ml (SD1265.64) representing a mean EBL of 45.13 % (SD27.1). In 24.1 % of the patients, EBL was over 80 % of their estimated blood volume. Preoperative fibrinogen did not correlate with EBL (p = 0.747, r = -0.035); transfusion requirements (p = 0.47) did not correlate either. However surgical time correlated with more extensive intraoperative EBL (p \ 0.0001), and EBL correlated with more postoperative complications (p = .0081). Conclusion: Contrary to what has been found in adolescents, preoperative fibrinogen plasma concentration is not a good predictor of intraoperative bleeding and transfusion requirements in ASD surgery. A reliable variable to predict excessive bleeding and avoid associated complications in ASD surgery is still needed. Disclosures: author 1: none; author 2: grants/research support; Company = Depuy Synthes; author 3: grants/research support; Company = DePuy Synthes, K2M, consultant; Company = DePuy Synthes, Biomet; author 4: none; author 5: grants/research support; Company = DePuy-synthes; author 6: grants/research support; Company = DePuy Synthes, consultant; Company = De Puy Synthes; author 7: grants/research support; Company = Depuy Synthes, consultant; Company = Stryker; author 8: grants/research support; Company = Medtronic, Depuy Synthes, stock/shareholder; Company = IncredX; author 9: grants/research support; Company = DePuy Synthes; author 10: grants/research support; Company = DePuy Synthes.

Eur Spine J (2015) 24 (Suppl 6):S711–S742 QF67 THE EFFECT OF LUMBAR DISC DEGENERATION ON POSITIONAL CHANGES IN THE LUMBAR LORDOSIS: A CROSS-SECTIONAL COMPARISON WITH HEALTHY CONTROLS Bjarke Brandt Hansen, Tom Bendix, Jacob Juel Grindsted, Henning Bliddal, Robin Christensen, Philip Hansen, Robert C.G. Riis, Mikael Boesen The Parker Institute, Department of Rheumatology, Copenhagen University Hospital, Frederiksberg, Denmark Purpose: To examine the influence of lumbar disc degeneration (LDD) on the lumbar lordosis in weight-bearing positional magnetic resonance imaging (pMRI). Background: It is generally assumed that age related LDD leads to a flattened lumbar-lordosis. Methods and Materials: Patients with a low back pain (LBP) score above 40 on a 0-100 mm visual analogue scale (VAS) both during activity and rest; and a sex and age-decade matched control group without LBP were scanned in the supine and standing position in a 0.25 T open MRI unit (G-Scan). All images were evaluated and LDD was graded by Pfirrmann’s LDD classification on a scale from 1 to 5. Subsequently, the L2-to-S1 lumbar lordosis angle was measured. Results: Thirty-eight patients with an average VAS of 58 (±13.8) mm during rest and 75 (±5.0) mm during activity and 38 healthy controls were included. MRI disc changes were common in both groups, whereas, the summation of the lumbar Pfirrmann’s grades (LDD score) was significantly higher in the patients (Mean Difference (MD): 1.44, CI:0.80- 2.10; P \ 0.001). No difference was found for supine-to-standing lordosis angle changes (MD: 0.8o, CI: -1.8 to 3.3; P = 0.57) between groups. In patients, the LDD scores were not correlated with the lordosis angle neither in standing nor supine position, but for the controls LDD did correlate with the supine-tostanding lordosis angle change (Spearman: r = -0.627, P \ 0.001). Conclusion: Lumbar disc degeneration was not associated with lordosis angle in patients with LBP, neither in supine nor in standing position. In individuals without LBP, LDD scores were inversely correlated with the lordosis-angle change from supine to standing position, suggesting that age-related disc degeneration may lead to increased lumbar stiffness. Disclosures: author 1: none; author 2: none; author 3: none; author 4: none; author 5: none; author 6: grants/research support; Company = Danish Research Council; author 7: none; author 8: consultant; Company = Non-paid Research adviser Esaote ltd, Genoa Italy, other financial report; Company = Travel support for ESSR 2010, 2014 and ECR 2014, Esaote ltd, Genoa Italy.

QF68 THE ACCURACY OF THORACIC PEDICLE SCREW PLACEMENT IN SPINAL DEFORMITIES WITH THE USE OF SPECIFIC RAPID PROTOTYPING 3D TEMPLATES Janez Mohar, Robert J. Cirman Department of Spine Surgery and Paediatric Orthopaedics, Valdoltra Orthopaedic Hospital, Ankaran, Slovenia Introduction The anatomy of the deformed thoracic vertebrae makes the pedicle screw placement with established techniques imprecise, possibly dangerous. The literature review shows an average of 11 %

S739 pedicle wall violations by screws using the free hand technique, increased accuracy with the use of navigation over fluoroscopically guided technique and a recommendation of navigation over the free hand technique when placing pedicle screws in the mid thoracic deformed spine. A prospective usability study has been conducted in order to assess the accuracy of a novel technique of thoracic pedicle screw placement in spinal deformities with the use of specific rapid prototyping 3D templates. Methods The technique was firstly tested on plastic bone models, followed by cadavers and lastly, patients. A CT scan of thoracic spine was performed and its DICOM files were exported to a software application where the screw entry point, its trajectory and its diameter as well as length, were planned for each instrumented vertebra. After an ideal position of the screw had been planned, a 3D printing of a biocompatible photopolymer template, designed as to tightly fit the posterior elements of the vertebra, followed. A pedicle screw was inserted with the guidance aid of a specific 3D template, incorporating a pin hole defining the screw entry point and trajectory. A screw position was measured on a postoperative CT scan and compared to the planned screw position. For each instrumented pedicle the Zdichavsky pedicle breach grade was evaluated. Results 85 screws were placed as described in 15 cases of deformed thoracic spine. Compared to the preoperative planning, the average instrumented screw entry point offset on X and Y axes were 0.73 mm and 0.95 mm, respectively. The average difference in trajectory between planned and actual screw on the sagittal and transverse planes were 2.76 ° and 2.66 °, respectively. According to Zdichavsky classification, 80 screws (94 %) were grade I (optimal placement), 5 screws (6 %) were grade II (fair placement) and there were no grade III screws (dangerous placement). Conclusion The guidance aid of specific rapid prototyping 3D templates shows high accuracy of the thoracic pedicle screw placement in spinal deformities. Further high quality studies are needed to assess the technique’s pedicle screw placement accuracy and usability compared to established techniques. Disclosures: author 1: none; author 2: none.

QF69 PROXIMAL JUNCTIONAL FAILURE WITH NEUROLOGICAL DEFICIT FROM A RETROSPECTIVE DATABASE Charles Sansur, Andrew Frost, Niall Craig, John Schmidt, Jennifer McCool, Megan Dumas University of Maryland Medical Center, NHS Grampian, Leesburg, USA Introduction: A literature review found only one small 7 patient study with neurological deficits associated with PJK [1] . We believe this is a first of its kind study to specifically examine the rates of neurological complications in a large cohort of patients diagnosed with PJK. Methods: Data from 2344 patients was entered either prospectively or retrospectively into a multi-center database from 2009-2013. 48 separate sites entered data including 5 overseas sites. Patients were included in this study if they were diagnosed with PJK. Diagnosis was made by the implanting surgeon and did not include a set number of degrees. Results: 554 patients within the DB had a diagnosis of PJK at one or two levels from the UIV. This included 275 adolescents and 279 adults. 103 patients (18.6 %) were also diagnosed with a neurological deficit. 67 patients were revised.

123

S740 The most common complications reported were 56 right and 58 left bicep reflex ratings of either a 4 (very brisk, hyperreflexive, with clonus) or 5 (sustained clonus); 13 patients had abnormal bladder function, 19 had abnormal bowel function, and 17 were diagnosed with decreased anterior thigh response. Conclusion: Within our database, 418 patients of the 2344 (18.7 %) exhibited some form of neurological deficit. Our analysis also shows that the percentage of patients with a neurological deficit who also have PJK is 18.6 %. Therefore this cohort showed no difference in neurological events from that of the general spine patient population. References 1. McClendon, J., Jr., et al., Techniques for operative correction of proximal junctional kyphosis of the upper thoracic spine. Spine (Phila Pa 1976), 2012. 37(4): p. 292-303. Disclosures: author 1: consultant; Company = Depuy synthes; author 2: none; author 3: grants/research support; Company = Cerapedics, consultant; Company = Q Spine; author 4: employee; Company = K2M; author 5: stock/shareholder; Company = K2M, inc., employee; Company = K2M, inc.; author 6: employee; Company = K2M, inc.

Eur Spine J (2015) 24 (Suppl 6):S711–S742 the AIS patients is 17 times higher than expected compared to the age-matched Danish population. The pathology of cancers among the AIS patients are different from a Nordic cohort of Airline Pilots, exposed to an equal radiation dose during their adult period of life. Disclosures: author 1: grants/research support; Company = Bevica Foundation; author 2: grants/research support; Company = Danish Rheumatism Association; author 3: none; ; author 4: grants/research support; Company = Norton Healthcare, AO Spine, OREF, SRS, other financial report; Company = OREF, NIH, University of Louisville, Center for Spine Surgery and Research, Region of Southern Denmark, Children’s Tumor Foundation, Nuvasive, Medtronic, employee; Company = Norton Healthcare; author 5: none.

QF71 INCIDENTAL DUROTOMY: HOW TO INCREASE THE RATE OF SUCCESSFUL SINGLE STAGE CLOSURE Luca Papavero, Nils Engler Clinic for Spine Surgery, Schoen Klinik Hamburg Eilbek, Germany

QF70 INCIDENCE OF CANCER AND INFERTILITY, IN PATIENTS TREATED FOR ADOLESCENT IDIOPATHIC SCOLIOSIS 25 YEARS PRIOR Ane Simony, Steen Bach Christensen, Karl Erik Jensen, Leah Y. Carreon, Mikkel O Andersen Middelfart Hospital, Middelfart, Denmark Introduction: Adolescent females with idiopathic scoliosis are exposed to substantial amounts of radiation during treatment and follow-up for AIS.The purpose of this study was to determine the amount radiation exposure patients received during treatment for AIS and report the incidence of infertility and cancer in adulthood. Method: 219 consecutive AIS patients treated at Rigshospitalet, Copenhagen between 1983 and 1990 were identified and requested to return for clinical and radiographic examination. The incidence of cancer was determined through chart review and follow-up interviews. In addition, the subjects and age-matched controls were queried regarding infertility, age at first pregnancy and spontaneous abortions. Using X -ray reports that included patient position, mAs and kV used and the number of x-rays taken, a radiation physicist calculated the total radiation dose during treatment and follow-up adjusted for BMI and sex. Results: 159 (78 %) patients participated in the follow up study, and medical charts were available in 209 patients. 2 patients had passed away, one due to cardiac arrest and one to breast cancer. 8 patients had emigrated. Radiation information was available in 211 patients. The mean calculated mean total radiation exposure was 1.58 mSv (0, 44-6, 9).An average of 16.3 (range, 8-34) x-rays were taken during treatment. The rate of infertility (10 %) and spontaneous abortion (23 %) is similar to the normal controls. The AIS patients had 1.4 children, which is 1 child less than the 2.5 children in the control cohort. 9 (4.3 %) AIS patients developed cancer, mostly breast (3) and endometrial (4). The patients with endometrial cancers have a low BMI 16.4 (15.9-16.6) at first radiation exposure, and the patients with breast cancers were older with BMI 20.4 (18.0-22.4) The incidence of cancer in this cohort is 17 times greater than the incidence of 0.25 % in an age-sex matched cohort from the Danish Cancer Society Annual Report. Conclusion: The infertility and spontaneous abortion rate was similar between AIS patients and an age matched cohort. The cancer rate in

123

Introduction: Incidental durotomy (ID) is the most common complication of spine surgery. Intraoperative management is often based on a steep and emotional learning curve rather than on a structured scheme. Traditionally, ID closure has not been a structural part of the teaching program for spinal trainees at our institution. In 2010, a tenstep (10ST) guideline with the acronym ‘‘Bird Dove, MD’’ was introduced as a tutorial in the residency program. The aim of this retrospective cohort study was to compare the efficacy of the new surgical protocol (10ST) with the ‘‘individual’’ closure technique (InT) of the ‘‘old days’’. Methods: A database of 4020 consecutive surgeries for lumbar degenerative disease over the past 4 years was searched for ID. A total of 176 patients (100 females, mean age 58 years) enclosing 13 external referrals with persistent lumbar CSF-fistula were treated in 195 procedures. Several patients needed a second and few even a third look to get a successful closure of the ID. Six orthopedic surgeons and two neurosurgeons performed the intraspinal part of the surgery routinely with the aid of a microscope. Two dural repair techniques were compared: the ‘‘individual’’ technique (InT) and the 10ST.The 10 ST enclosed the following steps: 1) Bone removal until you see the whole dural tear; 2) Intradural look; 3) Repone the fibers; 4) Do an intradural patch if necessary; 5) Dural closure; 6) Outside patch, 7) Valsalva maneuver; 8) Epidural pedicled muscle flap, 9) Multilayer closure; 10) Drainage of CSF if necessary. In 149 procedures, the surgeons repaired the ID with the InT with which they felt most confident. The 10ST was applied in 46 procedures. Results: The overall prevalence of ID was 4.4 %. The prevalence was lowest in virgin micro-diskectomies (1.7 %) and ranged from 3.6 % in decompression for spinal canal stenosis up to 14.5 % in revision procedures. Among 107 primary surgeries, the InT achieved a single-stage closure of the ID in 96 procedures (89.7 %). Among 20 virgin surgeries, the 10ST was successful in all cases (p = 0.21). Among 42 revision surgeries for persistent CSF- leakage, the InT achieved single-stage closure in 36 procedures (85.7 %). The 10ST was successful in all 26 cases (p = 0.03). The follow-up was one year. Conclusions: In revision surgery for persistent CSF-leakage a structured treatment scheme (10 ST) improves the surgical performance. A cadaver-lab training before the ID occurs is recommended for trainees. Disclosures: author 2: employee; Company = Scho¨n Klinik Hamburg Eilbek, Clinic for Spine Surgery.

Eur Spine J (2015) 24 (Suppl 6):S711–S742 QF72 RELATIONSHIP BETWEEN PSOAS MUSCLE POSITION AND SPINO-PELVIC ALIGNMENT: AN ANATOMICAL ANALYSIS USING MRI FOR LATERAL INTERBODY FUSION SURGERY. 1

Akira Kondo, 2Yuichiro Abe, 1Kentaro Yamada, Takanobu Otsuka, 1Shigenobu Sato

2

1

Dept. of Orthopedics Surg. Eniwa Hospital, 2Dept. of Orthopedics Surg. Nagoya City University, Japan

In the lateral interbody fusion (LIF) procedures, approaching path passes anterior to or through the psoas muscle. To prevent the neural injury during the procedures, the anatomy of the lumbar plexus, which exists within the psoas muscle, has been reported in numerous publications. There are however, only few reports of the relative relationship between psoas muscle position and spinal alignment. In the present study, we analyzed the positional relationship between the psoas muscle and intervertebral discs and described the relationship between the psoas muscle position and sagittal spino-pelvic alignment. We included 100 cases who were taken standing whole spine radiography and lumbar magnetic resonance imaging (MRI). Using the axial view of the MRI, we measured the positions of the anterior and posterior psoas muscle rims. When the psoas rims were located anterior to the rim of intervertebral disc, the cases were defined as positive. The amount of overhang of the psoas muscle is expressed as the ratio to the anteroposterior disc diameter. We analyzed the correlations between the overhangs and sagittal spino-pelvic alignment (pelvic tilt [PT], pelvic incidence [PI], sacral slope [SS], and lumbar lordosis [LL]). The psoas anterior overhang at L2/3/4/5/S were -21, -12, 9 and 51 %, respectively. The posterior overhang at L2-S were -13, -11, 2 and 38 %, respectively. There were significant positive correlations between the anterior overhang and PI (r = 0.43), PT (r = 0.45), and PI-LL (r = 0.37), particularly at the L4/5 level. Furthermore, there were significant negative correlations between the posterior overhang and LL at the L2/3 and L3/4 levels. The psoas muscle originates from the lumbar transverse process and passes in front of the hip joint. Therefore, if the femoral head is located the more anterior, psoas muscle is located the more anterior. In this study, we demonstrated that high PI, high PT, and high PI-LL, all mean the situation that the femoral head is located more anterior of the spine, are associated with greater psoas anterior overhang. The LIF procedure might possibly be adapted to cases with imbalanced spino-pelvic alignment, high PI, high PT, or high PI-LL. In these cases, we should keep in mind that the psoas muscle is located more anterior. Disclosures: author 1: none; author 2: none; author 3: none; author 4: none; author 5: none.

QF73 CAN A FORMULA REALLY PREDICT THE THEORETICAL LUMBAR LORDOSIS RELIABILITY THROUGH A RADIOGRAPHIC STUDY IN 296 HEALTHY VOLUNTEERS Fe´thi Laouissat, Pierre Roussouly CMCR Massues-Croix Rouge Franc¸aise, Lyon, France Hypothesis and design: Prospective radiographic study of 296 healthy volunteers without spinal pathology. To assess the reliability of 2 formulas that aimed to predict tLL with the measured L1S1

S741 lordosis, and quantify general fluctuations in the sagittal alignment of the lumbar spine and pelvis. Introduction: L1S1 lordosis measurement is considered as a gold standard for assessing lumbar curve magnitude. Moreover, previous studies established different formulas to predict tLL. Enhancing their statistical significance, in an asymptomatic population, remains relevantMethod: PI, SS and the lumbar parameters L1S1 lordosis, inflexion point (InP), global LL and total number of lordotic vertebra (LL verteb) were evaluated in 296 healthy volunteers (126 male, 170 female; mean age, 27 years; range, 18-48 yrs). Two formulas (F1: tLL = 1/2(PI+TK)+10 and F2: tLL = PI+9) were compared to L1S1 using Bland-Altman Plot test. Comparison between the 4 types of sagittal spinal shape on the Roussouly classification used Student, ANOVA and Tukey tests for quantitative variables and chi, Fischer and Holm tests for qualitative variables. Results: Reliability between tLL and L1S1 was weak (Figure 1). Mean LL verteb was 2.9 for type 1, 4.2 for type 2, 4.5 for type 3, and 5.4 for type 4 (p \ 0.0001). The more proximal the InP to the thoraco-lumbar junction, the greater SS and PI. There were significant differences between the 4 types in terms of LL (51° for type 1, 48° for type 2, 58° for type 3, and 69° for type 4; p \ 0.001) and of L1S1 (46° for type 1, 45° for type 2, 56° for type 3, and 67° for type 4; p \ 0.001). However, LL significantly differed between types 1 and 2 (p \ 0.04) while L1S1 did not (p = 0.7). Conclusion: Sagittal spinal alignment varied significantly in normal volunteers. Lumbar spine shape was assessed more precisely by global LL angle, by enhancing the difference in transition between lordotic and kyphotic curves in the thoraco-lumbar junction. A formula that strongly rely tLL and spino-pelvic parameters is still lacking. Disclosures: author 1: none; author 2: consultant; Company = Kisco International, stock/shareholder; Company = SMAIO, royalties; Company = Medtronic.

QF74 COMPUTER-ASSISTED SPINOPELVIC TUMOR RESECTION USING ULTRASONIC OSTEOTOME INTEGRATED STEREOTACTIC NAVIGATION Justin Bird, Patrick Lin, Robert Satcher, Bryan Moon, Valerae Lewis MD Anderson Cancer Center, Houston, USA Objectives: Multiplanar osteotomies are complex surgical procedures often performed for en bloc resection of spinopelvic tumors. Conventional ostetomies are performed utilizing surface anatomy as a guide and can result in large variations between the planned and actual osteotomies performed. Computer-assisted navigation may serve as a helpful adjunct to complex osteotomy by registering an osteotome or drill to the navigation system. Recently, ultrasonic cutting devices have been introduced to spine surgery that are designed to perform osteotomies more safely than the conventional osteotome or drill. However, the feasibility of integrating the ultrasonic osteotome with 3D navigation and its usefulness in spinopelvic tumor resections remain unknown. Materials and Methods: A retrospective review was performed on all surgical cases in which the resection technique employed the use of an ultrasonic osteotome integrated into a stereotactic navigation platform. The primary outcome measures were 1) ability to perform the desired osteotomy 2) technique-related operative complications and 3) margin status. Results: 14 cases of en bloc tumor resection involving the spine and/ or sacrum were identified for review. Osteotomies were performed in the following locations: 11 sacrum, 2 thoracic spine, and 1 cervical spine. In all cases the ultrasonic osteotome was successfully

123

S742 registered to the navigation system and integration with the stereotactic CT-guided navigation system provided real-time visualization of the trajectory and location of the ultrasonic osteotome tip. Osteotomies were performed safely in all cases without complication. Osteotomies were limited to 20mm cutting depth, the working length of the osteotome tip. In 4 cases, osteotomies greater than 20mm were required. In these cases the osteotomies were initiated with the navigated ultrasonic osteotome and completed with a traditional ostetome. Surgical complications unrelated to the use of the described technique were noted in 5 cases: durotomy (2), wound complication (3), postoperative hemorrhage (1), partial nerve root laceration (1). Microscopic margins were positive in 2/14 cases (sacrum: 1 bone margin, 1 soft tissue). Conclusions: Integration of the ultrasonic osteotome with the stereotactic CT-guided navigation system appears safe and technically feasible for performing complex spine and pelvic osteotomies for en bloc tumor resection. Disclosures: author 1: consultant; Company = DePuy Synthes Spine, BrainLab; author 2: none; author 3: none; author 4: none; author 5: none.

QF75 INCIDENCE AND RISK FACTORS FOR VENOUS THROMBOEMBOLISM AFTER SPINE SURGERY: A PROSPECTIVE COHORT STUDY Hideki Murakami, Satoru Demura, Satoshi Kato, Moriyuki Fujii, Takashi Igarashi, Noritaka Yonezawa, Hiroyuki Tsuchiya Department of Orthopaedic Surgery, Kanazawa University, Kanazawa, Japan Background: There have been no comparative studies on venous thromboembolism (VTE) after spinal surgery with respect to screening patients for both deep venous thrombosis (DVT) and pulmonary thromboembolism (PE). The goal of this study is to

123

Eur Spine J (2015) 24 (Suppl 6):S711–S742 demonstrate the incidence and identify the associated risk factors of VTE after spine surgery. Patients and Methods: A total of 604 patients who underwent elective spine surgery between 2007 and 2012 at the authors’ institution were considered for inclusion in this study and divided into 7 groups. : Group 1) 108 patients with cervical degenerative disease treated with laminoplasty; Group 2) 59 patients with cervical degenerative disease treated with cervical fusion surgery; Group 3) 93 patients with thoracic/lumbar degenerative disease treated with three levels or more posterior fusion surgery; Group 4) 79 patients with lumbar degenerative disease treated with one level posterior lumbar interbody fusion (PLIF); Group 5) 120 patients with lumbar degenerative disease treated with decompression surgery without fusion; Group 6) 53 patients with spinal tumor treated with piecemeal excision with stabilization; Group 7) 92 patients with spinal tumor treated with total en bloc spondylectomy. DVT and PE screening was performed for all 604 patients 7-10 days after surgery. The binomial logistic regression analysis was used to assess the association of risk factors. Results:. The overall incidence of VTE was 9.3 % (64/604 patients). 1.8 % (11/604 patients) showed PE, and no DVT was found in 8 of these 11 (72.7 %). The incidence of VTE was 2.8 %, 3.4 %, 10.8 %, 10.1 %, 12.5 %, 15.1 %, and 19.6 % in group 1, 2, 3, 4, 5, 6, and 7, respectively. All five VTE-positive patients had only distal DVT without proximal DVT or PE in the cervical spine surgery (Group 1, 2). On the other hand, 10 (55.6 %) of 18 patients had proximal DVT or PE in group 7. The statistical analysis showed that female (P = 0.007) and advanced age (P = 0.009), spinal tumor (P \ 0.001), operative time (P = 0.018), nerurologic deficit (P = 0.001), duration of postoperative bed rest (P = 0.003) were risk factors for VTE. On the other hand, cervical spine surgery was the low risk of VTE (P \ 0.001). Conclusion:. The current study demonstrates that the rate of VTE varies with different spinal levels, surgical procedures, and pathologies. In particular, cervical spine surgery has a low risk attached to it. Spinal tumor surgery carries a high risk of critical VTE. Disclosures: author 1: none; author 2: none; author 3: none; author 4: none; author 5: none; author 6: none; author 7: none.