Eur Spine J (2017) 26 (Suppl 2):S335–S405 DOI 10.1007/s00586-017-5225-1

ABSTRACTS

EUROSPINE Meetings 2017: e-Poster Abstracts EUROSPINE 2017 Scientific Programme Poster presentations

Conclusion: These results suggest that this CaP coating could reduce the production of specific inflammatory cytokines that enhance fibrous tissue formation, while enhancing osteoblastic differentiation on PEEK implants. This biologic response to coated PEEK and CF-PEEK potentially translates to a reduced potential for the ‘‘halo effect’’. Disclosures: Author 1: stock/shareholder: Tissue Regeneration Systems, Inc., employee: Tissue Regeneration Systems, Inc.. Author 3: stock/shareholder: Tissue Regeneration Systems.

BASIC SCIENCE, BIOLOGY

P1 BIOMINERAL CAP COATING REDUCES INFLAMMATORY BEHAVIOR OF PEEK IMPLANTS Leena Jongpaiboonkit, Nicholas Caras, William Murphy Tissue Regeneration Systems, Inc., Plymouth, USA Introduction: Prior research has suggested that the peri-inflammatory response around PEEK implants, also described as a ‘‘halo effect’’, may be due to the stimulation of specific inflammatory cytokines that enhance fibrous tissue formation [Olivares-Navarrete R, Spine, 2015 40(6):399–404]. The purpose of this study was to determine if a nanostructured calcium phosphate (CaP) coating on PEEK cylinders has the potential to reduce production of inflammatory mediators, and enhance production of anti-inflammatory mediators. This in vitro study examined hMSCs behaviors on PEEK and carbon fiber reinforced (CF) PEEK substrates with and without the nanostructured CaP coating. Specifically, we examined (1) osteoblastic differentiation and (2) secretion of inflammatory interleukins. Methods: A nanostructured CaP coating was applied to PEEK and CFPEEK cylinders through a proprietary biomimetic precipitation process. Materials were subjected to surface pre-treatments and subsequently incubated in a pH adjusted mineral solution. An incubation protocol was followed for 4 days resulting in a continuous, plate-like nanostructure coating that provided high surface area favorable for cell attachment (image 1). Human mesenchymal stem cells were cultured on tissue culture polystyrene, PEEK and CF-PEEK with and without CaP coating. Osteoblastic differentiation and specific inflammatory cytokines that enhance fibrous tissue formation were assessed after 5 days. Results: The uncoated PEEK cylinders demonstrated relatively high levels of pro-inflammatory cytokines, specifically high levels of IL-1b (associated with fibrous tissue formation), and IL-1b, IL-6, and IL-8 (associated with chronic inflammation). These presence of the mineral coating eliminated the increases in pro-inflammatory cytokines. In addition, the coated PEEK cylinders promoted enhanced osteoblastic differentiation in comparison to uncoated PEEK cylinders.

P2 DRUG X CAN INHIBIT ANTI-APOPTOTIC CELL DEATH INDUCED BY OXIDATIVE STRESS IN VITRO AND IMPROVE LOCOMOTOR FUNCTION OF LOWER LIMBS IN SPINAL CORD INJURY MODEL MICE IN VIVO Mikito Tsushima, Shiro Imagama, Kei Ando, Kazuyoshi Kobayashi Department of Orthopaedic Surgery, Nagoya University, Nagoya, Japan Introduction: Drug repositioning is a notable strategy that comprehensive screening of drugs enable to be clinically indicated for noncurable diseases. Thus, it is a way provided low cost and short term because drugs had been used generally and properly, and searched side effects in detail. The object of this study is to detect some drugs which we will be able to use clinically for spinal cord injury (SCI), by the drug repositioning strategy. Methods: In this study, we screened 1186 FDA-approved drugs to identify a clinically applicable small molecular medical compounds

123

S336 that inhibit apoptotic cell death of cultured motor neurons and primary motor neurons (PMNs) from the spinal cord of fetal mice, induced by oxidative stress. Firstly, NSC-34, a hybrid cell-line of spinal motor neurons of mice, were treated by the drug library, and secondarily provided by hydrogen peroxide (H2O2). In this way, we led NSC-34 to be dead and searched absorbance of living neuronal cells treated by each chemical compound comprehensively by MTS assay kit. Next, we introduced PMNs to cell death in the same way. We evaluated living cells and narrowed down several targeting drugs. Furthermore, as an apoptotic trial, we evaluated the number of TUNEL-positive cells induced by oxidative stress after drug treatment in NSC-34 and PMNs. Thus, we made SCI-contusion model mice, and administered the drug for 4 weeks and measured the BMS score at any time. Result section: We detected ten drugs which inhibited cell death by the screening for several times in NSC-34. Additionally, we performed the same screening protocol more than ten times and choose Drug X which showed effective inhibition of cell death in PMNs significantly. The number of TUNEL positive cells was decreased after Drug X treatment compered to after control (DMSO). The BMS score of SCI model mice was measured and analyzed in time series, improvement of motor function was significantly observed in the drug administration group and the score of the administration group was 3.1, compared with the score of non-administration group, 1.2 (n = 5, p \ 0.01). Discussion: These results showed that drug X may have decreased cell death in primary motor neurons by anti-apoptotic effects resulted from oxidative stress. Drug X have been generally used in other areas and have been established safety for human. We hope that it will be clinically curable drugs for neuronal disorders, such as degenerative or ischemic diseases, as well as traumatic diseases. Significance: Drug X can inhibit anti-apoptotic cell death induced by oxidative stress in vivo trial of spinal cord injury of mice. Drug X is used widely and clinically as a safe drug, we expect to add further applications for SCI to Drug X. Disclosures: Author 1: none; Author 2: none; Author 3: none; Author 4: none;

Eur Spine J (2017) 26 (Suppl 2):S335–S405

BASIC SCIENCE, BIOMECHANICS

Design: Cadaveric study and finite element analysis (FEA). Introduction: PJK is not uncommon and can result in significant complication after spinal deformity surgery. Increased mechanical stress at the proximal junction due to the transition of a fixed to a mobile spine segment, along with osteoporosis have been suggested as risk factors for PJK. We utilized a tapered dose of cement in T10, T9, T8 to reduce the rate of proximal junctional fractures. Methods: Fifteen ligamentous, osteoporotic T6-pelvis specimens with screw/rod instrumentation from T10 to S1 were divided equally into three groups: group 1—no cement; group 2–4cc of cement in T10 and 4 cc in T9 (2cc through each pedicle); and group 3–4cc of cement in T10 (UIV), 3cc total in T9 (UIV + 1), and 2cc in T8 (UIV + 2). The pelvis and T6 vertebra were potted and compression was applied 10 mm anterior to the center of T6 using an MTS actuator. Maximum load to failure was measured in Newtons (N). The spines were evaluated using fluoroscopy and CT. Results: There was a significant reduction in fractures in group 3 vs 2 and 1 (0 vs 5 vs 5, p = 0.0019, respectively). Posterior ligamentous rupture occurred in 4 specimens in group 3; 3 in group 2; and 1 in group 1. The mean peak load-to-failure values showed an increasing trend from groups 1 to 3 (p = 0.38). There was no difference in specimen DEXA values (p = 0.71) and there was no hardware failure in any group. FEA mirrored the cadaveric data and the maximum load to failure increased from groups 1 to 3. Endplate stresses were reduced in group 3 vs 2 and 1. Conclusions: In both cadaveric and FEA models, tapering the dose of cement in the UIV, UIV + 1, UIV + 2 decreased endplate stresses, increased the load required for failure, and significantly reduced the incidence of vertebral fractures above long instrumented constructs. This technique may protect the spine from PJK due to fracture, but may increase the risk of posterior ligamentous failure. Disclosures: Author 1: none; Author 2: consultant: Rehab, Inc.; Zavation,other financial report: AO Foundation; Author 3: grants/ research support: AO Foundation, DePuy Spine and NSF, stock/ shareholder: OsteoNovus, Spinal Balance, royalties: Globus; Author 4: grants/research support: AO Spine, consultant: Depuy Synthes, Zimmer Biomet, Stryker, stock/shareholder: Surgical Solutions, Vivex, royalties: Zimmer Biomet.

P3 REDUCED RATE OF PROXIMAL JUNCTIONAL FRACTURES ABOVE LONG INSTRUMENTED CONSTRUCTS UTILIZING A TAPERED DOSE OF VERTEBRAL CEMENT. A BIOMECHANICAL STUDY

P4 IN VITRO ANALYSIS OF THE SEGMENTAL FLEXIBILITY OF THE THORACIC SPINE BY STEPWISE REDUCTION OF THE ANATOMICAL STRUCTURES

David Briski, Robert Mcquire, Vijay Goel, Joseph Zavatsky

Christian Liebsch, Stefan Grundler, Claudia Ottardi, Benedikt Schlager, Hans-Joachim Wilke

Ochsner Medical Center, New Orleans, LA, USA; Spine and Scoliosis Specialists, Tampa, FL, USA; University of Toledo, Toledo, OH, USA

Institute of Orthopaedic Research and Biomechanics, Ulm, Germany

Summary: Fifteen T6-pelvis specimens with screw/rod instrumentation from T10 to S1 were divided into three groups: group 1—no cement; group 2—equal cement in T10 and T9; and group 3—tapered dose of cement in T10, T9, and T8; and loaded to failure. There was a significant reduction in fractures in groups 3 vs 2 and 1 (0 vs 5 vs 5, respectively). Finite element analysis (FEA) mirrored the cadaveric data and maximum load to failure increased from groups 1–3. Title: Reduced Rate of Proximal Junctional Fractures Above Long Instrumented Constructs Utilizing a Tapered Dose of Vertebral Cement—A Biomechanical Study Hypothesis: In a T10-pelvis instrumented model, a tapered dose of vertebral cement in T10, T9, T8 will reduce the rate of fractures at the proximal junction.

123

Basic knowledge about the human thoracic spine is still very limited. However, spinal diseases, such as scoliosis, as well as traumatic spinal injuries are often associated with the thoracic spine. Numerical models of the thoracic spine provide a valuable tool to investigate biomechanical factors in pathogeneses as well as new therapeutic options. To calibrate these models, experimental data regarding the biomechanics of all thoracic spinal structures are essential. Therefore, the purpose of this in vitro study was to examine the stabilizing effect of the single anatomical structures of the thoracic spine by their stepwise reduction. Six human thoracic functional spinal units of the segmental levels T2–T3, T6–T7, and T10–T11, respectively, with an average age of 56 years (50–65) were tested quasi-statically in a custom-built spine tester and monitored using an optical motion tracking system. While applying 2.5 Nm pure moments in flexion/extension, lateral bending,

Eur Spine J (2017) 26 (Suppl 2):S335–S405 and axial rotation, the range of motion (ROM) and neutral zone (NZ) of the motion segments were measured (1) in intact condition (IC), then after severing (2) the supraspinous ligament (SL), (3) the interspinous ligament (IL), (4) the ligamentum flavum (FL), and (5) the facet joint capsules (FC), (6) after resecting the vertebral arcs (VA), then after severing (7) the posterior (PLL) and (8) anterior (ALL) longitudinal ligament, as well as (9) after nucleotomy (NC) according to a previous study on the lumbar spine [1]. Statistical significant differences between the single resection steps were evaluated using the Friedman test and Mann–Whitney U test (p \ 0.05). ROM and NZ generally increased after each resection step for every motion segment, whereby the T6–T7 segments showed higher ROM than T10–T11, but lower ROM than T2–T3. NC had the biggest effect on spinal stability; the ROM increased up to 58% in lateral bending, 94% in axial rotation, and 213% in axial rotation relative to IC as well as significantly compared to its previous condition (ALL) for every spinal level and loading direction. PLL and ALL mostly influenced the flexion/extension flexibility, whereas the posterior elements (SL, IL, FL, FC, VA) had a low effect on ROM and NZ compared to PLL, ALL, and NC. The single anatomical structures of the thoracic spine have a stabilizing effect. Since the load-deformation characteristics did not change through all resection steps, the anulus fibrosus can be seen as determinant for the thoracic spinal motion and the ligaments and nucleus as a quantitative motion limiter. The results also indicate that the removal of the posterior elements is not adequate for the surgical correction of spinal deformities. The generated data is sufficient for the calibration of numerical models of the thoracic spine. References 1. Heuer F, Schmidt H, Klezl Z, Claes L, Wilke HJ (2007). Stepwise reduction of functional spinal structures increase range of motion and change lordosis angle. J Biomech 40(2):271–280. Disclosures: Author 1: grants/research support: German Research Council (DFG), Project WI 1352/20-1; Author 2: grants/research support: German Research Council (DFG), Project WI 1352/20-1; Author 3: grants/research support: German Research Council (DFG), Project WI 1352/20-1, none; Author 4: grants/research support: German Research Council (DFG), Project WI 1352/20-1; Author 5: grants/research support: German Research Council (DFG), Project WI 1352/20-1, none.

S337 (e) cage with internal fixator. The resulting range of motion (RoM) was normalised to the native RoM. During the flexibility tests, plain X-Ray images in maximal inclination and maximal reclination were performed to determine the CoR. Results: In lateral bending and axial rotation the devices had the least effect on the RoM (Fig. 1). In flexion/extension motion all instrumentations caused a significant reduction in RoM (p \ 0.011). The SMS device without and with the cage reduced the RoM of the native state to 58% (SD 17.8) and 53% (SD 15.7), respectively. With the X-plate the RoM was further reduced to 27% (SD 16.8). The highest reduction of RoM to 17% (SD 17.2) was seen for the cage with additional pedicle screws. Compared to the native state the COR for the all instrumentations showed a small shift towards cranial. In anterior–posterior direction the COR for the SMS was comparable to the native state (central of vertebral body), whereas for the Xplate and the internal fixator the COR shifted to the posterior wall towards the instrumentation. Conclusion: As an add on for lumbar interbody fusion procedures the interspinous devices significantly reduced the RoM in flexion/extension, while in axial rotation and lateral bending only the internal fixator had a significant effect on the RoM. Further biomechanical and clinical evidence is required to investigate the effect of the changes in CoR and RoM in fusion procedures in patients. Disclosures: Author 1: none; Author 2: grants/research support: Q-spine; Author 3: none; Author 4: none; Author 5: grants/research support: St Teresa Medical, Q Spine, consultant: Q Spine, stock/ shareholder: Q Spine.

P5 FUNCTIONAL AND RADIOGRAPHIC EVALUATION OF AN INTERSPINOUS DEVICE AS AN ADJUNCT FOR LUMBAR INTERBODY FUSION PROCEDURES Anna Spicher, Werner Schmoelz, Rene Schmid, Hannes Stofferin, Niall J. A. Craig Department of Trauma Surgery, Tirol Kliniken, Innsbruck, Austria Introduction: In order to minimize surgical trauma for interbody fusions recently interspinous implants are discussed as an adjunct stabilization instead of pedicle screw instrumentation. The purpose of the proposed study is to evaluate the effect of the newly designed interspinous device with bands (SMS) on the range of motion (RoM) and the center of rotation (CoR) of a treated motion segment and to compare it with an established interspinous device as well as with pedicle screw instrumentation. Methods: Flexibility tests with pure moment loading of ± 7.5 Nm in the three main loading planes were conducted in six thoracolumbar monosegmental motion segments (mean age 56.2, BMD 96.1 mg/ ccm), in the following states: (a) native, (b) native with SMS device, (c) intervertebral cage with SMS device and (d) cage with X-plate,

P6 QUALITY AND QUANTITY OF MOTION OF A MOBILE AXIS OF ROTATION POLYCRYSTALLINE DIAMOND CERVICAL DISC PROSTHESIS AFTER 1- AND 2-LEVEL ARTHROPLASTY Leonard Voronov, Saeed Khayatzadeh, Robert Havey, Gerard Carandang, Kenneth Blank, Avinash Patwardhan Department of Orthopaedic Surgery and Rehabilitation, Loyola University Chicago, Maywood, IL, USA; Edward Hines Jr. VA Hospital, Hines, IL, USA Background context: Anterior cervical discectomy and fusion has been associated with the development of adjacent segment degeneration (ASD). Cervical total disc arthroplasty (TDA) is an alternative

123

S338 to fusion to prevent ASD, as studies have demonstrated that TDA can replicate physiologic motion. An innovative disc prostheses made of polycrystalline diamond, one of the hardest and most durable substances known, can replicate physiologic motion while minimizing wear debris. The purpose of this study was to assess the effect of cervical TDA on the motion of the human cervical spine in response to moments applied in flexion–extension, lateral bending, and axial rotation after 1- and 2-level arthroplasty. Materials and methods: Nine cadaveric C3–T1 specimens (mean age 38.3 ± 5.8 years) were tested. The kinematic testing apparatus allowed continuous cycling between specified maximum moment endpoints in flexion–extension, lateral bending, and axial rotation to ±1.5 Nm. Compressive preload (150 N) was used in flexion–extension. Vertebral motion was measured using a 3-D optoelectronic motion measurement system. A six-axis load cell was used to measure applied follower preload and moments. Specimens were tested: Intact, after C5–C6 TDA, and C6–C7 TDA. Repeated-measures analysis of variance (ANOVA) with Bonferroni correction was used for statistical analyses. Significance is shown by p \ 0.05. Results (Figure): 1-level TDA, results from C5 to C6, ROM () changed from: Flexion–extension: 12.8 ± 2.5 to 10.5 ± 2.1 (p = 0.03). Lateral bending: 8.5 ± 2.8 to 3.7 ± 1.0 (p \ 0.01). Axial rotation: 10.4 ± 1.1 to 6.2 ± 1.9 (p \ 0.01). Change in segmental stiffness (Nm/deg): Flexion: 0.09 ± 0.03 to 0.21 ± 0.09 (p = 0.004). Extension: 0.08 ± 0.03 to 0.18 ± 0.07 (p = 0.003). Change in neutral zone ():

Eur Spine J (2017) 26 (Suppl 2):S335–S405

P7 OPTIMAL CEMENT DOSAGE AND CONFIGURATION FOR PROPHYLACTIC VERTEBROPLASTY ABOVE LONG THORACOLUMBAR FUSION CONSTRUCTS TO REDUCE PROXIMAL JUNCTION KYPHOSIS (PJK): A FINITE ELEMENT STUDY Joseph Zavatsky, David Briski, Robert McGuire, Vijay Goel Ochsner Medical Center, New Orleans, LA, USA; Spine and Scoliosis Specialists, Tampa, FL, USA; University of Toledo, Toledo, OH, USA; University of Mississippi Medical Center, Jackson, MS, USA

Flexion–extension 1.8 ± 0.7 to 1.8 ± 0.8 (p = 0.966). 2-level TDA, results from C6 to C7, ROM () changed from: Flexion–extension: 10.0 ± 3.4 to 11.4 ± 3.0 (p = 0.07). Lateral bending: 7.5 ± 2.8 to 5.1 ± 2.3 (p = 0.07). Axial rotation: 7.7 ± 1.7 to 5.3 ± 0.9 (p = 0.02). Change in segmental stiffness (Nm/deg): Flexion: 0.13 ± 0.06 to 0.15 ± 0.08 (p = 0.424). Extension: 0.12 ± 0.05 to 0.11 ± 0.04 (p = 0.736). Change in neutral zone (): Flexion–extension 1.5 ± 1.0 to 2.1 ± 0.9 (p = 0.304). Conclusions: This disc prostheses restored ROM in flexion–extension to intact levels. In lateral bending the TDA maintained 68% of ROM at C6–C7 and 43% at C5–C6. In axial rotation 60% of the ROM was maintained at C5–C6 and 69% at C6–C7. All other biomechanically tested designs of TDA have shown similar reduction in lateral bending and greater reduction in axial rotation. The decrease in lateral bending and axial rotation after TDA may be a multifactorial phenomenon. Device kinematics, placement and tensioning of the remaining lateral annulus fibers during prosthesis insertion may play a role in maintained motion. Overall, the data suggest that this TDA provides similar cervical spine kinematics as compared to the preoperative condition. Disclosures: Author 1: grants/research support: Dymicron, Orem, UT, USA; Author 2: none; Author 3: none; Author 4: none; Author 5: none; Author 6: stock/shareholder: Spinal Kinetics, Sunnyvale, CA, USA.

123

Introduction: Long segment fusion is often utilized in the treatment of symptomatic spinal deformity. The incidence of proximal junction kyphosis (PJK) is high and the etiology is multifactorial, including vertebral fracture. There is a high incidence of osteoporosis in this patient population and prophylactic vertebroplasty (PVP) has been shown to reduce the risk of PJK. As demonstrated in our previous work, a tapered dose of prophylactic vertebral cement in UIV (T10), UIV + 1 (T9) and UIV + 2 (T8) in an instrumented T10–S1 model eliminated the incidence of fracture and PJK. The tapered dose of cement may buffer axial forces allowing a smoother load transfer through the segments. Our objective in this study, various combination of tapered bone cement dosage along with different locations of the bone cement were simulated in an effort to obtain the optimal bone cement dosage and its location for the prophylactic vertebroplasty. Optimizing volume and location of cement within the vertebra may further reduce endplate stresses observed at the proximal junction above long instrumented deformity constructs, thereby further reducing the risk of fracture, PJK, and revision surgery. Methods: A validated FE model from T6 to pelvis (Fig. 1) was used for the analyses. An osteoporotic model was developed and modified by insertion of screws and rods from T10 to S1, therein simulating the standard surgical procedure in silico. The nine different cases (Table 1) of tapered bone cement dosage were simulated in UIV (T10), UIV + 1 (T9) and UIV + 2 (T8) with the anterior, center and also the combination of anterior center location of bone cement. The load was applied to a metal bock 10 mm anterior to the center of the vertebra to simulate flexion moment and the pelvis was fixed (Fig. 2). The stresses at the end plates of T7–T10, as well as strains at the posterior ligaments were recorded to quantitatively evaluate the effect of different dosage and positions of bone cement.

Eur Spine J (2017) 26 (Suppl 2):S335–S405 Results: Increasing dosages of cement (cases 1–9) resulted in decreasing stresses at both the superior and inferior endplates of the cemented vertebra (T8, T9, and T10) (Figs. 2, 3) but resulted in increasing stress at the inferior endplate of T7. Although, anterior cement placement resulted in lower stresses observed in the superior and inferior endplates of the cemented T8, T9, and T10 vertebrae, compared to central placement, anterior placement resulted in a 4% increase in stress at the inferior endplate of T7. The % change in the stress between T7 inferior and T8 superior was lowest (30.8% for anterior (T8 stress \ T7 stress) and -7.6% (T8 stress [ T7 stress for center placement) for case 3 (T10-2.5cc, T9-2cc and T8-1 cc). The combination of anterior and center placement (T10-anterior, T9-anterior central and T8-center), resulted in low endplate stresses in both cemented (T10, T9, and T8) and non-cemented T7 vertebra. Disclosures: Author 1: grants/research support: AO Spine, consultant: DePuy Synthes, Zimmer Biomet, Stryker, stock/shareholder: Surgical Solutions, Vivex, royalties: Zimmer Biomet; Author 2: none; Author 3: consultant: Rehab, Inc, Zavation, other financial report: AO Foundation.

CERVICAL SPINE

P8 COMPARISON OF 7-YEAR RESULTS OF ONE-LEVEL VERSUS TWO-LEVEL CERVICAL DISC ARTHROPLASTY AND FUSION Matthew F. Gornet, Todd H. Lanman, J. Kenneth Burkus, Scott D. Hodges, Jeffrey R. McConnell, Randall F. Dryer, Francine W. Schranck, Anne G. Copay The Orthopedic Center of St. Louis, Chesterfield, St Louis, USA Background: Two independent clinical trials have concluded that cervical disc arthroplasty (CDA) is as safe and effective as anterior cervical discectomy and fusion (ACDF) for treating symptomatic cervical disc disease (SCDD) at 1 and 2 levels. Purpose: This study was to compare the safety and effectiveness at 7-year follow-up for the subjects treated with 1- and 2-level, respectively using CDA and ACDF. Materials and methods: Retrospective analysis of prospectively collected, combined data of 1 and 2-level FDA IDE clinical trials of a titanium ceramic composite cervical artificial disc. A total of 545 and 397 patients were studied in the 1- and 2-level trials, respectively: CDA (n = 280 and 209), ACDF (n = 265 and 188). The 84-month safety and effectiveness outcomes were compared between 1-level and 2-level CDA and ACDF, specifically: NDI, neck and arm pain (0–20 scale), SF36 PCS, neurological status, adverse events, and secondary surgeries at index and adjacent levels. The 1-level vs 2-level comparisons were done across studies, and a propensity score method was used to adjust for potential confounding effects and adjusted means were reported. Results: There were no preoperative differences between 1 and 2-level, respectively for CDA and ACDF patients for NDI, neck and arm pain, and SF-36 PCS scores. All patient groups significantly improved for NDI, neck and arm pain, and PCS scores from preoperative to 84-month. Comparison of 1 vs 2-level CDA: there were no significant differences between 1 and 2-level CDA for NDI improvement (38.2 vs 39.0, p = 0.768), neck pain (11.7 vs 12.3, p = 0.374), arm pain (11.3 vs 11.0, p = 0.736), SF-36 PCS (12.6 vs 14.5, p = 0.220), or proportions of patients who maintained neurological status (92.8 vs 91.6%, p = 0.867). The rate of secondary surgeries was numerically (but not significantly) higher for 1-than 2-level CDA at the index and adjacent levels (7.3 vs 4.2%, p = 0.566) and (11.6 vs 6.5%, p = 0.056), respectively. The rate

S339 of serious AEs was significantly higher for 1 than 2-level CDA (67.8 vs 56.7%, p = 0.004). Comparison of 1 vs 2-level ACDF: there were no significant differences between 1 and 2-level ACDF for NDI improvement (31.1 vs 31.6, p = 0.859), neck pain (9.7 vs 9.9, p = 0.796), arm pain (9.9 vs 10.1, p = 0.848), SF-36 PCS (10.8 vs 12.1, p = 0.424), proportions of patients maintaining or improving neurological status (79.7 vs 82.1%, p = 0.421), or rates of secondary surgeries at index levels (13.6 vs 14.7%, p = 0.631) or adjacent levels (10.9 vs 12.5%, p = 0.366). The rates of serious AEs were similar for 1 and 2-level ACDF (61.8 vs 68.2%, p = 0.200) but the rates of all AEs (94.5% vs 98.2%, p \ 0.001) and device-related AEs (18.9% vs. 27.7%, p = 0.036) were significantly lower for 1-level than 2-level ACDF. Conclusions: One and 2-level CDA appear to be equally safe and effective in the treatment of SCDD at 7-year. Two-level ACDF was equally effective as 1-level, but 2-level ACDF had a higher rate of device-related AEs. Disclosures: Author 1: consultant: Medtronic, K2M, stock/shareholder: Bonovo, International Spine and Orthopedic Institute, LLC, Nocimed, OuroBoros, Viscogliosi Bros Venture Partners LLC, royalties: Medtronic; Author 2: consultant: Medtronic, Nuvasive, royalties: Nuvasive, Stryker, Medtronic; Author 3: grants/research support: Medtronic, consultant: Medtronic; Author 4: grants/research support: Medtronic, consultant: Medtronic; Author 5: consultant: Globus Medical Inc., stock/shareholder: Globus Medical Inc., royalties: Globus Medical Inc.; Author 6: consultant: Innovasis, royalties: Globus, Nuvasive; Author 7: grants/research support: Predicted, Reported, and Observed Outcomes Foundation, stock/shareholder: Nocimed, other financial report: Medtronic (vendor); Zimmer/Biomet (vendor); Nocimed (vendor); Aesculap (vendor); Author 8: none.

P9 MICROENDOSCOPIC LAMINOTOMY VERSUS CONVENTIONAL LAMINOPLASTY FOR CERVICAL SPONDYLOTIC MYELOPATHY: A 5-YEAR FOLLOW-UP STUDY Akihito Minamide, Munehito Yoshida, Andrew K. Simpson, Hiroshi Yamada, Hiroshi Hashizume, Yukihiro Nakagawa, Hiroshi Iwasaki, Shunji Tsutsui, Masanari Takami, Yasutsugu Yukawa Department of Orthopaedic Surgery, Wakayama Medical University, Wakayama, Japan Background: Some problems after cervical expansive laminoplasty (ELAP) for cervical myelopathy have been reported, due to damage to the cervical posterior soft tissues, including muscles and ligaments. Well described potential complications of ELAP include persistent axial pain, restriction of neck motion, and loss of lordotic curvature. Therefore, minimally invasive spinal (MIS) decompression surgery, which can allow for multilevel segmental decompression while preserving the paraspinal muscles and the posterior elements, have been developed. One type of endoscope-assisted decompressive technique for cervical myelopathy is the cervical microendoscopic laminotomy (CMEL). The compression of the cervical spinal cord in cervical spondylotic myelopathy (CSM) consists of a pincer mechanism due to bulging disc and a hypertrophied ligamentum flavum. The concept of CMEL is to remove the dorsal compressive elements of the articular segment, including the ligamentum flavum and superior and inferior edge of the lamina. CMEL may provide some benefit with regard to maintaining lordosis and decreasing axial symptoms. Nonetheless, the long-term clinical benefits of CMEL in patients with CSM have not yet been elucidated. Therefore, the purpose of this study was to clarify the efficacy of posterior articular segment

123

S340 decompression by investigating the long-term clinical results of CMEL for patients with CSM, and to compare postoperative clinical and radiographic outcomes between ELAP and CMEL techniques. Methods: Consecutive patients with CSM who required surgical treatment were enrolled. All enrolled patients (n = 78) underwent CMEL or ELAP. All patients were followed postoperatively for greater than 5 years. The preoperative and 5-year follow-up evaluation included neurological assessment [Japanese Orthopaedic Association (JOA) score], JOA recovery rates, axial neck pain (VAS), SF-36 and cervical sagittal alignment (C2–7 subaxial cervical angle). Results: 61 patients (CMEL group: 31, ELAP group: 30) were included for analysis. The preoperative JOA score was 10.1 points in the CMEL group and 10.9 points in the ELAP group (p [ 0.05). The JOA recovery rates were similar, 57.6% in the CMEL group and 55.4% in the ELAP group (p [ 0.05). The axial neck pain in the CMEL group was significantly lower than that in the ELAP group (p \ 0.01). At 5-year followup, cervical alignment was more favorable in the CMEL group, with an average 2.6 gain in lordosis [versus 1.2 loss of lordosis in the ELAP group (p \ 0.05)] and lower incidence of postoperative kyphosis. Conclusions: CMEL is a novel, less invasive, technique that allows for multilevel posterior cervical decompression for treatment of CSM. Our 5-year follow-up data demonstrates that patients after CMEL have similar neurologic outcomes to conventional laminoplasty, with significantly less postoperative axial pain and improved subaxial cervical lordosis when compared to their traditional laminoplasty counterparts. 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.

P10 POSTOPERATIVE CHANGES OF CERVICAL SAGITTAL BALANCE IN PATIENTS WITH CERVICAL SPONDYLOTIC MYELOPATHY: A COMPARISON BETWEEN C3–6 LAMINOPLASTY AND C3–7 LAMINOPLASTY-

Eur Spine J (2017) 26 (Suppl 2):S335–S405 compressive myelopathy. However, there has been no report on the influence of the extent of decompression by laminoplasty (LP) on postoperative changes of CSB. The aim of this study was thus to compare postoperative changes of CSB after C3–6 LP for cervical spondylotic myelopathy (CSM) with those after C3–7 LP. Methods: 88 patients who had undergone LP for CSM since January 2010 (C3–6 LP, n = 56; C3–7 LP, n = 32) and were followed for at least 2 years after surgery were included in this study. Age at the time of surgery, gender, and the Japanese Orthopaedic Association (JOA) score before surgery showed no significant differences between the C3–6 and the C3–7 groups. Neurological outcome was assessed using JOA score at 2-year follow-up. The presence of postoperative axial neck pain was also examined. Cervical lateral X-ray images taken in the neutral standing position were evaluated preoperatively and at 2-year follow-up. The radiographic measurements included C2–7 angle and C1–C7 sagittal vertical axis (SVA). Results: The mean JOA score improved significantly from 11.0 points before surgery to 13.1 points at 2-year after surgery in the C3–6 group (mean recovery rate 40.3%), and from 11.2 points to 13.1 points in the C3–7 group (mean recovery rate 30.4%). Recovery rate of JOA score showed no significant difference between the 2 groups. The mean C2–7 angle significantly decreased from 10.7 before surgery to 8.1 at 2-year after surgery in the C3–6 group, and from 9.0 to 6.4 in the C3–7 group. Both preoperative and postoperative C2–7 angles were not significantly different between the two groups and did not significantly affect postoperative neurologic improvement. The mean C1–C7 SVA decreased from 28.0 mm preoperatively to 26.1 mm at 2-year after surgery in the C3–6 group (p = 0.23), on the other hand, it increased significantly from 30.4 to 39.6 mm in the C3–7 group (p = 0.006). The mean recovery rate of JOA score was 30.0% in the imbalance (postoperative C1–C7 SVA C30 mm) group and 41.9% in the balance (C1–C7 SVA\30 mm) group, respectively. Recovery rate of JOA score in the imbalance group tended to be lower than that in the balance group, although the difference was not significant. With respect to the presence of postoperative axial neck pain, there was no significant difference between the imbalance and the balance groups. Conclusions: C1–C7 SVA significantly increased after surgery in the C3–7 group, whereas it decreased in the C3–6 group. We presumed that preservation of the funicular section of nuchal ligament attached to the C7 spinous process which is a significant static cervical flexion limiter could prevent postoperative worsening of CSB in the C3–6 group. The mean recovery rate of JOA score in the imbalance group tended to be poorer than that in the balance group, although the difference was not significant. Disclosures: Author 1: none; Author 2: none; Author 3: none; Author 4: none.

P11 PROMIS PHYSICAL FUNCTIONING CORRELATION WITH NDI AND MJOA IN THE SURGICAL CERVICAL MYELOPATHY PATIENT POPULATION Robert Owen, Steven McAnany, Luke Zebala

Atsunori Ohnishi, Hironobu Sakaura, Yusuke Kuroda, Tetsuo Ohwada

Department of Orthopedic Surgery, Washington University in St Louis, St. Louis, USA

Department of Orthopaedic Surgery, Kansai Rosai Hospital, Amagasaki, Japan

Background context: Legacy patient reported outcome measures such as Neck Disability Index (NDI) and modified Japanese Orthopedic Association Score (mJOA) have become essential for analyzing treatments for cervical myelopathy. Significant associated administrative burdens impose limits on completion of these measures. The

Purpose: Several studies have demonstrated the importance of cervical sagittal balance (CSB) in surgical outcomes of cervical

123

Eur Spine J (2017) 26 (Suppl 2):S335–S405 Patient Reported Outcomes Measurement Information System (PROMIS) group developed a patient outcome measure in order to improve reporting of patient symptoms, function, and health and to reduce administrative burden. Despite early success in orthopedics, NDI and mJOA scores have not been compared with PROMIS in patients with spondylotic cervical myelopathy. The aim of this study is to compare NDI and mJOA with PROMIS to determine their correlations in a surgical patient population longitudinally. Methods: 61 patients with a diagnosis of cervical spondylotic myelopathy that went on to surgery were included in the study. All patients were treated at the same university spine center by four spine surgeons. PROMIS, NDI, and mJOA measurements were collected preoperatively, at 1–4 months and at 6 months. Data was extracted from a central Redcap database. Correlations between NDI, mJOA and PROMIS were quantified using Pearson correlation coefficient measurements. Students T-tests were used to demonstrate correlation significance with alpha set at 0.05. Results: All 61 (100%) of patients completed baseline preoperative questionnaires. 38 (62%) of patients completed all questionnaires at 1-4 months. 21 (34%) of patients completed all questionnaires at 6 months. The PROMIS physical function assessment and NDI demonstrated a strong negative longitudinal correlation, with Pearson r values of (-0.7, -0.75, and -0.68) at baseline, initial follow up and 6 months. The PROMIS physical function assessment and mJOA demonstrated a moderate positive correlation, with Pearson r values of (0.61, 0.67, 0.57) at baseline, initial follow up and 6 months. Students T test demonstrated a P value of \0.0001 for all Pearson correlation calculations. Conclusions: PROMIS physical function scores have a strong negative correlation with NDI scores both at baseline and in the early postoperative course in patients undergoing surgery for cervical myelopathy. PROMIS physical function scores have a moderate positive correlation with mJOA scores both at baseline and in the early postoperative course in patients undergoing surgery for cervical myelopathy. Surgeons may factor these outcomes into the delivery and interpretation of patient reported outcome measures in patients with cervical myelopathy, both at baseline and in the postoperative course. Use of PROMIS physical function assessments for this patient population may improve completion of outcome measures in the office and reduce administrative burden while still providing reliable outcomes data. Disclosures: Author 1: none; Author 2: none; Author 3: grants/research support: AO Spine, consultant: Medtronic, K2M, Depuy Synthes.

P12 INFLUENCE OF DIABETES MELLITUS ON SURGICAL OUTCOMES IN CERVICAL MYELOPATHY: A PROSPECTIVE, MULTI-CENTER STUDY Shinji Tanishima, Atsushi Tanida, Tokumitsu Mihara, Masaaki Murata, Tsugutake Morishita, Toshiaki Takahashi, Yasuo Morio, Yoshirou Nanjo, Toshiyuki Dokai, Hideki Nagashima Department of Orthopedic Surgery, Tottori University, Yonago, Japan Objective: To investigate the effects of diabetes mellitus on surgical outcomes in cervical myelopathy. Methods: This was a multi-center, prospective study. Subjects were patients who underwent surgery to treat cervical spondylotic myelopathy and ossification of the posterior longitudinal ligament (OPLL) from March 2012 to May 2015. There were a total of 61 cases (37 men and 24 women) with a mean age of 69.5 years. Surgical outcomes were evaluated preoperatively and 1 year postoperatively

S341 using the Japanese Orthopedic Association (JOA) score and JOA Cervical Myelopathy Evaluation Questionnaire (JOACEMQ). The investigation items were the fasting blood glucose (FBG) and glycated hemoglobin (HbA1c) measured preoperatively, and one week, one month and one year postoperatively; the F-wave conduction velocity, latency, rate of occurrence and M-wave latency in the ulnar and tibial nerves measured preoperatively and one year postoperatively. We also evaluated the disease duration, age, sex and presence or absence of diabetic treatment. Subjects were divided into a normal group (N group; 42 subjects) and a diabetic group (DM group; 19 subjects). We then evaluated the relationship between the surgical outcomes and each of the investigation items. Results: No significant difference was observed between the N and DM groups in terms of the conduction velocity in the peripheral nerves. The JOA score improved significantly in both groups, with improvement rates of 40.1% in the N group and 26.3% in the DM group, although no significant difference was noted between the two groups (p = 0.17). There was no correlation between the improvement in JOA score and the preoperative or postoperative blood glucose levels or HbA1c levels. There was no significant improvement noted in the JOACMEQ score, which evaluated cervical function, upper and lower limb function and bladder function in the two groups. We then divided the DM group into a group with normal FBG after 1 year (DMG group; 11 subjects) and a group with an FBG exceeding 140 mg/dL after one year (DMB group; 8 subjects), then compared the surgical outcomes in the two groups. The JOACMEQ score for upper limb function improved significantly in the DMG group (p \ 0.01). Compared to the DMB group, there was a non-significant increase in lower limb function scores in the DMG group (p \ 0.06). Worsening of upper limb and lower limb function were each defined as dependent variables, while glycemic control, disease duration, sex, body mass index (BMI) and age were defined as explanatory variables. We performed a logistic analysis and found that the disease duration was a factor for upper limb function in functional disorders of the upper limbs [p = 0.02, odds ratio 1.1, 95% confidence interval (CI) 1.11–1.02]. We also found that poor glycemic control may be a risk factor for functional disorders of the lower limbs (p = 0.07, odds ratio 1.1, 95% CI 1.0–1.1). Discussion: Improvement in lower limb function tends to be poor when glycemic control is poor. We believe that poor glycemic control may prevent functional recovery of the spinal cord. 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: grants/research support: Astellas Pharma Inc, Takeda Pharmaceutical Co.

P13 VERTEBRAL ARTERY AND OSSEOUS ANOMALIES CHARACTERISTIC AT THE CRANIOCERVICAL JUNCTION DIAGNOSED BY CT AND 3D CT ANGIOGRAPHY IN NORMAL CZECH POPULATION: ANALYSIS OF 511 CONSECUTIVE PATIENTS Petr Vanek, Ondrej Bradac, Renata Konopkova, Vladimir Benes Department of Neurosurgery and Neurooncology, Military University Hospital, Prague, Czech Republic; Charles University, First Medical Faculty, Prague, Czech Republic Introduction: There are numerous indications for stabilization using instrumentation of the upper cervical spine. This area is comprised of

123

S342

Eur Spine J (2017) 26 (Suppl 2):S335–S405

sophisticated anatomy. There is no study describing bony and vascular anomalies of this area in the middle European population. The main aim of this study was to investigate prevalence of any vertebral artery (VA) variations and osseous anomalies in the region of the craniocervical junction in a large sample of Czech patients based on three-dimensional computed tomographic angiography (3D CTA). Materials and methods: The VA has a variable course through C2 before it passes above its groove on the posterior arch of C1. The artery can course more medially, more posteriorly or more superiorly, thus limiting the diameter of the bony elements used as landmarks for the safe insertion of metalwork. This is known as a high riding VA (HRVA). The VA was considered HRVA in this study if the thickness of the C2 isthmus was less than 5 mm and/or the C2 internal height was less than 2 mm and/or the width of the C2 pedicle was less than 4 mm. The prevalence of ponticulus posticus (PP) was also identified. Following the VA variations in the V3 segment of the artery were persistent first intersegmental artery (FIA), fenestration (FEN) of the VA and the posterior inferior cerebellar artery (PICA) branch originating from the C1/2 part of VA. Results: Records of 511 patients from our institution were analysed. The mean age of the patients was 63.6 years. One hundred and twenty-three (24.1%) patients were identified to have HRVA, 30 (6%) present on both sides. The age of patient over 70 years and female sex were found to be significant risk factors for HRVA presence. The prevalence of a nearby PICA branch was 4%, FIA was 0.4% and FEN was 0.2%. The presence of PP was identified in 14.3% of patients. Conclusion: The HRVA and PP are common anomalies in the Czech population and routine preoperative high resolution CT evaluation is mandatory to prevent the VA injury when C1–C2 instrumentation is planned. The female sex and age over 70 years were found to be the most important factors for HRVA presence. The FIA and the FEN VA were rare in our study contrary to reports published from Asia, showing as many as a 10% the VA presence over the starting point for C1 lateral screw. On the basis of the infrequent occurrence of these anomalies, we don’t recommend routine CT angiography when upper cervical spine instrumentation in the normal population is planned. Disclosures: Author 1: none; Author 2: none; Author 3: none; Author 4: none.

Methods: Between 2002 and 2015, 52 patients who underwent thoracic OLF spine surgery was taken into study. Patients were observed for a minimum of 1 year with mean follow up period of 17.84 months (range 12–72 months). The recovery rate was calculated as follows: (postoperative JOA score - preoperative JOA score)/(11 - preoperative JOA score) 9 100. Results: Fifty-two patients with thoracic OLF induced myelopathy. There were 35 men and 17 women with mean age of 49.5 (range 30–75). The mean duration of symptoms between initial onset and diagnosis was 13 months (range 1 month–5 years). Pre operative JOA score is 5.8 (range 3–9).MRI and CT were performed on all patients. There were 11 patients (21.2%) decompressed at the upper thoracic spine (T1–T4) and 10 patients (19.2%) decompressed at the middle thoracic spine. Majority of the affected levels [31 patients (59.6%)] were lower thoracic spine (T9–T12). The number of decompressed levels were 1–3 in 44 (84.6%) patients and more than 3 in 8 (15.3%) patients. Intramedullary MRI T2 signal changes was observed in 20 patients (38.5%).Morphological classifications as determined by axial CT images were lateral in 11 (21.1%) patients, enlarged and extended in 8 (15.38%) patients, fused in 19 (36.53%) patients and 14 (26.92%) patients are of tuberous type. Associated cervical OPLL which doesn’t warrant surgery were in 15 patients (28.8%). Radiological signs tram track and comma sign of dural ossification was found in 11 (21%). The mean JOA score was 5.8 pre operatively and improved to 7.4 postoperatively at 1 year follow up. Mean JOA recovery rate was only 30.29% (range -20 to 83%). Good recovery rate ([50%) was found in 18 patients. 24 patients had poor recovery \50%). 7 patients remained unchanged. 3 patients had worsened from the preoperative status and remained the same. In all patients laminectomies was done. Intraoperatively, we found adhesions to duramater (ossification of duramater) in 19 (36.5%). Conclusions: The duration of preoperative symptoms and lower preop JOA score along with tuberous type OLF, dural ossification and T2 signal intensity in MRI were the risk factors in predicting poor outcome. Early diagnosis and sufficient surgical decompression improved the functional outcomes of DOLF patients. Disclosures: Author 1: none; Author 2: none; Author 3: none; Author 4: none.

DEGENERATIVE THORACOLUMBAR SPINE

P14 PREDICTIVE FACTORS OF POOR SURGICAL OUTCOME IN THORACIC OSSIFIED LIGAMENTUM FLAVUM; AN INDIAN INSTITUTIONAL STUDY

P15 ADJACENT SEGMENT DEGENERATION AND REVISION SURGERY RATES AFTER HYBRID FUSIONSTABILIZATION OF THE LUMBAR SPINE AS COMPARED TO CLASSICAL CIRCUMFERENTIAL FUSION. STUDY OF 2 COHORTS OF PATIENTS WITH DEGENERATIVE DISCOPATHY AT 2 AND 5-YEAR FOLLOW-UP

Prabu Raj Andi Perumal Raj, Indira Devi, Dhananjay Bhat, Dhaval Shukla

Carlos Barrios, Jose´ Ignacio Maruenda, Felipe Garibo, Juan Solaz, Borja Maruenda

Department of neurosurgery, NIMHANS, Bangalore, India Introduction: Dorsal ossification of the ligamentum flavum (DOLF) is a pathological condition that affects the ligament and causes slowly progressive myelopathy in adults. Because of long-standing compression of the spinal cord by OLF, a patient’s functional prognosis may not always be favorable, and attempts have been made in recent studies to identify clinical factors that are predictive of the surgical outcome of patients with thoracic OLF. Surgical decompression is the most common treatment of choice for patients with compressive myelopathy due to DOLF. However, the surgical outcome is not always satisfactory.

123

Valencia Catholic University, Valencia, Spain; University Clinic Hospital, Valencia, Spain; Hospital de la Ribera, Alzira, Spain Background/introduction: It has been suggested that hybrid dynamic fusion-stabilization of the lumbar spine could reduce the incidence of adjacent segment degeneration (ASD). Purpose of the study: This study compares the occurrence of radiological degeneration of the adjacent disc and the rate of revision surgery in two cohorts of patients with degenerative discopathy undergoing circumferential lumbar fusion (historical series) and hybrid posterolateral fusion with topping-off dynamic stabilization. Materials and methods: The historical cohort with circumferential fusion consisted of 54 patients (mean age 49 years) and the hybrid

Eur Spine J (2017) 26 (Suppl 2):S335–S405 fusion-stabilization cohort of 41 patients (mean age 52 years). In both cases, a maximum of two lumbar segments were involved. In the historical cohort a circumferential fusion with TLIF was used at the inferior level. Clinical and radiological outcomes were evaluated at 2 and 5-year follow-up. Changes in adjacent discs above the instrumentation were assessed using conventional radiographs (UCLA scale) and MRI (Pfirrmann grading). Results: In the cohort of 360 fusion, ASD appeared at the upper adjacent segment in 2.7% and in 18.6% at 2 and 5-year follow-up, respectively. In cases with hybrid fusion-stabilization, these figures were 2.4 and 7.3% of ASD at 2 and 5-year follow-up. Regarding re-interventions, in the historical cohort there was 1.4% in the first 2-year and 7.4% at 5-year follow-up. In the hybrid fusion-stabilization cohort these numbers were 0 and 4.8%, respectively. In the historical cohort with circumferential fusion, the causes of reoperations at 5-year follow-up were symptomatic ASD. However, in the cohort with hybrid fusion-stabilization cases, revision surgeries were caused by implant failures. Conclusions: Hybrid fusion-stabilization reduces ostensibly the development of ASD. The rate of reinterventions was lower using the hybrid system, and was caused by implant failures and not by ASD. Although monitoring of these two cohorts is relatively long, it seems necessary to confirm at long-term the benefits of dynamic fusionstabilization techniques. Disclosures: Author 1: none; Author 2: none; Author 3: none; Author 4: none; Author 5: none.

P16 PREDICTING PHYSICAL HEALTH-RELATED QUALITY OF LIFE USING NUMERIC, SEMANTIC AND GRAPHIC METHODS IN THE ASSESSMENT OF CHRONIC LOW BACK PAIN PATIENTS Christine Cedraschi, Orly Sarid, Michel Kossovsky, Jules Desmeules, Anne-Franc¸oise Allaz, Vale´rie Piguet Division of Clinical Pharmacology and Toxicology, Multidisciplinary Pain Centre, Geneva, Switzerland Purpose: The physical component summary (PCS) score of the SF36, a widely-used health-related quality of life (HRQoL) questionnaire is related to function in chronic low back pain (LBP) patients. This study aimed to examine the associations between PCS and LBP intensity as measured by visual analogue scales (VAS), pain drawing characteristics, pain words (PW), psychological and sociodemographic variables; and to predict PCS using these variables. Methods: Cross-sectional study design investigating chronic pain patients referred to the Geneva University Hospital Pain Center (2013–2015). Patients completed self-administered questionnaires: SF36 [Physical Component Summary (PCS) + Mental Component Summary (MCS) Scores], McGill Pain Questionnaire, VAS, Hospital Anxiety and Depression Scale, sociodemographic variables. The number of pain drawing (PD) characteristics (lines, arrows and crosses) was summed as an indicator of the patient’s depiction of the extent of pain. Full data were available for 82.8% (n = 236/285); only patients with LBP comprised the final sample: N = 109 (46.2%); 67.6% were women, 62.9% had B10 years of education; 88% were living with someone; 49.5% had French as their mother-tongue; mean pain duration was 10 years (SD ±13). Descriptive statistics were conducted for continuous variables. Associations between PCS PD characteristics, PW, VAS and psychological measures were performed with Pearson correlations. Linear regression analysis was performed with PCS as the dependent variable.

S343 Results: Sociodemographic variables (gender, age, education level, and living conditions) and pain duration were not related to PCS. PCS was positively correlated with PD characteristics (r = -.240, p = 0.026), PW (r = -.327, p = 0.003), HADS depression (r = -.539, p = 0.000), and VAS scores (r = -.429, p = 0.000). PCS and MCS were correlated (r = 0.80, p = 0.000), which indicated multicollinearity. The final model explained 41.0% of the variance in PCS with those using more Pain Drawing characteristics to describe their pain r = -.221; p = 0.027), expressing depression (r = -.443, p = 0.000) and higher VAS values (r = -.263, p = 0.009) displaying lower PCS scores. Conclusions: The physical component summary score of the SF-36 is impacted by graphic, psychological, and numeric aspects of pain assessment, pointing to the contribution of symbolic pain assessment to the understanding of patients’ appraisal of their ability to function. Disclosures: Author 1: none; Author 2: none; Author 3: none; Author 4: none; Author 5: none; Author 6: none.

P17 PREOPERATIVE EVALUATION OF LEFT COMMON ILIAC VEIN USING LUMBAR SPINE MRI IN OBLIQUE LATERAL INTERBODY FUSION AT L5–S1 Nam-Su Chung, Chang-Hoon Jeon, Han-Dong Lee, Young-Uk Seo Department of Orthopaedic, Ajou University Medical Center, Suwon, South Korea Purpose: We aimed to evaluate the anatomical configuration of left common iliac vein (LCIV) and the risks of mobilization of LCIV during oblique lateral interbody fusion (OLIF) at L5–S1. Introduction: The LCIV represents the vascular structure that is injured most frequently in the anterior approach to L5–S1, because it is often located medially and may obstruct the anterior surface of the disc. The influence of the anatomical position and morphological characteristics of the LCIV on the risks of mobilization of LCIV during the operation has not been comprehensively studied. Methods: Sixty-seven consecutive patients who underwent anterior lumbar fusion (ALIF, n = 39; OLIF, n = 28) at L5–S1 were included. The anatomical configuration of the LCIV at the L5–S1 disc on axial magnetic resonance images of the lumbar spine. The LCIV was classified into three types according to the difficulty of mobilization: type I (no requirement for mobilization; LCIV runs laterally for more than twothirds of the length of the left side of the L5–S1 disc), type II (easy mobilization; LCIV obstructs the L5–S1 disc space, but the perivascular adipose tissue is present under the LCIV), and type III (potentially difficult mobilization; no perivascular adipose tissue under the LCIV). The patient records were reviewed for vascular complications. Results: There were 23 men and 44 women in this study, with a mean age of 64.1 years (range 19–83 years). Type I LCIV configuration was found in 34 (50.7%) patients, type II in 18 (26.9%), and type III in 15 (22.4%). There were 7 (10.4%) patients with LCIV injury (type I; n = 0, type II; n = 2, type III; n = 5) (P = 0.003). Intra-observer reliability for the LCIV classification ranged from substantial to excellent and inter-observer reliability ranged from moderate to excellent. Conclusions: The type III LCIV configuration (no perivascular adipose tissue under the LCIV) showed a high rate of vascular injury. Preoperative LCIV evaluation, based on its position and difficulty of mobilization, is a reliable and valid means of predicting vascular injury during OLIF at L5–S1. Disclosures: Author 1: none; Author 2: none; Author 3: none; Author 4: none.

123

S344 P18 SPINOPELVIC SAGITTAL ALIGNMENT AFTER MINIMALLY INVASIVE DECOMPRESSION SURGERY WITHOUT FUSION IN PATIENTS WITH LUMBAR DEGENERATIVE SPONDYLOLISTHESIS Sho Dohzono, Hiromitsu Toyoda, Yusuke Hori, Shinji Takahashi, Akinobu Suzuki, Hidetomi Terai, Hiroaki Nakamura

Eur Spine J (2017) 26 (Suppl 2):S335–S405 P19 LUMBAR TOTAL DISC REPLACEMENT: PREDICTORS FOR LONG-TERM OUTCOMES Ha˚vard Furunes1,2,3, Christian Hellum1, Helga Brøgger1, Ivar Rossvoll4, Jens Ivar Brox1,2, Linda Berg5, Milada Cvancarova2, Ansgar Espeland5, Kjersti Storheim1 1

Department of Orthopaedic Surgery, Osaka University, Osaka, Japan Purpose: Spinopelvic sagittal balance is important in managing lumbar diseases and low back pain (LBP). Anterior translation of the C7 plumb line, high pelvic incidence (PI), and pelvic retroversion were reportedly marked in degenerative spondylolisthesis (DS) patients compared with a normal population. The purpose of this study was to evaluate the change in spinal sagittal alignment after decompression alone in patients with low-grade DS. Methods: We retrospectively reviewed the records of 87 patients (48 men, 39 women; mean age 69.2 ± 9.3 years) who underwent microendoscopic laminotomy. We enrolled 35 patients with DS and 52 patients without DS. Spinopelvic parameters were evaluated, including the sagittal vertical axis (SVA), lumbar lordosis (LL), sacral slope (SS), pelvic tilt (PT), and PI. Primary outcome was a change in spinopelvic alignment between the baseline and latest follow-up values (DS group versus non-DS group). Secondary outcomes were the relations between the improved global sagittal alignment and the preoperative spinopelvic parameters. Results: Both groups showed significantly alleviated LBP, leg pain, and leg numbness. There were no significant intergroup differences in the JOA score or VAS at the latest follow-up. Preoperative SVA and PI were significantly higher in the DS group than in the non-DS group (p \ 0.05). SVA significantly decreased and LL significantly increased in the DS group (p \ 0.05), whereas those parameters did not differ significantly from before to after surgery in the non-DS group (Figure). No significant differences existed between the groups for SS, PT, or PI—(=minus) LL. In both groups, the SVA improvement correlated significantly with the preoperative SVA (DS: r = 0.702, non-DS: r = 0.397). There was also a significant intergroup difference in the correlation coefficient (z = 1.98 r = 0.048). Discussion: In the present study, the results in the DS group, SVA and LL did not significantly improve in the non-DS group, indicating that the preoperative kyphotic compensation to avoid neuralgic symptoms is greater in DS patients than in those without DS. Our results suggest that global sagittal imbalance in patients with DS can be treated by decompression without requiring a corrective procedure in some patients whose malalignment was induced by compensatory reduction of the LL. Conclusions: SVA and LL significantly improved after microscopic laminotomy in patients with low-grade DS and neurologic symptoms. SVA improvement in the DS group was correlated with preoperative spinopelvic sagittal imbalance. It should be noted that the strength of those correlations was greater than for those in the non-DS group. Disclosures: Author 1: none; Author 2: none; Author 3: none; Author 4: none; Author 5: none; Author 6: none; Author 7: none.

123

Oslo University Hospital, Oslo, Norway; 2University of Oslo, Oslo, Norway; 3Gjøvik Hospital, Gjøvik, Norway; 4St Olav’s University Hospital, Trondheim, Norway; 5Haukeland University Hospital, Bergen, Norway Background: Total disc replacement (TDR) is a surgical option for selected patients with low back pain (LBP) and degenerative intervertebral discs when non-operative treatment fails. Data on predictors associated with favorable long-term outcome after TDR are still warranted. Purpose of the study: We aimed to identify predictors for favorable long-term outcomes after TDR that are easy for clinicians to assess prior to treatment. Materials and methods: We analysed a cohort of 82 patients with degenerative discs and chronic LBP who were treated with TDR and originally participated in a randomised trial comparing TDR and multidisciplinary rehabilitation. Patients randomised to surgery were analysed eight years postoperatively. In addition, 11 patients randomised to rehabilitation who crossed over and were treated with TDR (median time since surgery 72 (range 41–88) months) were included. Potential predictors were measured at baseline, and the outcomes assessed at 8-year follow-up. At baseline, median age was 41 (range 25–54) years, 49% were women, 24% had comorbidity, 32% were employed, 22% had only primary school, median Oswestry Disability Index (ODI) was 40 (range 28–70) points and median duration out of work was 12 (range 0–70) months. Outcome measures were dichotomised according to whether the participants had a clinically important functional improvement (15 points or more on ODI) (outcome 1) and to whether they were working at 8-year follow-up (outcome 2). Associations between potential predictors and outcomes were modelled using logistic regression. For outcome 2, the results were also organised in a prediction matrix and expressed as probabilities. Results: Of all assessed variables, only presence of Modic changes (type 1 and/or 2) was statistically significantly associated with outcome 1, patients with pre-treatment Modic changes had 5 times higher probability of achieving at least 15 points improvement on ODI compared to patients without Modic changes (OR = 5.04, 95% CI 1.39–18.21, p = 0.01). Variables significantly associated with outcome 2 were comorbidity (no/yes) (OR 2.38, 95% CI 1.43–13.78, p = 0,01), education level (higher education/only primary school) (OR 3.56, 95% CI 1.13–11.19, p = 0.03), ODI (\50 points/C50 points) (OR 3.60, 95% CI 1.04–12.48, p = 0.04) and duration of sick leave before treatment (\12 months/C12 months) (OR 4.14, 95% CI 1.62–10.59, p = 0.003). The probability of working at the last followup was 1% for patients with C12 months sick leave, comorbidity, ODI C50 and low education prior to treatment and 87% for patients with \12 months sick leave, no comorbidity, ODI \50 and higher education (Fig. 1). Conclusion: Patients with Modic changes prior to the TDR surgery were more likely to report a clinically important functional improvement at long-term follow-up. Patients with lower ODI-score, no comorbidity, higher education and less time out of work prior to surgery were more likely to work at follow-up. Disclosures: Author 1: none; Author 2: none; Author 3: consultant: Curato, Norway; Author 4: none; Author 5: none; Author 6: none; Author 7: none; Author 8: none; Author 9: none.

Eur Spine J (2017) 26 (Suppl 2):S335–S405

S345

Figure 1. Prediction matrix

≥ 12 months sick leave < 12 months sick leave

ODI ≥ 50 ODI < 50 ODI ≥ 50 ODI < 50

Only primary school Comorbidity No comorbidity 1% 9% 4% 25 % 7% 38 % 22 % 68 %

Higher education Comorbidity No comorbidity 4% 24 % 12 % 52 % 20 % 67 % 47 % 87 %

Probability of working at eight-year follow-up after total disc replacement based on a probability matrix model including education level, comorbidity or not, duration of sick leave before treatment (\12 or C12 months) and Oswestry Disability Index (ODI \50 points or C50 points).

P20 LONG-TERM OUTCOME OF POSTEROLATERAL FUSION WITH DYNAMIC TRANSPEDICULAR TOPPING-OFF FOR THE ADULT DEGENERATIVE LUMBAR SPINE. REPORT OF 30 CASES WITH MORE THAN 9-YEAR FOLLOW-UP Felipe Garibo, Jose´ Ignacio Maruenda, Borja Maruenda, Carlos Barrios University Clinic Hospital, Valencia, Spain; Hopsital de la Ribera, Alzira, Spain; Valencia Catholic University, Valencia, Spain Background/introduction: It is speculated that hybrid one-level or multi-level fusion constructs with dynamic topping-off stabilization could reduce the incidence of adjacent segment degeneration (ASD), but there are currently no long-term data. Purpose of the study: This study analyzes the 9-year follow-up outcomes of this type of hybrid systems, with special emphasis on the adjacent disc radiological degeneration and the rate of reintervention. Materials and methods: A series of 30 patients with lumbar degenerative disease (mean age, 49 years) were retrospectively reviewed. Patients underwent posterior fusion (1–4 levels) with dynamic topping-of stabilization (hybrid construct) using a transpedicular device. Clinical outcomes were assessed with: lumbar and radicular VAS, Oswestry Disability Index (ODI) and degree of patient satisfaction. Changes in adjacent discs were assessed using conventional radiographs (UCLA scale) and MRI (Pfirmann). Results: A significant initial postoperative clinical improvement was found: lumbar and radicular VAS passed from 6.8 to 5.6 before surgery to 3.2 and 2.8, respectively at immediate postoperative evaluation; ODI decrease from 59.9 preop to 29.1 at the postoperative period. A 80% of patients were satisfied with the procedure in the immediate postoperative period. At 9-year follow-up, worsening of the clinical parameters was found: mean lumbar and radicular VAS were 4.5 and 4.3, and ODI increased to 39.3. The patient satisfaction rate decrease to 60%. Adjacent segment radiographic changes (loss of disc height) were found in two cases (6.6%) that also showed a progression of disc degeneration on MRI (Pfirrmann grade 3–4). Osteolysis around the topping-off screws was detected in other two cases (6.6%). Revision surgery was required in three cases (9.9%): one with screw loosening and two with discitis of adjacent upper segment. There were no rod or screw breakages. Conclusions: Long-term outcomes of posterior lumbar fusion with topping-off dynamic transpedicular stabilization show a low rate of ASD and revision surgery. The re-operations were due to implant failures (lysis around the pedicle screw). There were no rod or screw breakages. Although follow-up is relatively long, it seems still necessary further long term randomized studies to confirm the efficacy of hybrid constructs in preventing adjacent segment degeneration. Disclosures: Author 1: none; Author 2: none; Author 3: none; Author 4: none.

P21 DEGREE OF MOBILITY AT L4–L5 IN PREOPERATIVE AND INTRAOPERATIVE IMAGES IN PATIENTS WITH LUMBAR DEGENERATIVE SPONDYLOLISTHESIS Federico P. Girardi, Roland Duculan, Alex Fong, Manuela C. Rigaud, Carol A. Mancuso Hospital for Special Surgery, New York, USA Background context: Flexion and extension images have been used to determine mobility at the level of the disease; however, the degree of mobility in relation to other images, particularly intraop images has not been characterized. Purpose: To characterize the degree of mobility in preoperative and intraoperative images in LDS patients undergoing surgery. Study design/setting: Cross-sectional preop and intraop study, tertiary spine center. Patient sample: 100 consecutive patients undergoing surgery for LDS at L4–5 Outcome measures: Change in the degree of mobility on preop and intraop images. Methods: Consecutive patients were enrolled several days preop. CT scans and radiographs (lateral upright, flexion and extension) were obtained preop (4) while sagittal fluoroscopic images (supine and prone) were taken intraop (2) after anesthesia induction but before incision. Mobility is a change in translation and angulation. A change in mobility is the difference between the changes of movement from the minimum. Degree of mobility is an incremental change (range 0 none—6 most). We used the Pearson’s Chi square test to assess the overall difference across images and the Poisson regression to identify images that best show an association with degree of mobility. Instability defined as translation C3.5 mm or angulation C11. Results: Mean 68, 56% women, mean BMI 28, 59% had a smoking history, mean back and leg pain were 6.1 and 5.2, respectively. Of the 100 patients, 54 met criteria for instability: 42% by translation, 4% by angulation, 8% by both; the remaining were classified as stable by these criteria. According to the Poisson regression, the flexion image has a 1.87 (p \ 0.001) times higher rate in identifying the degree of mobility in translation while the prone image has a 2.32 (p \ 0.001) times higher rate in identifying degree of mobility in angulation compared to the CT image as the reference group. The lateral and supine images combined had the highest rate in identifying the degree of mobility in translation difference (p \ 0.001) compared other images. In identifying degree of mobility in angulation difference, the CT and prone images have a rate ratio of 2.32 (p \ 0.001) compared to CT and extension images as the reference group. The overall difference across all images is statistically significant (p \ 0.001) Conclusions: Preoperative flexion and extension images are not sensitive enough in identifying the degree of mobility in patients with LDS at L4–L5. Mobility in other images such as prone, lateral upright/supine and CT/prone, gives us the highest rate ratio in identifying mobility in angulation, translation difference and angulation difference, respectively. Furthermore, other variables such as facet morphology, degree of diastasis, and disc height might also help to avoid missing instability in these patients. Disclosures: Author 1: grants/research support: Aesculap, MiMedx, consultant: DePuy Spine, LANIX, Inc; NuVasive, Inc.; Ortho Development Corp.; SpineArt USA, Inc., stock/shareholder: Centinel Spine; Spinal Kinetics; Lifespan; Liventa Bioscience; Paradigm Spine LLC; Pioneer Spine Technology, Inc; Small Bones Innovations, LLC, royalties: LANIX, Inc; Ortho Development; NuvVasive Inc.; DuPuy Spine, employee: Hospital for Special Surgery; Author 2: none; Author 3: none; Author 4: none; Author 5: none.

123

S346 P22 AMONG PATIENTS WITH STABILITY ON FLEXION– EXTENSION RADIOGRAPHS, PAIN IN FLEXION AND EXTENSION DOES NOT PREDICT SUBSEQUENT INTRAOPERATIVE INSTABILITY Federico P. Girardi, Alex Fong, Manuela C. Rigaud, Roland Duculan, Carol A. Mancuso

Eur Spine J (2017) 26 (Suppl 2):S335–S405 P23 COMPLEXITY OF LUMBAR SURGERY DOES NOT DICTATE SURGEONS’ ASSESSMENT OF FUNCTIONAL OUTCOME Federico P. Girardi, Manuela C. Rigaud, Alex Fong, Roland Duculan, Carol A. Mancuso Hospital for Special Surgery, New York, USA

Hospital for Special Surgery, New York, USA Background context: We previously showed that 36% of patients with lumbar degenerative spondylolisthesis (LDS) classified as stable by preop flexion–extension radiographs were unstable under anesthesia by intraop images. Pain, muscle contraction, and other anatomic features have been hypothesized to limit flexion and extension, which results in under reporting of instability when measured by standard preop flexion–extension radiographs. Purpose: Our goal was to ascertain if preop symptoms predicted true instability when standard radiographs indicated stability. Among LDS patients classified as stable by flexion–extension radiographs, we compared preop symptoms between those who remained classified as stable (RS) versus those who were re-classified as unstable (RCU) based on intraop post-anesthesia images. Study design/setting: Cross-sectional preop and intraop study, tertiary spine center Patient sample: 100 consecutive patients undergoing surgery for LDS at L4–5 classified as stable by preop flexion–extension radiographs Outcome measures: Comparison of preop pain measurements between RS and RCU patients. Methods: Consecutive patients were enrolled preop and rated severity of back and leg pain (0 = none, 10 = worst) and completed the Oswestry Disability Index (ODI) (0 = best, 100 = worst). Presence of pain in flexion and extension was reported by surgeons. Instability was defined as translation C3.5 mm or angulation C11. Additional images included preop CT scans and intraop fluoroscopic images after anesthesia but before incision. Using similar translation and angulation thresholds, all patients were re-assessed for instability using intraop images. 64 were RS and 36 were RCU. Self- and surgeon-reported pain of these groups were compared with t-tests and Chi square tests. Results: Mean age was 67, 57% were women, mean BMI was 29, mean ODI score was 48 (range 6–78), and 53% had a smoking history. Compared to RS patients, RCU patients did not differ in self-reported back pain (mean 6.4 vs 6.3, p = 0.85) or leg pain (mean 6.5 vs 6.9, p = 0.49). They did not differ in pain on flexion (25 vs 19%, p = 0.53) or pain on extension (58 vs 61%, p = 0.75). There were also no differences according to age, BMI, ODI score, or smoking status. However, more men (47%) than women (28%) were RCU (p = 0.06). Conclusions: Of patients with L4-5 LDS who did not have instability on preop flexion–extension radiographs, back and leg pain and pain on flexion and extension did not predict true instability. These findings indicate that relying on symptoms and physical exam is not sufficient to predict intraop instability. Future studies will combine several concurrent preop radiographic features to predict true intraop instability. Disclosures: Author 1: grants/research support: Aesculap, MiMedx, consultant: DePuy Spine, LANIX, Inc; NuVasive, Inc.; Ortho Development Corp.; SpineArt USA, Inc., stock/shareholder: Centinel Spine; Spinal Kinetics; Lifespan; Liventa Bioscience; Paradigm Spine LLC; Pioneer Spine Technology, Inc; Small Bones Innovations, LLC, royalties: Lanx Inc.; Ortho Development; Nuvasive; DuPuy Spine, employee: Hospital for Special Surgery; Author 2: none; Author 3: none; Author 4: none; Author 5: none.

123

Background context: Follow-up assessments after lumbar surgery provide surgeons with necessary technical and patient-centered information to advise advancement of physical activity. Although patients with more complex surgery may be more likely to limit their activities, more complex surgery may not necessarily predispose to more functional limitations if the surgery was technically successful. Purpose: To evaluate surgeons’ postop functional assessment in terms complexity of surgery, technical outcome, and patient characteristics. Study design/setting: Cross-sectional study, tertiary spine center. Patient sample: 134 patients approximately 3 months after complex lumbar surgery Outcome measures: Surgeons’ ratings of functional outcome Methods: Surgeons were asked at the 3-month postop visit to rate current outcome of surgery based on functional outcome (0 = poor, 10 = excellent) and technical outcome (0 = worst, 10 = best). Patients were asked about current symptoms and pain medications, function with the Oswestry Disability Index (ODI); and psychological well-being with the Patient Health Questionnaire-8 (PHQ) a standard measure of depressive symptoms (score range 0–24, higher = more symptoms). Complexity of surgery was assigned which a standard Surgical Invasiveness Index (SII) summed value was calculated with a maximum of 10 points per vertebral level; higher value = greater complexity. Results: Mean age 63, 50% men, 31% working, 30% on disability. 90% had multilevel surgery, 85% had fusion, and the mean SII value was 11 (range 1–41), median 10 (interquartile range 7–14). Time since surgery was 2.9 months. Compared to preop pain, 90% reported their pain had improved, 5% the same, and 5% worsened. Before surgery 43% were taking opioids; of these 81% were taking opioids at 3 months, and an additional 6% patients were newly taking opioids. Mean PHQ score was 5. Surgeons’ mean ratings were eight for functional outcome and eight for technical outcome. Based on ordinal logistic regression, functional outcome was not associated with complexity (SII value) for the entire range (p = 0.38) nor at the lower (OR 0.8, CI 0.4–1.6; p = 0.47) and higher range (OR 1.2, CI 0.5–2.2; p = 0.95). Variables associated with worse functional outcome in multivariable analysis were: worse technical outcome (OR 2.4, CI 1.7–3.3; p \ 0.0001), current opioid use (OR 2.0, CI 1.0–4.0; p = 0.04), worsened pain (OR 1.8, CI 1.4–2.4; p \ 0.0001), and more depressive symptoms (OR 1.2, CI 1.1–1.3; p \ 0.0001). Patient-reported disability (ODI) was correlated with surgeons’ functional rating (r = 0.62, p \ 0.0001). Conclusions: Complexity of lumbar surgery does not automatically dictate functional outcome. Patients should be informed that more complex surgery is not necessarily a restriction to advancing physical activity if the technical outcome is successful. Disclosures: Author 1: consultant: DePuy Spine, LANIX, Inc; NuVasive, Inc.; Ortho Development Corp.; SpineArt USA, Inc., stock/shareholder: Centinel Spine; Spinal Kinetics; Lifespan; Liventa Bioscience; Paradigm Spine LLC; Pioneer Spine Technology, Inc; Small Bones Innovations, LLC, royalties: Lanx Inc.; Ortho Development; Nuvasive; DuPuy Spine, employee: Hospital for Special Surgery; Author 2: none; Author 3: none; Author 4: none; Author 5: none.

Eur Spine J (2017) 26 (Suppl 2):S335–S405

S347

P24 SURGICAL OUTCOME CATEGORISATION USING GLOBAL ASSESSMENT (GA) Amanda Hansson Hedblom, Emma Jonsson, Peter Fritzell, Olle Ha¨gg, Fredrik Borgstro¨m Quantify Research, Stockholm, Sweden; LIME/MMC, Karolinska Institutet, Stockholm, Sweden; Neuro-orthopedic clinic, Futurum Academy, Jo¨nko¨ping, and Capio St Go¨ran Hospital, Stockholm, Sweden; Spine Center Go¨teborg, Gothenburg, Sweden Objective: To explore the relationship of Global Assessment (GA) and minimum clinical important differences (MCIDs) for other patient reported outcome measures (PROMs) and to identify an appropriate surgical outcome classification based on GA for spine surgery. Methods: Global assessment is a PROM developed to evaluate pain relief after spine surgery. The GA asks patients to respond to how their pain has changed compared to pre-surgery categorised in five levels (1: disappeared; 2: much improved; 3: somewhat improved; 4: unchanged; and 5: worsened). Since the GA does not collect pre-surgery pain status, changes are not quantifiable and it is not possible to establish MCID cut-offs. Using data from the Swedish spine register (Swespine) we estimated the changes by GA level from baseline (surgery) to 2 years post-surgery for the following PROMs: Oswestry Disability Index (ODI), Visual Analog Scale (VAS) for leg and back pain, and EQ-5D. A literature review was conducted to identify published MCID thresholds for the ODI, EQ-5D and VAS-pain. These were then compared with the estimated mean changes for each GA-level to inform on an appropriate categorisation of surgical outcome (e.g. success or failure) using GA. The data used for the analysis was based on patients included in the Swespine having lumbar spine surgery during years 2000–2012. GA-leg and VAS-leg were used as the relevant measure for spinal stenosis and disc herniation patients; GA-back/ VAS-back were used for spondylolisthesis and degenerative disc disease (DDD). Results: In the diagnosis group spinal stenosis, the proportion of patients who had successful outcome was 57%, in the disc herniation group it was 76%, in spondylolisthesis 65% and in DDD 73%. The corresponding proportions who had undetermined outcomes were 19, 13, 18, 16% for the respective diagnosis groups. Estimated changes by GA level and the identified MCID thresholds are presented in Fig. 1. The maximum and minimum MCID threshold values are indicated by the red lines. Patients who reported GA levels 1 and 2 generally exceeded the improvement thresholds for ODI (-5 to -20) and VAS (-18 to -35), whereas levels 4 and 5 did not. Level 3 outcomes were more ambiguous. Looking at EQ-5D, the range of MCID improvement thresholds is notably wide, rendering less precise results, but a similar pattern in terms of GA levels was distinguishable. Conclusion: Categorising patients reporting GA levels 1 and 2 as having a successful outcome of surgery (meaningful clinical improvement) seems to be consistent with reported changes in other PROMs. Those reporting level 3 are less clear and could be categorised as having undetermined outcome and levels 4 and 5 as failed outcome. Disclosures: Author 1: grants/research support: Medtronic; Author 2: grants/research support: Medtronic; Author 3: none; Author 4: none; Author 5: consultant: Medtronic.

P25 PATHOGENESIS OF SCIATICA IN LUMBAR DISC HERNIATION. CHANGES IN INTRARADICULAR OXYGEN SATURATION AND ELECTOPHYSIOLOGICAL VALUE BY INTRAOPERATIVE STRAIGHT LEG RAISING (SLR) TEST Shigeru Kobayashi, Yuichi Hashishin, Hisanori Matsuoka, Katsuhiko Hayakawa, Adam Meir Department of Orthopaedics and Rehabilitation Medicine, Fukui University, Fukui, Japan Introduction: Sciatica most commonly occurs when a herniated disk compresses part of the nerve root. The straight-leg-raising (SLR) test has been one of the most significant of clinical signs when making a clinical diagnosis of lumbar disc herniation. This study is to investigate the relationships among nerve root movement, intraradicular oxygen saturation and electrophysiological values during an intraoperative SLR test in vivo. Methods: The subjects were 15 patients with lumbar disc herniation who underwent micro-discectomy [10 men and 5 women aged 33.1 years on average (range 25–43)]. The patients’ legs were allowed to hang down to the angle at which sciatica had occurred and the change of nerve root action potentials and intraradicular oxygen saturation were measured. After removal of the hernia, a similar procedure was repeated. During the SLR test performed pre-operatively, 7 patients developed sciatica at an angle of 10 and 8 patients at 30. This test was performed only on those who gave informed consent. Results: The intraoperative SLR test showed that the hernia compressed the nerve roots to increase their flatness, resulting in a clear disturbance by which gliding distance was reduced to only few millimeters. During the SLR test, amplitude of action potential and intraradicular oxygen saturation showed a sharp decrease at the angle that produced sciatica. When the angle of the legs was returned to zero degrees, amplitude and oxygen saturation showed an immediate improvement and the value recovered to that obtained before the SLR test. After removal of the hernia, the nerve roots showed smooth gliding in all patients. The intraoperative SLR test conducted after removal of the hernia showed no significant decrease of amplitude and oxygen saturation in the nerve roots. When the SLR test was performed at 1 week after the operation, all the patients were negative and did not develop sciatica, unlike the results obtained preoperatively.

123

S348 Discussion: When sciatica was experienced during the SLR test, hernia mass caused relatively strong mechanical stretch to develop at the nerve roots, resulting in about 70–80% decrease in intraradicular oxygen saturation. Our data suggest that temporary ischemic changes in the nerve root cause transient conduction disturbances. It was predicted that disturbed blood flow caused by this mechanical stress led to development of edema and anoxemia, constituting an inductive factor of ectopic discharge responsible for pain 1, 2, 3. Conclusion: When sciatica was experienced during the SLR test, the amplitude and intraradicular oxygen saturation deteriorated significantly after the SLR test. Acknowledgements: This work was supported by Grant-in Aid from the Ministry of Education, Science and Culture of Japan (25460719). Disclosures: None.

P26 3D POROUS LAMELLAR TITANIUM LATERAL INTERBODY CAGES: SUBSIDENCE RATES COMPARED TO PEEK AND EARLY CLINICAL OUTCOMES

Eur Spine J (2017) 26 (Suppl 2):S335–S405

Two level PEEK lateral cages showing minor subsidence. Paent has no back or leg pain

Single level lamellar tanium lateral cage with plate with no subsidence

P27 CLINICAL AND RADIOGRAPHIC RESULTS FOLLOWING THE USE OF LATERAL INTERBODY CAGES TO RESTORE LUMBAR LORDOSIS IN REVISION FUSION SURGERY Robert Lee

Robert Lee Department of Spinal Surgery, Royal National Orthopaedic Hospital NHS Trust, Stanmore, UK Introduction: Whilst the elastic modulus of PEEK cages is low, they have been shown to promote fibrosis and an inflammatory reaction. Increased rates of non-union and subsidence have been reported. Titanium has more favourable biological properties but the elastic modulus and stiffness of solid titanium is much greater than cancellous bone and subsidence has also been reported. 3D porous lamellar titanium cages are a less stiff implant and hence combine more favourable biological properties with better implant design. This study demonstrates that lamellar titanium lateral interbody cages show significantly less subsidence with good early clinical outcomes. Subsidence can lead to loss of lordosis and loss of indirect decompression. Hypothesis: 3D porous lamellar titanium lateral interbody cages have significantly less subsidence than PEEK lateral cages with good early clinical outcomes. Design: Retrospective review of prospectively collected data from a single surgeon series. Methods: There were 2 matched groups with 20 patients in each group: PEEK (group 1) and lamellar titanium (group 2). A retrospective review of prospectively collected outcome data (VAS leg, VAS back, EQ-5D, EQ-5D VAS and ODI) was then performed and radiographs reviewed for subsidence immediately post op, 6 weeks post op and 6 months post op. Results: Each patient group had similar diagnoses including degenerative scoliosis and spondylolisthesis. 43 lateral cages were inserted in each group with 1 level in 7 patients, 2 levels in 6 patients, 3 levels in 4 patients and 4 levels in 3 patients. Subsidence was 2.0 mm in the PEEK group. There was no subsidence in any of the lamellar titanium patients. This difference was highly significant. Outcomes at 6 months were as follows. Group 1: VAS leg 2.4, VAS back 2.9, EQ-5D 26.5, EQ-5D VAS 71.9, ODI 37. Group 2: VAS leg 1.5, VAS back 2.0, EQ-5D 23.5, EQ-5D VAS 82.3, ODI 18. Difference in parameters was statistically significant (p \ 0.05) for ODI and EQ-5D-VAS. Conclusion: 3D porous lamellar titanium lateral interbody cages show no subsidence and compared to PEEK lateral cages with better ODI and EQ-5D VAS outcome scores at 6 months. The lack of subsidence may be due to the decreased stiffness of lamellar titanium. Disclosures: Author 1: grants/research support: K2M, consultant: K2M, Medtronic, Signus, SI-Bone.

123

Department of Spinal Surgery, Royal National Orthopaedic Hospital NHS Trust, Stanmore, UK Introduction: Patients with previous fusions can present with iatrogenic flat back deformity, non-union, progressive deformity and proximal level degeneration. There can be issues with recurrent compression or stenosis of the central canal, lateral recess or neural foramen. Traditional open approaches include direct revision decompression and extension of the fusion, sometimes with posteriorly inserted interbody cages. The risk here is of nerve damage and difficulty negotiating scar tissue. Minimally invasive lateral cages can effectively indirectly decompress nerves and restore lumbar lordosis. Hypothesis: Using multiple minimally invasive lateral cages in revision lumbar surgery can restore lordosis, indirectly decompress nerves and produce excellent outcomes with less complications Design: This is a single surgeon case series with retrospective review of prospectively collected data. Between January 2013 and January 2016, patients were included who had previous lumbar decompression or fusion surgery, treated with minimally invasive lateral interbody cages and who had a minimum 12 months follow-up. Methods: Complications were recorded together with the following radiographic parameters and clinical outcome scores. Pre and postoperative radiographic parameters: lumbar lordosis (LL), pelvic incidence–lumbar lordosis mismatch (PI–LL), sagittal vertical axis (SVA) and pelvic tilt (PT). Outcome scores: VAS back, VAS leg, EQ5D, EQ-5D VAS, ODI. Results: There were 25 patients: 15 cases with a previous lumbar fusion (of which 4 had a non-union), 13 cases with degenerative scoliosis, 15 cases with positive sagittal balance and 5 cases with spondylolisthesis. 15 cases were purely MIS and 10 cases hybrid with open posterior fusion. A total of 49 lateral cages were inserted with 1 level in 7 patients, 2 levels in 13 patients, 3 levels in 4 patients and 4 levels in 1 patient. The distribution of levels was as follows: L1/2—5 cages, L2/3—15 cages, L3/4—19 cages, L4/5—7 cages. Average post-operative parameters showed improvement of SVA from 90 to 44 mm and PI–LL from 27 to 8. Average 6 month outcome scores were: VAS back 9–3, VAS leg 8–2, EQ-5D 0.181–0.638, EQ-5D VAS 33–72 and ODI 68–41. These scores were maintained in those patients reaching the 2 years follow-up mark. Conclusion: Minimally invasive lateral cages in revision lumbar surgery can restore lordosis and produce good early outcome measures. Disclosures: Author 1: grants/research support: K2M, consultant: K2M, Medtronic, Signus, SI-Bone.

Eur Spine J (2017) 26 (Suppl 2):S335–S405

S349

P28 3D POROUS LAMELLAR TITANIUM VERSUS PEEK AND SOLID TITANIUM INTERBODY TLIF CAGES: A PROSPECTIVE CLINICAL AND RADIOGRAPHIC COMPARISON

P29 SURGERY FOR HERNIATED LUMBAR DISC IN INDIVIDUALS 65 YEARS OF AGE OR OLDER: A MULTICENTER OBSERVATIONAL STUDY Mattis Madsbu, Øyvind Salvesen, Øystein Nygaard, Tore Solberg, Sasha Gulati

Robert Lee, Lester Wilson Department of Spinal Surgery, Royal National Orthopaedic Hospital NHS Trust, Stanmore, UK Introduction: Whilst the elastic modulus of PEEK cages is low, they have been reported as promoting fibrosis and an inflammatory reaction. Increased rates of non-union and subsidence have been reported. Titanium has more favourable biological properties but the elastic modulus and stiffness of solid titanium is much greater than cancellous bone and subsidence has also been reported. 3D porous lamellar titanium cages are a less stiff implant and hence combine more favourable biological properties with better implant design. This study demonstrates that lamellar titanium TLIF cages show significantly less subsidence with good evidence of fusion between 6 and 12 months. Hypothesis: 3D porous lamellar titanium TLIF cages have less subsidence and better fusion rates compared to PEEK and solid titanium cages. Design: Retrospective review of prospectively collected data from a two surgeon series. Methods: A standard MIS Wiltse approach with the same screws and operative technique was used by both surgeons to perform the TLIF. There were 3 matched groups with 20 patients in each group: PEEK (group 1), solid titanium (group 2), lamellar titanium (group 3). Diagnoses included primary cases as well as revision TLIFs and up to grade 3 spondylolisthesis. A retrospective review of prospectively collected outcome data (VAS leg, VAS back, EQ-5D, EQ-5D VAS and ODI) was then performed. Patients had CT scans at either 6 or 12 months depending on the institution. Results: Subsidence was 2.3 mm in the PEEK group, 1.5 mm in the solid titanium group and 0.6 mm in the lamellar titanium group. This difference was significant. Clinically all groups reported excellent outcomes (average post op scores: VAS leg 2.0, VAS back 2.3, ODI 20, EQ-5D 26.5, EQ-5D VAS 79.9). However the lamellar titanium group showed a greater initial decrease in VAS back at 6 weeks. Non-union occurred in two of the PEEK cases and in one solid titanium case. All the lamellar titanium cases showed fusion at either 6 or 12 months. Conclusion: The initial results of 3D porous lamellar titanium cages show less subsidence and better fusion rates compared to PEEK and solid titanium. Disclosures: Author 1: grants/research support: K2M, consultant: K2M, Medtronic, Signus, SI-Bone; Author 2: royalties: K2M.

Department of Neurosurgery, St. Olavs Hospital, Trondheim, Norway Background: Lumbar microdiscectomy is the most common surgical treatment, but data on surgical outcomes in the elderly are limited. The aim of this study was to compare patient-reported outcomes following lumbar microdiscectomy in patients at least 65 years of age with younger patients. Methods: Data were collected through the Norwegian Registry for Spine Surgery. Patients were eligible if they had a primary diagnosis of LDH and underwent non-emergency single-level lumbar microdiscectomy between 2007 and 2013. Patients were excluded who had undergone previous lumbar spine surgery or had coexisting spondylolisthesis or scoliosis. The primary endpoint was change in Oswestry disability index (ODI) 1 year after surgery. Secondary endpoints were generic quality of life (EuroQol; EQ-5D), back pain numerical rating scale (NRS), leg pain NRS, and complications. Results: There were 5195 patients\65 years and 381 patients C65 years. Loss to follow-up at 1-year was 16.5% (n = 63) in patients C65 years and 31.9% (n = 1658) in patients\65 years (p \ 0.001). For all patients there was a significant improvement in ODI (-31.04 points, 95% CI -30.34, -31.74, p \ 0.001). There were no differences between age cohorts in mean changes of ODI, EQ-5D, or leg pain, but elderly patients experienced more improvement in low back pain (Table 1). Patients C65 years experienced both more perioperative complications (4.2 vs 2.3%, p = 0.019) and complications occurring within 3 months of hospital discharge (12.4 vs 5.4%, p \ 0.001), mainly due to more urinary tract infections, micturition problems, and unintentional durotomies. Younger patients had shorter hospital stays than patients C65 years (2.7 vs 1.8 days). Conclusion: Although they had more minor complications and longer hospital stays, individuals C65 years experienced improvement after lumbar microdiscectomy that was similar to that of younger individuals. Age alone should not be a contraindication to surgery, as long as the individual is fit for surgery. Disclosures: Author 1: none; Author 2: none; Author 3: none; Author 4: none; Author 5: none.

Table: Comparison of surgical results with complication rates between SHILLA & MdGR for EOS at Minimum Follow-up of FIVE Years

Parameter

34 (13 males & 21 females)

Age at surgery

6.92 years (range: 1.92 – 11.83)

8.68 years (range: 5.00 t o 7.00)

Patients aged >10 years at surgery

3

15

No.of patients with preop Cobb <50o

4

6 (CGR to MdGR conversions)

No. of patients conve rted from CGR

No. of single rods inserted No. of patients who ha d two stage surgeries

Zero

10

77 (n=33)

40 (n=34)

Zero

13

6

Zero

Follow-up duration

7.00 years (range: 4.75 – 10.75)

6.00 years (range: 5.00 t o 7.00)

No. of Graduates

18 (4 males & 14 females)

17 (7 males & 10 females)

No. of patients who ha d definitive spinal fusion

15

No of patients who ha d PFT

6

6

31o (from 69o to 38 o)

9o (from 64o to 55o)

T1-S1 height gain (pre-op to final follow-up)

79mm

65mm

Complication Rate

73 %

12 %

Cobb angle correction

Infection Solid Titanium Cage: 6 month CT showing non-union and subsidence

MdGR

33 (M:F ratio – Not reported)

Total surgeries performed during growth

PEEK Cage: 6 month CT showing non-union and subsidence

SHILLA

Number of patients

Rod breakage

Lamellar Titanium Cage: 6 month CT showing solid fusion and no subsidence

11

6 / 33

4 / 34

18/33 (all were DR)

12/34 (6 each in SR & DR)

Proximal junctional kyphosis*

3

9*

Conversion to C GR or VEPTR

2

Zero

*

SRS definition was adhered to when evaluating for PJK in MdGR group unlike in the SHILLA group

123

S350

Eur Spine J (2017) 26 (Suppl 2):S335–S405

P30 LUMBAR MICRODISCECTOMY FOR SCIATICA IN ADOLESCENTS: A MULTICENTER OBSERVATIONAL REGISTRY-BASED STUDY

continued Last value carried forward analysis Adolescents (n = 77)

Mattis Madsbu, Tore Solberg, Øystein Nygaard, Sasha Gulati Department of Neurosurgery, St. Olavs Hospital, Trondheim, Norway Background: Lumbar disc herniation (LDH) is rare in the adolescent population. Factors predisposing to LDH in adolescents differ from adults with more cases being related to trauma or structural malformations. Further, there is limited data on patient-reported outcomes after lumbar microdiscectomy in adolescents. Our aim was to compare clinical outcomes at one year following single-level lumbar microdiscectomy in adolescents (13–19 years) compared to younger adults (20–50 years) with LDH. Methods: Data were collected through the Norwegian Registry for Spine Surgery. Patients were eligible if they had radiculopathy due to LDH, underwent single-level lumbar microdiscectomy between January 2007 and May 2014, and were between 13 and 50 years old at time of surgery. The primary endpoint was change in Oswestry disability index (ODI) 1 year after surgery. Secondary endpoints were generic quality of life (EuroQol; EQ-5D), back pain numerical rating scale (NRS), leg pain NRS, and complications. Results: A total of 3245 patients were included (97 patients 13–19 years old and 3148 patients 20–50 years old). A significant improvement in ODI was observed for the whole population, but there was no difference between groups (0.6, 95% CI -4.5 to 5.8, p = 0.811). There were no differences between groups concerning EQ-5D (-0.04, 95% CI -0.15, 0.07, p = 0.442), back pain NRS (-0.4, 95% CI -1.2, 0.4, p = 0.279), leg pain NRS (-0.4, 95% CI -1.2, 0.5, p = 0.374) or perioperative complications (1.0% for adolescents, 5.1% for adults, p = 0.072). Conclusion: The effectiveness and safety of single-level microdiscectomy are similar in adolescents and the adult population at 1-year follow-up. Disclosures: Author 1: none; Author 2: none; Author 3: none; Author 4: none. Primary and secondary patient reported outcomes at 1 year Complete case analysis Adolescents (n = 59)

ODI

Baseline

One year

Mean change

35.0

8.5

26.5

EQ-5D

0.36

0.83

0.48

Back pain NRS

5.7

2.2

3.5

Leg pain NRS

6.3

1.4

4.9

Complete case analysis Adults (n = 2013)

ODI EQ-5D

Difference in mean Baseline One Mean change year change between groups (95% CI)

P-value

41.3

0.811

0.34

14.1 0.78

27.2 0.43

0.6 (-4.5, 5.8)

-0.04 (-0.15, 0.07) 0.442

Back pain NRS

5.7

2.7

3.1

-0.4 (-1.2, 0.4)

0.279

Leg pain NRS

6.5

2.0

4.5

-0.4 (-1.2, 0.5)

0.374

123

Baseline ODI

33.7

One year

Mean change

8.3

25.4

EQ-5D

0.38

0.85

0.47

Back pain NRS

5.5

2.0

3.5

Leg pain NRS

6.4

1.3

5.0

Last value carried forward analysis Adults (n = 2470)

ODI EQ-5D

Difference in mean Baseline One Mean change year change between groups (95% CI)

P-value

41.2

0.593

0.34

14.6 26.6

1.2 (-3.3, 5.8)

0.77 0.43 -0.04 (-0.13, 0.05) 0.386

Back pain NRS 5.8

2.7

3.1

-0.4 (-1.1, 0.3)

0.261

Leg pain NRS

2.0

4.5

-0.5 (-1.2, 0.2)

0.173

6.5

P31 RELATIONSHIP BETWEEN SPINO-PELVIC PARAMETERS AND CLINICAL AND FUNCTIONAL OUTCOMES IN 104 PATIENTS UNDERGOING LUMBAR FUSION SURGERY Bruno Maia, Pedro Carvalhais, Luis Teixeira Spine Center, Cirurgia da Coluna, Coimbra, Portugal Background: As our understanding of sagittal balance has improved, we have become more aware of harmful affects that surgical interventions may have on it. It has been shown that lumbar fusion may change the sagittal balance, decreasing the sacral slope (SS) and lumbar lordosis (LL). Fused spinal levels are locked in a position that is dependent on operative position and the surgical approach. Additionally, post fusion pain has been shown to be related to a decreased SS, increased pelvic tilt (PT), and decreased LL. There is growing interest in use of spinopelvic parameters to predict clinical outcomes. Objectives: The aim of this study was to determine whether consideration of sagittal balance parameters is correlated with better clinical results. Study design and methods: We retrospectively reviewed 104 patients undergoing lumbar arthrodesis by different surgical techniques (MISS–TLIF, PLIF and PLF). Obtaining pre and post surgery X-ray images, the following radiological parameters were analysed: SS, PT and LL. Clinical evaluation of pain was performed according to ‘‘visual analogue score’’ (VAS) and the associated disability through ‘‘Oswestry Disability Index ‘‘ (ODI). Using a special software program, we measured spino-pelvic parameters on digital radiographs. Relationship between radiological and clinical evaluation were analysed in order to determine statistically significant differences. Results: A positive change on SS, resulted in statistically significant improvement of postoperative clinical scores and lumbar lordosis (p \ 0.05). Decrease in PT parameters, was associated with a positive variation of clinical scores in the postoperative period and increased lumbar

Eur Spine J (2017) 26 (Suppl 2):S335–S405 lordosis, resulting in statistically significant differences (p \ 0.05). Increase in LL has also resulted in better clinical scores. We also analysed differences between radiological parameters in different surgical techniques, verifying that the best clinical and radiographic results were obtained by PLIF technique. Conclusions: It was found that increases in PT and decreases in SS are associated with loss of lumbar lordosis and worse clinical results. These changes translate in significant overload of adjacent levels causing pain and worst outcomes. In this study the best radiological results resulted in statistically significant clinical improvements, and the PLIF technique was the most effective. Preservation of posterior arch and of the spine in minimally invasive surgery (MISS-TLIF) complicates the reversal of compensatory mechanisms in degenerative spine diseases, influencing the final results. In this study we also preform a literature review to determine the ideal lumbar lordosis and indications for minimally invasive fixation, in order to obtain better clinical outcomes. The success of lumbar fusion surgery depends on reversal of compensatory mechanisms observed in patients with degenerative diseases. Disclosures: Author 1: none; Author 2: none; Author 3: none.

P32 HIGH RATE OF STUDIES WITH LEVEL OF EVIDENCE I–II AMONG THE 100 MOST CITED ARTICLES IN LUMBAR SPINAL STENOSIS Juan Martı´nez-Andre´s, Carlos Barrios, Joaquin Alfonso Valencia Catholic University, Valencia, Spain Background/introduction: Lumbar spinal stenosis (LSS) is currently the most common indication for spinal surgery in older adults. Controversies still remain concerning the underlying causes and the optimal rationale for treatment. One of the pillars of scientific progress is the ability to keep abreast from the most important achievements in the scientific literature. To date, no study has used bibliometric analysis to review the most influential articles in lumbar spinal stenosis. Objective: To identify and analyze the characteristics of the 100 most cited articles on lumbar spinal stenosis with special reference to the evolution of the level of evidence. Methods: This is bibliometric study of current literature. The Thomson Reuters Web of Science was accessed to find the 100 most cited articles on lumbar spinal stenosis. The number and density of citations, authors, countries, journal and year, department, level of evidence, being part of multicenter studies and type of study, were recorded for each paper. The level of evidence was classified according to the guidelines by the Oxford Centre for Evidence-Based Medicine. Results: Until January 2017, the top-100 articles accumulated 11136 citations (average 259.05/year), ranging individually between 442 and 50 (average 111.36). The first reference was published in 1974 in clinical orthopedics and related research. The most frequent type of study focus on therapeutic issues (n = 40). The leading period was 1994-2003 (n = 44). The most prolific author was Katz JN (n = 10). Many studies were multicentric (n = 44). The USA (n = 51) and Spine (n = 48) have the publishing hegemony. Papers with level of evidence II were predominant (n = 49). Prior to 1985 until the decade 2006–2015, the mean level of evidence improved substantially (3.0 ± 0.9 versus 1.7 ± 0.8, p \ 0.01). The highest citation index (9, 16) corresponds to articles at the level of evidence I and the lowest (4.43) at level IV (p \ 0.01). Conclusions: This bibliometric analysis reveals a good level of evidence in the published clinical series and includes 100 articles useful for the approach of lumbar canal stenosis. Disclosures: Author 1: none; Author 2: none; Author 3: none.

S351 P33 RISK FACTORS OF ADJACENT SEGMENT DISEASE AFTER LUMBAR FUSION FOR DEGENERATIVE DISEASE Sergei Masevnin, Dmitry Ptashnikov, Dmitry Mikhaylov, Oleg Smekalenkov, Nikita Zaborovskii, Olga Lapaeva Department of Spine Surgery, Vreden Russian Research Institute of Traumatology and Orthopedics, St. Petersburg, Russia Introduction: Spinal segment arthrodesis has become a widely accepted treatment for degenerative disorders of the lumbar spine. For today rigid internal fixation with 360 fusion has been viewed as the gold standard for spine stabilizing surgery. However spinal fusion alters the normal biomechanics of the spine and eliminates mobile segments causing overload of adjacent segments. At the same time many patients with degenerative disorders have abnormalities of spino-pelvic parameters, which remain postop. Thus, spinal fusion, according to some authors can accelerate the degeneration of adjacent segments, especially under unfavorable biomechanics of the spine. Materials and methods: This retrospective study evaluated 120 patients underwent 360 short fusion lumbar surgery (one and two levels) from 2008 to 2013 for the treatment of degenerative conditions of the lumbar spine. The mean follow-up period was 42.2 months. Long cassette standing anteroposterior and lateral radiographs were performed on the preoperative, postoperative and follow-up visits. In all cases several spino-pelvic parameters were measured, including sagittal vertical axis (SVA), lumbar lordosis (LL), pelvic tilt (PT), pelvic incidence (PI) and PI–LL. Multivariate logistic regression analysis and receiver operating characteristic analysis (ROC) were performed to evaluate these parameters as risk factors of adjacent segment disease (ASD). Results: Radiographic ASD was found in 36 patients (30%). We have found a significant difference comparing ASD patients with no ASD cases in PI–LL parameter (p = 0.007). Statistical analysis showed that preoperative PI–LL of more than 11 was a significant risk factor of the ASD appearance after short lumbar fusion (p = 0.007; odds ratio 4,2; 95% CI 1.46–12.25). Conclusion: Patients with spino-pelvic alignment disturbances such as PI–LL mismatch have statistically significant increasing risks of the incidence ASD. Preoperative PI–LL of more than 11 increases the risk of occurrence of ASD 4.2 times. Level of evidence: 3. Disclosures: Author 1: none; Author 2: none; Author 3: none; Author 4: none; Author 6: none.

P34 MINIMALLY INVASIVE DECOMPRESSION SURGERY FOR LUMBAR SPINAL STENOSIS WITH DEGENERATIVE SCOLIOSIS: PREDICTIVE FACTORS OF RADIOGRAPHIC AND CLINICAL OUTCOMES Akihito Minamide, Munehito Yoshida, Andrew K. Simpson, Hiroki Iwahashi, Ryohei Kagotani, Hiroshi Yamada, Hiroshi Hashizume, Yukihiro Nakagawa, Hiroshi Iwasaki, Shunji Tsutsui Department of Orthopaedic Surgery, Wakayama Medical University, Wakayama, Japan Background: There is ongoing controversy regarding the most appropriate surgical treatment for lumbar spinal stenosis (LSS) with concurrent degenerative lumbar scoliosis (DLS): decompression alone, decompression with limited spinal fusion, or long spinal fusion

123

S352 for deformity correction. The coexistence of degenerative stenosis and deformity is a common scenario; Nonetheless, selecting the appropriate surgical intervention requires thorough understanding of the patients clinical symptomatology as well as radiographic parameters. Minimally invasive spinal (MIS) decompression surgery has demonstrated clinical outcomes for neurologic recovery equivalent to traditional open decompression, while allowing the distinct advantage of maintaining the posterior stabilizing structures. The authors theorized that MIS decompression, through the preservation of these posterior stabilizing structures, may provide an effective means of decompressing the spinal nerves in patients with stenosis and coexisting deformity, without causing deformity progression. The aims of this study were (1) to investigate the clinical outcomes of MIS decompression in LSS patients with DLS, and (2) to identify the predictive risk factors for both radiographic and clinical outcomes after MIS decompression in this patient population. Methods: 438 consecutive patients were enrolled in this study if they had an evidence of LSS. All enrolled patients underwent the endoscope-assisted decompression surgery for LSS. Inclusion criteria was evidence of LSS and DLS with coronal curvature measuring greater than 10. The Japanese Orthopaedic Association (JOA) score, JOA recovery rate, low back pain (LBP), and radiographic features were evaluated preoperatively and at over 2 years postoperatively. Spinopelvic parameters, including lumbar lordosis (LL), sacral slope (SS), pelvic tilt (PT), and pelvic incidence (PI) were measured on lateral standing radiographs of the entire spine. Results: Of the 438 patients, 122 were included in final analysis, with a mean follow-up of 2.4 years. The JOA recovery rate was 47.6%. LBP was significantly improved at final follow-up. Cobb angle was maintained for 2 years postoperatively (p = 0.159). Clinical outcomes in foraminal stenosis patients were significantly related to sex, preoperative high Cobb angle and progression of scoliosis (p = 0.008). In the severe scoliosis patients, the JOA recovery was 44%, and was significantly depended on progression of scoliosis (Cobb angle: preoperation 29.6, 2-year follow-up 36.9) and mismatch between the pelvic incidence (PI) and the lumbar lordosis (LL) (preoperative PI–LL 35.5 ± 21.2) (p = 0.028). Conclusions: Clinical outcomes of MIS decompression surgery were generally excellent/good, without the progression of degenerative scoliosis. For LSS patients with more severe deformity, Cobb C30, degree of coronal deformity and PI–LL mismatch were predictive risk factors of JOA recovery and relief of LBP. 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.

P35 IS ABDOMINAL AORTIC CALCIFICATION SEEN ON PREOPERATIVE LATERAL LUMBAR X-RAY A PREDICTOR OF FUSION FOLLOWING LUMBAR SPINAL FUSION SURGERY? Naoki Okubo, Masayuki Nakahara Department of Orthopedic Surgery, Kitasuma Hospital, Kobe, Japan Objective: Abdominal aortic calcification (AAC) is known to be closely associated with cardiovascular disease and lower limb arteriosclerotic disease. The decreased lumbar nutrient blood flow accompanying arteriosclerosis may affect bone fusion after lumbar spinal fusion surgery. We investigated the relationship between AAC and bone fusion following lumbar spinal fusion.

123

Eur Spine J (2017) 26 (Suppl 2):S335–S405 Subjects and methods: Sixty-six consecutive patients were observed for at least 1 year following posterior lumbar interbody fusion (PLIF) for lumbar degenerative disc disease. Sixteen patients under 60 years old and 3 patients with suspected early intervertebral infection were excluded, and the remaining 47 patients (26 males, 21 females; mean age, 71.7 years) were analyzed. The presence/absence of bone fusion was judged from computed tomography or flexion–extension radiographs at 1 year after surgery, and patients were divided into fusion and non-fusion groups. AAC was also evaluated using the AAC 24-point scale score (Kauppila, Atherosclerosis, 1997) based on plain lateral X-ray images of the lumbar spine. The Mann–Whitney U-test was performed, and differences showing values of p \ 0.05 were regarded as significant. Results: The fusion group included 26 patients (14 males, 12 females; mean age, 72.0 years) and the non-fusion group included 21 patients (12 males, 9 females; mean age, 71.2 years). No significant differences were seen between groups in terms of gender or age. The bone fusion rate at 1 year after surgery was 55.3%. Median AAC 24-point scale score was 3.5 (range 0–14) in the fusion group and 8.6 (range 1–21) in the non-fusion group, with the latter being significantly higher (p = 0.001). Discussion: Correlation of AAC with systemic arteriosclerosis and decreased bone mineral density of the thoracolumbar vertebrae has been pointed out, and each of these can affect bone fusion after PLIF. The AAC 24-point scale score can be easily evaluated from lateral lumbar X-rays. Postoperative bone fusion may be delayed in patients with AAC and a high AAC 24-point scale score, and it is important in such cases to carefully monitor the patient’s course while paying close attention to postoperative therapy and osteoporosis treatment. Conclusion: Arterial calcification seen on preoperative lateral lumbar X-ray is a predictor of fusion following lumbar spinal fusion surgery. Disclosures: Author 1: none; Author 2: none.

P36 PROMIS PHYSICAL FUNCTIONING AND PAIN CORRELATION WITH ODI AND VAS IN THE SURGICAL LUMBAR DISC HERNIATION PATIENT POPULATION Robert Owen, Steven McAnany, Luke Zebala Department of Orthopedic Surgery, Washington University in St Louis, St. Louis, USA Background: Legacy patient reported outcome measures such as Oswestry Disability Index (ODI) and Visual Analog Score (VAS) have become essential for analyzing treatments for lumbar disc herniations. Significant associated administrative burdens impose limits on completion of such measures. The Patient Reported Outcomes Measurement Information System (PROMIS) group developed a patient outcome measure in order to improve reporting of patient symptoms, function, and health and to reduce administrative burden. Despite early positive results, ODI and VAS scores have not been compared with PROMIS in patients with lumbar disc herniations undergoing surgery. The aim of this study is to compare ODI and VAS with PROMIS physical function and pain assessments to determine their correlations in a surgical patient population longitudinally. Methods: 82 patients with a diagnosis of lumbar disc herniation that went on to surgery were included in the study. All patients were treated at the same university spine center by four spine surgeons. Patients were seen and PROMIS, ODI, and VAS leg scores were collected preoperatively, at 1–4 months, and at 6 months. Data from

Eur Spine J (2017) 26 (Suppl 2):S335–S405 the mentioned outcome measures was extracted from a central Redcap database. Correlations between ODI and VAS with PROMIS physical function and pain, respectively were quantified using Pearson correlation coefficient measurements. Student’s T-tests were used to demonstrate correlation significance with alpha set at 0.05. Results: All 82 (100%) of patients completed all baseline preoperative questionnaires. 57 (70%) of patients completed all questionnaires at the 1–4 months follow up. 19 (24%) of patients completed questionnaires at 6 month follow up. PROMIS physical function and ODI demonstrated a strong negative longitudinal correlation, with Pearson r values of (-0.86, -0.82, -0.79) at baseline, initial follow up and 6 months, respectively. PROMIS pain and VAS leg pain demonstrated a weak positive correlation, with Pearson r values of (0.58, 0.48, 0.49) at baseline, initial follow up and 6 months. Students T-test demonstrated a P value of \0.0001 for all Pearson correlation calculations. Conclusions: PROMIS physical function scores have a strong negative correlation with ODI scores at baseline and in the postoperative course in patients undergoing surgery for lumbar disc herniations. PROMIS pain scores have a weak positive correlation with VAS leg pain scores at baseline and in the postoperative course. Surgeons may factor these outcomes into the delivery and interpretation of patient reported outcome measures in patients with lumbar disc herniations undergoing surgery. Use of PROMIS physical function for this patient population may improve completion of outcome measures in the office and reduce administrative burden while still providing reliable outcomes data, while use of PROMIS pain scores may not represent a reliable alternative for pain assessment in this patient population. Disclosures: Author 1: none; Author 2: none; Author 3: grants/research support: AO Spine, consultant: Medtronic, K2M, Depuy Synthes.

P37 PROMIS PHYSICAL FUNCTION AND PAIN CORRELATION WITH ODI AND VAS PAIN IN THE SURGICAL PATIENT POPULATION WITH LUMBAR STENOSIS AND CLAUDICATION

S353 the Pearson correlation coefficient. Students T-tests were used to show correlation significance with alpha set at 0.05. Results: All 108 (100%) of patients completed baseline preoperative questionnaires. 65 (60%) of patients completed questionnaires at 1–4 months. 34 (31%) of patients completed questionnaires at 6 months. PROMIS physical function and ODI demonstrated a strong negative correlation, with Pearson r values of (-0.69, -0.73, -0.62) at baseline, initial follow up and 6 months. PROMIS pain and VAS back pain demonstrated an inconsistent correlation, with Pearson r values of (0.45, 0.53, 0.74) at baseline, initial follow up and 6 months. PROMIS pain and VAS leg pain demonstrated an inconsistent correlation, with Pearson r values of (0.39, 0.41, 0.61) at baseline, initial follow up and 6 months. Students T-test showed a P value of\0.0001 for all correlation calculations. Conclusions: PROMIS physical function scores have a strong negative correlation with ODI scores at baseline and in the postoperative course in patients undergoing surgery for lumbar stenosis with claudication. PROMIS pain scores have an inconsistent correlation with VAS back and leg pain scores at baseline and postoperatively. Surgeons may factor these outcomes into the delivery and interpretation of patient reported outcome measures in patients with lumbar stenosis with claudication undergoing surgery. Use of PROMIS physical function for this patient population may improve completion of outcome measures in the office and reduce administrative burden while still providing reliable outcomes data, while use of PROMIS pain scores may not represent a consistent reliable alternative for pain assessment in this patient population. Disclosures: Author 1: none; Author 2: none; Author 3: grants/research support: AO Spine, consultant: Medtronic, K2M, Depuy Synthes.

P38 INFLUENCE OF PSYCHOSOCIAL DISTRESS IN THE RESULTS OF ELECTIVE LUMBAR SPINE SURGERY Luiz Pimenta, Vivian Amaral, Heber Martim, Rodrigo Amaral, Joes Nogueira, Luis Marchi

Robert Owen, Steven McAnany, Luke Zebala

Instituto de Patologia da Coluna (IPC), Sao Paulo, Brazil

Department of Orthopedic Surgery, Washington University in St Louis, St. Louis, USA

Summary: In this study, all patients undergoing elective lumbar fusion were referred to presurgical psychological screening. As result of the screening, patients received one of the following recommendations: favorable (green group), favorable with recommendation (yellow group) or non-favorable (surgeon reconsidered surgery and conservative measures). The patients who had received a ‘‘yellow flag’’ from the psychological assessment have higher level of depression, secondary gains and/or work compensation and evolve with worse results following elective lumbar spine surgery. Hypothesis: Presurgical psychological screening can point out which patients who tend to obtain poorer clinical outcomes following elective lumbar spine fusion. Design: Single center, retrospective and comparative study. Introduction: Low back pain is caused by several pathological entities and its perception can be altered by external factors, for example by some psychological and social factors. The objective of this study was to analyze surgical results following systematic presurgical screening in patients undergoing elective lumbar fusion. Methods: Patients with indication to elective lumbar spine surgery were evaluated in a presurgical psychological assessment. Patients received a favorable recommendation (green group) or a favorable recommendation with reservation (yellow group). The groups were compared using the following psychosocial parameters and clinical outcomes: demographic and clinical history, depression (HAD-D),

Background: Legacy patient reported outcome measures such as Oswestry Disability Index (ODI) and Visual Analog Score (VAS) have become essential for analyzing treatments for lumbar stenosis with claudication. Significant administrative burdens impose limits on completion of such measures. The Patient Reported Outcomes Measurement Information System (PROMIS) group developed a patient outcome measure to improve reporting of symptoms, function, and health and to reduce administrative burden. Despite early positive results, ODI and VAS have not been compared with PROMIS in patients with lumbar stenosis with claudication. The aim of this study is to compare ODI and VAS with PROMIS physical function and pain to determine their correlations in a surgical patient population longitudinally. Methods: 108 patients with a diagnosis of lumbar stenosis with claudication that went on to surgery were included in the study. All patients were treated at the same university spine center by four spine surgeons. Patients were seen and PROMIS, ODI, and VAS scores were collected preoperatively, at 1–4 months, and at 6 months. Data was extracted from a Redcap database. Correlations between ODI and VAS with PROMIS physical function and pain were quantified using

123

S354

Eur Spine J (2017) 26 (Suppl 2):S335–S405

anxiety (HAD-A), pain levels (VAS), disability (ODI) and quality of life (EQ-5D) at preop and at a 6/12-month follow-up. Results: Total 136 patients were included (51% female). 62.5% of total were allocated at the green group, and 37.5% in the yellow group. Between groups, similar mean values for pain were seen at the preop point, but the green group evolved with lower pain levels after surgery (p = 0.003). For the ODI and EQ-5D scales, the green group already had shown lower clinical impairment at preop (p = 0.009 and p = 0.003, respectively) and evolved with better outcomes at the final evaluation (p = 0.049 and p = 0.017). According to the all the clinical parameters studied, both groups have experienced similar and statistically significant clinical benefits after surgery. Conclusion: The result from presurgical screening can successfully assess the presence of psychological distress, which correlates with worse self-assessed clinical outcomes, both in the preoperative period and following the surgery. Despite the differences between groups, even the patients with mild psychosocial impairment did experience clinical benefits following the surgery. Disclosures: Author 1: consultant: Nuvasive, stock/shareholder: Nuvasive, royalties: Nuvasive; Author 2: none; Author 3: none; Author 4: consultant: Nuvasive; Author 5: none; Author 6: none.

Conclusion: In this study, it was possible to observe that WC is associated with worse clinical results following elective surgical treatment of the lumbar spine. Disclosures: Author 1: consultant: Nuvasive, stock/shareholder: Nuvasive, royalties: Nuvasive; Author 2: none; Author 3: none; Author 4: consultant: Nuvasive; Author 5: none.

P39 WORKER’S COMPENSATION DETECTED BEFORE THE SURGERY IS ASSOCIATED WITH WORSE CLINICAL RESULTS AFTER LUMBAR FUSION

Background: Lumbar spinal stenosis (LSS) is a prevalent degenerative spinal disease in the aged population. LSS is typically treated with medication and physiotherapy in the primary care setting. When conservative treatment fails for LSS, surgery is considered in integrated evaluation, including neurological findings, severity of symptoms, and scoring systems, such as the Japanese Orthopaedic Association (JOA) score and the Japanese Orthopaedic Association Back Pain Evaluation Questionnaire (JOABPEQ). However, which factors may be associated with the need for surgical intervention has not been established. This study aimed to identify factors that affect surgical outcome in patients with LSS. Methods: Among 75 patients who were diagnosed with LSS and underwent surgery, 34 (17 males and 17 females) were enrolled in this study. The mean age at surgery was 68.3 years and the follow-up was 1 year. According to the recovery rate of the JOA score, the patients were divided into two groups: satisfactory ([50%) and unsatisfactory (\50%) surgical outcome groups. Univariate analyses were performed to compare the two groups regarding various clinical findings, age, sex, comorbidity (hypertension, diabetes mellitus, and kidney disease), the waiting period from symptomatic onset to surgery, and each item of the JOA score and JOABPEQ. Multiple logistic regression analysis was then performed to identify the independent factors associated with an unsatisfactory surgical outcome in LSS. Results: Among the 34 patients, 11 (32.4%) had an unsatisfactory surgical outcome (2 had decompression alone and 9 had decompression with posterior fixation). Univariate analyses showed that lower extremity/bladder function of the JOA score and lumbar function, walking ability, and psychiatric factors of the JOABPEQ tended to be deteriorated in the unsatisfactory outcome group. There were no significant differences in age, sex, the presence of comorbidity, and the waiting period to surgery between the two groups. Multiple logistic regression analysis showed that decreased bladder function of the JOA score (odds ratio 2.0, 95% confidence interval 1.1–3.7) was independently associated with an unsatisfactory surgical outcome in LSS. Conclusion: This study shows that decreased bladder function of the JOA score is an independent factor associated with an unsatisfactory surgical outcome in LSS. This suggests that early surgical intervention should be considered once bladder dysfunction occurs. Our findings could provide clinical information for family physicians to refer patients with LSS to spinal surgeons without missing appropriate timing for surgery. Disclosures: Author 1: none; Author 2: none; Author 3: none; Author 4: none; Author 5: none; Author 6: none.

Luiz Pimenta, Joes Nogueira-Neto, Vivian Amaral, Rodrigo Amaral, Luis Marchi Instituto de Patologia da Coluna (IPC), Sao Paulo, Brazil Summary: This study aims to evaluate the association of surgical results with preoperative worker’s compensation (WC) in patients submitted to spine fusion. In this study, it was possible to observe that WC is associated with worse clinical results following elective surgical treatment of the lumbar spine. Objective: This study aims to evaluate the association of surgical results with preoperative worker’s compensation (WC) in patients submitted to spine fusion. Design: This is a retrospective, comparative, single center study. Introduction: Studies have shown that worker’s compensation (WC) can correlate with a negative prognosis after surgical treatment in several orthopedic pathologies, including in spine. A study from our country related the postoperative WC with inferior results in spine surgery, but did not analyze the presence of WC before the surgery was related to different results. Methods: Patients who underwent lumbar spine arthrodesis were included. The endpoints were pain scores (VAS), physical restraint (ODI) and quality of life (EQ-5D). Outcomes were analyzed before surgery and after surgery (minimum 6 and maximum 12 months of follow-up). Two groups were compared: with or without WC at the preoperative visit. Results: 132 cases (mean age 54 years old and 51% female) were analyzed, 29 (22%) allocated in the WC group. The groups had matched values for age, gender, and preoperative depression scores. In the preoperative evaluation, both groups showed similar scores for pain and physical disability, however the CT group presented poorer quality of life (p = 0.05). Although both groups showed improvement in clinical outcomes after surgery compared to baseline (p \ 0.05), worse scores were observed for the WC group compared to the non-WC group, respectively: EVA 4.9 vs 3.2 (p = 0.02), ODI 34.7 vs 23.4 (p = 0.002) and EQ-5D 0.56 vs 0.75 (p = 0.01).

123

P40 PREOPERATIVE BLADDER FUNCTION CAN BE ASSOCIATED WITH SURGICAL INDICATION FOR PATIENTS WITH LUMBAR SPINAL STENOSIS Hideo Sakane, Eiji Takasawa, Yasunori Sorimachi, Yoichi Iizuka, Haku Iizuka, Hirotaka Chikuda Department of Orthopeadic Surgery, Japanese Red Cross Maebashi Hospital, Maebashi, Japan; Department of Orthopeadic Surgery, Japanese Red Cross Fukaya Hospital, Fukaya, Japan; Department of Orthopeadic Surgery, Gunma University Graduate School of Medicine, Maebashi, Japan

Eur Spine J (2017) 26 (Suppl 2):S335–S405

S355

P41 TWO-LEVEL PLIF WITH CORTICAL BONE TRAJECTORY SCREW FIXATION VERSUS 2-LEVEL PLIF USING TRADITIONAL PEDICLE SCREW FIXATION FOR 2-LEVEL DEGENERATIVE LUMBAR SPONDYLOLISTHESIS: A COMPARATIVE STUDY

P42 SURGICAL OUTCOMES OF PLIF WITH CORTICAL BONE TRAJECTORY SCREW FIXATION FOR DEGENERATIVE LUMBAR SPONDYLOLISTHESIS. A COMPARATIVE STUDY WITH PLIF USING TRADITIONAL PEDICLE SCREW FIXATION

Hironobu Sakaura, Toshitada Miwa, Tomoya Yamashita, Yusuke Kuroda, Atsunori Ohnishi, Tetsuo Ohwada

Hironobu Sakaura, Toshitada Miwa, Tomoya Yamashita, Yusuke Kuroda, Atsunori Ohnishi, Tetsuo Ohwada

Department of Orthopaedic Surgery, Amagasaki, Japan

Department of Orthopaedic Surgery, Amagasaki, Japan

Objectives: Cortical bone trajectory (CBT) screw technique is a new alternative strategy to obtain improved pedicle screw (PS) fixation and several biomechanical studies have demonstrated favorable fixation strength of the CBT screw. Moreover, screw insertion through a caudomedial starting point provides advantages in limiting dissection of the superior facet joints and reducing muscles dissection and retraction and reduces the risk of superior facet violation by the screw. However, the biomechanical behavior of multi-level CBT screw-rod fixation remains unclear and surgical outcomes of 2-level posterior lumbar interbody fusion (PLIF) using CBT screw fixation have not been reported. We thus examined the surgical outcomes after 2-level PLIF with CBT screw fixation for 2-level degenerative lumbar spondylolisthesis (DS) and to compare these results with those following 2-level PLIF using traditional PS fixation. Methods: The subjects were 22 consecutive patients who underwent 2-level PLIF with CBT screw fixation for 2-level DS and were followed for 39 months in the mean (CBT group). As a historical control group, 20 consecutive patients who underwent 2-level PLIF with traditional PS fixation for 2-level DS and were followed for 35 months in the mean were enrolled (PS group). Clinical symptoms were evaluated using the Japanese Orthopaedic Association (JOA) score. Fusion status was assessed by dynamic plain radiographs and computed tomography. Surgery-related complications including symptomatic adjacent segment disease (ASD) were examined. Results: The mean operative duration and intraoperative blood loss were 192 min and 495 ml in the CBT group, and 218 min and 612 ml in the PS group, respectively (p \ 0.05 and p [ 0.05). The mean JOA score improved significantly from 12.3 points before surgery to 21.1 points (mean recovery rate 54.4%) at final follow-up in the CBT group, and from 12.8 points preoperatively to 20.4 points (mean recovery rate 51.8%) at final follow-up in the PS group (p [ 0.05). Solid spinal fusion was achieved at 90.9% of segments in the CBT group and 95.0% of segments in the PS group (p [ 0.05). Symptomatic ASD developed in 2 patients of the CBT group (9.1%) and 4 patients of the PS group (20.0%, p [ 0.05). Conclusions: Two-level PLIF with CBT screw fixation for 2-level DS could be less invasive and resulted in equal improvement of clinical symptoms to 2-level PLIF using traditional PS fixation. The incidence of symptomatic ASD tended to be lower in the CBT group than in the PS group, although the differences between the 2 groups were not significant. Disclosures: Author 1: none; Author 2: none; Author 3: none; Author 4: none; Author 5: none; Author 6: none.

Objective: Several biomechanical studies have demonstrated favorable fixation strength of the cortical bone trajectory (CBT) screw. Moreover, screw insertion through a caudomedial starting point provides advantages in limiting dissection of the superior facet joints and reducing muscles dissection and retraction and reduces the risk of superior facet violation by the screw. We expected that the less invasive approach associated with posterior lumbar interbody fusion (PLIF) using CBT screw fixation could result in better clinical outcomes such as less postoperative low back pain and the lower incidence of adjacent segment disease (ASD) than PLIF using traditional pedicle screw (PS) fixation. We thus investigated surgical outcomes after PLIF with CBT screw fixation for degenerative lumbar spondylolisthesis (DS) and compared these results with those after PLIF using traditional PS fixation. Methods: The subjects were 95 consecutive patients who underwent single-level PLIF with CBT screw fixation for DS and were followed for 35 months in the mean (CBT group). As a historical control group, 82 consecutive patients who underwent single-level PLIF with traditional PS fixation for DS and were followed for 40 months in the mean were enrolled (PS group). Clinical status was assessed using the Japanese Orthopaedic Association (JOA) score. Fusion status was assessed by dynamic plain radiographs and CT scan. Surgery-related complications and additional surgery were also evaluated. Results: The mean operative duration was 123 min in the CBT group and 145 min in the PS group, respectively (P \ 0.01). The mean estimated intraoperative blood loss was 205 ml in the CBT group and 204 ml in the PS group, respectively (P [ 0.05). The mean JOA score improved significantly from 13.7 points before surgery to 23.3 points at the latest follow-up (mean recovery rate 64.4%) in the CBT group, whereas from 14.4 points preoperatively to 22.7 points at final followup (mean recovery rate 55.8%, P \ 0.05) in the PS group. Successful fusion rate was 88.4% in the CBT group, and 96.3% in the PS group (P [ 0.05). Symptomatic ASD was developed in 3 patients of the CBT group (3.2%), on the contrary, in 9 patients of the PS group (11.0%, P \ 0.05). Conclusions: PLIF with CBT screw fixation for DS, compared with PLIF using traditional PS fixation, provided favorable postoperative improvement of clinical symptoms and reduced the incidence of symptomatic ASD after surgery. However, the fusion rate tended to be lower in the CBT group than in the PS group, although there was no significant difference. Disclosures: Author 1: none; Author 2: none; Author 3: none; Author 4: none; Author 5: none; Author 6: none.

123

S356 P43 PERCUTANEOUS VS OPEN PEDICLE SCREW FIXATION FOR TRANSPSOAS LATERAL LUMBAR INTERBODY FUSION Kotaro Satake, Tokumi Kanemura, Hiroaki Nakashima, Naoki Segi, Jun Ouchida Department of Orthopaedic Surgery, Konan Kosei Hospital, Konan, Japan Introduction: Percutaneous pedicle screw (PPS) is used increasingly as a supplemental fixation in transpsoas lateral interbody fusion (LLIF) due to its less invasive aspect. This retrospective review of a prospective cohort was performed to compare the radiological and clinical outcomes of the two pedicle screw (PS) fixation types [PPS and open PS (OPS)] for LLIF. Materials and methods: 51 patients (20 males and 31 females, 68.1 ± 10.6 years, 81 segments) who underwent LLIF (B3 segments) were enrolled in this study. All of them were followed for a minimum 2 years postoperatively. LLIF procedure was done with use of polyetheretherketone cages packed with allogenic bone, and all segments were supplemented with bilateral PSs. PPS placement was performed for 21 patients (28 segments) and OPS for 30 patients (53 segments). There were no significant differences between the two groups in patient backgrounds. Posterior bone graft was not performed in OPS group. The estimated blood loss (EBL) and complications related to PS placement were reviewed. Visual analogue scales (VASs) and the Japanese Orthopedic Association Back Pain Evaluation Questionnaire (JOABPEQ) were obtained from two groups and were compared. PS malposition, late-onset cage subsidence (C2 mm) and bone bridge formation in- and outside the cage (bony fusion) of each segment were evaluated by computed tomography (CT) 2 years postoperative. Results: EBL/segment were reduced significantly in PPS compared to OPS (34.0 ml ± 26.8 vs 169.1 ml ± 171.5, p = 0.001). Complications were 1 intraoperative and 1 postoperative back out in PPS, and 2 postoperative back outs in OPS. There were no significant differences in terms of VASs, improvements in all domains of JOABPEQ and the malposition rate of PSs (1.6 vs 1.0%) between the two groups. PPS demonstrated significantly higher incidence of late-onset cage subsidence (21.4 vs 1.9%, p = 0.006) and lower fusion rate (60.7 vs 90.6%, p = 0.002) than OPS. Conclusion: Although PPS reduced blood loss significantly, it demonstrated a significantly higher rate of late-onset cage subsidence and lower fusion rate than OPS though these did not affect the clinical outcomes. We should be aware of the possible inferiority of PPS compared to OPS for segment stabilization. Disclosures: Author 1: none; Author 2: consultant: NuVasive, Medtronic; Author 3: none; Author 4: none; Author 5: none.

P44 PROSPECTIVE COHORT STUDY ON THE EFFICACY OF A SINGLE POSTERIOR OBLIQUE CAGE (POLIF) IN DEGENERATIVE LUMBAR SPINE DISEASE: RESULTS AT MID-TERM FOLLOW-UP Laura Scaramuzzo, Fabrizio Giudici, Marino Archetti, Leone Minoia, Eleonora Caboni, Antonino Zagra I.R.C.C.S. Galeazzi Orthopedic Institute, Milan Italy Introduction: The use of a single oblique cage associated to posterior lumbar fusion in degenerative diseases is still a controversial issue.

123

Eur Spine J (2017) 26 (Suppl 2):S335–S405 Aim of the study was to evaluate in a prospectic way at midterm follow-up, the efficacy of a lumbar interbody fusion obtained by single oblique cage (POLIF) implanted by a posterior approach. Methods: From October 2013 to June 2016, 278 patients (124 male— 154 female) underwent interbody fusion by POLIF with posterolateral fusion and pedicle screw fixation. Preoperative diagnosis, intraoperative data (surgical time, blood loss), length of stay, complications were recorded in a dedicated database. Clinical outcome was evaluated at regular intervals by Visual Analog Scale (VAS) and SF-12. Radiographic evaluation included preoperative standard and dynamic X-ray, preoperative MRI, immediate postoperative X-ray and at regular intervals, associated to dynamic X-ray at 6 months and last follow-up. Cage subsidence, calculated as difference between mean disc height at immediate post-operative and last follow-up time, fusion grade and segmental lordosis were evaluated. A statistical analysis was conducted for parametric data with student t-test, with Chi square test for non-parametric data; significance was established for p \ 0.05. Results: 135 discopathies, 76 degenerative spondylolisthesis grade I according to Meyerding, 67 post-discectomy syndrome were treated. Minimum follow-up was 6 months, maximum 3 years. Intraoperative blood loss were 250 ml, mean length of stay was 5.7 days. Mean preoperative VAS score was 7.5 (5.7–8.8) decreased to 2.1 (1–5) p = 0.003 at 6 months minimum follow-up and to 1.2 (0.5–1.7) p = 0.004 at last follow-up. Mean preoperative SF-12 value was 34.5% (25.7–50.4%) increased to 75.4% (68.2–99.4%) p = 0.003 at 6 months minimum di follow-up and to 93.5% (73.5–99.6%) p = 0.002 at last follow-up. At last follow-up, all patients showed a good fusion rate, no cage subsidence, maintenance of a good segmental lordosis and no hardware failure. Conclusions: POLIF associated to instrumented posterolateral fusion is a viable and safe surgical technique. It allows reduction of operative time, complications, and costs needing a short learning curve ensuring satisfactory clinical and radiographic results at mid-term follow-up. Disclosures: Author 1: none; Author 2: none; Author 4: none; Author 5: none; Author 6: none.

P45 POST TRAUMATIC STRESS DISORDER SYMPTOMS FOLLOWING ELECTIVE SPINE SURGERY. SURGERY IS NOT THE REASON Ehab Shiban, Youssef Shiban, Bernhard Meyer Department of Neurosurgery, Technical University of Munich, Germany; Experimentelle Psychologie (Klinische Psychologie und Psychotherapie) Universita¨t Regensburg, Regensburg, Germany Objective: Recently, posttraumatic stress disorder symptoms (PTSD) were attributed to spine surgery. Furthermore, PTSD symptoms were associated with reduced clinical benefit. However, PTSD is seen in up to 10% of the normal population and there was no preoperative evaluation for PTSD symptoms. Aim of this study was to assess the incidence and influence of PTSD Symptoms on clinical outcome 1 year after surgery. Methods: A prospective study of patients undergoing elective spine surgery for degenerative disc disease was performed. The patients were evaluated for PTSD using the PTSS score before, 3 and 12 months after surgery. In addition SF36 physical composite score (PCS), Oswestry Disability Index (ODI), EuroQOL 5D questioner and pain visual analog scale (VAS) were completed preoperatively, 3 and 12 months after surgery. Incidence and influence of PTSD symptoms on clinical outcome were examined.

Eur Spine J (2017) 26 (Suppl 2):S335–S405

S357

Results: 184 patients met the inclusion criteria. 52.7% were male, mean age was 59.4 years. Abnormal PTSS scores were observed in 43.7, 20 and 22% of cases before, 3 and 12 months following surgery, respectively. However, only 8% developed PTSD symptoms only after surgery. At 1 year follow up patients with abnormal PTSS scores had a worse clinical outcome compared to their counterparts (SF36 PCS: 33.1 vs 40.0, P = 0,002; EuroQOL Index: 0.64 vs 0.78, P = 0.11, VAS Pain 5.6 vs 3.7, p = 0.001). Conclusion: PTSD symptoms are associated with worse outcome following elective spine surgery. However, the vast majority of patients that exhibit PTSD symptoms had already exhibited symptoms before surgery. Thereby the PTSS score seems nonspecific in the preoperative setting. Disclosures: Author 1: none; Author 2: none; Author 3: grants/research support: Relievant, Medtronic, Ulrich Medical, consultant: Relievant, Medtronic, Nexstim, royalties: Spineart.

Disclosures: Author 1: none; Author 2: none; Author 3: grants/research support: Relievant, Medtronic, Ulrich, Icotec, consultant: Relievant, Medtronic, Ulrich, Icotec, royalties: Spineart.

P46 ELECTIVE SPINE SURGERY FOR PATIENTS OLDER THAN 90-YEARS OLD: IS 90 THE NEW 80?

Study design: Retrospective and prospective study. Purpose: To find out which radiographic parameters are associated with successful indirect decompression with lateral lumbar interbody fusion (LLIF) and to confirm effect of indirect decompression with LLIF. Summary of background: Lateral lumbar interbody fusion is one of methods to treat lumbar degenerative diseases. Indirect decompression effect of lateral lumbar interbody fusion was reported in improvement of postoperative radiographic and clinical outcome. Although it indirect decompression was confirmed, few reports concerned about which parameters are associated with successful indirect decompression. Methods 60 patients presenting with lumbar degenerative conditions underwent lateral lumbar interbody fusion at 84 lumbar levels. Preoperative and postoperative plain film and MRI were obtained to access disc height (DH), foraminal height (FH), translation (TL), segmental disc angle (SDA), spinal canal dimeter (SCD) and cross-sectional area of the thecal sac (CSA). In preoperative images different between standing (plain film) and supine (MRI) of D DH, D FH, D TL and D SDA were recorded. Increase postoperative CSA [30% is criteria for successful indirect decompression. The relationships between DH, FH, TL, SDA, D DH, D FH, D TL, D SDA and successful indirect decompression were assessed by correlational analysis. Results 84 lateral lumbar interbody fusions were performed successfully without neural complications. Disc height, foraminal height, spinal canal diameter and cross-sectional area of the thecal sac improved significantly. Multivariate regression analysis demonstrated that the preoperative standing foraminal height (P = 0.044), supine disc height (P = 0.047) and CSA (P = 0.045) were factors that correlated with successful indirect decompression. By adding cut point to standing FH [ 20.19 mm. (sensitivity = 83%, specificity = 86%), supine DH [ 7.97 mm. (sensitivity = 83%, specificity = 71%), CSA \60.9 mm2. (sensitivity = 83%, specificity = 71%). Mean CSA increase was 20.12%. Conclusion: Preoperative standing high foraminal height, high supine disc height and low cross-sectional area of the thecal sac are correlate with radiographic successful indirect decompression with LLIF. Disclosures: Author 1: none; Author 2: none; Author 3: consultant: Medtronic; Author 4: none.

Ehab Shiban, Jens Lehmberg, Bernhard Meyer Department of Neurosurgery, Technical University of Munich, Munich, Germany Introduction: Demographic trends make it incumbent on spine surgeons to recognize the special challenges involved in caring for older patients. Aim of this study was to identify variables that may predict early mortality in geriatric patients over the age of 90. Methods: Retrospective analyses of all patients over the age of 90-year, which were treated between 2006 and 2014 at our department for degenerative spine disease, were performed. Patient characteristics, type of treatment and comorbidities were analyzed with regards to the 30-day mortality rate. Results: 25 patients were identified. Mean age was 92.8 years (range 91–101), 21 (84%) patients were female. 16 (64%) patients were on anticoagulation therapy. 17 (68%) patients were treated operatively. Mean hospital stay was 14 days (range 2–40). Mean Charlson comorbidity index was 5.5 (range 0–12) and mean diagnosis count was 12 (range 2–24). The 30-day mortality rate was 17% in the surgically treated group compared to 0% in the conservatively treated group (p = 0.2). Gender (p = 0.42), diagnosis count (p = 0.65), charlson index (p = 0.65) and anticoagulation therapy (p = 0.9) did not correlate with the 30-day mortality rate. Cause of death was pulmonary embolism in two cases and was unknown in one case. Conclusion: 30-day mortality rate in patients over 90-year-old following elective spine surgery is very high. Standard geriatric prognostic scores seem less reliable for these patients. Prospective validations studies are needed in order to establish treatment recommendations for such patients.

P47 PREDICTIVE PARAMETERS FOR SUCCESSFUL INDIRECT DECOMPRESSION WITH LATERAL LUMBAR INTERBODY FUSION Weerasak Singhatanadgige, Suthee Laokomen, Worawat Limthongkul, Wicharn Yingsakmongkol Department of Orthopedic Surgery, Faculty of Medicine, Chulalongkorn University and King Chulalongkorn, Memorial Hospital, Thai Red Cross Society, Bankok, Thailand

123

S358 P48 RADIOGRAPHIC OUTCOME OF MIS LATERAL AND MIS TRANSFORAMINAL LUMBAR INTERBODY FUSION IN THE TREATMENT OF DEGENERATIVE LUMBAR DISEASE: A SINGLE INSTITUTE STUDY Weerasak Singhatanadgige, Worawat Limthongkul, Wicharn Yingsakmongkol Chulalongkorn University and King Chulalongkorn Memorial Hospital, Thai Red Cross Society, Bangkok, Thailand Abstract Study design: Retrospective, Single institute, institutional review board cohort study. Objective: The purpose of this study was to compare radiographic outcomes between minimally invasive transforaminal lumbar interbody fusion (MIS TLIF) or MIS lateral lumbar interbody fusion (LLIF) in the treatment of patients with degenerative lumbar disease Background: Few studies compared radiographic results between MIS TLIF and LLIF. This study is the single-institute comparative study operated by the same and consistent surgical technique. Methods: Patients who were operated by MIS TLIF or LLIF for one or two levels degenerative lumbar disease between 1 January 2015 and 31 December 2016 were included. Radiographic parameters include translation, anterior disc height (ADH), posterior disc height (PDH), foraminal height (FH), segmental angle, lumbar lordosis (LL), pelvic incidence (PI), pelvic tilt (PT), sacral slope (SS) were measured at preoperative, immediate post op, 1 month, 3 months, 6 months, 12 months, 18 months and last follow up. Computer tomography (CT) scan was done at 12 months post-operative to evaluate fusion. Magnetic resonance imaging (MRI) at preoperative and 3 months post-operative were used to evaluate changing of spinal canal area. Fisher exact and ANOVA were used to compare the measuring parameters from both groups. Results: Totally 97 patients (45 MIS TLIF and 52 LLIF) with 114 operative levels (51 MIS TLIF and 63 LLIF) were selected. LLIF resulting in higher foraminal height restoration at immediate and 1 month follow up (p \ 0.01). MIS TLIF group showed higher segmental angle and LL (p \ 0.05). There were higher PT restoration and lower SS for MIS TLIF group. There was no difference of ADH and PDH between groups. Conclusion: MIS TLIF showed better correction of segmental angle, LL and Spinopelvic parameters (PT and SS). LLIF resulting in better foraminal height restoration. Disclosures: Author 1: none; Author 2: consultant: Medtronic; Author 3: none.

P49 A COMPARISON BETWEEN REPEAT DISCECTOMY VERSUS FUSION FOR THE TREATMENT OF RECURRENT LUMBAR DISC HERNIATION: SYSTEMATIC REVIEW AND META-ANALYSIS Weerasak Singhatanadgige, Chotawan Tanavalee, Worawat Limthongkul, Wicharn Yingsakmongkol Department of Orthopedic Surgery, Faculty of Medicine, Chulalongkorn University and King Chulalongkorn, Memorial Hospital, Thai Red Cross Society, Bangkok, Thailand Abstract Background: Recurrent lumbar disc herniation is a common negative sequelae from primary discectomy procedure. Repeat discectomy

123

Eur Spine J (2017) 26 (Suppl 2):S335–S405 versus spinal fusion were debated. Successful outcomes from both treatments remain unclear. Several meta-analysis studies did not mention to re-operative rate between two treatment methods. Hence, this point was our interest. Study design: Systematic review and meta-analysis. Objective: The primary objective was to compare re-operative rate between repeat discectomy and fusion treatment. The secondary objectives included comparing clinical improvement, operative time, blood loss, complications and post-operative hospital stay between repeat discectomy and fusion treatment. Methods: Study’s source from PubMed Medline, Ovid Medline, Cochrane database and Google scholar of articles. Only comparative studies in human were selected. Two reviewers selected studies and classified quality of the studies. Data analysis using comprehensive meta-analysis version 2 software. Result: There were initially 182 citations found. After the analysis, we found four studies which were compatible with our inclusion criteria. The meta-analysis showed that the re-operative rate was seem to be higher in discectomy group (9.09%) compared to fusion group (2.00%). However, there was no statistically significant difference. The main reason for reoperation in discectomy group was recurrent disc herniation. The main reason for reoperation in fusion group was implant failure. The operative time and postoperative stay were significant less in discectomy group. The improvement rate was not difference between groups. Conclusion: The reoperation after fusion seem to be less than repeat discectomy for recurrent disc herniation, however there was no statistical significant difference. Both treatment technique have equal improvement rate and complications. Disclosures: Author 1: none; Author 2: none; Author 3: consultant: Medtronic; Author 4: none.

P50 ANALYSIS OF RISK FACTORS FOR WORSENING LOW BACK PAIN IN RESIDENTS IN YONAGO, JAPAN: WORSENED IN THE REGION RESIDENTS Shinji Tanishima, Hiroshi Hagino, Hiromi Matsumoto, Hideki Nagashima Department of Orthopedic Surgery, Faculty of Medicine, Yonago, Japan Introduction: Several factors, such as osteoporosis, obesity, and lack of exercise, contribute to low back pain. This longitudinal study aimed to investigate the risk factors for low back pain in local residents of Yonago, Japan. Materials and methods: Our study conducted in 2014 comprised 96 participants (34 males and 62 females; mean age of 73.9 years at beginning of the study) of general medical examination, living in Yonago, Japan, who provided informed consent. We assessed low back pain using the visual analog scale (VAS) and divided local residents into two groups. Group A included residents with worsening

Eur Spine J (2017) 26 (Suppl 2):S335–S405 low back pain experienced from 2014 to 2016. Group B included residents who experienced no change in low back pain during the study period. We compared body mass index (BMI), bone mineral density (%YAM), skeletal muscle mass, standing posture, and exercise frequency, which was defined as physical activity at least twice a week. We then used logistic regression analysis to identify the risk factors for low back pain. Results: 40.8% of group A residents exercised regularly and 81.2% of group B residents exercised regularly. There was significantly difference between group A and B (P \ 0.001). There were no significant differences in BMI, skeletal muscle mass, standing posture, and bone density between the two groups. Logistic regression analysis revealed lack of exercise, as assessed using VAS, as a significant risk factor for worsening low back pain (odds ratio 0.19; P = 0.002; 95% confidence interval 0.07–0.53). In addition, as per VAS assessment, average Bone mineral density of residents who worsened low back pain in spite of getting regular exercise was 75.4% (%YAM) and that of residents did not worsen low back pain without getting exercise was 83.4%. There was significantly difference (P \ 0.04).The residents with who had low bone density did not get benefit from its preventive effects. Conclusion: Our results indicate that regular exercise is useful in preventing low back pain. However, people with low bone density did not get enough exercise to benefit from its positive, preventive effects. As the residents in this region tend to be elderly, we speculate that the low back pain is associated with osteoporosis. Although exercise is typically suggested to prevent low back pain in patients, it may not be effective in preventing low back pain associated with osteoporosis. Disclosures: Author 1: none; Author 2: none; Author 3: none; Author 4: grants/research support: Nippon Zoki Pharmaceutical Co., Ltd., Teijin Pharma, Ltd., Taisho Toyama Pharmaceutical Co., Ltd.

S359 statistical comparison between cohorts. Pseudoarthosis was determined by 2-D computed tomography at [1 year f/u. Results: 184 pts were included in the analysis. mTLIF had 80 pts, 35 females/45 males. Mean age 51.5 (23–75), 6 revisions. TLIF had 104 pts, mean age 51 years (14–74), 62 males/42 females, 12 revisions. No significant differences were seen between the two cohorts with respect to mean age (51.5 vs 51.0 years), sex, medical comorbidities, number of levels fused, or revision procedures. Respective pseudoarthrosis rates were 13.7 and 11.7% (p = 0.07). There was a significantly higher percentage of pseudoarthosis in pts undergoing revision procedures (p = 0.02). No statistical differences were seen between choices of interbody bone graft material. No statistical differences were seen with ODI or VAS scores at 2 years. Conclusion: mis-TLIF and open TLIF demonstrated comparable outcomes with regards to pseudaoarthrosis rates. Revision surgery was the only identifiable risk factor for pseudoarthrosis. Disclosures: Author 1: none; Author 2: grants/research support: K2M, consultant: K2M, DePuy-Synthes; Author 3: none; Author 4: none; Author 5: none; Author 6: none; Author 7: none; Author 8: none.

TRAUMA, THORACOLUMBAR SPINE

P52 SURGICAL OUTCOME OF THORACOLUMBAR BURST FRACTURE WITH SEVERE NEUROLOGICAL DEFICIT: A COMPARISON BETWEEN DIRECT DECOMPRESSION BY ANTERIOR CORPECTOMY AND INDIRECT DECOMPRESSION BY POSTERIOR LAMINECTOMY CHIEN-JEN HSU

P51 PSEUDOARTHROSIS RATE IN MINIMALLY INVASIVE TRANSFORAMINAL LUMBAR INTERBODY FUSION(MTLIF): 2-YR OUTCOMES VS OPEN TLIF Daniel Thibaudeau, Kimona Issa, Nikhil Sahai, Conor Dunn, Sina Pourtaheri, Michael Faloon, Kumar Sinha, Ki Soo Hwang, Arash Emami Seton Hall University, School of Health and Medical Sciences, South Orange, USA Summary: This study compared the pseudarthrosis rates between a two consecutive cohorts of patients that had either undergone mTLIF or open TLIF. Statistical analysis identified only revision surgery as a relevant risk factor for pseudoarthrosis. Hypothesis: Differences in the rate of pseudoarthrosis may be seen between open and mis-TLIF Design Retrospective clinical cohort Introduction MIS-TLIFs has shown similar long-term clinical outcomes with decreased perioperative morbidity and earlier return to work as compared to open TLIFs. However, the rate of pseudarthrosis and the impact of various demographic or comorbid factors have not been evaluated. Methods: Between 2006 and 2012, 230 consecutive pts underwent one or two-level mTLIF or open TLIF at a single institution. Clinical and radiographic data was reviewed. Pts with complete medical records and minimum 2 years f/u were included in the analysis. Demographic data, medical comorbidities, surgical and radiographic data as well as pt assessed outcomes scores (ODI, VAS) were quantified. Binomial continuous and categorical tests were used for

Department of Orthopaedics, VGHKS, Kaohsiung, Taiwan Purpose: To investigate the long-term outcomes of patients with severe neurological impairment caused by thoracolumbar burst fractures after either direct decompression by anterior corpectomy or indirect decompression by posterior laminectomy. Materials and methods: We conducted a retrospective, cohort study on 42 patients with severe neurological impairment caused by thoracolumbar burst fractures that were operated from January 2003 to December 2010. The injury mechanisms were 21 accidental fall, 19 traffic accidents, and 2 direct blows by heavy objects. The injured levels were 3 T11, 14 T12, 16 L1, 6 L2 and 3 L3. We excluded the patients with pre-operative neurological function of Frankel D and E. Twenty patients that were operated by anterior decompression with vertebral reconstruction after posterior instrumentation were allocated into group I. The other 22 patients that were treated by posterior laminectomy after posterior instrumentation were allocated into group II. No significant difference was disclosed in gender distribution, injured level, and mean age between two groups. The vertebral reconstructions utilized autogenous iliac tricortical bone for 6 patients and mesh cage containing fractured bone fragments for 38 patients. Rehabilitation program was arranged by 2 weeks postoperatively. The postoperative follow-up included SF-36 scores and radiographic assessment at every 6-week interval in first 3 months and thereafter every 3-month at least for 2 years. Results: The average follow-up time was 5.5 years. No operative complication required further surgical treatments except two superficial wound infection in group I. The preoperative neurological status according to the Frankel classification was 4 Frankel C, 10 Frankel B and 6 Frankel A in group I and 5 Frankel C, 11 Frankel B and 6

123

S360

Eur Spine J (2017) 26 (Suppl 2):S335–S405

Frankel A in group II. At the latest follow-up, patients of group I restored to 5 Frankel E, 8 Frankel D, 1 Frankel C, 3 Frankel B, 3 Frankel A. Thirteen patients recovered independent ambulation. The average recovery is 1.5 Frankel grades. In group II, the patients restored to 1 Frankel E, 1 Frankel D, 7 Frankel C, 7 Frankel B, 6 Frankel A. Two patients recovered independent ambulation. The average recovery is 0.3 Frankel grade. No screw loosening or cage dislodgement developed throughout the follow-up in all patients. Conclusion: Anterior corpectomy provided adequate decompression to eliminate the compression on spinal cord. The neurological function therefore recovered more than indirect decompression by posterior laminectomy. The vertebral reconstruction with posterior instrumentation maintained long-term spinal stability and provided patients with favorable conditions for rehabilitation. Patients with severe neurological deficit after thoracolumbar burst fractures achieved significantly better neurological function recovery by direct decompression of anterior corpectomy in the long-term follow-up. Disclosures: Author 1: none.

P53 RANDOMIZED, CONTROLLED, CLINICAL TRIAL COMPARING OF THREE SURGICAL TREATMENTS FOR UNSTABLE FRACTURES OF THE THORACOLUMBAR SPINE IN CHINA Xiongke Hu, Qinpeng Zhao, Xinhua Yin

P54 LUMBOPELVIC STABILIZATION IN SACRAL FRACTURE Mona Khoury, Kristian Welle, Cornelius Jacobs, Christof Burger, Koroush Kabir

Spinal Surgery, Honghui Hospital, Xi’an, China

Department of Orthopaedics and Trauma Surgery, University Hospital Bonn, Bonn, Germany

Study design: Prospective, randomized, controlled. Objective: To compare the outcomes of three surgical treatments for unstable fractures of the thoracolumbar spine. Summary of background data: Unstable thoracolumbar fractures can lead to pain and limited mobility of patients, and surgical treatment can restore the stability of the spine and avoid kyphosis. Their management may be challenging and controversy remains over the ‘‘best’’ surgical treatment. Methods: 90 patients with thoracolumbar fractures were randomly divided into 3 groups. The surgical treatments included open reduction and internal fixation (group 1), conventional percutaneous pedicle screw fixation system (group 2), novel percutaneous pedicle screw fixation system (the screws have a certain angle and the tail side with hat ring, group 3). Results: There were no statistical differences between three groups in age, injury site, Cobb angle, anterior height of fracture vertebral body, VAS and ODI (P [ 0.05). Group 1 had significant differences in operative time, intraoperative blood loss, hospitalization time and postoperative VAS score compared with group 2 and group 3 (P \ 0.01). Group 1 and group 3 had significant differences in Cobb angle restoration and fracture height restoration rate after surgery compared with group 2 (P \ 0.01). Group 2 and group 3 had no complications, whereas group 1 had one case with superficial skin incision infection. There was no significant difference in VAS score and ODI between the three groups at the last follow-up (P [ 0.05). Conclusion: The novel percutaneous pedicle screw not only has a good effect of fracture reduction, but also significantly reduce the operation time, intraoperative blood loss, hospitalization time and postoperative VAS score. Under the premise of strictly grasp indications, this pedicle screw system can also serve as a good choice. Keywords: Thoracolumbar fractures; Minimally invasive surgery; Percutaneous pedicle screw fixation; Pedicle screws. Disclosures: Author 1: none; Author 2: none; Author 3: none.

Background: Lumbopelvic stabilization allows the fixation of sacral fractures. Aim of our retrospective study was to define the outcome of after lumbopelvic stabilization in patients who suffered a sacral fracture. Methods: A retrospective analysis was performed by searching the database of our clinic for ICD-10-code S32 in the space of time from 01-01-2013 to 12-15-2016 followed by a selection for lumbopelvic stabilization. Respecting those criteria 18 patients could be included in this study. Results: The indication for open procedures was a fracture Denis type II or III with a neurological deficit or a fracture type 2 to 4 grouped after the Roy–Camille classification. 7 patients had an open access operation, while the procedure was performed with a minimal-invasive technique in 11 patients. The mean age of patients with an open procedure was 54 years; percutaneous operated Patient exhibited a middle age of 64 years. A type II fracture after Denis was seen in 3 cases; a type III fracture was diagnosed in 15 times. Only five patients had a monotrauma. In 11 cases a fall causes the trauma, in 6 cases a traffic accident and in one case an entombment. In the group with open procedure, four of seven patients showed a neurological deficit preoperative. In all patients a postoperative neurological improvement could be documented and in one case a postoperative paresis of the dorsiflexion occurred. Only in open procedure group postoperative infections arises (n = 2). The impairment by pain or implants was approximately equal in both groups. The mean inpatient residence time of patients receiving a minimalinvasive procedure was lower. Blood loss and operating time was significant lower in the minimalinvasive group. In all cases the correct position of implants could be verified in postoperative computed tomographies.

123

Eur Spine J (2017) 26 (Suppl 2):S335–S405 Conclusions: Lumbopelvic stabilization allows a sufficient therapy for complex fracture of the sacral fractures with spinopelvic dissociation patterns. The indication for open or minimal-invasive procedure depends on the grade of dislocation or possible neurological deficits. An inferior outcome after open procedures relating to persisting neurological deficits, postoperative infections and residence time is recorded. But it is important to reflect that especially those patients suffered the more complex trauma particularly accompanied by other severe injuries. Above all patients with a preoperative neurological deficit were only treated with an open procedure in our study. A direct comparison of both methods should not be performed because of different indications for both procedures. However a percutaneous procedure should be recommended for patient without any neurological deficit, because of reduced operating time and reduced blood loss. Disclosures: Author 1: none; Author 2: none; Author 3: none; Author 4: none; Author 5: none.

P55 SURGICAL TREATMENT OF GROSS RIGID POSTTRAUMATIC DEFORMITIES OF THORACIC AND LUMBAR SPINE Aleksey Shul’ga, Igor’ Norkin, Vladimir Zaretskov, Aleksey Smolskin Research Institute of Traumatology, Orthopedics and Neurosurgery of Saratov State Medical University of Razumovsky, Saratov, Russia Objectives: The most common cause of rigid posttraumatic deformities of thoracic and lumbar spine is the non-substantiated use of conservative methods and inadequate surgical methods. The aim of the study is to evaluate surgical treatment results of gross rigid posttraumatic deformities in thoracic and lumbar spine. Methods: We conducted surgical treatment results analysis of 23 patients with gross rigid posttraumatic deformities in thoracic and lumbar spine. 14 patients were indicated to two-stage and 9 patients—to three-stage surgical interventions. While choosing the treatment we considered damage level, deformity character and postinjury time. Standard investigation procedure included neurological status assessment (ASIA/IMSOP), trauma character (AO/ASIF), pain intensity (VAS) and quality of life score (ODI). Results: We achieved adequate spinal cord and roots decompression, full deformity correction and stable spine fixation in all surgical cases. 16 patients (69.5%) noticed complete pain regression postsurgically and 7 patients (30.5%) reported decreased pain syndrome scoring 1-2 VAS. A number of injuries with complicated trauma (12 patients) showed significant (2–3 points on Harrison scale) and moderate (4 patients—1–2 points Harrison) increase of muscular strength in lower limbs. Conclusion: Surgical treatment of rigid posttraumatic deformities of spinal column is laborious and traumatic. Positive result is determined by adequate mobilization of the damaged segment, vast spinal cord decompression, deformity correction with the respect to anatomic and biomechanical features of thoracic and lumbar spine. Following these rules it is possible to restore sagittal balance of spinal column without the risk of neurological symptoms in patients. Disclosures: Author 1: none; Author 2: none; Author 3: none; Author 4: none.

S361 P56 MINIMALLY INVASIVE PEDICLE SCREW SHORTSEGMENT FIXATION COMBINED WITH PERCUTANEOUS VERTEBROPLASTY FOR THORACOLUMBAR BURST FRACTURE CAN PRESERVE THORACOLUMBAR MOTION AND THE ACTIVITIES OF DAILY LIFE REGARDLESS OF POST-TRAUMATIC DISC DEGENERATION Eiji Takasawa, Yasunori Sorimachi, Yoichi Iizuka, Haku Iizuka, Hirotaka Chikuda Department of Orthopaedic Surgery, Japanese Red Cross Maebashi Hospital, Maebashi, Japan; Department of Orthopaedic Surgery, Gunma University Graduate School of Medicine, Maebashi, Japan Background: For early rehabilitation from spinal trauma, patients are treated with minimally invasive spinal surgery (MIS), such as percutaneous pedicle screw (PPS) fixation and vertebroplasty (PVP). Although early clinical results for short-segment posterior fixation with PPS and PVP for thoracolumbar burst fractures have been satisfactory, few studies have reported the clinical outcomes for patients after implant removal. This study evaluated the post-traumatic disc/facet degeneration and range of motion (ROM) of the thoracolumbar spine after implant removal. Methods: Thirteen patients with thoracolumbar burst fractures, who were operatively treated within 72 h of admission, were enrolled. In addition to indirect reduction of the vertebral body with ligamentotaxis following pedicle screw insertion, transpedicular direct endplate reduction and intracorporeal hydroxyapatite (HA) grafting for the fractured vertebral body were performed under fluoroscopic imaging. Implants were removed about 1 year after surgery. Patient age, body mass index (BMI), fractured vertebral level, kyphotic deformity, disc/facet degeneration on MRI, spinal ROM, and back pain were recorded and evaluated postoperatively. Results: The mean age and BMI at surgery were 54.4 years and 24.3, respectively. Fractured vertebral levels were T11: 1, T12: 4, L1: 4, and L2: 4. Sagittal alignment was improved from a mean preoperative kyphosis of 18.5 to 11.5 with surgery but this deteriorated to 20.6 at the final follow-up after implant removal. The ratio of anterior and posterior vertebral body height was also improved from a mean preoperative ratio of 0.61–0.79 post-operatively; this improvement was preserved at 0.74 at the final followup. In terms of disc degeneration, the signal intensity of the cranial disc adjacent to the fractured vertebral body decreased by one grade from the preoperative to the follow-up MRI after implant removal in 12/13 cases (92.3%). Caudal disc degeneration was found in only 3/13 cases (23.0%) and was associated with higher BMI. Facet degeneration was found in 3/13 cases (23.0%) and was associated with older age; these changes were already observed at injury. With respect to back pain in activities of daily life, nine patients did not report back pain, three reported occasional, minimal pain, and one reported moderate pain; none reported severe pain or needed daily use of analgesics. After implant removal, flexion–extension radiographs revealed a mean ROM of 6.7 (range -0.1 to 10.6). Conclusion: Vertebral body height is preserved after implant removal following MIS by PPS fixation and PVP with posterior reduction. Our results demonstrate that this procedure does not require fusion to an affected segment, thereby preserving thoracolumbar motion and daily life activities regardless of post-traumatic disc degeneration. Disclosures: Author 1: none; Author 2: none; Author 3: none; Author 4: none; Author 5: none.

123

S362

Eur Spine J (2017) 26 (Suppl 2):S335–S405

P57 SARCOPENIA, BUT NOT FRAILTY, PREDICTS THE OCCURRENCE OF ADVERSE EVENTS AFTER EMERGENT SURGERY FOR METASTATIC DISEASE OF THE SPINE E´tienne Bourassa-Moreau, Anne Versteeg, Raphaele Charest-Morin, Charles Fisher, Micheal Boyd, Brian Kwon, Marcel Dvorak, Tamir Ailon, Scott Paquette, John Street

TUMOR, THORACOLUMBAR SPINE

P58 APPARENT DIFFUSION COEFFICIENT AND EXTRAMEDULLARY, INTRADURAL SPINAL CORD TUMORS: PRE-OPERATIVE ASSESSMENT OF TUMOR AGGRESSIVE BEHAVIOR

Vancouver Acute Spine Program, Vancouver, Canada Vadim Byvaltsev, Ivan Stepanov, Andrei Kalinin, Evgeny Belykh Introduction: Frailty and Sarcopenia are often considered synonymous, and both have been shown to predict adverse events (AE’s) in a number of surgical populations. Patients with metastatic disease to the spine may be either frail or sarcopenia or both. The aim of this study was to investigate the independent prognostic values of frailty and sarcopenia for the occurrence of AE following emergent surgery for metastases. Methods: An ambispective study of 108 patients, undergoing emergent surgery for spinal metastases from 2009 to 2016. Data included: demographics, tumour type and burden, neurological status, surgical and nonsurgical treatment. AE’s, including mortality were identified using SAVES V2. Sarcopenia was measured by normalized total psoas area (NTPA) on pre-op CT. Frailty was defined by both Modified Frailty Index (mFI) and the Metastatic Spinal Tumor Frailty Index (MSTFI), calculated from extensive detailed co-morbidity data. Logistic regression were also used for statistical analysis. Results: 85% of patients had an post-op AE including (SSI 4%, delirium 19%, UTI 35%). 25% of the patients died within 3 months of surgery. NTPA measurement took approximately 30 s per patient. There was a statistically significant relation between NTPA quartiles and the occurrence of AE (p \ 0.01) and early mortality (p \ 0.05). Post-op AE’s occurred in 70% of non-sarcopenic (NTPA ratio[0.93) and 93% of sarcopenic (NTPA ratio\0.93) patients (p \ 0.05). 15% of non-sarcopenic patients and 31% of sarcopenic patients within 3 month after surgery (p \ 0.05). mFI did not correlated with intra-op or post-op AE’s or mortality. Conclusions: Sarcopenia, as measured rapidly by psoas muscle area on CT, independently predicts post-operative mortality and post-op adverse events in patients undergoing emergent surgery for spinal metastasis. The Modified Frailty Index, requiring data often not available in emergency situations, did not predict AE or life-expectancy. Sarcopenia, rather than Frailty, may be a more appropriate indicator of vulnerability and adverse outcome in patients with metastatic disease of the spine. Disclosures: Author 1: none; Author 2: grants/research support: AOSpine Start Up Grant, consultant: AOSpine International; Author 3: none; Author 4: grants/research support: OREF—paid to institution, consultant: Medtronic, Nuvasive, royalties: Medtronic; Author 5: none; Author 6: consultant: Acorda Therapeutics; Author 7: grants/ research support: Medtronic, AO Spine, Synthes, DePuy, consultant: Medtronic, royalties: Medtronic; Author 8: none; Author 9: none; Author 10: grants/research support: MEDTRONIC.

AEs Rate 1

P=0,01 0,8

Mortality < 3month 40,0%

P=0,03

1

Irkutsk State Medical University, Irkutsk, Russia; 2Railway Clinical Hospital, Irkutsk, Russia; 3Scientific Center of Surgery and traumatology, Irkutsk, Russia; 4Irkutsk State Medical Academy of Continuing Education, Irkutsk, Russia Introduction: Spinal tumors comprise 15% of all CNS tumors. Their annual incidence is 2–10 per 100,000. Extramedullary, intradural spinal cord tumors (EISCTs) are rare. They comprise about 53–68.5% of all spinal cord tumors. The main biomarkers of aggressive behavior have been identified in the pathology literature, including the proliferative marker Ki-67. In the literature, diffusion-weighted imaging (DWI) and low apparent diffusion coefficient (ADC) values provide similar markers of aggressive behavior in brain tumors. The aim of this study was to determine if there is a correlation between ADC and Ki-67 in EISCTs. Materials and methods: A retrospective analysis of immunohistochemical characteristics and diffusion-weighted imaging of EISCTs was performed. The specimen indices and ADC values were measured. Linear regression analysis of ADC values and Ki-67 was used to compare these numerical parameters. Results: There were 31 patients with Ki-67 indices and ADC maps. The signs of tumor aggressive behavior (severe deformation of subarachnoid space, blurring nerve roots, cauda equina asymmetry) was confirmed by standard T1- and T-2 WI and intraoperative visualization in nine patients. The mean Ki-67 index for the nonaggressive group was 0.68%, with a mean ADC value of 1137 mm2/s. In the aggressive group, the mean ADC value was 734 mm2/s, with a mean Ki-67 index of 2.5%. Linear regression analysis of ADC values and Ki-67 demonstrates a strong negative correlation (p = 0.002, r = -0.79). Conclusions: Thus, the present study found a strong correlation between Ki-67 and low ADC values. This correlation demonstrates the potential of DWI as a possible biomarker for aggressive malignant behavior EISCTs, which may ultimately affect the surgical approach and postoperative management. Disclosures: Author 1: none; Author 2: none; Author 3: none; Author 4: none.

P59 SCORING SYSTEMS AND SURGICAL CANDIDACY IN VERTEBRAL METASTASIS: A REVIEW

30,0%

John Tristan Cassidy, Joseph F. Baker, Brian Lenehan

0,6 20,0% 0,4

Department of Trauma and Orthopaedics, University Hospital Limerick, Limerick, Ireland

10,0%

0,2

0,0%

0 Below Median

Above Median

Below Median Above Median

PSOAS U Median= 0,93

123

Introduction: Numerous scoring systems have been developed to help differentiate surgical candidates amongst patients with vertebral metastases. This review considers those scoring systems most relevant to contemporary orthopaedic practice.

Eur Spine J (2017) 26 (Suppl 2):S335–S405 Methods: Eight scoring systems were selected for review including seven prognostic systems, four designed for surgeons (Tokuhashi, Tomita, Bauer and Oswestry Spinal Risk Index) along with three other widely used scoring systems (Katagiri, Rades and Van der Linden). The eighth system was the Spinal Instability Neoplastic Score (SINS) which represents a novel approach compared to historical scoring systems which focus on prognosis. Selection criteria for scoring systems is outlined in Table 1. In 2017 literature search was performed for each scoring system and all identified papers were examined. Results: The Bauer and Oswestry Spinal Risk Index (OSRI) have the most accurate prognositic predictive ability, with the newly OSRI being favoured by the contemporary literature as it demands less investigation and is therefore more readily accessible. There was a growing trend in articles designed to customize scoring systems for individual cancer pathological subtypes. The SINS is increasing awareness of spinal instability in this cohort and shows good reliability amongst surgeons and oncologists. Conclusion: The increased understanding of cancer pathology and subsequent development of customized treatments has led to prolonged survival. This not only impacts surgical candidacy on the basis of prognosis but also provides prolonged opportunity for the development of spinal instability. In order for contemporary scoring systems to remain relevant ongoing review, development and revalidation is mandatory. Disclosures: Author 1: none; Author 2: consultant: 3D4Medical; Author 3: none. Table 1 Selection criteria for scoring systems in vertebral metastases

S363 Methods: Totally 42 consecutive aggressive VH cases were diagnosed and treated in our department between 2001 and 2015. Trocar biopsy is indicated in suspected malignant cases. Radiotherapy was usually our first choice, if the neurological deficit was mild or developed slowly. Surgery was indicated, if the neurological deficit was severe or developed fast, or the radiotherapy was not effective. Pathologic finding from operative specimen confirmed the diagnosis. Results: This series included 15 males and 27 females, and the mean age at diagnosis was 46.6 (range 10–69) years old. 34 cases had myelopathy, 7 patients had radiculopathy and one had caudal equina syndrome. 20 patients had decompression and 22 had decompression with vertebroplasty. In the 20 patients with decompression without vertebroplasty, the average estimated blood loss was 1768.0 (range 260–4000) ml and the average surgical time was 276.0 (range 180–450) min; in 22 patients with decompression with vertebroplasty, the blood loss was 1081.8 ml (range 300–3000) ml (P \ 0.05) and the surgical time was 253.6 (range 150–360). All these case had an average 50.1 (range 12–144) months follow-up. There was no recurrence in those cases with radiotherapy and/or vertebroplasty, while two had local recurrence in those cases treated by decompression alone, which was successfully treated repeated decompression and radiotherapy. Conclusions: In aggressive vertebral hemangiomas, epidural soft tissue compression was usually the main reason for neurological deficit. Biopsy is indicated in cases with untypical image findings. In cases with rapid progressive and/or severe myelopathy, decompression could be combined with intraoperative vertebroplasty to reduce blood loss. Disclosures: Author 1: none; Author 2: none; Author 3: none; Author 5: none; Author 6: none.

Clinically utility/user friendliness Contributes to decision re: surgical candidacy Validation in the literature Examined by literature/updated since 2006 Robust derivation methods and data

P61 THE CLINICAL FEATURES AND TREATMENT OUTCOMES OF LANGERHANS’ CELL HISTIOCYTOSIS OF SPINE Liang Jiang1, XiangYu Xu1, Song Bo Han2, ShaoMin Yang3, XiaoGuang Liu1, ZhongJun Liu1

P60 DECOMPRESSION AND INTRAOPERATIVE VERTEBROPLASTY FOR ENNEKING STAGE 3 AGGRESSIVE VERTEBRAL HEMANGIOMAS WITH NEUROLOGICAL DEFICIT Liang Jiang1, Ben Wang1, Xiao Guang Liu1, Shao Min Yang2, Na Meng3, Zhong Jun Liu1 1 Department of Orthopedics, Peking University Third Hospital, Beijing, China; 2Department of Pathology, Peking University Health Center, Beijing, China; 3Department of Radiotherapy, Peking University Third Hospital, Beijing, China

Background: Vertebral hemangiomas (VHs) are called benign tumor, but actually just vascular malformation. Most of them are latent (Enneking S1), and only 1% expended into the epidural space with neurological deficit (Enneking S3). The treatments for aggressive VHs are still controversial for their rarity. Different treatments have been reported to provide good clinical results, including radiotherapy alone, vertebroplasty, direct alcohol injection, and surgical decompression with/without radiotherapy, vertebrectomy and total en bloc spondylectomy. But intraoperative excessive bleeding might be life threatening, and we applied intraoperative vertebroplasty to minimize the blood loss. To retrospectively evaluate the safety and efficiency of this present treatment strategy.

1 Department of Orthopedics, Peking University Third Hospital, Beijing, China; 2Department of Radiology, Peking University Third Hospital, Beijing, China; 3Department of Pathology, Peking University Health Center, Beijing, China

Background: Langerhans cell histiocytosis (LCH) of the spine is a relatively rare condition with unknown etiology. The diagnosis and treatment protocols for spine LCH remain controversial. In this study, we evaluated the efficacy and safety of our proposed diagnosis and treatment protocol introduced in 2009. Methods: We retrospectively reviewed 110 patients with spine LCH who had been diagnosed and treated in our hospital from October 1997 to November 2015. All cases were analyzed in terms of age, gender, clinical and radiologic presentation, therapy, and outcome. The indications for computed tomography (CT)-guided biopsy and surgery for spine LCH have become more stringent since 2009. In cases of a solitary spinal lesion, immobilization and/or observation were first suggested. Chemotherapy was suggested for cases with multifocal LCH lesions, and low-dosage radiotherapy was restricted to recurrent solitary lesion. Result: This series included 69 male and 41 female patients (age range 1–52 years). Pain was the most common symptom (93.6%, 103/110). Pathologic diagnosis was achieved in 72 cases (65.5%). Ninety-eight cases (89.1%) were followed up for an average of 63.8 months (range 12–159). Approximately 77.8% (42/54) cases had CT-guided biopsy before 2009 and 73.2% (41/56) cases had biopsy

123

S364 after 2009 (P = 0.57). Immobilization and/or observation were performed in 25.9% (14/54) and 75.0% (42/56) cases before and after 2009, respectively (P \ 0.001). Approximately 35.2% (19/54) and 10.7% (6/56) cases had surgery (P = 0.002) before and after 2009, respectively. During the follow-up, no significant difference was found in the outcomes between the two groups treated before and after 2009 (P = 0.58). Conclusion: Biopsy is not mandatory for typical spine lesions of LCH. Given the self-healing tendency of spine LCH, immobilization and/or observation remain the first-choice treatments for LCH lesions. Surgery should be indicated for patients with muscle strength grade less than 3/5, rapid deterioration of neurological function within 2 weeks, persistent neurological symptoms over 2 months, severe deformity, and/or segmental instability. Disclosures: Author 1: none; Author 2: none; Author 4: none; Author 5: none; Author 6: none.

P62 ASYMPTOMATIC CONSTRUCT FAILURE IN METASTATIC SPINE TUMOUR SURGERY Naresh Kumar, Dhiraj Sonawane, Aye Sandar Zaw, Aravind Kumar National University Health System, Singapore, Singapore Introduction: The patients with spinal metastases need surgical instrumentation due to pain, instability and/or neurological symptoms. The construct usually comprises of pedicle screw and rod systems with or without anterior cage allograft/cement replacement or augmentation. Fusions are not performed in MSD due to poor patient survival, increased morbidity, poor host graft bed compounding detrimental effects from adjuvant RT and CT. These results in construct failure which could either be symptomatic or asymptomatic. Methods: We retrospectively analysed a total of 288 patients who underwent surgery for MSD in a tertiary referral institution between 2005 and 2015. Clinical data was collected from electronic medical records. Radiological data included SINS, type of lesion and levels of fixation. Surgical data included MIS/open surgery, levels and type of stabilization. FOF was defined as breakage of the rod/screw, screw back out/cut out/loosening, increase in angular deformity, reduction in anterior column height, tilting/subsidence of cage. The endpoint was last follow-up/death. Construct length was categorised as short (\6), intermediate (6–9) or long ([9). We excluded patients with previous spine surgery, infection of implant leading to failure, survival \30 days post-op and/or age \18 years. Results: Out of 288 patients, 264 were included in the final analysis. Lung (23.2%) was the most common primary followed by breast (16.7%). The average follow-up was 11 months (1–92 months). A total of 53 patients showed construct failure out of which, 35 were asymptomatic i.e. having no pain, no worsening of neurology despite construct failure. The most common failure was asymptomatic change in angular deformity 86.5%, screw ploughing 64%, back out 5%, cut out 5%, cage tilt 2.8%. All FOF cases were recorded in the SIN score range of 7–14, 54% of the patient with FOF had SIN score [9, while no failure was noticed in patients with SIN score \7. Conclusions: We conclude that most common FOF cases showed increase in angular deformity and reduction in anterior column height. All these patients are asymptomatic and do not require revision though they may indicate early sign of the failure of construct.

123

Eur Spine J (2017) 26 (Suppl 2):S335–S405 Disclosures: Author 1: none; Author 2: none; Author 3: none; Author 4: none.

P63 DOES PERIOPERATIVE BLOOD TRANSFUSION INFLUENCE POSTOPERATIVE COMPLICATIONS IN METASTATIC SPINE TUMOUR SURGERY Naresh Kumar, Aye Sandar Zaw, Shashidhar Kantharajanna, Barry Tan National University Health System, Singapore, Singapore Background context: There have been numerous advances in surgical techniques for spinal metastases. Despite the advances, blood loss still remains a significant problem in surgery for spinal metastases. The substantial blood loss is presently addressed by transfusion of allogeneic blood products. The effect of allogeneic blood transfusion (ABT) in surgical oncology has been most studied and debated in colorectal and hepatobiliary oncological surgeries. This literature has raised the controversy regarding the adverse effects of ABT influencing survival and/or postoperative complications. Previous studies pertaining to surgery in spinal metastases investigated the influence of ABT on survival, but not on postoperative complications. We aimed to evaluate the influence of perioperative ABT on postoperative complications and infections in patients undergoing metastatic spine tumour surgery (MSTS). Methods: This is a retrospective cohort study including 247 patients who underwent MSTS in a tertiary university hospital between 2005 and 2014. Variables recorded from electronic records and case notes included patient demographics, tumour characteristics, operative details and postoperative complications. The primary outcome measures were postoperative complications including cardiovascular, neurological, respiratory, renal, infection, haematological and electrolyte imbalance within 30 days after MSTS. Postoperative infections which included surgical site infection, systemic sepsis, chest infection and urinary tract infection were also subsequently analyzed. Multivariate logistic regression analyses were performed to assess influence of blood transfusion on postoperative complications and infections. Further adjustment for potential confounders was made in these analyses. There are no funding source and any conflict of interest to be declared. Results: Of 247 patients, 133 patients (54%) received ABT. The overall median unit of blood transfused was 2 units (range 0–10 units). The mean age at the time of surgery was 60 years (range 25–87 years). The adjusted odds of developing any postoperative complication was 2.27 times higher in patients with transfusion (95% CI 1.17–4.38, P = 0.01) and 1.24 times higher odds per every unit increase in blood transfusion (95% CI 1.05–1.46, P \ 0.01). Exposure to blood transfusion also increase the odds of having overall postoperative infections (OR 3.58, 95% CI 1.15–11.11, P = 0.02) and there were 1.24 times higher odds per every unit increase in transfusion (95% CI 1.01–1.54, P = 0.04). Conclusions: The current study adds evidence to the literature implicating ABT to be influential on postoperative complications and infections in patients undergoing MSTS. Appropriate blood management measures should, therefore, be given a crucial place in the care of these patients so as to reduce any putative effect of blood transfusion. Disclosures: Author 1: none; Author 2: none; Author 3: none; Author 4: none.

Eur Spine J (2017) 26 (Suppl 2):S335–S405

S365

P64 INCIDENCE AND PROGNOSIS OF SPINAL METASTASES AS THE FIRST MANIFESTATION OF MALIGNANT TUMORS: ANALYSIS OF 338 PATIENTS UNDERGOING THE SURGICAL TREATMENT Chong-Suh Lee, Se-Jun Park, Sung-Soo Chung, Kyung-Joon Lee, Chan-Duk Park, Dong-Uk Kim, Jin-Sung Park Department of Orthopedic Surgery, Samsung Medical Center, Seoul, Korea Background: Most studies have included the prognosis of spinal metastases from primary malignancy. There are few reports studying spine metastasis as the initial manifestation of malignancy. So, the aim of this study is to evaluate the incidence and prognosis of inaugural spine metastasis in patients who were diagnosed with spinal metastases and underwent the surgical treatment Methods: From January 1994 to December 2014, we reviewed retrospectively 338 patients who underwent surgical treatment for metastatic spinal disease. Through the patients’ electronic medical records, enrolled patients were divided into two groups. Group A is that patients who had no history of malignancy and were diagnosed with primary malignancy site through spinal metastases. On the other hand, group B is that patients who had history of treatment for primary origin malignancy of spine metastasis and develop spine metastasis. The incidence of group A according to primary malignancy was examined and the authors evaluated whether there is a difference between the surgical treatment prognosis of the group A and group B or not. Result: Among the enrolled 338 patients, 209 patients were male and 129 patients were female. The patients’ mean age at the surgical treatment was 57.3 years. During the follow-up period, 264 patients were died and 74 patients were alive. Mean survival was 11.5 months (range 0.6–105.4 months) after surgical treatment. The incidence of group A was 94 patients (27.7%). According to primary malignancy site, lung cancer was in 35/103 patients (34.0%), liver cancer in 8/45 patients (17.8%), kidney cancer in 10/33 patients (30.3%), colorectal cancer in 3/29 patients (10.3%), breast cancer in 3/22 patients (13.6%), prostate cancer in 3/10 patients (30%), thyroid cancer in 4/8 patients (50%) and others carcinoma were in 28/88 patients (31.8%). Among the others, non-Hodgkin’s lymphomas were the highest ratio of group A in 6/6 patients (100%). In Kaplan–Meier survival analysis, the mean survival was 23.3 months and median survival was 11.6 in group A. And the mean survival was 11.6 months and median survival was 6.9 in group B. Group A showed significant longer survival than group B (p = 0.012). According to primary malignancy site, only lung cancer with highest prevalence showed significant difference. Conclusion: Among the enrolled 338 patients who underwent surgical treatment for spine metastases, the incidence of spine metastases as the initial manifestation of primary malignancy was 94 patients (27.7%). And spinal metastases as the first manifestation of malignancy were good prognostic factor when considering the surgical treatment. Disclosures: Author 1: none; Author 2: none; Author 3: none; Author 4: none; Author 5: none; Author 6: none; Author 7: none.

P65 AGGRESSIVE VERTEBRAL HEMANGIOMAS. REGIONARY BLOODFLOW FEATURES Sergey Mizyurov, Sergey Likhachev, Vladimir Zaretskov, Vladislav Arsenievich, Alexey Norkin, Sergey Stepukhovich, Igor Norkin Research Institute of Traumatology, Orthopedics and Neurosurgery of Saratov State Medical University of Razumovsky, Saratov, Russia Objectives: Arterial steal syndrome might be the cause of neurological deficit in vertebral hemangioma (VH) patients. Ultra-sound localization of segmentary blood vessels for arterial bloodflow disorders assessment in spinal cord is a method of spinal bloodflow analysis in VH. The objective of the present research is to study aggressive hemangiomas influence on the reginary bloodflow. Methodology: USDG of segmentary vessels bloodflow at the damage level has been done to 26 VH patients. The diameter of intercostal and lumbar arteries and veins, peak and terminal systolic speed, mean bloodflow speed, systolic-diastolic ratio, pulsatility and resistance indexes and maximum venous bloodflow speed were the parameters of measurement in the research. Volumetric bloodflow was calculated automatically. USDG of segmentary vessels was obtained from 30 healthy volunteers as well. We used Microsoft Office Excel 7.0 and STATISTICA 6.1 (StatSoft) for statistical data analysis. Results: There were 26 patients with aggressive vertebral hemangiomas. 5 patients (19.2%) showed lower paraparesis, at the same time there was no compression of medullary canal contents found and USDG revealed decreased volumetric bloodflow in the arteries at the damage level. 21 patients (80.8%) without neurologic deficit did not have any significant changes in volumetric bloodflow in segmentary

123

S366 arteries. There was increased maximum venous bloodflow in the veins at the damage level in all 26 patients. All VH patients showed signs of venous bloodflow arterialization at the hemangioma level. By-pass shunt from the arterial system to venous one resulted in pseudo pulsation of venous bloodflow. All patients including five with lower paraparesis underwent transcutaneous vertebroplasty. 1 day postsurgically there was moderate increase in volumetric bloodflow of segmentary arteries at the damage level in neurological deficit patients and 1 month postsurgically the same patients showed volumetric bloodflow increasing to normal values. Venous bloodflow pseudo pulsation has vanished in all 26 patients after the operation. Conclusion: Vertebral hemangiomas can cause arterial steal syndrome in spinal cord. Tumour filling with cement eliminates arteriovenous shunt. Disclosures: Author 1: none; Author 2: none; Author 3: none; Author 4: none; Author 5: none; Author 6: none; Author 7: none.

P66 INTRODUCTION OF PROGNOSTIC SCORING FOR PATIENTS WITH SPINAL METASTASES: THE NORTHERN IRELAND EXPERIENCE Kathryn Rooney, Stacey Thompson, Nagy Darwish Fracture Department, Royal Victoria Hospital, Belfast, Ireland Aim: To establish if a prognostic scoring system could be introduced when assessing patients presenting with metastatic spinal cord compression to aid decision making with regards to surgical management. Method: A 3-month retrospective review of referrals with metastatic spine lesions was undertaken in the Royal Victoria Hospital, Belfast. Using the modified Tokuhasi score for prognosis each referral was scored. The scoring system is used to predict prognosis and therefore can be correlated with management options. \6 months—conservative; 6 months–1 year palliative surgery; [1 year excisional surgery. The score was then compared to management that patient received to determine if current practice reflected that advised by prognostic scoring. Results: There were 32 referrals for patients with metastatic spinal cord compression between October 2016 and December 2016. 3 patients had incomplete data and were therefore excluded. 66% of our patients underwent the appropriate treatment for their prognosis as per the scoring system. 13 patients scored \8 and were categorised for conservative management; in our study 3 actually proceeded to surgery. Of those categorised for palliative surgery 50% underwent posterior stabilisation and decompression while the other seven were treated conservatively. Only 2 patients scored[12 and both patients proceeded to palliative surgery rather than excisional due to other patient factors. Conclusions: There can be difficulties when assessing patients who fall into the palliative group. These are a complex group of patients requiring both a multidisciplinary and holistic approach to decision making. When assessing a patient for treatment of their metastatic spinal tumour, awareness of the prognosis is extremely important in therapy selection. We advocate the use of oncological scoring systems as an adjunct when deciding upon a management plan for patients with spinal cord compression. Further follow-up of patients stratified to palliative surgery should be carried out to establish the benefit of current management and allow further refinement of the scoring system. Disclosures: Author 1: none; Author 2: none; Author 3: none.

123

Eur Spine J (2017) 26 (Suppl 2):S335–S405 P67 STEREOTACTIC BODY RADIOTHERAPY FOLLOWED BY SURGERY WITHIN 24 HOURS FOR THE TREATMENT OF UNSTABLE SPINAL METASTASES; A SAFETY AND FEASIBILITY STUDY Anne Versteeg, Joanne van der Velden, Wietse Eppinga, Nicolien Kasperts, Sophie Gerlich, Jochem Hes, Cumhur Oner, Jorrit-Jan Verlaan Department of Orthopaedic Surgery, University Medical Center Utrecht, The Netherlands Introduction: The standard treatment of unstable spinal metastases consists of surgical stabilization followed by conventional radiotherapy at a minimum of 2 weeks later to allow for wound healing. This time interval however delays the time before radiotherapy-induced pain relief and local tumour control can be achieved. Furthermore, up to 40% of the patients experience incomplete pain relief and most conventional radiotherapy fractionation schemes require multiple hospital visits. An alternative treatment strategy with a better and earlier pain response requiring less hospital visits is desirable. Stereotactic body radiotherapy (SBRT) allows for a higher and more accurately delivered radiation dose while actively sparing the surgical area to prevent wound complications. Hence, the time interval between the two treatments may be minimized. The aim of this study is therefore to assess the safety and feasibility of combining high dose single fraction SBRT followed by surgical stabilization within a 24-h window for the treatment of painful unstable spinal metastases. Patients and methods: A phase I/IIa study was conducted according to the IDEAL criteria for the evaluation of new complex interventions. Patients with painful unstable spinal metastases from a solid tumor requiring surgical intervention, a Karnofsky performance status C50 and an ASIA E or D without progressive neurological deficits were eligible for inclusion. Information on demographic characteristics, treatment, toxicity (according to the CTCAE 4.0), adverse events and survival were systematically collected. The main adverse event of concern was the occurrence of wound complications. First, SBRT treatment was simulated on a planning CT and MRI to deliver a dose of 18 Gy to the spinal metastasis and 8 Gy to the rest of the vertebral body while actively minimizing the radiation dose in the posterior surgical area. After SBRT treatment was delivered, surgical stabilization followed within 24 h according to routine practice. Results: A total of 13 patients, 7 females and 6 males, were included in the study. The most common primary tumor was breast cancer followed by renal cell and lung cancer. The combination of SBRT and surgery was completed for all patients within 24 h. A posterior approach was used for all surgical procedures, with 2 conventional open and 11 percutaneous procedures. The mean operating time was 80 min (range 46–162) with a median blood loss of 50 ml (range 50–300). The median length of hospital stay was 5 days. None of the patients experienced a wound complication. One patient suffered a grade 3 complication (cement leakage leading to nerve root compression and irritation), which was deemed not related to the combination of the two procedures but a known surgical complication. Conclusion: The combination of single fraction SBRT and surgical stabilization within 24 h is safe and feasible for the treatment of unstable spinal metastases. Disclosures: Author 1: grants/research support: AOSpine Start Up Grant, consultant: AOSpine International; Author 2: none; Author 3: none; Author 4: none; Author 6: none; Author 7: none; Author 9: grants/research support: DePuy-Synthes; Author 10: none.

Eur Spine J (2017) 26 (Suppl 2):S335–S405

S367 Disclosures: Author 1: none; Author 2: none; Author 3: none; Author 4: none; Author 5: none.

INFECTION, THORACOLUMBAR SPINE

P69 PERCUTANEOUS SUCTION AND IRRIGATION FOR THE TREATMENT OF RECALCITRANT SPONDYLODISCITIS

P68 METASTATIC SPINE TUMOUR SURGERY: DOES PERIOPERATIVE BLOOD TRANSFUSION INFLUENCE THE SURVIVAL AND CANCER PROGRESSION? Aye Sandar Zaw, Shashidhar Kantharajanna, Barry Tan, Balamurugan Vellayappan, Naresh Kumar National University Health System, Singapore, Singapore Introduction: Allogeneic blood transfusion (ABT) has been the main replenishment method for lost blood in patients metastatic spine tumour surgery (MSTS). However, the impact of ABT on cancer related outcomes has been controversial in various studies. Till date, there have been no studies looking at the association between perioperative blood transfusion and disease progression in MSTS. We aimed to evaluate the influence of perioperative ABT on disease progression and survival in patients undergoing MSTS. Methods: We did a retrospective analysis of 247 patients who underwent MSTS at a single tertiary institution between 2005 and 2014. The impact of the use of perioperative ABT (either exposure to or quantities of transfusion) on disease progression and overall survival was assessed using Cox regression analyses while adjusting for potential confounding variables. Results: Of 247 patients, 133 patients (54%) received ABT. The overall median unit of blood transfused was 2 units (range 0–10 units). Neither blood transfusion exposure nor quantities of transfusion were found to be associated with overall survival [hazard ratio (HR) 1.15, P = 0.35] and (HR 1.10, P = 0.11) and progression-free survival (HR 0.87, P = 0.18) and (HR 0.98, P = 0.11), respectively. The factors influencing overall survival were primary tumour type and preoperative ECOG while primary tumour type was the only factor having impact on progression-free survival. Conclusions: This is the first study providing evidence that disease progression and survival in patients undergoing MSTS are less likely to be influenced by perioperative ABT. The worse oncological outcomes are more likely to be caused by the clinical circumstances necessitating blood transfusion but not transfusion itself. However, given that ABT can have propensity towards developing postoperative infections including surgical site infection, utilization of patient blood management interventions would be worthwhile rather than relying solely on allogeneic blood transfusions for these patients, if and whenever possible.

William Griffiths-Jones, Luigi Aurelio Nasto, Oliver M Stokes, Hossein Mehdian The Centre for Spinal Studies and Surgery, Nottingham, UK Background/introduction: The primary management of pyogenic spondylodiscitis is conservative. Once the causative organism has been identified, by blood culture or biopsy, administration of appropriate intravenous (IV) antibiotics is started. Occasionally patients do not respond to antibiotics and surgical irrigation and debridement is needed. The treatment of these cases is challenging and controversial. Furthermore, many affected patients have significant comorbidities often precluding more extensive surgical intervention. Hence the need for less invasive yet effective techniques for treatment of spinal infections. Purpose of the study: To report on the feasibility and safety of a novel percutaneous technique for irrigation and debridement of the intervertebral disc space in patients with pyogenic spondylodiscitis. Materials and methods: A series of ten consecutive patients diagnosed with pyogenic spondylodiscitis received percutaneous disc irrigation and debridement. The procedure was performed inserting two Jamshidi needles percutaneously into the disc space. Needles were inserted with aid of intraoperative image intensifier. Indications for surgery were: poor response to antibiotic therapy (eight patients), and need for more extensive biopsy (two patients). Pre- and post-operative white blood cell count (WBC), C-reactive protein (CRP), erythrocyte sedimentation rate (ESR), Oswestry Disability Index (ODI), and Visual Analogue Score (VAS) for back pain were collected. Minimum follow-up was 18 months, with regular interval assessments. Results: There were seven males and three females with mean age of 67 years. The mean WBC before surgery was 14.63 9 109/L (10.9–26.4) and dropped to 7.48 9 109/L (5.6–9.8) after surgery. Mean preoperative CRP was 188 mg/L (111–250) and decreased to 13.83 mg/L (5–21) after surgery. Similar improvements were seen with ESR. All patients reported significant improvements in ODI and VAS scores after surgery. The average hospital stay after surgery was 8.17 days. All patients had resolution of the infection, and there were no complications associated with the procedure. Conclusions: Our study confirms the feasibility and safety of our percutaneous technique for irrigation and debridement of pyogenic spondylodiscitis. Percutaneous irrigation and suction offers a truly minimally invasive option for managing recalcitrant spondylodiscitis or for diagnostic purposes. The approach used is very similar to discography and can be easily adapted to different hospital settings. Disclosures: Author 1: none; Author 2: none; Author 3: none; Author 4: none.

123

S368 P70 A CLINICAL STUDY OF INTERNAL FIXATION, DEBRIDEMENT AND INTERBODY THORACIC FUSION TO TREAT LONG SEGMENTAL THORACIC TUBERCULOSIS VIA POSTERIOR APPROACH: MINIMUM 3 YEAR FOLLOW UP Xiongke Hu, Qinpeng Zhao, Dingjun Hao Spinal Surgery, Honghui Hospital, Xi’an, China Purpose: We evaluated the clinical efficacy and feasibility of onestage posterior internal fixation, debridement and interbody thoracic fusion in the treatment of long segmental thoracic tuberculosis. Methods: Seventy-six patients with long segmental thoracic tuberculosis were studied retrospectively: 35 men and 41 women with an average age of 38.5 years. Operating time, blood loss, time in bed, complications, neurological function, rate of deformity correction and rate of interbody fusion were investigated. Results: All cases were followed up for 42.6 months on average. Average mean operating time was 270 min, evaluated blood loss during operation 840 ml, rate of kyphosis correction 85%, corrected kyphosis angle 28 and loss of corrected angle 3.2. Patients whose neurological function improved accounted for 92.1%. Erythrocyte sedimentation rate (ESR) or C-reactive protein (CRP) decreased to normal levels three months after operation. Bone fusion was achieved within 3–6 months (average 5.5 months). No severe complications or spinal cord injury occurred. Conclusions: Posterior approach can successfully remove the focus of tuberculosis with complete interbody thoracic fusion after operation, which restores spinal stability Keywords: Thoracic vertebrae; Spinal tuberculosis; Posterior; Long segmental; Internal fixation. Disclosures: Author 1: none; Author 2: none; Author 3: none.

Eur Spine J (2017) 26 (Suppl 2):S335–S405 quality of life, sagittal deformity, and complications after percutaneous instrumentation. Materiel and methods: 28 patients (average age 67.8 years) were prospectively analyzed preoperatively and postoperatively at day 5, 6 weeks, 3 and 12 months. Clinical results were evaluated by VAS back pain and EQ-5D. Radiographic deformity was assessed using the sagittal index. Fusion was evaluated on CT at 12 months. Duration of hospitalization, evidenced bacteria, length of antibiotic treatment, course of CRP, and complications were monitored. A Bayesian model was used for statistical evaluation: a probability [0.95 indicated a significant change. Results: The average VAS was 7.04 preoperatively, 3.15 at day 5, 2.17 at 6 weeks, 1.86 at 3 months, 1.64 at 12 months. Probabilities of decrease were [0.95 within the first 6 weeks. The average EQ-5D was 0.229 preoperatively, 0.563 at 6 weeks, 0.687 at 3 months, 0.755 at 12 months. Probabilities of improvement were [0.95 at each time point within 12 months. The average sagittal index was 15.1 preoperatively, 9.6 at day 5 (probability of increase = 1.000), and 11.0 at 1 months (probability of decrease = 0.0037). CT at 12 months evidenced inter-body fusion in 59.1%, partial fusion in 18.2% and pseudarthrosis in 22.7%. Radiographic results did not predict clinical results. The average hospital stay was 4.3 ± 2.1 days. Antibiotic treatment was stopped at 6 weeks in 23 patients (CRP\5 mg/l) and at 3 months in 5 patients. Complications were: two pulmonary, one pedicle screw migration, one hematoma. Septic complications did not occur. Conclusion: Percutaneous instrumentation is safe and efficient in the treatment of spondylodiscitis. It improves back pain and quality of life significantly in the early postoperative period and prevents kyphotic deformity. Minimal invasive instrumentation does not lead to septic complications and the course of antibiotic treatment is not influenced. Comparative studies with brace treatment are pending. Disclosures: Author 1: none; Author 2: grants/research support: Stryker, Clariance, consultant: Stryker, Medtronic, LDR Me´dical, Ceraver, royalties: Clariance; Author 3: consultant: Clariance, Stryker; Author 4: none; Author 5: none; Author 6: none; Author 7: none; Author 8: grants/research support: Clariance, Stryker, LDR, Integra, Medtronic, consultant: Clariance.

P72 LUMBOSACRAL DESTRUCTION DUE TO SPINAL TUBERCULOSIS: EVALUATION OF SURGICAL FIXATION, HEALTH-RELATED-QUALITY-OF-LIFE OUTCOME, AND SPINO-PELVIC PARAMETERS Didik Librianto, Ifran Saleh, Yoshi Djaja P71 SAFETY AND EFFICACY OF PERCUTANEOUS INSTRUMENTATION AND ANTIBIOTIC TREATMENT PYOGENIC SPONDYLODISCITIS Arnauld Lambert, Yann Philippe Charles, Sebastien Schuller, Axel Walter, Nicolas Lefebvre, Yves Hansmann, erik Sauleau, Jean-Paul Steib CHU Strasbourg, Strasbourg, France Introduction: Pyogenic spondylitis is usually treated by antibiotics in combination with a brace in patients presenting back pain and a risk for vertebral body collapse. Recent retrospective data suggests that patient’s autonomy might improve rapidly after a percutaneous instrumentation. The purpose of this study was to analyze back pain,

123

Department of Orthopaedics and Traumatology, Fatmawati General Hospital, Universitas Indonesia, Jakarta, Indonesia Introduction: Lumbosacral spine tuberculosis is a rare entity that represents only 2–3% of all spinal tuberculosis cases. Management of lumbosacral tuberculosis is always a surgical challenge, due to complex anatomy and biomechanics of the lumbosacral and pelvic region. Eradication of the tuberculosis lesion, restoration of the lumbar lordosis, spino-pelvic balance and sound stabilization are the aim of each surgery for these cases. This study was aimed to evaluate the outcome of surgical outcome of lumbosacral tuberculosis. Methods: This study retrospectively reviewed 52 patients with lumbosacral tuberculosis (L4–S2) with vertebrae destruction who were treated surgically in our institution between January 2011 and January 2016. The surgical procedures were mainly posterior debridement and instrumentation by various distal fixation sites

Eur Spine J (2017) 26 (Suppl 2):S335–S405 (sacral, iliac or both) with or without anterior debridement and interbody-fusion. Outcome data included were neurologic status, functional outcome measured by HRQoL form and SF-36 questionnaire, lumbosacral angle, spino-pelvic parameters (pelvic tilt, sacral slope) and implant failure/loosening. Results: Fourth lumbar vertebra was the commonest involved segment in our series. Neurologic status improvement was found in almost all cases regardless the initial lumbosacral angle and type of treatment. Functional outcome improvement was also found increased in proportion to the neurological improvement. Involvement of the S1 segment was the main indicator for spino-pelvic imbalance. The risk of revision surgery due to any cause was 10% in our series, in which the quality of construct in the damaged bone was the main cause of revision. Conclusion: Regardless the type of fixation, improvement of functional outcome and neurological state in lumbosacral tuberculosis following an adequate debridement and anti-tuberculosis drug is expected. In determining the type of construct, the quality of the anchoring bone should be considered to prevent failure/loosening of the implants. Disclosures: Author 1: none; Author 2: none; Author 3: none.

S369 PCS score significantly (p \ 0.05) (Fig. 1A), with an odds ratio of 1.017(unit by unit) of improving SF-36 PCS score on multivariate analysis (p \ 0.05) (Fig. 1B). The breaking point in age for this effect was calculated to be 37.5 years (AUC = 58.0, p = 0.05). On the other hand, gender was found not to have a significant effect on any of the HRQOL scores. Discussion–Conclusion: This study demonstrates that patient age may have a positive effect on treatment outcome parameters in surgically treated patients with ASD and the breaking point of this effect may be earlier than generally anticipated. Gender on the other hand, does not seem to affect results. This information may be important in patient counseling for the anticipated outcomes of surgery. Disclosures: Author 1: grants/research support: Depuy Synthes, Medtronic, stock/shareholder: IncredX, royalties: Zimmer Biomet, AOSpine; Author 2: none; Author 3: none; Author 4: none; Author 5: grants/research support: DEPUY SYNTHES, consultant: DEPUY SYNTHES, MEDTRONIC, royalties: ALPHATEC, SPINEART, CLARIANCE; Author 6: grants/research support: DEPUY SINTHES, consultant: K2M. DEPUY SINTHES; Author 7: grants/research support: Depuy-Synthes; Author 9: grants/research support: DePuy Synthes Spine, Medtronic, consultant: DePuy Synthes Spine, Biomet; Author 10: grants/research support: DePuy-synthes.

ADULT DEFORMITY

P73 THE INFLUENCE OF AGE AND GENDER ON TREATMENT OUTCOMES IN SURGICALLY TREATED PATIENTS WITH ADULT SPINAL DEFORMITY Emre Acaroglu, Selim Ayhan, Selcen Yukse, Vugar Nabiyev, Ibrahim Obeid, Francisco Javier Sanchez Perez-Grueso, Frank Kleinstuck, Ahmet Alanay, Ferran Pellise, European Spine Study Group ESSG 1

ARTES Spine Center, Ankara, Turkey; 2Department of Biostatistics, Yildirim Beyazit University, Ankara, Turkey; 3Spine Unit, Bordeaux University Hospital, Bordeaux, France, 4Spine Unit, Hospital Universitari La Paz, Madrid, Spain; 5Schulthess Klinik, Zu¨rich, Switzerland; 6Comprehensive Spine Center, Acibadem Maslak Hospital, Istanbul, Turkey, 7Spine Unit, Hospital Universitari Vall d’Hebron, Barcelona, Spain Background: Factors affecting the natural history and treatment outcomes are being investigated in order to identify the best and proper clinical approach in ASD. European Spine Study Group’s recent research suggests the critical age for the surgery to become more complex to be at early 30 s. However, the evidence is still insufficient on whether age by itself, as well as gender has any effect on treatment outcomes. Aim: To determine the clinical impact of age and gender on treatment results in surgically treated ASD patients. Patients and methods: Prospectively collected data from a multicentric ASD database was analyzed and all surgical patients with a minimum follow-up of 1 year were included and analyzed for demographic, clinical, radiological and health related quality of life (HRQOL) parameters. Patients were separated into two groups based on improvement in HRQOL parameters by minimum clinically important difference (MCID). Student’s t-test and Chi square test were used to analyze the effect of age and gender on outcome measurements followed by a multivariate binary logistic regression model for these results with statistical significance. Results: A total of 186 patients (157F, 29M) with a mean age of 49.2 ± 19.1 years were analyzed. Age was found to affect only SF-36

P74 ADULT SPINAL DEFORMITY OVER 70 YEARS OF AGE: A MINIMUM TWO-YEARS FOLLOW-UP STUDY Emre Acaroglu, Cem Karabulut, Selim Ayhan, Selcen Yuksel, Francisco Javier Sanchez Perez-Grueso, Frank Kleinstuck, Ibrahim Obeid, Ahmet Alanay, Ferran Pellise, European Spine Study Group ESSG 1

ARTES Spine Center, Ankara, Turkey; 2Department of Biostatistics, Yildirim Beyazit University, Ankara, Turkey; 3Spine Unit, Hospital Universitari La Paz, Madrid, Spain; 4Schulthess Klinik, Zu¨rich, Switzerland; 5Spine Unit, Bordeaux University Hospital, Bordeaux, France; 6Comprehensive Spine Center, Acibadem Maslak Hospital, Istanbul, Turkey; 7Spine Unit, Hospital Universitari Vall d’Hebron, Barcelona, Spain Background: Treatment of adult spinal deformity (ASD) in elderly patients remains controversial. Past research demonstrated favourable results by surgery based on relatively small series. There is virtually no data available on the results of non-surgical treatment as well as the factors involved in the decision of surgical treatment. Aim: To analyse;

123

S370 •

• •

the factors leading to surgery by comparing the baseline characteristics of operative vs non-operative patients and the cut-off values of these (if significant), to evaluate the safety and efficacy of surgery, and, to compare the treatment results of surgical and non-surgical management of patients with ASD over 70 years of age with a follow-up period of 2 years.

Methods: Patients over 70 years of age with a minimum follow-up of 2 years were included and evaluated with demographic, clinical, surgical, radiological data and HRQOL (SF-36 MCS, SF36-PCS, SRS-22 and ODI) parameters. Peri/postoperative complications, classified as major (life threatening or requiring additional surgery) and minor were also investigated for their effects on treatment results and HRQOL parameters. Results: Database had 181 patients of whom 90 (F: 71, M: 29; surgical: 61, non-surgical: 29) had 2-year follow-up. Comparison of surgical vs non-surgical groups at baseline demonstrated significant differences for all HRQoL parameters and major coronal Cobb angle (p \ 0.05). Calculated optimal cut-off values that separates groups for COMI, ODI, SF-36 PCS and SRS-22 were 5.7, 37.0, 37.5 and 3.2, respectively (p \ 0.05 for all). All operative patients (n = 61; F: 46, M: 15) had been treated with posterior surgery and a total of 39 osteotomies and 24 interbody fusions had been performed. Overall, 135 complications in 46 patients (51%) (71 major, 64 minor) and 1 death (1.1%) were observed. Reoperation rate was 62.3% at the end of 2 years. For all HRQOL parameters except SF-36 MCS, surgically treated patients were found to be significantly improved at 2 years even in the presence of complications (p \ 0.05) (Fig. 1). Conclusion: This study on patients with ASD over 70 years of age followed for a minimum of 2 years has demonstrated that; the decision for surgical treatment is based on the level of disability with significant cut-off values (37.5 for ODI); surgical treatment was associated with a high rate of complications and re-operation rate, but despite this, only surgery (compared to non-surgical treatment) provided improvement in HRQOL in elderly. Disclosures: Author 1: grants/research support: Depuy Synthes, Medtronic, stock/shareholder: IncredX, royalties: Zimmer Biomet, AOSpine; Author 2: none; Author 3: none; Author 4: none; Author 5: grants/research support: DEPUY SINTHES, consultant: K2M. DEPUY SINTHES; Author 6: grants/research support: Depuy-Synthes; Author 7: grants/research support: DEPUY SYNTHES, consultant: DEPUY SYNTHES, Medtronic, royalties: alphatec, spineart, clarariance; Author 9: grants/research support: DePuy Synthes Spine, Medtronic, consultant: DePuy Synthes Spine, Biomet; Author 10: grants/research support: DePuy-synthes.

123

Eur Spine J (2017) 26 (Suppl 2):S335–S405 P75 CATASTROPHIC PROXIMAL JUNCTIONAL FAILURE FOLLOWING SURGICAL TREATMENT OF ADULT SPINAL DEFORMITY Joo-Hyun Ahn, Kee-Yong Ha, Young-Hoon Kim, Sang-Il Kim, Hyeong-Youl Park Department of Orthopaedic Surgery, Catholic University of Korea, Seoul, Korea Introduction: Correction surgery for adult spinal deformity (ASD) often need long segmental constructs. However, problems at the proximal junction of long segmental constructs, including proximal junctional kyphosis (PJK) and proximal junction failure (PJF), sometimes develop within a few months after surgery. Among various forms of proximal junctional problems, acute morphologic changes can cause neurologically catastrophic symptoms and are more likely to lead to revision surgery. Catastrophic proximal failures (CPFs) were defined as several forms in previous studies. However, they included heterogeneous forms and seemed to have clinical and radiographic differences from each other. The purpose of this study was to evaluate the prevalence and risk factors of each form of CPF. Materials and methods: From 2004 to 2014, ASD patients who underwent correction surgeries with long segment fixation at out institute were included. We performed clinical data review and radiographic measurements on X-rays at preoperative, immediate postoperative, and every 1-year follow-up visits. CPF was defined as fracture at the upper instrumented vertebra (UIV) or UIV + 1, metallic failure at UIV, or 15 or more PJK. Patients with CPF were compared with non-CPF group. Also, we compared each type of CPF with others. Results: CPF developed in 14 of 116 UIVs at TL junction and 4 of 42 UIVs at thoracic region group. Failure modes were UIV fracture in four patients, UIV + 1 fracture in six patients, soft tissue failure in four patients, and screw pull-out in four patients. Screw pull-out developed within postoperative 3-month and others did later. Two fracture groups showed lower body mass index (BMI) and bone mineral density (BMD). Soft tissue failure showed much higher thoracic kyphosis and proximal junctional angle on preoperative films. Fixation failure group had marked correction of lumbar lordosis. Five of 18 patients underwent revision surgery due to pain and neurologic deficits. After revision, 1 patients did not recover from neurologic compromise. Conclusions: CPF was identified in 11.4% of ASD patients undergoing long segmental fixation surgery. Among four modes of failures, screw pull-out developed earlier and UIV + 1 fracture did much later. Osteoporosis and sagittal malalignment were marked preoperatively in UIV fracture group. Soft tissue failure seemed to be frequent when UIV was at the kyphotic region. Fixation failure was likely to develop after marked correction of lumbar lordosis. Disclosures: Author 1: none; Author 2: none; Author 3: none; Author 4: none; Author 5: none.

Eur Spine J (2017) 26 (Suppl 2):S335–S405

S371 Image 1: Image 1a shows the screw positioned through the entry point, image 1b is an axial plane reconstruction and image 1c is an oblique plane reconstruction. These images show, that the screw is completely within the iliac bone. Disclosures: Author 1: stock/shareholder: EOS Imaging; Author 2: employee: Institute of Anatomy and Cell Biology, Ludwig-Maximilians-University, Germany; Author 3: none.

P76 THE TEARDROP TECHNIQUE: SAFE AND EASY ILIAC SCREW PLACEMENT Christof Birkenmaier, Axel Unverzagt, Carolin Melcher Department of Orthopedics, Physical Medicine and Rehabilitation, Ludwig-Maximilian-University Munich, Munich, Germany Background: Freehand placement iliac screws can be difficult for novices. Typically it requires dissecting the posterior superior iliac spine (PSIS), the use of connectors and may leave palpable screw heads. Choosing a more medial and inferior entry point eliminates these problems, but makes correct positioning more challenging. Despite its advantages, this technique has found only limited acceptance. Methods: In a human pelvis the medial and lateral cortices of the ilium were marked with pieces of wire and a CT scan was obtained. The pelvis was placed on a carbon operating table and a C-arm was positioned in the same way as in a surgical situation to achieve a proper projection of the iliac teardrop. Correlated photographic and C-arm images were obtained. The CT data was entered into the SpineAssist software. Virtual screws were placed through the defined entry point and along the visual axis. The trajectory was evaluated in all three anatomical planes as well as in a drive-through mode. Results: In the teardrop view, the lateral iliac cortex becomes almost one solid line, leaving a visual safety margin towards the hip joint and the correct entry point is located very close to the PIIS. In the simulation, we placed virtual screws (120 9 8 mm). Their position was examined in the three planes and always exhibited a trajectory that was completely within the ilium (Image 1). It is possible by means of this technique to position a screw very tightly along the lateral and the inferior iliac cortices, thus providing superior anchoring. Conclusions: The outline of the iliac teardrop figure is made up entirely by the iliac cortices, making iliac screw placement using the teardrop technique safe and reproducible. It requires only a C-arm and even novices should be able to safely place iliac screws in this way. Screws placed according to this technique can easily be connected without offset connectors. Supported by this data, the teardrop technique should be taught to residents. Figure legend:

P77 QUALITY OF LIFE AND SAGITTAL ALIGNMENT AFTER SURGICAL CORRECTION OF ADULT SPINAL DEFORMITY WITH POSTERIOR INSTRUMENTATION Evalina Burger, Christopher J. Kleck, Vikas V. Patel, David Calabrese, Andriy Noshchenko, Sean Molloy, Michael S. Chang, Vincent Fiere Department of Orthopedics, University of Colorado, Aurora, USA Introduction: The role of improved sagittal alignment in long term clinical outcomes for correction of ASD is not well established. The evaluation of the relationship between QOL outcomes and long term

123

S372 radiographic correction of spinal imbalance, were the goals for this study. The study confirms that there is a strong clinical treatment effect after surgical correction of adult spine deformity. Sagittal alignment did not change significantly, due to the absence of sever sagittal imbalance pre operatively in the majority of the enrolled patients. Hypothesis: Improvement of QOL measurements has a strong correlation with correction of sagittal imbalance in ADS. Design: Multi-center, prospective cohort study. Study design: Multi-center prospective cohort study. Sponsor: MDEDICREA USA Corp. (New York, NY 10013). Enrolled Medical Facilities: Department of Orthopaedics University of Colorado Denver (USA); Center of Orthopedics Santy (France); Royal National Hospital (UK); Sonoran Spine (USA). Local institutional approval was obtained for each facility. Inclusion Criteria: Cobb angle C30; age C21 years old; indication for four or more operated levels. Exclusion criteria: vertebral column resection; neuromuscular scoliosis; recent trauma; spinal cord abnormalities; metabolic spinal pathology; pathologic obesity; osteomyelitis; pregnancy; and insulin dependent diabetes. Follow-ups: 3, 6, 12, and 24 months. Quality of life assessment: Oswestry Disability Index (ODI) and Scoliosis Research Society Scale (SRS22). Sagittal alignment indices (plain radiography): sagittal vertical axis (SVA); lumbar lordosis (LL); thoracic kyphosis (TK); pelvic incidence–LL mismatch (PI–LL); and pelvic tilt (PT). ANOVA was used for statistical analysis. Results: 54 patients were included (14 male); mean age, 62.1 (SD 11.3), 22% had osteotomy; dropout, 18.5% at 24 month. Preoperative mean ODI, 21.1 (SD 7.5) and SRS22, 2.68 (SD 0.9) were improved to 14.5 (ODI) and 3.64 (SRS22) at 24 months follow up (P \ 0.0001); 95% of patients positively evaluated the treatment outcomes. Preoperative mean values of sagittal alignment were: SVA, 51 (SD 42); LL, 41 (SD 19.1); TK, 45 (SD 15.1); PI–LL, 15 (SD 21.6); and PT, 23 (SD 10.5). Significant changes of these characteristics were not revealed. Conclusion: Significant improvement of the quality of life can be explained by correction of the spinal deformity, stabilization of spine, and effect of decompression. Sagittal alignment did not change significantly due to absence of preoperative imbalance in 50% of patients (SVA \|50|) and high variability of the indices. Disclosures: Author 1: grants/research support: 4. Activ-L Artificial Disc. CT, Aesculap, Inc., Industry, 11/28/2007 to 6/30/2016, $1,448,783, consultant: DSM. Paradigm Spine, Signus, X-Spine, Medicrea, Medacta, Spine Wave; Author 2: grants/research support: DePuySynthes, Pfizer, Globus, Medicrea, OREF, OmEGA, consultant: Medicrea, Medtronic, Medacta; Author 3: grants/research support: Pfizer, Medicrea, Aesculap, Medtronic, Orthofix, Vertiflex, consultant: Stryker, Zimmer, royalties: Aesculap; Author 4: none; Author 5: none; Author 6: grants/research support: Medtronic, consultant: K2M, Zimmer, De Puy, Medtronic; Author 7: consultant: Stryker/Seaspine/DePuy/Spinewave; Author 8: consultant: Medicrea, Clariance, royalties: Medicrea, Clariance.

P78 THE RESULTS OF SPINAL DEFORMITY SURGERY IN PATIENTS WITH PARKINSONS DISEASE Ari Demirel, Kestutis Valancius, Kristian Hoey, Peter Helmig, Haisheng Li, Cody Eric Bu¨nger, Ebbe Stender Hansen Department of Orthopaedics, Aarhus University Hospital, Aarhus, Denmark Summary: Retrospective study of 18 patients with Parkinsons disease (PD) who have undergone extensive spinal deformity surgery. Questionnaires, pre-and postoperative radiographs, and revision surgeries were analysed to understand the success level after surgery.

123

Eur Spine J (2017) 26 (Suppl 2):S335–S405 Treatment acceptance, deformity correction and level of satisfaction were high despite frequent revisions. Introduction: Parkinsons disease (PD) is the second most common neurodegenerative disorder. Truncal dystonia in long standing PD often causes loss of spinal balance. The literature on deformity surgery in patients with PD is remarkably scarce with many reports of complications and failures, and it is still unclear weather patients with spinal deformity due to PD benefit from operative treatment. Purpose of the study: To investigate Parkinsons patients after their spinal deformity surgery with focus on technical issues, functional status and treatment satisfaction. Materials and methods: We investigated all patients with PD and extensive deformity and spinal fusion surgery between 2002 and 2016. Patients answered PDQ-39, SRS-22, EQ-5D, SF-36 questionnaires and the question ‘‘If you had the knowledge that you have today, would you have said yes to your first operation’’. Pre- and postoperative radiographs were analysed by plumb line displacement. Complications and revision surgeries were noted. Results: 18 patients, 13 men and 5 women, were operated upon. Mean age was 71.7 (SD 4.5). Satisfaction score in SRS-22 was 4.07 (SD 0.8). According to questionnaires, the patients had high social and emotional results, while they scored lower in mobility and functional parameters. Sixteen patients answered the question regarding treatment acceptance positively. No patient had chronic worsening of Parkinsons symptoms. Revision surgeries were frequent but declined with gained experience. Revision incidents were caused by: rod breakage (12), screw loosening (7), and wound revision (1). Plumb line displacement and number of revisions per patient are shown in the table. Conclusion: Deformity surgery in PD must be extensive with restoration of sagittal and coronal balance and fusion to the ileum. Instrumental complications and revisions must be expected. The satisfaction level of Parkinson patients operated on for spinal deformity appears high despite frequent instrumental problems and revisions. Prospective studies comparing life quality before and after surgery and health economics are warranted. Disclosures: Author 1: none; Author 2: none; Author 3: none; Author 5: none; Author 6: none; Author 7: none.

Eur Spine J (2017) 26 (Suppl 2):S335–S405

S373

P79 EFFECT OF DEFORMITY SURGERY ON BACK AND LEG PAIN IN PATIENTS WITH ADOLESCENT IDIOPATHIC SCOLIOSIS (AIS)

P80 DEMOGRAPHIC VARIABLES INFLUENCE OPTIMAL SPINOPELVIC PARAMETERS IN ADULT SPINAL DEFORMITY SURGERY

Tamas F Fekete, Anne F Mannion, Daniel Haschtmann, Frank Kleinstueck, Markus Loibl, Dezsoe Jeszenszky

Sleiman Haddad, Susana Nu´n˜ez Pereira, C¸aglar Yilgo¨r, Emre R. Acaroglu, Frank S. Kleinstueck, Francisco Javier Sanchez PerezGrueso, Ibrahim Obeid, Ahmet Alanay, Ferran Pellise´, ESSG European Spine Study Group

Schulthess Clinic, Zu¨rich, Switzerland Background: Surgery for adult idiopathic scoliosis (AIS) aims to prevent curve progression but in some patients it also relieves pain. However, the association between back pain and AIS is controversial. Our clinical experience is that a proportion of AIS patients, especially young adults, have also relevant back pain. Whether this is related to their deformity and, hence, whether deformity surgery is associated with a relevant reduction in their pain is unclear. The influence of age at surgery on back pain relief also remains to be investigated. Our study aims to investigate the pain relief after surgery in patients with AIS in young adults (19–30 years) and in adolescents (12–18 years). Materials and methods: A retrospective analysis of prospectively collected data from patients aged 12–30 years, operated for AIS in our hospital from 2005 to 2014 and registered in our local patient outcomes database linked to EUROSPSINE’s Spine Tango Registry. Preoperatively and up to 2 years’ postoperatively, patients completed the Core Outcome Measures Index, which includes two 0–10 scales for back pain and leg/buttock pain. A score of 4/10 or more is considered ‘‘relevant pain’’. Results: We identified 85 AIS patients [74 (87%) females] fitting the inclusion criteria. Of these, 60 were aged 12–18 years (mean 15.5 ± 1.7 years) and 25 were 19–30 years (mean 22.5 ± 3.1 years). There were no significant differences (p [ 0.05) between the age-groups for coronal Cobb angles of the main curves or Lenke curve types and these curves showed no correlation with pain intensity (p [ 0.05). Back pain was correlated with age (r = 0.31, p = 0.004). Preoperatively, 42% patients had a back pain score of C4/10 (52% in adults, 38% in adolescents). Just 8% patients had a leg pain score of C4/10 (16% in adults, 5% in adolescents). Those with notable back pain showed a significant (p \ 0.0001) improvement 2 years after surgery. There was no significant difference in the extent of improvement in older and younger patients (p = 0.22;) Conclusion: Young adults with idiopathic scoliosis more frequently have relevant back pain (C4/10 on pain scale) and have pain of higher average intensity than do adolescents. In patients with relevant back pain at baseline, surgery is associated with a statistically and clinically significant alleviation of pain, independent of age. Skeletally mature young adults benefit from deformity surgery as much as adolescents in terms of back pain relief. Disclosures: Author 1: none; Author 2: none; Author 3: none; Author 4: grants/research support: Depuy-Synthes; Author 5: none; Author 6: consultant: DePuy Synthes Spine,royalties: DePuy Synthes Spine.

Spine Surgery Unit, Vall d’Hebron University Hospital, Barcelona, Spain Summary: In this multicenter study the postoperative sagittal spinopelvic parameters in adult spinal deformity (ASD) patients with good clinical results and no mechanical failure were dependent on PI as well as age, weight and gender. Hypothesis: Sagittal alignment in the fused ASD with ideal results varies with patient´s demographics. Design: Retrospective analysis of prospectively collected data from an international multicenter ASD database Introduction: Often the widely used Schwab-SRS alignment goals are associated to poor clinical outcomes and mechanical failures. A better understanding of alignment characteristics is desirable to improve overall outcomes after ASD surgery. Our goal was to identify the individual alignment characteristics predicting optimal outcome after ASD surgery. Methods: Retrospective analysis of a prospectively collected data from 6 deformity centers sharing a comprehensive ASD database. Patients with PSFI [3 segments, [2 years f-up and optimal clinical outcome (ODI \22 at final f-up and no mechanical failures) were analysed. We used multivariate logistic regression analysis with a stepwise procedure to find the best combination of non-modifiable patient and radiological variables that would best predict each of the dependent radiological outcome variables [LL, SS, PT, global tilt (GT), SVA]. These dependent radiological variables were then modeled by the least squares method. The difference between predicted and actual LL was also modeled as a function of demographic variables. Results: At the time of the study 596 surgical patients were included in the database. 87 met inclusion criteria. Mean age was 40 (SD 18.2; range 18–84); 72% were women. Mean number of fused segments was 9.7 (SD 3.5). 33% had pelvic fixation. Multivariate logistic regression analyses consistently found that individual alignment was dependent on age and PI (p \ 0.0001; Table 1). Weight, but not BMI, determines optimal SS, PT and GT (p \ 0.001), stressing the importance of load rather than load distribution as a predictor of failure. SVA is dependent on age and gender. Also, optimal sagittal alignment is proportional to pelvic anatomy (PI) but with a coefficient of 0.67 in line with recent literature. Our models were more predictive than published ones (R [ 0.637; R2 [ 0.406). Older and heavier patients were compensating more (had higher GT and PT) for a fixed PI and LL and could deviate less from their calculated ideal values before failing. Finally, the difference between actual and predicted LL was found to vary with age confirming lower deviation from predicted value with age.

123

S374 Conclusion: Our findings question the universal validity of the PI–LL concept and favor a more sophisticated approach where pelvic anatomy, age, gender and weight all play a role in patients with good results. Surgical planning in the elderly and obese should be strict to achieve optimal clinical results as they tolerated worst any deviation from their ideal values Disclosures: Author 1: none; Author 2: grants/research support: Depuy Synthes; Author 3: none; Author 4: grants/research support: Depuy Synthes, Medtronic, stock/shareholder: IncredX, royalties: Zimmer Biomet, AOSpine; Author 5: grants/research support: DepuySynthes; Author 6: grants/research support: DEPUY SINTHES, consultant: K2M. DEPUY SINTHES; Author 7: grants/research support: DEPUY SYNTHES, consultant: DEPUY SYNTHES, MEDTRONIC, royalties: ALPHATEC, SPINEART, CLARIANCE; Author 9: grants/research support: DePuy Synthes Spine, Medtronic, consultant: DePuy Synthes Spine, Biomet; Author 10: grants/research support: DePuy-synthes.

P81 RESTORATION OF THORACIC KYPHOSIS IN AIS PATIENTS WITH THORACIC HYPOKYPOSIS OR LORDOSCOLIOSIS USING MULTIPLE PONTE OSTEOTOMIES WITH OR WITHOUT ADDITIONAL BILATERAL RIB OSTEOTOMIES Sinan Kahraman, Alim Can Baymurat, Selhan Karadereler, Cem Sever, Feride Gokce Demir, Isik Karalok, Ayhan Mutlu, Tunay Sanli, Meric Enercan, Azmi Hamzaoglu

Eur Spine J (2017) 26 (Suppl 2):S335–S405 the preoperative values and f/up values, were 19.3 and 17.8 for group A, and 24.2 and 21.2 for group B. Mean number of MPO were 3 (2–5) in group A, and 4 (2–6) in group B. Mean correction in pts who underwent MPO more than three levels was significantly higher than those with osteotomies less than three levels (p \ 0.05). In group B, the mean number of BRO levels was 5 (3–8). Kyphosis restoration in the sagittal plane was better in group B, however there was no statistically significant difference between the groups (p [ 0.05). Comparison of preop and f/up PFT in both groups showed statistically significant improvement (p \ 0,05). In f/up, SRS22r and ODI values were 4.4 and 6, respectively for group A, and 4.2 and 8, respectively for group B. Conclusion: In AIS patients with TH and TL, the addition of BRO to MPO provides better kyphosis restoration.Better restoration of thoracic kyphosis was achieved when MPO performed more than three levels. The addition of BRO to MPO will increase the pulling vertebral column posteriorly segment by segment in the TH or TL segments, which facilitates correction.PFT showed similar improvement at the end of 2 years 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: MEDTRONIC.

P82 THE FATE OF CERVICAL KYPHOSIS WITH NEGATIVE SAGITTAL MALALIGNMENT FOLLOWING SURGICAL CORRECTION OF ADULT SPINAL DEFORMITY: IS IT PERMANENT WHEN SEEN AFTER SURGICAL CORRECTION OF SAGITTAL MALALIGNMENT?

Istanbul Spine Center, Istanbul, Turkey Background: The addition of bilateral rib osteotomies (BRO) to multiple Ponte osteotomies (MPO) enables better kyphosis restoration in AIS pts with thoracic hypokyphosis (TH) or thoracic lordosis (TL). In both techniques, pulmonary function tests (PFT) show improvement at the end of min. 2 years. Purpose of the study: The aim of this study was to investigate the clinical, radiologic, and pulmonary functions at the end of min. 2 years follow up in pts with TH or TL who underwent either only MPO or additional BRO to MPO Materials and methods: 62 AIS pts (11M, 51F) with TH (n = 49) or TL (n = 13) who underwent MPO with or without BRO were included. Group A: pts who underwent only MPO (n = 40), group B: pts who underwent BRO (between T4 and T10) in addition to MPO (n = 22). Mean corrections in the coronal plane, and the increases in the sagittal plane parameters (T2–T12), (T5–T12) were compared between the two groups. Preop and f/up FVC, FEV1, AND FEV1/ FVC values on PFT were compared between the two groups. The clinical improvements were compared by changes in the SRS-22 and ODI values. Statistical analyses were performed with repeated measures T test and two way anova for mixed measures Results: Mean age was 15.5 (13–18). Mean f/up was 57.5 months (24–126). Correction rates for MT and TL/L curves were 88 and 78%, respectively for group A, and 84% and 76%, respectively for group B. Mean increases in T2–T12, T5–T12 angles in the sagittal plane, between

123

Jung-Hee Lee, Sung Joon Shin, Dong-Gune Chang, Jin-Soo Kim, Chul-Hee Lee, Tae-Jin Kim, Won-Ju Shin Kyung Hee University, Seoul, Korea Introduction: Several studies have been published with regard to postoperative cervical kyphosis resulting from an effort to achieve an optimal sagittal balance. Therefore, our purpose was to evaluate postoperative cervical kyphosis according to the amount of postoperative lumbar lordosis correction compared to the pelvic incidence (PI–LL), and to the postoperative and last follow up sagittal balance. Methods: Degenerative lumbar kyphosis (n = 122) who underwent surgical correction with a minimum of 2-year follow up were analyzed. Four related studies were performed by comparing spinopelvic parameters evaluated by categorizing the patients into three different ways (postoperative PI–LL, postoperative C7PL, last follow-up C7PL) and performing a Pearson’s correlative study between spinopelvic parameters of the postoperative and last follow up to assess the relationship of each spinopelvic parameters. Results: There were significant differences or direct relationship in assessing cervical lordosis and T1 slope with regard to postoperative C7PL (negative: n29, \50 mm; n = 60, [50 mm; n = 33), but not with postoperative PI–LL (\-10; n = 50, 10 \ ±10; n = 43, [10; n = 29), there were no significant correlation between postoperative

Eur Spine J (2017) 26 (Suppl 2):S335–S405 thoracic kyphosis and postoperative cervical lordosis, which showed significant correlation of those of the last follow up. However, cervical kyphotic change with postoperative SVA (C7PL \0 or \50 mm) was restored to cervical lordosis at last follow-up. Conclusion: Therefore, postoperative cervical kyphosis or hypolordosis is temporary during the adjusting period of thoracic kyphosis restoration, after postoperative negative sagittal malalignment. However, the degree of cervical lordosis may differ from the postoperative or last sagittal balance status. Disclosures: Author 1: none; Author 2: none; Author 3: none; Author 4: none; Author 5: none; Author 6: none; Author 7: none.

P83 HOW TO PREVENT PROXIMAL JUNCTIONAL KYPHOSIS FOLLOWING COMBINED ANTERIOR COLUMN REALIGNMENT AND POSTERIOR SHORT FUSION TO THE SACRUM IN DEGENERATIVE LUMBAR KYPHOSIS Jung-Hee Lee, Sung Joon Shin, Dong-Gune Chang, Jin-Soo Kim, Chul-Hee Lee, Tae-Jin Kim, Won-Ju Shin 1

Department of Orthopaedic Surgery, College of Medicine, Kyung Hee University, Seoul, Korea; 2Department of Orthopaedic Surgery, St. Paul’s Hospital, The Catholic University of Korea College of Medicine, Seoul, Korea (*only the third author)

Study design: A prospective study. Introduction: Optimal sagittal vertical axis (SVA) following surgery is one of the most important factors to predict success in adult spinal sagittal deformity. Although three-column osteotomy is a powerful method to achieve optimal SVA, many complications have been reported. Purpose: To minimize complications as well as maintaining optimal SVA, combined short anterior column realignment (ACR) and posterior spinal fusion (PSF) might be an alternative surgical strategy. Our purpose was to demonstrate the ideal indication for this method. Materials and methods: This study included 27 patients (average age 66.6 years) with degenerative lumbar kyphosis and compensated thoracolumbar (TL) lordosis, underwent short PSF with posterior column osteotomy (PCO) following ACR. The minimum follow up period was 5 years. Postoperative lumbar lordosis (LL) was compared to the ideal LL, which was estimated by a formula based on the pelvic incidence (PI) (overcorrection: postoperative LL [ ideal LL, undercorrection: postoperative LL \ ideal LL). Results: Comparing both groups, the optimal SVA group (\50 mm, n = 16), showed an average PI of 44.4, and 53.8 in the suboptimal SVA group (C50 mm, n = 11). The average LL was +4.4 vs +4.2 preoperatively, and -48.1 vs -35 postoperatively with an average LL correction of 47.1 as a total. The prevalence of proximal

S375 junctional kyphosis (PJK) was significantly higher in the suboptimal SVA group (p = 0.0402). Conclusion: The ideal indication for this procedure should include low PI (\50) and TL lordosis for preventing PJK in degenerative sagittal malalignment with an average LL correction of 47.1. Disclosures: Author 1: none; Author 2: none; Author 3: none; Author 4: none; Author 5: none; Author 6: none; Author 7: none.

Table. Comparison between Optimal SVA group and Suboptimal SVA group Optimal SVA(n=16) Suboptimal SVA (n=11) Variables Mean ± SD Mean ± SD 65.1 ± 6.3 68.8 ± 5.7 Age( year) Postoperative optimal SVA 16/0 9/2 (Yes/No) SVA (mm) 157.0 ± 61.6 116.7 ± 64.1 Preoperative -0.4 ± 26.6 18.6 ± 44.5 Postoperative 20.9 ± 23.0 88.9 ± 53.0 Last follow-up 44.1 ± 6.9 53.8 ± 11.0 Pelvic incidence Pelvic tilt (°) 28.4 ± 14.1 42.5 ± 12.6 Preoperative 15.3 ± 8.1 21.4 ± 11.6 Postoperative 17.2 ± 11.5 32.3 ± 11.7 Last follow-up PT ratio (PT/PI x100, %) 62.5 ± 23.9 83.0 ± 30.0 Preoperative 31.9 ± 14.5 39.0 ± 22.4 Postoperative 32.5 ± 18.6 54.8 ± 16.9 Last follow-up Sacral slope (°) 15.7 ± 10.0 10.8 ± 17.5 Preoperative 32.3 ± 8.0 35.5 ± 15.3 Postoperative 32.7 ± 9.5 26.9 ± 11.3 Last follow-up Thoracic kyphosis (°) 4.0 ± 16 . 1 -0.7 ± 12.3 Preoperative 14.9 ± 14.8 9.3 ± 8.0 Postoperative 19.5 ± 13.2 5.4 ± 9.4 Last follow-up Thoracolumbar junction (°) -9. 7 ± 6.7 -9.7 ± 7.4 Preoperative -3.9 ± 11.8 -0.6 ± 11.6 Postoperative 5.9 ± 10 . 8 9.1 ± 15.6 Postoperative change -4.2 ± 9.4 16.0 ± 16.1 Last follow-up -0.4 ± 6.3 16.5 ± 13.1 Last change Lumbosacral junction (°) 2.3 ± 8.8 1.3 ± 21.3 Preoperative -35.3 ± 11.1 -32.5 ± 12.7 Postoperative -34.3 ± 12.8 -26.7 ± 15.0 Last follow-up Lumbar lordosis(°) 4.4 ± 15 . 1 4.2 ± 23.2 Preoperative -48.1 ± 4.6 -35.0 ± 12.9 Postoperative -45.2 ± 4.4 -20.7 ± 14.7 Last follow-up -49.6 ± 5.0 -56.1 ± 8.0 Estimated ideal 1.5 ± 5.7 21.1 ± 10.3 Postoperative - Ideal 7/9 0/11 Overcorrection (Yes/No) Fused segments angle(°) 2.2 ± 7.4 5.2 ± 20.6 Preoperative -44.4 ± 5.4 -40.3 ± 9.9 Postoperative 46.6 ± 7.9 45.5 ± 17.2 Postoperative correction -42.7 ± 5.3 -38.1 ± 10.4 Last follow-up PJA (°) - 3. 2 ± 7.7 -4.1±3.4 Preoperative -2.7 ± 5.9 4.0±11.1 Postoperative 1.5 ± 9.9 17.7 ± 11.6 Last follow-up 3/13 7/4 PJK prevalence (Yes/No) SVA: sagittal vertical axis ; PJA: proximal junctional angle; PJK: proximal junctional kyphosis; 1): Independent sample t-test ; 2): Wilcoxon's signed rank test, 3): Fisher's exact test

P-value 0.1832) 0.15673) 0.11271) 0.1761) 0.00161) 0.00921) 0.01311) 0.07982) 0.00172) 0.05931) 0.32881) 0.00332) 0.3581) 0.53441) 0.15991) 0.50522) 0.26411) 0.00521) 0.9811) 0.47981) 1.00002) 0.00041) 0.00032) 0.90182) 0.55631) 0.16881) 0.6392) 0.00731) 0.00021) 0.01581) 0.00001) 0.02163) 0.08852) 0.22781) 0.83861) 0.20091) 0.67111) 0.15232) 0.00071) 0.04023)

P84 MINIMALLY INVASIVE ANTERIOR COLUMN RECONSTRUCTION IN PRIMARY ADULT DEGENERATIVE SCOLIOSIS CORRECTION SURGERY: LATERAL CAGES ARE ABLE TO RESTORE LORDOSIS AND PROVIDE INDIRECT DECOMPRESSION Robert Lee Department of Spinal Surgery, Royal National Orthopaedic Hospital NHS Trust, Stanmore, UK Introduction: Patients with adult degenerative scoliosis often present with leg pain (deformity causing neural compression), back pain and issues with sagittal balance. Complications following open correction surgery can be high and sagittal balance correction poor using a posterior only technique. Multiple anterior cages inserted via a minimally invasive lateral or anterior technique is an effective way of correcting the coronal deformity as well as restoring sagittal balance in patients with adult degenerative scoliosis. It also provides indirect decompression of neural structures. Hypothesis: The use of minimally invasive lateral cages in primary adult degenerative scoliosis surgery is an effective surgical strategy

123

S376 with good outcome scores, restoration of lumbar lordosis and correction of sagittal balance. Design: This is a retrospective review of prospectively collected data of a single surgeon case series of 48 patients with adult degenerative scoliosis with no previous fusion surgery and minimum 6 months follow-up. Methods: Pre and post-operative radiographic parameters: lumbar lordosis (LL), pelvic incidence–lumbar lordosis mismatch (PI–LL), sagittal vertical axis (SVA), pelvic tilt (PT) and Cobb angle. Outcome scores: VAS Back, VAS leg, EQ-5D, EQ-5D VAS, ODI, Roland Morris Disability Score (RMD). Results: There were 30 females and 18 males. Average age was 67.0yrs (54.9–83.4 years). Positive sagittal balance was an issue in 31 patients. 42 cases were purely MIS and 6 cases hybrid with open posterior fusion. 30 cases were two stage corrections. A total of 128 lateral cages were inserted with 1 level in 7 patients, 2 levels in 14 patients, 3 levels in 15 patients and 4 levels in 12 patients. The distribution of levels was as follows: L1/2—12 cages, L2/3—34 cages, L3/4—44 cages, L4/5—38 cages. Average post-operative scores showed improvement of SVA 90.1–35.3 mm, PI–LL 26–2 and Cobb angle 23–5. Average 6 month outcome scores were: VAS back 8–3, VAS leg 8–2, EQ-5D 0.257–0.720, EQ-5D VAS 44–74, ODI 64–28, RMD 16–11. These scores were maintained in patients reaching the 1 and 2 years followup mark. Conclusion: The use of minimally invasive lateral cages in primary adult degenerative scoliosis surgery is an effective surgical strategy with good restoration of lordosis. Disclosures: Author 1: grants/research support: K2M, consultant: K2M, Medtronic, Signus, SI-Bone.

Eur Spine J (2017) 26 (Suppl 2):S335–S405 cages. Careful choice of cage angle size at each lumbar level enables restoration of good sagittal alignment. Software is available to not only measure patient alignment parameters but also to simulate the operation. However there is often difficulty in executing the preoperative plan with techniques such as osteotomies. This study shows that surgical planning in anterior column reconstruction can accurately predict the eventual outcome. Hypothesis: Pre-operative surgical planning software can reliably predict the alignment outcome in anterior column reconstruction surgery. Design: A retrospective review of prospectively collected data of a single surgeon case series of 40 patients with positive sagittal balance (both compensated and uncompensated). Methods: All patients had erect pre-operative and post-operative whole spine X-rays including C7 and both femoral heads. Images were loaded onto pre-operative surgical planning software. Pre-operative surgical planning software was then used to measure the following parameters: lumbar lordosis (LL), pelvic incidence–lumbar lordosis mismatch (PI–LL), SVA and pelvic tilt (PT). Surgery was simulated using multilevel lateral or anterior cages. Pelvic tilt was then adjusted to normal values for each pelvic incidence to determine the predicted SVA. Predicted parameters were then compared to the final outcome. Results: There were 26 degenerative scoliosis, 2 spondylolisthesis, 8 iatrogenic flat back and 12 patients with proximal level degeneration. Surgimap planning was accurate to within 15.13 mm for SVA, 4 for LL, 4 for PI–LL and 6 for PT. Conclusion: Pre-operative surgical planning software provides a reliable way of predicting alignment outcomes in anterior column reconstruction surgery. Disclosures: Author 1: grants/research support: K2M, consultant: K2M, Medtronic, Signus, SI-Bone.

70 year Female with Degenerative Scoliosis and Positive Sagittal Balance. 3 level lateral , L5/S1 ALIF and T11 to S2AI MIS fusion

P85 RELIABILITY OF PRE-OPERATIVE SURGICAL PLANNING SOFTWARE IN PREDICTING POSTOPERATIVE ALIGNMENT IN PATIENTS UNDERGOING MINIMALLY INVASIVE MULTILEVEL ANTERIOR COLUMN RECONSTRUCTION FOR POSITIVE SAGITTAL BALANCE DEFORMITY Robert Lee Department of Spinal Surgery, Royal National Orthopaedic Hospital NHS Trust, Stanmore, UK Introduction: Surgical correction of positive sagittal balance can be performed via multilevel minimally invasive lateral and anterior

123

P86 DOES ARTHRODESIS STOPPING AT L5 CAUSE SAGITTAL DECOMPENSATION? Hyeong-Youl Park, Kee-Yong Ha, Young-Hoon Kim, Sang-Il Kim, Joo-Hyun Ahn, Jae-Won Lee Department of Orthopaedic Surgery, Catholic University of Korea, Seoul, Korea Introduction: The cause of sagittal decompensation after long instrumented lumbar fusion could be multifactorial. Advanced L5–S1 disc degeneration after long fusion stopping at L5 developed significant positive sagittal balance. However, there is limited information on the role of disc degeneration in the development of sagittal

Eur Spine J (2017) 26 (Suppl 2):S335–S405 decompensation after the treatment of degenerative sagittal imbalance. The purpose of this study was to identify any correlation of L5– S1 disc status and sagittal decompensation after long fusion stopping at L5. Methods: Fifty-nine consecutive patients with an average age of 68.6 years who underwent fusion of at least five segments from L5 were included. Preoperative grade of disc degeneration was evaluated with plain radiographs by Kuhns grades and MRI by Pfirrmann classification. For postoperative grade of disc degeneration was evaluated only with plain radiographs. Spinopelvic parameters were also measured. Results: Nine out of 41 patients had preoperative advanced degeneration of the L5–S1 disc (Weiner grade 2–3). Thirty-two out of 41 patients were assessed as ‘‘healthy discs’’ preoperative (Weiner grade 0–1) and were evaluated for SAD. By latest follow-up, L5–S1 SAD developed in 8 of these 32 patients (25%). Preoperative sagittal balance was significantly different between the ‘‘healthy’’ and SAD group. Sagittal balance at postoperative 3-month had been improved in most patients, but sagittal decompensation re-developed at the final follow-up: C7 plumb [5 cm in 87.5% of SAD patients (7/8) and 45.8% of ‘‘healthy’’ disc patients (11/24). There was a trend toward inferior Scoliosis Research Society Patient Questionnaire-22 pain scores at the final follow-up in SAD patients (P = 0.13). One out of 41 patients had extension of fusion to the sacrum. An additional 5 of 8 SAD patients were considered for extension to the sacrum but comorbidities precluded surgery (1 patient) or the patients declined revisional surgery (4 patients). Conclusion: Advanced L5-S1 DDD developed in 25% of ASD patients after long fusions to L5 within 2–10 years-follow-up. SAD patients had preoperative sagittal imbalance and presented sagittal decompensation at follow-up despite postoperative improvement of sagittal balance. Disclosures: Author 1: none; Author 3: none; Author 4: none; Author 5: none.

P87 THE PATTERN OF LOSS OF CORRECTION IN ANKYLOSING SPONDYLITIS-RELATED THORACOLUMBAR KYPHOSIS AFTER POSTERIOR WEDGE OSTEOTOMY: A LONG-TERM FOLLOW-UP INVESTIGATION Bangping Qian, Mu Qiao, Yong Qiu, Bin Wang Spine Surgery, Drum Tower Hospital of Nanjing University Medical School, Nanjing, China Background: Pedicle subtraction osteotomy (PSO) has been widely used as an effective procedure to correct thoracolumbar kyphosis caused by ankylosing spondylitis (AS). Short-term studies have demonstrated good surgical outcomes, but there is a paucity of literature focusing on middle-term results. Purpose of the study: To assess the durability of radiographic and clinical outcomes in thoracolumbar kyphosis secondary to AS following lumbar PSO with over 5-year follow-up. Materials and methods: We performed a retrospective review of 155 consecutive AS patients undergoing lumbar PSO from January 2001 to December 2011 at our institution. Twenty-four patients who fulfilled the inclusion and exclusion criteria were included in this study with an average follow-up of 6.9 years (range 5–15 years). Radiographical evaluations included global kyphosis (GK), lumbar lordosis (LL), sagittal vertical axis (SVA), spinal-pelvic angle (SSA), kyphosis of proximal non-fused segments (KPNS), angle of fused

S377 segments (AFS), proximal junctional angle, osteotomized vertebra angle (OVA), distal intervertebral disc wedging (DIDW) and proximal junctional angle. Meanwhile, clinical outcomes were assessed by the Oswestry disability index (ODI) and Numerical rating scale (NRS). Results: The average correction per PSO segment was 34.9. Significant improvement in sagittal parameters were found postoperatively (P \ 0.05), and no obvious deterioration was noticed during the minimum 5-year follow-up. Mild loss of correction in GK (2.82) and LL (3.77) were observed at the final follow-up (P \ 0.05). The KPNS and DIDW increased from 26 and -5.0 postoperatively to 30 and -2.2 at the latest follow-up (P \ 0.05), respectively. In contrast, no significant diminishment was identified in OVA and AFS (P [ 0.05). The ODI and NRS improved significantly from 20.6 and 6.6 preoperatively to 5.9 and 2.3 at the final follow-up (P \ 0.05). Conclusion: PSO is an effective procedure for treatment of thoracolumbar kyphosis caused by AS and can maintain sustained surgical outcomes without long-term complications. The detected loss of correction was mainly attributable to non-instrumented segments without fully ossified bridging syndesmophyte. Disclosures: Author 1: none; Author 2: none; Author 3: none; Author 4: none.

P88 SPONTANEOUS BONE UNION OF VERTEBRAL PSEUDARTHROSIS AFTER PEDICLE SUBTRACTION OSTEOTOMY FOR THORACOLUMBAR KYPHOSIS SECONDARY TO ANKYLOSING SPONDYLITIS Yong Qiu, Xinkun Cheng, Bangping Qian, Jun Jiang, Zhe Qu, Yunpeng Zhang, Zezhang Zhu Drum Tower Hospital of Nanjing University Medical School, Nanjing, China Background: Pseudarthrosis appears as a nonunion occurring transversely through intervertebral disc and/or vertebral body in AS. PSO through pseudarthrosis has been confirmed to be effective and safe to correct thoracolumbar kyphosis for these patients. Purpose of the study: To evaluate the effectiveness and safety of pedicle subtraction osteotomy (PSO) with posterior fusion crossing pseudarthrosis in correcting thoracolumbar kyphosis secondary to ankylosing spondylitis (AS), and to investigate the spontaneous bone healing in pseudarthrosis after surgery. Materials and methods: Twelve AS patients (10 males/2 females) with pseudarthrosis underwent PSO below/above the pseudarthrosis with posterior fusion (group P) and 35 AS patients (32 males/3 females) without pseudarthrosis underwent PSO (group NP) were included in the study. The Oswestry Disability Index (ODI) scores and visual analogue scale (VAS) were used to assess life quality. The sagittal spinal-pelvic parameters including global kyphosis (GK), Cobb angle of pseudarthrosis levels (CPL), thoracic kyphosis, lumbar lordosis (LL), pelvic incidence, pelvic tilt, sacral slope and sagittal vertical axis (SVA) were measured and compared between the two groups. Results: All of the GK, LL, SVA, VAS and ODI scores improved significantly in group P (p \ 0.05). No significant difference was noted between the two groups (p [ 0.05). There was no obvious loss of correction (1.8 ± 1.8) in group P. The loss of CPL is 0.3 ± 5.3 in group P. Solid radiological fusion was observed in the segment of pseudarthrosis in 27.5 months on average. Conclusions: For AS patients with pseudarthrosis not complicating with neurological deficits, PSO below/above the pseudarthrosis with

123

S378

Eur Spine J (2017) 26 (Suppl 2):S335–S405

only posterior instrumentation is a method effective for spontaneous bone healing, if the instrumentation crosses the pseudarthrosis. This important finding implies that anterior debridement of pseudarthrosis and supplemental fusion are not indispensable for these patients before or after PSO surgical correction for kyphosis. Disclosures: Author 1: none; Author 2: none; Author 3: none; Author 4: none; Author 5: none; Author 6: none; Author 7: none.

P89 AN ANALYSIS FOR SAFETY INSERTION OF SACRAL ALAR-ILIAC SCREW ON THE FLUOROSCOPY, USING THREE-DIMENSIONAL ANALYZING SOFTWARE Kentaro Yamada, Yuichiro Abe, Shigenobu Satoh Department of Orthopaedic Surgry, Wajokai Eniwa Hospital, Hokkaido, Japan Purpose: S2 alar-iliac (SAI) screws have been common anchor recently in cases of lumbosacral fixation. The screw deviation to anterior or caudal direction has potential risk for major vessel injury, internal iliac artery or superior gluteal artery. Therefore, use of navigation system or intraoperative CT is recommended to avoid such injuries. On the fluoroscopy, the tear-drop-view has been recommended to confirm adequate screw insertion. The flaw of tear-dropview is difficult to set the screws and review the fluoroscopy simultaneously due to the setting of the fluoroscopy unless special surgical table. The pelvic inlet view is another recommended methods, however, no study have investigated how the beam should be tilted. The purpose of this study was to investigate the condition of the fluoroscopy which could provide accurate information of SAI screw deviation under anteroposterior or lateral view using three-dimensional analyzing software. Materials and methods: A total of 50 patients who were over 40 years old and taken abdominal thin-slice CT in the outpatient clinic of internal medicine were included in this study. The obtained CT images were reconstructed using 3D analyzing software for simulation of spinal screws. Ideal SAI screws (7.0 mm 9 80 mm) were individually set on three-dimensional analyzing software from entry point of 1 mm inferior and 1 mm lateral to the S1 dorsal foramen. Anterior or caudal deviated screws were defined as deviated half thread of screw by rotation anteriorly or caudally from the entry point of the ideal screw. The conditions which provide proper recognition for the screw deviation were investigated on lateral view and anteroposterior view with beam tilted caudally using the raysum methods in order to virtualize fluoroscopic images. Results: The anterior deviated screws were not recognized on lateral view in all cases, but recognized on pelvic inlet view (Figure: arrow) in all cases. The anterior deviated screw were recognized in specific range of beam tilted angle from S1 slope: 32.8 ± 12.1–101.8 ± 15.0. If the beam tilt angle set by 60, all anterior deviated screw, except one case, were recognized. The recognizable beam tilt angle ranges for caudal deviated screw were 15.9 ± 12.0–46.6 ± 11.4. However, 21% of screws were not able to be recognized in any inlet angle. Conclusions: The safety margins of SAI screws were typically smaller in anterior direction than in caudal direction. Intraoperative fluoroscopic setting was recommended 60 inlet from S1 slope to avoid anterior screw deviation in cases which fail to obtain clear images by the tear-drop-view. The lateral view was recommended to make sure that SAI screw was not deviated caudally. Disclosures: Author 1: none; Author 2: none; Author 3: none.

123

P90 THE DIFFERENCES OF THE SAFETY MARGINS OF SACRAL ALAR-ILIAC SCREW BETWEEN TWO ENTRY POINTS Kentaro Yamada, Yuichiro Abe, Shigenobu Satoh Department of Orthopaedic Surgry, Wajokai Eniwa Hospital, Hokkaido, Japan Purpose: S2 alar-iliac (SAI) screws have been common anchor recently in cases of lumbosacral fixation. The screw deviation to anterior or caudal direction has potential risk for major vessel injury, internal iliac artery or superior gluteal artery. Two entry points have been reported as insertion of SAI screw. However, no study have investigated difference between the two entry point of SAI screws. The purpose of this study was to evaluate safety margins of SAI screw by two entry point and investigate characteristics of patient required special attention on the screw insertion, using three-dimensional (3D) analyzing software. Materials and methods: A total of 50 patients who were over 40 years old and taken abdominal thin-slice CT in the outpatient clinic of internal medicine were included in this study. The obtained CT images were reconstructed using 3D analyzing software for simulation of spinal screws. The ideal SAI screws (7.0 mm 9 80 mm) were set on software by two entry point (A: the midline between the S1 and S2 forsal foramen meet the lateral sacral crest, B: 1 mm inferior and 1 mm lateral to the S1 dorsal foramen). Anterior or caudal deviated screws were defined as deviated half thread of screw by rotation anteriorly or caudally from the entry point of the ideal screw. The safety margins of S2AI screw were measured by rotational angle between ideal and deviated screw on anterior or caudal direction in two entry points. The differences of safety margin were evaluated in terms of direction, laterality, sex, and entry point. Moreover, the cases with small safety margin were investigated in the point of the view of location of L5 vertebral body; Deep-seated L5 was defined that the superior endplate of L5 existed below intercrestal line.

Eur Spine J (2017) 26 (Suppl 2):S335–S405 Results: The anterior safety margin of the two entry points were A: 9.1 ± 1.6, B: 9.7 ± 1.5. The caudal safety margin of the two entry points were A: 10.9 ± 3.8, B: 13.9 ± 4.1. In both entry point A and B, the safety margin in anterior direction were smaller than caudal direction (p \ 0.001). In both anterior and caudal direction, screws with entry point A were indicated smaller safety margin than those with entry point B. Patients with Deep seated L5 (n = 16) were showed smaller caudal safety margin in both entry points, especially mean safety margin in entry point A was only 8.38. The ratio of subjects with caudal safety margin \10 was significantly higher in patients with deep-seated L5 than those without deep-seated L5 (75 and 27.9%, p \ 0.001). Laterality and sex difference of each screw did not differ. Conclusions: The safety margins of SAI screws were smaller in anterior direction than in caudal direction, and by entry point A than B. The entry point B is recommended in terms of safety screw insertion. Safety margin of SAI screw tended to be small in caudal direction in patients with deep-seated L5. The SAI screw should be set toward cranial from entry point B in patients with deep-seated L5. Disclosures: Author 1: none; Author 2: none; Author 3: none.

P91 RELATIVE LUMBAR LORDOSIS (RLL): A NEW PI-BASED PROPORTIONAL PARAMETER THAT QUANTIFIES LUMBAR LORDOSIS MORE PRECISELY COMPARED TO PI–LL CONCEPT Caglar Yilgor, Nuray Sogunmez, Yasemin Yavuz, Ibrahim Obeid, Frank S. Kleinstueck, Francisco Javier Sanchez Perez-Grueso, Emre R. Acaroglu, Ferran Pellise´, Ahmet Alanay, ESSG European Spine Study Group

S379 mismatch for different PI values. RLL was not effected by PI (r = -0.093, p [ 0.05) and quantified the divergence from ideal lordosis for all PI values. When analyzed by RLL, each of PI–LL ‘0’, ‘+’ and ‘++’ groups were further divided into 2 or 3 distinct subgroups of patients that have different PI values (p = 0.000, p = 0.000 and p = 0.003, respectively). RLL subgroups within the same PI–LL category displayed different mechanical complication rates (p = 0.000, p = 0.000 and p = 0.001, respectively) (Figure). RLL had better correlations to ODI, COMI, SF-36 PCS and MCS and all SRS-22 subdomains when compared to PI–LL (p \ 0.05). Conclusion: As a linear value, PI–LL may be insufficient or misleading in quantifying the target lordosis for the whole spectrum of pelvic incidence. PI-based proportional parameter of relative lumbar lordosis reveals that each Schwab PI–LL group consists of inhomogeneous subgroup of pts with different mean PI values and mechanical complication rates. RLL offers a proportional quantification of lumbar lordosis for all PI sizes. RLL predicts mechanical complications more precisely and has better correlations to HRQoL. Disclosures: Author 1: none; Author 2: grants/research support: DePuy Synthes; Author 3: none; Author 4: grants/research support: Depuy Synthes, consultant: Depuy Synthes, medtronic, royalties: Alphatec, spineart, clariance; Author 5: grants/research support: Depuy-Synthes; Author 6: grants/research support: DEPUY SINTHES, consultant: K2M. DEPUY SINTHES; Author 7: grants/ research support: Depuy Synthes, Medtronic, stock/shareholder: IncredX, royalties: Zimmer Biomet, AOSpine; Author 8: grants/research support: dePuy Synthes Spine, Medtronic, consultant: dePuy Synthes Spine, Biomet; Author 9: grants/research support: Depuy Synthes, consultant: Medtronic; Author 10: grants/research support: DePuy-synthes.

School of Medicine, Acibadem University, Istanbul, Turkey Background: Quantification of the spinopelvic mismatch with the simplistic criterion of PI–LL within 10 has limitations and needs to be adapted to the intrinsic pelvic morphology of each patient. Although PI–LL is useful for average PI sizes, it fails to quantify spinopelvic mismatch for patients with upper and lower normal PI values. PI-based proportional parameter of relative lumbar lordosis (RLL) is described as ‘measured minus ideal LL’ and indicates the amount of lordosis relative to the ideal lordosis in proportion to PI (ideal LL = 0.62 9 PI + 29) RLL \-25 indicates severe hypolordosis, -24 B RLL \ -14: moderate hypolordosis, -14 B RLL B 11: aligned and RLL [11: hyperlordosis. Purpose: The aim was to compare PI–LL and RLL according to mechanical complication rates and HRQoL score correlations. Methods: This study was a retrospective analysis of a prospectively collected multicentric data of adult spinal deformity patients. Inclusion criteria were C4 levels fusion, and C2 years follow-up. 222 patients (168 female, 54 male) met the inclusion criteria. Mean age was 52.2 ± 19.3 (18–84) years. Mean follow-up was 28.8 ± 8.2 (24–62) months. ODI, COMI, SF-36 PCS, MCS and SRS-22 pain, function, mental health, self-image and subtotal scores were used as patient-reported health related quality of life (HRQoL) scores. PJK/ PJF, DJK/DJF, rod breakage and implant-related complications were recorded as mechanical complications. Correlations between PI–LL, RLL, PI and HRQoL were analyzed using Pearson correlation coefficient. PI values and mechanical complication rates in RLL subgroups for each PI–LL category were compared using one-way ANOVA, Independent Samples t and Chi squared tests. Results: Changes in PI–LL were effected by changes in PI (r = 0.441, p \ 0.001) reducing its ability to quantify spinopelvic

123

S380 P92 RELATIVE SPINOPELVIC ALIGNMENT (RSA): A NEW PIBASED PROPORTIONAL PARAMETER THAT QUANTIFIES SAGITTAL ALIGNMENT MORE PRECISELY COMPARED TO SVA

Eur Spine J (2017) 26 (Suppl 2):S335–S405 grants/research support: Depuy Synthes, Medtronic,stock/shareholder: IncredX, royalties: Zimmer Biomet, AOSpine; Author 8: grants/research support: DePuy Synthes Spine, Medtronic,consultant: DePuy Synthes Spine, Biomet; Author 9: grants/research support: Depuy Synthes, consultant: Medtronic; Author 10: grants/research support: DePuy-synthes.

Caglar Yilgor, Nuray Sogunmez, Yasemin Yavuz, Ibrahim Obeid, Frank S. Kleinstueck, Francisco Javier Sanchez Perez-Grueso, Emre R. Acaroglu, Ferran Pellise´, Ahmet Alanay, ESSG European Spine Study Group School of Medicine, Acibadem University, Istanbul, Turkey Background: Use of SVA with Schwab modifier cutoffs can be misleading since it is dependent on patient positioning and pelvic rotation. SVA gives an estimation of a trunk’s general alignment by defining the positioning of C7 in reference to the sacrum and it has been reported to be one of the most reliable radiographic predictors of Health Related Quality of Life (HRQoL) scores. Yet, the goal of SVA \4 cm can be misleading as it ignores negative malalignment as a potential cause of failure and it can be masked by compensation. Relative spinopelvic alignment (RSA) is described as ‘measured minus ideal global tilt (GT)’ and indicates the amount of malalignment relative to the ideal GT in proportion to PI. (ideal GT = 0.48 9 PI-15) RSA [18 indicates severe positive malalignment, 10 B RSA \ 18:moderate positive malalignment -7 B RSA B 10: aligned and RSA \-7: negative malalignment. Purpose: The aim was to compare SVA and RSA according to mechanical complication rates and HRQoL score correlations. Methods: This study was a retrospective analysis of a prospectively collected multicentric data of adult spinal deformity patients. Inclusion criteria were C4 levels fusion, and C2 years follow up. 222 patients (168 female, 54 male) met the inclusion criteria. Mean age was 52.2 ± 19.3 (18–84) years. Mean follow-up was 28.8 ± 8.2 (24–62) months. ODI, COMI, SF-36 PCS, MCS and SRS-22 pain, function, mental health, self-image and subtotal scores were used as patient-reported health related quality of life (HRQoL) scores. PJK/ PJF, DJK/DJF, rod breakage and implant-related complications were recorded as mechanical complications. The Chi squared test and oneway ANOVA were performed to compare the mechanical complication rates and HRQoL scores in SVA ‘0’ subgroups classified by RSA. Results: 141 patients were post-operatively SVA ‘0’. Of those, 12 had negative alignment, 78 were aligned, 32 were moderate positively and 19 were severe positively malaligned according to RSA. Mechanical complication rates were 41.7, 23.1, 50.0 and 84.2%, respectively (p = 0.000). Mean ODI, COMI, SF-36 PCS and SRS-22 subtotal scores were different in these subgroups (p = 0.000, p = 0.001, p = 0.008 and p = 0.001, respectively) (Figure). Conclusion: SVA ‘0’ consists of both aligned and negatively/positively malaligned subgroups of patients with different mechanical complication rates and HRQoL scores. SVA is insufficient in quantifying the amount of spinopelvic alignment. Analysis of SVA ‘0’ patients with RSA revealed that SVA \4 cm patients consist of inhomogeneous group of patients with different mechanical complication rates and HRQoL scores. Relative spinopelvic alignment better distinguishes aligned and malaligned patients offering a ‘one-size-fitsall’ solution with proportional quantification of sagittal alignment for all PI sizes and pelvic compensation conditions. Disclosures: Author 1: none; Author 2: grants/research support: DePuy Synthes; Author 3: none; Author 4: grants/research support: DEPUY SYNTHES, consultant: DEPUY SYNTHES, MEDTRONIC, royalties: ALPHATEC, CLARIANCE, SPINEART; Author 5: grants/ research support: Depuy-Synthes; Author 6: grants/research support: DEPUY SINTHES, consultant: K2M. DEPUY SINTHES; Author 7:

123

P93 POST-TUMOR SPINAL DEFORMITY: NON-OPERATIVE VERSUS OPERATIVE MANAGEMENT Nikita Zaborovskii, Dmitrii Ptashnikov, Dmitrii Mikhaylov, Sergei Masevnin, Oleg Smekalenkov, Olga Lapaeva, Zabioulah Mooraby, Yang Le, Elnara Zamanova R.R.Vreden Russian Research Institute of Traumatology and Orthopedics, Saint-Petersburg, Russia Introduction: In adult patients with spinal deformity, pain and disability are the most relevant findings to take the decision of surgical treatment. As we know, importance of sagittal balance on healthrelated quality of life (HRQOL) parameters have been proved. Spinal deformity can also present in the adult life as the result of tumor, radiation, pathological fractures or other causes. Purpose of the study: The purpose of our research is evaluation of patients with post-tumor spinal deformity. Materials and methods: Data were collected for 44 patients with spinal deformity (SRS-Schwab sagittal modifiers: two grade and more) and confirmed spinal oncology by biopsy. All patients have had oncology remission with a 3-year minimum follow-up. Spinal deformity was developed in the period between 1 and 6 years after chemotherapy or radiotherapy of tumor. Patients were divided into

Eur Spine J (2017) 26 (Suppl 2):S335–S405 three groups. The first non-operative group (NOP) included 18 patients [plasmacytoma (n = 12), multiple myeloma (n = 6)] who underwent nonoperative management (non-steroid anti-inflammatory medication, muscle relaxants, pain medication, muscle exercises, steroid blocks, brace). The second operative group with selected spinal fusion and correction of regional deformity (SOP) consisted of 12 patients [plasmacytoma (n = 10), multiple myeloma (n = 2)]. The third operative group with long spinal fusion and correction of global spinal alignments (LOP) consisted of 14 patients [plasmacytoma (n = 12), cervix uteri cancer metastasis (n = 1)] which were treated surgically. Low scores of HRQoL and unsuccessful non-operative approach were indications for surgical intervention in the operative groups. Repeated biopsy in the operative groups did not show tumor cells, this was evaluated as local control of tumor. Radiographical, HRQoL-parameters (VAS, ODI, SF36) and complications were analyzed and compared between groups. Kruskal–Wallis test, Fisher’s exact test were performed with R 3.3.2. Results: Patients in the LOP group showed a restoration of radiographical global spinal alignments after surgery compared with SOP group (p \ 0.001). The HRQoL-scores after 2 years are significantly higher in the operative groups than in the NOP group (p \ 0.05). Hovewer we did not find significant difference between LOP and SOP groups. It is important that no patients had local recurrence or metastasis of tumor during follow-up in the operative (LOP and SOP) groups. Conclusion: Spinal tumor under local control is not contraindication for spinal deformity surgery. Correction of sagittal alignments may have potentially benefits. 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.

P94 LUMBAR DISC GEOMETRY AFFECTS THE RISK FOR ROD FRACTURE IN ADULT SPINAL DEFORMITY (ASD) SURGERY Joseph Zavatsky, Brandon Cook, David Briski Ochsner Medical Center, New Orleans, LA; Spine and Scoliosis Specialists, Tampa, FL, USA Summary: Retrospective review of adult spinal deformity (ASD) surgical patients with X-rays and 2-year follow-up were included. Patients were divided into two Groups: Rod fracture (RF) vs No RF. There was no difference in pedicle subtraction osteotomy (PSO), sagittal vertical axis (SVA), or previous spinal surgery between the groups. Vertical stature and male gender were significantly higher in the RF group. Lumbar disc geometry, including disc heights, diameters, and volumes, were significantly larger in the RF vs no RF group. Hypothesis: Larger disc geometry results in micro-motion and increased rod stresses adjacent to the PSO or apical lumbar vertebra increasing the risk of rod fracture. Design: Single center retrospective review. Introduction: Rod fracture (RF) has significant consequences for patients, including pain, loss of correction, and revision surgery. Risk factors associated with RF include previous spine surgery, insufficient sagittal vertical axis (SVA) correction, and pedicle subtraction osteotomy (PSO). Increased lumbar disc geometry, regardless of PSO, SVA, or previous surgery, may be a risk factor affecting RF rates. Methods: All patients with ASD having open posterior fusion constructs crossing the T/L junction and pelvis with scoliosis X-rays and 2-year follow-up were included. Patients were divided into two groups: group 1—patients with RF; group 2—no RF.

S381 Results: 39 of 52 patients met inclusion criteria. Group 1—included 15/39 (38%) patients with RF all requiring revision surgery. Group 2—included 24 (62%) patients without RF. There was no difference in age, construct length, number of non-fused lumbar discs, previous spinal surgery, pre or postop SVA, or PSOs between the groups. Vertical stature and male gender were significantly higher in the RF Group. There were significant differences between the RF vs no RF groups in L1/2 disc heights (H) (8.5 vs 6.5 mm, p = 0.03), L2/3 H (9.6 vs 6.7 mm, p = 0.015); L1/2 diameters (D) (41.4 vs 34 mm, p \ 0.01), L2/3 D (43.2 vs. 36.2 mm, p = 0.01); and L1/2 volumes (V) (11714 vs 7817mm3, p = 0.02), L2/3V (14025 vs. 8955mm3, p = 0.02), respectively. L3/4 and L4/5 disc geometry was larger in the RF Group but did not reach statistical significance. Conclusions: The overall prevalence of RF was 38% with no differences in previous spinal surgery, SVA, or PSOs between the two Groups. Vertical stature and male gender were significantly higher in the RF Group. Patients with RF had significantly larger non-fused disc heights, diameters, and volumes adjacent to the PSO or apical lumbar vertebra. Increased disc geometrics may allow for increased micro-motion and increased rod stresses and rates of RF. Disclosures: Author 1: grants/research support: AO Spine,consultant: DePuy Synthes, Zimmer Biomet, Stryker,stock/shareholder: Surgical Solutions, Vivex, royalties: Zimmer Biomet; Author 2: none; Author 3: none.

GROWING SPINE

P95 ACCURATE RESTORATION OF THORACIC KYPHOSIS IN LENKE 1 CURVES USING BILATERAL VERTEBRAL COPLANAR ALIGNMENT Carlos Barrios, Karen Weismann, Philippe Mazeau, Guahua Lu¨, Lei Kuang, Jose´ Ignacio Maruenda Valencia Catholic University, Valencia, Spain Background/introduction: All pedicle screw (PS) techniques obtain satisfactory coronal correction, but fail to accurately restore the frequent thoracic hypokyphosis of lenke 1curves. Bilateral vertebral coplanar alignment (BVCA) is a recently described technique for surgical treatment of adolescent idiopathic scoliosis (AIS), particularly focused on restoration of the normal thoracic kyphosis. Purpose of the study: The aim of this study was to report the postoperative restoration of the thoracic sagittal profile in Lenke 1curves using BVCA. Materials and methods: Review of a prospective multi-center registry. A total of 49 AIS patients underwent surgical correction using BVCA technique. BVCA reduction maneuvers include the correction of the deformity while descending an implantable rod within slotted tubes attached to the convex side screws. Restoration of physiologic kyphosis is achieved by spreading the distal ends of the tubes at the thoracic spine using different sized spacers. Coronal and sagittal curve correction was assessed at preoperative and 2-year FU. Results: Mean preoperative Cobb of the main thoracic curves was 68.5 ± 17.8 and was corrected to 20.1 ± 14.3 (70.7%). The preoperative average apical vertebral rotation of 18.7 ± 4.2 and was corrected to 8.8 ± 4.6 (52.9%). T5–T12 kyphosis showed a significant improvement from a mean Cobb angle of 18.8 ± 18.4 (95% CI 13.3–24.8) to 23.5 ± 9.9 (95% CI 20.1–26.1) in the last postoperative evaluation (25% increase) (Wilcoxon rank test, Z = -2.058; p = 0.040). Patients with (-) sagittal modifier gained

123

S382 13.9 Cobb on average; those with (+) modifier decreased 13.5, and patients with N sagittal modifier remained almost unchanged (2.1) (Figure 1). Differences in the changes of thoracic sagittal profile between the three groups were statistically significant (p \ 0.001). Before surgery, 57.1% of cases were normokyphotic (Lenke modifier N), and this percentage increase significantly to 90.5% after surgical correction using BVCA (p \ 0.001). Only 4.8% remained with thoracic hypokyphosis after surgery. Conclusions: BVCA allows a three-dimensional correction in a standardized fashion. Coronal correction of the curve was similar to other PS techniques. However, BVCA permits a satisfactory and relevant restoration of thoracic kyphosis. BVCA is therefore particularly recommendable for adolescents and young adults with idiopathic scoliosis showing thoracic hyperkyphosis or hypokyphosis. Disclosures: Author 1: none; Author 2: consultant: Spineway; Author 3: consultant: Spineway; Author 4: none; Author 5: none; Author 6: none.

Eur Spine J (2017) 26 (Suppl 2):S335–S405 taking rotation and tilt of the endplates of each vertebra in 3-D into account using a previously validated image processing technique, the height of the vertebral bodies and discs on the exact anterior and posterior side and height of the laminae and interlaminar space were measured semi-automatically. Total length was calculated from thoracic Cobb to Cobb end vertebra in AIS patients and matched levels in controls. Results: The anterior side of the scoliotic vertebral bodies and discs (mean ± standard deviation) was 3.4 ± 2.7% longer as compared to the posterior side, and 6.4 ± 6.0% longer as compared to the posterior elements (p \ 0.001). Compared to the controls, the non-osseous structures (discs and interlaminar space) contributed most to the segmental wedging (at least three times more than osseous parts; p \ 0.001). Conclusion: This study of the surgical cases confirms the presence of length discrepancy between the anterior and posterior spinal columns in scoliosis that is contributed by both the osseous and non-osseous structures with relatively greater proportion from the non-osseous structures, suggesting an adaptive secondary phenomenon over that of an exaggerated growth phenomenon. Disclosures: Author 1: none; Author 2: grants/research support: AO Spine European Young Researcher Award; Author 3: none; Author 4: stock/shareholder: MRIguidance BV; Author 5: none; Author 6: employee: The Chinese University of Hong Kong; Author 7: none; Author 8: none; Author 9: grants/research support: K2M.

P97 RIB VERTEBRA ANGLE DIFFERENCE (MEHTA’S ANGLE) IS A REPRESENTATION OF ROTATION AND APICAL LORDOSIS Rob Brink, Tom Schlo¨sser, Marijn vanStralen, Koen Vincken, Moyo Kruyt, Winnie Chu, Jack Cheng, Rene´ Castelein Department of Orthopaedic Surgery, University Medical Center Utrecht, Utrecht, The Netherlands

P96 THE RELATIVE CONTRIBUTIONS OF THE OSSEOUS AND NON-OSSEOUS STRUCTURES TO ANTERIOR SPINAL OVERGROWTH IN ADOLESCENT IDIOPATHIC SCOLIOSIS Rob Brink, Tom Schlo¨sser, Koen Vincken, Marijn vanStralen, Moyo Kruyt, Steve Hui, Winnie Chu, Jack Cheng, Rene´ Castelein Department of Orthopaedic Surgery, University Medical Center Utrecht, Utrecht, The Netherlands Introduction: The anterior part of the spine is longer than the posterior part in adolescent idiopathic scoliosis (AIS). This has been described as a potential driver for the onset and progression of the deformity. Purpose of the study: To quantify the length discrepancies in the osseous and non-osseous structures in the anterior as well as the posterior elements. Materials and methods: A consecutive series of prone high-resolution CT scans of 88 AIS patients (Cobb angle 46–109), acquired for navigation purposes, and 30 non-scoliotic controls were included. By

123

Introduction: The rib vertebra angle difference (RVAD), Mehta’s angle, describes apical rib asymmetry and was introduced as a prognostic factor for curve severity in early onset scoliosis, and later applied to other types of scoliosis as well. It is based on a conventional upright coronal radiograph of the spine. Purpose of the study: To establish the relationship between the conventional 2-D RVAD, the 3-D RVAD and the complex 3-D apical morphology in scoliosis. Materials and methods: An existing idiopathic scoliosis database of CT scans, acquired for spinal navigation, was used. 88 high-resolution CT scans (Cobb angle 46–109) were included. A previously validated, semi-automatic image processing technique was used to acquire complete 3-D spinal reconstructions for measurement of the 3-D RVAD, axial rotation and excess anterior length of the apical segments. Additionally, digitally reconstructed coronal radiographs of the CT scans were acquired for measurement of Mehta’s 2-D RVAD and coronal Cobb angle. Results: The 2-D RVAD was on average 24.4 ± 12.5, but in the true coronal view, hardly no asymmetry remained (3-D RVAD: 3.1 ± 12.5; p \ 0.001). 2-D rib asymmetry in the coronal plane of the patient did not correlate with the same RVAD measurements in the 3-D reconstructed coronal plane of the apex (r = 0.145; p = 0.178). The 2-D RVAD appeared to be a manifestation of the 3-D morphology of the apical region of the spine: a larger 2-D RVAD corresponded with increased axial rotation (r = 0.611; p \ 0.001) and apical lordosis (r = 0.452; p \ 0.001).

Eur Spine J (2017) 26 (Suppl 2):S335–S405 Conclusion: The 2-D RVAD is a projection-based reflection of the apical morphology in scoliosis and is a result of the 3-D deformation of the apex. This suggests that the 2-D RVAD could be used as indicator of the severity of the 3-D deformity of the scoliotic spine. Clinical implication should be further explored. Disclosures: Author 1: none; Author 2: grants/research support: AO Spine European Young Researcher Award; Author 3: stock/shareholder: MRIguidance BV; Author 4: none; Author 5: none; Author 6: none; Author 7: none; Author 8: grants/research support: K2M.

S383 Disclosures: Author 1: none; Author 2: none; Author 3: none; Author 4: none; Author 5: none; Author 6: none; Author 7: none.

P99 THE NATURAL HISTORY OF IDIOPATHIC SCOLIOSIS DURING GROWTH: A META-ANALYSIS Francesca Di Felice, Fabio Zaina, Sabrina Donzelli, Stefano Negrini

P98 EFFECT OF DIRECT VERTEBRAL ROTATION IN SINGLE THORACIC ADOLESCENT IDIOPATHIC SCOLIOSIS: BETTER DEFORMITY CORRECTION, MORE ROTATIONAL CORRECTION WITH LIMITED FUSION SEGMENTS Dong-Gune Chang, Jae Hyuk Yang, Jung-Hee Lee, Tae-Jin Kim, Seung-Woo Suh, Jin-Hyok Kim, Se-Il Suk Inje University Sanggye Paik Hospital; Korea University Guro Hospital; Kyung Hee University Hospital; Seoul, Korea Background/introduction: Optimum DVR (direct vertebral rotation) is very important factor for deformity correction, vertebral body rotation, and preservation of motion segments in the treatment of AIS (adolescent idiopathic scoliosis). Inappropriate maneuver during the DVR may result in under or over-correction of the major and compensatory curves. It may aggravate the unfused curve and cause trunk imbalance and decompensation. However, there have been no reports on the effect of DVR regarding the surgical outcomes in the treatment of thoracic AIS. Purpose of the study: To analyze the effect of DVR on radiologic outcomes in the treatment of thoracic AIS following selective thoracic fusion (STF) with pedicle screw instrumentation (PSI). Materials and methods: AIS patients with single thoracic curves (n = 110) treated by STF from neutral vertebra (NV) to NV or NV-1 with a minimum 2-year follow-up were retrospectively analyzed. The patients were divided into two groups; non-DVR (n = 63) and DVR groups (n = 47). Patients in non-DVR group underwent STF with bilateral rod derotation maneuver (RD) while patients in DVR group underwent STF with bilateral RD and DVR maneuver. Results: There was significant difference in the number of fused segments between the non-DVR and DVR groups (P \ 0.000). There was significant difference in the curve magnitude of main thoracic (MT) curve postoperatively (P = 0.001) and at the last follow-up (P = 0.006) between the non-DVR and DVR groups. However, there was no significant difference in proximal thoracic (PT) and lumbar curve postoperatively (PT curve: P = 0.186, lumbar curve: P = 0.155) and at the last follow-up (PT curve: P = 0.250, lumbar curve: P = 0.060) between the two groups. There was significant improvement of LIV tilt and disc angle and relatively well maintained during the follow-up period in both groups. There was no significant difference of rotation of AV and EV preoperatively (P [ 0.05). However, there was significant difference postoperatively (P \ 0.05), and at the last follow-up (P \ 0.05). Conclusions: DVR could effectively achieve better deformity correction, and more rotational correction with reduced number of fusion segments. However, it is important that DVR should be applied in proper direction with adequate force.

ISICO (Italian Scientific Spine Institute), Milan, Italy; Department of Clinical and Experimental Sciences, University of Brescia, Italy; IRCCS Fondazione Don Gnocchi, Milan, Italy Introduction: The real risk of progression of idiopathic scoliosis is considered to vary during growing phases, but so far data concerning this issue refer to a few studies and narrative reviews. Today no metaanalysis are available to pool the results of different studies. Purpose of the study: The aim of the study is to provide a metaanalysis of the current literature concerning the natural history of idiopathic scoliosis during growth. Materials and methods: We searched the MEDLINE, EMBASE and SCOPUS databases up to November 2016 to retrieve articles reporting about natural history of scoliosis during growth. Eligible studies were prospective or retrospective studies that enrolled patients with infantile, juvenile or adolescent idiopathic scoliosis followed up without any treatment from the time of detection. The studies were included only if they reported the progression rates during growth of untreated patients. Two authors independently reviewed each article for data extraction and quality assessment. The main outcome measure was the rate of progression. For the meta-analysis, the studies were grouped according to diagnosis: infantile idiopathic scoliosis (IIS), juvenile idiopathic scoliosis, (JIS) and adolescent idiopathic scoliosis (AIS). Due to expected heterogeneity, we applied a random effect model to pool data together. Results: Of the 1797 citations screened, we assessed 61 full-text articles and included 13 of these (2301 participants). Three studies included IIS patients (347 participants), 5 studies included a mixed population of JIS and AIS (1330 participants), 5 studies included AIS patients only (624 participants) The random pooled estimated progression rate was 49% (95% CI 1–97%) for IIS; 44% in a mixed group of patients affected by juvenile or adolescent IS (95% CI 16–71%) and 42% in AIS (95% CI 11–73%). The criteria to define progression were slightly different among studies, being this a change of 5, 6 or 10 Cobb, or a progression over the threshold of 50 Cobb. Risser, age and clinical features varied among studies. During growth, idiopathic scoliosis tends to progress in a high percentage of cases. The progression rate varies according to the age at diagnosis, with infantile scoliosis being the most unpredictable. There are many confounders like age, Risser sign and baseline Cobb angles that were not consistent among studies, and this makes the data quite heterogeneous as reported in our analysis. These features, together with the different definitions of progression can explain the variability of results among different studies. Conclusion: What is clear from almost all the studies is the risk of progression of the Cobb angle during growth, even if the rate of scoliosis progression is extremely variable among studies. We suggest that future research about natural history looks in a more detailed way at the clinical parameters that can predict progression, and give more homogeneous definition of progression.

123

S384 Disclosures: Author 1: none; Author 2: none; Author 3: none; Author 4: consultant: Medtronic—Janssen Pharmaceutical, stock/shareholder: ISICO.

Eur Spine J (2017) 26 (Suppl 2):S335–S405 r = 0.378) were significantly correlated with T1 to T12 height. Blood oxygen saturation at rest (P \ 0.001, r = 0.642) and at maximum exercise (P = 0.002, r = 0.537) were also significantly correlated with T1–T12 height. Most of the parameters of cardiovascular system, including heart rate, pulse pressure at rest and at maximum exercise, and the increase of them, were significantly correlated T1–T12 height (P \ 0.05, r from 0.361 to 0.472) and thoracic transverse diameter (P \ 0.05, r from 0.454 to 0.620). Conclusion: Overall exercise tolerance did not correlate with the severity of the thoracic cage deformities. However, disorders of the thoracic development, especially retardation of longitudinal growth and loss of spinal height, did influence the function of both respiratory and cardiovascular systems. Disclosures: Author 1: none; Author 2: none; Author 3: none; Author 4: none; Author 5: none; Author 6: none.

Table Correlations Between Radiographic Parameters and Pulmonary Function Test/Cardiopulmonary Function Test in Patients with Congenital Scoliosis

T1 to T12 Height (%)

P100 CARDIOPULMONARY FUNCTION TEST IN PATIENTS WITH CONGENITAL SCOLIOSIS: DO THORACIC CAGE DEFORMITIES AFFECT CARDIOPULMONARY FUNCTION?

Pulmonary Function Test FEV1(% of pred.) FVC(% of pred.) PEF(% of pred.) VC(% of pred.) TLC(% of pred.) RV(% of pred.) RV/TLC(%) Metabolic gas exchange Load(% of pred.) VO2peak(% of pred.) VO2peak (mL/kg/min) AT(%) RER Ventilation VE(% of pred.) Vt(% of pred.) RR(per min) BRr(%) BRt(%) Pulmonary gas exchange SpO2r (%) SpO2t (%) SpO2(%) Vd/Vt Cardiovascular system HR(% of pred.) HR(per min) PPr(mmHg) PPt(mmHg)

Youxi Lin, Xingye Li, Wangshu Yuan, Hui Cong, Haining Tan, Jianxiong Shen Department of Orthopaedic Surgery, Peking Union Medical College Hospital, Beijing, China Background: Deformities of thoracic dimensions, as a result of congenital scoliosis, causes pulmonary dysfunction. However, it is still unknown whether they affect patients’ cardiopulmonary exercise capacity. The aim of this study is to investigate the correlation of chest deformity and exercise tolerance in patients with congenital scoliosis. Methods: 40 patients with congenital scoliosis were included in this prospective study from January 2014 to December 2016. All patients had radiological assessment of spine and rib cage, as well as pulmonary function test and cardiopulmonary cycle ergometer test. 2-tailed Pearson correlation test was performed to investigate the correlation of thoracic cage parameters and pulmonary function and physical capacity. Results: 26 female aged 17.5 years (10–39) and 14 male subjects aged 18.9 years (13–32) were included. All radiographic parameters of thoracic dimension were calibrated by the patient’s own pelvic inlet width. Most of static pulmonary function parameters were significantly correlated with T1–T12 height, difference of hemithorax height and thoracic transverse diameter, respectively, as demonstrated by forced expired volume in one second, forced vital capacity, peak expiratory flow, vital capacity, total lung capacity and residual volume/total lung capacity ratio (P \ 0.05). In cardiopulmonary exercise test, most of the parameters of ventilation, including tidal volume (P \ 0.001, r = 0.647), respiratory rate (P = 0.001, r = -0.532) and breathing reserve both at rest (P = 0.002, r = 0.490) and maximum exercise (P = 0.021,

123

PP(mmHg) O2/pulse(% of pred.)

Le Hemithorax Height

Right Hemithorax Height

(%)

(%)

Difference of Hemithorax Height (%)

Thoracic Transverse Diameter (%)

Thoracic Anteroposterior Diameter (%)

r

P

r

P

r

P

r

P

r

P

r

P

0.651 0.532 0.369 0.596 0.645 0.367 -0.482

<0.001** <0.001** 0.019* 0.001** <0.001** 0.039* 0.005**

0.461 0.296 0.071 0.326 0.463 0.414 -0.287

0.003** 0.064 0.662 0.098 0.008** 0.018* 0.112

0.366 0.265 -0.089 0.142 0.409 0.621 0.108

0.020* 0.098 0.585 0.480 0.020* <0.001** 0.558

-0.370 -0.368 -0.144 -0.442 -0.457 -0.178 0.450

0.019* 0.020* 0.376 0.021* 0.008** 0.329 0.010*

0.477 0.331 0.321 0.469 0.329 -0.067 -0.523

0.002** 0.037* 0.043* 0.014* 0.066 0.714 0.002**

-0.132 -0.231 0.007 -0.069 -0.211 -0.342 -0.048

0.418 0.152 0.964 0.734 0.247 0.055 0.795

0.184 0.116 0.077 0.078 0.111

0.268 0.489 0.646 0.642 0.508

0.221 0.050 0.096 0.039 0.160

0.182 0.766 0.568 0.816 0.336

0.274 0.147 0.127 0.233 0.058

0.096 0.379 0.446 0.159 0.730

-0.127 -0.144 -0.114 -0.124 -0.018

0.448 0.389 0.497 0.460 0.917

0.274 0.037 0.102 -0.201 0.368

0.095 0.826 0.544 0.225 0.023*

-0.261 -0.366 -0.089 -0.442 0.096

0.113 0.024* 0.594 0.006** 0.566

0.284 0.647 -0.532 0.490 0.378

0.084 <0.001** 0.001** 0.002** 0.021*

0.150 0.442 -0.388 0.367 0.085

0.369 0.005** 0.016* 0.024* 0.618

0.249 0.401 -0.266 0.239 0.011

0.132 0.013* 0.106 0.148 0.949

0.036 -0.306 0.429 -0.547 -0.360

0.832 0.062 0.007** <0.001** 0.029*

0.355 0.317 -0.079 0.227 0.079

0.029* 0.053 0.639 0.170 0.642

-0.274 -0.176 0.139 -0.196 -0.068

0.096 0.292 0.405 0.239 0.689

0.642 0.537 -0.330 -0.103

<0.001** 0.002** 0.075 0.539

0.635 0.502 -0.278 0.048

<0.001** 0.005** 0.137 0.775

0.465 0.365 -0.191 0.170

0.007** 0.047* 0.311 0.307

-0.268 -0.308 0.262 0.199

0.139 0.098 0.162 0.231

0.412 0.324 -0.160 -0.048

0.019* 0.080 0.398 0.773

-0.199 -0.289 0.226 -0.099

0.276 0.121 0.230 0.554

0.367 0.385 0.432 0.472 0.361 -0.109

0.023* 0.017* 0.017* 0.009** 0.050 0.516

0.256 0.309 0.310 0.433 0.375 -0.042

0.121 0.059 0.095 0.017* 0.041* 0.804

0.186 0.094 0.169 0.309 0.293 -0.021

0.264 0.573 0.371 0.097 0.116 0.902

-0.163 -0.451 -0.205 -0.266 -0.223 -0.053

0.327 0.004** 0.278 0.156 0.236 0.750

0.454 0.492 0.456 0.620 0.531 -0.038

0.004** 0.002** 0.011* <0.001** 0.003** 0.822

-0.195 -0.117 0.097 0.227 0.229 -0.141

0.240 0.484 0.612 0.228 0.223 0.400

2-tailed Pearson test. *<0.05,**<0.01. Calibrated by the percentile of "pelvic inlet width". FEV1 = forced expiratory volume in 1 second, FVC = forced vital capacity, VC = vital capacity, TLC = total lung capacity, RV = residual volume, RV/TLC = the ratio of residual volume and total lung capacity. VO2peak = peak oxygen intake, AT = anaerobic threshold, RER = respiratory exchange ratio, VE = maximum ventilation volume per minute, Vt = maximum tidal volume, RR = maximum respiratory rate, BR = breath reserve (r= at rest, t=at maximum exercise, same below.), SpO2 = oxygen saturation, O2/pulse = oxygen pulse.

= change, Vd = dead space ventilation, HR = maximum heart rate, PP = pulse pressure,

P101 THE USE OF MAGNETICALLY-CONTROLLED GROWTH RODS (MCGR) AND TRADITIONAL GROWTH RODS (TGR) IN A PEDIATRIC POPULATION Herng Mak, Aine Redmond, Sarah Quidwai, Patrick Carroll, Patrick Kiely Department of Orthopaedics, Our Lady’s Children’s Hospital Crumlin, Dublin, Ireland Introduction: We conducted a single-centre prospective study comparing magnetically controlled growth rods (MCGR) and traditional growth rods (TGR) in a paediatric scoliosis population. The purpose was to evaluate patient experiences and clinical outcomes in both cohorts. Patient experience is the ultimate measure of success in modern medicine. Currently, there is a paucity of data on patient related experiences with MCGR. Methods: The MCGR sample comprised patients who had undergone MAGEC rod insertion from 2014 onwards. The TGR group comprised patients who had undergone insertion of VEPTR, Shilla or Legacy growth rods between 2012 and 2013. We reviewed length of hospital stay, number of surgeries, number of outpatient attendances and surgical complication rates. In addition, we used a questionnaire to determine patient-reported outcomes: the Scoliosis Research

Eur Spine J (2017) 26 (Suppl 2):S335–S405

S385

Society questionnaire (SRS-30). We reviewed radiographs to measure major curve correction using Cobb measurement, thoracic spine height (T1–T12), and T1–-S1 height. A single independent observer measured all radiographs for standardization purposes. Results: We reviewed 47 patients, of whom 23 (48.9%) had MAGEC rods inserted, 7 (14.9%) were converted from TGR to MAGEC rods and 17 (36.2%) had TGR inserted. 28 (59.6%) of subjects in the study were male and 19 (40.4%) were female. A significant proportion (51.06%) were children with complex medical needs. The mean number of open surgical procedures for the MAGEC rod cohort was 1.56, compared with 4.53 for the TGR group and 6.80 for those who converted from TGR to MAGEC rods. The average number of days spent in hospital was 10.69 for patients in the MAGEC rod cohort, 20 days for TGR group and 27.2 for the conversion group, which may reflect the complexity of these patients. Of the MAGEC rod patients, 5 (21.7%) developed wound-related complications and 4 (17.4%) had respiratory complications. In a similar trend, the TGR cohort showed 5 (29.4%) and 2 (11.8%), respectively. 10 (43.5%) MAGEC rod patients did not show any postoperative complications, compared to only 5 (19.4%) in TGR patients. In the MAGEC rod cohort, rod revision surgeries were noted in 4 patients (17.4%), and 9 (52.9%) in the TGR group. Rod failures were also noted, with 2 TGR patients (11.8%) experiencing postoperative rod fracture and breakage. In contrast, no rod failures were observed in the MAGEC rod group. We found that the sample size of the conversion cohort (7) to be too small for analysis of complications. Conclusions: Our study demonstrates that the use of MCGR in paediatric scoliosis patients reduces hospital stay and number of open surgical procedures. MCGR patients were less likely to develop postoperative complications and less likely to experience rod failure, demonstrating its potential to be an efficacious and patient friendly treatment for scoliosis in a paediatric population. Disclosures: Author 1: none; Author 2: none; Author 3: none; Author 4: none; Author 5: none.

plumb line and central sacral vertical line at standing and supine positions were also measured. The dynamic changes D) of these parameters between the standing and supine positions were calculated. From full-length standing lateral radiographs, thoracic kyphosis, lumbar lordosis (LL), pelvic incidence (PI), pelvic tilt, and T1-pelvic angle were measured, and PI–LL was calculated. Age, body mass index, Risser grade, brace treatment, and the SRS-22r scores in self-image and mental health were obtained from the medical charts. From the SRS-22r questionnaire, we defined the patients who answered that they had experienced moderate, moderate to severe, or severe back pain over the last month as patients with moderate or higher back pain; those who answered that they had back pain at rest sometimes, often, or very often as patients with back pain at rest. Multivariate logistic regression analysis was performed to identify radiographic and psychological risk factors for moderate or higher back pain and back pain at rest. Results: The prevalence of moderate or higher back pain was 18.8% (13/69 patients); back pain at rest, 24.6% (17/69 patients). Overall, the mean Cobb angle at the standing position were 17.5 (0–43.7) in PT curve, 37.0 (6.9–74.7) in MT curve, and 31.3 (3.9–63.5) in TL/L curve, respectively. Multivariate analysis revealed that the risk factor for moderate or higher back pain was low SRS-22r score in self-image [odds ratio (OR), 0.21; 95% confidence interval (CI), 0.06–0.66; p = 0.008]. The risk factors for back pain at rest were large DAVT at MT curve (OR 1.24; 95% CI 1.09–1.40; p = 0.001) and low SRS-22r score in mental health (OR 0.43; 95% CI 0.20–0.91; p = 0.03). Conclusions: Our study revealed that large change of AVT at MT curve was the risk factor for back pain at rest, and this suggested that dynamic changes of curve in patients with AIS related to back pain. Furthermore, psychological factors could play an important role to induce moderate or higher back pain and back pain at rest. Disclosures: Author 1: none; Author 2: none; Author 3: none; Author 4: consultant: Kyocera Medical, B-Braun, Medacta, Asahi-Kasei, Robert-reid; Author 5: none.

P102 RADIOGRAPHIC AND PSYCHOLOGICAL RISK FACTORS FOR BACK PAIN IN PATIENTS WITH ADOLESCENT IDIOPATHIC SCOLIOSIS WITHOUT CORRECTIVE SURGERY

P103 CONTINUOUS LOW BONE MINERAL DENSITY MAY ASSOCIATES WITH SEVERITY OF ADOLESCENT IDIOPATHIC SCOLIOSIS-LONGITUDINAL STUDY

Takahiro Makino, Yusuke Sakai, Shota Takenaka, Takashi Kaito, Hideki Yoshikawa

Mitsuhiro Nishida, Nobuyuki Fujita, Mitsuru Yagi, Naobumi Hosogane, Takeshi Fujii, Osahiko Tsuji, Narihito Nagoshi, Masaya Nakamura, Morio Matsumoto, Kota Watanabe

Department of Orthopaedic Surgery, Osaka University Graduate School of Medicine, Suita, Japan

Department of Orthopaedic Surgery, Tokyo Saiseikai Central Hospital, Tokyo, Japan

Purpose: The prevalence of back pain is not low among the patients with adolescent idiopathic scoliosis (AIS); however, the difference in radiographic parameters or patient backgrounds between patients with AIS who experienced back pain and those who do not remains unclear. The purpose of this study was to identify the radiographic and psychological risk factors of back pain in patients with AIS without corrective surgery, especially focused on dynamic changes of radiographic parameters. Methods: 69 consecutive female AIS patients [mean age, 14.0 years (10–18)] who had initially visited our outpatient clinic since July 2013 were included. From full-length standing and supine posteroanterior radiographs, Cobb angles, apical vertebral translation (AVT), and apical vertebral rotation measured by Cerny’s method were measured at proximal thoracic (PT), main thoracic (MT), and thoracolumbar/lumbar (TL/L) curves. The distance between C7

Introduction: The association of low bone mineral density (BMD) with occurrence or progression of adolescent idiopathic scoliosis (AIS) have been reported. However, few reports have evaluated the effect of low BMD on AIS in longitudinal study. The purpose of this study was to analyze factors related with low BMD in AIS patients based on the BMD data followed until bone maturity. Materials and methods: 61 AIS patients (55 females, 6 males) who underwent posterior correction and fusion surgery before the age of 20 years with Risser grade 4 or below were included. All of them were followed for more than 2 years after surgery and have reached over 20 years of age, and Risser grade 5. We measured BMDs of proximal femurs before surgery and at final follow-up. The low BMD was defined as less than mean-1SD. 24 patients had low BMD before surgery (L group 0.82 ± 0.06 g/cm2) and 37 patients had normal BMD before surgery (N group 1.02 ± 0.08 g/cm2). The mean

123

S386 preoperative age was 16.6 ± 1.9 years, and the mean follow-up period was 4.9 ± 1.7 years. We evaluated the factors that might be potentially related to low BMD. Results: 14/24 patients (58.3%) in the L group showed continuous low BMDs at final follow-up (L–L group, preoperative 0.79 ± 0.05 g/cm2 and final follow-up 0.78 ± 0.05 g/cm2). No patient presented with low BMD at the final follow-up in the N group. In the L group, the mean preoperative Cobb angle of 67.8 ± 11.2 in the L–L group was significantly larger than that in the normal BMD patients (L–N group) at final follow-up (55.6 ± 11.8) (p = 0.02). The mean preoperative BMI in the N group (19.8 ± 1.9 kg/m2) was significantly higher than that in the L group (18.0 ± 2.0 kg/m2). However, the mean preoperative BMI and age at surgery were not significantly different between the L–N group (18.7 ± 2.5 kg/m2, 16.7 ± 1.9 years) and the L–L group (17.5 ± 1.5 kg/m2, 16.8 ± 1.4 years). Conclusions: The patients with continuous low BMDs at final followup (L–L group) had significantly larger preoperative Cobb angle. Therefore, the low BMD might have association with the severity of scoliosis. Treatment to increase BMD may be potentially effective for patients with low BMDs to attenuate progression of scoliosis. 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: Kyocera, Japan.

P104 SCOLIOSIS CONVEXITY IS RELATED TO ORGAN ANATOMY Tom Schlo¨sser, Tom Semple, Siobha´n Carr, Simon Padley, Michael Loebinger, Claire Hogg, Rene´ Castelein University Medical Center Utrecht, Utrecht, The Netherlands Objectives: The aim of this cohort study was to explore the causeeffect-relationship between organ anatomy and curve convexity by studying the prevalence and convexity of idiopathic scoliosis in primary ciliary dyskinesia (PCD) patients with and without situs inversus. PCD is a respiratory syndrome characterized by defects in the embryonic nodal cilia and ‘random’ organ orientation; approximately 46% of patients have situs inversus. Information on curve convexity in PCD patient with situs inversus versus normal anatomy could increase our understanding of factors that are involved in the pathogenesis of scoliosis. Methods: Chest radiographs of all PCD patients that were part of the national annual PCD screening in the UK were systematically screened for existence of significant lateral spinal deviation using the Cobb angle. Positive values represented right-sided convexity. Curve convexity and Cobb angles were compared between PCD patients with situs inversus and normal anatomy. Results: A total of 196 PCD patients were adequately screened. A total of 108 (55%) subjects had normal organ anatomy (levocardia), eighty-four (42%) had situs inversus and four (2%) dextrocardia. The prevalence of scoliosis (Cobb [10 degrees) and significant spinal asymmetry (Cobb 5–10) was 8 and 20%, respectively. Curve convexity and Cobb angle were significantly different within both groups between situs inversus patients and patients with normal anatomy (P B 0.009). Moreover, curve convexity correlated significantly with organ orientation (P \ 0.001; Phi = 0.882): In all, except for one case, 16 PCD patients with scoliosis (8 situs inversus and 8 normal anatomy cases with a Cobb angle larger than 10) matching of curve convexity and orientation of organ anatomy was observed. Conclusions: This study confirms the hypothesis on the correlation between organ anatomy and curve convexity in scoliosis. The convexity of the primary curve is predominantly to the right in patients

123

Eur Spine J (2017) 26 (Suppl 2):S335–S405 with normal organ anatomy and to the left in patients with situs inversus. Disclosures: Author 1: grants/research support: AO Spine Europe Young Researcher Award; Author 2: other financial report: Speaker fees from Vertex; Author 3: grants/research support: Cystic Fibrosis Trust, NIHR, other financial report: Teva, Vertex, PTC, Cheisi, Pharmaxis; Author 4: none; Author 5: consultant: Bayer; Author 6: none; Author 7: grants/research support: K2M.

P105 POSTERIOR SCOLIOSIS CORRECTION DOES NOT INFLUENCE TONSILLAR LOCATION IN CHIARI I MALFORMATION WITH SYRINGOMYELIA Zezhang Zhu, Weiguo Zhu, Zhen Liu, Yong Qiu Spine Surgery, Drum Tower Hospital of Nanjing University Medical School, Nanjing, China Background/introduction: Concerning the increased risk of iatrogenic neurologic deterioration during deformity correction, patients with Chiari I malformation (CMI) or syringomyelia (SM) and typical neurological dysfunction are advocated posterior fossa decompression surgery by most surgeons as an initial step. However, for scoliosis patients with only CMI and minimal neurological defects, the management strategy is controversial. Purpose of the study: We report a retrospective series of patients with scoliosis secondary to CMI in the absence of SM to investigate the safety and effectiveness of one-stage posterior fusion and to explore the influence of correction surgery upon the natural evolution of cerebellar tonsil in these patients. Materials and methods: A total of 14 patients meeting the inclusion criteria at our spine center from 2010 to 2015 were enrolled in this study. One-stage posterior correction and fusion with segmental pedicle screw instrumentation was performed for all patients without additional neurosurgical intervention for their Chiari malformation. Patients’ medical records (age, gender, length of follow-up, neurologic signs and symptoms), intraoperative neuromonitoring (SEP and MEP), MRIs measurement (cerebellar tonsillar position) and radiographic studies were reviewed. Results: The mean age of the 14 patients at the time of surgery was 18.7 ± 11.2 years. The primary curve magnitude averaged 61.71 ± 20.93 and the cerebellar tonsillar position averaged 8.8 ± 2.6 mm (3 grade II and 11 grade I) below foramen magnum preoperatively. Throughout operation, no neuromonitoring difficulties (lack of baseline or loss of signal) were observed. Immediately after operation, the mean percent correction was 67.7 ± 11.5% of the major curve and the mean value of tonsillar position was 6.3 ± 2.7 mm (range 2.1–11.6 mm) below foramen magnum, with cerebellar tonsil upward shifting was observed in 9 patients (64.3%), whereas downward shifting in 5 patients (35.7%). During follow-up, the deformity correction was maintained satisfactorily and neural physical examination of each patient was not worse than the preoperative situation. In addition, no complications of deep-wound infection, pseudarthrosis, additional surgery, implant failure or neurologic deficits in this group were found. Conclusion: One-stage spinal fusion with instrumentation is a safe and effective procedure for treatment of scoliosis associated with only CMI in the setting of minimal neurological dysfunction, with satisfactory correction on coronal and sagittal planes without neurologic problems. Immediately after surgery, more than half (64.3%) of cerebellar tonsils shifted upwardly, which indicated that correction surgery did not exert a deleterious effect upon the behavior of cerebellar tonsils. Disclosures: Author 2: none; Author 3: none; Author 4: none. NEW TECHNIQUES

Eur Spine J (2017) 26 (Suppl 2):S335–S405 P106 REDUCTION IN FLUOROSCOPY TIME AND RADIATION DOSAGE USING AN INNOVATIVE GUIDE WIRE FOR PERCUTANEOUS PEDICLE SCREWS

S387 P107 NUANCES IN LUMBAR INTERBODY FUSION. REVIEW OF OUTCOME AND COMPLICATIONS ASSOCIATED WITH OBLIQUE LUMBAR INTERBODY FUSION (OLIF) AND EXTREME LATERAL INTERBODY FUSION (XLIF)

Brandon Cook, David Briski, Joseph Zavatsky Ochsner Medical Center, New Orlean, USA Introduction: Despite the benefits of minimally invasive spine surgery (MIS), inherent risks to the patient and surgeon exist. Numerous studies demonstrate increased radiation exposure with MIS secondary to the need for increased fluoroscopy, thereby increasing the risk of cataracts and malignancy. Additionally, inadvertent advancement of standard guide wires through the vertebral body can occur, placing vital ventral structures at risk. This study evaluates the benefit of utilizing a novel split-tip guide wire for percutaneous pedicle screw placement. Methods: Thirty consecutive cases of MIS transforaminal lumbar interbody fusion (TLIF) at L5-S1 were retrospectively evaluated. Group 1: standard straight guide wire, 15 patients; group 2: split-tip guide wire, 15 patients. Except for the type of guide wire utilized, the same operative technique was used in each case, including bicortical S1 screw fixation. Total fluoroscopy time, radiation dosage, operative time and complications were evaluated. Results: Mean total fluoroscopy time per case for group 1 was 231.1 vs. 154.2 s for group 2. Mean radiation dosage for group 1 was 16.22 vs. 8.69 rads in group 2. There was no significant difference in operative time. Inadvertent advancement of two S1 guide wires occurred in two different patients in group 1. Postoperative abdominal CT scans with contrast were negative. Conclusion: Utilizing a split-tip guide wire for percutaneous pedicle screw placement significantly decreased fluoroscopy time by 33% and radiation dosage by 46%. Cannulating the sacral promontory allows for bicortical S1 screw purchase, but removes the mechanical stop preventing inadvertent guide wire advancement. The split-tip guide wire may prevent wire advancement and decreases the need for fluoroscopic surveillance. This, in turn, reduces the risk of injury to structures ventral to the spine, while reducing the exposure to harmful radiation. Disclosures: Author 1: none; Author 2: none; Author 3: grants/research support: AO Spine, consultant: Depuy Synthes, Zimmer Biomet, Stryker, stock/shareholder: Surgical Solutions, Vivex, royalties: Zimmer Biomet.

Maria Grazia Di Benedetto, Sujay Deeherendra, Simon Clark, Marco Teli Department of Neurosurgery, Bari, Italy; Department of Neurosurgery, The Walton Centre, Liverpool, UK Background: Lumbar interbody fusion has been accepted as an effective procedure in the management of spinal conditions such as degenerative disease, spondylolisthesis, disc herniation and deformities. Over the past decade the trend in lumbar spinal fusion surgery has progressively evolved and recently minimally invasive lateral/ anterolateral retroperitoneal approach for anterior lumbar interbody fusion have gained popularity providing another corridor to access the lumbar spine with several advantages over conventional open anterior and posterior/transforaminal lumbar interbody fusion. Purpose: This study investigated advantages, pitfalls and complications of minimally invasive approach for anterior lumbar interbody fusion by using a lateral/anterolateral retroperitoneal approaches including OLIF and XILIF/DLIF Methods: A comprehensive review of available literature has been performed. The keywords ‘‘XLIF’’, ‘‘extreme lateral approach lumbar spine’’, ‘‘lateral approach lumbar interbody fusion’’, ‘‘OLIF’’, ‘‘oblique lumbar interbody fusion’’ were used to search for relevant articles. The literature were analyzed for radiographic and clinical outcomes (neurological outcomes, fusion rate, misalignment restoration and rate of adverse events). Results: OLIF and XILIF/DLIF are relatively new minimally invasive techniques that allow to access the intervertebral space from a unilateral approach anterior to or through the psoas muscle. These procedures minimize vascular and visceral injury related with anterior approach and completely preserve the posterior musculature. Successful fusion rate, significant pain scores improvement, good sagittal and coronal deformity correction have been reported. Operative time and blood loss are decreased with an early patient mobilization and reduced length of stay in hospital. XILIF/DLIF approach offer an easy access to lumbar multiple levels except for L5–S1 due to obstruction by the iliac crest at that level. The transpsoas retroperitoneal corridor is used to access the disc space with potential risks of lumbar plexus. The majority of the studies report a transient postoperative anterior thigh symptoms including pain, numbness and paresthesias as a result of lumbar plexus injury or indirect compessive neuropathy. The OLIF technique is suitable for levels L1–S1 and preserves psoas muscles as the approach is anterior to it. However, potential risks with OLIF surgery are vascular and indirect lumbar plexus injury. Conclusion: Based on the existing literature, lateral approaches to anterior lumbar spine including OLIF and XILIF/DLIF are promising and feasible minimally invasive surgical method for treatment of a range of spinal degenerative disorders requiring lumbar spine fusion. However, further clinical studies and long term outcomes are needed to assess the reliability of this technique, in addition further data and biomechanical studies are required to establish their effectiveness of restoring the sagittal misalignment Disclosures: Author 1: none; Author 3: grants/research support: Medtronic, consultant: Medtronic; Author 4: grants/research support: medtronic, consultant: Medtronic.

123

S388

Eur Spine J (2017) 26 (Suppl 2):S335–S405 significantly increased postoperatively with insignificant loss of correction at 1 year follow up. Both middle and posterior parts of disc height significantly restored by mean 37% while mean loss of anterior, middle and posterior disc height at 1-year follow-up was 0.9, 0.5, and 0.6 mm for each parameter, respectively. No cage subsidence or migration occurred. No postoperative infection occurred. Radiographically a 100% fusion rate with no screw loosening was found. Conclusions: TLIF using modular cages demonstrated no incidence of cage subsidence or migration, high fusion rate, and no screw loosening. Also, it was effective in restoring lumbar lordosis angle, segmental disc angle and disc height and maintaining this correction, which is attributed to the larger footprint of the modular cage. The modular TLIF-cage seems to be a safe method for interbody fusion in patients with risk of subsidence. Future studies should investigate prospectively the clinical and radiological outcome of the modular cage compared to traditional TLIF cages. Disclosures: Author 1: none; Author 2: none; Author 3: other financial report: DePuy Synthes, Medtronic.

P108 CLINICAL AND RADIOLOGICAL EVALUATION OF TLIF FOR DEGENERATIVE DISC DISEASES USING A NOVEL MODULAR CAGE Mohamed Elmekaty, Emad El Mehy, Yohan Robinson 1

Department of Surgical Sciences, Uppsala University Hospital, Uppsala, Sweden; 2Orthopedic Surgery Department, Tanta University, Tanta, Egypt Abstract: Study design: Retrospective cohort study. Introduction: Recently, there is an acceleration in the development and utilization of Minimally Invasive Surgical (MIS) techniques for lumbar spine fusion. Although several MIS techniques have been reported, it was challenging to insert a TLIF cage with large footprint to enhance the union rate and minimizing cage subsidence. The fundament of the novel modular cage is to provide large surface area for fusion with a limitation of bone and soft tissue damage and dural manipulation which is usually associated with other conventional TLIF cages. The modular cage consists of an integral rail and slot multi-segmental system inserted through unilateral foraminotomy and assembled within the disc space. Objective: to assess clinical and radiographic outcomes of a modular TLIF cage for degenerative disc diseases. Methods: Patients treated with TLIF using a novel modular cage between 2013 and 2016 analysed retrospectively with a 1-year minimum follow-up period. Changes in lumbar lordosis angle and segmental disc angle were measured using digitized radiograph analysis. Also, changes of anterior, middle and posterior disc height were analyzed preoperatively, postoperatively and at one year follow-up by using lateral plain X-ray in a neutral position. Cage subsidence, fusion rate, screw loosening, complications and % of endplate coverage were also assessed. Subsidence is considered if more than 3 mm of endplate settling occurred. Results: 20 patients were included in the study (age 66 ± 10 years, 65% female, BMI 28 ± 5 kg/m2). In 14 cases degenerative spondylosis and in 6 cases degenerative scoliosis was treated. A total of 37 lumbar levels was treated with the modular TLIF cage. Endplate coverage of the TLIF cage surface was 65 and 61% along antero-posterior and transverse diameters, respectively. Lumbar lordosis angle and segmental disc angle

123

P109 SURGICAL APPROACH CORRIDOR OF MINIMALLY INVASIVE ANTEROLATERAL OBLIQUE LUMBAR INTERBODY FUSION FOR L5–S1 Hae-Dong Jang, Jae Chul Lee, Sung-Woo Choi, Byung-Joon Shin Department of Orthopaedic Surgery, Soonchunhyang University Hospital, Seoul, Republic of Korea Background/introduction: In the minimally invasive anterolateral oblique lumbar interbody fusion (OLIF) L5–S1 widely performed in recent, the success of the surgical procedure depends on skillful vascular access without the complication [in particular, common iliac vein (CIV)]. Since most L5–S1 anterior approaches usually use the central corridor (medial side of the CIV), excessively medialized vessel course makes the surgery difficult and even impossible to perform. However, when considering the anatomical characteristics of left CIV (more medially located and can be easily damaged compared to arteries) and various course of the vessels, it is reasonable to decide approach side based on the analysis of the vascular window (VW) in preoperative magnetic resonance image. Purpose of the study: To report the surgical technique for the approach corridor which helps to prevent vascular injury and obtain sufficient surgical field of view in OLIF L5–S1. Materials and methods: We included 16 patients (4 men and 12 women) who underwent OLIF L5-S1 for the degenerative lumbar disease in our spine center from July 2015. As for the diagnosis, there were five patients with adjacent segment disease, six with degenerative scoliosis, three with lumbar degenerative kyphosis, and two with multilevel spinal stenosis. According to the standard OLIF technique, left anterolateral skin incision and retroperitoneal approach was performed on all patients. We defined central VW as the length between two medial borders of the bilateral CIV, and lateral VW as the length from the lateral border of the left CIV to the medial border of the left psoas muscle at L5–S1 level. Actual vascular access side (central or lateral) was decided by the length of VW preoperatively. Results: Mean central VW was 29.7 mm and the lateral was 13.9 mm, and there was significant difference (P \ 0.001). Four patients had extremely narrow central VW because of the left CIV that is located with transverse direction and cross the midline. Actual surgical approaches using central VW were nine cases, and lateral were 7. In two cases, we asked vascular surgeon’s help to dissect the vessel. In nine cases, we ligated branch vessels; middle sacral vein in four patients, ascending lumbar vein in 3, and iliolumbar vein in 2.

Eur Spine J (2017) 26 (Suppl 2):S335–S405 There were two vascular injuries; one traction injury occurred during the soft tissue dissection around CIV for OLIF L5–S1 and one laceration injury of CIV caused by trapping between the retractor and osteophyte while approaching L4–5 for combined OLIF surgery. In both cases, an experienced vascular surgeon repaired them immediately, and there was no remained complication. Conclusion: Since anatomical location and course of the left CIV was observed in various patterns, VW was determined based on individual’s vascular pattern. In OLIF L5–S1, we considered it is a technically critical point to decide approach corridor (lateral or medial to the CIV) through the comprehensive vascular assessment using preoperative MRI. Disclosures: Author 1: none; Author 2: none; Author 3: none; Author 4: none.

P110 COMPARISON OF INTERLAMINAR STABILIZATION WITH DECOMPRESSION VS FUSION WITH DECOMPRESSION IN PATIENTS REQUIRING 2 LEVELS OF SURGICAL TREATMENT FOR SPINAL STENOSIS

S389 patients from 21 sites in the USA between 2006 and 2010. It is a prospective, randomized, controlled trial conducted of patients diagnosed with spinal stenosis with up to a grade 1 spondylolisthesis. Patient sample: Of the 322 patients enrolled, 116 patients required surgical treatment at 2-levels. Patients were randomized to treatment groups in a 2:1 ratio. The ILS group consisted of 77 patients and the fusion group consisted of 39 patients. Outcome measures: Efficacy was measured using a Composite Clinical Success (CCS). Patients achieve CCS if they achieve all four of the following outcomes: C15 point improvement from baseline Oswestry Disability Index (ODI), no reoperation or epidural injections, no persistent, new, or increasing neurological deficits, and no major device-related complications. Methods: Data was collected pre-operation, and then at 6 weeks, 3, 6, 12, 18, 24, 36, 48, and 60 months post-operatively. The results of the four assessments were combined to calculate the percentage of patients achieving CCS. Results: There was a difference trending toward significance between groups for the absence of reoperation or epidural injection, with 68.8% of ILS patients and only 51.3% of fusion patients meeting this criteria (p = 0.065). 13% of ILS patients and 25.7% of fusion patients had a reoperation (p = 0.088). These components influenced the total percentage of patients achieving CCS to be much greater in the ILS group than the fusion group, with 55.1% (38/69) of ILS patients and only 36.4% (12/33) of fusion patients achieving CCS at month 60 (p = 0.077). There was no significant difference in ODI (p = 0.532), or absence of persistent, new, or increasing neurological deficits (p = 0.939). Each group had 2 patients experience major device-related complications (p = 0.480). With regard to ODI, VAS back and worse leg pain, SF-12, and ZCQ, both groups had significantly better results at every follow-up time point when compared to their respective baseline scores. Conclusions: The 2-level ILS patient group performed as well, if not better, than the 2-level fusion group, demonstrating both clinical outcome success and favorably low reoperation rates in patients who received 2-level ILS surgery. Disclosures: Author 1: none; Author 2: none; Author 3: none; Author 4: grants/research support: Bacterin, Integra, NuTech, Vertical, Spinal Kinetics, Centinel, Mallinckrodt, consultant: 4Web Medical, Alphatec Spine, Inc., Healthpoint Capital Partners, LP., IVY Healthcare Partners, LP., Nuvasive, Inc., Paradigm Spine, LLC., Spinal Kinetics, Spinal Partners III, Vertebral Technologies, Inc., Vertical Spine, Viscogliosi Brothers, LLC, stock/shareholder: Alphatec Spine, Inc., BI Members, LLC, BioAssets Development Corp., Healthpoint Capital Partners, LP., Ivy Healthcare Partners, LP., Nuvasive, Inc., Orthopaedic Investment Partners, LP., Paradigm Spine, LLC., Promethean Surgical Devices, Scient’x USA, Spinal Kinetics, Vertebral Technologies, Inc., Viscogliosi Brothers, LLC., Vertical Spine, royalties: Nuvasive, Inc, other financial report: Bonovo Orthopedics, Inc., Liventa BioSciences, Inc., MMF Systems, Inc., Royer Biomedical, Inc., Woven Orthopedic Technologies; Author 5: consultant: Paradigm Spine.

Rachel Simon, Christina Dowe, Samuel Grinberg, Frank Cammisa, Celeste Abjornson Hospital for Special Surgery, New York, NY, USA Background: Interlaminar stabilization (ILS) with decompression has been introduced as an alternative method of surgical treatment of lumbar spinal stenosis. Many studies examine the efficacy of ILS devices in single level procedures, but fewer studies focus on their efficacy in two level procedures. Purpose: This study compares microsurgical decompression with instrumented posterolateral fusion to microsurgical decompression with ILS, using coflex (Paradigm Spine, NY, NY), in patients requiring surgical treatment at two levels for lumbar spinal stenosis. Study design/setting: This patient cohort was part of the Investigational Device Exemption (IDE) study which enrolled a total of 322

P111 CARBON-FIBER ENHANCED PEEK (CF/PEEK) SCREWS IN SPINAL TUMOR SURGERY Olaf Suess, Bjoern Kuehn, Sven Mularski Spine and Neurotrauma Center, DRK Kliniken Berlin Westend, Berlin, Germany Introduction: Spinal instrumentation with metal implants may cause disturbing artifacts in post-surgical imaging, making the evaluation of the implant positioning, bony decompression of the intraspinal neural structures and the evaluation of tumor progression in destructive spinal

123

S390 tumor cases difficult. This retrospective analysis was performed to evaluate a newly developed carbon-fiber enhanced polyetheretherketone (cf/PEEK) screw- and rod-system for transpedicular osteosyntheses that is expected to compensate for these drawbacks. Materials and methods: A novel cf/PEEK-screw system with titanium-coated threads (PSI 2.0 system; icotec, Altsta¨tten, Switzerland) was used in a series of 14 thoracic and 13 lumbar spinal metastases. Pathologic fractures and epidural tumor invasion was located between T4 and L5. After microsurgical decompression surgery with epidural tumor excision, a dorsal instrumentation including 1–2 levels above and below the target level was implanted. Postoperatively, radiotherapy was initiated. CT and MRI scans, as well as prospective clinical data were collected and analyzed over a period of up to 24 months. Results: 168 screws were implanted in 19 cases of spinal lung, breast, colon and prostate cancer metastases + 8 cases of plasmozytoma (total n = 27 cases). Intra-operative handling of the cf/PEEK screws (5.5 and 6.5 9 35–50 mm) was analogous to full-titanium implants due to similar screw design/geometry and surgical implantation technique. Follow-up time was 4–25 months. Post-operative CT scans demonstrated sharp delineation at the screw thread- and bone-interface with radiological signs of screw loosening in 5.4% (8/168 screws in 5/27 cases; with need for surgical revision in 1 case). Complication rates: deep wound infect: 1/27; screw breakage: 0%. Irradiation foci and dose distribution were easy and precise to specify, in contrast to the planning with usual artifactdistorted scans (irradiation dose: 30–36 Gy). Conclusion: cf/PEEK pedicle screws showed to decrease artifacts in post-surgical imaging with consequent improved assessment of anatomical structures such as neuroforamen, dural sac and bony structures around the screws. Decreased artifacts helped to improve isodose calculation in post-surgical radiation therapy, allowing focused therapy planning and surveillance. In addition, an E-module close to that of cortical bone, as well as the ultra-thin titanium coating of the screw threads may be reason for good osseointegration. Randomized multicenter studies should give further information about material behaviour and complication rates, as well as long-term clinical results. Disclosures: Author 1: none; Author 2: none; Author 3: none.

Eur Spine J (2017) 26 (Suppl 2):S335–S405 only approach, with a 2-RC across the osteotomy sites. Whereas, correction rate and junction problems remain as the major concerns. Purpose of the study: To compare the radiographic results and clinical outcomes with use of a standard 2-rod construct (2-RC) versus a complex of 2-RC plus bilateral satellite rods (S-RC) across Ponte osteotomy levels in a matched-cohort cases with Scheuermann kyphosis (SK) with a minimum 2-year-follow-up. Materials and methods: Between 2012 and 2014, a consecutive series of 22 patients with SK had undergone posterior correction with S-RC across multilevel Ponte osteotomies by a single surgical team. Twenty-two control patients with 2-RC were identified at the same period, being matching for age, fusion levels, and kyphosis magnitude. Comparisons were made in terms of deformity, correction results, complications, junctional alignment and SRS-22 scores between the two groups. Results: No significant difference was found in age, gender, followup period, blood loss, osteotomy levels, or SRS-22 scores. Group S-RC tended to have longer operation time but without statistical significant difference (257.3 vs. 239.6 min, P = 0.20). Both groups had similar kyphosis angle before surgery, but group S-RC had higher correction rate 55.4 vs. 46.2%, P \ 0.01) and less correction loss (1.0 vs 2.4, P \ 0.01) than group 2-RC during the follow-up. Group S-RC had significantly greater improvement in domains of management satisfaction and pain. No cases were detected with pseudarthrosis or implant failure in either group. There were 2 cases with wound problems in each group, but proximal junctional kyphosis was more frequently seen in group 2-RC (8 of 22) than group S-RC (1 of 22). Conclusion: As a safe and simple method, use of S-RC is effective in providing increased correction force across multiple Ponte osteotomy levels as well as preventing proximal junctional kyphosis. Disclosures: Author 1: none; Author 2: none; Author 3: none; Author 4: none; Author 5: none.

P113 SAFETY AND EFFICACY OF CONCURRENT SPINAL DEFORMITY AND INTRA-SPINAL SURGERY: CASE SERIES Caglar Yilgor, Nuray Sogunmez, Mohamed Dalla, Gu¨lden Demirci Otluoglu, Memet Ozek, Ahmet Alanay School of Medicine, Acibadem University, Istanbul, Turkey

P112 ADDING SATELLITE RODS TO STANDARD 2-ROD CONSTRUCT: AN EFFECTIVE TECHNIQUE IN IMPROVING SURGICAL OUTCOMES AND PREVENTING PROXIMAL JUNCTIONAL KYPHOSIS FOR POSTERIOR CORRECTION OF SCHEUERMANN KYPHOSIS Xu Sun, Xi Chen, Bin Wang, Zezhang Zhu, Yong Qiu Spine Surgery, Drum Tower Hospital of Nanjing University Medical School, Nanjing, China Background/introduction: Multiple Ponte osteotomies are frequently employed to correct Scheuermann kyphosis via a posterior-

123

Background: Traditionally, coexisting spinal and intra- spinal pathologies are treated with staged neurosurgical and orthopedic interventions. Today, the recent advances in multimodality intraoperative neurophysiologic monitoring (IONM) techniques have allowed for concurrent surgeries. Potential advantages of concurrent surgery are: increased surgeon comfort via a single exposure, and increased patient comfort via single anesthetic exposure and hospitalization as well as decreased cost. Purpose: The aim of this is study was to evaluate the safety and efficacy of concurrent spinal and intra-spinal surgeries. Methods: A descriptive analysis was done on prospectively collected single center data. Neurosurgical diagnoses were 5 tethered cord, 3 diastematomyelia, 2 lipomyelomeningocele and 9 seconder tethered cord cases that were previously operated for diastematomyelia or meningomyelocele. Orthopedic diagnoses were ten congenital scoliosis three of which were previously operated, five neuromuscular, three spondylolistheses and one AIS. Neurosurgical procedures included 4 bone spur excisions, 13 detetherings and 2 lipomyelomeningocele repairs with duraplasty. Orthopedic

Eur Spine J (2017) 26 (Suppl 2):S335–S405

S391

procedures included 11 posterior instrumented fusions, 4 hemivertebrectomies, 3 magnetically-controlled growing rods and 1 vertebral colon resection. Results: 9 (47%) patients had pre-op neurological deficits. Mean pre-op coronal Cobb was 55.6 degrees (34–137). After the completion of intraspinal procedures, a mean of 7.4 levels (2–16) were fused. Mean surgical time was 426 (240–700) minutes. Mean EBL was 654 (200–1300) ml. Mean post-op coronal Cobb was 25.8 (7–56). Mean day to discharge was 5.8 (3–8) days. 3 patients (15.7%) had IONM loss that returned back to baseline during surgery. 2 of them had motor radiculopathy that were resolved within 6 months. 2 patients (10.5%) had CSF leakage. One of them had a superficial wound infection and had a revision surgery for debridement. One patient had motor radiculopathy after spondylolisthesis reduction that was revised the same day and reduction percentage was decreased. One patient had a superior mesenteric artery syndrome. A total of 6 patients (31.6%) experienced complications that are resolved without sequela as of the day of this report. Conclusion: Concurrent orthopedic and neurosurgery for pediatric deformity patients with intra- spinal pathology may be safe and efficient without significant long-term sequela. Disclosures: Author 1: none; Author 2: grants/research support: DePuy Synthes; Author 3: none; Author 4: none; Author 5: none; Author 6: grants/research support: Depuy Synthes, consultant: Medtronic.

tolerate the standing position for upright MRI scanning. Patients scanned in the Standing position had reliably measureable and significantly lower cross sectional area (p \ 0.01), disc height (p \ 0.001), and foramen diameter (p \ 0.001) than on supine imaging. The magnitude of volumetric change was significantly greater in grade II than in grade I slips (p \ 0.0001). Absolute slip percentage changes were not observed in grade I DLS (n = 5, p = 0.0638), but were noted in grade II DLS (n = 5, p \ 0.0001). The magnitude of change in disc angle was the same for both grade I and II DLS. Conclusion: In this pilot study, Standing MRI scanning reliably detects dynamic morphometric differences in a number of clinically important radiographic parameters in patients with DLS. There were also observed differences between grade I and grade II DLS absolute slip changes. This study provides further evidence for the belief that many patients may be appropriately surgically treated without a fusion. Disclosures: Author 1: none; Author 2: none; Author 3: grants/research support: Paramed Medical Systems; Author 4: none; Author 5: grants/research support: OREF, MEDTRONIC.

DIAGNOSTICS AND IMAGING

Jun Jiang, Bangping Qian, Yong Qiu, Bin Wang, Yang Yu, Zezhang Zhu

P114 DYNAMIC CHANGES IN DEGENERATIVE LUMBAR SPONDYLOLISTHESIS: A PILOT STUDY OF STANDING MAGNETIC RESONANCE IMAGING

Spine Surgery, Drum Tower Hospital of Nanjing University Medical School, Nanjing, China

Raphaele Charest-Morin, Michael Bond, David Wilson, Honglin Zhang, John Street Combined Neurosurgical and Orthopedic Spine Program, University of British Columbia, Vancouver, Canada Background: Degenerative lumbar spondylisthesis (DLS) presents with a wide spectrum of discrete anatomical pathologies that can lead to slips ranging from dynamic movement between vertebrae to static stable slips. Patients with L4/5 DLS typically have leg symptoms when standing. Recent studies have questioned the necessity for instrumented fusion in all cases of DLS and have recommended a more individualized approach to patient care. The Canadian Spine Outcomes and Research Network (CSORN) is currently undertaking an observational study of the treatment of this common condition. This pilot study examines the utility of an upright standing MRI in identifying dynamic morphometric changes that may allow refinement of individualized surgical treatment of DLS. Methods: Patients with single level, grade I or II L4/5 DLS, on the surgical waitlist at a single institution, were consented to participate. Participants were imaged in an upright MRI scanner (both sagittal and axial T2 images) in both the supine and standing positions, using an imaging protocol developed with a group of healthy volunteers (n = 15). The following morphometric parameters were measured: disc height, magnitude of slip, disc angle, presence of facet effusion, volume of subarticular space, diameter of mid foramen and cross sectional area of thecal sac. Measures from supine and standing images were compared using paired t-test statistics. Results: Ten patients (mean age 66.4 years) with DLS were included, 7 of which were female. Patients were evenly divided between Grade I (n = 5, 50%) and II (n = 5, 50%) DLS. All patients were able to

P115 PREOPERATIVE SHOULDER DIRECTIONALITY DEPENDS ON THE PROFILE OF MAIN THORACIC CURVE IN LENKE TYPE 2 ADOLESCENT IDIOPATHIC SCOLIOSIS PATIENTS

Background: Preoperative directionality of shoulder tilting seems to be independent of the radiographic features of proximal thoracic (PT) curve in adolescent idiopathic scoliosis (AIS) patients. To date, no study had investigated the mechanisms underlying the variety of preoperative directionalities of shoulder tilting in AIS patients. Purpose of the study: The purpose of this study was to evaluate the differences of radiographic features between Lenke type 2 (double thoracic curve) AIS patients with different preoperative directionalities of shoulder tilting. Materials and methods: A total of 130 Lenke type 2 AIS patients were included in this study and were divided into 2 groups according to the value of radiographic shoulder height (RSH). There were 78 cases (71 females and 7 males) with RSH less than 0 cm in group A and 52 cases (44 females and 8 males) with RSH equal to or more than 0 cm in group B. Preoperative standing anteroposterior X-ray films of the spine were obtained in all these subjects and were analyzed with respect to the following parameters: T1 tilt, PT Cobb angle, main thoracic (MT) Cobb angle, the apical level of PT curve, the apical level of MT curve and RSH. These parameters were compared between these 2 groups and the correlations between RSH and the other parameters were analyzed in all of these subjects. Results: No significant difference was found between these 2 groups with respect to PT Cobb angle or the apical level of PT curve (p [ 0.05). The apical level of MT curve was significantly more proximal in group A compared with group B (p \ 0.05). The MT Cobb angle was significantly larger in group A compared with group B (p \ 0.05). Both the T1 tilt and the PT Cobb angle/MT Cobb angle ratio in group A were significantly smaller in group A than those in group B (p \ 0.05). The RSH was positively associated with T1 tilt, the apical level of MT curve and the PT Cobb angle/MT Cobb angle ratio but was negatively associated with MT Cobb angle (p \ 0.05). Conclusion: The directionality of shoulder tilting is diverse in Lenke type 2 AIS patients. The preoperative directionality of shoulder mainly depends

123

S392 on the profile of MT curve rather than PT curve. The RSH should be carefully evaluated before making a surgical plan in these patients. Disclosures: Author 1: none; Author 2: none; Author 3: none; Author 4: none; Author 5: none.

P116 DOES THE POSITION OF THE HIP IN THE RIGHT LATERAL DECUBITUS POSITION AFFECT THE RETROPERITONEAL OBLIQUE CORRIDOR? AN MRI EVALUATION FROM L2 TO L5 Vit Kotheeranurak, Worawat Limthongkul, Weerasak Singhatanadgige, Wicharn Yingsakmongkol Spine Unit, Orthopedic Department, King Chulalongkorn Memorial Hospital, Bangkok, Thailand Purpose: To determine whether the retroperitoneal oblique corridor will be affected by the hip position in the right lateral decubitus position. Methods: Thirty-six consecutive patients undergoing diagnostic MRI were enrolled and MRI scan from L2 to L5 was performed in supine, right lateral decubitus positions with hip flexion, and right lateral decubitus with hip neutral, respectively. The retroperitoneal oblique corridor (ROC) was defined as the distance between the left lateral border of the aorta or the iliac artery (A) and the anterior medial border of the psoas (B), the ROC width at each level was measured and compared to each other. Additionally, the relative psoas cross-sectional area (CXA) and the psoas thickness (anterior 1/3) were also evaluated at each level. Results: ROC of hip in neutral position were significantly larger than flexion in all levels (All p \ 0.05); there was no significantly difference of ROC between all levels (p = 0.22), ROC was largest at L2/3 followed by L3/4 and L4/5, respectively. CXA and the psoas thickness (anterior 1/3) were smallest when the hip was in neutral position (All p \ 0.05). Conclusions: The retroperitoneal oblique corridors of L2–L5 were significantly increased when the hip is in neutral position, while the psoas cross-sectional area and thickness were minimized in this position. We recommend the neutral position of the left hip in the oblique lateral lumbar interbody fusion procedure. Disclosures: Author 1: none; Author 2: consultant: Medtronic; Author 3: none; Author 4: none.

123

Eur Spine J (2017) 26 (Suppl 2):S335–S405 P117 DIAGNOSING NECK PAIN: WHICH TESTS PROVIDE USEFUL INFORMATION? EVIDENCE-BASED RECOMMENDATIONS Nade`ge Lemeunier, Pierre Coˆte´, Heather Shearer, Jessica Wong, Louis-Rachid Salmi, Margareta Nordin Institut Franco-Europe´en de Chiropraxie, Toulouse, France Background: Neck pain and its associated disorders (NAD) are common in the general population. An accurate diagnosis is necessary to guide patient management and inform prognosis. Objectives: To develop recommendations for the clinical assessment of patients with NAD. Method: We updated the systematic review of the literature on the assessment of neck pain published by the 2000–2010 Bone and Joint Decade Task Force on Neck Pain and its Associated Disorders in 2008. We conducted 11 systematic reviews to synthesize the evidence on the validity and reliability of tests. A multidisciplinary panel of 18 clinical and scientific experts used the evidence to develop recommendations. Clinical tests and tools with adequate validity and reliability informed the development of recommendations. The recommendations were adapted to the French health care system following a consultation with stakeholders. Results: We recommend the following tests and tools for the assessment of NAD: (1) extension–rotation test, upper limb tension tests and neurological examination to assess anatomical integrity of the cervical spine; (2) C-spine rules and NEXUS criteria to rule out cervical spine injury; (3) Visual Analogic Scale or Numerical Rating Scale to assess pain intensity; (4) Neck Disability Index, Whiplash Disability Index, Neck Pain Driving Index or ProFitMap-Neck to assess disability; (5) static manual palpation of joints, trigger and tenders points to locate pain; (6) range of motions tests to evaluate cervical spine mobility. Clinicians should use these tests and tools to classify patients according to the 2000–2010 Bone and Joint Decade Task Force on Neck Pain and its Associated Disorders Classification. NAD Grade I is defined as pain with no signs or symptoms suggestive of major structural pathology and no or minor interference with activities of daily living. NAD Grade II is associated with pain with no signs or symptoms of major structural pathology, but major interference with activities of daily living. NAD Grade III refers to pain with no signs or symptoms of major structural pathology, but presence of neurologic signs such as decreased deep tendon reflexes, weakness, and/or sensory deficits. NAD Grade IV is pain associated with signs or symptoms of major structural pathology, such as fracture, myelopathy, neoplasm, or systemic disease, requiring prompt investigation and treatment. Conclusion: Our recommendations provide a useful approach for the assessment of patients with neck pain. We recommend that clinicians focus their clinical assessment on the tests and tools with adequate reliability and validity. Our results will help to standardize the diagnostic approach of patients with neck pain. Conflict of interest: The authors declare that they have no conflict of interest. Disclosures: Author 1: grants/research support: Association Franc¸aise de Chiropraxie, Institut Franco-Europe´en de Chiropraxie and Fond de Dotation en Recherche chiropratique, France,employee: Institut Franco-Europe´en de Chiropraxie; Author 2: grants/research support: (1) Ontario Ministry of Finance and Financial Services Commission of Ontario; (2) Canada Research Chair Program—Canadian Institutes of Health Research; (3) Ontario Trillium Foundation; (4) ELIB Reserach. Foundation, consultant: Eurospine Task Force on Research,other financial report: Expert witness on medical

Eur Spine J (2017) 26 (Suppl 2):S335–S405 malpractice; Author 3: none; Author 4: none; Author 5: none; Author 6: consultant: Medtronics, royalties: Wolters Kluwer, Springer.

P118 CARDIOPULMONARY FUNCTION TEST IN PATIENTS WITH CONGENITAL SCOLIOSIS: IMPACT OF CONGENITAL THORACIC SPINAL DEFORMITIES ON CARDIOPULMONARY FUNCTION Youxi Lin, Jinmei Luo, Wangshu Yuan, Hui Cong, Zheng Li, Jianxiong Shen Department of Orthopaedic Surgery, Peking Union Medical College Hospital, Beijing, China Background: Congenital scoliosis led to dysfunction of respiratory system. However, little is known about its impact on exercise capacity of patients. This study aims to investigate the correlation of spinal deformity and exercise tolerance in patients with congenital scoliosis. Methods: 40 patients with congenital scoliosis were included in this prospective study from January 2014 to December 2016. All patients had radiological assessment of the spine, as well as pulmonary function test and cardiopulmonary cycle ergometer test. The radiographic parameters of the spine were measured, and indices of pulmonary function and exercise test was collected. 2-tailed Pearson and Spearman correlation test were performed. Results: 26 female aged 17.5 years (10–39), with average coronal thoracic curvature of 77.5 (3–145) and 14 male subjects aged 18.9 years (13–32), with average coronal thoracic curvature of 68.5 (17–153), were included. Major thoracic curvature, thoracic apical vertebral rotation, thoracic apical vertebral translation, number of thoracic vertebra involved and of thoracic vertebra with congenital deformities were significantly correlated with most of static pulmonary function parameters, respectively, as shown in forced expired volume in one second (P \ 0.01, r from -0.629 to -0.521), forced vital capacity (P \ 0.01, r from -0.688 to -0.546), peak expiratory flow (P \ 0.05, r from -0.482 to -0.366), vital capacity (P \ 0.001, r from -0.707 to -0.621), total lung capacity(P \ 0.001, r from -0.705 to -0.611), residual volume (P \ 0.05, r from -0.425 to -0.351) and residual volume/total lung capacity ratio (P \ 0.05, r from 0.421 to 0.514). In cardiopulmonary exercise test, radiographic parameters were significantly correlated with most of the parameters of ventilation, including tidal volume (P \ 0.05, r from -0.604 to -0.379), respiratory rate (P \ 0.01, r from 0.441 to 0.621) and breathing reserve both at rest (P \ 0.01, r from -0.681 to -0.438) and maximum exercise (P \ 0.05, r from -0.584 to -0.371), but not with minute ventilation. Blood oxygen saturation at maximum exercise (P \ 0.05, r from -0.524 to -0.374) and its decrease (P \ 0.05, r from 0.363 to 0.511) were also significantly correlated with radiographic parameters. Conclusion: Although exercise capacity did NOT correlate to the severity of the thoracic deformity, static pulmonary function test demonstrated respiratory dysfunction, and cardiopulmonary exercise tests revealed decompensation and changes of breathing pattern as thoracic deformities worsened. Disclosures: Author 1: none; Author 2: none; Author 3: none; Author 4: none; Author 5: none; Author 6: none.

S393

Table Correlations Between Radiographic Parameters and Pulmonary Function Test/Cardiopulmonary Function Test in Patients with Congenital Scoliosis Major thoracic curvature (deg) (P)

Pulmonary Function Test FEV1(% of pred.) FVC(% of pred.) PEF(% of pred.) VC(% of pred.) TLC(% of pred.) RV(% of pred.) RV/TLC(%) Metabolic gas exchange Load(% of pred.) VO2peak(% of pred.) VO2peak (mL/kg/min) AT(%) RER Ventilation VE(% of pred.) Vt(% of pred.) RR(per min) BRr(%) BRt(%) Pulmonary gas exchange SpO2r (%) SpO2t (%) SpO2(%) Vd/Vt Cardiovascular system HR(% of pred.) O2/pulse(% of pred.)

Major thoracic apical vertebral rotation (NashMoe grade) (S)

Major thoracic apical vertebral translation (cm) (P)

Number of thoracic vertebra involved (S)

Number of thoracic vertebra with congenital deformities (S)

r

P

r

P

r

P

r

P

r

P

-0.629 -0.585 -0.398 -0.664 -0.699 -0.425 0.463

<0.001** <0.001** 0.011* <0.001** <0.001** <0.001** 0.008**

-0.521 -0.548 -0.366 -0.617 -0.684 -0.339 0.432

0.001** <0.001** 0.020* <0.001** <0.001** 0.058 0.014*

-0.548 -0.428 -0.378 -0.621 -0.611 -0.351 0.388

<0.001** 0.006** 0.016* 0.001** <0.001** 0.049* 0.028*

-0.694 -0.688 -0.482 -0.681 -0.675 -0.240 0.514

<0.001** <0.001** 0.002** <0.001** <0.001** 0.185 0.003**

-0.583 -0.546 -0.414 -0.707 -0.705 -0.348 0.421

<0.001** <0.001** 0.008** <0.001** <0.001** 0.051 0.016*

-0.170 -0.067 0.044 -0.105 -0.040

0.308 0.690 0.794 0.531 0.810

-0.015 0.135 0.165 -0.027 -0.109

0.930 0.419 0.321 0.872 0.515

-0.109 -0.067 0.158 -0.126 0.050

0.513 0.691 0.344 0.450 0.766

0.113 0.130 0.383 0.029 -0.146

0.498 0.437 -0.145 0.861 0.382

-0.027 0.000 0.085 -0.080 -0.065

0.871 0.999 0.613 0.634 0.696

-0.158 -0.466 0.534 -0.557 -0.433

0.342 0.003** 0.001** <0.001** 0.007**

-0.256 -0.461 0.494 -0.438 -0.408

0.121 0.004** 0.002** 0.006** 0.012*

-0.021 -0.379 0.529 -0.439 -0.371

0.902 0.019* 0.001** 0.006** 0.024

-0.213 -0.537 0.441 -0.681 -0.584

0.198 0.001** 0.006** <0.001** <0.001**

-0.240 -0.604 0.621 -0.537 -0.526

0.146 <0.001** <0.001** 0.001** 0.001**

-0.455 -0.524 0.438 0.187

0.009** 0.003** 0.015* 0.260

-0.285 -0.464 0.511 0.221

0.114 0.010* 0.004** 0.182

-0.501 -0.575 0.483 0.217

0.003** 0.001** 0.007** 0.191

-0.333 -0.374 0.363 0.065

0.063 0.042* 0.049* 0.699

-0.266 -0.471 0.497 0.209

0.141 0.009** 0.005** 0.208

-0.244 0.043

0.139 0.796

-0.114 0.155

0.494 0.354

-0.172 0.052

0.301 0.756

-0.196 0.023

0.239 0.892

0.001 -0.088

0.998 0.600

(P), 2-tailed Pearson test, (S), 2-tailed Spearman test. *<0.05,**<0.01. FEV1 = forced expiratory volume in 1 second, FVC = forced vital capacity, VC = vital capacity, TLC = total lung capacity, RV = residual volume, RV/TLC = the ratio of residual volume and total lung capacity. VO2peak = peak oxygen intake, AT = anaerobic threshold, RER = respiratory exchange ratio, VE = maximum ventilation volume per minute, Vt = maximum tidal volume, RR = maximum respiratory rate, BR = breath reserve (r= at rest, t=at maximum exercise, same below.), SpO2 = oxygen saturation, pulse.

= change, Vd = dead space ventilation, HR = maximum heart rate , O2/pulse = oxygen

P119 RELATION OF LUMBAR SYMPATHETIC CHAIN TO THE OPEN CORRIDOR OF RETROPERITONEAL OBLIQUE APPROACH TO LUMBAR SPINE: AN MRI STUDY Akaworn Mahatthanatrakul1, Thun Itthipanichpong2, Chindarat Ratanakornphan3, Numphung Numkarunarunrote3, Weerasak Singhatanadgige2, Wicharn Yingsakmongkol2, Worawat Limthongkul2 1 Department of Orthopaedics, Naresuan University Hospital, Phitsanulok, Thailand; 2Department of Orthopaedics, Faculty of Medicine, Chulalongkorn University, Bangkok, Thailand; 3 Department of Radiology, Faculty of Medicine, Chulalongkorn University, Bangkok, Thailand

Background: Minimally invasive retroperitoneal oblique approach to lumbar spine used surgical corridor between psoas muscle and aorta for exposure to anterior part of lumbar spine. Lumbar sympathetic chain (LSC) runs anterolateral to vertebral body and anterior to psoas muscle which make it structure at risk for injury for this approach. Injury to LSC can lead to decreased quality of life. Purpose of the study: Identify LSC relationship with surgical corridor for minimally invasive retroperitoneal anterolateral oblique approach different in each lumbar intervertebral disc level. Materials and methods: Magnetic resonance imaging (MRI) of lumbar spine of 143 patients were included. Exclusion criteria were patients with history of spinal injury, infection, malignancy or previous spinal instrumentation. The MRIs were reviewed by two orthopedists and one musculoskeletal radiologist independently. Left LSC was identified in T2-weight axial image at L2–3, L3–4 and L4–5 intervertebral disc levels. Distances between LSC and left psoas muscle, aorta or common iliac artery were recorded. Patients were grouped into non-scoliosis group, levoscoliosis group and dextroscoliosis group for data analysis. Results: Mean age of the patients was 62.3 years with 75.7% female. LSC were identifiable in 90.9% of levels. (92.3% in non-scoliosis group and 91.7% in levoscoliosis group and 81.25% in dextroscoliosis group) Distance between LSC and psoas muscle were 4.0 mm (95%CI 3.4–4.5) at L2–3, 4.7 mm (95%CI 4.0–5.3) at L3–4, 5.2 mm (95%CI 4.6–5.9) at L4–5. Statistical different was found between L2–3 and L4–5 level (p = 0.006). Distance between LSC and aorta or

123

S394 iliac artery were 12.4 mm (95%CI 11.3–13.4) at L2–3, 12.3 mm (95%CI 11.2–13.5) at L3–4, 10.6 mm (95%CI 9.2–12.0) at L4–5. No statistical different was found between each level. LSC size was 2.4 mm with no different between each level. In non-scoliosis group distance between LSC and psoas muscle were 3.1 mm (95%CI 2.7–3.5) at L2–3, 3.3 mm (95%CI 3.0–3.7) at L3–4, 4.0 mm (95%CI 3.5–4.5) at L4–5. Statistical different was found between L2–3 and L4–5 level (p = 0.012) and between L3–4 and L4–5 level (p = 0.041). Distance between LSC and aorta or iliac artery were 11.9 mm (95% CI 11.1–12.7) at L2–3, 11.4 mm (95% CI 10.5–12.2) at L3–4, 10.2 mm (95% CI 9.1–11.2) at L4–5. Statistical different was found between L2–3 and L4–5 disc level (p = 0.039) Conclusion: In non-scoliosis patients, LSC move away from psoas muscle and become closer to aorta in L4–5 disc level. LSC have more variation and more difficult to identify by MRI in scoliosis patient. Pre-operative planning for minimally invasive retroperitoneal oblique approach should include location of LSC. Disclosures: Author 1: none; Author 2: none; Author 3: none; Author 4: none; Author 5: none; Author 6: none; Author 7: consultant: Medtronic.

P120 CLINICAL IMPORTANCE OF POSTERIOR VERTEBRAL HEIGHT LOSS ON PLAIN RADIOGRAPHY WHEN CONSERVATIVELY TREATING OSTEOPOROTIC VERTEBRAL FRACTURES Jun-Yeong Seo, Yong-Suk Kwon, Young-Hoon Kim, Kee-Yong Ha Department of Orthopaedic Surgery, Jeju National University Hospital, Jeju, Republic of Korea Purpose: To predict spinal canal compromise, the assessment of plain radiography with magnetic resonance imaging (MRI) can aid the detection of vertebral body collapse and prevent the development of neurological deficits. Methods: Patients who suffered osteoporotic vertebral fractures (OVFs) between June 2010 and January 2012 underwent consecutive radiological assessments, including measurements of anterior height loss (AHL), posterior height loss (PHL), and the kyphotic angle (KA). The fracture morphology was classified by AOSpine thoracolumbar spine injury classification system. MRI was performed at the initial assessment and the extent of canal encroachment (CE) was calculated in all patients. Follow-up computed tomography (CT) or MRI was performed in patients exhibiting significant height loss in follow-up radiography. The fracture patterns in T1- and T2-weighted MRI were also assessed. Results: A total of 141 patients visited our institute for treatment of OVFs and 97 were enrolled; 15 were male and 82 were female. The mean age at initial visit was 70.3 ± 14.6 years. The initial spinal CE was correlated with the initial PHL and the initial AHL. The follow-up CE was correlated with age, the initial PHL, and the difference between the initial and last PHL DPHL (initial-last). OVFs with both endplate fractures have a greater tendency of posterior wall collapse than those with single endplate fracture. On initial T1-weighted sagittal MRI, a diffuse low signal change pattern of the fractured vertebra was correlated with PHL. Delayed neurological deficits developed in four patients. These patients underwent surgical intervention. Conclusions: In patients with simple compression fractures, attention should be paid to the posterior vertebral body and both endplates as well as the T1-weighted MRI findings to allow early detection of spinal canal compromise, which can have devastating consequences.

123

Eur Spine J (2017) 26 (Suppl 2):S335–S405 Disclosures: Author 1: none; Author 2: none; Author 3: none; Author 4: none.

P121 NEUROLOGIC DEFICITS AND MRI CHARACTERISTICS OF SYRINX IN IDIOPATHIC SYRINGOMYELIA RELATED SCOLIOSIS Haining Tan, Fan Feng, Youxi Lin, Xingye Li, Chong Chen, Zheng Li, Jianxiong Shen Department of Orthopaedic Surgery, Orthopaedic Surgery Beijing, Beijing, China Purpose: To explore the relationship between syrinx of idiopathic syringomyelia (IS) with scoliosis or neurological deficits. Background: Limited studies have shown that syrinx deviation is correlated with both scoliosis curve convexity and neurological abnormality. This study aims to further demonstrate the relationship between neurologic deficits, scoliosis curve and syrinx features. Methods: Fifty-five scoliosis patients secondary to IS were identified after ruling out all other causes of syringomyelia (Chiari malformation, spinal cord tumor, trauma, infection, tethered cord, etc.) and reviewed retrospectively. Patients with syrinx less than two vertebra levels or diameter less than 1 mm also were excluded. Location, syrinx/cord ratio (S/C), length and morphological appearance of the syrinx were systematically assessed on MR images. Neurologic symptoms were recorded through detailed physical examination of nervous system. Three subgroups (none, minor and severe) were classified according to reflex, sensory and motor disturbance. Results: The major curve Cobb was 69.8 ± 25.2 (range 33–132). The maximal S/C and length of the syrinx in IS averaged 0.58 ± 0.20 (range 0.20–0.98) and 8.4 ± 4.7 (range 2–19) vertebral levels. 36 (65.5%) patients had various neurological deficits, including tendon or superficial abdominal reflex abnormality (36, 65.5%), sensory and/ or motor disturbances (22, 40.0%). The S/C, length and morphological features had no correlation with degree of neurological deficit and scoliosis curve parameters, such as Cobb, flexibility, and apex vertebra translation (AVT). The major curve convexity wasn’t coincident with side of syrinx significantly (27.2% concordance rate, P = 0.52) or neurologic deficit (16.3% concordance rate, P = 0.21). Location of syrinx wasn’t correlated with range of major curve. Conclusion: None significant relationship is detected among neurologic deficits, scoliosis curve parameter and MRI features of syrinx. Correlation between syrinx of IS with either scoliosis curve or neurological deficit has been controversy. Disclosures: Author 1: none; Author 2: none; Author 3: none; Author 4: none; Author 6: none; Author 7: none.

Table 1 Correlation between major curve direction and the side of deviated syrinx Major curve direction Left Right Fisher exact test, P=0.52 Major curve direction Right Left Fisher exact test, P=0.29

Syrinx deviation Left Non-Left 6 13 8 28

Syrinx deviation Right Non-Right 9 27 2 17

Eur Spine J (2017) 26 (Suppl 2):S335–S405 P122 INFLUENCE OF SPINOPELVIC ALIGNMENT AND MORPHOLOGY ON DEVIATION IN THE COURSE OF THE PSOAS MUSCLE

S395 P123 IS THE COBB TECHNIQUE THE MOST RELIABLE AND VALID TO ASSESS THE SCOLIOTIC DEFORMITY IN 3D?

Shimei Tanida, Shunsuke Fujibayashi, Bungo Otsuki, Kazutaka Masamoto, Shuichi Matsuda

Fares Yared, Ayman Assi, Nour Khalil, Aren Joe Bizdikian, Ziad Bakouny, Joeffroy Otayek, Gerard Bakhos, Chris Labaki, Mohammad Karam, Ismat Ghanem

The Department of Orthopaedic Surgery, Graduate School of Medicine, Kyoto University, Kyoto, Japan

Laboratory of Biomechanics and Medical Imaging, Faculty of Medicine, University of Saint-Joseph, Beirut, Lebanon

Background: In some people, the psoas major rises laterally or anteriorly at the L4/5 disc level and detaches from the most posterior aspect of the disc despite the absence of transitional vertebrae; this is called the ‘‘rising psoas sign.’’ There are no reports of the relationship between spinopelvic parameters and rising psoas sign. The objective of this study was to investigate the relationship between spinopelvic parameters and deviations in the location and shape of psoas major muscle at the L4/5 disc level. Methods: We investigated the preoperative location and shape of both psoas major muscles in 64 patients treated with lateral lumbar interbody fusion. Spinopelvic parameters were measured on X-ray films. The morphology of psoas major at the L4/5 disc level was investigated with magnetic resonance images. The morphological measurements were normalized by the anteroposterior diameter of the center of the L4 vertebral body, which was measured by computed tomography. The rising psoas sign was classified into 2 types: bilateral or unilateral (Figure). Results: The pelvic incidence (PI) was significantly larger for the bilateral type than the others (normal and unilateral types) (60.2 ± 11.0 vs. 46.7 ± 8.7, p \ 0.001). The PI correlated significantly with the normalized anteroposterior diameter of the pelvis (R = 0.66, p \ 0.001). The receiver-operator characteristic curve showed an optimal cutoff value of PI = 54, with 75% sensitivity and 78.5% specificity. The coronal L1–4 Cobb angle was significantly larger in the unilateral type than the others (normal and bilateral types) (p \ 0.0001). In the unilateral type, the Cobb angle in the recumbent position correlated significantly with the normalized distance of the lateral deviation of psoas major (R = 0.60, p = 0.0085). Conclusion: The rising psoas sign was related to a higher PI and lumbar scoliosis. It was firstly elucidated that the spinopelvic alignment and morphology influence the deviation of the course of the psoas major muscle. Disclosures: Author 1: none; Author 2: none; Author 3: none; Author 4: none.

Introduction: Adolescent idiopathic scoliosis (AIS) is a 3D deformity of the spine commonly assessed on frontal X-rays using the Cobb angle technique [1]. Some authors have found this technique to be poorly representative of the extent of the deformity, suggesting that alternatives, such as the Ferguson [2] and centroid techniques [3], could better characterize this deformity. Our hypothesis was that the Cobb is the most reliable and valid technique to assess the scoliotic deformity. Purpose: The aim was to evaluate the reliability and validity of different measurement techniques of the scoliotic deformity. Methods: 73 subjects with AIS [55F, age 14 years (9–18)] underwent full body biplanar X-rays with 3D spinal reconstructions. The subjects were stratified into 4 groups with increasing Cobb values: Gr1 (10– 20) N = 20, Gr2 (20–40) N = 20, Gr3 (40–70) N = 20, Gr4 ([70) N = 13. 2D frontal X-rays were used to measure the Cobb, Ferguson and centroid angles (Fig. 1a) digitally using a software, and manually on a film using a goniometer. 3D reconstructions were used to calculate the frontal, sagittal and axial deformity indices (FDI, SDI and ADI: averages of absolute values of upper endplate inclinations or axes of each vertebra in the scoliotic segment, in the corresponding plane) which were considered as gold standard for deformity evaluation. Reliability was assessed by measuring each parameter twice manually then digitally, by three operators. Intraclass correlation coefficient (ICC) and reproducibility variance (SR) for each parameter were evaluated. As for the validity, stepwise multiple linear regressions (SMLR) were used to investigate which measurement technique is the most representative of the deformity indices in the three planes. Results: Overall, SR was higher for the manual technique regardless of the method, and increased with the severity of the deformity. In Groups 1, 2 and 3, the Cobb method showed the lowest SR (2.2; 2.8; 3.6 in groups 1–3, respectively) compared to the Ferguson and Centroid methods (from 2.6 to 4.8 and from 3.8 to 5.4, respectively). In Gr4, the Ferguson method showed the lowest SR (5.8 for Ferguson, vs 7.6 for the Cobb and 10 for the centroid). ICC was high in all groups and techniques ([0.89). The SMLR models found that the only determining factor for all techniques was the FDI. The slope of the linear regressions between the FDI and each parameter (Fig. 1b) was: 1.9 for Ferguson (R2 = 0.94), 2.79 for centroid (R2 = 0.93) and 2.97 for Cobb (R2 = 0.95). Conclusion: The Cobb was the most reliable and representative of the 3D deformity and shows that it would better differentiate between spines with different curvatures (highest slope of correlation). Although some studies had criticized the Cobb angle for inadequately representing the scoliotic deformity, this study revealed that the Cobb angle was the most appropriate measurement technique to assess the 3D scoliotic deformity. References (1) Cobb J, 1948. (2) Ferguson A, 1930. (3) Chen YL, 2007 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.

123

S396

Eur Spine J (2017) 26 (Suppl 2):S335–S405 PATIENT SAFETY

P125 THE INFLUENCE OF CONCOMITANT SYRINGOMYELIA ON PATIENT REPORTED OUTCOME FOLLOWING HIND BRAIN DECOMPRESSION Peter Janous, Neil Buxton, Tim Pigott, Marco Teli, Simon Clark Department of Neurosurgery, Orthopaedic Surgery Osaka, The Walton Centre, Liverpool, UK P124 WHERE TO STOP DISTALLY IN LENKE MODIFIER C AIS WITH LUMBAR CURVE MORE THAN 60? Zezhang Zhu, Xiaodong Qin, Yong Qiu Spine Surgery, Drum Tower Hospital of Nanjing University Medical School, Nanjing, China Introduction: The selection of LIV in AIS patients with large lumbar curve remains controversial. Stopping the distal fusion at L3 could save more mobile lumbar spinal segments but may increase the risk of decompensation. Purpose: The purpose of the study was to evaluate pre-operative radiographic factors that were associated with the selection of either L3 or L4 as LIV in posteriorly treated AIS patients with large lumbar curve ([60). Methods: 84 AIS patients with lumbar curve more than 60were included with a minimum of 2-year follow-up after posterior spinal fusion for lumbar curves (Lenke type 3C, 5C or 6C). Patients were grouped according to the selection of LIV, either L3 group or L4 group. All radiograph parameters were measured pre- and post-operatively including lumbar Cobb angle, lumbar flexibility and L3 translation and rotation on upright posteroanterior film and supine side-bending film, etc. The SRS-22 score was used to assess clinical outcomes. Radiographic and clinical parameters were compared between the two groups. Multivariate regression analysis was performed to determine the factors most predictive of LIV selection. Results: There were 24 patients in L3 group and 60 patients in L4 group. The average duration of follow-up was 3.1 years. At last followup, no difference was found in the clinical and radiographic parameters between the two groups. Preoperatively, the L3 group had lower L3 translation (28.6 vs. 35.1, p = 0.024), L3 translation on concave sidebending film (5.5 vs. 13.5, p \ 0.001), L3 rotation on convex sidebending film (1.3 vs. 2.0, p = 0.001) and larger lumbar flexibility (65.2% vs. 53.9, p = 0.022). Multivariate regression found that L3 translation on concave side-bending film was the single most important predictor of LIV selection. Specifically, concave bending L3 translation \10 mm was a potential threshold for selecting L3 as LIV. Conclusion: For AIS patients with lumbar curve larger than 60, one can reliably stop at L3 if preoperative L3 translation on concave sidebending film was less than 10 mm, with the same radiographic and clinical outcomes as fusing to L4. Disclosures: Author 1: none; Author 2: none; Author 3: none.

123

Objective: To evaluate the impact of concomitant syringomyelia on patient reported outcome measures and complications in patients undergoing hindbrain decompression for a Chiari 1 malformation. Design: Prospective data collection within the Spine Tango Registry data for one institution. Subjects: 95 patients undergoing Foramen Magnum Decompression (FMD) between March 2011 and March 2015. Methods: Outcome evaluation was performed using the neck Core Outcome Measures Index (COMI neck) and Gestalt impression (to assess improvement of headaches) was applied retrospectively at 12 months. Patients were split into two cohorts, those with and those without syringomyelia. Both cohorts were compared in all domains of the COMI neck questionnaires, headaches (Gestalt impression) and complications. Non-parametric data were analysed with Wilcoxon signed rank test (baseline and 12 months data), Man–Whitney U test and Fisher exact test (comparison of 12 months outcomes). Parametric data (age) were analysed with Student T-test. SPSS Software was used for above analysis. Results: 79 out of 95 (83%) patients returned 1 year follow-up COMI neck questionnaires. Thirty-three had concomitant syringomyelia and 46 had no syringomyelia present. There was no statistical difference in age, gender or ASA grade (p [ 0.05) between cohorts. Significantly more patients (11 vs. 4; p = 0.0186) had repeat surgery in syrinx group. There was no statistically significant difference in patient reported outcomes (COMI neck index median 4.5 ± 3.3 vs 4.2 ± 3.2; p = 0.376) between the syrinx and non-syrinx cohorts. However postoperative neck pain (median 4 ± 3.35 vs 1 ± 3.17; p 0.041) and arm/shoulder pain scores (2 ± 3.38 vs. 0 ± 2.628; 0.049) were significantly lower in the non-syrinx cohort. In both cohorts 57% patients had an improvement in headaches. 92% patients were ‘satisfied’ with treatment and 63% stated that the operation ‘helped’. 54% and 59% of patients in the syrinx and non-syrinx cohorts, respectively self-reported complications as a consequence of the operation. Statistical difference was not reached when comparing COMI neck index of patients with self-reported complications and without (p = 0.121). Overall, 25% patients required insertion of ventriculo-peritoneal shunt. Conclusions: This study demonstrates that clinical effectiveness of FMD is lower and self-reported complications are higher when evaluated by patients via patient reported outcome measures. Patients with and without concomitant syrinx showed equal overall outcomes, although neck and arm pain was higher in syrinx patients. High patients satisfaction with treatment reflects high standards of provided care in our institution. Self-reported complications did not have effect on outcomes.

Eur Spine J (2017) 26 (Suppl 2):S335–S405 Disclosures: Author 1: none; Author 2: none; Author 3: none; Author 4: none; Author 5: grants/research support: medtronic, consultant: medtronic.

S397 Disclosures: Author 1: none; Author 2: grants/research support: DePuy Synthes; Author 3: none; Author 4: none; Author 5: none; Author 6: grants/research support: DePuy, consultant: Medtronic.

P126 STANDARDIZED MULTIMODAL RECOVERY PATHWAY ACCELERATES INPATIENT RECOVERY Peri Kindan, Nuray Sogunmez, Yasemin Yavuz, Binnaz Ay, Caglar Yilgor, Ahmet Alanay School of Medicine, Acibadem University; Istanbul, Turkey Background: Posterior spinal fusion (PSF) is often associated with prolonged operative times, extensive soft tissue dissection, significant blood loss and blood transfusion. Implementation of standardized clinical pathways is reported to reduce pain and shorten length of hospital stay. Purpose: The aim was to compare the effects of standard care, adapted and tailored clinical pathways on post-operative inpatient recovery period. The hypothesis is standardized anti-fibrinolytic, multimodal analgesic and recovery pathway for adolescent idiopathic scoliosis (AIS) improves inpatient recovery after PSF. Methods: A retrospective comparative cohort study was conducted. In total, 74 patients (64F, 10M) with AIS diagnosis and underwent PSF were included as the patient sample group. A published clinical pathway was adapted and implemented in our hospitaland it was used in 27 consecutive AIS patients (23F, 4M). After the first experiences and completion of adaptation, the pathway was tailored in accordance with the feedbacks from the relevant disciplines and was used in 25 consecutive patients (23F, 2M). Age and curve matched 22 patients (18M, 4F) operated before implementation of adapted clinical pathway formed the control group. Protocols in the pre-pathway, adapted and tailored pathways are summarized in table 1. Fisher’s exact, Kruskal–Wallis, One-way ANOVA and Chi squared tests were used in comparisons. Results: Mean age, curve type, curve magnitude, curve correction, number of levels fused and % of chevron osteotomies were not different in three groups (Table). Estimated blood loss, usage of drains, blood transfusion rates, % of ICU stay, time to first defecation and time to discharge were significantly reduced with the adapted and tailored pathways (Table). Post-operative SRS22 scores were not different between groups (p [ 0.05). One patient in the pre-pathway group had PJK. One patient in the adapted pathway had an ileus. The tailored pathway was started with no-drains strategy, however there were two wound hematomas in the first ten patients, of whom one required debridement. One was a male patient and both had [300 ml EBL. Then the no-drains strategy was changed to drains for only male patients and/or [300 ml bleeders. Conclusion: Adaptation of a multimodal external recovery pathway may be difficult but good start. Further tailoring the adapted pathway by a multidisciplinary consensus may increase effectiveness. Adaptation and tailoring of a standardized multimodal recovery pathway resulted in reduced bleeding and transfusion, faster return to normal bowel functions and earlier discharge.

P127 BLOOD TRANSFUSION IN SURGICAL TREATMENT OF ADOLESCENT IDIOPATHIC SCOLIOSIS: ONE-CENTER EXPERIENCE OF PATIENT BLOOD MANAGEMENT IN 210 CASES Søren Ohrt-Nissen, Naeem Bukhari, Casper Dragsted, Martin Gehrchen, Pa¨r Johansson, Jesper Dirks, Jakob Stensballe, Benny Dahl Department of Orthopedic Surgery, Spine Unit, Rigshospitalet, Copenhagen, Denmark Background/introduction: Surgical treatment of adolescent idiopathic scoliosis (AIS) is associated with a substantial risk of perioperative transfusion of allogeneic red blood cells (RBC). Patient blood management (PBM) is an evidence-based multidisciplinary approach including several preventive measures to manage bleeding risks, reduce iatrogenic blood loss as well as modifying decision thresholds for the appropriate administration of blood therapy. Purpose of the study: This study aimed to describe the effect of perioperative PBM and access predictors of RBC in a large AIS cohort. Materials and methods: Patients with AIS treated with posterior instrumented fusion were consecutively enrolled from 2011 to 2016. PBM strategies were implemented in 2011 including prophylactic tranexamic acid (25 mg/kg bolus, 5 mg/kg/h during surgery), intraoperative permissive hypotension (target mean arterial pressure 40–50 mmHg during bleeding), restrictive fluid therapy including avoidance of synthetic colloids, restrictive RBC trigger according to

123

S398 institutional standardized protocol (Hb \ 7.3 g/dL in 2011–13; Hb \ 7.0 g/dL from 2014), and use of cell salvage. Results: A total of 210 patients were included. Sixty-four patients (31%) received RBC transfusion. A decline in perioperative RBC transfusion rate was observed from 77% in 2011 to 13% in 2016 (p \ 0.001). The RBC group had a significantly higher preoperative Cobb angle, a lower preoperative Hb, higher estimated blood loss (EBL) and a higher level of crystalloid volume (p \ 0.027). Multiple logistic regression including these variables as well as year of surgery (indicating effect of PBM) showed that significant predictors for RBC transfusion were preoperative Hb (OR 0.23, 95% CI 0.12–0.40), EBL (OR 1.27, 95%CI 1.15–1.42) and year of surgery (OR pr. year 0.76, 95%CI 0.57–0.99). Patients with a preoperative Hb B 13.0 g/dL (n = 29) had 3.16 times higher odds of RBC. Conclusion: We demonstrated an effect of perioperative PBM with a substantial reduction in the rate of RBC transfusion. A preoperative evaluation of anemia should be considered as an integrated part of PBM to further minimize the use of perioperative blood transfusion. Disclosures: Author 1: grants/research support: K2M; Author 3: none; Author 4: grants/research support: K2M, Medtronic; Author 5: none; Author 6: none; Author 7: none; Author 8: grants/research support: Medtronic, K2M.

Eur Spine J (2017) 26 (Suppl 2):S335–S405 is known of the effect of institutional or systematic factors on the occurrence of AE’s. Objective: This study examined adherence to BPTT’s at a single institution and the effect of failed adherence on the occurrence of AE’s. Methods: An ambispective study of all emergent spine surgery at a Quaternary Trauma Center between Jan 1, 2009 and December 31, 2013. Emergent surgery BPTT’s are \1, \4, \8, or \24 h. Patients were prospectively enrolled in an institutional database with adverse events identified using the Spine Adverse Events Severity System (SAVES). Operating room data was available from hospital database records which stores information on surgical time, BPTT, compliance to BPTT, wait time hours, surgical time and ASA score. Statistical analysis was done using Chi square and multivariate logistic regression modeling comparing adverse events by meeting BPTT or exceeding BPTT. Results: A total of 1323 patients were enrolled. 9.5% of these patients were booked as \1 h, 4.6% as \4 h, 67.1% as \8 h, and 18.6% as \24 h. The mean waiting time in these groups was 2.1 h (+1.1 h, SD 3.08), 3.9 h (-0.1 h, SD 3.75), 7.8 h (-0.2 h, SD 14.3) and 23.5 h (-0.5 h, SD 23.9), respectively. In total, 35.9% of patients failed to meet target surgical times, 77.8% of \1 h, 16.1% of \4 h, 29.1% of \8 h, and 26.7% of\24 h did not meet BPTT. Failure to meet BPTT was significantly associated with any post-operative adverse event (p \ 0.001) but not with intraoperative adverse events (p [ 0.05). On multivariate adjusted analysis pneumonia, neurologic deterioration, electrolyte disturbances, post-operative neurogenic pain, pulmonary embolism, UTI, transfusion requirement, and in-hospital mortality were all found to be associated with not meeting BPTT (p \ 0.05). Conclusion: Emergency spine surgical cases that did not meet BPTTs were associated with a significant increase in major post-operative adverse events including in-hospital mortality. These AE’s could potentially be avoided with improved access to the operating room in institutions where emergency spinal services are provided. Disclosures: Author 1: none; Author 2: none; Author 3: grants/research support: Hospira, Inc.; Canadian Anesthesiologists’ Society; Author 4: grants/research support: MEDTRONIC.

P129 REDUCTION IN WOUND COMPLICATIONS AND REVISION SURGERY WITH PLASTIC SURGERY CLOSURE IN THE TREATMENT OF NEUROMUSCULAR SCOLIOSIS COMPLICATIONS

P128 DOES EXCEEDING SURGICAL BOOKING PRIORITY TARGET TIME LEAD TO AN INCREASE IN ADVERSE EVENTS IN EMERGENCY SPINE SURGERY? Michael Bond, Raphaele Charest-Morin, Alana Flexman, John Street Combined Neurosurgical and Orthopedic Spine Program and Department of Anesthesia, University of British Columbia, Vancouver, BC, Canada Background: In health systems of limited resources, institutions utilize ‘booking priority target times’ (BPTT’s) to facilitate emergency surgeries. The decision to choose a specific BPTT is based on perceived clinical need and informed by the literature and evidence based medicine. Adverse events (AE’s) are common in emergency spinal surgery, with reported rates ranging from 9.5 to 38%. While the importance of patient and surgical factors have been evaluated, little

123

Brandon Cook, David Briski, Andrew King, Joseph Zavatsky Ochsner Medical Center, Childrens Hospital of New Orleans, New Orleans, USA Introduction: A Multi surgeon retrospective review looking at postoperative deep space infections after posterior spinal fusion can be difficult to manage and potentially devastating. The overall rate of infection after posterior spinal fusion has been reported as high as 23% in patients with neuromuscular scoliosis. A multilayered plastic surgery closure decreases potential dead space, protecting the spinal instrumentation. We compared surgically treated neuromuscular scoliosis patients with and without plastic surgery multilayered wound closure. Materials/methods: All neuromuscular scoliosis patients treated with posterior spinal fusion from 2008 to 2014 were analyzed. Patients with 2-year follow-up and completed charts were reviewed. Patients were categorized into 2 Groups: Group 1 [Plastic Closure (PC)]— included patients with a multilayered closure and advancement flaps when necessary; group 2—standard closure (SC). Differences in demographic, radiographic, and clinical parameters were analyzed. Results: 50 patients met inclusion criteria for the database, of which 39 had complete 2-year data. Group 1 had 11 patients, each having a

Eur Spine J (2017) 26 (Suppl 2):S335–S405 multilayered plastic surgery wound closure. Group 2 included 28 patients who had a standard wound closure. There was no difference in age, male gender, number of levels fused, or postop max coronal Cobb angles between the groups (Table 1). There was a significant difference in deep space infections (0 vs 7, p = 0.0057), revision surgeries (0 vs 7, p = 0.0057), EBL (2425 vs 644cc, p = 1.46E-06), OR time (467 vs 245 min, p = 1.97E-08), iliac screw fixation (58 vs. 21%, p = 0.022), and preop max coronal Cobb angle (58.29 vs 71.99, p = 0.043) in the PC vs SC groups, respectively. Conclusion: Plastic surgery closure resulted in a statistically significant decrease in infection and revision surgery rates despite this patient cohort having significant increases in blood loss, operative time, and iliac screw fixation, all of which have been shown to increase the risk of infection. Utilizing a plastic surgery closure can reduce dead space, providing better soft-tissue coverage of the spinal instrumentation reducing infections and revision surgery rates. Disclosures: Author 1: none; Author 2: none; Author 3: consultant: medicrea, stock/shareholder: nocimed; Author 4: grants/research support: AO Spine,consultant: Depuy Synthes, Zimmer Biomet, Stryker,stock/shareholder: Surgical Solutions, Vivex, royalties: Zimmer Biomet.

S399 combined fractures. Therefore most patients were undergoing surgery of the upper cervical spine (C1/2: 50%, C1–3: 27.5%, C0–3: 5%). Most patients had severe systemic diseases (52.5% ASA-3, 7.5% ASA-4). Furthermore 55% of the patients had accompanying injuries, which needed surgery in 45% of the cases. The mean stay in hospital was 18 days (IQR 9–27). No intraoperative complications or postoperative neurologic deficits were detected. The postoperative CT examinations showed proper alignment and correct implant position in all cases. In 3 cases (7.5%) a revision of the wound had to be performed due to complications of wound healing. In 19 cases (47.5%) general complications were registered: respiratory dysfunction due to pneumonia (20%), postoperative delirium (12.5%), cardial complication (12.5%), urinary tract infection (7.5%) and complications associated to accompanying injuries (5%). In two cases (5%) the complications led to death due to sepsis and cardiogenic shock. In a univariate analysis a significant correlation was shown between the occurrence of complications and duration of stay in hospital (p \ 0.001), ASA-score (p = 0.03), number of comorbidities (p = 0.022), number of drugs (p = 0.019), duration of stay at intensive care unit (p = 0.004). There was no significant difference of patients with or without complications regarding age, gender, operated levels and surgery time. Conclusion: Our findings show a high rate of general complications (47.5%) after posttraumatic dorsal cervical spine instrumentation. On the other hand, the surgical revision rate (7.5%) and 6-week mortality (5%) are low. Regarding our findings dorsal cervical fusion should be taken into consideration as a therapy option. Nevertheless the patients belong to a high-risk group, in which in particular the general complications must be recognized and addressed early. Disclosures: Author 1: none; Author 2: none; Author 3: none; Author 4: none; Author 5: none.

P131 DOES SCOLIOSIS AFFECT SLEEP BREATHING QUALITY? Xingye Li, Haiwei Guo, Zheng Li, Fan Feng, Jianxiong Shen P130 COMPLICATIONS AFTER POSTTRAUMATIC DORSAL CERVICAL SPINE FUSION Adnan Kasapovic, Kristian Welle, Cornelius Jacobs, Christof Burger, Koroush Kabir Department of Orthopedic and Trauma Surgery, University Hospital Bonn, Bonn, Germany Background: At least 25% of spine injuries affect the cervical spine. The aim of this study was to evaluate the complication rate of dorsal cervical spine fusion after cervical spine injuries without neurological deficits. Methods: 40 patients (52.5% male, 47.5% female) with fractures of the cervical spine were undergoing dorsal cervical spine instrumentation in the period of 2012–2016. Surgical and general complications during the first six postoperative weeks were registrated. Following data were obtained: age, gender, accompanying injuries, operated segments, stay at intensive care unit, stay in hospital, CT-results, medication, comorbidities, complications and revision operations. Fisher’s exact test, Chi square test and Mann–Whitney-U test were used for statistical analysis. Results: The mean age of the patients was 71.6 years (IQR 59–88). Fractures of the upper cervical spine were the most frequent injuries: 90% affected the axis, 22.5% affected the atlas and 32.5% were

Department of Orthopedics, Peking Union Medical College Hospital, Beijing, China Introduction: Scoliosis, especially thoracic curves, usually causes poor pulmonary function and lower oxygen saturation. In this way, scoliosis may impair sleep breathing quality. Literature regarding on the relationship between scoliosis and sleep breathing quality has not been reported. Study design: A case–control study. Hypothesis: Scoliosis may impair sleep breathing quality. Methods: A total of 62 scoliotic patients with thoracic curve and 25 healthy controls were included. Watch-PAT200 wrist sleep monitor was used to record oxygen saturation (SpO2), sleeping body position and sleep stages. Sleep breathing quality was described with following parameters: (1) pRDI (Respiratory Disorders Index) indicating mean respiratory events per hour of sleep including apnea, hypoxia and respiratory effort related arousal; (2) pAHI (Apnea and Hypopnea Index) expressing the number of apnea and hypopnea per hour of sleep; and (3) mean and minimal SpO2 during sleep. Mann–Whitney U test was used to compare sleep breathing indexes between two groups. Paired-t test was used to compare indexes in different body positions. Results: No difference of age, gender distribution and BMI was found between two groups. Scoliotic patients have statistical significant higher pRDI (median 10.10 vs. 8.65, p = 0.039) and pAHI (median 1.60 vs. 0.72, p = 0.029) than that of control group. The

123

S400 minimal SpO2 of scoliotic patients is lower (median 93 vs. 94%, p = 0.005) while no difference was found in mean SpO2 during sleep. In scoliotic patients, their pAHI were higher when lying on convex side of thoracic curve during sleep than concave side (2.52 vs. 2.27, p = 0.045) while no difference while no such difference was observed in control group. Distribution of sleep stages was same in two groups. Conclusions: Scoliotic patients have more respiratory events, apnea and hypopnea during sleep than control group. The minimal oxygen saturation in scoliotic patients is lower than normal population while no difference in mean oxygen saturation. Sleeping on the convex side of thoracic curve results higher apnea and hypoxia index than concave side. Disclosures: Author 1: none; Author 2: none; Author 3: none; Author 4: none; Author 5: none.;

P132 PERIOPERATIVE COMPLICATIONS IN SPINAL OSTEOTOMY FOR THORACOLUMBAR KYPHOSIS CAUSED BY ANKYLOSING SPONDYLITIS IN 323 PATIENTS: INCIDENCE AND RISK FACTORS

Eur Spine J (2017) 26 (Suppl 2):S335–S405 of patients with perioperative complications was significantly larger than that of patients without perioperative complications (P = 0.047). Patients who underwent two-level PSO (OR 2.9, 95% CI 1.2–7.2, P = 0.033) were found to have an increased likelihood of perioperative complications compared to those who underwent one-level PSO. The number of instrumented vertebra was significantly larger in patients with preoperative complications (P = 0.019). Conclusion: The overall incidence of perioperative complications in spinal osteotomy for thoracolumbar kyphosis secondary to AS was 13.0%. Increased age, larger preoperative GK, two-level PSO and increased number of instrumented vertebra were found to be risk factors probably. Disclosures: Author 1: none; Author 2: none; Author 3: none; Author 4: none; Author 5: none; Author 6: none.

P133 RISK FACTORS FOR NONUNION OF TRANSPSOAS LATERAL LUMBAR INTERBODY FUSION Kotaro Satake, Tokumi Kanemura, Hiroaki Nakashima, Naoki Segi, Jun Ouchida

Bangping Qian, Jichen Huang, Yong Qiu, Bin Wang, Yang Yu, Zezhang Zhu

Department of Orthopaedic Surgery, Konan Kosei Hospital, Konan, Japan

Spine Surgery, Drum Tower Hospital of Nanjing University Medical School, Nanjing, China

Introduction: Although numerous papers have reported the fusion rate of transpsoas lateral lumbar interbody fusion (LLIF), few described the risk factors for nonunion. This retrospective review of a prospective cohort was performed to evaluate the influence of nonunion on clinical outcomes and to identify risk factors for it. Materials and methods: 68 consecutive patients who underwent LLIF (68.6 ± 11.0 years, 25 males and 43 females, 136 segments) were followed for a minimum 2 years postoperatively. All LLIF segments were applied with polyetheretherketone cages packed with allogenic cancellous bone, and were supplemented with bilateral pedicle screws (PS) [open PS for 107 segments and percutaneous PS (PPS) for 29 segments]. For each segment, bone bridge formation inside and outside (including the facets) the cage was evaluated by CT 2 years postoperative, and a segment without any bridge formation was determined to be a nonunion. Participants were classified into two groups according to CT evaluation; group N which contained 1 or more nonunion segments and Group F which contained no nonunion segment. VAS scores and the improvement of the five domains of the Japanese Orthopedic Association Back Pain Evaluation Questionnaire (JOABPEQ) were compared between Groups N and F. The risk factors for nonunion were determined by uni- and multivariate analyses. Results: The nonunion rate was 16.9%. Improvement in the domain of pain-related disorders of JOABPEQ was significantly reduced in Group N compared to F (41.2 vs 71.1%, p = 0.04). Univariate analysis revealed that age (73.0 ± 8.9 vs 68.5 years ± 11.9, p = 0.049), the incidence of previous multiple vertebral fractures (17.5 vs 3.5%, p = 0.03), and PPS usage (39.1 vs 16.8%, p = 0.02) were significantly higher in the nonunion segments. Multivariate analysis identified previous multiple vertebral fractures (OR 6.0, 95% CI 1.3–27.2, p = 0.02) and PPS usage (OR 3.3, 95% CI 1.2–8.9, p = 0.02) as significant risk factors for nonunion. Conclusion: Nonunion of LLIF correlated significantly to the increase in JOABPEQ pain-related disorders. We should be aware of the higher nonunion rate in patients with multiple vertebral fractures and in the segments supplemented with PPS. Disclosures: Author 1: none; Author 2: consultant: NuVasive, Medtronic; Author 3: none; Author 4: none; Author 5: none.

Background/introduction: Several studies have shown acceptable and good radiological and clinical outcomes after spinal osteotomy for AS with thoracolumbar kyphosis. However, the incidence and risk factors of perioperative complications have not been reported in a large series of AS patients who underwent spinal osteotomy for the correction of thoracolumbar kyphosis. Purpose of the study: To describe the incidence of perioperative complications (within 6 weeks after the surgery) in spinal osteotomy for thoracolumbar kyphosis secondary to ankylosing spondylitis (AS) and to investigate the risk factors of these complications. Materials and methods: From April 2000 to February 2017, 323 consecutive AS patients with an average age of 35.5 ± 10.1 years (range 17–71 years) who underwent spinal osteotomy were enrolled. Patients with perioperative complications were identified. Demographic, radiological and surgical data were compared in patients with or without perioperative complications. Perioperative complications were classified into intraoperative and early postoperative (within 6 weeks after the surgery) complications. The complications were further stratified into surgical, anesthesia and general complications. Results: Two hundred and eighty-six consecutive pedicle subtraction osteotomy (PSO) and 37 multiple Smith Petersen osteotomies (SPOs) procedures were performed. Overall, 42 complications were identified in 42 patients (13.0%), among which there were 27 intraoperative complications and 15 early postoperative complications. The surgical, anesthesia and general complications occurred in 35 cases (10.8%), 2 cases (0.6%), and 5 cases (1.5%), respectively. In terms of surgical complications, the rate was 10.8% with 15 cases (4.6%) of neurological complications, 10 cases (3.1%) of vertebral subluxation, 9 cases (2.9%) of dura tears, and 1 case (0.3%) of wound dehiscence. Major complications including motor deficit, inadvertent extubation, acute pulmonary edema, acute pancreatitis, duodenal perforation, and wound dehiscence were observed in 8 cases (2.5%). Patients with perioperative complications were older than those without perioperative complications (P = 0.036). Preoperative global kyphosis (GK)

123

Eur Spine J (2017) 26 (Suppl 2):S335–S405 P134 M-C ANGLE AS A PREDICTOR OF DYSPHAGIA AFTER POSTERIOR FIXATION FOR TRAUMATIC LOWER CERVICAL SPINE INJURY Eiji Takasawa, Yasunori Sorimachi, Yoichi Iizuka, Haku Iizuka, Hirotaka Chikuda Department of Orthopaedic Surgery, Japanese Red Cross Maebashi Hospital, Maebashi, Japan; Department of Orthopaedic Surgery, Gunma University Graduate School of Medicine, Maebashi, Japan

S401 P135 POSTOPERATIVE DECREASE IN INTERVERTEBRAL LORDOSIS AND DISC SPACE DISTRACTION INDUCE NEUROLOGICAL COMPLICATIONS FOLLOWING POSTERIOR LUMBAR INTERBODY FUSION AS REVISION SURGERY Tomoya Yamashita, Shinya Okuda, Tomiya Matsumoto, Takafumi Maeno, Tsuyoshi Sugiura, Motoki Iwasaki, Masayuki Furuya, Hiroyuki Aono Department of Orthopaedic Surgery, Osaka, Japan

Background: Postoperative dysphagia (PoD) is an unwanted complication after anterior spinal decompression and fusion or posterior occipitocervical (O-C) fusion and may lead to an unfavorable course after surgery. PoD has been reported to be caused by iatrogenic esophageal perforation and postoperative cervical malalignment; however, there is little information regarding the prevalence of PoD after lower cervical spine injury (C5–C7). In cervical spine injury, it is difficult to select an appropriate cervical angle for treatment because an evaluation of the native cervical alignment is not possible after trauma. Previous work has described the use of a new indicator, the M-C angle, to ascertain the degree of adequate craniocervical alignment in O-C fusion for patients with rheumatoid cervical disorders. The M-C angle is formed by the McGregor line and the cervical spinal column line (i.e., a straight line connecting the center of the vertebral bodies of C2 and C7). The purpose of this study was to reveal the incidence of PoD and confirm the impact of the M-C angle on dysphagia after posterior fixation for the lower cervical spine injuries. Methods: Twenty-one patients who had undergone posterior fixation for lower cervical spine injuries were retrospectively reviewed. Injuries were classified using the AOSpine classification, the Allen scheme, and the subaxial cervical spine injury classification (SLIC) score. Cervical alignment parameters were analyzed by M-C, occipito-C2 (O-C2) and C2–7 angles. The Bazaz dysphagia score was used to evaluate the severity of PoD. Patients were divided into PoD (group D) and no PoD (group N) groups. Statistical analyses were performed using the Mann–Whitney U test, with significance set at p \ 0.05. Results: All patients had undergone posterior fixation with pedicle screws and/or lateral mass screws. After surgery, two patients (9.5%) complained of PoD (one severe, one moderate), and percutaneous endoscopic gastrostomy was performed for the patient with severe dysphagia. There was no statistical difference in the AOSpine type, the Allen classification, SLIC score, O-C2 angle or C2–7 angle between the two groups. We found that the M–C angles in group D were less than 90, whereas those in group N were maintained over 90 (p = 0.023). Conclusion: The M-C angle has a substantial impact on dysphagia after lower cervical posterior fusion, and appears to be a universal alignment parameter for measuring O-C sagittal alignment as compared with O-C2 and C2-7 angles. Regardless of the injury type, the M-C angle is easily measured during surgery and can be a practical indicator to avoid PoD. Disclosures: Author 1: none; Author 2: none; Author 3: none; Author 4: none; Author 5: none.

Objective: To clarify the risk factors, especially the effects of nerve stretching, for postoperative neurological complications in posterior lumbar interbody fusion (PLIF) without excessive nerve retraction by bilateral total facetectomy as revision surgery. Summary of background data: Although the main concern with revision lumbar surgery is the possibility of neurological complications, no study has evaluated the risk factors for neurological complications following PLIF as revision surgery. Methods: Between 2005 and 2014, 50 consecutive patients (31 males, 19 females) underwent revision PLIF for recurrent stenosis or recurrent herniation at the previously decompressed segments. New or worsening leg pain or motor loss, which was observed until five days after revision PLIF, was classified as a neurological complication. Radiological evaluations associated with the magnitude of nerve stretching included the following measurements on plain radiographs: (1) anterolisthesis at flexion before surgery (a, in mm) and just after surgery (b, in mm); (2) reduction of slippage (a–b, in mm); (3) intervertebral lordosis at the neutral position before surgery (c, in degrees, lordosis is a positive value) and just after surgery (d, in degrees); (4) decrease in intervertebral lordosis (c–d, in degrees); (5) posterior disc height at the neutral position before surgery (e, in mm) and just after surgery (f, in mm); and (6) distraction of the posterior disc space (f–e, in mm). The patients were divided into two groups: patients with neurological complications (NC group) and patients without neurological complications (non-NC group). Results: Sixteen patients (32%) had symptomatic neurological complications. The NC group had a significantly higher percentage of progressive spondylolisthesis than the non-NC group (69 vs 26%, p \ 0.005). Decrease in intervertebral lordosis was significantly greater in the NC group than in the non-NC group (0.8 vs. -1.5, p = 0.03). Distraction of the posterior disc height was significantly greater in the NC group than in the non-NC group (5.0 vs. 2.6 mm, p \ 0.005). Neurological complications were seen in 100% of the patients with decrease in intervertebral lordosis[3 and distraction of the posterior disc height [3 mm. In contrast, neurological complications were not seen in 88% with both decrease in intervertebral lordosis \3 and distraction of the posterior disc height \3 mm. Conclusions: Preoperative anterolisthesis, decrease in intervertebral lordosis, and distraction of the posterior disc height appear to be risk factors for neurological complications following revision PLIF. In revision PLIF, surgeons should achieve segmental lordosis without excessive disc height distraction.

123

S402

Eur Spine J (2017) 26 (Suppl 2):S335–S405

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.

P137 IS CHIROPRACTIC CARE ASSOCIATED WITH THE DEVELOPMENT OF CAUDA EQUINA SYNDROME? A CASE-CROSSOVER STUDY

NONOPERATIVE TREATMENT

Pierre Coˆte´, Eleanor Boyle, J. David Cassidy, Stephanie Choi, Cesar Hincapie´

P136 DOES DISC WEDGING CONTRIBUTE TO THE CLINICAL RESULTS OF BRACE TREATMENT IN ADOLESCENT IDIOPATHIC SCOLIOSIS? Bangping Qian, Hao Liu, Yong Qiu, Zezhang Zhu, Jack Cheng, TP Lam, Bobby Ng Spine Surgery, Drum Tower Hospital of Nanjing University Medical School, Nanjing, China Summary: Disc Wedging was previously reported as a contributory factor for progression of idiopathic scoliotic (IS) curves. And initial disc wedging percentage has been considered as a new prognostic factor of curve progression. However, it remains unknown whether there is a correlation between the initial disc wedging and the outcome of brace treatment. Introduction: The objective of this study is to investigate whether the initial disc wedging percentage could serve as an important factor in predicting the curve regression, stabilization, or progression during bracing in adolescent idiopathic scoliosis (AIS). Methods: A retrospective review was performed on the patients with standardized bracing treatment from January 2009 to July 2014. Standardized SRS criteria for bracing were utilized in the case selection. The demographic data, growth status, and Cobb angle of each visit were recorded. The initial disc wedging percentage was calculated. Patients were divided into group A (progressive, C6) and group B (non-progressive, \6) based on their final bracing outcome. Differences between two groups were analysis in terms of Unpaired-t text and the correlation between the bracing outcome and the parameters before bracing, including anthropometric measurements (BMI), curve magnitude, and disc wedging percentage were evaluated using univariate analysis and regression analysis. Results: Sixty-one patients were included. There were 15 (24.59%) girls in group A and 46 (75.41%) girls in group B, respectively. Significant differences between progressive and non-progressive groups were found in terms of BMI (17.4 ± 2.6 vs. 15.3 ± 3.2, P \ 0.05), initial disc wedging percentage (31.7 ± 18.3 vs. 63.4 ± 22.4%, P = 0.001), and Cobb angle (29.4 ± 6.7 vs. 25.5 ± 5.6, P \ 0.05). As revealed by the multiple logistic regression analysis, initial disc wedging percentage (P = 0.001) was identified as an independent risk factor in curve progression in AIS girls, during the total duration of bracing treatment. Conclusion: The initial disc wedging percentage is a new independent risk factor in the curve progression during the duration of bracing treatment. The evaluation of initial disc wedging percentage before bracing may help to predict the clinical results of brace treatment. Disclosures: Author 1: none; Author 2: none; Author 3: none; Author 4: none; Author 5: none; Author 6: none; Author 7: none. Epidemiology

123

Faculty of Health Sciences, University of Ontario Institute of Technology, Oshawa, Ontario, Canada Background: Cauda equina syndrome (CES) is a serious neurological disorder and surgical emergency that can lead to permanent paraplegia, and bowel, bladder or sexual dysfunction. Case reports suggest that chiropractic manipulation is a cause of CES. We investigated the association between chiropractic care and CES; we compared it to the association between primary physician care and CES. Methods: We conducted a population-based case-crossover study. The source population included all adults aged 18 years and older who were residents of the province of ON, Canada. This study used three data sources: (1) the Discharge Abstract Database; (2) the Ontario Health Insurance Plan billing database; and (3) the Ontario Registered Persons Database. Cases had a diagnosis of CES as defined by either 334.6 in ICD9-CM or G 83.4 in ICD10-CA and were treated surgically between 1994 and 2004. We used a time-stratified approach to randomly select four control periods during the year prior to hospital discharge. The exposures included all ambulatory encounters made to chiropractors, primary care physicians using the specialty codes submitted to the Ontario Health Insurance Plan. We tested different hazard periods (1 day, 3 days, 1 week, 2 weeks, and 1 month before the index date. Exposure occurred if any chiropractic or primary care physician visits were recorded during the designated hazard periods. Conditional logistic regression was used to estimate the association between CES after chiropractor and primary care physician visits. Separate models were built using different prespecified hazard periods, stratified by age and gender. Results: We identified 720 cases over a 10-year period. 52.6% were male and the mean age was 44.2 years (SD 14.5). 20.7% of cases had visited a chiropractor within 30 days of the index date, while 72.4% had visited a PCP within that time. We found positive associations between chiropractic care and CES for all exposure periods. For example, individuals \60 years were 4 times more likely to have visited a chiropractor within 2 weeks of their CES [OR 4.16; 95% CI (3.06, 5.65)]. Similarly, we found positive associations between primary physician care and CES for all exposure periods, age-groups and gender. For example, individuals \60 years were 6 times more likely to have visited a physician within 2 weeks of their CES [OR 6.55; 95% CI (5.29, 8.10)]. Overall, the associations were stronger for physician care than for chiropractic care. Conclusions: Our results suggest that the development of CES is associated with both chiropractic and physician care. The increased risks of CES associated with chiropractic and physician care is likely due to patients with low back pain secondary to lumbar disc herniation seeking care for their pain before the clinical diagnosis of CES occurring. We found no evidence of excess risk for CES associated with seeking care from chiropractors, when compared with primary care physicians.

Eur Spine J (2017) 26 (Suppl 2):S335–S405 Disclosures: Author 1: grants/research support: (1) Canada Research Chair Program, Canadian Institutes of Health Research; (2) Ontario Ministry of Finance; (3) Canadian Chiropractic Association; (4) Canadian Chiropractic Protective Association; (5) Ontario Trillium Foundation, consultant: (1) Expert witness in court cases, Canadian Chiropractic Protective Association, other financial report: Honorarium and travel expenses to teach at Eurospine TFR Reserach course and Grant Panel; Author 2: employee: University of southern Denmark; Author 3: grants/research support: Canadian Chiropractic Protective Association; Author 4: none; Author 5: grants/research support: Canadian Institutes of Health Research; Canadian Chiropractic Research Foundation.

P138 EUROPEAN COMPARISON OF SPINAL SURGERY HOSPITALIZATIONS FROM 2010 TO 2013 ACCORDING TO PATIENT PROFILES Lucie de Le´otoing, Ce´cile Blein, Pascale Brasseur, Alexandre Vainchtock HEVA, Lyon, France Objectives: This study was performed to compare hospitalizations for spinal surgery development across France, Spain, Germany and Belgium from 2010 to 2013 and to analyze patient’s characteristics. Methods: A retrospective analysis was conducted from hospital databases PMSI for France, CMBD for Spain, SHI for Germany and RHM for Belgium between 2010 and 2013. All spinal surgery hospitalizations were collected based on procedure codes according to the respective classification of each country (CCAM for France, ICD-9 for Spain and Belgium, and OPS for Germany). The mapping of ICD-9 and OPS codes was undertaken from the French procedures. Standardization of rates of spinal surgery patients were based on age from the EU population. Results: In 2013, crude rates of hospitalized patients with spinal surgery were 6.43 per 10,000 in Spain, 18.95 per 10,000 in France, 66.27 per 10,000 in Germany and 67.50 per 10,000 in Belgium. All countries experienced an increase of this number of patients from 2010 to 2013: ?14% in Spain, ?17% in France and ?18% in Germany except for Belgium with a decrease of 13%. The gender distribution was similar between all countries except for Belgium, with slightly more women treated in Spain and Germany: 52% versus 50% in France. However, in Belgium the number of man treated is more important (64%). Mean age was lower in Spain (53 ± 16 years) France (54 ± 17 years) and Belgium (52 years) than in Germany (59 ± 15 years); the (70–80) year group was overrepresented in Germany (24% of patients versus 15% in France and Spain) to the detriment of the (30–40) year group (6% of patients in Germany versus 15% in France and Spain). The age standardized rates of spinal surgery patients were higher in Germany (60.55 per 10,000) and Belgium (67.36 per 10,000) than in France (19.91 per 10,000) and Spain (6.53 per 10,000). Conclusion: Between 2010 and 2013, spinal surgery was marked by a progression of more than 14% in each country except for Belgium with a decrease of 13%. The standardized rate of spinal surgery patients varied significantly between the 4 countries, Germany and Belgium having the highest. Disclosures: Author 1: employee: HEVA; Author 2: grants/research support: Medtronic; Author 3: stock/shareholder: Medtronic plc, employee: Medtronic plc; Author 4: consultant: HEVA.

S403 P139 CLINICAL ANALYSIS OF MENTAL HEALTH SCORE IN PATIENTS WITH LOW BACK PAIN USING JOABPEQ AND PAINDETECT Akihiko Hiyama, Masahiko Watanabe Department of Orthopaedic Surgery, Tokai University School of Medicine, Kanagawa, Japan Background: In patients with low back pain (LBP) who combine serious psychosocial factors with clinical findings of pain, there is a possibility that the psychosocial factors modify the pain. Therefore, the purpose of this study is to investigate the relationship between the Japanese Orthopaedic Association Back Pain Evaluation Questionnaire (JOABPEQ) mental health score and the characteristics of LBP, with a focus on lower limb symptoms. Methods: 331 patients included in this study. All patients between the ages of 20 and 79 were asked to complete a set of questionnaires including the Japanese version of the painDETECT (PDQ- J), a pain Numeric Rating Scale (NRS), the JOABPEQ, and the Short Form 36 (SF-36). Based on their JOABPEQ mental health scores, the patients were divided into two groups for the comparative study: those with JOABPEQ mental health score of \50 were classified as low score group, and those scoring [50 were classified as the high score group. To identify any differences between the two groups, age, sex, PDQ-J score, NRS score, duration of symptoms, percentage of pain components, percentage of lower limb symptoms and self-reported general health were compared between the two groups. Results: 196 patients (59.2%) were classified into the low score group, and 135 (40.8%) into the high score group. The mean PDQ- J and NRS scores and percentage of LBP patients with neuropathic pain and lower limb symptoms were higher in the low score group. We also evaluated the relationship between lower limb symptoms and JOABPEQ mental health scores in 49 patients with neuropathic LBP. The results show that there was no significant difference between the incidence of lower limb symptoms and JOABPEQ mental health scores in patients with neuropathic LBP. Conclusion: We found that in patients with neuropathic LBP, psychological factors may modify pain intensity and may lead to an exaggerated or histrionic presentation of the pain. Disclosures: Author 1: none; Author 2: none.

P140 COST-EFFECTIVENESS-ANALYSIS TLIF VERSUS PLIF Inge Caelers, Kim Rijkers, Henk van Santbrink, Suzanne de Kunder Department of Neurosurgery, Maastricht, The Netherlands Title: Cost-effectiveness-analysis TLIF versus PLIF. Introduction: Degenerative spine disease leading to neurological symptoms is progressively treated using instrumented spine surgery, which leads to an increase in health care costs. The two most frequently performed instrumented procedures for the lumbar spine in The Netherlands are transforaminal lumbar interbody fusion (TLIF) and posterior lumbar interbody fusion (PLIF), their usage depends on the surgeons’ preference. More knowledge on cost-effectiveness of these procedures is important to be able to constrain costs in the future. Objective: To describe aspects of a cost-effectiveness-analysis and to analyze the cost-effectiveness of TLIF and PLIF based on existing literature.

123

S404 Methods: Based on the criteria of The National Health Care Institute of the Netherlands we have critically reviewed existing literature on cost-effectiveness of TLIF and PLIF. We conducted a search using i.e. Pubmed and Embase for studies reporting on TLIF or PLIF, lumbar spondylolisthesis or lumbar instability and cost. First, hits were screened on title and abstract, followed by full text screening on all eligibility criteria. A systematic review was performed to compare cost-effectiveness. Differences in cost-items for calculation of costeffectiveness and use of ‘charges’ and ‘costs’, were taken into account. Results: 266 studies were identified, 16 studies were included (13 TLIF and 3 PLIF). Because of heterogeneity among studies, costs were presented in ranges instead of means. Calculated healthcare related costs for TLIF and PLIF studies were higher when ‘charges’ were used (TLIF $25,539–$47,029, PLIF $14,081–$86,112) than when ‘costs’ were used (TLIF $25,871–$39,314, PLIF $29,699). Costs for productivity loss and cumulative QALY-gain were determined in TLIF studies only ($6,717–$36,537 and 0.67–0.86 QALY). Cost-effectiveness was determined in two TLIF studies; $46,475/ QALY and $66,914/QALY. Conclusion: Many different factors have to be taken into account in order to perform a cost-effectiveness-analysis. For this reason, healthcare-economists are valuable members of a team conducting cost-effectiveness research. Based on our review, we conclude that information on cost-effectiveness of instrumented spine surgery is sparse. The little information that is available indicates that this type of treatment is expensive, especially when compared to other types of surgery carried out in a similar patient group, for instance total hip replacement. More research is necessary to elucidate whether instrumented spine surgery is cost-effective, and to designate whether differences in techniques result in differences in cost-effectiveness. Disclosures: Author 1: none; Author 2: none; Author 3: grants/research support: Academic Grant, Academic Medical Centre Maastricht, the Netherlands.

MEDICAL ECONOMICS

P141 AN ASSESSMENT OF ANNUAL COSTS OF PATIENTS HOSPITALIZED FOR SPINAL TUMORS IN FRANCE: ANALYSIS USING THE PMSI DATABASE Lucie de Le´otoing, Je´roˆme Fernandes, Charle`ne Tournier, Baptiste Jouaneton, Alexandre Vainchtock HEVA, Lyon, France Objective: To assess hospitalization costs of patients with spinal tumors in France. Method: All hospital stays with spinal tumors were extracted from the PMSI-MCO 2012 and 2014 database (French Medical Information System Program- Medicine, Surgery, Obstetric): ICD-10 codes C41.2 or C41.4 or C41.8 as principal/related or significantly associated diagnosis for primary tumors; association of ICD-10 codes C79.5 and M49.5 as principal/related or significantly associated diagnosis or M49.5 alone as principal/related diagnosis for secondary tumors. Patients were followed during 1 year from their first stay (e.g. March

123

Eur Spine J (2017) 26 (Suppl 2):S335–S405 2012/March 2013; October 2014/October 2015). An algorithm and a medical review excluded non spinal tumor related stays. Associated costs during this period were added up: a total annual cost (‘‘burden’’) as well as a mean annual cost per patient were estimated. Valuation was performed considering French official tariffs for 2012–2015. Results: In 2012, 9415 stays were extracted and considered as directly related to spinal tumors corresponding to 3284 patients. Patients were 66 ± 19 years old on average, 55% were female. 94% of stays occurred in public hospitals. 56% of stays were ambulatory (length of stay, LOS 0 day) and mean LOS associated with full hospitalizations was 11 ± 14 days. Spine fracture management, chemotherapy and radiotherapy accounted for 30, 30 and 28% of stays, respectively. Overall, the economic annual burden of hospitalizations for spinal tumor reason was €26 million; expensive drugs and implants funded in addition to DRGs (diagnosis related groups) accounted for 6% of this burden. The mean annual cost was €7817 ± 8858 per patient. Results from 2014 show an increase in the number of stays, patients and associated costs. Conclusion: Spinal tumors are mainly managed in public hospital. Hospital-related costs associated with spinal tumors are high. New spinal tumors treatment modalities could decrease these costs. Disclosures: Author 1: employee: HEVA; Author 2: consultant: MSD; Author 3: employee: HEVA; Author 5: consultant: HEVA.

P142 COST OF LOW BACK PAIN: RESULTS FROM A REGISTER STUDY IN SWEDEN Emma Jonsson, Gylfi Olafsson, Peter Fritzell, Olle Ha¨gg, Fredrik Borgstro¨m Quantify Research, Stockholm, Sweden; LIME/MMC, Karolinska Institutet, Stockholm, Sweden; Futurum Academy, Jo¨nko¨ping, Sweden; Capio St Go¨ran, Stockholm, Spine Center Go¨teborg, Gothenburg, Sweden Objectives: The objective of this study was to estimate the societal costs of low back pain with/without radiant leg pain (LBP). Methods: Patients visiting healthcare providers and/or having sick insurance benefits with a registered LBP diagnosis in the Western region of Sweden (Va¨stra Go¨taland) in 2008–2011 were identified in national registers and an administrative database. Patient level data from registers including inpatient and outpatient care, drug prescriptions, socioeconomics and social security were extracted. Direct healthcare costs and indirect costs in terms of sick leave and early retirement were summarised over time periods called LBP episodes. An LBP episode started with a LBP related healthcare contact or work absence and ended when six months had elapsed without a LBP related healthcare contact or work absence. Results: More than 130,000 patients were identified with over 160,000 LBP episodes. Mean age at episode start was 52 years and 59% of patients were female. The mean total cost per LBP episode was estimated at €2,761 (SD 12,809). The mean duration of episodes was 51 days (SD 140). There was a sharp increase in total cost at the first month after LBP episode starts and a marked decrease from the second month after episode start. Total cost levelled off at a higher level during the 2 years after episode start compared with the 2 years before episode start. The total economic burden of LBP in Sweden in all LBP episodes that started in 2011 was estimated at €739 million,

Eur Spine J (2017) 26 (Suppl 2):S335–S405

S405

or €78 per capita (Table 1). Indirect costs in terms of sick leave and early retirement accounted for 65% of total cost. Non-surgical inpatient care (€13 per capita) accounted for almost half of the direct costs, followed by medical visits (€10 per capita). Spine surgery (€8 per capita) corresponded to 10% of the total economic burden. Conclusion: LBP has an apparent impact on the overall resource use and work loss. The results indicate that there is a high short term cost increase at the beginning of an LBP episode, but also that the costs decrease in the long term after the LBP symptoms have come to clinical attention. Disclosures: Author 1: grants/research support: Medtronic; Author 2: consultant: Medtronic; Author 3: none; Author 4: none; Author 5: consultant: Medtronic.

123