Oral Presentations

OP01: Degenerative-Cervical 1

A001: Pain Is the Recovery Priority for Patients With Degenerative Cervical Myelopathy

Benjamin Davies¹, Oliver Mowforth¹, Iwan Sadler², Bizhan Aarabi³, Brian Kwon⁴, Shekar Kurpad⁵, James Harrop⁶, Jefferson Wilson⁷, Robert Grossman⁸, and Michael Fehlings⁷

¹University of Cambridge, Cambridge, UK

²Myelopathy.org

³University of Maryland, College Park, MD, USA

⁴University of British Columbia, Vancouver, British Columbia, Canada

⁵Medical College of Wisconsin, Milwaukee, WI, USA

⁶ Thomas Jefferson University Hospital, Philadelphia, PA, USA

⁷University of Toronto, Toronto, Ontario, Canada

⁸ Houston Methodist Hospital, Houston, TX, USA

Introduction: Degenerative cervical myelopathy (DCM) is a very common and disabling condition caused by chronic compression of the cervical spinal cord from surrounding degenerative changes of the spine. At present the only treatment available is surgical decompression. While this is able to stop disease progression, recovery after surgery often remains incomplete and many patients are left with life-long disability. Individuals with DCM suffer among the lowest quality of life scores of all chronic diseases, yet their suffering is poorly understood. The present study therefore aimed at establishing relevant functional domains and recovery priorities for DCM sufferers. Material and Methods: Functional domains were established as part of qualitative interviews and a consensus process. A cross-sectional survey was advertised via a DCM charity (Myelopathy.org) and Google AdWords. Individuals were asked about their disease characteristics, including limb pain (visual analogue scale) and functional disability (patient-derived modified Japanese Orthopaedic Association score). They were also asked to rank the established recovery domains (arm and hand function, walking, upper body/trunk function, sexual function, elimination of pain, sensation and bladder/bowel function) in order of priority. Priorities were analyzed as the modal first priority and mean ranking. The influence of demographics on selection was analyzed, with significance P < .05. Results: Of 659 survey responses obtained, 481 were complete. Overall, pain was the most popular recovery priority (39.9%) of respondents, followed by walking (20.2%), sensation (11.9%), and arm and hand function (11.5%). Sexual function (5.7%), bladder and bowel (3.7%), or trunk function (3.5%) were chosen less frequently. When considering the average ranking of symptoms, while pain remained the priority (2.6 ± 2.0), this was closely followed by walking (2.9 ± 1.7) and arm/hand function (3.0 ± 1.4). Sensation ranked much lower (4.3 ± 2.1). With respect to disease characteristics, overall pain remained the recovery priority, with the exception of patients with greater walking impairment (P < .005) who prioritized walking, even among patients with lower pain scores. Conclusion: This is the first study investigating patient priorities in DCM. The patient priorities reported provide an important framework for future research and will help ensure that it is aligned with patient needs.

A002: Surgical Outcome of Anterior Versus Posterior Approach in the Treatment of Cervical Spondylotic Myelopathy: A Propensity-Score-Matched Analysis

Hui Yu Koh¹, Kenneth Cheung¹, and Kenny Kwan¹

¹The University of Hong Kong, Hong Kong

Introduction: Surgical decompression is the only known effective intervention to relieve cervical spondylotic myelopathy (CSM) patients of their symptoms and can be broadly divided into the anterior and posterior approaches. Several studies have compared these 2 approaches. However, no definitive conclusion has been drawn due to methodological bias and inappropriate comparisons. The aim of this study was to compare the surgical outcome in CSM patients undergoing anterior versus posterior approach using a propensity score-matched analysis. Material and Methods: A retrospective cohort study was performed, and medical records were reviewed. Inclusion criteria were patients with a clinical and appropriate radiological diagnosis of CSM with no prior surgical intervention who underwent surgical decompression in our institution between January 2005 and December 2015, aged above 18 years, and a minimum follow-up of 2 years. Exclusion criteria were loss to follow-up, revision surgery, and incomplete clinical data set in the medical records. A 1:2 propensity score matching of the anterior and posterior group was performed. Matching criteria were age at surgery, number of spinal levels involved, preoperative mJOA (modified Japanese Orthopaedic Association) score, preoperative Nurick score, C2-7 SVA (sagittal vertical axis), and C2-7 lordosis. A nearest-neighbor method within 0.02 standard deviations of the calculated score without replacement was performed to adjust for the patients’ baseline characteristics. All statistical analyses were performed using SPSS Statistics software version 24.0 (IBM). A significance level of <.05 was assumed to be statistically significant. Results: A total of 242 patients matched the inclusion criteria and underwent surgical intervention at our institution. Eighty-seven patients were excluded due to incomplete records. The study group thus comprised of 155 patients (43 anterior and 112 posterior surgery patients). A total of 82 patients were matched, consisting of 32 anterior and 50 posterior surgery patients. There was no significant difference in for all covariates after propensity score matching. Overall, both anterior and posterior surgical approach groups did not differ in postoperative Nurick scores, mJOA scores, length of hospital stay, blood loss, and complication rates. Patients who underwent anterior surgery had better maintenance of C2-7 lordosis compared with the posterior group (−8.37 ± 12.1° vs −1.87 ± 13.4°, P < .05). In addition, recovery rate (62.1 ± 26.9% and 49.8 ± 32.3%, respectively) was higher in the anterior group as compared to the posterior group, but this was not found to be statistically significant. Further subgroup analysis based on the cervical sagittal alignments showed that patients with C2-7 SVA and C2-7 lordosis misalignment had better surgical outcomes in terms of postoperative mJOA score (14.7 ± 1.66 vs 13.2 ± 1.92, P < .05) and 1.35 times reduction in relative risk of reoperation (P = .031) if they underwent anterior surgery compared with posterior surgery. Conclusion: Our findings show that there was no difference in terms of neurologic outcomes between anterior and posterior approaches, which is in line with some of the findings of the recent meta-analyses. However, we found that patients who had undergone anterior surgery had better maintenance of their cervical sagittal alignment. Moreover, patients who had preoperative sagittal malalignment had better clinical outcomes and reduced reoperation rates if they underwent anterior surgery.

A003: Treatment of Cervical Spinal Canal Stenosis With Myelopathy With Corpectomy and Expandable Cage Implantation: 5 Years Outcomes

Vadim Byvaltsev¹, Andrei Kalinin¹, Marat Aliev¹, Evgenii Belykh¹, Ivan Stepanov¹, Bakhyt Aglakov¹, and Bobur Yussupov¹

¹Irkutsk State Medical University, Irkutsk, Russian Federation

Introduction: Cervical myelopathy is one of the severe pathological conditions leading to a significant deterioration in the quality of life, a decrease in the ability to work, and disability of patients. The development of clinical symptoms of cervical myelopathy is based on congenital or acquired narrowing of the vertebral canal. To date, an optimal method for treating the stenosis of the cervical spinal canal is the anterior decompression of the spinal cord with interbody fusion by various materials. According to the published data, in a number of cases complications are associated with nontransplantation, repeated compression of neural structures due to migration, difficulty in preparing the required implant size for the restoration of cervical lordosis with prolonged decompression of the spinal cord. The purpose of our study was to analyze treatment results of patients with the degenerative cervical spinal canal stenosis with myelopathy after corpectomy and telescopic expandable cage implantation. Materials and Methods: Authors reviewed 60 patients (35 males, 25 females), mean age 47.3 years (range: 39-62 years), that were available at 5-year follow-up after surgery. Neck Disability Index (NDI), Visual Analogue Scale (VAS), Macnab and Nurick scales were used to assess the preoperative status and clinical outcomes. X-rays and MSCT (multislice computed tomography) scans were used to assess the fusion. Results: All patients had failed conservative treatment within the mean period of 10 months (6-18 months) and proceeded with surgery. One-level corpectomy was performed in 29 (48.3%) patients, 2 levels in 22 (36.7%), and 3 levels in 9 (15%). Significant decrease of pain was revealed after surgery in cervical spine and in upper extremities (mean VAS decreased from 54 to 14, P = .02; and from 62 to 10, P = .015 consequently). Mean NDI decreased significantly at 1-year follow-up from 68 to 16, P = .03, and did not change at the 5-year follow-up. According to the Nurick scale 5 years after surgery full regress of symptoms was achieved in 24 (40%), 29 (48.3%) patients improved, 6 (10%) unchanged, and 1 patient get worse (1.7%). Macnab results were excellent in 19 (31.6%), good in 30 (50%), fair in 10 (16.7%), poor in 1 (1.7%) cases. Fusion was achieved at 1-year follow-up in 48 (80%) cases, at 2 years in 53 (88.3%), at 3 years in 56 (93%), at 4 years in 57 (95%), at 5 years in 58 (97%) patients. MRI (magnetic resonance imaging) showed complete regression of cord lesion in 14 (23.3%) and partial regression in 32 (53.3%) cases. Electromyography showed improvement of F-wave and M-response in upper extremities in 41 (68.3%) patients. There were no complications associated with corpectomy or implant itself but 3 postoperative wound infections and 2 hematomas which were successfully treated conservatively. Conclusion: This retrospective review showed effectiveness of anterior decompression and expandable cages implantation in the patients with degenerative cervical spinal canal stenosis and myelopathy. The robust fixation allowed achieving fusion in the 97% of patients after single and multilevel corpectomies at 5-year follow-up without significant postoperative complications.

A004: Do Cervical Spine Patients Recall Their Preoperative Status? A Cohort Study of Recall Bias in Patient-Reported Outcomes

Ilyas Aleem¹, Bradford Currier², Heidi Poppendeck², Paul Huddleston², Jason Eck³, John Rhee⁴, Mohamad Bydon², Brett Freedman², and Ahmad Nassr²

¹University of Michigan, Ann Arbor, MI, USA

²Mayo Clinic, Rochester, MN, USA

³Center for Sports Medicine and Orthopaedics, Chattanooga, TN, USA

⁴Emory Spine Center, Emory University, Atlanta, GA, USA

Introduction: Recall bias is a well-known source of systematic error in clinical research. The accuracy of patient recall following cervical spine surgery, however, remains unknown. We sought to characterize the accuracy of patient recollection of preoperative symptoms after cervical spine surgery. Material and Methods: Consecutive patients undergoing cervical spine surgery for myelopathy or radiculopathy were enrolled. Neck and arm numeric pain scores and Neck Disability Indices were recorded preoperatively. Patients were asked to recall their preoperative status at both short (<1 year) and long-term (≥1 year) follow-up. Actual and recalled scores were compared using paired t-tests and relations were quantified using Pearson correlation coefficients. Multivariable linear regression was used to identify factors impacting recollection. Results: Seventy-three patients with a mean age of 58.2 years were included. Compared to their preoperative scores, patients showed significant improvement in neck pain (mean difference [MD] = −2.9, 95% CI = −3.5 to −2.3), arm pain (MD = −3.4, 95% CI = −4.0 to −2.8), and disability (MD = −12.4%, 95% CI = −16.9 to −7.9). Patient recollection of preoperative status was significantly more severe than actual for neck pain (MD = +1.5, 95% CI = 0.8 to 2.2), arm pain (MD = +2.3, 95% CI = 1.6 to 3.0), and disability (MD = +5.8%, 95% CI = 2.4 to 9.2). Moderate correlation between actual and recalled scores with regard to neck (r = 0.41), arm (r = 0.50) pain, and disability (r = 0.67) was seen. This was maintained across age, gender, and time between date of surgery and recollection. Over 30% of patients switched their predominant symptom on recall. Conclusion: Relying on patient recollection does not provide an accurate measure of preoperative status following cervical spine surgery. Prospective and not retrospective collection of patient-reported outcomes remains the gold standard to measure and interpret outcomes following cervical spine surgery. Recall bias has the potential to affect patient satisfaction and requires further study.

A005: Minimum Clinically Important Difference and Patient Acceptable Symptom State of Japanese Orthopaedic Association Score in Degenerative Cervical Myelopathy Patients

So Kato¹, Yasushi Oshima¹, Yoshitaka Matsubayashi¹, Yuki Taniguchi¹, Sakae Tanaka¹, Katsushi Takeshita²

¹University of Tokyo, Tokyo, Japan

²Jichi Medical University, Shimotsuke, Japan

Introduction: Despite the worldwide popularity of modified Japanese Orthopaedic Association score (mJOA), the original Japanese Orthopaedic Association (JOA) score is still commonly used in East Asian countries, including Japan. However, unlike mJOA score, the psychometric properties of JOA score remain poorly understood. The objective of the present study was to elucidate the psychometric properties of the original JOA score, including the minimum detectable change (MDC), minimum clinically important difference (MCID), and patient accepted symptom state (PASS). Material and Methods: We retrospectively reviewed a consecutive series of laminoplasty for degenerative cervical myelopathy patients in a single academic institution. Pre- and postoperative JOA scores were collected, and the recovery rate was calculated. The patients were also asked to answer an anchor question analyzing their postoperative health transition used for the MDC and another question assessing the patient satisfaction used for the MCID and PASS. Anchor-based methods were used to determine the cutoff values by a receiver operating characteristic (ROC) curve analysis. Results: A total of 101 patients were included in the analysis. The mean preoperative JOA score was 10.3 (standard deviation [SD]: 2.4), and the mean postoperative JOA score was 13.4 (SD: 2.5). The mean recovery rate was 44%. A total of 68% of the patients admitted that their health condition was at least “somewhat better” than their preoperative condition, and 66% were at least “somewhat satisfied” with the treatment results. Based on ROC curve analyses, the MDC and MCID for JOA score in degenerative cervical myelopathy patients were calculated to be 2.5. The PASS was estimated to be 14.5, and the MCID for JOA recovery rate was 52.8%. Conclusion: The MDC, MCID, and PASS for JOA score for degenerative cervical myelopathy patients were reported by anchor-based ROC curve analyses.

A006: Epidemiology and Outcomes of Neck Pain Following Surgery for Cervical Radiculopathy

Mark A. Maclean¹, Ayoub Dakson¹, Fred Xavier¹, Sean D. Christie¹, and CSORN Investigators²

¹Dalhousie University, Halifax, Nova Scotia, Canada

²Canadian Spine Outcomes Research Network, Toronto, Ontario, Canada

Introduction: Degenerative cervical radiculopathy (DCR) is commonly associated with severe neck pain, in addition to arm pain and moderate levels of disability. Many studies have demonstrated improved arm pain following surgery; however, axial neck pain is generally not felt to improve. The purpose of this study was to determine whether surgery for cervical radiculopathy improves axial neck pain. Materials and Methods: This retrospective study utilized data from the Canadian Spine Outcomes and Research Network (CSORN) registry for patients who received surgery for degenerative cervical radiculopathy. Subgroups were comprised of patients that underwent 1-level, 2-level, 3-level ADCF (anterior cervical discectomy and fusion) or cervical disc arthroplasty (CDA). The primary outcome was 12-month postoperative reduction in Visual Analogue Scale for Neck Pain (VAS-NP). Secondary outcomes included Neck Disability Index (NDI), Visual Analogue Scale for Arm Pain (VAS-AP), Short-Form Physical Health Composite Scale (SF-36-PCS), and Mental Health Composite Scale (SF-36-MCS). Results: We identified 79 patients with cervical radiculopathy: 1-level ACDF (38%), 2-level ACDF (35%), 3-level ACDF (8%), and CDA (19%). Preoperative clinical outcome scores were similar for all patients, regardless of procedure. Mean VAS-NP and VAS-AP improved by 3.6 ± 2.7 and 4.4 ± 3.2 points, respectively. Mean change in VAS-NP, VAS-AP, SF-36-MCS, SF-36-PCS, and NDI were not statistically different across surgical groups (for all outcome measures, P > .05). Although VAS-AP was significantly reduced in all surgical groups (P < .05), only patients undergoing 2-level ACDF and CDA demonstrated significant improvement in VAS-N (P < .001). Mean change in VAS-NP was not significant for 1-level (P = .098) and 3-level ACDF (P = .132) groups, respectively. Conclusion: Reduction in VAS-NP score was similar for all procedures. However, only patients that underwent a 2-level ACDF or CDA demonstrated a statistically significant 12-month postoperative improvement in VAS-NP. This study provides new information on the improvement in neck pain following surgery for cervical radiculopathy and may allow surgeons to more accurately prognosticate patient’s convalescence and aid surgical decision making.

OP02: Trauma-Cervical 1

A007: The Clinical Implications of Adding Computed Tomography Angiography in the Evaluation of Cervical Spine Fractures: A Propensity-Matched Analysis

Daniel Tobert¹, Hai Le², Justin Blucher², Mitchel Harris², and Andrew Schoenfeld²

¹University of Utah, Salt Lake City, UT, USA

²Harvard University, Boston, MA, USA

Introduction: Screening asymptomatic blunt trauma patients for cervical arterial injury is controversial. Vertebral artery injury (VAI) is most commonly associated with cervical spine fracture and many guidelines advocate indiscriminate screening of all cervical spine fractures. The purpose of this study is to determine whether the addition of computed tomography angiography (CT-A) results in a change in management for patients with cervical spine fractures. Materials and Methods: Adult patients with acute cervical spine fractures after blunt trauma between 2000 and 2015 were retrospectively identified. Patients with penetrating trauma, neoplasm, or prior cervical spine surgery were excluded. The following variables were recorded: age, biologic sex, medical comorbidities, Injury Severity Score (ISS), mechanism of injury, whether CT-A of the neck was obtained in addition to CT, cervical spine fracture characteristics, and the presence of VAI. Recommendation for a change in management with antithrombosis was the primary outcome measure. Detection of stroke and VAI were secondary outcomes. Propensity score matching was performed to negate the significant baseline demographic and clinical characteristics. Results: There were 3943 patients screened and 2831 patients eligible. Propensity score matching yielded one cohort receiving CT-A and one cohort that did not, both with 644 patients and equivalent demographic and clinical characteristics. CT-A identified definite or indeterminate VAI in 113 patients and 62 patients had antithrombosis recommended. In the cohort without CT-A, VAI was discovered in 11 patients incidentally through other imaging and 8 were recommended antithrombosis. Two patients in the CT-A group had major adverse bleeding events as a result of antithrombosis initiation. There were no preventable strokes in either group. Conclusion: The addition of CT-A increased detection of VAI and antithrombosis recommendation. There was a high incidence of indeterminate CT-A findings. There were no preventable strokes in either cohort and 2 major adverse bleeding events as a result of recommended pharmacologic antithrombosis. Nonselective screening is not warranted and should be limited to a high-risk subset of patients.

A008: The Window of Opportunity for Surgical Decompression in Patients With Acute Traumatic Cervical Spinal Cord Injury

Marko Jug¹, Nataša Kejžar², Matej Cimerman¹, Fajko Bajrovic¹

¹University Medical Center Ljubljana, Ljubljana, Slovenia

²University of Medicine, Ljubljana, Slovenia

Introduction: Acute traumatic spinal cord injury (tSCI) is a devastating event. Recent clinical practice guidelines suggest early surgical decompression (SD), that is, within 24 hours after injury, but the optimal time for SD remains unknown. In this study we aimed to analyze the effect of timing of SD within the first 24 hours after injury and to find the optimal time window for SD in patients with acute cervical tSCI and persistent spinal cord compression in order to achieve a significant neurologic recovery of at least 2 AIS grades. Additionally, we evaluated the association of timing of SD adjusted for the degree of spinal canal compromise (SCC) and severity of SCI (complete or incomplete). Material and Methods: A prospective cohort study of consecutive patients with AIS grades of A through C and fracture and/or dislocation of the subaxial cervical spine with MRI-confirmed spinal cord compression operated on within the first 24 hours after injury was performed. The primary outcome was the change in AIS grade at the 6-month follow-up. Receiver operating characteristic (ROC) curve was used to determine the best cut-point for the time form injury to SD in which the chances for significant neurologic recovery are the highest (Youden index). Multivariate logistic regression was used to model significant neurologic recovery with time to SD, percentage of SCC and injury severity. Results: Of the 73 enrolled patients, 64 patients concluded the study. Forty-one patients presented as (ASIA Impairment Scale) AIS A, 9 as AIS B, and 14 as AIS C injuries. The median time to decompression was 7 hours. Twenty-five percent of patients improved by at least 2 AIS grades, 28% by 1 AIS grade, and 45% experienced no improvement, with 1 patient deteriorating by 1 AIS grade. Time of SD and SCC were statistically significant predictors of neurologic recovery of at least 2 AIS grades, that is, increasing the delay in time from injury to SD or increasing the degree of SCC before SD lowered the chances for significant neurologic improvement. The injury severity was a marginally significant predictor of neurologic recovery with strong correlation to SCC. The optimal time window for SD to achieve a neurologic improvement of at least 2 AIS grades for our population of patients was within 4 hours form injury (95% CI = 4, 9). Sensitivity and specificity at optimal cutoff point were 0.58 and 0.92, respectively. Conclusion: Our results show that the effect of SD on neurologic recovery within the first 24 hours after acute cervical tSCI is time dependent and the optimal time window for SD in our cohort is between the time of injury and the first 4 to 9 hours after injury, depending on the degree of SCC and injury severity. Therefore, due to heterogeneity of tSCI no uniform cutoff for the timing of SD can be set for all patients. Nevertheless, SD should be achieved as soon as possible.

A009: Blunt Cerebrovascular Injury: Incidence and Long-Term Follow-up

Dennis Hundersmarck¹, Willem-Bart Slooff¹, Jelle Homans¹, Quirine Van der Vliet¹, Nizar Moayeri¹, Falco Hietbrink¹, Gert De Borst¹, Cumhur Öner¹, Sander Muijs¹, and Luke Leenen¹

¹University Medical Center Utrecht, Utrecht, Netherlands

Introduction: Blunt cerebrovascular injuries (BCVI) are identified in 1% to 2% of all blunt trauma patients and can cause ischemic stroke. Computed tomography angiography (CTA) scanning is the diagnostic modality of choice in BCVI suspected patients. Although CTA-scanning capabilities are ever improving, digital subtraction angiography remains the golden standard. Furthermore, data about long-term functional outcomes after BCVI is limited. The aim of this study was to determine BCVI incidence in relation to imaging modality improvements and determine long-term functional outcomes. Material and Methods: This retrospective cohort study was performed at the University Medical Center Utrecht, a Dutch level 1 trauma center. During a 10-year period, from January 2007 till December 2016, all consecutive BCVI patients ≥18 years old were identified. BCVI incidence among the overall blunt trauma group and specific trauma population subgroups (polytrauma, basilar skull fracture, and cervical trauma) was calculated before and after improvements in CTA scanning and contrast administration protocol. Three periods were identified where CTA diagnostic modalities were improved. In the first period (I), from 2007 up to and including 2009, trauma patients received a full body CT angiogram using a 64-slice CT scanner. In the second period (II), during the course of 2010 up to and including 2013, trauma patients received a full body CT angiogram using a 256-slice CT scanner. In the third period (III), during the course of 2014 up to and including 2016, trauma patients received a full body CT angiogram using the same CT scanner used in period II. However, split bolus intravenous contrast administration flow during the second phase of administration was increased from 3 mL/s to 6 mL/s, providing a higher contrast density. In addition, trauma patients presenting in period III with isolated maxillofacial, cervical, or head injuries, visible at CT scanning, received additional CTA scanning. Long-term functional outcomes following BCVI, using the EuroQol 6-dimensional (EQ-6D) telephonic questionnaire, were determined. Functional outcomes were compared to reference populations, using the EQ-5D index score. Results: Seventy-one BCVI patients were identified among approximately 12 000 (0.59%) blunt trauma patients. In the first period (I) BCVI incidence among the overall blunt trauma group and polytrauma, basilar skull fracture, and cervical trauma subgroups was found to be 0.3%, 0.9%, 1.2%, 4.6%, respectively, which more than doubled toward the third period (III), resulting in 0.8%, 2.4%, 1.9%, 8.5%, respectively. Ischemic stroke as a result of BCVI was found in 20 patients (28%). Six in-hospital deaths were BCVI related (8%). Functional outcomes were determined by the EQ-5D score, with a follow-up rate of 83%. BCVI patients reported lower functional outcomes compared with the Dutch reference population of the same age (P < .001), as well as the general trauma population presenting at the same institution (P < .01). Ischemic stroke was identified as a major cause of functional impairment at long-term follow-up. Conclusion: Improved CTA diagnostic modalities have increased BCVI incidence. Furthermore, BCVI patients reported significant functional impairment at long-term follow-up.

A010: Unbiased Recursive Partitioning to Stratify Patients With Acute Traumatic Spinal Cord Injuries: External Validity in an Observational Cohort Study

Nathan Evaniew¹, Nader Fallah¹, Carly Rivers¹, Vanessa Noonan¹, Tova Plashkes¹, Zeina Waheed¹, Brian Kwon¹, Charles Fisher¹, and Marcel Dvorak¹

¹University of British Columbia, Vancouver, British Columbia, Canada

Introduction: Clinical trials of novel therapies for acute traumatic spinal cord injuries (SCI) are extremely challenging because variability in spontaneous neurologic recovery can make discerning actual treatment effects challenging. Unbiased Recursive Partitioning regression with Conditional Inference Trees (URP-CTREE) is a novel approach to this challenge that was developed through analyses of a large European SCI database (EMSCI). URP-CTREE uses early neurologic impairment to predict how much motor recovery individual patients will achieve, and it holds potential to optimize clinical trial design by efficiently stratifying patients and decreasing required sample sizes. We performed an external validation study to determine how well a previously reported URP-CTREE model stratified patients into distinct homogeneous subgroups and predicted subsequent neurologic recovery when applied to an independent cohort. Methods: We included all eligible patients from an ongoing prospective cohort study that presented to a single academic quaternary care referral center from May 2004 to February 2018 with complete American Spinal Injury Association (ASIA) Impairment Scale (AIS) A acute traumatic cervical spinal cord injuries and baseline motor levels at C4 to C6. We applied the previously reported URP-CTREE model and evaluated Upper Extremity Motor Score (UEMS) recovery at final follow-up. We tested univariate associations with the Pearson correlation coefficient, and differences between medians with the Mann-Whitney U test. We considered recovery to be correctly predicted when final UEMS scores were within a prespecified threshold of 9 points from the median in each group. Results: We included 101 cervical AIS A patients, whose mean times from injury to baseline and follow-up neurological examinations were 6.1 days (SD 17) and 235 days (SD 71), respectively. Median UEMS recovery was 7 points (interquartile range [IQR] = 2 to 12). We applied the previously reported URP-CTREE model, which partitioned our cohort into 5 stratified groups of predicted UEMS recovery, according to 4 predictor variables. One of the predictor variables was not statistically significant in our independent sample of patients, one of the stratified groups did not fit a consistent sequence of progressively improving UEMS scores, and 3 of the 5 stratified groups had medians that were not significantly different from their adjacent groups. However, the model correctly predicted UEMS recovery in 86 patients (85%). Prediction accuracy decreased to 75% in a sensitivity analysis that included only those patients whose time from injury to baseline neurological examination was less than 24 hours (n = 63). Conclusions: A previously reported URP-CTREE model had limited ability to stratify patients into distinct homogeneous subgroups, but reasonably promising accuracy for predicting final motor recovery in an independent cohort of acute cervical AIS A patients. Our cohort had earlier baseline examinations than the majority of patients in the EMSCI database, but their examinations were still later than that which is typical for enrolment in acute clinical trials. Further research is warranted to evaluate the external validity of URP-CTREE among patients with incomplete injuries and to investigate additional strategies for accurately stratifying patients with acute SCI.

A011: Early Complications, Morbidity, and Mortality in Octo- and Nonagenarians Undergoing Posterior Intraoperative Spinal Navigation Based C1/2 Fusion for Acute Traumatic Odontoid Type II Fractures

Basem Ishak¹, Till Schneider¹, Valerie Gimmy¹, Andreas Unterberg¹, and Karl Kiening¹

¹Heidelberg University Hospital, Heidelberg, Germany

Introduction: Odontoid type II fractures are the most common cervical spine injury in the elderly. Recent studies confirm that external stabilization is associated with high mortality and complication rates. The decision for surgical treatment is still controversial, particularly with regard to elevated perioperative risk owed to frequent comorbidities and poor bone quality. Modern navigation may help improve perioperative safety and efficacy and thereby reduce complication rate. The aim of this study was to assess short-term mortality as well as mid-term clinical and radiological outcome in elderly. Material and Methods: Thirty-five patients with an acute traumatic odontoid type II fracture who underwent posterior atlanto-axial instrumentation with a modified Goel-Harms technique at our institution between January 2007 and December 2015 were retrospectively analyzed and prospectively examined clinically and radiologically. Comorbidities were stratified using the age-adjusted Charlson Comorbidity Index (AACCI). Mortality, length of ICU (intensive care unit) and hospital stay were determined. The neurological and radiological outcome, blood loss, the necessity of blood transfusion as well as medical and surgical complications were evaluated. Quality of life was measured using the EQ-5D and SF-36 questionnaires at final follow-up (FU). Results: Average age was 86.5 years (range 80-96 years). Mean AACCI was 7.1, which is classified as severe. In-hospital mortality was nonexistent and no patient showed new permanent neurological deficits after surgery. Average length of hospital stay was 13.8 days and 2 days for ICU. Blood transfusion was necessary in one patient. Three patients developed cardiopulmonary complications postoperatively (2 pneumonia, 1 heart attack). One wound infection occurred. Twenty-five patients were available for final FU with a mean FU of 22 months (range 6-72 months). The quality of life measured by EQ-5D showed a good outcome (0.7 ± 0.1). All SF-36 domains were reduced in comparison with the German population. Solid bony fusion could be achieved in all patients. Conclusion: Our current study confirms that modification of the C1/C2 posterior fusion technique by using intraoperative CT navigation is a safe and effective procedure in the elderly with few complications and preservation of favorable postoperative quality of life. Implant-related complications such as screw loosening or migration could be avoided under navigation guidance. The overall complication rate was 11%. Surgery in the very old should be considered as first-choice treatment.

A012: Management in Acute Traumatic Cervical Spinal Cord Injury

Andrei Stefan Iencean¹ and Stefan Mircea Iencean²

¹Clinic Emergency N Oblu Hospital, Iasi, Romania

²Gr T Popa University of Medicine and Pharmacy, Iasi, Romania

Introduction: Acute traumatic cervical spinal cord injury (SCI) is a very important medical emergency. Many traumatic spinal cord injuries are initially incomplete and the secondary damage completes the destruction of the spinal cord later. Materials and Methods: We present the retrospective analysis of over 1100 cases of acute traumatic cervical spinal cord injuries treated over a 30-year period (1988-2018). These cases of complete or incomplete acute traumatic cervical spinal cord injuries have been treated differently in relation to the evolution of knowledge and the setting standards of treatment. Results: The results were analyzed according to the complete or incomplete lesion; corticoid treatment; surgical timing (early or late surgical decompression); and other types of neuroprotection (correction of hypotension, hypoxia, decrease of intrathecal pressure by lumbar puncture, bone marrow tissue implant, etc). Our patients with acute traumatic cervical spinal cord injuries were treated in this period in relation to the treatments recommended at that time. For example, we used methylprednisolone for a while, or after that we used another corticosteroid scheme; also we have established correlations between the evolution of patients with cervical SCI and the level of the phosphorylated neurofilament NF-H subunit in cerebrospinal fluid as a biomarker. Conclusions: Analysis of over 1000 patients showed better results after early surgical decompression; correction of systemic hypotension and reduction of intrathecal pressure by lumbar puncture; and using an individualized corticotherapy scheme.

OP03: Deformity-Thoracolumbar (Adult) 1

A013: Is Fresh Whole Blood Superior to Blood Components Transfusion in Major Spine Surgeries? A Randomized Controlled Study

Shanmuganathan Rajasekaran¹, P. Keerthivasan¹, Ajoy Prasad Shetty¹, and Rishi Mugesh Kanna¹

¹Ganga Hospital, Coimbatore, India

Introduction: Major spine surgeries are associated with significant morbidity that are complemented by increased risks of significant perioperative blood loss with need for multiple blood transfusions. Any possibility to decrease the morbidity and improve outcome is beneficial and should be explored. Transition from fresh whole blood (FWB) to component transfusion in the 1990s was based on convenience rather than on scientific evidence. FWB transfusion has proved superior to blood components in reducing coagulopathic and inflammatory response and improving survival rate in hemorrhagic shock and cardiovascular surgery. However, the primary pathology itself in those studies confound the coagulation and inflammatory markers. Whether this superiority of FWB over component transfusion stands true in major spine surgeries too remain unknown and needs to be evaluated. We did a prospective randomized control study in our hospital from September 2017 to August 2018 to compare the effectiveness of FWB and blood component transfusion in improving clinical outcome and serological parameters in early postoperative period following major spine surgery. Materials and Methods: Sixty-five patients (adolescent idiopathic scoliosis 31, early onset scoliosis 16, neuromuscular scoliosis 15, and neurofibromatosis 3) planned for major spine surgery(>6 levels of fusion, expected blood loss >750 mL, duration of procedure >4 hours) were randomized into 2 groups. FWB group received fresh whole blood, whereas COMP group received components transfusion. Preoperative parameters including demographics, diagnosis, and facial measurements to assess puffiness, intraoperative parameters including level of fusion, type of graft, duration of procedure, volume of transfusion, and blood loss, and postoperative parameters including postoperative drain collection, vitals, duration of oxygen support and ICU (intensive care unit) stay, volume of transfusion, complications, and well-being score were assessed. In addition, biochemical parameters such as PT (prothrombin time), ABG (arterial blood gas), calcium, electrolytes, lactate, and interleukin-6 were also compared between the 2 groups. Results: Mean age of study group was 14.17 years in FWB and 16.48 in components (COMP) group. Mean number of segments fused were 10 in both groups, duration of procedure was 219.52 minutes (FWB) and 206.50 minutes (COMP). Intraoperative blood loss was 926 mL versus 845 mL, and mean transfusions made was 1.90 units versus 1.61 units. Postoperative drain collection was more in FWB (516.84 mL vs 406.19 mL). Duration of ICU stay was less in FWB (40.44 hours vs 45.53 hours), and duration of oxygen dependence is also less (36.34 hours vs 43.29 hours). Mean pH was 7.445 in FWB and 7.397 in COMP (P < .01). Interleukin-6 was 30.63 and 34.48, respectively. Facial puffiness was present in 18 out of 35 in component group and only 7/30 in FWB group. Other biochemical parameters were similar. Conclusion: Patient transfused with fresh whole blood achieved better clinical outcome in the immediate postoperative period in terms of decreased duration of ICU stay and oxygen dependence and had lesser inflammatory response compared to those who had component transfusion.

A014: Low-Cost Indigenous Implants Versus High-Cost Multinational Company Implants in Treatment of Adolescent Scoliosis Surgery

Rishi M. Kanna¹, Ajoy P. Shetty¹, and S. Rajasekaran¹

¹Ganga Hospital, Coimbatore, India

Introduction: Corrective surgery for adolescent idiopathic scoliosis (AIS) involves huge costs with average per patient cost being approximately $100 000 in the United States. The awareness about the disease and the benefits of surgery are increasingly being recognized in developing countries like India. However, lower per capita income, socioeconomic status, and meagre health care insurance often make the surgeon and the patient to opt for low-cost Indian pedicle screws for deformity correction surgeries (7-10 times less expensive than imported screws). These titanium screws conform to the American Standards for Testing and Materials (ASTM) and are widely used in the Indian subcontinent. However, there is no available data regarding the safety, efficacy, durability, and cost-effectiveness of low-cost Indian screws in comparison to imported screws in scoliosis surgeries. Methods: We performed a retrospective review of case records of consecutive patients who underwent posterior deformity corrective surgery for AIS and completed a minimum of 2-year follow-up. Patients with congenital scoliosis, deformity >100°, pathological curves, need for complex osteotomy, and age >20 years were excluded from the study. Both groups of patients had a uniform surgical technique operated by the same surgical team, pedicle screws placed strategically, and curve correction achieved by rod rotation technique. Patients’ demographics, pre- and postoperative Cobb angle, Lenke curve type, curve flexibility, number of screws used, total vertebral segments fused, and cost analysis based on implants, length of stay, and complications were studied. The patients chose an Indian versus imported implant based on their affordability. Results: Ninety-two patients (female–male = 82:10) formed the study group and were divided into 2 groups: Group A (Indian)—n = 30; and Group B (Imported)—n = 62. The distribution of curves was as follows: Lenke I—40, II—3, III—13, IV—1, V—25, VI—10. The mean age in years was comparable (16.7 ± 3.8 vs 15.2 ± 3.3, P = .05). There was no significant difference in the various study parameters including the mean preoperative Cobb angle: 59.4 ± 14.4° versus 56.8 ± 13.5° (P = .3990); curve flexibility—39.9 ± 13.5% versus 41.5 ± 14.5% (P = .6); mean postoperative Cobb angle—21.6° versus 17.8° (P = .5); mean curve correction achieved—65 ± 23.8% versus 69 ± 25.7% (P = .46); mean levels fused—10.9 ± 2.3 versus 10.4 ± 2.6 (P = .36); the screw density ratio—0.68 versus 0.69 (P = .9); and the mean loss of correction at final follow-up, 2.4° versus 2.6° (P = .71) between the 2 groups. But the mean cost per patient of $2339.42 in Group A was significantly less than in Group B ($9045.75) (P < .001). Conclusion: The significantly lower cost Indian pedicle screws provided similar deformity correction, which was maintained at 2 years. The mean levels fused, number of screws used, and screw density ratio was similar in both groups. With concerns about raising health care costs, a low-cost model for spine deformity correction surgery is a welcome need of the hour.

A015: One-Stage Posterior Multiple-Level Asymmetrical Ponte Osteotomies Versus Single-Level Posterior Vertebral Column Resection for Severe and Rigid Adult Idiopathic Scoliosis: A Minimum 2-Year Follow-up Comparative Study

Yangpu Zhang¹ and Yong Hai¹

¹Capital Medical University of China, Beijing, China

Introduction: To evaluate and compare the efficacy and safety of one-stage posterior multiple level asymmetrical Ponte osteotomies (MAPO) and single-level posterior vertebral column resection (VCR) in the treatment of severe and rigid adult idiopathic scoliosis (AIS), a retrospective comparison study was performed. Material and Methods: A retrospective comparative study was conducted to evaluate radiological and clinical results in patients with severe and rigid AIS. A total of 38 patients who underwent one-stage posterior MAPO or VCR and fusion surgery at our hospital with minimum 2-year follow-up between February 2009 and November 2015 were enrolled and divided into 2 groups according to the procedures they received. Twenty-six patients were included in MAPO group and 12 patients in VCR group with an average age of 26.65 ± 8.40 years and 27.92 ± 7.50 years. The average follow-up was 30.24 ± 10.55 months. The surgical details and complications were recorded. The radiological parameters before and after the surgery, as well as the final follow-up was collected and analyzed. Clinical outcome including Oswestry Disability Index (ODI) and Scoliosis Research Society-22 (SRS-22) questionnaire scores were obtained preoperatively and at final follow-up. Results: The main curve in MAPO and VCR group were corrected from an average of 98.52 ± 16.50° to 44.11 ± 17.72° and 108.91 ± 16.56° to 56.49 ± 18.82°, with a mean correction rate of 56.72 ± 13.38% and 49.18 ± 12.00% with no significant difference among the 2 groups. No significant loss of correction at final follow-up was found for major curves in both groups. The postoperative coronal and sagittal parameters of the 2 groups were all improved and it showed no significant differences between the 2 groups. The mean operative time and blood loss of VCR group was significantly greater than those of MAPO group. All the clinical scores were significantly improved at final follow-up, with no statistical difference. The incidence of complications in VCR group was 33.33%, which was significantly higher than that in MAPO group. Conclusion: The surgical procedure of multiple asymmetrical Ponte osteotomy is a safe, easy-to-operate, and effective technique that can correct scoliosis and restore the sagittal alignment. It can gain similar correction outcome to VCR, offering the advantages of reduced operation time, reduced blood loss, and greatly reduced the complication and may go a long way toward solving the problems of severe and rigid adult idiopathic scoliosis.

A016: The Role of the Fractional Lumbosacral Curve in Persistent Coronal Malalignment Following Adult Thoracolumbar Deformity Surgery

Alekos Theologis¹, Thamrong Lertudomphonwanit², Lawrence Lenke³, Keith Bridwell⁴, and Munish Gupta⁴

¹University of California–San Francisco, San Francisco, USA

²Ramathibodi Hospital, Mahidol University, Bangkok, Thailand

³Columbia University Medical Center, New York, NY, USA

⁴Washington University in St Louis, St Louis, MO, USA

Introduction: Achieving appropriate coronal alignment is less reliable in adults with coronal malalignment due to trunk shift ipsilateral to degenerated thoracolumbar scoliosis’ apex. The goal of this study is to compare radiographs/surgical techniques for thoracolumbar deformity with varying severity and direction of coronal imbalance. Material and Methods: Review of adults who underwent posterior spinal fusions to pelvis (≥5 levels) for thoracolumbar scoliosis. Exclusion: revisions, no coronal deformity, thoracic Cobb >30°, and anterior operations. Patients were divided into 3 groups, as proposed by Rao et al: (1) Type A, CSVL (central sacral vertical line) <3 cm; (2) Type B, CSVL >3 cm, and C7 plumb shifted to scoliosis’ concavity; (3) CSVL >3 cm and C7 plumb shifted to scoliosis’ convexity. Radiographic parameters and surgical techniques were compared. Results: A total of 144 patients (male 6; female 118; avgerage age 58 ± 10 years; Type A 87; Type B 19; Type C 18). Type C had significantly greater lumbosacral fractional curves. Twenty-eight percent of Type C were treated with fractional curve TLIFs (transforaminal lumbar interbody fusion), while all, but one, Type B had TLIFs of the fractional curve. Deformity parameters after surgery were similar, except Type C had persistently greater fractional curves/coronal malalignment. All preoperative Type B were appropriately corrected postop. For preoperative Type C, 67% remained Type C and 33% became Type A postoperative. Compared to those who became Type A, persistently undercorrected and malaligned (Type C) patients had significantly greater preoperative lumbosacral fractional curves, greater preoperative coronal Cobb angles, and more commonly involved TLIFs of lumbosacral fractional curves. Conclusion: In adults who underwent primary, posterior-only operations for thoracolumbar spinal deformity, the majority of Type C coronal deformities remained coronally undercorrected and malaligned postoperative. For these patients, an alternative surgical strategy should be considered to more adequately correct lumbosacral fractional curves and maintain and/or restore coronal balance.

A017: Successful Creation of Deployable Preoperative Predictive Risk Calculators for Individual Patient Event-Free Survivorship for Major Complications, Hospital Readmissions, and Unplanned Surgery Following Adult Spinal Deformity (ASD) Surgery

Ferran Pellisé¹, Miquel Serra-Burriel², Sleiman Haddad¹, Alba Vila-Casademunt³, Francisco Javier Sánchez Pérez-Grueso⁴, Shay Bess⁵, Emre R. Acaroglu⁶, Justin S. Smith⁷, Frank Kleinstück⁸, Virginie Lafage⁹, Ibrahim Obeid¹⁰, Frank J. Schwab⁹, Christopher I. Shaffrey⁷, Ahmet Alanay¹¹, Christopher P. Ames¹², International Spine Study Group ISSG⁹, European Spine Study Group ESSG³

¹Hospital Vall d’Hebron, Barcelona, Spain

²Universitat Pompeu Fabra, Barcelona, Spain

³Vall d’Hebron Institute of Research, Barcelona, Spain

⁴Hospital Universitario La Paz, Madrid, Spain

⁵Presbyterian St. Luke’s/Rocky Mountain Hospital for Children, Denver, CO, United States

⁶Ankara ARTES Spine Center, Ankara, Turkey

⁷University of Virginia Medical Center, Charlottesville, VA, USA

⁸Schulthess Klinik, Zürich, Switzerland

⁹Hospital for Special Surgery, New York, NY, USA

¹⁰Bordeaux University Hospital, Bordeaux, France

¹¹Acibadem University, Istanbul, Turkey

¹²University of California San Francisco, CA, USA

Introduction: Accurate preoperative risk stratification in adult spinal deformity (ASD) surgery is indispensable to improve patients’ selection, accordingly adjust surgical invasiveness and optimize outcomes. Few predictive models allow for proper patient selection, adjustment of invasiveness, and patient frailty optimization to predict and reduce postoperative major complications (MC), hospital readmissions (READMIT), and unplanned surgery (UNPLAN). We aim to create accurate predictive models for the occurrence and timing of these MC, READMIT, and UNPLAN following ASD surgery. Material and Methods: Surgically treated ASD patients with >2 year follow-up were identified. Patient demographic, radiographic, operative, baseline patient-reported outcome measures, and complications data were analyzed to build event-free survival curves for MC, READMIT, and UNPLAN, and to create predictive models by means of a random survival forest with 80/20 train/test sets. A total of 101 variables were used to train the models. Goodness of fit was assessed in the test set. Missing value imputation was performed with the miss-Forest package. R software was used for analysis. Results: A total of 1018 ASD patients operated before September 2014 (77.7% women, 55.5 years mean age, 10.7 levels fused segments, 55.5% pelvic fixation, 21.2% 3CO) by 57 surgeons at 24 sites in 5 countries (2 continents), with 2047.9 observation-years, were included in the analysis used to build MC, READMIT, and UNPLAN risk calculating models with proved successful model fit. Missing value imputation was 14.59%. C-statistic value (70.6%) proved successful model fit. Models demonstrate that 87.9% of patients are MC-free at 10 days postoperative, 78.5% at 90 days, and 63% at 2 years. Surgical invasiveness (LIV-pelvic fixation, length of fusion, prior surgery), age, magnitude of sagittal deformity, patient frailty (walking and lifting capacity), and blood loss most strongly predict MC. Surgeon and site most strongly predict READMIT and UNPLAN. Curves show a continued survivorship decrease for event free MC, READMIT, and UNPLAN beyond >2-year follow-up. Conclusion: Risk calculating models for event-free MC, READMIT, and UNPLAN following ASD surgery demonstrate that patient-related factors, >1/3 of which are modifiable, account for 55% of the MC predictive model weight. Surgeon and site represent 4% for MC, but are most relevant for READMIT and UNPLAN.

A018: Rocket Incision and Approach: A Novel Surgical Approach in Posterior Correction of Neuromuscular Scoliosis

Saurabh Kapoor¹, Keiran O’Boyle¹, and Masood Shafafy¹

¹Queen’s Medical Center, Nottingham, UK

Introduction: Surgical treatment of neuromuscular scoliosis is associated with high risk of infection as incision extends into the nappy area. We describe the safety and efficacy of a new approach developed to reduce these risks. Methods: Data were prospectively collected for a cohort of patients with neuromuscular scoliosis requiring posterior correction in which this novel surgical approach was utilized. A single surgeon at a single institution operated all patients. Technique: Patient is positioned as standard for a posterior approach. Both posterior superior iliac crests (PSIS) are marked on the skin. The positions of the superior gluteal arteries (SGA) are marked bilaterally using pulsations located approximately 1 inch below and at the junction of outer and middle third of the iliac crest. A semilunar line is drawn from one PSIS to the other curving upward with the apex centered at L2/L3. From the apex a vertical line centered on the spine is drawn to the proximal thoracic spine. The overall shape of this looks similar to a rocket or inverted wine glass. The semilunar line is incised directly to through the skin, subcutaneous fat, and the deep fascia. This flap is then raised inferiorly as a whole away from the posterior paravertebral muscles as far as necessary but staying short of the origin of the SGA. Detaching paravertebral muscles distally from middle portion of posterior iliac crest follows subperiosteal exposure. This allows relatively blood less exposure of the posterior pelvis for instrumentation, and when they are reattached, they cover the screws making them less prominent. For closure, the muscles are first reattached to using nonabsorbable sutures or absorbable with long half-life followed by standard closure technique. Results: Study cohort included 4 males and 1 female, with average age of 16.2 years (range 8-19). Average period of follow-up was 15 months (range 3-24). Four patients had cerebral palsy and one Duchene muscular dystrophy. All surgical wounds healed with no complications including superficial or deep infection or metal work prominence. Conclusion: Rocket incision and approach appears to provide a safe and effective alternative approach in posterior correction of neuromuscular scoliosis.

OP04: Biomechanics 1

A019: The Effects of Added Compression on Sagittal Plane Lumbosacral Junction Rod Strain and Sacral Screw Bending Moment During In Vitro Loading

Brian P. Kelly¹, Anna Newcomb¹, Jennifer Lehrman¹, Bernardo DeAndrada Pereira¹, Jake Godzik¹, Jay Turner¹, and Randy Hlubek¹

¹Barrow Neurological Institute, Phoenix, AZ, USA

Introduction: Variations in load distribution based on surface strains of instrumentation spanning the lumbosacral junction during in vitro loading are often used to evaluate biomechanical effects of different pedicle screw and rod (PSR) configurations. The gold standard for biomechanical comparisons of spinal instrumentation involves the application of pure moment loads to induce bending of the instrumented spine. However, spinal constructs in vivo are subjected to substantial vertical compressive forces due to gravitational loading of upper body mass. The effects of added compressive loads during in vitro bending of the PSR instrumented lumbar spine, as determined by surface strains at the lumbosacral level, are not well understood. The purpose of this study was to analyze the biomechanical differences in lumbosacral rod strain, and sacral screw bending moments during sagittal plane bending, with and without a compressive pre-load. Material and Methods: Fourteen cadaveric specimens were tested in a 6 DOF (degree of freedom) robot under continuous real-time loading in 2 PSR configurations (L2-S and L2-IL) using 5 types of loads: 7.5 N m flexion (FL), 7.5 N m extension (EX), 400 N compression (C), combined C + FL, and combined C + EX. Strain gauges were used to measure rod strains (RS) at L5-S and net sacral screw bending moments (SS). Data were analyzed using one-way repeated-measures ANOVA followed by Holms-Sidak paired analysis (P < .05). Results: RS significantly increased with L2IL versus L2S (FL: +77%; EX: +65%; C + FL: +46%; C: +40%; P < .004), but without significance during C + EX (+20%, P = .224). SS significantly decreased with L2IL versus L2S (FL: −78%; EX: −81%; C + FL: −52%; C+ E X: −33%; C: −50%; P = .049). Compression loading alone induced significantly greater RS and SS than pure moment FL and EX in both L2S and L2IL groups (P < .001). Adding C to FL significantly increased RS in L2S (+170%) and L2IL groups (+123%) (P < .001), and SS in L2S (+86%) and L2IL (+292%) (P < .001). Adding C to EX significantly increased RS in L2S (+76%; P < .001) and L2IL groups (+27%; P = .043), as well as SS in L2IL (+277%; P < .001) only. Compared to C alone, C + FL increased RS and SS in both groups but without significance, while C + EX tended to reduce RS with significance for L2IL only (−21%; P = .007), and reduce SS with significance for L2S only (−31%; P = .002). In the presence of C, SS was greater and more sensitive to the direction of loading (C + FL vs C + EX) with L2S (P < .001) versus L2IL (P = .510), while RS was more sensitive to the direction of loading with L2IL (P < .001) versus L2S (P = .430). These differences were not present for magnitudes of RS or SS with pure moment loads (FL vs EX; P > .2). Conclusion: The relative effects of L2IL versus L2S (increased RS and reduced SS) were consistent among all types of loads. However, the additive effects of compression on RS and SS varied with type of instrumentation (L2S vs L2IL) and direction of bending (FL vs EX). Compression loading significantly influences instrumentation performance and can help provide improved understanding of lumbosacral junction load distributions with PSR constructs.

A020: Dual Pedicle and Cortical Screws Using Robotic Navigation Improves Load to Failure in the Osteopenic Lumbar Spine: An In Vitro Biomechanical Analysis

Michael Steinmetz¹, Jessica Riggleman², Jonathan Harris², John Butler¹, Chelsea Wright¹, Mir Hussain², Bryan Ferrick³, and Brandon Bucklen²

¹Cleveland Clinic Lerner College of Medicine, Cleveland, OH, USA

² Musculoskeletal Education and Research Center (MERC), a Division of Globus Medical, Inc, Audubon, PA, USA

³Drexel University, Philadelphia, PA, USA

Introduction: Compression fractures due to severe osteopenia may result in kyphotic deformity and/or neurological deficit requiring surgical intervention. Multi-rod reconstruction techniques have been shown to maximize posterior fixation, but at present, utilize a single pedicle screw anchor point to support multiple rods. Robotic navigation technologies have been developed to improve pedicle screw (PS) accuracy and may allow for the insertion of both PS and cortical screws (CS) within the same pedicle, providing 4 points of bony fixation per vertebra. Recent studies have demonstrated the radiographic feasibility for such a dual screw trajectory for posterior lumbar spinal instrumentation; however, the biomechanical characterization of the technique is presently lacking. Investigators quantified the ultimate load to failure and construct stiffness of 4-rod constructs of (1) bilateral pedicle screw and (2) bilateral dual pedicle and cortical screw (PSCS) techniques in an osteopenic model. Material and Methods: Fourteen cadaveric lumbar vertebrae (L2) were divided into 2 groups (n = 7) of similar bone density (t-score: −2.30 ± 1.38 and −2.16 ± 1.17 for PS and PSCS, respectively). Group 1 was instrumented with PS, and group 2 was instrumented with PS and CS screws within the same pedicle. Operative constructs (L2) included (1) bilateral pedicle screw fixation with 4 rods using rod-to-rod connectors (PS 4-rod); and (2) bilateral PSCS dual screw fixation with 4 rods affixed at PS and CS (PSCS 4-rod). Static compression testing was performed. The construct was connected to four 5.5 mm diameter titanium rods, and the cranial ends of 2 rods were secured to pedicle screws placed within the polyblock. The specimen was loaded in the axial direction to failure at a constant displacement rate of 5 mm/min. Ultimate load to failure (N) and mechanical stiffness (N/mm) were determined. Comparisons were made between groups (significance at P < .05). Results: PSCS 4-rod reconstruction significantly improved the ultimate load-to-failure and mechanical stiffness outcomes. The average load to failure between groups was 892.50 ± 288.15 N and 1798.04 ± 329.23 N for PS 4-rod and PSCS 4-rod, respectively. The average stiffness between groups was 329.67 ± 79.35 and 429.64 ± 84.02 for PS 4-rod and PSCS 4-rod, respectively. A significant difference was noted between groups for both outcome measures (P < .05). Conclusion: The present study provided the first biomechanical evaluation of a novel dual screw technique using robotic navigation. Four-rod reconstruction with dual pedicle and cortical screws significantly improved pedicle fixation and construct stiffness. Increased stiffness and resistance to failure could be an important component, particularly in patients with severe osteopenia, in hardware maintenance in the absence of fusion, or until bone fragment healing is finalized following stabilization of comminuted bone. Further clinical work is required to determine the efficacy of dual screw fixation in resisting failure in patients with osteopenia.

A021: Functional Requirements for the Long-Term Stabilization of Annulus Fibrosus Lesions

Dmitriy Alexeev¹, Benedikt Helgason¹, Melanie Tschopp¹, and Stephen Ferguson¹

¹ETH Zurich, Institute for Biomechanics, Zurich, Switzerland

Introduction: The use of thin membranes has been proposed for the repair and stabilization of annulus fibrosus defects. Poly(∊-caprolactone) (PCL) is widely used in the production of bioresorbable electrospun nanofibrous membranes for tissue engineering and regenerative medicine applications. In the context of applying PCL scaffolds to repair annulus fibrosus defects, specific mechanical properties that guide cellular activity and resist deformation are required. Ideally, the mechanical properties of the membrane should match those of the native tissue and remain stable over a relevant time period to allow tissue regeneration and repair. Factors such as increased temperature and hydration, in combination with the temporally changing material properties of PCL, may lead to undesirable alterations in scaffold properties following prolonged exposure. The aims of this study were to define the required properties of electrospun PCL membranes and to study the changes in mechanical properties over a 6-month period in a simulated physiological environment. Material and Methods: To define the required material properties of PCL membranes, a parametric finite element model of the spinal motion segment was created. The dimensions of the model were varied to reflect the natural population variation in disc height, width, and lordosis angle. The range of stress-strain profiles in the outer posterolateral annulus were determined for physiological bending and compression. Based on the predicted material properties, 3 types of PCL scaffolds were produced under variable electrospinning conditions with distinct fiber diameters: (1) 1.5 ± 0.3 μm, (2) 3.5 ± 0.2 μm, and (3) 6.5 ± 0.5 μm. The scaffolds were submerged in DMEM at 37°C for 180 days. Control scaffolds were kept dry at 37°C over the same period. Cyclic extension tests, dynamic scanning calorimetry to determine crystalline content, and infrared spectroscopy to determine ester bond degradation were performed at 0, 14, 30, 60, 90, and 180 days. Results: Required modulus values in the range of 10 to 20 MPa were predicted. These properties were best replicated by the membranes with 3.5 μm and 6.5 μm fiber diameters. No significant difference in the mechanical properties between the submerged and control scaffolds was observed at any time point. A small but significant increase in the crystalline content of the scaffolds was observed at the 180-day time point (∼42% to 47%). All samples were found to have a significant increase in E modulus over the incubation time, which correlated with the change in crystalline fraction. No significant changes in ester bond density were observed. Conclusion: Electrospun PCL membranes represent a promising strategy for annulus fibrosus repair, providing mechanical properties similar to native tissue, with a degree of flexibility in fiber diameter which can be used to tailor cell response. Contrary to expectations, no significant reduction in mechanical properties was observed over a 6-month period, which corresponds to the typical spontaneous tissue healing time reported in literature. Linear degradation of bulk PCL samples by ester bond cleavage has been observed in in vivo rat models. As this behavior was not seen in the submerged scaffolds, we conclude that the in vivo degradation is driven by enzymatic processes.

A022: A Biomechanical In Vitro Study of a Novel Radio Translucent Expandable CF/PEEK Vertebral Body Replacement

Daniel Adler¹ and Michael Akbar¹

¹Ruprecht-Karls-University Heidelberg, Heidelberg, Germany

Introduction: Vertebral body replacement (VBR) combined with dorsal instrumentation is well-established to stabilize tumor related, unstable (>AOSpine A3) vertebral fractures in patients with adequate general condition and life expectancy. The vast majority of implants are manufactured of metallic alloys, which lead to artifacts in postoperative radiological diagnostics as well as in adjunct radiation therapy. The purpose of this study was to evaluate biomechanical data of a novel CF/PEEK (carbon fiber–reinforced PEEK) VBR in combination with a CF/PEEK screw and rod system compared to an established titanium 360° treatment. Material and Methods: Six fresh frozen human thoracolumbar specimens (Th11-L3) were tested in a 6 degree of freedom spine tester in flexion/extension, lateral bending, and axial rotation with pure bending moments of ±7.5 N m. CF/PEEK pedicle screws (6.5 mm) were used in Th11/12 and L1/2 in all tested specimens. Variations in the instrumentations consisted of 2 different 6 mm rod materials (CF/PEEK vs titanium) with and without cross-connectors and 2 different VBR materials (CF/PEEK prototype vs titanium). Results: In lateral bending the ROM (range of motion) of the treated index segments (Th12-L2) was significantly (P < .05) reduced in all instrumented states compared to native condition (instrumentations: 0.24-0.30° [SD 0.18-0.21] vs native: 11.2° [SD 3.26]). There were no significant differences (P > .05) between the tested instrumentations in all specimens. In flexion/extension the ROM of the instrumented index (Th12-L2) segments’ was significantly (P < .05) reduced in all instrumented states compared to segments’ native condition (instrumentations: 0.29-0.35° [SD 0.12-0.31] vs native: 10.3° [SD 2.4]) without significant differences (P > .05) between the tested instrumentations. In axial rotation the ROM of the instrumented index (Th12-L2) segments’ was also reduced in all instrumented states (P < .05) compared to native condition (instrumentations: 1.84-3.95° [SD 0.33-0.9] vs native: 4.83° [SD 0.95]). Less rigid CF/PEEK rods in combination with CF/PEEK VBR resulted in the highest ROM (3.95°, SD 0.90), whereas titanium rods with titanium VBR and 2 additional cross-connectors resulted in the lowest ROM (1.84°, SD 0.33). Two additional cross-connectors reduced ROM in axial rotation for all instrumentations independently to VBR or rod material (CF/PEEK vs titanium). VBR’s material (CF/PEEK vs titanium) had only small effect on ROM while posterior rod material (CF/PEEK vs titanium) had a greater effect on ROM in axial rotation. Conclusion: The novel CF/PEEK VBR meets the expectations regarding load capacity and stability. Two cross-connectors are recommended as they reduce ROM independently of rod and VBR material. In axial rotation CF/PEEK rods provide a higher ROM than titanium rods. In lateral bending and flexion/extension there were no significant differences between the tested instrumentations and VBR. The stiffness of rod material affects the ROM rather than the stiffness of VBR material.

A023: Dual Fluoroscopy for Measurement of Spinopelvic Parameters Through the Gait Cycle

Iain S. Elliott¹, Keisuke Uemura¹, Penny Atkins¹, Niccolo Fiorentino¹, Nicholas Spina¹, Darrel Brodke¹, Andrew Anderson¹, and William Ryan Spiker¹

¹University of Utah, Salt Lake City, UT, USA

Introduction: Spinopelvic parameters measured on lateral long-cassette standing radiographs are performed with the patient in an upright standing posture, in which they are asked to hold a specific postural position for a length of time sufficient to acquire a radiograph that will encompass the entire spine and pelvis. Normal ranges for pelvic tilt (PT), pelvic incidence (PI), and sacral slope (SS) have been defined, and multiple studies have shown that aberrant radiographic metrics correlate with health-related quality of life measures. However, we have yet to understand how these parameters might vary in a patient’s daily life, and whether there might be some ability to compensate when asked to take long-standing cassette films in clinic. It is currently unknown whether spinopelvic parameters vary significantly at other points in a patient’s life, and whether abnormal degrees of motion are associated with pathology. In this study, we used high-speed dual fluoroscopy to quantify in vivo motion of the pelvis to analyze spinopelvic parameters dynamically in healthy volunteers. Material and Methods: Eleven healthy volunteers underwent computed tomography (CT) scans of their pelvis, and walked at their self-selected speed on a treadmill. A previously validated dual fluoroscopy system acquired images of a randomly selected hemi-pelvis and proximal femur. Three-dimensional reconstructions of the hip were generated from the CT image data. Model-based tracking was used to calculate the in vivo position of the hip in each frame of the high-speed fluoroscopic video. PT, PI, and SS were quantified as dynamic parameters throughout the full gait cycle. Results: The range of PI was from 33.9° to 73.9°. Pelvic tilt at static position ranged from 1.72° to 22.2°. The average subject had a variation in the pelvic tilt of 4.7° through the course of the gait cycle and ranged from 1.8° to 10.9°. The maximum that the pelvic tilt changed in any subject was 10.9°, and the least was 1.8°. Conclusion: The sacral slope and pelvic tilt vary throughout the gait cycle, even for healthy individuals. More research is needed to understand the implications of high degrees of pelvic mobility and whether abnormally high or low motion is associated with pathology. Although the sample size is low due to the time-intensive nature of data collection with dual fluoroscopy, the data provided by this study will allow comparison to patients with pathologic sagittal imbalance, and inform interpretation of static radiographs of spinopelvic parameters.

A024: Variations Among Human Spine Segments and Their Relationships to In Vitro Kinematics: A Retrospective Analysis of Experimental Data Including 581 Cervical Motion Segments From 147 Donor Spines

Anna Newcomb¹, Jennifer Lehrman¹, Bernardo DeAndrada Pereira¹, Neil Crawford², and Brian P. Kelly¹

¹Barrow Neurological Institute, Phoenix, AZ, USA

²Globus Medical, Phoenix, AZ, USA

Introduction: Kinematic data from in vitro flexibility tests involving cadaveric spines are often used to evaluate medical device performance and to validate finite element models of human spine segments. However, data describing relationships between motion segment features (ie, dimensions of vertebrae and discs, subject height) and kinematics are lacking. The goal of the current study was to retrospectively investigate the relationships between cadaveric cervical spinal motion segments and their biomechanical characteristics, including coupled motion. Materials and Methods: 581 cadaveric cervical motion segments (between C1 and T1) from 147 donor spines (86 males/61 females, range: 20-69 years, mean age 53.5 ± 10.6 years, bone mineral density [BMD] 0.558 ± 0.101 g/cm², mean height 173.4 ± 12.0 cm) were studied. General spine tissue donor information (age, gender, height, and weight at time of death) was obtained from medical histories provided by tissue banks. BMD were obtained from DEXA scans at C4. Dimensions of vertebrae and discs were measured using ImageJ. Kinematic data were retrieved from studies involving intact testing of the fresh cervical spine segments (Occiput-T1), with all tests performed in the same lab using the same protocol. Loads of 1.5 N m were applied while optoelectronically measuring intervertebral rotations in 3 planes (FE, LB, and AR) at all spinal levels. Relationships between donor information (vertebrae and disc dimensions, subject height, weight, gender), bone quality (BMD), and kinematic data (range of motion [ROM], CF LB/AR, and CF AR/LB) were studied using Pearson correlation and multiple regression analyses (P < .05). Results: Overall values for ROM for intact motion segments were within typical values as reported in the literature (Figure 1). There were significant negative correlations between BMD and ROM, starting at C4-C5 and continuing caudally (R < −0.18, P < .02). Other significant correlations with ROM include the following: cervical level (all directions of motion: R < −0.18, P < .001]), vertebral body height (FL-EX: R = −0.15, P = .004), vertebral body width (AR: R = −0.22, P < .001; LB: R = −0.28, P < .001), lateral mass height (AR: R = 0.11, P < .05; LB: R = −0.18, P < .03), disc height (FL: R = −0.14, P = .014), donor age (LB: R = −0.11, P < .009), gender (FL-EX and LB: male < female, P < .001), weight (FL-EX and LB: R = −0.10, P < .02), height (FL-EX and LB: R < −0.13, P < .003), and body mass index (BMI; LB: R = −0.132, P = .004). There were significant correlations between both CMs and cervical level (Figure 2), with AR/LB decreasing caudally and LB/AR following the trend of primary motion ratios (LB/AR). There were no correlations between CM and donor information (P > .4). Multiple linear regression analysis showed that ROM of cervical motion segments can be predicted using relationships between subject cervical level, gender, age, weight, and/or BMD (FL-EX: R = 0.321, P < .001; AR: R = 0.643, P < .001; and LB: R = 0.535, P < .001). Conclusion: Significant relationships exist between cervical spine segments’ dimensions, BMD, and biomechanical properties (ROM and CM) at mid- and lower cervical levels. These relationships are of value especially during the cervical spine model validation process, and in studies focusing on motion preservation.

OP05: Arthroplasty-Cervical

A025: Reconsideration of Patient Selection for Cervical Disc Replacement (CDR) Based on Minimum 10-Year Follow-up Results

Feifei Zhou¹ and Yu Sun¹

¹Peking University Third Hospital, Beijing, China

Introduction: Cervical disc replacement (CDR) can be complicated by heterotopic ossification (HO) and unwanted ankylosis at the index level, which some authors considered it as an inevitable consequence of cervical nonfusion surgery. Through a tough learning curve of CDR, we noticed that patients with more severe preoperative cervical spondylosis had higher rates of postoperative ossification. Most of indications for different cervical artificial prostheses were described qualitatively, but lack of quantitative criteria. The objective of our study was to explore the quantitative preoperative degenerative changes so as to control the incidence of HO at a relatively low level based on a minimum of 10-year follow-up data. Methods: A retrospective review was performed on 54 patients (66 levels) who underwent Bryan disc replacements by a single group within one institution between December 2003 and August 2008. All patients had at least 10 years of clinical follow-up. Postoperative bone formation at the index level was graded on lateral cervical spine radiographs using the McAfee classification. Preoperative degeneration of the index level was measured by quantitative scoring system based on neutral lateral radiographs including disc height loss, anterior osteophytes with respect to the AP diameter of the corresponding vertebral body, and endplate sclerosis (Table 1). The appearance of ossification of anterior longitudinal ligament (OALL), the range of motion (ROM) at the target level, age, and gender were also collected. Descriptive analyses were presented using means and standard deviations for normally distributed variables and t-test was used to compare between 2 groups (age and ROM). The data sets that were not normally distributed were evaluated with the Mann-Whitney U-test. Fisher’s exact χ² test was used to compare the categorical variables (disc height, anterior osteophytes, endplate sclerosis, and OALL) between the groups. According to univariate analysis, the significant factors were analyzed with the multifactor logistic regression analysis. The receiver operating characteristic curve (ROC) was conducted to identify the optimal cutoff point for the significant factors. Results: The study patients had an average age of 44 years, with a mean follow-up of 120.3 months. The overall incidence of postoperative bone formation was 68.2% (45/66 levels). According to the univariate analysis, gender, disc height, anterior osteophyte, and endplate sclerosis were included for multivariate analysis. The multivariate logistic regression identified disc height, anterior osteophyte, and endplate sclerosis as the independent risk factors for postoperative bone formation while the optimal cutoff point for disc height, anterior osteophyte, and endplate sclerosis were 0.5, 1.5, and 1.5 in terms of ROC curve. Conclusion: The overall incidence of postoperative bone formation after CDR was relatively high when patients are followed for greater than 10 years. Most notably, patient selection in preoperative degeneration of the target level plays an important role in postoperative ossification. Rigorous indication, nearly normal disc height, mild anterior osteophyte, and endplate sclerosis should be considered in Bryan disc replacement.

Table 1.

Scoring System of Cervical Disc Degeneration Based on Neutral Lateral Radiographs (Walraevens J, Liu B, Sloten JV, Goffin J. Qualitative and quantitative assessment of degeneration of cervical intervertebral discs and facet joints. Eur Spine J. 2009;18:358-369.)

A026: Clinical Significance of Bone Loss Phenomenon of Vertebral Bodies at the Operative Segment After Cervical Arthroplasty

Dong Hwa Heo¹ and Choon Keun Park¹

¹The Leon Wiltse Memorial Hospital, Suwon, Republic of Korea

Introduction: Heterotopic ossification and spontaneous fusion are complications of cervical arthroplasty. In contrast, bone loss phenomenon of vertebral bodies at the operation segment after cervical arthroplasty has also been observed. The purpose of this study is to investigate a potential complication—bone loss of the anterior portion of the vertebral bodies at the surgically treated segment after cervical total disc replacement (TDR)—and discuss the clinical significance. Material and Methods: All enrolled patients underwent follow-up for more than 24 months after cervical arthroplasty using the Prodisc-C, Prodisc-vivo, Mobi-C, Discocerv, and Baguera C disc. Clinical evaluations included recording demographic data and measuring the Visual Analog Scale and Neck Disability Index scores. Radiographic evaluations included measurements of the functional spinal unit’s range of motion and changes such as bone loss and heterotopic ossification. The grading of the bone loss of the operative segment was classified as follows: Grade 1, disappearance of the anterior osteophyte or small minor bone loss; Grade 2, bone loss of the anterior portion of the vertebral bodies at the operation segment without exposure of the artificial disc; or Grade 3, significant bone loss with exposure of the anterior portion of the artificial disc. Results: A total of 247 patients were enrolled in this study. Among them, bone loss developed in 114 patients (Grade 1 in 40 patients, Grade 2 in 38 patients, and Grade 3 in 36 patients). Grade 3 bone loss was significantly associated with postoperative neck pain (P < .05). Incidence of bone loss was significantly higher in Baguera-C than other types of artificial discs (P < .05). In contrast, incidence of heterotopic ossification was significantly higher in Prodisc-C (P < .05). Bone loss was related to the motion preservation effect of the operative segment after cervical arthroplasty in contrast to heterotopic ossification. Conclusion: Bone loss may be a potential complication of cervical TDR and affect early postoperative neck pain. However, it did not affect mid- to long-term clinical outcomes or prosthetic failure at the last follow-up. Also, this phenomenon may result in the motion preservation effect in the operative segment after cervical TDR.

A027: Correlation Between Heterotopic Ossification After Artificial Cervical Disc Replacement and Intervertebral Disc Degeneration

Xiongsheng Chen¹, Yin Zhao¹, Shengyuan Zhou¹, and Bo Yuan¹

¹Changzheng Hospital, Second Military Medical University, Shanghai, China

Introduction: There were kinds of factors influencing the incidence of heterotopic ossification (HO) after artificial cervical disc replacement (ACDR), while not including reports about the effects of intervertebral disc degeneration on the occurrence of HO. In view of this, the current research retrospectively analyzed the data of 120 patients having accepted ACDR, to investigate the correlation between the HO after ACDR and the degeneration of cervical intervertebral disc. Material and Methods: From January 2009 to June 2016, 120 patients who had undergone ACDR (Discover, Depuy Spine) were included into this study. There were 87 males and 33 females with an average age of 43.83 ± 8.76 years (range 27-67 years). There were 100 patients who had undergone single-level implant and 20 patients double-level implant. Cervical spine flexion-extension X-rays were taken to assess the range of motion (ROM) of the surgical level. HO situation were evaluated by McAfee classification through cervical lateral X-rays. Degeneration degree of patients’ cervical intervertebral discs were evaluated by Pfirrmann grading standards with the preoperative cervical magnetic resonance imaging (MRI). The occurrence and McAfee classification of HO in the replacement segment at the last follow-up were done. According to whether HO occurred, the patients were divided into HO group and non-HO group. ROM and preoperative Pfirrmann score were compared between the 2 groups of patients. Results: All 120 patients were followed-up. The average follow-up time was 52.29 months (13-102 months). The Visual Analogue Scale (VAS) score, Japanese Orthopaedic Association (JOA) score, and Neck Disability Index (NDI) index all improved at the last follow-up. At the last follow-up, a total of 27 patients developed HO, and the incidence of HO was 22.5%. ROM of the HO group (7.40 ± 1.75°) was significantly smaller than that of the non-HO group (9.79 ± 1.40°). Pfirrmann score of preoperative disc degeneration (3.31 ± 1.11) was significantly higher in HO group than in the non-HO group (2.40 ± 0.96). The correlation analysis showed that the Pfirrmann degeneration score of the preoperative replacement segment of the HO group was positively correlated with the postoperative HO McAfee classification (R = 0.765, P < .01), and negatively correlated with ROM of the replacement segment (R = −0.866, P < .01). Conclusion: The occurrence of HO after ACDR was correlated with the degree of preoperative degeneration of intervertebral disc. The degree of degeneration of cervical disc in HO group was significantly higher than that of non-HO group. In the HO group, the higher the degree of Pfirrmann degeneration of the intervertebral disc preoperatively was, the higher the McAfee grade at the last follow-up was and the lower ROM of the replacement segment was.

A028: Single-Level Cervical Arthroplasty With ProDisc-C Artificial Disc: 10-Year Follow-up Results in One Center

Yanbin Zhao¹, Feifei Zhou¹, and Yu Sun¹

¹Peking University Third Hospital, Beijing, China

Introduction: The aim of this study was to evaluate the clinical and radiographic outcomes of cervical arthroplasty using the ProDisc-C prosthesis. Material and Methods: Clinical and radiographic evaluations, including dynamic flexion-extension lateral images, were performed at baseline and at final follow-up. Results: Twenty-one patients who had single-level ProDisc-C arthroplasty were followed up for a mean period of 122 months. The range of motion at the operated level was 9.1 ± 4.0° at baseline and 6.3 ± 3.7° at final follow-up. Fifteen of 21 levels (71.4%) developed heterotopic ossification: 3 were classified as grade I or grade II heterotopic ossification, 12 were classified as grade III or grade IV heterotopic ossification, according to McAfee’s classification. Two patients developed adjacent segment diseases and received the revision surgeries. Conclusion: ProDisc-C arthroplasty had acceptable clinical and radiographic results at 10-year follow-up. Heterotopic ossification was common after ProDisc-C arthroplasty, which decreased the range of motion.

A029: Clinical and Radiographic Outcomes for a Next-Generation Artificial Cervical Disc: Two-Year Follow-up at 5 Investigational Centers Participating in an US FDA-Approved Study

Richard Guyer¹, Dom Coric², Frank Phillips³, Rick Sasso⁴, Andew Sama⁵, Frank Cammisa⁵, Scott Blumenthal¹, Todd Albert⁵, and Jack Zigler¹

¹Center for Disc Replacement at Texas Back Institute, Plano, TX, USA

²Carolina Neurosurgery & Spine Associates, Charlotte, NC, USA

³Rush University Medical Center, Chicago, IL, USA

⁴Indiana Spine Group, Carmel, IN, USA

⁵Hospital for Special Surgery, New York, NY, USA

Introduction: Ball-and-socket design cervical disc replacements have been previously described regarding their clinical and radiographic outcomes. A next-generation artificial cervical disc (M6-C^; Spinal Kinetics, an Orthofix Company), which mimics the natural disc morphology and biomechanical characteristics, is currently the subject of a US FDA clinical trial. The unique design of this cervical disc provides 6 independent degrees of freedom including axial compression, flexion-extension, lateral bending, translation, and axial rotation. Materials and Methods: The 5 investigational centers participated in a prospective, multicenter, FDA IDE study of patients with cervical radiculopathy who had not improved after at least 6 weeks of nonsurgical treatment. Skeletally mature patients were implanted with the cervical disc at one level from C3 to C7. Patients were enrolled according to specified entry criteria including preoperative Neck Disability Index (NDI) score of ≥30%, neck or arm pain Visual Analogue Scales (VAS) ≥4, with additional study entry criteria similar to prior FDA IDE studies. Patient evaluation included the NDI, Neck and Arm VAS, neurological assessment, and radiographic assessments. Radiographs were obtained for both quantitative and qualitative assessments using validated methods (QMA; Medical Metrics, Inc). Results: Eighty-three patients (43 males) with a mean age at surgery of 44 years and a mean length of symptoms prior to surgery of 24 months were included in this analysis. Mean blood loss was 31 cc. Surgical time was 83 minutes and mean hospital stay was 0.5 days. Baseline mean NDI was 55.4 and at 2-year follow-up was significantly decreased to 11.3 (P < .001). There was a significant reduction in VAS neck pain from 7.3 preoperative to 1.3 at 2 years (P < .001). Mean predominant preoperative arm VAS pain was 7.2 and improved to 0.5 at 2 years (P < .001). Radiographically there were no migrations, no cases of neurological deterioration as a result of the procedure, and one mild radiolucency at 24 months. Additionally, there were no reports of revision or supplemental fixation during the follow-up period. There was one removal due to suspected infection and one reoperation 1 month postoperatively due to incomplete decompression at the index surgery. The mean index level ROM (range of motion) improved from 7.8° preoperative to 8.1°. Conclusion: The results from this 2-year follow-up of cervical disc replacement subjects implanted with a next-generation design demonstrate the expected clinical outcomes. The results to date may be comparable to, or better than, prior reports of other cervical discs but a limited cohort population does not allow a full comparison at this time.

A030: Influence of Prosthesis Design on the Relationship Between Preoperative and Postoperative Flexion-Extension Range of Motion After Cervical Disc Arthroplasty

Avinash Patwardhan¹, Robert M Havey², and Saeed Khayatzadeh²

¹Department of Orthopaedic Surgery and Rehabilitation, Maywood, IL, USA

²Edward Hines Jr. VA Hospital, Hines, IL, USA

Introduction: Biomechanical goals of cervical disc arthroplasty are to (1) restore physiologic motion and (2) maintain stability at the index segment. In a patient with limited mobility at the index, segment disc arthroplasty should restore normal physiologic motion. Conversely, if a motion segment is hypermobile, the prosthesis should restore stability by eliminating hypermobility. We investigated the influence of prosthesis design on the relationship between preoperative and postoperative FE-ROM (flexion-extension range of motion) after cervical disc replacement using fixed-core, mobile-core, and restrained compressible-core prostheses. Material and Methods: Sixty-Five C5-C6 and C6-C7 motion segments were tested using identical methodology. The flexion-extension motions of cervical segments before and after disc replacement were investigated using 5 different cervical disc prostheses classified into 3 design groups: Fixed-Core (N = 33): PRODISC C, DISCOVER, and PCM; age: 56 ± 10, 19 males/9 female, 28 C5-C6 and 5 C6-C7; Mobile-Core (N = 16): Mobi-C; Age: 42 ± 6, 5 males/3 female, 8 C5-C6 and 8 C6-C7; Restrained Compressible-Core (N = 16): M6 C; Age: 43 ± 5, 5 males/3 females, 8 C5-C6 and 8 C6-C7. Each cervical spine specimen underwent kinematic testing (FE moment = ±1.5 N m, compressive follower preload = 150 N) using the same testing apparatus and protocol. Results: Fixed-Core: FE-ROM after TDA was positively correlated with preoperative FE-ROM (R ² = 0.65, P < .0001). Preoperative FE-ROM ranged from 5° to 19°, whereas motion after disc arthroplasty ranged from 2° to 23°. Mobile-Core: FE-ROM after TDA was positively correlated with preoperative FE-ROM (R ² = 0.36, P = .014). Preoperative FE-ROM ranged from 6° to 20°, whereas motion after disc arthroplasty ranged from 8° to 24°. Restrained Compressible-Core: FE-ROM after TDA was not correlated with preoperative FE-ROM (R ² = 0.17, P = .12). Preoperative FE-ROM ranged from 5° to 21°, whereas motion after disc arthroplasty ranged from 8° to 17°. Conclusion: We observed that disc replacement using either a fixed-core or moving-core design yielded postoperative FE-ROM that was positively and significantly correlated with preoperative FE-ROM (P < .05). Arthroplasty with the mobile-core prosthesis allowed segmental motions far exceeding the physiologic range, with 6/16 implanted segments yielding motions of 20° or greater. Following disc arthroplasty using a restrained compressible-core prosthesis, segments with limited mobility had physiologic ROM whereas hypermobile segments showed more controlled postoperative motion. These results suggest that a restrained compressible-core prosthesis that provides progressive resistance to angular motion like the natural anatomy can better achieve both goals of restoring physiologic motion without sacrificing stability.

OP06: Trauma-Cervical 2

A031: Systemic Protein Kinase Inhibition Mitigates Immune Cells Infiltration After Spinal Cord Injury

Mohammad-Masoud Zavvarian¹, James Hong¹, Jian Wang², and Michael Fehlings¹

¹University of Toronto, Toronto, Ontario, Canada

²Toronto Western Hospital, University Health Network, Toronto, Ontario, Canada

Introduction: Traumatic spinal cord injury (SCI) is a debilitating and multifaceted condition that limits the quality of life for millions of patients worldwide. Despite recent advances in the posttraumatic care, the available treatment options for SCI patients are still limited. The disruption of blood-spinal cord barrier (BSCB) by the mechanical trauma is a major challenge that limits the potency of most SCI treatments. Infiltration of pro-inflammatory immune cells following the BSCB disruption leads to further permanent damage to the injury epicenter. Therapeutic stabilization of BSCB can potentially attenuate the immune cells migration and improve SCI recovery. Protein kinase C (PKC) is a key regulator of the junctional complexes in BSCB, and its inhibition has been shown to enhance the integrity of the BSCB in non-SCI injury models. The aim of this study is to analyze the effect of PKC inhibition using midostaurin—a clinically approved protein kinase C inhibitor—on BSCB integrity, and to determine their efficacy as a treatment for SCI. Materials and Methods: SCI was induced in 12 Wistar rats using the clip-compression injury model at C6-7. The injured rats were randomly assigned to 2 cohorts (n = 6), receiving either midostaurin or vehicle control. In addition, 6 laminectomized shams were used to constitute a noninjured cohort. All rats were sacrificed at 24 hours postoperation, and the total RNA and protein were extracted from the spinal cord to evaluate the molecular changes. Western blotting was used to assess the phosphorylation of downstream target molecules. Differentially expressed genes were identified using reverse transcription qualitative polymerase chain reaction. Multiplex Luminex assay was used to examine the inflammatory response after SCI. Results: Administration of 25 mg/kg midostaurin reduced the phosphorylated GSK3 and STAT3 at the injury epicenter (1 day postinjury). This demonstrates the penetrance of the midostaurin into the spinal cord. The transcriptional analysis reveals downregulation of adhesive and migratory genes including JAM2, THY1, and ITGB1. This ultimately leads to the mitigation of pro-inflammatory markers, such as fractalkine, IL-1a, and IL-5 at 1-day postinjury. Conclusions: This study demonstrates that systemic protein kinase C inhibition is an effective strategy for preventing secondary SCI damage, which can have a significant impact on the enhancement of neuroprotective regime applied on traumatic SCI. In the future, a dose-response study will be performed to determine the most effective dosage of midostaurin for reducing BSCB permeability after SCI. The results obtained from this study will further our understanding of molecular changes in SCI and enhance the availability of clinical treatments in order to improve the quality of life for SCI patients.

A032: Early Surgical Decompression Improves Neurological Outcome After Complete Traumatic Cervical Spinal Cord Injury: A Meta-Analysis

Paula Valerie ter Wengel¹, P. C. de Witt Hamer², J. C. Pauptit², N. A. Van der Gaag³, F. C. Oner⁴, and W. P. Vandertop²

¹Slotervaart Hospital, Amsterdam, Netherlands

²Amsterdam UMC, VUmc, Amsterdam, Netherlands

³LUMC, The Hague, Netherlands

⁴UMCU, Utrecht, Netherlands

Introduction: In patients with traumatic spinal cord injury (tSCI), a distinction in surgical urgency is made on the basis of severity of initial neurological injury. The optimal timing of surgical decompression, as well as its impact on neurological recovery, is as of yet undetermined. This study addresses neurological improvement after early and late surgery for complete and incomplete cervical traumatic spinal cord injury. Material and Methods: A systematic search retrieved 15 publications of observational studies reporting outcome measurements after surgery in 1.126 patients with cervical tSCI from PubMed and Embase databases. Surgery was considered early within 24 hours, and late thereafter. An improvement of at least 2 grades on the ASIA (American Spinal Injury Association) scale was considered clinically meaningful. The MOOSE guidelines were followed. Improvement rates were summarized using individual patient data in a Bayesian random effects model and compared for those with complete and incomplete tSCI after early and late surgery. Results: In patients with complete cervical tSCI (n = 422), improvement was more frequent after early surgery than after late surgery (respectively, 22.6%, 95% confidence interval [CI] = 16.6% to 28.7%, and 10.4%, 95% CI = 5.6% to 15.8%; odds ratio [OR] = 2.6 [95% CI = 1.4-5.1]). Whereas in patients with incomplete cervical tSCI (n = 636), improvement was similar between early and late surgery (respectively, 30.4%, 95% CI = 19.8% to 41.6%, and 32.5%, 95% CI = 21.4% to 45.8%; OR = 0.9 [95% CI = 0.4-1.9]). Conclusion: These data suggest a paradigm shift in the treatment of patients with complete cervical tSCI, as surgical decompression within 24 hours is more frequently associated with clinically meaningful improvement. In incomplete cervical tSCI, neurological outcome is similar between early and late surgery.

A033: Does the Cervical X-Ray and CT Provide False Security After Trauma? Sensitivity and Specificity of X-Ray, CT, and MRI in Patient With Cervical Trauma

Sebastian Bigdon¹, Pascale Amrein¹, Sven Hoppe¹, Moritz Deml¹, Aristomenis Exadaktylos¹, Lorin Benneker¹, and Christoph Albers¹

¹Inselspital University of Bern, Bern, Switzerland

Introduction: Trauma with consecutive neck pain is a common reason for admission to the emergency department. The Canadian C-Spine rules and the NEXUS criteria are well accepted to guide decision making toward radiographic workup. The modality of imaging depends on the patients’ history and clinical symptoms; however, there is no consensus whether conventional X-ray, computed tomography (CT), or magnetic resonance imaging (MRI) is superior to confirm or rule out C-spine injuries. Materials and Methods: We retrospectively reviewed 4110 consecutive patients with neck pain after a trauma that were admitted to the emergency department (ED) of our level 1 trauma center. The available imaging as well as the modality of imaging obtained during the presentation in ED was documented. Thereafter, radiographic findings documented by the radiologist were compared to the report from the spine surgeon, and the presence or absence of any kind of injury was compared between the 2 reports. In case of disagreement between the 2 reports, 2 orthopedic surgeons specialized in spine surgery blinded to the former report independently reviewed the images to find a consensus. Results: In 540 patients primary conventional X-ray of the C-spine was performed. An injury was detected in 10 patients (2%). In 390 patients additional CT scans or MRI was conducted. The evaluation of all acquired images revealed that 32 injuries to the cervical spine were missed on plain films leading to a 24% sensitivity of conventional X-ray. There were no false positive injuries diagnosed on conventional X-ray as compared to CT or MRI, resulting in a specificity of 100%. In 2193 patients, primary CT scan was performed. One hundred sixty-four patients had an injury of the C-spine (7%). Ninety out of the 2193 patients received additional MRI scans (4%) revealing 10 additional injuries, of which 5 required immediate surgical therapy. One initially suspected injury on CT could not be confirmed on MRI. The resulting overall sensitivity of CT was 94% and specificity of 99%, respectively. Sixty-three patients underwent MRI without previous X-ray or CT to assess possible C-spine injuries. Three injuries were found. A total of 160 patients had MRI with additional CT and/or X-ray. Thirty-four injuries were detected on MRI with no missed injury and no false positive report as compared to the other 2 imaging modalities, leading to a 100% sensitivity and specificity of MRI. Conclusion: Our study is to this point the largest evaluation of patients with traumatic neck pain admitted to a level 1 trauma center. We found that X-ray examinations have a poor sensitivity to find injuries to the cervical spine, missing nearly 7 out of 10 injuries in this series. While CT scan has a very good sensitivity and is routinely performed to confirm or rule out an injury, there are still relevant injuries missed using only the CT. In our retrospective analysis, the MRI was the best diagnostic modality with a sensitivity and specificity of 100% to detect C-spine injury following trauma.

A034: Temporary Atlantoaxial Segmental Fixation and Reduction for Grauer Type IIB Dens Fractures

Nanjian Xu¹ and Weihu Ma¹

¹Sixth Hospital of Ningbo, Ningbo, China

Introduction: The treatment of Grauer type IIB dens fractures not amenable to anterior screw fixation remains a challenge for spinal surgeons because of preservation of atlantoaxial rotation. Although most of the atlantoaxial fixation techniques used to address dens fractures achieve satisfactory stability and bone fusion, atlantoaxial rotation is rarely preserved. The purpose of this study is to assess the clinical results of temporary atlantoaxial segmental fixation and reduction for Grauer type IIB dens fractures, which may traditionally require posterior atlantoaxial fusion. Material and Methods: From May 2010 to June 2015, 23 patients with Grauer type IIB dens fractures not amenable to anterior screw fixation were enrolled and treated using 1-stage temporary atlantoaxial segmental fixation and reduction without fusion. There were 15 males and 8 females, aged from 14 to 68 years with an average of 47.3 ± 16.19 years. The internal fixations were removed after the consolidation of the fractures was radiologically confirmed. Dynamic CT (computed tomography) was performed >1 year after implant removal to evaluate atlantoaxial rotation. Visual analogue scale (VAS) and Neck Disability Index (NDI) were used to evaluate the clinical efficacy. Results: Two patients had numbness of occipital cervical, and both of them recovered within 2 months. All patients were followed-up. The follow-up time ranged from 18 to 25 months (average 21.61 ± 2.35). The healing time of bone fusion ranged from 6 to 10 months (average 8.26 ± 1.25). At a mean of 8 months after the initial operation, fracture union was confirmed and implants were removed. There were no complications associated with instrumentation. Fracture healing was observed in 23 patients and the instrumentation was removed. After removing the instrumentation, the NDI had improved (P < .01). The average preoperative VAS score was 6.78 ± 0.67, while the score of last follow-up was 0.43 ± 0.51, with statistically significant differences (P < .01). The mean atlantoaxial rotation was 15.80 ± 5.16° to the left and 15.80 ± 5.31° to the right. The average total atlantoaxial rotation was 31.60 ± 9.87°. Conclusion: Temporary atlantoaxial segmental fixation and reduction could effectively avoid the loss of atlantoaxial rotation caused by posterior atlantoaxial fusion.

A035: Description of a New Odontoid Fracture Variant: Proposal for a Modification of the Anderson and D’Alonzo Odontoid Fracture Classification

David Gendelberg¹, Haitao Zhou¹, Amy Cizik¹, Nathan Wigner¹, Christina Cheng¹, Quynh Nguyen¹, and Carlo Bellabarba¹

¹University of Washington, Harborview Medical Center, Seattle, WA, USA

Introduction: Odontoid fractures are common cervical spine injuries in the adult population. The classification system described by Anderson and D’Alonzo is the most frequently used to describe odontoid fractures. Type 3 odontoid fractures are usually amenable to conservative management and type 2 fracture management is controversial. We have observed an odontoid fracture variant that consists of an odontoid fracture and a concurrent fracture in the coronal plane that separates the pars-interarticularis from the rest of the C2 body. Therefore, the purpose of this paper is to describe this odontoid fracture variant and designate it as a type 4 odontoid fracture in the Anderson and D’Alonzo classification. Materials and Methods: A more precise distinction between odontoid fractures with and without a fracture that separates the pars-interarticularis from the C2 body is proposed. Criteria that define this fracture variant include the following: (1) coronal fracture line separating pars-interarticularis from rest of the C2 body; (2) any type of odontoid fracture; and (3) unilateral versus bilateral pars involvement. To evaluate for reproducibility, a Cohen κ test with a power analysis was performed to evaluate interobserver reliability among 2 spine surgeons for the proposed classification system modification. Results: Sixteen odontoid fracture variants were reviewed, of which 13 were acutely treated operatively. Of the 3 fractures treated conservatively, 2 cases failed and displaced and 1 case failed halo management and was subsequently treated operatively. Cohen κ test for interobserver reliability was found to be 0.686, indicating “good” agreement. Conclusion: Odontoid fractures in which one or both pars-interarticularis are separated from the vertebral body are clinically distinct fracture patterns, which differ from the other odontoid fracture types in the Anderson and D’Alonzo classification. These fractures are relatively common. In one study, where they were described as “complex C2 fractures,” and they compromised 19% of all C2 fractures. We feel that this odontoid fracture variant indicates a greater degree of instability and may therefore benefit from surgical intervention. Furthermore, identification of this fracture variant has been reproduced among different surgeons. In light of its frequency and its influence on clinical decision making we believe that it should be designated as a type 4 odontoid fracture.

A036: Radiologic Factors to Predict Injury of Transverse Atlantal Ligament in Unilateral Sagittally Split Fracture of C1 Lateral Mass

Jong-Beom Park¹ and Do-Gyun Kim¹

¹Uijeongbu St. Mary’s Hospital, The Catholic University of Korea, Uijeongbu, Republic of Korea

Introduction: Unilateral sagittally split fracture (USSF) of C1 lateral mass (LM) is a variant type of C1 atlas fracture. Recently, it is recognized as unstable fracture, which causes late deformity of occipitocervical junction that requires extensive reconstructive surgery. Since USSF of C1 LM is rare, its definite treatment guideline has not been established. The integrity of transverse atlantal ligament (TAL) is a key factor to determine whether to treat surgically or nonsurgically in C1 atlas fracture. However, no information is available about which type of USSF of C1 LM is associated with injury of TAL. Therefore, we performed the current study to investigate radiologic factors to predict injury of TAL in USSF of C1 LM. Material and Methods: Twenty-six consecutive cases of USSF of C1 LM were included from 5 trauma centers of tertiary university hospitals. The fractures associated with other cervical spines, such as C2 and occiput, were excluded from the study. The mean age was 52 years. Sixteen were males and 10 were females. Two radiologists determined presence of TAL injury in MRI (magnetic resonance imaging) using Dickman’s classification and divided into 2 groups: TAL injury and TAL intact. If the results of 2 judgements were not identical, the third radiologist reevaluated. Three spine surgeons measured radiologic parameters and the averages were used as final results: Total LM displacement (LMD), unilateral LMD at fracture side, atlanto-dental interval (ADI), fracture gap, clivus canal angle (CCA), atlanto-occipital joint axis angle (AOJAA), and basion-dens interval (BDI). The radiologic results were compared between 2 groups. The incidence of associated other C1 fractures was also investigated and compared between the 2 groups. Results: Sixteen were TAL injury group (9 type I and 7 type II) and 10 were TAL intact group. Total LMD and unilateral LMD at fracture side were higher in TAL injury group than TAL intact group (5.9 mm vs 1.2 mm, P < .001, and 4.3 mm vs 1.0 mm, P < .001, respectively). ADI and fracture gap were higher in TAL injury group than TAL intact group (2.0 mm vs 1.5 mm, P < .05, and 6.9 mm vs 2.1 mm, P < .001, respectively). However, CCA, AOJAA, and BDI were not statistically different between the 2 groups (155.6° vs 154.9°, P = .824; 107.8° vs 105.9°, P = .676; and 4.4 mm vs 4.2 mm, P = .751. Total LMD was positively correlated to unilateral LMD at fracture side (CC = 0.937, P < .001), ADI (CC = 0.449, P < .01), and fracture gap (CC = 0.658, P < .001), but not CCA, AOJAA, and BDI (CC = −0.221, P = .279; CC = −0.042, P = .837; and CC = −0.138, P = .502). Incidence of associated other C1 fractures was higher in TAL injury group that TAL intact group (87% vs 20%, P < .001). Conclusion: Our results suggest that total LMD more than 5.9 mm and unilateral LMD at fracture side more than 4.3 mm are radiological factors to predict injury of TAL in USSF of C1 LM. For such cases, we recommend early surgical stabilization to avoid coronal and sagittal malalignment of occipitocervical junction, resulting in unsatisfactory clinical outcomes.

OP07: Infections 1

A037: Hemoglobin A1c as a Predictor of Surgical Site Infection Following Single-Level Lumbar/Lumbosacral Posterior Fusion in Patients With Diabetes

Dong Wuk Son¹, Geun Sung Song¹, Sang Weon Lee¹, Soon Ki Sung¹, Jun Seok Lee¹, and Doo Kyung Son¹

¹Pusan National University Yangsan Hospital, Pusan, Republic of Korea

Introduction: Diabetes mellitus (DM) is a prevalent disease of glucose dysregulation that has been associated with an increased risk for postoperative infection following spine surgery. The status of preoperative blood glucose management may affect the risk of surgical site infection (SSI). In the present study, we evaluated the association between preoperative glycemic control as indicated by hemoglobin A1c (HbA1c) in patients with diabetes and incidence of postoperative SSI following single-level lumbar/lumbosacral posterior fusion. Furthermore, we calculated a threshold level of HbA1c above which the risk of SSI after posterior lumbar/lumbosacral arthrodesis increases significantly in diabetic patients. Material and Methods: From January 2009 to December 2015, a total of 92 patients who underwent single-level lumbar/lumbosacral posterior fusion with a diagnosis of DM who had a preoperative HbA1c recorded within 4 weeks of surgery were included in the study. Patients were divided into 2 groups according to whether they had SSI, and then demographic/clinical data were compared. A receiver operating characteristic (ROC) and area under the curve (AUC) analysis was conducted to define the cutoff value of HbA1c above which the risk of SSI was significantly increased. Including this value, potential variables were verified by multiple logistic regression analysis. Results: Twenty-four patients were treated for SSI and 68 patients maintained noninfectious condition within 1 year. Three of the 24 (12.5%) patients developed SSI in the deep layer requiring operative irrigation and debridement. The preoperative HbA1c value was significantly higher in patients with SSI (6.8%) than in those without SSI (6.0%; P = .008). The results of ROC analysis determined that HbA1c ≥6.9% could serve as a threshold for significantly increased risk of SSI (P = .003, AUC = 0.708, sensitivity = 62.5%, specificity = 70.6%). After adjusting for potential confounders, there was a significant association between preoperative HbA1c and occurrence of SSI (P = .008, odds ratio = 4.500, 95% confidence interval = 1.486-13.624). Conclusion: In patients with diabetes, the preoperative glycemic control as indicated by HbA1c is an independent risk factor for SSI following single-level lumbar/lumbosacral posterior fusion. Particularly when preoperative HbA1c exceeded 6.9%, the risk of SSI significantly increased.

A038: Factors Associated With an Increased Risk of Developing a Postoperative Infection Following Spine Surgery

Mina Aziz¹, Greg McIntosh², Michael Johnson³, Charles Fisher⁴, Michael Weber¹, and Michael Goytan³

¹McGill University, Montreal, Quebec, Canada

²Canadian Spine Outcomes and Research Network, Mandrake, Ontario, Canada

³University of Manitoba, Winnipeg, Manitoba, Canada

⁴University of British Columbia, Vancouver, British Columbia, Canada

Introduction: Postoperative infection is a serious complication of spine surgery and can contribute to the strain on the health care system’s resources; some studies have estimated the cost of such an infection to be $200 000 per patient. The purpose of this study is to determine what factors affect the risk of developing postoperative infection. We hypothesize that female gender, smoking, diabetes, having thoracolumbar procedures, having a neurological deficit, increased age, body mass index (BMI), American Society of Anesthesiologists (ASA) score, blood loss, and number of operative levels increase the patients’ risk of developing a postoperative infection. Material and Methods: A retrospective review of prospectively collected data within the Canadian Spine Outcome Registry Network (CSORN) was conducted. Data were analyzed using IBM-SPSS. ANOVA was used to analyze continuous variables, while odds ratios were used to analyze categorical variables. Multivariable logistical regression analysis was conducted (odds ratios) to determine any association between the outcome and independent factors. Significance level was P < .05. Results: There were 7747 patients identified from the registry that had completed at least 12 weeks of follow-up. There were 199 infections recorded representing a 2.6% risk of infection. There were no association found between the risk of developing a postoperative infection and gender, age, smoking status, neurological status, and diabetes. Having a higher ASA score was associated with an increased risk of developing a postoperative infection (adjusted odds ratio [OR] = 1.427, 95% confidence interval [CI] = 1.123-1.812, P < .004). Increased intraoperative blood loss was associated with an increased risk of developing a postoperative infection (adjusted OR = 1.0, 95% CI = 1.0-1.001, P < .008). Increased number of operative levels was associated with an increased risk of developing a postoperative infection (adjusted OR = 1.076, 95% CI = 1.003-1.154, P < .004). Having an elevated BMI showed a trend toward an association with developing postoperative infection but did not reach statistical significance (adjusted OR = 1.025, 95% CI = 0.999-1.052, P < .056). Undergoing thoracolumbar procedures also showed a trend toward an association with developing postoperative infection but did not reach statistical significance (adjusted OR = 1.517, 95% CI = 0.953-2.413, P < .079). Conclusion: There is a 2.6% overall rate of postoperative spine infection across 20 Canadian centers. The factors that were associated with an increased risk of developing a postoperative infection were higher ASA scores, increased blood loss, and having more operative levels. Undergoing thoracolumbar procedures and having a higher BMI showed a trend toward association that did not reach statistical significance. This study establishes a benchmark against which the effectiveness of future interventions to reduce infection can be compared.

A039: Prevention of Surgical Site Infection in Major Spinal Surgery: Evidence Update With Focus on Wound Care

Goldenberg Yoni¹, Martin K Hunn¹, Susan Liew¹, and Jin W Tee¹

¹Alfred Hospital, Melbourne, Victoria, Australia

Introduction: Surgical site infection (SSI) is a major strain on health care resources. Rates of SSI in spinal operations are reported to be anywhere between 0.6% and 16%, each with huge economic burden and morbidity to the patient. While the literature appears to suggest that major spine surgery (≥3 levels) increases the risk of SSI, there has been significant variability in the techniques reported in the literature aimed at preventing these wound complications. This review has considered new research and systematic reviews over the last 18 years that focus on wound care techniques which would reduce SSIs, in particularly in major spine surgery, and whether or not those should be incorporated in future guidelines. The domains that were considered are (1) preoperative bathing, (2) skin preparation, (3) skin incision prophylaxis, (4) fascia and skin closure techniques, (5) dressings, and (6) use of drains. Material and Methods: This review was conducted in accordance with the preferred reporting items for systematic reviews and meta-analysis (PRISMA) guidelines. For each domain, electronic databases and reference lists of key articles were searched from January 1, 2000, to August 31, 2018, to identify studies meeting inclusion criteria. Results: Initial search combined for all domains yielded 818 studies. After applying the study inclusion criteria 52 studies were retrieved. Examining independently each domain, 20 publications were selected. The review has revealed mounting evidence in support of preoperative chlorhexidine gluconate bathing and nasal screening for Staphylococcus aureus preoperatively. Statistical significance was not always demonstrated and none were focused on major spine surgery. The most efficacious skin preparation is not established in the literature with some studies showing superiority of chlorhexidine-alcohol solution over aqueous povidone iodine. Other fields in the literature further support alcohol-based antiseptics. There has been strong recommendation for the use of antibiotic prophylaxis. Prolonged antimicrobial prophylaxis, defined as >24 hours after first preoperative dose, does not reduce SSI risk, and major spine surgery was not studied separately in relation to risk-reduction stratification. Limited level I evidence was available for wound closure techniques with only a single systematic review. In that study there was felt to be a lack of evidence for optimal closure technique in posterior spinal surgery. Lower quality studies demonstrated significant reduction in SSI using 2-octyl-cyanoacrylate. No data were available to support a particular technique of fascia closure; however, in other specialties, interrupted and continuous suture fascial closure were found to demonstrate no difference in terms of wound infection or dehiscence while antimicrobial sutures significantly reduced SSI in these cohorts. Evidence regarding dressings and drains in reducing SSI are limited but emerging evidence suggests some utility in those treated with silver-impregnated dressings. Drains have demonstrated to neither increase nor reduce the risk of SSI. There is no evidence to support the use of prophylactic antibiotics while the drains are in situ. Conclusion: This review enables tertiary spine centers to structure a comprehensive wound care plan that is evidence based while acknowledging that more high-quality research is required in order to validate such wound care plans in reducing institutional SSI in major spine surgery patients.

A040: Application of Simultaneous PET-MR Imaging in Pyogenic Spondylodiscitis: Preliminary Study

Ikchan Jeon¹, Eun Jung Kong¹, and Chul Pyo Hong²

¹Yeungnam University Hospital, Daegu, Korea

³Catholic University of Daegu, Daegu, Korea

Introduction: There is still no definite correlation between follow-up magnetic resonance (MR) imaging findings and clinical status for monitoring treatment response in spinal infection. Furthermore, several factors related with general condition can have an influence on blood inflammatory indexes. Recently, fluorine-18 fluorodeoxyglucose positron emission tomography (F-18 FDG PET) shows great potential as a new option. Simultaneous F-18 FDG PET and MR (PET/MR) imaging make us to expect a huge synergic effect in a view of anatomic and metabolic advantages. Material and Methods: Prospective study. From January 2017 to February 2018, 34 patients with pyogenic thoracolumbar spondylodiscitis were enrolled. Decision making for treatment response (good or poor response) was performed based on clinical symptoms, blood inflammatory marker (erythrocyte sedimentation rate [ESR] and C-reactive protein [CRP]), and 1 or 2 times of PET/MR imaging in the patients after at least 3 weeks of antibiotic therapy. We compared the differences of clinical data including ESR, CRP, and PET/MR imaging between good-response group (n = 28, group A) and poor-response group (n = 16, group B) (10 patients of group B were also included to group A after a successful treatment), and analyzed the recurrence rate at least 6 months of follow-up period in group A. Results: Good-response group showed lower ESR (45.64 ± 29.33 vs 72.31 ± 29.20), CRP (0.95 ± 1.38 vs 2.83 ± 2.64), and SUV_max (4.15 ± 1.42 vs 6.84 ± 2.23) in group A than group B (P < .05). There was a difference in the distribution of F-18 FDG uptake depending on treatment response. F-18 FDG uptake was limited to destroyed disc in group A compared with the broad involvement to vertebral body in group B. SUV_max was the most reliable diagnostic method with cutoff value 5.14, sensitivity 93.8%, specificity 82.1%, positive predictive value 75%, and negative predictive value 95.8%. There was one patient with recurrence during the follow-up period (1/28, 3.6%). We found the possibility of DWI and ADC as the specific MR sequences which can reflect the result of PET. Conclusion: PET/MR imaging is an independent valuable method that is unaffected by other conditions with higher sensitivity and negative predictive value for evaluating treatment response in pyogenic spondylodiscitis. We need to obtain the best MR sequences that can reflect the result of PET to be more helpful.

A041: Multidisciplinary Management of Pyogenic Spondylodiscitis: Epidemiological and Clinical Features, Prognostic Factors, and Long-Term Outcomes

Enrico Pola¹, Giovanni Autore¹, Valerio Cipolloni¹, Luca Piccone¹, Massimo Fantoni¹, and Francesco Ciro Tamburrelli¹

¹Catholic University of the Sacred Heart, Rome, Italy

Introduction: Pyogenic spondylodiscitis (PS) is a severe infective disease burdened by high morbidity rates. Although there is rising incidence, the proper management of PS is still much disputed. Objectives of this study were to delineate the clinical features of PS and to analyze the prognostic factors and the long-term outcomes of a large population of patients. Material and Methods: A total of 207 cases of PS treated from 2008 to 2016 with a 2-year follow-up were enrolled. Clinical data from each patient were reported. The outcomes was initially the rate of healing without residual disability and secondary outcomes the length of stay, healing from infection, death, relapse, and residual disability. Binomial logistic regression and multivariate analysis were used to evaluate prognostic factors. Results: Median diagnostic delay was 30 days and the rate of onset neurological impairment was 23.6%. Microbiological diagnosis was established in 155 patients (74.3%), and the median duration of total antibiotic therapy was 148 days. Orthopedic treatment was conservative for 124 patients and surgical in 47 cases. Complete healing without disability was achieved in 142 patients (77.6%). Statistically confirmed negative prognostic factors were negative microbiological culture, neurologic impairment at diagnosis, and underlying endocarditis (P ≤ 0.05). Healing from infection rate was 90.9%, while residual disabilities occurred in 23.5%. Observed mortality rate was 7.8%. Conclusion: The leading predictive factor for successful treatment is microbiological diagnosis. Multidisciplinary management and early diagnosis are necessary to identify underlying aggressive conditions and to reduce neurological complications associated with poorer long-term outcomes. Despite good prognosis rates, PS may lead to major disabilities still representing a challenging disease.

A042: One-Stage Reconstruction Surgery in Tuberculous Spondylitis With Kyphotic Deformity

Asrafi Rizki Gatam¹, Phedy Phedy¹, Fachrisal Ipang¹, Ajiantoro Aji¹, and Luthfi Gatam¹

¹ Fatmawati General Hospital, Jakarta Selatan, Indonesia

Introduction: In managing tuberculous spondylitis with kyphotic deformity, 2-stage combined anterior-posterior approaches were performed to reconstruct both anterior and posterior column. Anterior procedure is known for increase morbidity due to very extensive approach such as thoracotomy or thoracophrenicolumbotomy. Single -stage posterior only with lateral extracavitary approach for anterior column reconstruction is one of the options in managing the cases with less morbidity. The objective of this study is to evaluate clinical outcome and radiologic result kyphotic deformity patients who were treated using one-stage reconstruction surgery. Method: This is a prospective study of tuberculous spondylitis with kyphotic deformity managed with single-stage anterior and posterior reconstruction with lateral extracavitary posterior only approach. Comparison of pre- and postoperation Visual Analogue Scale (VAS), Oswestry Disability Index (ODI), kyphotic angle, and sagittal balance were measured in all the cases. Result: There were 20 tuberculous spondylitis cases managed with this technique. All of the cases were located on thoracal or thoracolumbar junction area. The mean age of the patients was 37.1, 12 males and 8 females. Procedures were performed with lateral extracavitary approach. The mean estimated blood loss and length of surgery were 750 mL and 194 minutes. The mean preoperation local kyphotic angle was 54.1°, and regional kyphotic angle was 37.5°. The mean postoperation local kyphotic angle was 23.9°, and regional kyphotic angle was 16.8°. Two patients had postoperative hypoesthesia in the subsequent dermatome level. All of the patients have good clinical outcomes with no neurologic deficit, with mean 1 year postoperation VAS of 0.3 and ODI of 6.7. Radiology of 1-year follow up showed good fusion on the anterior segment of the spine. Conclusion: Single-stage posterior only with lateral extracavitary approach for anterior column reconstruction and posterior instrumentation is a viable alternative method to treat kyphotic deformities. This surgical procedure will produce a good alignment, strong fixation, better fusion, and reduced morbidity with 1-stage surgery.

OP08: Biomechanics 2

A043: Sequential Changes in Lumbar Lordosis and Segmental Stability Following Lateral Interbody Cage Placement, Smith-Peterson Osteotomy, and Anterior Longitudinal Ligament Release

Richard K. Hurley Jr¹, Amy A. Claeson², Jason A. Inzana², Anup Gandhi², and Zachary Child³

¹Brooke Army Medical Center, Fort Sam Houston, TX, USA

²Zimmer Biomet, Westminster, CO, USA

³Tahoe Orthopedic Institute, South Lake Tahoe, CA, USA

Introduction: Interbody cage systems can be used in isolation to achieve lumbar lordosis correction, but can also be deployed using techniques to shorten the posterior or lengthen the anterior spinal column to yield even larger corrections. The Thomasen pedicle subtraction osteotomy (PSO) yields ∼30° of correction, but with high degree of difficulty and surgical complication rates.^1,2 A lateral lordotic cage with Smith-Petersen Osteotomy (SPO), a hyper-lordotic lateral cage with anterior longitudinal ligament (ALL) release,^3,4 or a lateral cage with combined techniques³ show potential for substantial lordosis correction. The removal or release of local spinal anatomy, however, can increase segmental motion and may inhibit fusion. The objective of this biomechanical cadaver study was to identify the lordosis correction and resulting segmental stability achieved with implantation of lateral lordotic cages of increasing angle coupled with either or both minimally invasive surgical techniques of ALL release and SPO. We hypothesize that a combination of lordotic interbody cage placement, ALL release, and SPO would yield more lordosis at a single level than with these interventions in isolation and that the biomechanical stability of the construct will be sufficiently rigid to promote fusion. Material and Methods: Cadaveric lumbar spines (n = 6) were divided into L1-L2 and L3-L4 segments, potted in plaster with pedicles exposed and randomly distributed (n = 6 segments/group); Group 1: SPO followed by ALL release; Group 2: ALL release followed by SPO. Lateral cages of 8°, 20°, and 30° were placed with posterior spinal fixation at each step. A 1-hole modular fixation plate was added with ALL release and replaced by a 2-hole plate at the last intervention. Sagittal fluoroscopic images for lordosis angle measurement and range of motion (ROM) data were collected after each intervention. A total of 7.5 N m pure moments were applied to segments in flexion-extension (FE), lateral bending (LB), and axial rotation (AR), while motion capture recorded ROM. Linear mixed models with repeated measures provided statistical comparisons at α = .05 within each group for each metric (angle, FE-, LB-, AR-ROM); data reported as percent of native. Results: The combination of SPO with ALL release yielded the largest lordotic angle regardless of procedure order (correction angle: 28.5 ± 7.3°), but also greatest ROM in all planes (FE: 61.5 ± 48.4%, LB: 42.2 ± 42.0%, AR: 162.8 ± 178.3%). A 30° cage provided no additional gain in lordotic angle and increased ROM in FE and AR with ALL release and SPO (FE: 36.8 ± 27.1%, AR: 144.3 ± 138.6%). An 8° lordotic cage with SPO achieved the largest angle correction (10.1 ± 6.8°) while reducing ROM in all planes. Addition of a 2-hole plate does not decrease correction and tends to reduce ROM in all planes compared to the 1-hole plate. Conclusion: Lordosis >15° is possible through minimally invasive techniques of SPO and ALL release and performing a SPO without ALL release yields substantial correction without compromising stability. Performing a SPO after ALL release increases segment ROM in FE, LB, AR, thus a 2-hole plate is recommended to aid in stability with ALL release.

References

1. Bridwell, Lewis SJ, Edwards C, et al. Complications and outcomes of pedicle subtraction osteotomies for fixed sagittal imbalance. Spine (Phila Pa 1976). 2003;28:2093-2101.

2. Bridwell, Lewis SJ, Lenke LG, Baldus C, Blanke K. Pedicle subtraction osteotomy for the treatment of fixed sagittal imbalance. J Bone Joint Surg Am. 2003;85-A:454-463.

3. Melikian R, Yoon ST, Kim JY, Park KY, Yoon C, Hutton W. Sagittal plane correction using the lateral transpsoas approach: a biomechanical study on the effect of cage angle and surgical technique on segmental lordosis. J Neurosurg Spine. 2016;41:E1016-E1021.

4. Deukmedjian AR, Dakwar E, Ahmadian A, Smith DA, Uribe JS. Early outcomes of minimally invasive anterior longitudinal ligament release for correction of sagittal imbalance in patients with adult spinal deformity. Scientific World J. 2012;2012:789 698.

A044: Fixation Strength of Cortical Versus Traditional Pedicle Screws in High-Quality Bone: A Biomechanical Cadaver Study

Steve Chang¹, Amy A. Claeson², Anup Gandhi², and Yoshiyuki Yato³

¹Barrow Brain and Spine, Phoenix, AZ, USA

² Zimmer Biomet, Westminster, CO, USA

³Murayama Medical Center, Musashimurayama, Tokyo, Japan

Introduction: The cortical bone trajectory (CBT) aims to maximize contact between the screw thread and dense cortical bone within the landmarks surrounding the pedicle.^1-3 The medial-to-lateral trajectory allows for reduced exposure and minimized muscle disruption^1,4,5; thus, the technique has been gaining traction as a minimally invasive solution. Additionally, this technique may pose beneficial for osteoporotic patients because cortical bone may be less affected by the osteoporotic cascade than cancellous bone. The CBT typically dictates a shorter screw than the traditional pedicle screw trajectory; thus, the biomechanical fixation strength of the bone-screw interface has been assessed via axial pullout and cyclic toggle to failure using either finite element analyses⁷ or ex vivo cadaveric studies.^1,6,8 We hypothesize cortical screws in the CBT have equivalent fixation strength to traditional screws in the traditional trajectory in both failure modes and superior fixation strength when normalized by the smaller shank volume. Material and Methods: Fresh-frozen cadaveric L1-L5 vertebrae (n = 20) of high-quality bone (bone mineral density [BMD] > 0.80 g/cm²) were stratified into axial PULLOUT and cyclic TOGGLE to failure groups, such that lumbar levels were evenly distributed and bone quality was not significantly different. Zimmer Biomet cortical (5.5 mm Ø × 30-45 mm length) and traditional (6.5 mm Ø × 45 mm length) pedicle screws were bilaterally inserted by a spine surgeon in their intended trajectories. For PULLOUT testing, vertebrae were fixed with the screw of interest aligned axially and coupled to the actuator, which pulled the screw from the pedicle at a rate of 5 mm/min. For TOGGLE testing, vertebral bodies were clamped axially. A rod, secured orthogonal to the screw axis and affixed to the actuator, applied a load-controlled sinusoidal waveform beginning at −100/−10 N and increasing −50/−5 N each 50 cycles until −2 mm of displacement was achieved. Test order (cortical vs traditional) was randomized for each vertebra. Biomechanical outcomes (PULLOUT: peak load, TOGGLE: failure load) were compared using Wilcoxon signed-ranks tests with significance set to α = 0.05 both as recorded and normalized by the shank volume. Results: Cortical screws had 15.7% less shank volume and comparable shank surface area to traditional screws of matched length. Resistance to toggle and pullout strength were comparable between both screws as recorded. Normalized to shank volume, cortical screws in the CBT failed at significantly higher loads in TOGGLE (166 ± 58.4%, P < .05) and significantly higher peak loads in PULLOUT (131 ± 40.6%, P < .05) compared to those of traditional screws. Conclusion: Per volume, cortical pedicle screws in the CBT exhibit superior fixation strength to traditional screws in the traditional trajectory in high-quality bone. Cortical screws in the CBT pose advantages over traditional screws including use as a minimally invasive technique, minimized incidence of medial breach, and reduced artifact from clinical imaging. Furthermore, this technique provides an additional means of posterior fixation in revision or in patients with lower bone quality.

References

1. Santoni BG, Hynes RA, McGilvray KC, et al. Cortical bone trajectory for lumbar pedicle screws. Spine J. 2009;9:366-373.

2. Matsukawa K, Yato Y, Kato T, Imabayashi H, Asazuma T, Nemoto K. In vivo analysis of insertional torque during pedicle screwing using cortical bone trajectory technique. Spine (Phila Pa 1976). 2014;39:E240-E245.

3. Baluch DA, Patel AA, Lullo B, et al. Effect of physiological loads on cortical and traditional pedicle screw fixation. Spine (Phila Pa 1976). 2014;39:E1297-E1302.

4. Matsukawa K, Yato Y, Nemoto O, Imabayashi H, Asazuma T, Nemoto K. Morphometric measurement of cortical bone trajectory for lumbar pedicle screw insertion using computed tomography. J Spinal Disord Tech. 2013;26:E248-E253.

5. Mobbs RJ. The “medio-latero-superior trajectory technique”: an alternative cortical trajectory for pedicle fixation. Orthop Surg. 2013;5:56-59.

6. Sansur CA, Caffes NM, Ibrahimi DM, et al. Biomechanical fixation properties of cortical versus transpedicular screws in the osteoporotic lumbar spine: an in vitro human cadaveric model. J Neurosurg Spine. 2016;25:467-476.

7. Matsukawa K, Yato Y, Imabayashi H, Hosogane N, Asazuma T, Nemoto K. Biomechanical evaluation of the fixation strength of lumbar pedicle screws using cortical bone trajectory: a finite element study. J Neurosurg Spine. 2015;23:471-478.

8. Ueno M, Sakai R, Tanaka K, et al. Should we use cortical bone screws for cortical bone trajectory? J Neurosurg Spine. 2015;22:416-421.

A045: Advanced Degeneration of Articular Facet Joint Cartilage in Adolescent Idiopathic Scoliosis: Cause of Pain

Abdulaziz Bin Shebreen¹, Daniel Bisson², Lisbet Haglund², and Jean Ouellet¹

¹McGill University, Montreal, Quebec, Canada

²McGill Scoliosis & Spinal Research Unit, Montreal, Quebec, Canada

Introduction: Adolescent idiopathic scoliosis (AIS) is a poorly understood deformity of the thoracolumbar spine that affects the vertebrae, intervertebral discs, and the articular facet joints. Thus, in this study, a comprehensive investigation was performed to determine the impact of abnormal biomechanics within the scoliotic spine on the intrinsic composition of the cartilage cells and its extracellular matrix. Material and Methods: Surgically removed articular facet joint tissue were collected from patients undergoing spinal corrective surgery for AIS deformities, while an asymptomatic non-scoliotic articular facets joint tissue were obtained from cadaveric organ donors. Alterations in cartilage tissue structure were evaluated histologically with safranin-O fast green and a modified OARSI grading scale. Evaluation of pro-inflammatory mediators, matrix-degrading proteases, and fragmented matrix molecules associated with cartilage degradation was analyzed by immunohistochemistry, mass spectrometry and western blotting. Results: Safranin-O fast green staining revealed that young scoliotic facet joints show clear signs of degeneration with substantial proteoglycan loss. The proteoglycan levels were significantly lower than in healthy asymptomatic non-scoliotic control individuals. In comparisons to controls, scoliotic articular facets showed increased cell density and immunopositivity to Ki-67, a proliferative marker, along with higher expressions of IL-1b, VEGF, angiogenin, and MMP-13. Expression and fragmentation of the small leucine rich proteins (SLRPs) chondroadherin, decorin, biglycan, lumican, and fibromodulin were analyzed with western blot. Chondroadherin and decorin were fragmented in cartilage from patients with a curve greater than 70° whereas biglycan and fibromodulin did not show curve related fragmentation. Furthermore, this study shows that articular facets in young AIS patients have features comparable to age- or injury-related osteoarthritic changes. Conclusion: AIS facet joints show advanced cartilage deterioration as seen in osteoarthritis.

A046: Epidemiological Examination of Trunk Muscle Strength in Healthy Adults

Philipp Flößel¹, Stefan Zwingenberger¹, Alexander Thomas¹, Achim Walther¹, Heidrun Beck¹, Klaus-Dieter Schaser¹, and Alexander C. Disch¹

¹University Hospital Carl Gustav Carus at Technische Universität Dresden, Dresden, Germany

Introduction: Insufficient trunk muscle strength and muscular dysbalance between flexion and extension are according to literature associated with lower back pain. There are only a few cross-sectional studies with small sample sizes in which isokinetic measurement of trunk muscle strength was performed. The aim of the study was to determine a basic value of trunk muscle strength for healthy people. A model ought to be developed for screening risk-patients with inadequate power or disturbed trunk muscle balance. Material and Methods: A total of 230 healthy probands without complaints took part in the prospective longitudinal study (n = 106/124 male/female; age: 41.6 ± 13.1 years, height: 174 ± 9 cm; weight 73.4 ± 13.1 kg). The maximal trunk muscle strength was determined in 5 measuring days within 2-year follow-up in isokinetical and concentrical manner with an IsoMed 2000 dynamometer. According to the model of supercompensation, the sample was divided into an active group and an inactive group. The maximal relative turning moment for trunk muscle flexion (Flexabs) and extension (Extabs) was analyzed. The ratio between absolute flexional turning moment and body weight (Flexabs/kg = Flexnorm) as well as the ratio between absolute extensional turning moment and body weight (Extabs/kg = Extnorm), and the ratio Flexabs/Extabs (RKquot) were calculated. The t-test was used for statistical comparison of the groups. The significance level was defined as P < .01. Results: This is the first longitudinal study with over 1150 isokinetic measurements of trunk muscle strength. Because of the high number of measurements multiple cohorts were formed (sex, active/inactive, 3 age groups). Absolutely and regarding the body weight, male probands reached higher values of trunk muscle strength than female. Female probands reached significant higher RKquot (P < .01) caused by less extension power. Intraindividually the RKquot appeared to be stable within the 2 years. No difference was shown between active and inactive probands concerning the RKquot. However, active probands had a better power-load ratio. Conclusion: RKquot rises with growing age in both, female and male. This could be explained through reduced mobility and a posterior rotational axis shift with growing age. Further research of combined trunk muscle strength diagnostics and motion analysis are necessary for explaining the reduction of trunk extension power functionally and structurally. Results of this study can be used in the future as control data for classification of different spinal pathologies.

A047: Anti-Axial Rotation Strength of Long-Segment Fixation With Cortical Bone Trajectory and Traditional Pedicle Screw: A Biomechanical Study

Zhenlei Liu¹, Fengzeng Jian¹, Hao Wu¹, Xingwen Wang¹, Wanru Duan¹, Kai Wang¹, Jian Guan¹, Yu Qian¹, and Zan Chen¹

¹Xuanwu Hospital, Capital Medical University, Beijing, China

Introduction: Several studies have reported that cortical bone trajectory (CBT) screw has similar, if not stronger, fixation strength compared with traditional pedicle screw (PS). However, anti-axial rotation strength of long-segment instrumentation with CBT screws has been of concern and has not been validated. Material and Methods: Ten human cadaveric lumbar vertebrae (L1-L5) underwent computed tomography (CT) scan to validate the integrity of the vertebrae. Based on the Hounsfield unit at pars interarticularis, these specimens were paired and divided into 2 groups, one of which were instrumented at L1 through L3 (3 segments fixation) with CBT screws and the other with PS. Bionix Servohydraulic Test System (MTS Systems Corporation) was used to compare both static anti-axial rotation strength and fatigue resistance of the 2 constructs. Static torque was measured at rotation rate of 0.3° per second. Maximum torque of each cycle (0.5 Hz) was recorded during the fatigue test. Maximum rotation of construct failure was examined. Results: Within ±10° range, rotation torque of both CBT and PS construct exhibit good linear relation with displacement, y(N m) = 2.02 × (degree) (R ² = .9993) for CBT and y(N m) = 3.57 × (degree) (R ² = 0.9988) for PS, respectively (P < .01). After ±5° for 300 cycles and then ±10° for 300 cycles, mean maximum torque of CBT construct decreased from 20.52 N m to 18.54 N m, and that of PS construct decreased from 31.74 N m to 26.01 N m (P < .01). The mean maximum rotation of construct failure for CBT was ±19.6°, compared with ±13.4° for PS construct (P < .01). Conclusion: In long-segment fixation, CBT construct has less anti-axial rotation strength than PS construct. However, CBT construct has better fatigue resistance and bigger range of rotation. Clinical outcomes should be followed to verify if CBT construct is superior to the rigid PS fixation.

A048: Morphological and Mechanical Evaluation of Newly Formed Bone After Spinal Fusion Treatment Using a Biomimetic Peptide Enhanced Bone Graft

Jacobus Arts¹, Scott Johnson², Jerome Connor², and Bert van Rietbergen³

¹Maastricht University Medical Center, Maastricht, Netherlands

²Cerapedics Inc, Westminster, CO, USA

³ Eindhoven, Netherlands

Introduction: The aim of spinal fusion is to mechanically stabilize the motion segment using a combination of instrumentation and targeted bone regeneration. Iliac crest bone graft (ICBG) remains the gold standard for fusion as it exhibits osteoconductive, osteoinductive, and osteogenic properties, but there are now numerous synthetic bone graft options available. It is expected that these substitute materials progress toward near-normal trabecular organization as fusion progresses; however, there is a lack of information regarding the morphological changes and development of subsequent mechanical stability during the progression to fusion. The goal of this study is to quantify the morphology and mechanical properties of bone that is formed in the intervertebral region after single-level spinal fusion treatment using a biomimetic peptide enhanced bone graft (P15L) in an animal model. Material and Methods: Mature female sheep (<1 year old) underwent 1-level interbody surgery (retroperitoneal approach; L2-L3) using a PEEK interbody spacer. Each animal cohort was divided into 2 groups: one received a cage filled with P15L and the other a cage filled with ICBG. Animals were sacrificed at 30 (n = 6), 90 (n = 12), and 180 (n = 12) days following surgery. Animal care complied with the Canadian Council on Animal Care (CCAC) guidelines for care and use of experimental animals. A micro-CT (computed tomography) (25 μm voxel size) was acquired at sacrifice. A total of 30 micro-CT scans were evaluated from the 3 time points. Morphological analysis was performed for the region of newly formed bone at the center of the interbody cage and parameters included metrics for the volume of new bone formed, trabecular properties, and connectivity. As the mid-transversal plane of the cage included metal markers that caused severe metal artefacts in the micro-CT scans, only the regions well above and below these markers could be analyzed. An analysis for mechanical stability was undertaken to assess the load-bearing capacity of the graft-vertebrae region using micro-finite element (FE) analysis. Micro-FE models were generated for a region of 4 mm in height above and below the marker-affected region and the stiffness of these bone regions was calculated and averaged. Statistical significance between treatment groups was assumed for P values <.05. Results: At D30 there was no significant bone formation in the ICBG group whereas in the P15L group significant bone formation was found. At D90 substantial amounts of bone were formed in both groups, but significantly more (72%) in the P15L group than in the ICBG group. At D90, both groups developed similar amounts of bone and differences were no longer significant. The calculated stiffness linearly increased over time for the P15L group, while for the ICBG group it started slower. Significant differences between the groups were found only for D90. Conclusion: Although both groups reached similar amounts of bone volume and mechanical stiffness, the P15L peptide enhanced bone graft group did so significantly faster than the gold standard (ICBG).

OP09: Surgical Complications 1

A049: A Comparison of 30-Day Readmission, Reoperation, and Morbidity in Patients That Underwent Anterior Cervical Spinal Surgery, Multi-Level ACDF, and Anterior Cervical Corpectomy and Fusion Approaches Compared Using 10 Years of ACS-NSQIP Data

Nickolas Mancini¹, Austen Katz¹, Teja Karukonda¹, Mark Cote¹, and Isaac Moss¹

¹University of Connecticut, Farmington, CT, USA

Introduction: Multi-level cervical pathology can often be addressed with either anterior cervical corpectomy and fusion (ACCF) or multilevel anterior discectomy and fusion (ML-ACDF). The American College of Surgeons National Surgical Quality Improvement Program (ACS-NSQIP) database was examined to determine 30-day outcomes from these procedures to help guide future surgical decision making. Material and Methods: Patients who underwent ACCF or ML-ACDF from 2006 to 2015 procedures were identified in the ACS-NSQIP database. Thirty-day outcomes including readmission, reoperation, and complication rates were compared between the 2 groups. Independently associated predictors of readmission, reoperation, and morbidity were analyzed using multivariate logistic regression. Bonferroni adjustment was utilized to correct for multiple comparisons. P-values less than the Bonferroni-adjusted α level were considered significant, P-values were between .05 and the Bonferroni-adjusted P-value were considered trends. Results: The search identified 15 600 total patients, 11 944 underwent ML-ACDF, and 3656 underwent ACCF. The ACCF group was significantly older (56.2 vs 54.9 years; P < .001), had poorer overall health (American Society of Anesthesiologists [ASA] class ≥3: P < .001), had less female patients (48.90% vs 51.80%; P = .002), and more African-American patients (14.86% vs 9.97%; P < .001) compared to the ML-ACDF group. The ACCF group had longer operating time (151.51 vs 143.20; P < .001), greater length of stay (LOS) (3.19 vs 1.98; P < .001), was performed less often as an outpatient (10.6% vs 21.3%; P < .001), and performed more often as an emergency (2.30% vs 0.80%; P < .001) when compared to ML-ACDF. The ACCF group had higher rates of diabetes (17.04% vs 15.31%; P = .012), ventilator dependency (0.27% vs 0.01%; P < .001), dialysis (0.88% vs 0.23%; P < .001), disseminated cancer (1.39% vs 0.10%; P < .001), steroid requirement (3.97% vs 3.08%; P = .009), unintentional 6-month weight loss (0.63% vs 0.23%; P < .001), preoperative blood transfusion (0.41% vs 0.07%; P < .001), sepsis/septic shock (P < .001), increased blood urea nitrogen (BUN) (15.97 vs 15.24; P < .001), increased creatinine (0.97 vs 0.90; P < .001), and decreased hematocrit (40.88 vs 41.53; P < .001) compared to the ML-ACDF group. ACCF predicted greater rates of complications in both univariate (5.99% vs 2.59%; P < .001) and multivariate (odds ratio [OR] = 1.675; P < .001) analyses, greater rates of reoperation in univariate analysis (3.34% vs 1.52%; P < .001), and trended in multivariate analysis (OR = 1.687; P = .005). Readmission rates were similar between the 2 groups. Increasing age and ASA class ≥3 predicted readmission (P < .001), while outpatient surgery trended. Sepsis/septic shock, LOS, and outpatient surgery predicted reoperation (P < .001), while congestive heart failure and ASA class ≥3 trended. Female gender trended toward reduced reoperation. Increasing age and ASA class ≥3 predicted morbidity (P < .001), and disseminated cancer, weight loss in the past 6 months, emergency surgery, and outpatient surgery trended toward increasing morbidity. Female gender and increasing hematocrit predicted reduced morbidity (P < .001). Conclusion: After controlling for patient factors, undergoing ACCF was an independent risk factor for higher 30-day morbidity (OR 1.675) and trended toward higher reoperation (OR 1.687) compared to ML-ACDF. There was no correlation between type of procedure and 30-day readmission. Older age, higher ASA class, greater length of stay, and specific comorbidities predicted higher complication rates, while female gender and increasing preoperative hematocrit were protective, regardless of procedure. These findings can help guide surgical treatment decision making in specific patient populations.

A050: Intraoperative Complications of Posterior Internal Fixation of Upper Cervical Spine

Vadim Manukovsky¹, Konstantin Tiulikov¹, Tahir Tamaev¹, Valery Serikov¹, and Irina Afanasyeva¹

¹Dzhanelidze Research Institute of Emergency Care, Saint-Petersburg, Russian Federation

Introduction: Over the past decade, operations of internal fixation of the upper cervical spine have moved from the category of “exclusive surgery” into usual practice. Nevertheless, the risk of intraoperative complications is quite high and is associated primarily with the manual skills of the surgeon, operating equipment, and surgical anatomy of the affected zone. The problem of prevention of iatrogenic complications is extremely important. The Aim of the Study: To develop a conception of the optimal target surgery, depending on the anatomical features of craniovertebral transition and upper cervical spine, character of damages, and the degree of destruction of bone structures. Taking into account anatomical features of the upper cervical spine will minimize intraoperative risks when dosage traction, reduction, and reliable fixation is needed. Methods and Materials: The analysis of 127 patients with injuries of upper cervical spine, which were treated at the Dzhanelidze Research Institute of Emergency Care and the Military Medical Academy from 2011 to 2017. In all cases we use various options of posterior instrumental fixation using DePuy implants. Spiral computed tomography (SCT) were performed to all patients with suspected cervical spine injury before and after the operative treatment. Selective angiography of brachiocephalic vessels and the circle of Willis was carried out according to indications to assess the collateral blood flow. In case of malposition of screws on the control spiral CT the control angiography was performed in postoperative period. In 80 (63%) cases only posterior fixation by J. Harms was used. Fixation was performed with 2, 4, and 6 screw structures in the C1-C1, C1-C2, C2-C3, C1-C3, C1-C2-C3 segments. Occipitospondylodesis was applied in 47 (37%) cases. Control CT of cervical spine was performed immediately after the operation and before the elimination of a patient from anesthesia. After receiving the results of SCT, we took a decision of further tactics: to eliminate patient from anesthesia or perform a selective angiography, if malposition of screws was detected. In case of detecting screws malposition and sings of arterial occlusion, estimation of blood flow and collateral circulation was performed to make a decision for further tactics: replacement of screws, changing fixation method, or eliminate a patient from anesthesia. Results: In total 569 screws were placed to 85 patients. Malposition of screws was noted in 9 cases, of 7 patients. In 7 cases screws passed through the transverse foramen, and in one case screw passed through spinal canal. In one case, after intraoperative SCT, fixation was converted to occipitospondylodesis. Venous bleeding from plexus of C1 roots occurred in 32 cases and was not threatening complication. Sustainable hemostasis achieved by hemostatic sponge and coagulation. There was no injury of main arteries and related bleeding. In one case damage of the dura mater with intraoperative liquorrhea occurred. There was no complications of occipitospondylodesis (C0-C2, C0-C3 fixation), postoperative mortality and infectious complications. Discussion: Clinically significant intraoperative complication occurred in one case and manifested as Wallenberg’s syndrome, which regressed after a course of vasoactive therapy. To avoid incorrect placement of screws we offered modified Harms technique with translaminar entry point, which greatly simplified the manipulation technique and reduced risk of venous plexus damage and C1, C2 roots. Conclusion: The choice of fixation method is determined by the anatomical features, type of fracture, experience of the operating surgeon, and operating equipment. Knowledge of anatomical landmarks, execution of control spiral CT immediately after operation allows in due course identify complications of manipulation. The usage of DSA allows estimating the degree of compression of craniovertebral transition vessels and making the right tactical decision. Posterior internal screw fixation system is most desirable in craniocervical injury, and occipitospondylodesis is as an alternative. The most optimal conditions is the presence of an intraoperative CT combined with navigation tool.

A051: Low Bone Mineral Density Is a Major Risk Factor for Revision Surgery After Posterior Lumbar Fixation

Yu-Mi Ryang¹, Maximillain Loeffler², Anna Rienmüller³, Bernhard Meyer², and Jan S. Kirschke²

¹Helios Klinikum Berlin-Buch, Berlin, Germany

²Technical University Munich, Munich, Germany

³University Hospital Vienna, Vienna, Austria

Introduction: Reduced lumbar bone mineral density (BMD) seems to increase the risk of secondary screw loosening, nonfusion, and adjacent segment fractures after lumbar spinal fusion (LSF) necessitating revision surgery. Identifying possible risk factors might prevent surgical failure by using PMMA-augmented pedicle screw fixation. This study assessed the association of low BMD with revision surgery after LSF. Material and Methods: A retrospective observational study of a total of 1441 consecutive LSF was performed between 2010 and 2014. Eighty-one patients receiving LSF with a rigid spine system or with PMMA-augmentation were included with a minimum follow-up of 3 years. Attenuation measurements of preoperative MDCT (multidetector computed tomography) scans were performed for the retrospective assessment of lumbar volumetric BMD. Lumbar BMD values were calculated with asynchronous calibration by opportunistic QCT. Mean BMD values between patients with and without revision surgery and with or without PMMA-augmentation were compared. Results: Twenty-nine percent of patients without and 85% with PMMA-augmented LSF had prevalent osteoporosis (BMD < 80 mg/cm³). The rate of revision surgery was 15% in nonaugmented (7/48) and 12% (4/33) in augmented surgeries. Patients needing revision surgery in the nonaugmented group had a significantly lower mean BMD (73.0 ± 18.4 mg/cm³) than patients not needing reoperation (110.2 ± 37.8 mg/cm³; P = .005). The only significant risk factors for reoperation in the nonaugmented group were female sex and decreased BMD (adjusted odds ratio [OR] = 62.391 for female sex, P = .025; OR = 0.951 per 1 mg/cm³ increase in BMD, P = .031). The BMD threshold predicting reoperation after nonaugmented surgery was 83.7 mg/cm³. Conclusion: Despite a much lower mean BMD, PMMA-augmented showed no higher revision surgery rate than nonaugmented surgeries in our study. Since BMD was found to be an independent risk factor for revision surgery, it should be evaluated by opportunistic QCT before LSF, so augmentation is considered, if the BMD is in the osteoporotic range (<80 mg/cm³).

A052: Risk Factors Associated With Perioperative Complications After Lumbar Spine Surgery in Geriatric Patients

Oscar Bravo¹, Rodrigo Donoso¹, Manuel Valencia¹, Felipe Novoa¹, Juan Jose Zamorano¹, Ratko Yurac¹, Carlos Thibaut¹, Bernardo Merello¹, Rodrigo Varela¹, Alvaro Silva¹, and Bartolomé Marré¹

¹Clinica Alemana of Santiago, Santiago, Chile

Introduction: Perioperative complications can compromise outcomes, affecting patients’ quality of life, particularly in elderly patients. This ever-increasing population of patients has special needs due to physiological changes associated with aging. They are at risk of presenting adverse events such as infection, delirium, and excessive bleeding, all of which can contribute to functional decline, need for continuous assistance, and even death. Identifying risk factors for perioperative complications in these patients can help us to improve prevention strategies. We assessed perioperative complications in elderly patients (>65 years old) undergoing lumbar spine surgery at our institution and determined possible risk factors associated to them. Material and Methods: We identified elderly patients (>65 years old) who underwent lumbar spinal decompression and/or fusion procedures at our institution from January 2015 to June 2018. A retrospective review of their medical records and operating room notes was carried out to obtain demographic data, as well as intra- and postoperative complications (up to 30 days after discharge). Descriptive analysis together with logistic modelling using SPSS 22 software was performed to determine the risk of presenting a perioperative complication. Results: Seventy-three patients fulfilled our inclusion criteria, mean age 72.3 ± 5.5 years, 40 females. The most frequent indication for surgery was spinal stenosis (95.9%). The median Charlson Comorbidity Index was 3 (2-11) and the most frequent comorbidities were high blood pressure (60.3%), cancer (19.2%), type 2 diabetes (16.4%), and osteoporosis (5.5%). We also assessed the spine surgical invasiveness index obtaining a median of 6 points (2-22); most of the patients had open surgery, while only 13.7% had minimally invasive surgery. The mean number of decompressed levels was 1.23, while the mean number of fused levels was 1.34, with an overall median operating time of 190 minutes (55-1090) and median intraoperative bleeding of 300 cc (5-4500). Regarding complications, we identified a total of 52 perioperative complications in 26 patients (10 intraoperative complications in 8 patients and 42 postoperative complications in 24 patients, 6 of the patients presenting intraoperative complications also presented postoperative complications). The most frequent intraoperative complication was dural tears (6 cases), while wound-related complications were the most frequent postoperative one (13 cases). The mean hospital stay was 5.3 days (1-23). Using multivariate analysis, the following risk factors for presenting any type of perioperative complication were identified: having and intraoperative complication (odds ratio [OR] = 8.81, P = .012), Charlson Comorbidity index (OR = 1.54, P = .018), spine surgical invasiveness index (OR = 1.26, P = .014), number of decompressed levels (OR = 2.82, P = .02), number of instrumented levels (OR = 1.65, P = .047), and hospital stay (OR = 1.59, P = .006). Conclusion: We identified demographic and surgical factors that are significantly associated with the occurrence of perioperative complications in elderly patients undergoing lumbar spine surgery, particularly an 8-fold increase of this risk after presenting an intraoperative complication.

OP10: Trauma-Thoracolumbar 1

A053: Influence of Endplate Size and Implant Positioning of Vertebral Body Replacements on Biomechanics and Patients’ Outcome

Michael Kreinest¹, Sabine Kelka², Paul A. Grützner¹, Philipp Kobbe³, Miguel Pishnamaz³, and Sven Y. Vetter¹

¹Klinik für Unfallchirurgie und Orthopädie, Ludwigshafen, Germany

²Klinik für Allgemein-, Viszeral- und Gefäßchirurgie, Hamburg, Germany

³Klinik für Unfall- und Wiederherstellungschirurgie, Aachen, Germany

Introduction: Dorsoventral stabilization combining an internal fixator with an additional anterior vertebral body replacement (VBR) is frequently used in patients suffering from destructed vertebral bodies. To implant the VBR, a corpectomy in part or in total as well as a discectomy is performed. Thus, the VBR is in direct contact to the endplates of the adjacent vertebral bodies. Subsidence of VBR implants occurs mainly in the first year after implantation, and is the major postoperative complication of implanted VBRs. However, the influence of the VBR implant’s endplate size and of the VBR implant’s positioning in daily patient care is still unknown. Thus, the aim of the current study is (1) analyzing the choice of the endplate size and the positioning of VBR implants in daily patient care and (2) analyzing the influence of the VBR implants’ endplate size and the VBR implants’ positioning on clinical and radiological outcome. Material and Methods: All patients who received a VBR implantation between September 2009 and December 2010 were included in the study. Patients’ outcome was analyzed 3 years after VBR implantation using the Visual Analogue Scale (VAS) spine score and an additional questionnaire. A safe zone on the vertebral body’s endplate was defined and implant positioning within this zone as well as overall endplate covering were evaluated by computed tomography. VBR implants’ subsidence was evaluated by X-ray in standing position. Furthermore, biomechanical compression tests (BZ1-MM14450.ZV04, XForce Type HP 2.5 kN, TestXpert 2, Zwick GmbH 6 Co KG, Ulm, Germany) on 22 lumbar vertebral bodies with 2 different VBR endplate sizes have been performed to determine the maximum compression force F _max. Data management and statistical analysis were carried out using SAS 9.2 software (SAS Institute Inc, Cary, NC). Results: The follow-up was successful for n = 47 patients. The mean coverage of the vertebral body’s superior and inferior endplate by the VBR endplate was 27.8% and 30.8%, respectively. Based on the scientific literature, a safe zone was defined. The mean overlap of the safe zone by the VBR endplate was 49.8% at the inferior endplate and 40.6% at the superior endplate. The mean VBR subsidence was 1.1 ± 1.2 mm. The VBR’s subsidence did not have any effect on VAS spine score and on patients’ satisfaction with their outcome. Further analysis showed that the VBR’s subsidence seems to be increased if overlap of VBR’s endplate with the safe zone is reduced. In the compression tests, no significant difference (P = .468) was found between the median F _max using the 2 different endplate sizes. Conclusion: In daily patient care, the coverage of the vertebral body’s endplate by the VBR’s endplate as well as the positioning of the implant in the safe zone does not entirely comply with the recommendations given by the literature. But the recommended amount of endplate covering seems to have neither influence on the subsidence of the VBR implant nor on the patients’ long-term outcome in daily patient care. On the other hand, correct positioning of the VBR implant may influence the amount of the implant’s subsidence.

A054: Robotic Versus Open Fixation for Traumatic Spine Injuries: A Case-Control Study

Alexandra Satanovsky¹, Yehiel Gellman¹, Hananel Shear Yeshuv¹, Leon Kaplan¹, and Josh E Schroder¹

¹Hadassah Medical Center, Jerusalem, Israel

Introduction: Unstable traumatic fractures of the thoracolumbar spine require surgical stabilization. The necessity to restore stability and sagittal balance after trauma is crucial for early patient mobilization and long-term spinal function. The traditional fixation method is open fixation allowing reduction and stabilization of the broken segments (OF). In recent years, several studies have indicated percutaneous fixation (PF) for traumatic thoracolumbar fractures is advantageous in aspects of infection rate, blood loss, and hospitalization time. The challenge of the PF is that it is radiation heavy and requires a steep learning curve. Previous work showed that using robotic-assisted pedicle screws is safe and is low on radiation. In this study, we compare our results for PF in open versus percutaneous fixation for traumatic thoracolumbar fractures in a single trauma center. Material and Methods: All patients with traumatic vertebral fractures, who underwent fixation in a level 1 trauma center between 2006 and 2018 were enrolled. Retrospective data of demographics (age, sex, comorbidities), mechanism of injury, AO classification, level of injury, level and method of fixation, laboratory results (hemoglobin and hematocrit before and after surgery), requirement of blood products, length of hospital stay, discharge destination, change in Cobb angle, and complications were collected. Results: During the study period, 88 patients who underwent fixation were matched 44 for each group. Demographic data (age, sex, mechanism of injury) were similar (P > .05); 29/44 in each group were male, mean age 34 in OF and 39 in PF, with leading mechanisms of injury of MVA (motor vehicle accident) and fall from height. More levels of fixation were recorded in OF (4.57 levels and 3.7 in OF and PF accordingly, P < .01). Operative parameters were significantly better in PF group in theatre time, anesthesia time, and requirement of blood products (P < .0001). No difference was detected in pre- and postoperative hemoglobin levels. Postoperative parameters were also favorable to percutaneous group in hospital stay of 9.7 days in PF and 18.27 in OF. Normal neurological status was recorded in 40/44 patients in CF and only 27/44 in OF. Requirement of blood products, surgery, and anesthesia time and hospitalization duration was significantly better in PF group. More complications were recorded in OF group (PE, wound infection, pneumonia, adverse reaction to fresh frozen plasma). Change in Cobb angle was also higher in PF group (12.97° vs 6.76° in OF). Conclusion: We have found that robotic-assisted percutaneous fixation for traumatic thoracolumbar fractures is superior to open fixation in surgical parameters, hospital stay, and surgical results. In our experience, PSF with robotic-assisted percutaneous fixation is the preferred surgical method for treatment of unstable traumatic thoracolumbar fractures.

A055: Thoracoscopic Correction of Posttraumatic Kyphosis With an Expandable Cage: Radiologic and Patient-Reported Outcomes

Arjen Smits¹, Jaap Deunk¹, Fred Bakker¹, and Frank Bloemers¹

¹VU University Medical Center, Amsterdam, Netherlands

Introduction: Undesirable late consequences of traumatic thoracolumbar fractures are spinal deformity and pain. This is mostly due to loss of vertebral body height after insufficient conservative treatment of unstable fractures. Correction of a posttraumatic deformity is possible through a minimally invasive thoracoscopic approach, which prevents the morbidity of an open technique and improves visual exposure of the anterior column at the same time. The value of this approach has been proven in the treatment of unstable fresh spinal fractures. The outcomes of anterior correction with an expandable cage through a minimal invasive thoracoscopic approach for the correction of posttraumatic kyphosis (PTK) however have not yet been reported. A retrospective cohort study was performed to determine the long-term radiological and functional outcomes of thoracoscopic posttraumatic kyphosis correction (PTKC) using an expandable cage in patients without neurologic injury. Materials and methods: All patients that underwent thoracoscopic posttraumatic kyphosis correction with an expandable cage between 2007 and 2017 in one academic trauma center were further analyzed. Patient data were collected from the hospital information system and all radiologic material was reassessed for evolution of kyphosis and intervertebral body height (IBH). Quality of life (QOL) and functional outcome scores were determined using EQ5D and ODI (Oswestry Disability Index) questionnaires. Additionally satisfaction and subjective symptom improvement were determined. Results: Fourteen patients were treated for symptomatic PTK through a combined posterior and thoracoscopic approach. Nine of these patients received initial conservative fracture treatment and 5 underwent initial posterior fracture fixation which did not remain stable. All patients with PTK presented with pain and without neurologic injury. The traumatic fractures had been located between T6 and L2; the mean time between injury and PTKC was 15.4 months. The mean preoperative Cobb angle (CA) was 26.1° and improved with 10.6° immediately after PTKC. During the first follow-up 4.8° kyphosis correction was lost, but CAs remained stable at longer follow-up. Nine patients reported additional outcomes after a mean of 85 months after surgery. The majority reported improvement of symptoms, was satisfied, and willing to undergo the procedure again. The mean EQ-5D index score was 0.71 and the mean ODI score was 22.3. Conclusion: The radiologic and clinical results of minimal invasive thoracoscopic PTKC using an expandable cage are satisfactory. The majority of patients is satisfied after treatment. Functional and QOL scores are fairly good. The thoracoscopic approach minimizes surgical morbidity, does not lead to serious complications, and provides a good option for PTKC.

A056: Utility of Initial Weight-Bearing X-Rays for Predicting the Evolution of AOSpine Type A Thoracolumbar Spine Fractures

Oscar Bravo¹, Max Grez², Jaime Cancino³, Guillermo Izquierdo³, Felipe Novoa¹, and Manuel Valencia¹

¹Clinica Alemana of Santiago, Santiago, Chile

²Hospital Naval de Viña del Mar, Valparaíso, Chile

³Mutual de Seguridad de Santiago, Santiago, Chile

Introduction: In absence of neurological deficit or instability, thoracolumbar fracture surgery of A type AO (ATLF) depends on regional radiologic parameters of the compromised segment. So far, there is no information that describe their behavior through weight bearing and progress until it is consolidated. Our objective is to compare regional parameters in acute and medium-term imaging studies with or without weight bearing in A type thoracolumbar fractures orthopedically treated. Materials and Methods: Retrospective cohort with ATLF diagnosis between January 2015 and April 2018, in a Chilean Trauma Centre, orthopedically treated, imaging studies were evaluated with and without weight bearing (WB+/WB−), taken in acute and medium-term follow-up (3-6 months), measuring regional kyphosis (RK), anterior collapsing (AC), and vertebral wedging (VW). Student’s t test for paired samples was used for statistical analysis, variables were related through linear correlation tests. Results: A total of 44 patients with a mean age of 48 years (±14.56). A1 type fracture (n = 28) and L1 level (n = 22) were the most frequent. Differences in RK, AC, and VW in initial study WB− versus WB+ were as follows: RK 9° versus 13.09° (P .015), AC 19.39% versus 27.18% (P .005), and VW 11.66° versus 15.64° (P .0001). Comparing initial study WB+ versus 3 to 6 months, a linear correlation was seen (R ² = 0.774, P .001): RK 13.09° versus 14.25° (P .518), AC 27.18% versus 10.41% (P .278), and VW 15.64° versus 16.66° (P .321). Conclusions: Weight-bearing studies increase initial radiological parameters generating a minor difference with definitive radiological studies when compared with without weight bearing, in addition to having a positive linear correlation with the definitive study. This allows the C-C studies to improve the accuracy of the final result of the orthopedic treatment and maybe diminished the cost of control in this group of patients with this type of fracture.

OP11: Degenerative-Cervical 2

A057: Comparison of Inpatient Versus Outpatient Single-Level Anterior Cervical Discectomy and Fusion for Cervical Radiculopathy: Utilization, Safety, and Economics From a National Perspective

Christopher Witiw¹, Fabrice Smieliauskas², John O’Toole¹, Richard Fessler¹, and Vincent Traynelis¹

¹Rush University Medical Center, Chicago, IL, USA

²The University of Chicago, Chicago, IL, USA

Introduction: An anterior cervical discectomy and fusion (ACDF) is commonly performed to treat cervical radiculopathy. Recently, evidence has mounted to support the safety of single-level ACDF in the outpatient setting. Moreover, data are beginning to emerge which suggest that choosing the outpatient setting may represent a potential for cost-savings. Current data are from single-center studies or US databases at the state level; however, there is a paucity of nationally representative data. The aim of this study is to evaluate the utilization, safety, and economics of inpatient versus outpatient single-level ACDF from a national perspective. Material and Methods: Adults undergoing single-level ACDF between July 2003 and December 2014 for cervical radiculopathy were included from the MarketScan Commercial Claims and Encounters and Medicare Supplemental databases. This national commercial health insurance claims database is representative of the largest segment of US health care users. Outcomes of interest consisted of utilization, adverse events, and total payments to the health provider for the index procedure and over a 30-day horizon. Propensity score matching balanced the groups on observed baseline covariates. Results: A total of 45 963 patients were included. The inpatient cohort consisted of 24 282 subjects and the outpatient cohort consisted of 21 681 patients. The proportion of cases classified as outpatient increased from 33.5% in 2003 to 66.6% by 2014. Thirty-day readmission rates were 3.22% in the outpatient cohort, which was significantly higher than the rate of 1.49% in the inpatient cohort (P < .001). However, overall adverse events were not significantly higher in the outpatient cohort. Over a 30-day horizon the mean payment made for an inpatient procedure was $29 459 ± 17 780, compared with $25 321 ± 17 522 for an outpatient procedure. After propensity score matching, those in the outpatient cohort incurred payments of $4025 (95% confidence interval = 3580-4470; P < .001) less over a 30-day horizon than those in the inpatient cohort. Conclusion: This large-scale, nationally representative data suggests that outpatient single-level ACDF for cervical radiculopathy may represent a potential source of health care saving without increased risk when compared with inpatient ACDF for appropriately selected patients.

A058: The Influence of Cervical Spondylolisthesis on Clinical Presentation and Surgical Outcome in Patients With DCM: Analysis of a Multicenter Global Cohort of 458 Patients

Aria Nouri^1,2, SO Kato³, Jetan Badhiwala⁴, Michael Robinson¹, Mejia Munne Juan^1,2, George Yang^1,2, William Jeong^1,2, Rani Nasser^1,2, David Gimbel^1,2, Joseph Cheng^1,2, and Michael Fehlings⁴

¹University of Cincinnati, Cincinnati, OH, USA

²Yale University, New Haven, CT, USA

³University of Tokyo, Tokyo, Japan

⁴University of Toronto, Toronto, Ontario, Canada

Introduction: Cervical spondylolisthesis (CS) is common among patients with degenerative cervical myelopathy (DCM). However, its impact on clinical presentation and surgical outcome have not been well described. Herein, we compare patients with and without CS on MRI (magnetic resonance imaging) undergoing surgical treatment for DCM. Material and Methods: A total of 458 MRIs from the AOSpine North America and International Studies were reviewed. CS was identified using MRIs. Patients with DCM were divided into 2 cohorts, those with CS and those without, and propensity matching was performed. Patient demographics, neurological and functional status at baseline and 2-year follow-up were compared. Results: Compared to non-spondylolisthesis patients (n = 404), CS patients (n = 54) were 8.8 years older (P < .0001), presented with worse baseline neurological and function status (mJOA [modified Japanese Orthopaedic Association], P = .008; Nurick, P = .008; SF-36-PCS, P = .01), more commonly presented with ligamentum flavum enlargement (81.5% vs 53.5%, P < .0001), were less commonly from Asia (P = .0002), and tended to have more compressed levels (P = .052) and lower prevalence of ossification of the posterior longitudinal ligament (P = .098). There was no difference in sagittal alignment (P = .94). Surgical approach varied between cohorts (P = .0002), with posterior approaches favored in CS (61.1% vs 37.4%). CS patients also had more operated levels (4.3 ± 1.4 vs 3.6 ± 1.2, P = .0002), and tended to undergo longer operations (196.6 ± 89.2 minuets vs 177.2 ± 75.6 minutes, P = .087). The mean improvement of neurological function was lower with CS (mJOA [1.5 ± 3.6 vs 2.8 ± 2.7, P = .003]; Nurick [−0.8 ± 1.4 vs −1.5 ± 1.5, P = .002]), and CS was an independent predictor of worse mJOA recovery ratio at 2 years (B = −0.190, P < .0001). After propensity matching, the mean improvement of neurological function was still lower in patients with CS (mJOA [1.5 ± 3.6 vs 3.2 ± 2.8, P < .01]; Nurick [−0.8 ± 1.4 vs −1.4 ± 1.6, P = .02]). Conclusion: CS patients are older and present with worse neurological and functional impairment. Furthermore, they receive surgery on more levels and more commonly from the posterior. CS may indicate a more advanced state of DCM pathology and is more likely to result in a suboptimal surgical outcome.

A059: Does Preoperative T1 Signal Changes on Magnetic Resonance Imaging Affect the Postoperative Functional Recovery in Patients of Cervical Myelopathy?

Ayush Sharma¹, Vijay Singh¹, Romit Aggrawal¹, Sumit Mathapati¹, Priyank Deepak¹, Shourab Sinha¹, Nilesh Mangale¹, Jeet Savle¹, and Ajay Jaiswal¹

¹Dr B R Ambedker Central Railway Hospital, Mumbai, India

Aim: Aim was to determine effects of preoperative signal changes of magnetic resonance imaging (MRI) on postoperative functional recovery in patients of cervical myelopathy. Material and Methods: Prospectively collected data of the patients who underwent surgery for cervical myelopathy with more than 1 year of follow-up were included in study. Modified Japanese Orthopaedic Association scores (mJOA), preoperative and follow-up X-rays, and MRI (2 Tesla) were used to determine the functional and radiological outcomes. Sixty-three patients who fulfilled the selection criteria were finally included in the study. Result: The mean age was 52 years. The mean duration of symptoms was 9.8 months. Thirty patients underwent surgery by anterior approach while posterior approach was done in 33 patients. Mean preoperative mJOA score was 10.18. Significant improvement was noted in mean mJOA scores at 3 months (13.41), 6 months (14.45), and at 1 year follow-up (15.50). Improvement in mJOA scores was found to be independent of the approach of surgery. While preoperative mJOA scores were comparable, 1 year follow-up mJOA scores were significantly low for the patients (14.6) with preoperative T1 cord change when compared to the group with only T2 signal changes on preoperative MRI (15.89). Conclusion: The presence of preoperative T1 signal changes might indicate a less significant postoperative recovery in patients of cervical myelopathy.

A060: Patterns and Predictors of Return to Work After the Surgical Management of Cervical Spondylotic Myelopathy

Arun Paul¹, Rohit Amritanand¹, Justin Arockiaraj¹, Kenny David¹, and Venkatesh Krishnan¹

¹Christian Medical College, Vellore, India

Introduction: From a patients perspective, returning to work is arguably one of the most important outcomes following spinal surgery.¹ A number of studies have described return to work following lumbar surgery² and cervical discectomy.³ Cervical spondylotic myelopathy (CSM) is one of the leading causes of spinal cord dysfunction.⁴ Yet, to the best of our knowledge, there have been no reports in the English literature describing return to work among patients undergoing surgery for CSM. This study was conducted to determine the patterns as well as the predictors of return to work following surgical treatment of CSM. Methods: This observational study was conducted on patients who underwent surgical treatment for CSM (Nurick Grade >3) at a university hospital from January 2012 to December 2017. The study received clearance from the institutional review board. At follow-up, the patients were divided into 2 groups. Those who returned to work within 6 months of surgery were assigned to Group 1. Those who did not were assigned to Group 2. Their preoperative demographics and medical details were recorded. Functional status was assessed using Nurick grade and modified Japanese Orthopaedic Association score. Details of their preoperative and present occupational status was documented along with surgical details. Data from the 2 groups were compared using appropriate statistical tests. Results: A total of 34 patients were included in the study. At a mean follow-up of 32 months, 52.9% returned to work by 6 months (Group 1). Baseline characteristics such as preoperative pain, functional scores, and comorbidities were comparable between both the groups. The nature of work had a statistically significant association with return to work by 6 months (P = .005) with poorer rates seen in manual laborers. The service group and homemakers were more likely to return to work. Significantly better functional outcomes were seen in patients who returned to work at 6 months as calculated using Nurick grade (P = .000) and mJOA score improvement (P = .001). All patients (100%) who returned to work and 75% of patients who did not return to work were satisfied with the outcome of surgery (P = .039). Age, body mass index, duration of symptoms, preoperative absenteeism from work, smoking, diabetes mellitus, number of levels operated, surgical approach, postoperative complications, and cord signal changes in magnetic resonance imaging did not have any significant association with the probability of postoperative return to work. Conclusion: In spite of functional improvement, CSM is associated with poor return to work by 6 months after surgery with manual laborers being the most vulnerable group. Though the functional outcome is better in patients who return to work by 6 months, majority of the patients even in the non–return to work category are satisfied with the outcome of surgery. The results of this study will help surgeons modulate patient expectations as well as provide a platform for counselling them regarding their return to work.

OP12: Tumor 1

A061: Outcome After Surgery for Symptomatic Spinal Metastases: Results From a Prospective Database of 2621 Surgical Patients

David Choi¹ and Global Spine Tumour Study Group²

¹NHNN, London, UK

²United Kingdom

Introduction: The incidence of spinal metastases is increasing due to better treatment of the primary tumor, longer survival, better radiological imaging, and advances in spine surgery. We present the results of surgery for symptomatic spinal metastases, and conclusions drawn from the largest prospective series of surgical patients to date. Material and Methods: The Global Spine Tumour Study Group prospectively collected patient data, with appropriate ethical and institutional review board approvals, for patients who presented to 23 specialist spine surgery units in 12 countries across 3 continents. Data included admission data, neurological and functional status, surgical data, and postoperative outcomes (EuroQol EQ-5D quality of life, Karnofsky performance status, Frankel neurological grade, sphincter function, mobility, complications, and survival). Results: A total of 2621 patients were assessed, of whom 1681 patients had “completed follow-up” (defined as death or documented 2-year survival). Mean follow-up period was 355 days from date of surgery. Median Tokuhashi score was 9. Rapid improvements in pain were seen by time of discharge from hospital and sustained for at least 24 months. Mean quality of life improved after surgery in all age groups, including the over-80s. Strong predictors of survival were the tumor type, number of spinal levels affected, number of visceral metastases, and Karnofsky performance score (KPS). Low preoperative KPS was the strongest predictor of poor survival (less than 3 months) when surgery might be considered inappropriate. Factors that predicted good postoperative quality of life were KPS and preoperative EQ-5D. There was an overall improvement in survival in patients who were operated in the more recent 5-year time period, 2011 to 2016, compared to previous 5-year periods. Conclusion: Surgery for symptomatic spinal metastases is effective in improving pain and quality of life, in appropriately selected patients. Performance status is an important criterion for patient selection and postoperative quality of life, whereas indicators of disease load (type and number of metastases) are more associated with survival.

A062: Accuracy of Freehand Versus Navigated Thoracolumbar Pedicle Screw Placement in Patients With Metastatic Tumors of the Spine

Rafael De la Garza Ramos¹, Yaroslav Gelfand¹, Murray Echt¹, Michael Longo¹, Merrit Kinon¹, Jonathan Nakhla², and Reza Yassari¹

¹Montefiore Medical Center, New York, NY, USA

²Brown University, Providence, RI, USA

Introduction: Patients with metastatic spinal disease may require internal fixation for treatment of clinical instability. The purpose of this study is to compare the accuracy and breech rates of freehand (FH) versus navigated (NV) pedicle screws in the thoracic and lumbar spine in patients with metastatic spinal tumors. Methods: A retrospective review of our institutional neurosurgical spine database was conducted to identify adult patients who underwent pedicle screw fixation in the thoracic or lumbar spine for metastatic spinal tumors between 2012 and 2018. Only patients with postoperative computed tomography (CT) scans were included. Breeches were assessed based on the Gertzbein and Robbins classification and only screws placed >4 mm outside of the pedicle wall (lateral or medial) were considered breeched (defined as a potentially “unsafe” zone). Screws placed via the FH technique were compared to screws placed using 3D navigation after acquisition of intraoperative CT scans. Results: A total of 62 patients received 547 pedicle screws (average 8 per patient)—34 patients received 298 pedicle screws in the FH group and 28 patients received 249 screws in the NV group. Average age for all patients was 58.7 years and 58.1% were male. There were 40/547 breeches, corresponding to a breech and accuracy rate of 7.3% and 93.7%, respectively. Breeches were lateral in 34/40 cases (85%) and medial in 6/40 (15%). The breech rate was 9.7% in the FH group and 4.4% in the NV group (χ² test, P = .017); this corresponded to an accuracy rate of 90.3% and 95.6%, respectively. Only one patient from the overall cohort (in the FH group) required revision surgery due to a medial breech abutting the spinal cord (1.6% of all patients; 2.9% of FH patients); no patient suffered organ, vessel, or neurological injury from screw breeches. Conclusion: Navigated pedicle screw placement in patients with metastatic spinal tumors have a significantly higher radiographic accuracy compared to the FH technique. However, the revision surgery rate is extremely low and no patient suffered organ or neurovascular injury in either group. Nonetheless, navigation also offers the advantage of real-time localization of spinal tumors and aids in targeting and resection of these lesions.

A063: Early Results, Complication Rates, and Revision Rates Following Carbon Fiber–Reinforced Pedicle Screws for Stabilization of the Thoracolumbar Spine

Anna Rienmüller¹, Judith Bernett², Bernhard Meyer², and Yu-Mi Ryang²

¹Medical University Vienna, Vienna, Austria

²Technical University Munich, Munich, Germany

Introduction: Surgical treatment of spinal tumors and metastasis usually comprises decompression and posterior fixation of the spine by pedicle screw fixation with or without vertebrectomy and vertebral body replacement. Standard pedicle screw systems are made of titanium alloys, which lead to artifacts in postoperative and follow-up imaging as well as difficulties in postoperative radiation planning and execution. A few years ago, carbon fiber–reinforced PEEK (CFRP) pedicle screw systems were developed for spinal fixation with the purpose of reducing imaging artifacts and radiation absorption. We aim to report our experience of surgical stabilization using CFR PEEK implants for internal fixation of the spine for spinal metastasis and primary spinal tumors. Material and Methods: We performed a retrospective analysis of our prospectively collected database including all patients who underwent spinal stabilization using a CFR pedicle screw-based system. Patient demographics, diagnosis, intraoperative and postoperative complications, implant-related complications, as well as postoperative additional treatment, outcome, and revision surgeries were evaluated. Results: Between August 2015 and March 2018, a total of 100 patients (59 men, 41 women) with a mean age of 64 (range 20-93) received spinal fixation of the thoracolumbar spine using a CFR internal fixator, implanting a total number of 738 CFR pedicle screws. Mean BMI (body mass index) was 26.3 ± 4.9, mean duration of surgery was 203 minutes (78-384), and an average of 4.8 levels (range 2-8) was stabilized. In 92% navigation was used for implantation. In 20 patients additional cement-augmentation of pedicle screws was performed and 22 patients received additional vertebral body replacement. Intraoperatively we found dural tear in 4%, 2 screw breakages during pedicle screw insertion, and 22/738 screws were revised intraoperatively due to malpositioning without further sequelae. Postoperative surgical complications needing revision surgery were documented in 13 patients. No revision surgery was performed due to implant malposition or failure until last follow-up. Sixty-four percent of patients received postoperative radiation and/or chemotherapy. Conclusion: CFRP pedicle screws offer a valuable alternative to standard titanium alloy pedicle screw systems. Complication and revision rates are comparable to standard internal fixators. Therefore, CFRP pedicle screw fixators seem to be a valid and safe option for surgical stabilization of the thoracolumbar spine for tumor surgery. Further prospective randomized controlled trials are necessary to evaluate possible advantages of CFRP implants in relation to various pathologies and long-term outcomes.

A064: Asymptomatic Construct Failure After Metastatic Spine Tumor Surgery: A New Entity or a Continuum With Symptomatic Failure?

Naresh Kumar¹, Ravish Patel¹, Barry Tan¹, Aditya Singla¹, Lay Wai Khin¹, Nivetha Ravikumar¹, Helena Milavec¹, Aravind Kumar², Hwee Weng Dennis Hey¹, Leok-Lim Lau¹, Joseph Thambiah¹, and Gabriel Liu¹

¹National University Hospital, Singapore, Singapore

²Ng Teng Fong General Hospital, Singapore, Singapore

Introduction: Reported incidence of implant and/or construct failures after instrumentation in metastatic spine tumor surgery (MSTS) is low (1.9% to 16%). These incidence rates are based on clinical presentations or revisions required for symptomatic failures. The phenomenon of asymptomatic construct failure (AsCF) after MSTS has not been described. Hence, we aim to study the incidence, onset, underlying mechanism, natural history, and the associated factors leading to AsCF after MSTS. Methods: Ours is a retrospective review of the prospectively collected data of 288 patients who underwent MSTS at a single tertiary care institute from 2005 to 2015. Data collected were patient demographics, oncological, and operative and postoperative variables. Operative details were the number of spinal levels instrumented and/or decompressed and types of fixation used. Radiological evidence of construct failures was identified using serial radiographs. Patients with AsCF were analyzed for risk factors and survival duration. Competing risk regression analyses were done where AsCF was the event of interest, with symptomatic failure (SF) and death as competing events. Kaplan-Meier survival curves were obtained for patients with AsCF, SF, and no failures. Results: AsCF was observed in 41/246 patients (16.7%). Average onset of AsCF after MSTS was 2 months (1-9 months). Early AsCF (<3months from index surgery) accounted for 80.5%, while late AsCF (>3months) were observed in 19.5%. Increasing age (P < .02) and primary breast (13/41 = 31.7%; P < .01) tumors were associated with higher rates of AsCF. Most common radiologically detectable AsCF mechanism was angular deformity (increase in kyphosis) in 29 patients, followed by screw ploughing and screw loosening in 15 patients each. There was a trend toward AsCF in patients with SINS ≥7, instrumentation across junctional regions, and construct length of 6 to 9 levels, although the associations were not significant. Median survival of AsCF patients was 20 months (3-95 months) in patients with early failure and 41 months (11-92 months) in patients with late failure. Average follow-up duration was 20 months. Conclusions: Majority of early AsCF did not require any intervention. Late failure was seen in patients who survived longer and maintained ambulation for a longer period. This may be due to the failure of fusion and/or late recurrence of tumors. AsCF is not necessarily an indication for aggressive or urgent intervention. However, we recommend frequent follow-up with periodic investigations to detect progressive construct failure. Increasing age and patients with primary breast tumor have a higher possibility of AsCF after MSTS.

OP13: Minimally Invasive Spine Surgery 1

A065: Decreasing Thickness and Remodeling of Ligamentum Flavum After Oblique Lumbar Interbody Fusion

Akaworn Mahatthanatrakul¹, Guang-Xun Lin², Hyeun-Sung Kim³, and Jin-Sung Kim²

¹Naresuan University Hospital, Phitsanulok, Thailand

²Seoul St. Mary’s Hospital, The Catholic University of Korea, Seoul, Republic of Korea

³Nanoori Suwon Hospital, Suwon, Republic of Korea

Introduction: Ligamentum flavum and annulus fibrosus are 2 main non-bony structure that could compress the spinal canal. Studies after spinal fusion found atrophy of these tissues. This study aimed to demonstrate the reduction of ligamentum flavum size after oblique lumbar interbody fusion (OLIF) using magnetic resonance imaging (MRI). Material and Methods: Fourteen patients who underwent OLIF without direct decompression were included. The self-reported measurements used were the Oswestry Disability Index (ODI) and Numerical Rating Scale (NRS) for back pain and leg pain. The MRIs were obtained at the preoperative period, immediate postoperative period, and during the follow-up period. Disc height (DH) were measured in standing plain radiograph. MRI measurements were spinal canal cross-sectional area (SCSA), spinal canal width, ligamentum flavum thickness (LFT), ligamentum flavum area (LFA), and foraminal area (FA). Results: Mean age of the patients was 68.1 ± 11.7 years. Nine patients were female. The operation was done at L2-3 for 2 levels, L3-4 for 8 levels, and L4-5 for 6 levels for a total of 16 levels. ODI improved from 50.8 ± 14.6% to 35.5 ± 19.4% (P = .009). NRS for back pain improved from 5.8 ± 2.3 to 2.1 ± 1.9 (P = .001). NRS for leg pain improved from 6.5 ± 1.7 to 2.1 ± 1.6 (P < .001). Mean times between postoperative MRI and follow-up MRI was 1.7 ± 1.0 years. Mean disc height increased from 7.6 ± 1.8 mm to 11.3 ± 1.6 mm at the immediate postoperative period but decreased to 9.8 ± 1.5 mm during the follow-up period (P < .001). SCSA increased from 99.5 ± 59.6 mm² to 135.6 ± 72.9 at postoperative MRI and 174.0 ± 69.1 mm² during follow-up (P < .001). LFT decreased from 4.2 ± 1.1 mm to 3.3 ± 0.8 mm during the postoperative period (20.6%) and further decreased to 3.1 ± 0.6 mm during the follow-up (6.6%; P < .001). LFA decreased from 99.5 ± 37.6 mm² to 89.9 ± 39.0mm² during the postoperative period (9.6%) and further decreased to 83.7 ± 32.7 mm² during follow-up (7.0%; P = .008). Foraminal area increased from 65.5 ± 21.9 mm² to 96.1 ± 19.6 mm² during the postoperative period (46.8%) and increased to 107.4 ± 25.1 mm² (11.8%) during follow-up (P < .001). Conclusion: OLIF could decompress the spinal canal and foraminal canal indirectly. Despite the diminishing disc height during the follow-up period, the spinal canal and foraminal canal were further increased in size, partly due to remodeling of the ligamentum flavum.

A066: Antepsoas Access, the Best of Lateral Access Without Muscle Damage

Alberto P. Contreras¹, Fernando Calderon Carrillo¹, Erik Hernandez Vasquez¹, and Diana Chavez Lizarraga¹

¹Centro del Cerebro y Columna Vertebral Hospital Mexico, Mexico City, Mexico

Anterolateral accesses to the lumbar spine have a history dating back to 1960 with transperitoneal access, but it showed more progress with the integration of sedation, lighting, and optics systems giving description to the minimally invasive retroperitoneal access to L2 to L5 with microscope to the implementation of static cages and monitoring for transpsoas access in 1998. Description of antepsoas access and thus avoiding manipulating the lumbosacral plexus inside the muscle has been cataloged as the greatest technical advance for indirect decompression and stabilization of degenerative lumbar spine and degenerative column plus deformity. From January 2017 to June 2018 at Hospital Angeles Mexico of the Angeles Health Services Group, lumbar spine surgery was performed with antepsoas access to scheduled patients. They were selected from medical appointments, performing radiographic studies that included panoramic vertebral column for measurement of sagittal balance, dynamic studies, simple lumbar spine resonance, and bone densitometry from spine and hip. Inclusion criteria: Diagnosis of degenerative disease of symptomatic lumbar spine that did not improve with conservative treatment based on pain killers, weight control, and column hygiene indications including physical therapy in the previous 6 months. Patients also were assessed by Psychology avoiding additional benefits like laboral inability, the preoperative valuation was mandatory in those older than 45 years and grading ASA I-II/IV. The Start scale for lumbar pain and Oswestry scale for functionality were used. Exclusion criteria: L5S1 segment disease, active smoking, BMI (body mass index) greater than 30, and psychiatric and/or psychological disorder. Informed consent was signed for both the procedure and to obtain photographic and video material during the surgery. In total there were 54 patients who were asked to perform the procedure. The following demographic data were revealed: the female gender predominates in 35 patients (64%) and male gender in 19 patients (36%). Active working patients with no distinction of gender. Always performed left side access by the layout of the aorto-psoas corridor and in the image studies always look for fatty tissue between both anatomical structures. The access column level was from L2 to L5 and the levels where L5S1 was treated in the same position either by anterolateral access or by TLIF (transforaminal lumbar interbody fusion) were excluded. Time of the surgery was 25 minutes per level (indirect decompression that includes access, discoidectomy, and intersomatic cage placement). The posterior instrumentation that was performed in the same lateral position of the patient, guided by O-arm, increased the time in 30 minutes, no matter whether it was 1 or 3 levels. In total there were 84 levels in 42 patients (2 levels), 27 levels in 9 patients (3 levels), 8 levels in 2 patients (4 levels), and 1 level in 1 patient. The most affected level was L4L5, followed by L3L4. The position of the patient is lateral decubitus without flexion of the table, offering no traction, protection of bone prominences, and slight flexion of the left hip. The access is made under radiological planning in the year 2017 with 2D fluoroscopy, placing the incision 2 fingers cross-anterior to the vertebral body at the level of the intervertebral disc. In multilevel cases, there was variation of the incision. Blunt dissection of the muscular planes is carried out until reaching the peritoneal fat detecting the abdominal wall, psoas and aorta under direct vision, and the psoas muscle is later displaced allowing access to the intervertebral disc by placing a 30 mm diameter separator without bone fixation or neurophysiological monitoring. Discoidectomy was performed with contralateral fibrous annulus opening, using either the ALIF (anterior lumbar interbody fusion) cage or lateral cage, bone graft, or demineralized bone matrix. Postoperative pain control was with acetaminophen at a dose of 3 g per day with parecoxib 80 mg daily; antibiotic prophylaxis was with vancomycin single dose. Moving in and out of bed was recommended at 4 hours after the surgery, and they left for home with indication of dorsolumbar and lower limb stretching exercises. This kind of access to the spine without muscle damage has allowed rapid functional recovery, achieving correction of deformity cases associated with degeneration, offering improvement with indirect decompression, stabilization of the lumbar spine with shorter hospital stay, and early integration into their activities. This technique has proved to be a reasonable alternative to avoid the use of transoperative monitoring because we avoid navigating through the psoas muscle and thereby avoid complications related to the compression of the lumbosacral plexus, and no complications related to the peritoneal cavity were observed.

A067: Percutaneous Endoscopic Lumbar Discectomy for L5S1 Disc Herniation. Does Interlaminar Approach Give Better Results Compared to Transforaminal Approach?

Asrafi Rizki Gatam¹, Luthfi Gatam¹, and Harmantya Mahadhipta¹

¹Fatmawati General Hospital, Jakarta Selatan, Indonesia

Introduction: Symptomatic lumbar disc herniation often needs surgical decompression. The management itself has been developed through generation with microdiscectomy as the “gold standard” for surgical decompression. Transforaminal percutaneous endoscopic lumbar discectomy (PELD) under local anesthesia offers less invasive surgery with many advantages, but the transforaminal approach to the disc space is very challenging due to high lying iliac crest and narrow intervertebral foramen. The interlaminar approach to the disc space is another alternative, but the need of general anesthesia and mobilization of the nerve structure is one of the disadvantages of the procedure. The aim of this study is to share the outcome, complication, surgical technique, and obstacle of both approaches. Method: This is a cohort prospective studies. The inclusion criteria were patients with true herniated nucleus pulposus without any degenerative process who failed for conservative treatment. Visual analog scale for sciatica was compared pre- and postoperatively, and the patients’ satisfaction results were measured using modified MacNab’s criteria. All patients in the transforaminal group underwent transforaminal PELD under local sedation anesthesia, and those in the interlaminar group underwent interlaminar PELD under general anesthesia. Result: The mean age was 31.9 years (range 14-51), consisting of 27 males (45%) and 33 females (55%). All patients had L5S1 disc herniation; 27(45%) patients underwent transforaminal approach under local anesthesia and the remaining 33 (55%) patients underwent interlaminar approach under general anesthesia. The mean sciatica VAS decreased from 5.7 (range 4-7) to 2.5 directly after surgery in the transforaminal group and from 5.8 (range 4-8) to 1.8 in the interlaminar group. The straight leg raise test was still below 35° in 8 patients in the transforaminal group and negative in all in the interlaminar group. There were 5 patients on the transforaminal group that needed to be reoperated due to persistent symptoms. In the interlaminar group there were 2 patient who had cauda equina syndrome. On follow-up the transforaminal group had more back pain and dysesthesia on the L5 root and still not resolve until 1 year, in which the interlaminar group showed slight dysesthesia on S1 traversing root but resolve within 1.5 months and also resolution of cauda equina symptoms after 2 months. The overall satisfactory result in the transforaminal group was 81%, and 93% in the interlaminar group. Conclusion: Both the techniques are reliable in treating L5S1 disc herniation. They offer many advantages such as day care procedure, minimal soft tissue damage, and less possibility for nerve injury. Patient selection is very important before deciding which technique to choose. From the authors’ point of view, interlaminar approach for L5S1 is easier and gives more predictable outcome compare to the transforaminal approach.

A068: Minimally Invasive Direct Thoracic Interbody Fusion (MIS-DTIF): Technical Notes of a Single Surgeon Study

Hamid Abbasi¹

¹Inspired Spine, Tristate Brain & Spine Institute, Minneapolis, MN, USA

Introduction: Minimally invasive direct thoracic interbody fusion (MIS-DTIF) is a new single-surgeon procedure for fusion of the thoracic vertebrae below the scapula (T6/7) to the thoracolumbar junction. In this proof of concept study, we describe the surgical technique for MIS-DTIF and report our experience and the perioperative outcomes of the first 4 patients who underwent this procedure. Material and Methods: In this study we attempt to establish the safety and efficacy of MIS-DTIF. We have performed MIS-DTIF on 6 spinal levels in 4 patients with degenerative disk disease or disk herniation. We recorded surgery time, blood loss, fluoroscopy time, complications, and patient-reported pain. Throughout the MIS-DTIF procedure, the surgeon is aided by biplanar fluoroscopic imaging and electrophysiological monitoring. The surgeon approaches the spine with a series of gentle tissue dilations and inserts a working tube that establishes a direct connection from the outside of the skin to the disk space. Through this working tube, the surgeon performs a discectomy and inserts an interbody graft or cage. The procedure is completed with minimally invasive (MI) posterior pedicle screw fixation. Results: For the single-level patients the mean blood loss was 90 mL, surgery time 43 minutes, fluoroscopy time 293 seconds, and hospital stay 2 days. For the 2-level surgeries, the mean blood loss was 27 mL, surgery time 61 minutes, fluoroscopy time 321 seconds, and hospital stay 3 days. We did not encounter any clinically significant complications. Thirty days postsurgery, the patients reported a statistically significant reduction of 5.3 points on a 10-point sliding pain scale. Conclusion: MIS-DTIF with pedicle screw fixation is a safe and clinically effective procedure for fusions of the thoracic spine. The procedure is technically straightforward and overcomes many of the limitations of the current MI approaches to the thoracic spine. MIS-DTIF has the potential to improve patient outcomes and reduce costs relative to the current standard of care. We are currently expanding this study to a larger cohort and recording long-term outcomes and costs.

A069: Clinical, Radiological, and Patient-Reported Outcomes 13 Years After Pedicle Screw Fixation With Balloon-Assisted Endplate Reduction and Cement Injection for Traumatic Thoracolumbar Burst Fractures

Erin De Gendt¹, J. S. Kuperus¹, W. Foppen¹, F. C. Oner¹, and J. J. Verlaan¹

¹UMC Utrecht, Utrecht, Netherlands

Introduction: In the management of traumatic thoracolumbar burst fractures, short-segment pedicle screw fixation with balloon-assisted endplate reduction (BAER) and cement injection is a safe, feasible, and effective technique to maintain radiological alignment with minimum spinal segments involved. However, still 20% of patients reported daily discomfort despite good spinal alignment and fusion 6 years after trauma. This study provides clinical, radiological, and patient-reported outcomes after a minimum 13 years follow-up. Material and Methods: A cohort of 20 patients treated 13 to 14 years earlier with pedicle screw fixation, BAER, and cement injection for traumatic thoracolumbar burst fractures was invited at the outpatient clinic for clinical/radiological examinations. Patient-reported outcome measurements were obtained, prior to examinations. Current data were compared with previously obtained results. Results: Seventeen patients (median age 50; range 32-80) of 18 patients currently alive cooperated. No or minimal back pain was reported by 15 patients and 12 patients returned to their previous heavy labor work. Median visual analogue score (80%; 50% to 100%) was similar to the results at 6 years (80%; 60% to 100% P = .259). An Oswestry Disability Score of less than 20% (reflecting minimal disability) was reported by 14 patients, compared with 15 patients at 6-year follow-up. No significant differences were found in wedge or Cobb angle between the time points. Cement resorption was not observed. Conclusion: Results from this study suggest that, 13 years after pedicle screw fixation combined with BAER and cement injection for traumatic thoracolumbar burst fractures, functional performance, pain, and radiological outcomes of the current cohort were stable or had slightly improved.

A070: Prospective Randomized Trial Comparing Endoscopic Discectomy and Conventional Open Microdiscectomy for Radicular Pain Treatment Due to Lumbar Disc Herniation

Guilherme Meyer¹, Ivan Rocha¹, and Alexandre Cristante¹

¹São Paulo University, São Paulo, Brazil

Introduction: Microdiscectomy, despite the good results, may result in damages to the local tissue. In other fields, endoscopic surgeries are considered the gold standard due to the minimal collateral damage. There are not studies comparing both methods performed in Latin America. Objective: Compare the traditional microdiscectomy and percutaneous endoscopic lumbar discectomy for the treatment of disc herniation regarding pain, disability, and complications. Methods: Prospective randomized trial with patients with sciatica due to lumbar disc herniation comparing 2 different surgical techniques. Forty-seven patients were divided into 2 groups and monitored for 12 months. Oswestry Disability Index and Visual Analog Scale for pain were recorded. Results: After surgery the leg pain and disability improved significantly but without significant difference between the groups. There was significantly less back pain after surgery until the third month. After that the groups were statistically the same. There were no statistical differences regarding recurrence, infection, and subsequent surgeries. Conclusion: Endoscopic discectomy results are similar than conventional microdiscectomy according to pain and disability improvement; however, lumbar pain is less during the first 3 months. Endoscopic discectomy is a safe and efficient alternative to microdiscectomy.

OP14: Degenerative-Lumbar 1

A071: Unilateral Laminotomy for Bilateral Decompression Versus Posterior Decompression With Instrumented Fusion for Lumbar Degenerative Spondylolisthesis at 5-Year Follow-up

Calvin Kuo¹, Maqdooda Merchant¹, Alem Yacob¹, Kamran Majid¹, and Ravi Bains¹

¹Kaiser Permanente, Oakland, CA, USA

Introduction: Controversy exists regarding whether fusion should be used to augment decompression surgery in patients with symptomatic lumbar spinal stenosis with low-grade degenerative spondylolisthesis. For years, the standard has been fusion with laminectomy in order to prevent postoperative instability. However, instability and reoperations may be reduced or prevented using structure sparing decompression techniques without the need for fusion. Material and Methods: We identified 164 patients with degenerative spondylolisthesis and lumbar stenosis who underwent unilateral laminotomy for bilateral decompression (ULBD) from 2007 to 2011 in a large integrated health care system. These patients were propensity score matched on age, gender, race, smoking status, and Charlson Comorbidity Index with patients who underwent posterior lumbar decompression and instrumented fusion (Fusion) (n = 437). The primary outcome was 5-year reoperation rate. Secondary outcome measures included postoperative complication rates, blood loss during surgery, and length of stay. Results: The reoperation rate at 5-year follow-up was significantly lower at 10.4% for ULBD compared to 17.2% for Fusion (P = .0454). Patients that underwent ULBD had significantly less mean estimated blood loss compared to Fusion (82 vs 445 mL, P < .0001) and significantly shorter mean length of stay (2.3 vs 4.6 days, P < .0001). ULBD reoperations were more frequent at the index surgical level; Fusion reoperations were more common at an adjacent level. The 2 types of operations had similar postoperative complication rates, and both groups tended to have fusion reoperations. Conclusion: For patients with stable degenerative spondylolisthesis and lumbar stenosis, ULBD is a viable, durable option compared to Fusion with a lower reoperation rate within a 5-year follow-up period, as well as decreased blood loss and length of stay. Further prospective studies are required to determine the optimal clinical scenario for ULBD in the setting of degenerative spondylolisthesis.

A072: Are There Gender-Based Differences in Outcomes for Elective Lumbar Spine Surgery in Canada?

Henry Ahn¹, Abel Davtyan¹, Sean Christie², Raja Rampersaud³, Kenneth Thomas⁴, Chris Bailey⁵, Charles Fisher⁶, Eugene Wai⁷, Jerome Paquet⁶, Albert Yee⁸, Peter Jarzem⁹, and Philippe Phan⁷

¹St. Michael’s Hospital, University of Toronto Spine Program, Toronto, Ontario, Canada

²Dalhousie University, Halifax, Nova Scotia, Canada

³Toronto Western Hospital, Toronto, Ontario, Canada

⁴University of Calgary, Calgary, Alberta, Canada

⁵University of Western Ontario, London, Ontario, Canada

⁶University of British Columbia, Vancouver, British Columbia, Canada

⁷University of Ottawa, Ottawa, Ontario, Canada

⁸Sunnybrook Hospital, University of Toronto Spine Program, Toronto, Ontario, Canada

⁹McGill University, Montreal, Quebec, Canada

Introduction: Few studies have examined the impact of gender on elective spine surgery outcomes. The purpose of this study was to determine gender differences in the outcome for lumbar decompression, microdiscectomy, and lumbar decompression and fusion, utilizing a national Canadian spine database. Material and Methods: Retrospective analysis was performed on 1316 patients who underwent either a single-level microdiscectomy (n = 614), lumbar decompression (n = 343), or a single lumbar decompression with fusion surgery (n = 259). Patients were prospectively enrolled in a national multicenter database (Canadian Spine Outcomes Research Network database), between October 2008 and August 2017. Baseline measurements included Back and Leg Pain scales, ODI (Oswestry Disability Index), Health State, SF-36 PCS/MCS, and EQ5D, along with measurements at 3 and 12 months postoperatively. Results: In the microdiscectomy group at baseline (124 females, 135 males), there was a difference in the mean MCS score, higher in males (46.2 ± 8.3 vs 44.9 ± 7.9, P < .05). There were significantly more females on pain medications (97% vs 91.1% males, P < .012). One year postoperatively, males had a higher change in the mean PCS value (14.6 ± 9.7 vs 11.7 ± 10.3, P < .012). Otherwise both groups had similar improvements in outcomes. In patients who underwent lumbar decompression surgery, there was a higher mean baseline back pain scale score among females (7 ± 2.0 vs 6.3 ± 2.6, P < .016), and females had a higher baseline ODI score (48.1 ± 13.4 vs 42.0 ± 15.6, P < .001). There was a larger percentage of female patients on pain medications (95% vs 77%, P < .05). Twelve months following surgery, females had a larger change in health state score (16.3 ± 23.1 vs 9.4 ± 22, P < .05) and a trend toward larger PCS score improvement. In patients who underwent decompression and fusion surgery, baseline scores showed a higher Patient Health Questionnaire (PHQ9) scores in females (10.9 ± 6.4 vs 7.9 ± 5.6, P < .001). There was a trend toward higher baseline ODI scores and worse baseline MCS scores for females. There were no significant differences in outcomes between genders after surgery. Conclusion: Results show significant gender differences at baseline in patients undergoing all 3 forms of elective lumbar spine surgery. Females presented with lower MCS scores especially in the group undergoing microdiscectomy surgery indicating a worse baseline mental health state. Female patients also presented with higher PHQ9 scores when undergoing lumbar decompression and fusion surgery. Furthermore, female patients had higher baseline ODI scores in groups undergoing decompression surgery, along with higher baseline back pain scores. Although female patients presented with worse baseline scores, lumbar decompression and fusion surgery was effective, with no gender-based differences in outcomes, and in lumbar decompression surgery females had better health state change with other outcomes being similar between genders. In microdiscectomy operations, male patients improved significantly with higher PCS scores following microdiscectomy surgery. It appears that female patients present with worse baseline scores than their male counterparts for all 3 forms of common elective lumbar spine surgery. Reasons for this difference need further investigation including assessing for bias among family physicians when deciding to refer patients, along with investigating potential differences in pain tolerance.

A073: Clinical Outcomes Research in Spine Surgery: What Are Appropriate Follow-up Times?

Oliver Ayling¹, Charles Fisher¹, Tamir Ailon¹, and Nicolas Dea¹

¹Vancouver General Hospital, Vancouver, British Columbia, Canada

Introduction: There has been a generic dictum in spine and musculoskeletal clinical research that 2-year follow-up is necessary for patient-reported outcomes (PRO) to adequately assess the treatment effect of surgery; however, the rationale for this duration is not evidence based. The purpose of this study is to determine the PRO follow-up time necessary to ensure that the effectiveness of a lumbar surgical intervention is adequately captured. Material and Methods: Using the different dimensions of pain, physical function, and mental quality of life of PROs from the Canadian Spine Outcomes and Research Network (CSORN) prospective database, the time course to recovery plateau after lumbar spine surgery was assessed for lumbar disc herniation, degenerative spondylolisthesis, and spinal stenosis. One-way ANOVAs with post hoc testing were used to compare the following standardized PROs at baseline, and 3, 12, and 24 months postoperatively: Disability Scale (DS), Visual Analogue Scale (VAS) leg and back pain, and Short Form (SF-12) Mental and Physical Component Summary (MCS/PCS) scores. Results: There were significant differences determined by one-way ANOVAs for all spine pathologies and specific PROs (P < .0001). Time to plateaued recovery after surgery for lumbar disc herniation (n = 661), lumbar stenosis (n = 913), and lumbar spondylolisthesis (n = 563) followed the same course for the following PROs: VAS back and leg pain, 3 months; DS, 12 months; PCS, 12 months; and MCS, 3 months. Beyond these time points no further significant improvements in PRO were seen. Patients undergoing fusion surgery plateaued at 12 months on the DS and PCS, compared to 3 months in patients that did not have a fusion for degenerative spondylolisthesis or spinal stenosis. Conclusion: Specific health dimensions of PROs follow distinctly different recovery plateaus, indicating a 2-year postoperative follow-up is not required for all PROs to accurately assess the treatment effect of lumbar spinal surgery. Ultimately the clinical research question should dictate follow-up time and the health dimension of the outcome measure utilized; however, there is now evidence to guide the specific duration of follow-up for pain, function, and mental quality of life dimensions.

A074: Nanotechnology in Spinal Fusions: Efficient or Not?

Karel Willems¹ and Philippe Lauweryns²

¹Az Delta Roeselare, Roeselare, Belgium

²Regionaal Ziekenhuis Sint-Trudo, Sint Truiden, Belgium

Introduction: Posterior interbody fusion is frequently used to treat degenerative low back pain. Titanium (Ti) cages are considered the gold standard, but the more elastic PEEK cage seems to result in better clinical and radiological outcome. PEEK is an inert material that does not promote bony ingrowth. Several types of coating have been used to combine the mechanical characteristics of PEEK with a bioactive layer. Ti coating at low temperature and high energy results in coating with a thickness in the submicron range is a dense, nonporous metallic layer on the surface of the implant not harming the microsurface topography, the radiolucency, and the elasticity of the implant. An animal study, where long bones of sheep (femur and tibia) were implanted with coated and uncoated dowels, compared the osseo-integration of PEEK dowels with that of PEEK dowels coated with CaP or with titanium. Histology showed direct bone implant contact with the Ti-coated and the CaP-coated dowels. The surface was covered with bone trabeculae, whereas on the sections of the control PEEK dowel a fibrotic layer was seen between the dowel and the surrounding bone tissue. Comparative impaction tests with PEEK PLIF (posterior lumbar interbody fusion) cages with either no coating, CaP nanocoating, or Ti nanocoating showed that the uncoated cages lost 0.39 mg, CaP nanocoated cages lost 0.57 mg, and the Ti nanocoated cages lost 0.75 mg. The wear of Ti plasma spray coated cages was 2.02 mg. A randomized controlled, double-blind, multicenter, parallel, 3-arm study of the clinical and radiographic outcome of PLIF at one level was assessed for implantations of PEEK cages, with Ti coating (TSC) or CaP coating (osteoCon) or uncoated (reCreo) cages. Material and Methods: Patients between 18 and 75 years with chronic mechanical low back pain with or without radiation into the leg (>6 months) refractory to pharmacological and nonsurgical conservative treatment scheduled for stabilization and decompression via PLIF approach utilizing supplemental posterior fixation were randomly assigned to receive implantation of PEEK cages, with Ti coating (TSC) or CaP coating (osteoCon) or uncoated (reCreo) cages. The primary radiological outcome was the implant stability and fusion status assessed with X-ray, standing A/P and lateral radiographs, and computed tomography (CT) scan at 6 and 12 months. The CT scans were evaluated by an independent experienced spine radiologist, blinded to the used spacer. The primary clinical outcome was the evolution from baseline in pain, disability, and quality of life. Clinical evaluation was performed preoperatively, and at 6 weeks, 3 months, 6 months, and 12 months. Patients were asked to report pain for the leg and for the back on a 10-point visual analogue scale (0 = no pain and 10 = the worst imaginable pain), and to fill out the Oswestry Disability Index (ODI) and the SF-36 was used preoperatively and after 12 months. Results: In the group treated with the Ti nanocoated cages more patients had definite ingrowth at 6 and 12 months. No significant clinical differences between groups were observed. Conclusion: Although the clinical outcome was not significantly different between the groups, the higher rate of bony ingrowth is important. In a meta-analysis of 4 randomized controlled clinical trials showed that fusion had an impact on the longer term clinical outcome.

A075: History of Prior Lumbar Microdiscectomy Poorly Predicts Outcomes Following Open Posterior Lumbar Fusion

Jannat Khan¹, Bryce Basques¹, Jessica Gosse¹, Michael Berkowitz¹, Konstantin Tchalukov¹, Clayton Maschhoff¹, Gagan Grewal¹, Matthew Colman¹, and Howard An¹

¹Rush University, Chicago, IL, USA

Introduction: As lumbar fusion procedures increase in the United States, there has been more scrutiny in identifying patient factors that can predict short-/long-term outcomes. While revision and extension of fusions are more technically challenging than primary procedures, there is lack of information about effects of prior microdiscectomy on outcomes after posterior lumbar fusion. This study aims to compare the differences and identify influencing characteristics in outcomes between patients undergoing elective open posterior lumbar fusion with/without a history of a prior lumbar microdiscectomy. Materials and Methods: A retrospective cohort study of consecutive patients undergoing primary elective open posterior lumbar spinal fusion at one academic institution, 2014 to 2018. Exclusion criteria: <18 years at surgery time, had procedure with fusion of intervertebral level cranial to L1, or fusion to treat lumbar fracture, tumor, or infection. Patients with history of prior microdiscectomy at a planned fusion level were identified/separated. Patient and operative characteristics were compared between unilateral and bilateral radiculopathy using χ² analysis and independent sample t tests for categorical and continuous data, respectively. Preoperative and final postoperative Visual Analog Scale (VAS) Back pain, VAS Leg pain, Oswestry Disability Index (ODI), were collected. Preoperative, immediate postoperative, and final radiographs were assessed to measure lumbar lordosis (LL), pelvic tilt (PT), pelvic incidence (PI), with PI-LL calculated. Additionally, rates of dural tear, postoperative complications, postdischarge destination, re-operation, and achievement of minimally clinically important difference (MCID) for VAS back, VAS leg, and ODI, and pseudoarthrosis, along with operative time and postoperative length of stay were collected. Binary outcome variables were compared with multivariate logistic regression, and continuous outcome variables were compared using multivariate linear regression, with lower lumbar fusions as reference. Multivariate analyses controlled for baseline characteristics; threshold for statistical significance was set at P < .05. Results: A total of 444 patients were included in this study, including 42 (9.46%) that underwent a prior discectomy at the fusion level. On bivariate analysis, patients undergoing fusion after a prior discectomy were more likely to be male (P = .008), to be a current smoker (P = .045), had higher rates of interbody use (P = .012), and less likely to have grade 1 or 2 spondylolisthesis (P < .001 and P = .038, respectively), but had no increased rate of preoperative opioid use (P = .610) or ASA (American Society of Anesthesiologists) score (P = .064). On multivariate analysis, patients with a history of prior discectomy had no difference in VAS back, VAS leg, or ODI at final follow-up (P > .05). No differences were found between rates of dural tears, postoperative complications, reoperation rates, or rates of discharge to facility. On radiographic analyses there were no differences in preoperative parameters; however, at final follow-up, patients with a history of prior discectomy had a lower average pelvic tilt (P = .026). Conclusions: There were no difference in postoperative outcomes in patients undergoing lumbar fusion between those with and without a history of prior lumbar discectomy. In patients appropriately indicated for fusion procedures, surgeons can expect good outcomes despite patients having a prior laminotomy and microdiscectomy.

A076: The Impact of Degenerative Disc Disease on Regional Volumetric Bone Mineral Density (VBMD) Measured by Quantitative Computed Tomography

Ichiro Okano¹, Stephan Salzmann¹, Courtney Ortiz Miller¹, Colleen Rentenberger¹, Jennifer Shue¹, John Carrino¹, Andrew Sama¹, Frank Cammisa¹, Federico Girardi¹, and Alexander Hughes¹

¹Hospital for Special Surgery, New York, NY, USA

Introduction: It has been reported that degenerative disc disease (DDD) is associated with higher bone mineral density (BMD). Most of the previous studies utilized dual X-ray absorptiometry (DXA). However, DDD is often associated with proliferative bone changes and can lead to an overestimation of BMD measured with DXA. Trabecular volumetric BMD measured with quantitative computed tomography (QCT) is less affected by those changes and can be a favorable alternative to DXA for patients with degenerative spinal changes. The aim of this cross-sectional observational study is to investigate the effect of DDD on regional volumetric BMDs (vBMDs) measured by QCT in patients undergoing posterior lumbar fusion at a single academic institution. Material and Methods: Institutional ethics board approval was obtained for this study. We reviewed the data of consecutive patients undergoing posterior lumbar spinal fusion between 2014 and 2017 who had a routine preoperative CT scan and MRI (magnetic resonance imaging) within a 90-day interval. Patients on any anti-osteoporotic treatment were excluded and 132 patients were included in this study. QCT measurements were conducted in L1 to S1 vertebral trabecular bone. Any apparent sclerotic lesions that might affect vBMD values were excluded from the area of interest. Among 660 discs of the 132 patients, levels with spontaneous fusion, previous fusion surgery, or poor image quality were excluded and 626 discs were analyzed. The vBMDs of each level were defined as the average vBMD of the upper and lower vertebrae. To evaluate DDD, 5-grade Pfirrmann grade, Modic grade, and total end plate score were documented. Univariate regression analysis and multivariate analyses with a generalized linear mixed model adjusted with individual variability of segmental vBMDs were conducted with vBMD as the response variable. The statistical significance level was set at P < .05. Results: Mean age (±SD [range]) was 65.9 ± 11.3 (26-87). A total of 58.3% of the patients were female. Mean vBMD (±SD) was 119.0 ± 39.6 mg/cm³. Univariate analyses demonstrated that Pfirrmann grades showed negative associations with vBMD (P < .001 in all grades), whereas any Modic changes (type 1, P = .012; type 2, P = .002; type 3, P = .019) and high endplate score (score 10-12, P < .001) were associated with high vBMD. After adjusting with age, body mass index, race, disc level, gender, and previous surgery, Pfirrmann grade was not an independent contributor of vBMD (grade 2 vs 3, P = .09; vs 4, P = .14; vs 5, P = .98), but the presence of any Modic change (type 1, +8.3% P = <.001; type 2, +5.2%, P < .001, type 3, +25.9%, P < .001) and high total endplate score (score 6-7, +4.7%, P = .040; 10-12, +12.1%, P < .001) were shown to be independent contributors of vBMD. Conclusion: Our results demonstrate that the presence of Modic change and higher total end plate score were significantly associated with an increase of regional trabecular vBMDs in apparently normal bone areas, but no association was observed with Pfirrmann grade. This finding suggests that there is no direct association between vBMD and disc degeneration itself, but the concomitant endplate changes have positive effect on regional vBMD in this patient population.

OP15: Deformity-Thoracolumbar (Adolescent) 1

A077: Novel Intraoperative Technique for Selecting Optimal Lower Instrumented Vertebra in Adolescent Idiopathic Scoliosis

Vishal Borse¹, Peter Loughenbury¹, and Peter Millner¹

¹Leeds General Infirmary, Leeds, UK

Introduction: Surgical treatment of adolescent idiopathic scoliosis (AIS) is achieved with a posterior correction and fusion. Determination of fusion levels can be contentious as surgery aims to achieve a balanced correction that allows preservation of motion segments. Various preoperative methods (fulcrum bending, side bending, push-pull techniques) have been described to help with decision making. We describe a novel intraoperative technique that can be used to decide on the lowest instrumented vertebra (LIV) during surgical correction. Materials and Method: Patients with Lenke types III-VI AIS requiring instrumentation into the lumbar spine. Preoperative planned fusion levels (upper and lower instrumented levels) were determined using standard protocols for our unit (fulcrum bending views, AP traction views, remaining growth potential, and morphology of individual curves). Intraoperatively, the level proximal to the planned LIV was exposed and instrumented with bilateral pedicle screws. A short, temporary, transverse rod was attached to the screws. A coronal bending moment was applied to the rod using a standard “rod holder.” Real-time fluoroscopy was used to determine the flexibility of this level in both directions—to identify whether the disc space below the instrumented level becomes parallel or closed on the convexity. If the level below was flexible (fully correctable distal disc space) then this level was chosen as the LIV and one planned fusion level was “saved.” If there was limited flexibility then the preoperative plan was followed. Results: Description of a novel intraoperative technique with case examples and intraoperative imaging. Conclusion: We describe a novel method for the intraoperative evaluation of LIV. The method has the potential to improve effective determination of the LIV, producing shorter constructs with preservation of distal motion segments.

A078: Asymmetric Biomechanical Characteristics of the Paravertebral Muscle in Adolescent Idiopathic Scoliosis

Aixing Pan¹ and Yong Hai¹

¹Beijing Chaoyang Hospital, Beijing, China

Introduction: We sought to assess the biomechanical properties of the paravertebral muscles in adolescent idiopathic scoliosis (AIS) patients with Lenke type 1, 2, or 3 (Lenke 1-3) curves. Material and Methods: The MyotonPro and shear wave elasticity imaging (SWEI) system were used to assess the biomechanical features of the thoracic paravertebral muscles in AIS patients with Lenke 1-3 curves. Muscle tone (F), stiffness (S), relaxation time (R), Deborah number (C), and elasticity (D) of the paravertebral muscles on the concave and convex sides of the scoliosis curvature were detected at the following points: (a) apex of the curve, (b) upper and (c) lower limits of the curve. The Cobb angle of the main curve was measured using the anteroposterior whole spine radiograph in the standing position. Results: A total of 40 AIS patients with a mean Cobb angle of 66.49 ± 32.8° were included in this study. Muscle tone on the concave side was significantly greater than that on the convex side (a: 18.8 ± 2.2 Hz vs 17.4 ± 1.6 Hz, b: 18.1 ± 1.9 Hz vs 16.7 ± 1.6 Hz, c: 18.0 ± 2.3 Hz vs 16.8 ± 1.6 Hz, P < .05). Muscle stiffness on the concave side was significantly greater than that on the convex side (a: 401.3 ± 76.8 N/m vs 354.3 ± 75.6 N/m, b: 365.6 ± 68.5 N/m vs 326.3 ± 80.4 N/m, c: 380.7 ± 87.5N/m vs 346.3 ± 67.1N/m, P < .05). Relaxation time was significantly longer on the convex side than on the concave side (a: 15.0 ± 2.2 ms vs 13.6 ± 2.5 ms, b: 16.0 ± 2.3 ms vs 14.7 ± 2.4 ms, c: 15.4 ± 2.3 ms vs 14.5 ± 2.8 ms, P < .05). Deborah number was greater than that on the convex side at a and b points (a: 0.9 ± 0.1 vs 0.8 ± 0.1, b: 0.9 ± 0.1 vs 0.9 ± 0.1, P < .05). No statistically significant difference in muscle elasticity was observed between the concave side and the convex side (P > .05). Pearson correlation analysis demonstrated that stiffness on the concave side was moderately positively correlated with the Cobb angle (P < .05, r = 0.582); the Deborah number on both sides and the relaxation time on the concave side showed a moderate negative correlation with the Cobb angle (P < .05, r = −0.632; r = −0.432; r = −0.611). Conclusion: Concave paraspinal muscle tone and stiffness were greater than those on convex side in AIS patients. The asymmetric biomechanical characteristics of paraspinal muscles are closely related to the severity of scoliosis.

A079: Predictors for Postoperative Shoulder Balance in Lenke 1 Adolescent Idiopathic Scoliosis: A Prospective Cohort Study

Alberto Gotfryd¹, Maria Fernanda Caffaro¹, Robert Meves¹, and Osmar Avanzi¹

¹Santa Casa de São Paulo, São Paulo, Brazil

Introduction: Imbalanced shoulders are a major cause of dissatisfaction in AIS patients. In Lenke 1 curves, MT curve fusion is supposed to lead to spontaneous correction of the proximal thoracic (PT) curve and thereby promote shoulder balance. However, this is not always observed. The main goal of the present study is to determine the predictors of the shoulder balance after main thoracic (MT) fusion in patients with Lenke 1 adolescent idiopathic scoliosis (AIS). Material and Methods: Fifty-two Lenke 1 AIS patients who underwent MT fusion by a posterior approach were prospectively evaluated preoperatively, immediately postoperatively, and 2 years after the surgical procedure. The shoulder balance was determined using the biacromial angle. The clinical results were examined for their correlation with several radiographic measurements. Results: Spontaneous correction of the PT Cobb angle after MT fusion was noted in 52% of cases, similar to that observed on preoperative bending films. A total of 51% of patients had unbalanced shoulders before surgery (right side higher). Two years after surgery, 30.77% showed unbalanced shoulders (P < .001). However, 17.1% of patients presented with a higher left shoulder, a reversion of the initial deformity. This phenomenon was more common among the patients with mild or no shoulder asymmetry (biacromial angle inferior to 1°) before surgery (P < .001). It was also determined that for each degree measured for the clavicle angle, there was an elevation of 0.14° for the ipsilateral shoulder. Conclusions: In Lenke 1 cases with higher right shoulder and absence of abnormalities in the sagittal plane view, the correction of the main right thoracic curve could be enough to balance the shoulders. No correlation was found between shoulder balance and the amount of correction of the PT and MT curves.

A080: How Can We Predict the Development of Shoulder Imbalance After the Correction of Idiopathic Scoliosis?

Aleksandr Aleinik¹, Andrey Bokov¹, and Sergey Mlyavikh¹

¹Privolzhskiy Research Medical University, Nizhniy Novgorod, Russian Federation

Introduction: A number of factors influence the patient self-assessment after the surgical correction of idiopathic scoliosis: the amount of deformity correction, residual rib hump, balance of the vertebral column, and shoulder balance. Most of these factors can be controlled by the surgeon. However, previous studies have not revealed a strong prognostic factor for the development of persistent shoulder imbalance. The aim of our study is to identify the perioperative factors that reliably determines the development of persistent shoulder imbalance. Material and Methods: Ninety-four patients with idiopathic scoliosis (all Lenke types) were included in a retrospective study. All of them underwent corrective surgery with all-screw technique. Mean age was 17.8 ± 6.3 years. All patients underwent full spine biplanar radiography prior to surgery, in the early postoperative period and 12 months after. The main parameters of deformity and indicators of sagittal and frontal balance of the spine were measured. The difference of the clavicle shadows (RSHD) was measured. We considered the shoulder imbalance significant (SHI) with RSHD ≥20 mm. If it persists for 12 months, we considered persistent shoulder imbalance (PSHI). On the LP-radiographs after the operation the severity of asymmetry of the proximal and main thoracic spine was assessed. The presence of pronounced asymmetry was considered as “double rib hump” (DRH). Radiological signs of DRH: 2 lifts of the rib-shadows in the proximal and distal thoracic regions, crossing of the rods at the level of the transition of the main thoracic curve (MT) to the proximal thoracic curve (PT), a significant difference in the level of the screw-heads at the upper thoracic spine. Statistical analysis was carried out using Friedman-ANOVA, Spearman correlation analysis, and the Mann-Whitney U test. Results: SHI was registered in 27 cases (28.72%). The majority of patients had self-correction of the shoulder level in the first year after surgery. However, in 13 (13.83%) cases PSHI was registered. Among these patients were presented different types of deformity: 3 patients, Lenke 1; 3, Lenke 2; 3, Lenke 3; 2, Lenke 4; 1, Lenke 5; 1, Lenke 6. Thus, 5 (38.46%) patients had structural PT before surgery, and 8 (61.54%) no structural PT. In 6 (46.15%) PSHI patients, the upper instrumented vertebra was higher than the T3, and in 7 (53.84%) T3 and below. According to demographics and initial parameters of the deformity no significant differences between PSHI group and patients with balanced shoulders were revealed. PSHI correlated with the following factors: the presence of a structural PT (P = .041 642), the amount of residual PT, r = 0.22 (P = .03), DRH, r = 0.75 (P > .005). We did not see marked dependence of PSHI of the instrumentation level, initial magnitude of the deformity, or the initial parameters of the shoulder girdle. Conclusion: Not only frontal plane spinal deformity but also asymmetry of the chest affects the balance of the shoulders. LP-radiographs provide data on the rib chest deformity, which underlies the development of persistent shoulder imbalance. The proposed X-ray parameters can be reproduced in an operating room, which will reduce the likelihood of the shoulder imbalance after surgical correction of idiopathic scoliosis.

A081: A Novel Mobile App-Based Handheld Scanner Puts Scoliosis Follow-up Assessment Into the Hands of Patients/Relatives: A Feasibility Study

Cheuk Ki Chan¹, Michael Kai Tsun To¹, Kenny Yat Hong Kwan¹, Jason Pui Yin Cheung¹, Xin Yu¹, King Cheung Berry Cheung², Yuk Lung Tsang², Sunny Lut Hey Chu², Johnson Yiu Nam Lau², and Kenneth Man Chee Cheung¹

¹The University of Hong Kong, Hong Kong

³Avalon SpineCare (HK) Ltd, Hong Kong

Introduction: Radiographic measurements of Cobb angles is the current gold standard for assessment and follow-up of adolescent idiopathic scoliosis (AIS) patients. Progressive curvatures require bracing or surgical treatment. The regular attendance of the follow-up clinics is a burden not just to the patients and their family but also a burden on the health care system, in addition to the increased radiation exposure to the patients. SpineScan3¹ is a novel hand-held device that can record the shape of the back by the gyroscope and accelerometer embedded within a computer chip. It has been shown to be a reliable assessment tool for a patient’s back topography with good correlation with radiographic Cobb measurements.² The purpose of this prospective study was to compare the difference in measurements obtained by a well-trained technician (WTT) and newly trained person (NTT, ie, relatives of patients) using SpineScan3D, to determine the feasibility of this device for potential home assessment by non-medical person. Material and Methods: A total of 109 patients with AIS ranging from 8 to 20 years of age were recruited. Measurements of back topography were done in the forward bending position and the standing position, rolling the SpineScan3D from approximately lower cervical (C7) to lower lumbar (L5) level. Patients were first examined by WTT, also as a way to teach and perform once for NTT. The NTT then performed the examination from the beginning. This was a noninvasive examination with no potential for adverse events or risk involved. Data collected were analyzed using SPSS, version 25.0. Results: The axial plane data were displayed as maximum tilt angles in the forward bending position (S1 and S2) and in standing position (K1 and K2). Interrater reliability was assessed by Cronbach’s α: S1 = 0.629, S2 = 0.781, K1 = 0.622, K2 = 0.522. After removing 23 subjects (21%) in which the differences between the WTT and NTT were more than 2 SD of the mean difference from the analysis, the Cronbach’s α are S1 = 0.789, S2 = 0.884, K1 = 0.835, K2 = 0.738 (n = 86). The scanning results obtained are similar for both groups with mean absolute difference around 2° (S1 = 2.30°; S2 = 2.14°; K1 = 1.46°; K2 = 1.61°). Overall there was a strong positive correlation between the results obtained by WTT and NTT. Cronbach’s α coefficient between 2 groups for the 4 major domains were high. The mean absolute difference of 2° between WTT and NTT is not clinically significant. Interclass correlation was excellent for scanning in the forward-bending position and good for scanning in the standing posture. As there is a subgroup of relatives (23 individuals, 21% of total) who seem to have difficulty using the device, current plan is to further refine the training scheme to aim at further improve the performance of this subset of NTT. Conclusion: This study supports the potential for home monitoring by use of a novel hand-held scanner, to relieve parents’ concerns of curve progression and reduce the need for clinic follow-up and radiation exposure. Additional areas for further improvement were also identified.

A082: A New Modified Luque Trolley Technique for the Treatment of Early-Onset Scoliosis (EOS): A 13 Patient Case Series With a Minimum 5-Year Follow-up

Andrew Cottam¹, Bart Van Herwijnen¹, and Evan Davies¹

¹University Hospital Southampton, Southampton, UK

Introduction: We present a large single surgeon case series evaluation of a new growth guidance technique for the treatment of progressive EOS (early-onset scoliosis). A traditional Luque trolley construct uses wires to hold growth guidance rods together. We describe a new technique that uses Domino end to side connectors in place of the wires with the aim of providing a stronger construct to better limit curve progression while allowing longitudinal growth. Methods: We conducted a thorough retrospective review of patient records and radiological imaging. Sequential measurement of Cobb angle and length of rods was recorded, as well as any further surgical procedures and associated complications. This enabled us to quantify the techniques’ ability to limit curve progression and simultaneously allow growth of the construct. In total 28 patients (20 idiopathic, 4 syndromic, and 4 neuromuscular) have been treated with this technique, 25 of which have a minimum of 2-year follow-up and 13 with a minimum of 5-year follow-up. Results: The average correction of the preoperative Cobb angle was 48.9%. At 2-year follow-up the average loss of this initial correction was 15°, rising to only 20° at a minimum of 5 years (including 4 patients with 8 year plus follow-up). The growth of the constructs’ was limited. The average growth at 2 years was 3.7 mm rising to 19 mm at 5-year follow-up. Patients who underwent surgery with this technique before the age of 8 seem to do better. This group had only an 18% revision rate at an average time of 7 years after the index procedure. The average growth in this group was 22 mm. This compares to the group who had their index surgery after the age of 8, which had a 64% revision rate at an average of 3.2 years after surgery and an average growth of only 11.6 mm. Overall in the series there were 4 hardware failures (14%), 1 deep infection (3.5%), and only 10 patients had 1 extra surgery after the index procedure (36%). Only 2 of the 13 patients who are at 5-year plus follow-up have been revised. Conclusion: This technique has a good capacity for initial curve correction and for limiting further curve progression with limited longitudinal growth before 2 years and improved growth thereafter. It appears that this technique may not be so useful after the age of 8 years due to poor growth and a higher early revision rate. We have also demonstrated a low-cost technique with a low hardware failure rate that saves many future surgeries for the patient compared to other techniques used in the treatment of EOS.

OP16: Tumor 2

A083: Management of Cervical Spine Malignant Disease: A Case Series and Systematic Literature Review

Pedro Coutinho¹, Michael Paci², Andrew Fanous², and Glen Manzano²

¹Yale University, New Haven, CT, USA

²University of Miami, Miami, FL, USA

Introduction: The spinal column is the most common site of malignant bony tumors, whether primary or metastatic. Epidural spinal cord compression incidence may be up to 19% in certain types of malignancies such as breast cancer. Approximately in one quarter of patients with new diagnosis of metastatic spinal cord compression, spinal epidural compression is the first presentation of malignancy. The use of stereotactic body radiation therapy (SBRT) has demonstrated promising outcomes in the treatment of spinal bone metastases. However, the presence of tumoral lesions in the epidural space represents the major shortcoming for this type of treatment. For such situations, surgical treatment is usually the first line of treatment, aiming primarily to maintain adequate neurological function. The cervical spine is the least common site of vertebral column metastasis, with only 8% to 10% of all patients with vertebral metastatic disease afflicted with cervical lesions. For this reason, there is a paucity of experience and evidence in the current literature regarding the management of epidural malignant lesions in that region. Material and Methods: In this study, we present a retrospective analysis of our experience with the surgical management of cervical malignant tumors. The study represents our institutional experience between the years 2008 and 2016. Demographic data, history, physical exam, presenting symptoms, imaging, operative records, and follow-up for each patient were collected. Additionally, the type of primary cancer, stage at presentation, adjuvant therapies, operative approach and techniques, amount of blood loss, postoperative neurological status, and complications were examined. We also performed a systematic literature review with a detailed search of the PubMed database with the following key words: cervical spine, metastasis, radiosurgery, chemotherapy, and spine surgery. Results: A total of 11 patients with malignant cervical spine tumors of various pathologies were identified (melanoma—3; multiple myeloma—3; lung sarcomatoid carcinoma—1; small cell lung cancer—1; male breast adenocarcinoma—1; esophageal adenocarcinoma—1; and Hodgkin lymphoma—1). Each of them underwent surgery. Nine patients had laminectomies with instrumented fusion, one patient had a corpectomy with fusion, and one had minimally invasive resection of the right C7 facet and transverse process. No perioperative complications occurred for any of the cases. The average duration of follow-up was 21.9 months. Neck pain significantly improved in all patients. Of the 2 cases that presented with myelopathic signs, one had major improvement on follow-up evaluation, while the second had only partial improvement in myelopathy and weakness. Two patients had local recurrence of tumor, while other 2 died as a result of widespread multisystem disease. The remaining 7 patients were in remission with no evidence of local recurrence at the time of follow-up. Conclusion: Cervical spine malignant lesions are relatively rare and less frequent than in other segments of the spine. Based on findings in both our clinical series and systematic review of literature, aggressive management with cervical decompression or radiosurgical treatment plays an important role in improving patients’ quality of life.

A084: Comparing Complication Rates for Corpectomy Versus Posterior Approaches in Operations for Metastatic Cervical Spine Disease: A National Database Study

Yaroslav Gelfand¹, Rafael De La Garza-Ramos¹, Murray Echt¹, Jonathan Nakhla², Michael Longo¹, and Reza Yassari¹

¹Montefiore Medical Center, New York, NY, USA

²Rhode Island Hospital of Brown University, Providence, RI, USA

Introduction: Minimizing complications in oncologic spine surgery is crucial for maximizing the benefits to the patient. Often the nature of the disease and structural failure dictate the surgical approach; however, there are cases when both an anterior or posterior approach could achieve similar results in decompressing the spinal cord and stabilization. We compared the rate of perioperative complications in patients who underwent an anterior approach (corpectomy) versus a posterior approach (laminectomy with/without fusion) for metastasis in the cervical spine. Material and Methods: Patients undergoing surgery for cervical metastatic spine disease were identified in the American College of Surgeons National Surgical Quality Improvement Program (ACS-NSQIP) database from 2006 to 2016. Two groups were established as follows: (1) patients who underwent one or more level corpectomy; and (2) patients who underwent posterior approach to the cervical spine (laminectomy with/without fusion). A χ² test was used to assess the association between surgical approach and complication and mortality rates. A multivariate analysis was used to confirm independent association between these values. Results: A total of 288 patients were identified who underwent fusion surgery for metastatic cervical spinal disease: 36 (13%) underwent corpectomy surgery compared to 252 (87%) who underwent posterior laminectomy and fusion. Complication rate was 27.8% and 13% (P = .021) and the 30-day mortality rate was 16.7% and 11.5%, respectively, but the difference was not statistically significant (P = .37). After controlling for gender, history of hypertension, smoking, operative time, and emergency versus elective procedures, patients who underwent corpectomy for metastatic disease of the cervical spine were more likely to develop a complication (odds ratio [OR] = 2.86; 95% confidence interval [CI] = 1.16-7.02; P = .022). When comparing average operative times, anterior approaches were significantly faster than posterior approaches (3.3 hours vs 4.1 hours, P = .02) even when controlled for other comorbidities and emergency surgeries. Conclusion: In this study, we found that despite significantly shorter operative times anterior surgical approaches (corpectomy) for metastatic cervical spine disease had significantly higher rate of complications than posterior approaches. While more studies are required to further elucidate this relationship, a posterior approach for metastatic cancer patients may be more desirable if it can provide adequate stabilization.

A085: Cervical En Block Spondilectomy: Planning, Results, and Failures (16 Cases)

Alessandro Luzzati¹, Gennaro Scotto¹, and Luca Cannavò¹

¹Istituto Ortopedico Galeazzi, Milan, Italy

Introduction: Cervical localization of spinal primitive tumors is relatively rare. The choice of the best surgical strategy for the treatment of these lesions, in order to obtain an acceptable radicality, is very difficult in these sites. Cervical pain is the most common symptom. The early diagnosis of these tumor is not easy because frequently patients present nonspecific symptoms and because radiological examinations are usually negative. Only after performing more detailed examinations, such as computed tomography (CT) scan and magnetic resonance imaging (MRI), is it possible to have a diagnostic biopsy, which is mandatory in these patients. Material and Methods: The authors report the experience of a small group of patients (16) affected by cervical primitive malignant tumors: 5 osteosarcomas, 1 aggressive osteoblastoma, 5 chordomas, 3 chondrosarcomas, and 2 sinovial cell sarcomas. The mean age was 42 years (range 11-61 years), the mean follow-up was 21 months (range 8-42 months). The levels of resection were as follows: 1 level of vertebrectomy in 4 patients, 2 levels of vertebrectomy in 4 patients, 3 levels of vertebrectomy in 6 patients, and 4 levels of vertebrectomy in 2 patients. In every patient we performed a double approach, in 2 patients the second approach was performed 48 hours after the first one to minimize the surgical stress. In 2 cases we performed a trans-mandibular approach because of the rostral localization of the tumor. In the majority of patients we performed a long fixation (occipito-cervico-thoracic fixation) associated with cages filled with anterior autogenous cortico-spongiosus bone chips; in the others the anterior reconstruction was an autogenous graft (vascularized in 1 case). Results: Three patients had a local recurrence, respectively, at 20, 25, and 34 months after surgery and they died due to lung involvement after about 12 to 15 months from the local recurrence (only one patient underwent local surgery). One patient died 23 months after surgery for general progression without signs of local recurrence. One patient died within 1 week from surgery for vascular complications. The other patients are alive, with no signs of local disease (locally free-disease) and no signs of systemic disease (NED: no evidence disease). Conclusion: En bloc resection for primary cervical tumor of the spine is a challenge for the surgeons due to the complexity of the anatomy of this region: the presence of the vertebral artery (both resected in 2 cases without neurological damage), the contiguity of the aero-digestive tract and of the main encephalic vessels, the presence of medulla oblongata and spinal cord. Three patients had under-lesional damage after surgery, in partial remission after some months. All other patients did not have neurological damages. Our high percentage of local recurrences and of major complications (5 deaths) is probably due to anatomical complexity of the region, where sometimes is very difficult, or even impossible, to obtain acceptable resection margins.

A086: En Bloc Resection of Tumors of the Lumbar Spine: A Systematic Review of Outcomes and Complications

Morgan Jones¹, Okezika Uhiara¹, James Holton¹, Simon Hughes¹, Petr Rehouzek¹, Alistair Stirling¹, Mel Grainger², and Marcin Czyz¹

¹ The Royal Orthopaedic Hospital NHS Trust, Birmingham, UK

²University Hospitals of Birmingham NHS Foundation Trust, Birmingham, UK

Introduction: Total en bloc spondylectomy (TES) is an aggressive surgical technique that is indicated for primary malignant bone tumors, aggressive benign tumors, and infrequently solitary metastatic lesions. TES in the lumbar spine specifically represents a challenge due to its anatomy, which may necessitate combined anterior and posterior approaches, as well as the potential functional deficits patients may be left with as the result of nerve root sacrifice. Currently, the data informing practice is the result of large heterogeneous case series of variable quality, utilizing different techniques across multiple spinal levels for a range of pathologies. The aim of this study is to undertake a systematic review of the literature associated with en bloc resection of primary tumors of the spine in the lumbar and thoracolumbar region with regard to the predictors of favorable outcomes and complications. Material and Methods: We undertook a systematic review following the recommendations of the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA 2009) guidelines. Ethical approval was not required as our study only reviewed published articles and did not directly involve patient intervention. The literature review was conducted using the PubMed, Medline, OVID, EMBASE, and Cochrane Library databases on March 22, 2017. The following search terms were used: “lumbar,” “spine,” en-bloc,” “spondylectomy,” “tumour*,” “tumor*,” “malignant,” “primary.” Boolean operators (AND, OR) were combined with relevant keywords to refine the search. Results: An initial literature search of PubMed, Medline, OVID, EMBASE, and the Cochrane Library databases identified 55 papers; after applying exclusion criteria 9 studies describing a total of 87 patients remained for final analysis. The most common primary tumor was GCT (35%), followed by chordoma (22%) and chondrosarcoma (12%). Thyroid, renal, and breast metastases were the most common secondary malignancies (14% each). The levels resected were as follows: L4 (26%), L2 (21%), L5 (21%), L3 (20%), L1 (12%). The median length of follow-up was 42 months (interquartile range = 22-64). The presence of extracompartmental tumor was predictive of complications (odds ratio = 24.25, P < .05). No specific surgical approach was associated with increased risk of blood loss or complication although there was a significantly faster operative time utilizing the single-stage combined anterior-and-posterior approach. Complication rates were reported in up to 65% of cases although there was a reassuringly low rate of deep infection. Five-year survival was 92% and 64% in the primary and metastatic groups, respectively. Conclusion: This is the largest review of the available literature on the outcomes and complications of lumbar spondylectomy for primary bone tumors. As predicted, there was better survival in patients presenting with primary tumors and achieving wide margins intraoperatively. Due to the complexity of the procedure complication rates are high, though not all studies were able to comment on individual cases. A more complete understanding of morbidity, survivorship, and outcomes associated with TES in the lumbar spine will allow us to better manage this hugely heterogeneous group of patients.

A087: Transpedicular Vertebrectomy With Circumferential Spinal Cord Decompression and Anterior-Posterior Reconstruction for Patients With Thoracic Spinal Metastatic Tumors

Tarush Rustagi¹, Hazem Mashaly¹, Michelle Williams¹, David Kline¹, and Ehud Mendel¹

¹Ohio State University, Columbus, OH, USA

Introduction: Most spinal metastases causing cord compression are located in the vertebral body. Apart from spinal cord compression, frequent association include vertebral collapse, spine instability, and local kyphosis. The main objectives of surgical treatment are decompression, pain control, spine stabilization, and improvement in quality of life/survival. Anterior approaches are associated with increased morbidity, anesthetic risk, postoperative complications, and contraindicated with preexisting chest comorbidities. The transpedicular vertebrectomy (TPV) has recently gained popularity allowing circumferential spinal cord decompression and spinal column reconstruction from a posterior-only incision. Our study was to review our experience of consecutive cases of TPV including presentation, outcome, and complications. Methods: Study was conducted at Ohio State University, James Cancer Center, from 2008 to 2016. The inclusion criteria included consecutive cases (between age 15 and 100 years) with TPV between T2 and T12 following metastasis. All other causes of revision surgeries were excluded. Demographic, surgical, and clinical data were collected through chart review. The pre- and postoperative images were used to evaluate extent of disease, spinal cord compression (SCC), and degree of vertebral body collapse. Antero-posterior thecal sac diameter was measure preoperatively and postoperatively based on axial T2 weighted MRI (magnetic resonance imaging). The Cobb angle before and after surgery was calculated on the sagittal CT (computed tomography) scan images. Hall-Wellner confidence band was used for the survival curve. The Sign Test was used to assess the changes for Frankel Grade, cord diameter, and Cobb angle. Fisher test and Wilcoxon signed rank test were used to to assess differences by the construct type (cage vs cement). Results: Seventy-five patients included 38 females and 37 males. The mean age at the time of surgery was 56.63 (range 25-81 ± 11.62 SD). Forty-two patients (56%) presented with mechanical pain, 3 (4%) with radicular pain, 28 (37.3%) with both back pain and radicular pain, and 2 (2.7%) had potentially unstable asymptomatic lesion. Fifty-two percent cases had lung, breast, and prostate metastasis and the remaining was a mixed group. Operative: Single-level TPV was performed in 59 patients (78.7%), 2 levels in 15 cases (20%), and 3 levels in 1 case (1.3%). Anterior reconstruction included polymethyl methacrylate (PMMA) in 61 patients (81.3%), and expandable titanium cage in 14 patients (18.7%). The mean blood loss for the entire cohort was 2000 mL; average hospital stay was 8 days following surgery. Preoperative embolization was done in 27 cases. Complications: There were 5 (6.7%) intraoperative complications. One patient had a durotomy with loss of SSEPs and developed weakness in the postoperative period. Four patients had a durotomy during surgery with good outcome. Eighteen patients (24%) developed postoperative complications, and 3 patients (4%) died within 30 days of surgery. Two patients (2.7%) developed progressive weakness from epidural hematoma requiring emergent evacuation. One of the 61 patients (1.7%) had cement displacement into the spinal cord that was treated surgically. Frankel grade changed in 13 patients, 11 (84.6%) of those saw improvement (P = .0225). Radiological: SCC improved by mean of 5 mm (P < .0001). The Cobb angle improved by mean of 7° (P < .0001). Survival: The mean survival time was estimated to be 8 months with a 95% confidence interval of (5, 10). Cage/PMMA Comparison: Surgical outcome and complication rates are similar between the 2 construct types, we are unable to conclude equivalence between the 2 constructs. We did not conduct a direct cost comparison. Conclusion: Single-stage TPV can provide adequate access for circumferential spinal decompression and allow for 360° spinal instrumentation in patients with thoracic spinal metastases. No statistical difference in clinical or radiological outcomes or complication rates between patients who had anterior column reconstruction using the PMMA versus titanium cage.

A088: Patient Satisfaction After Surgery and/or Radiotherapy for Metastatic Spine Disease

Anne Versteeg¹, Arjun Sahgal², Laurence Rhines³, Daniel Sciubba⁴, James Schuster⁵, Michael Weber⁶, Aron Lazary⁷, Stefano Boriani⁸, Chetan Bettegowda⁴, Michael Fehlings⁹, Michelle Clarke¹⁰, Paul Arnold¹¹, Ziya Gokaslan¹², and Charles Fisher¹³

¹University Medical Center Utrecht, Utrecht, Netherlands

² Sunnybrook Odette Cancer Centre and University of Toronto, Toronto, Ontario, Canada

³The University of Texas MD Anderson Cancer Center, Houston, TX, USA

⁴ Johns Hopkins University, Baltimore, MD, USA

⁵Hospital of the University of Pennsylvania, Philadelphia, PA, USA

⁶McGill University and Montreal General Hospital, Montreal, Quebec, Canada

⁷National Center for Spinal Disorders and Buda Health Center, Budapest, Hungary

⁸ IRCCS Istituto Ortopedico Galeazzi, Milan, Italy

⁹ University of Toronto and Toronto Western Hospital, Toronto, Ontario, Canada

¹⁰ Mayo Clinic, Rochester, MN, USA

¹¹The University of Kansas Hospital, Kansas City, KS, USA

¹Brown University, Rhode Island Hospital, and The Miriam Hospital, Providence, RI, USA

¹³University of British Columbia and Vancouver General Hospital, Vancouver, British Columbia, Canada

Introduction: Health-related quality of life (HRQOL) has been recognized as one of the most important outcomes in evaluating care for patients with spinal metastases. In addition to HRQOL, a patient’s satisfaction with treatment is becoming more important. The objective of this study was to evaluate patient satisfaction with surgical and/or radiotherapy treatment for metastatic spine disease. Materials and Methods: Patients with spinal metastases treated with surgery and/or radiotherapy were enrolled in a prospective international multicenter observational study. Demographic, histologic, treatment, adverse event, and HRQOL data were collected. Evaluation of HRQOL included the NRS pain score, EQ-5D-3L, and the Spine Oncology Study Group Outcome Questionnaire (SOSGOQ2.0). Patient satisfaction was derived using the posttreatment SOSGOQ2.0 questions at 6, 12, and 26 weeks posttreatment. Patients were classified as satisfied, neutral, or dissatisfied. Last observation carried forward was used in case of missing data. Results: At 12 weeks posttreatment, 158 (84%) of the surgically treated patients were satisfied and 9 (5%) were dissatisfied compared to 95 (77%) of the patients treated with radiotherapy alone being satisfied and 7 (6%) being dissatisfied. Dissatisfaction after surgical treatment was associated with lower baseline values for leg strength (P = .031) and lower social functioning scores (P = .053). Significant improvements in pain, physical function, mental health, social function, leg function, and EQ-5D were associated with satisfaction after surgical treatment. Patients who were dissatisfied after treatment with radiotherapy alone were more often single (P = .030) and showed a trend for lower baseline social function scores (P = .069). Satisfaction after radiotherapy treatment was associated with significant improvements in pain, mental health, and overall SOSGOQ2.0 scores. Conclusion: High levels of satisfaction with treatment outcomes were observed after treatment with surgery plus radiotherapy or treatment with radiotherapy alone. Posttreatment satisfaction was associated with significant improvements in pain and the different dimensions of HRQOL including physical function, social function, and mental health. Funding: This study was funded by an Orthopaedic Research and Education Foundation grant and by AOSpine International through the AOSpine Knowledge Forum Tumor.

OP17: Minimally Invasive Spine Surgery 2

A089: Percutaneous Endoscopic Lumbar Discectomy for the Treatment of Multisegmental Lumbar Disc Herniation in Middle-Aged and Young Adults: A Retrospective Cohort Study

Haolin Sun¹

¹Peking University First Hospital, Beijing, China

Introduction: To compare the clinical effect and safety of different operation strategy (PELD [percutaneous endoscopic lumbar discectomy] of responsible segment or total segments) for the treatment of multisegmental disc herniation in middle-aged and young adults. Methods: A total of 49 cases of patients with lumbar disc herniation of double segments treated by PELD from January 2015 to December 2016 were retrospectively reviewed, in which 32 patients were treated with responsible segment and 15 patients were treated with double segments. The clinical outcomes were evaluated and compared by imaging parameters including lumbar lordosis angle, lumbar intervertebral height in lumbar spine X-ray, and Pfirrmann grades of disc in lumbar MRI (magnetic resonance imaging). Others include operation time, anesthesia time, hospitalization time, and postoperative complications. The surgical results were evaluated according to the Visual Analogue Scale (VAS) and Japanese Orthopaedic Association (JOA) score. Results: The follow-up period was 6 months in both groups. The operation time, anesthesia time in responsible segment group were significant shorter than that in the 2-segments group (P < .05). No difference was found in hospitalization time between the 2 groups. There was no significant difference between the 2 groups in the VAS score and JOA at 3 and 6 months postoperation (P > .05). One patient in the responsible segment group presented with an unrelieved postoperative pain, which did not disappear until after the other section of surgery. Conclusions: For the treatment of lumbar disc herniation of double segments, PELD for responsible segment could have the similar clinical efficacy compared with PELD for double segments, which could preserve the stable structure of the spine while reducing operating time and anesthesia time.

A090: Modified OLIF: Comparison OLIF and DILF

Hyung Chang Lee¹

¹Department of Anterior Spine Surgery, Busan, Republic of Korea

Introduction: The objective of this study is describing detailed surgical technique and short-term clinical and radiological outcomes of our new method, modified oblique lumbar interbody fusion (OLIF) and posterior screw fixation for revision surgery. Material and Methods: We compared 3 patients group. Group 1 consists of 20 patients who had underwent DILF. Group 2 consists of 21 patients who had underwent OLIF. Group3 consist of 57 patients who had underwent modified OLIF. All patients underwent OLIF surgery and used a cage and allobone graft without real-time neuromonitoring. The other patients had underwent DILF with neuromonitoring. After the insertion of the interbody cage, posterior fixation was performed. The key concept of our new method is approaching more perpendicular site than standard OLIF to achieve easy perpendicular procedure and get more margin from vessels and ureter. The second concept is dissection under the psoas muscle so that lesser injury to muscle. We use illumination retractor and Langenbeck. No specific device is needed. The radiological and clinical outcomes were assessed preoperatively and at 1, 3, 6, and 12 months postoperatively. Results: There was no perioperative complication in the modified OLIF group. There was 1 ureter injury in the OLIF group. DILF group had more postoperative complications such as leg elevation difficulty. Clinical and radiologic outcomes did not shown any difference between groups. No serious complications occurred in all cases. Bone fusion was successful in all cases. Conclusion: Our modified OLIF is useful and safe. It can avoid muscle and nerve injury due to undercut and retract whole psoas muscle. And it can get more distance from ureter and vessels so that we get lower complication rates.

A091: The Efficacy of Posterior Cervical Foraminal Decompression Using Unilateral Biportal Endoscopy

Jin Hwa Eum¹

¹Bumin Hospital, Busan, Korea

Introduction: Cervical foraminal stenosis is the narrowing of the hole that the cervical spinal nerves go through and exit the spinal column. Surgery remains the best therapy for patients under 65 years and patients with extensive clinical symptoms. Decompression of cervical foramen can be achieved through anterior, posterior, and combined approaches. All dorsal procedures share the same disadvantage of a broad midline incision often carried out from C3 to the T1 level. To avoid this large tissue trauma, minimally invasive procedures are currently used for 1- to 3-level laminectomy and foraminotomy. A new endoscopic technique that uses a biportal endoscopic approach has been applied to conventional arthroscopic systems for spinal disease. This technique was adopted for cervical foraminal soft disk protrusion and cervical foraminal stenosis with good clinical results. Material and Methods: There were 27 patients with radicular symptoms. There were 19 females and 8 males. Patients completed a questionnaire (Neck Disability Index [NDI], Visual Analogue Scale [VAS], and Modified Japanese Orthopedic Association [mJOA] score) preoperatively, immediate postoperatively, and during the follow-up visit. Results: There were 19 females and 8 males. Mean age was 52.4 years. Mean follow-up duration was 1.2 years. A 1-level procedure was performed in 22 cases and a 2-level procedure in 5 patients. Operative time ranged from 120 to 216 minutes (mean 164 minutes). The duration of symptoms before surgery was an average of 20.1 months (range 1-78 months), and there was no correlation between the duration of symptoms and recovery. No specific complications occurred during surgery and the immediate postoperative period. All patients showed improvement in their radicular symptoms. Average VAS improved from preoperative of 6.6 to 3.1 in the immediate postoperative period, 2.6 at 6 months postoperative and 2.25 at 12 months. Average NDI improved from preoperative value of 27.8 (56.6%) to 8.2 (16.4%) at 12 months. Conclusion: Posterior endoscopic cervical decompression with a unilateral biportal endoscopic approach is a good alternate surgical technique for cervical foraminal stenosis; however, the results must be confirmed with additional clinical cases and a longer follow-up period.

A092: Efficiency of Indirect Decompression in Minimally Invasive Oblique Lateral Lumbar Interbody Fusion (MIS-OLIF) Without Posterior Decompression for the Treatment of Degenerative Lumbar Disease

Jung-Woo Hur¹, Kyeong-Sik Ryu¹, Jin-Sung Kim¹, Ho-Jung Chung¹, and Myung-Soo Song¹

¹Seoul St. Mary’s Hospital, The Catholic University of Korea, Seoul, Republic of Korea

Introduction: Indirect decompression using posterior longitudinal ligament (PLL) ligamentotaxis with posterior stabilization was found to be a useful technique for some patients with stable thoracolumbar burst fractures. Recent studies advocate the usefulness of indirect decompression with large lateral interbody cage in some selected cases, but its efficiency has not been proven yet. The purpose of this study was to demonstrate the clinical and radiological results of indirect decompression using PLL ligamentotaxis in minimally invasive oblique lateral interbody fusion (MIS-OLIF) without posterior decompression for the treatment of degenerative lumbar disease. Material and methods: We have retrospectively reviewed 236 patients who underwent MIS-OLIF without posterior decompression for the treatment of single or 2-level degenerative lumbar diseases since November 2013 to May 2018. Clinical outcomes were measured using the Visual Analog Scale (VAS) scores and Oswestry Disability Index (ODI). Radiologic measurements were determined using plain lateral radiographs and sagittal and axial magnetic resonance imaging. Measurements included disc height, foraminal height, foraminal area, canal diameter, and cross-sectional area (CSA) of thecal sac at the disc level. Differences in preoperative and postoperative radiologic parameters were compared, and the relationship between the ratio of extension and that of the preoperative values were assessed. Results: XLIF (extreme lateral interbody fusion) procedure has been developed recently and used increasingly. XLIF is performed using a lateral approach that passes through the retroperitoneal space and psoas muscle. Oliveira et al reported that substantial dimensional improvement was evidenced with the increases of 13.5% in foraminal height, 24.7% in foraminal area, and 33.1% in central canal diameter. Kepler et al also reported improvement of foraminal area by as much as 35% in degenerative foraminal stenosis using XLIF. OLIF procedure was introduced more recently and has several theoretical advantages over XLIF procedure. These include less invasion of the psoas muscle and lumbar plexus, decreased need for neuromonitoring, and access to the L4-5 level with a high-riding pelvis and the L5-S1 level. Fujibayashi et al reported that the mean cross-sectional area of the thecal sac increased from 99.6 mm preoperatively to 134.3 mm postoperatively using OLIF. Sato et al also reported significant increases in spinal area (19%) and intervertebral foramen areas (21% to 39%). There was significant improvement of clinical results after the surgery. Substantial dimensional increment was demonstrated in all radiographic parameters, with increases of 49.1% in average disc height, 33.7% in foraminal height, 44.5% in foraminal area, 37.4% in central canal diameter, and 36.2% in CSA. The median extension ratios were inversely correlated with preoperative values. Six patients (2.5%) required additional posterior decompression for symptomatic remnant neural compression. Comparing slippage group with stenosis (without slippage) group, slippage group yielded better extension ratios in radiographic parameters with statistical significance. Conclusion: Significant improvements of foraminal and spinal canal area were demonstrated after MIS-OLIF without posterior decompression for the treatment of degenerative lumbar disease. Bulging of intervertebral discs was reduced through height reduction, and ligamentotaxis of the PLL may have decompressed the spinal canal. Patients with more severe degeneration and overt slippage preoperatively yielded better radiologic improvement.

A093: Cement Volume and Pattern of Distribution in Fenestrated Screws Used in Cement Augmented Spinal Stabilization: Its Relevance With BMD and Functional Outcome

Saumyajit Basu¹, Amitava Biswas¹, and Rohan Gala¹

¹Park Clinic, Kolkata, India

Introduction: Fenestrated screw is an important tool in a surgeon’s armamentarium in osteoporotic spine. Yet, there is limited evidence on radiological assessment of these screws as well as on their clinicoradiological outcome. The objective of our study is the assessment of cement volume and pattern of distribution in relation to the BMD (bone mineral density) and functional outcome. Material and Methods: Thirteen patients with at least 6-month follow-up were analyzed. They were evaluated in the form of CT (computed tomography) scan and parameters like cement morphology, zone, and pattern of distribution were assessed. The cement distribution in the vertebrae was divided into 4 zones on the axial CT cuts: anterior one third, middle one third and posterior one third of vertebral body, and the pedicle. Pattern of distribution was defined as concentrated or scattered. The leakage pattern was divided into anterolateral, posterolateral, and canal leakage. The relationship among BMD, volume, and pattern of cement distribution was studied and patients were clinically evaluated by comparing the preoperative and follow-up VAS (Visual Analogue Scale) and ODI (Oswestry Disability Index) scores. Results: Mean age of patients was 69 years with a mean follow-up of 14 months. Out of 92 screws, 84 (91.3%) had cement in the safe zone (1 and 2), 7 (7.6%) had in zones 1, 2, and 3, and only 1 (1%) had cement in zone 1. Concentrated cement distribution had higher BMD value than scattered distribution (P = .027). Preoperative mean ODI and VAS scores improved significantly (P < .001). Mean cement volume occupancy (3.66%) did not correlate with the mean improvement of VAS (65.5%) (R = −0.068), ODI (43.14%) (R = −0.088), or preoperative BMD value (R = −0.007). Seven screws had cement leakage in the anterolateral zone. Conclusion: Fenestrated screws are safe in osteoporotic spine. The volume of cement is independent of the BMD and functional outcome. However, the pattern of distribution relates well with the BMD.

A094: Surgical Treatment for Lumbar Lateral Recess Stenosis With the Percutaneous Endoscopic Technique: Interlaminar Approach Versus Transforaminal Approach

Yawei Li¹ and Bing Wang¹

¹The Second Xiangya Hospital of Central South University, Changsha, China

Introduction: Due to the minimally invasive technical development, percutaneous endoscopic lumbar decompression has become a representative minimally invasive spine surgery for lumbar spinal stenosis. For lateral recess stenosis, both interlaminar approach and transforaminal approach are feasible. Materials and Methods: From July 2015 to June 2016, 192 patients with lumbar lateral recess stenosis were randomly recruited into 2 groups: 102 patients underwent percutaneous endoscopic interlaminar decompression (PEID), and 90 patients underwent percutaneous endoscopic transforaminal decompression (PETD). The operation time, intraoperative radiation time, postoperative bed rest time, hospitalization time, and complications were compared between the groups. The surgical effectiveness was assessed using Visual Analog Scale (VAS), Oswestry Disability Index (ODI), and modified MacNab criteria. Results: In 2 groups, no significant differences were found in terms of the operative time, postoperative bed rest time, and hospitalization time (P > .05). The radiation time was 0.50 ± 0.36 seconds in the PEID group and 7.5 ± 1.65 seconds in the PETD group, and there was a significant difference between the 2 groups (P < .01). The postoperative ODI and VAS were obviously improved in both groups when compared with preoperation (P < .01). A mini-open reoperation was needed in 3.9% with PEID and in 7.8% with PETD because leg pain did not improve. According to the MacNab criteria, the satisfactory rates were 86.3% and 78.9% in the 2 groups, with a significant difference (P < .05). Conclusions: Percutaneous endoscopic technique via an interlaminar or transforaminal approach could obtain functional improvements in patients with lumbar lateral recess stenosis. However, PEID have advantages and include a faster puncture orientation, less intraoperative radiation exposure, and better clinical outcomes.

OP18: Degenerative-Lumbar 2

A095: Lumbar Disc Disease in Children and Adolescents: A Cross-Section Imaging Study

Jun Zhang¹, Zhi-Heng Liu², Fei Zhao³, Hai-Qiang Wang⁴, and Dino Samartzis⁵

¹The First Affiliated Hospital of Xi’an Jiaotong University, Xi’an, China

²Chinese PLA No.451 Hospital, Xi’an, China

³Fourth Military Medical University, Xi’an, China

⁴Shaanxi University of Chinese Medicine, Xi’an, China

⁵Rush University Medical Center, Chicago, IL, USA

Introduction: Lumbar disc disease (LDD) affects the health and welfare of people greatly, even children and adolescents. However, there is a paucity of knowledge on the prevalence and features of LDD among adolescents. The study aimed for filling in the blanks for children and adolescents (age less than 19 years). Material and Methods: Lumbar spinal diagnostic MR (magnetic resonance) imaging from a territory university hospital was collected during January 2008 to December 2010 with cross-section design. Both T1 and T2 weighted lumbar MRI images from L1/2 to L5/S1 were profoundly analyzed in axial and sagittal planes. Results: Overall, there were 599 children and adolescents (2995 lumbar discs) with lumbar MRI images among 5288 patients. The prevalence of LDD and LDH (lumbar disc herniation) was 2.2% and 3.5%, respectively, significantly lower than that of adults (P < .05). Modic changes, high intensity zone, spondylosis deformans, and lumbar disc sequestration were absent in these children and adolescents. Modic changes emerged at the third decade of cases; high intensity zone could occur as young as 19 years. Spondylosis deformans initiated at 48 years, whereas lumbar disc sequestration could occur as young as 20 years. Conclusion: The study provides the first line of evidence on the prevalence and features of LDD in children and adolescents. LDD emerges as early as the first decade of life without specific phenotypes, including Modic changes and high intensity zone.

A096: Effect of Spinal Decompression on Back Pain in Lumbar Spinal Stenosis: A Canadian Spine Outcomes Research Network (CSORN) Study

Shreya Srinivas¹, Charles Fisher², Greg Mcintosh³, and Nicolas Dea²

¹Alder Hey Children’s Hospital, Liverpool, UK

²Vancouver Spine Surgery Institute, Vancouver, British Columbia, Canada

³Canadian Spine Society, Toronto, Ontario, Canada

Introduction: Surgical decompression is usually offered for improvement of neurogenic claudication in patients with lumbar canal stenosis. These patients often have associated low back pain (LBP) and little is known about surgical results for this disabling symptom. The goal of the present study was to specifically quantify the improvement of LBP following surgical decompression for lumbar canal stenosis and to identify factors associated with change in LBP in this population. Materials and Methods: Consecutive patients who underwent surgical treatment for lumbar spinal stenosis without instability (1-2 level) were prospectively enrolled in the multicenter Canadian Spine Outcomes Research Network (CSORN) registry. Patient-reported outcomes were collected at baseline and at 3, 12, and 24 months after surgery. The primary outcome was change in LBP on the Numeric Rating Scale (NRS). Multivariable logistic regression was used to model the relationship between the outcome and potential factors associated with achieving minimal clinical important difference (MCID) in back pain using a backward selection procedure. Results: A total of 1221 patients were included in the analysis. Mean age was 64 years and 58% were males. Follow-up evaluations were available in 926 patients at 3 months and 651 patients at 12 months. LBP significantly improved 3 months after surgery and sustained at 24 months (P < .001). We found that 74% of patients reached MCID with regard to NRS back pain. Predictive factors for sustained improvement (12 and 24 months) in LBP after surgical intervention were absence of narcotic usage or compensation claims and increased severity of LBP prior to surgery (high NRS). Conclusion: Alleviation of clinically significant LBP was observed at 3 months after lumbar decompression surgery for lumbar spinal stenosis and this was maintained at 12 and 24 months after surgery in vast majority of patients. Increasing back pain severity, narcotic usage, and compensation claims were predictors of LBP improvement. Type of surgical intervention, including addition of fusion, did not have any association with improvement of LBP.

A097: Dynamic Pedicle-Based Stabilization Versus Fusion in Lumbar Degenerative Instabilities: A Prospective Randomized Multicenter Double-Blind Trial (Dynorfuse Trial)

Bernhard Meyer¹, Claudius Thome¹, Peter Vajkoczy¹, and Florian Ringel¹

¹Department of Neurosurgery, München, Germany

Introduction: 360° fusion is a standard treatment for symptomatic degenerative lumbar instabilities but is associated with relevant morbidity. Pedicle-based dynamic stabilization may be an alternative to reduce surgical complexity. This prospective randomized double-blind trial tested the hypothesis of clinical noninferiority of dynamic stabilization versus fusion for lumbar degenerative instabilities. Material and Methods: Patients with symptomatic low-grade lumbar degenerative disease (±stenosis) with either (1) ≥5 mm spondylolistheses or segmental instability of ≥3 mm or 10° in flexion/extension imaging and (2) failed conservative therapy for ≥3 months were randomized to TLIF (transforaminal lumbar interbody fusion) fusion or dynamic stabilization in a 1:1 fashion. Primary endpoint was the between treatment group difference in ODI (Oswestry Disability Index) 24 months after treatment. Patients and independent observers remained blinded to treatment allocation. Results: A total of 301 patients were randomized at 17 centers to fusion (FU) or dynamic stabilization (DY). Baseline data were not different except for a nonrelevant difference in ODI of 45.8 ± 17.2 versus 42.5 ± 17.1 (FU vs DY). Duration of surgery was significantly (P = .007) shorter for DY (188 vs 231 minutes) with reduced blood loss (370 vs 491 cc) (P = .038). ODI 24 months after treatment showed a significantly and clinically relevant reduction in both groups, without significant intergroup differences (30.8 ± 19.1 in FU vs 31.2 ± 18.1 in DY). Secondary outcome parameters also revealed no significant (P < .05) differences between groups, such as the incidence of serious adverse events. Early revisions were necessary in 4 cases in DY and in 2 cases in FU. Twenty-four months after surgery 34% versus 33% of patients had ongoing pain, and 0% versus 3% (DY vs FU) underwent hardware revision. Conclusion: Dynamic pedicle-based stabilization is noninferior to fusion in the treatment of low-grade lumbar degenerative instabilities, but surgically less complex with secondary failures being not different between groups.

A098: Abdominal Obesity and Paraspinal Muscularity Predict Proximal Junctional Kyphosis Requiring Surgical Revision Among Patients Undergoing Thoracolumbosacral Fusion

Zach Pennington¹, Ali Ahmed¹, Ethan Cottrill¹, Jeff Ehresman¹, Camilo Molina¹, Matthew Goodwin¹, C. Rory Goodwin², and Daniel Sciubba¹

¹Johns Hopkins School of Medicine, Baltimore, MD, USA

²Duke University, Durham, NC, USA

Background: Each year thousands of American adults undergo lumbosacral or thoracolumbosacral fusion operations for degenerative spine conditions. Prior research estimates that PJK (proximal junctional kyphosis) occurs in 20% to 40% of these patients, of which 50% require surgical revision, with an average cost of over $77 000. Elevated BMI (body mass index) and sagittal malalignment have been described as risk factors for PJK, but no literature exists describing weight distribution as a risk factor for PJK. Here we describe the influence of abdominal girth and paraspinal muscular sarcopenia as independent predictors of PJK. Materials and Methods: Retrospective review of all adult (>18 years old) neurosurgical spine patients undergoing lumbosacral or thoracolumbosacral fusion (T10-L4 to S1/pelvis) at the Johns Hopkins Hospital for degenerative conditions of the lumbar spine between January 1, 2013, and December 31, 2017. Variables considered were patient BMI, thoracolumbar muscularity, abdominal girth, operative details, and surgical history. All patients had a minimum of 6 months of radiographic follow-up and PJK was classified as having to undergo surgery with primary indication of PJK or adjacent segment disease. Results: Of the 81 patients with full records, 16 (19.7%) underwent surgery for PJK. These patients had significantly larger preoperative sagittal vertical axis, sacral slope, pelvic incidence, pelvic incidence-lumbar lordosis mismatch, postoperative SVA (sagittal vertical axis), postoperative change in LL (lumbar lordosis), and postoperative change in pelvic incidence (PI)-LL mismatch. They also had less robust muscularity at the top of their construct, larger girth-to-muscle ratios, and longer constructs. Multivariate logistic regression incorporating these factors successfully predicted surgical PJK with a C-statistic of 0.9028. Conclusions: Body habitus, body mass distribution, and construct characteristics all determine the risk of postoperative PJK in patients undergoing thoracolumbosacral or lumbosacral fusion for degenerative spine disease. The independent nature of construct and body habitus suggest that optimal surgical outcomes can be achieved by modifying the preoperative surgical plan to match patient body habitus.

A099: Functional and Radiological Outcomes of Combined Anterior-Posterior Approach Versus Posterior Alone in Management of Isthmic Spondylolisthesis: A Systematic Review and Meta-Analysis

Abduljababr Alhammoud¹, Gregory Schroeder², Osama Aldahamsheh¹, Kenan Alkhalili³, Mayan Lendner², Isam Sami Moghamis¹, and Alexander Vaccaro²

¹Department of Orthopedics, Doha, Qatar

² Rothman Institute, Thomas Jefferson University, Philadelphia, PA, USA

³Cairo University, Cairo, Egypt

Background: Lumbar isthmic spondylolisthesis (IS) in adults is defined as the forward slippage of a vertebra onto the top of the vertebra, resulting from a defect in the pars intraarticular, and can be low grade or high grade. Persistent back pain or neurological deficit are indications for surgical intervention. Surgery can be done from back, front, or both, with or without fusion, instrumentation, or decompression, and short or long segment. The aim of this study is to compare the fusion rate, functional outcomes, radiological outcomes, and complication rate of combined anterior-posterior approach (ALIF+PLF/PSF) [anterior lumbar interbody fusion + posterolateral fusion/pedicle screw fixation] versus posterior approach alone (TLIF/PLIF+PLF) [transforaminal lumbar interbody fusion/posterior lumbar interbody fusion + posterolateral fusion] in the management of isthmic spondylolisthesis. Methods: Following the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines, several databases were searched through August 2017 for any observational or experimental studies that evaluated combined anterior-posterior approach versus posterior alone in management of isthmic spondylolisthesis. Primary outcome was fusion rate, whereas secondary outcomes included functional outcomes (Visual Analogue Scale [VAS] and Oswestry Disability Index [ODI] score), complication rate (infection, neurological), and reoperation rate. Descriptive, quantitative, and qualitative data were extracted. Most of the cases were low-grade isthmic spondylolisthesis. Results: Of the 645 articles identified, 6 studies were eligible for the meta-analysis, with a total of 397 patients with isthmic spondylolisthesis, 198 in the combined (ALIF+PLF) group, and 199 in the posterior (TLIF/PLIF+PLF) group, average age of 47.2 years, and 185/212 male/female ratio. Although the fusion rate reached 100% in some studies, the pooled odds ratio [OR] of fusion rate (OR: 1.02, 95% confidence interval [CI]: [0.294, 3.552], [P .972]) did not reach statistical significant between (ALIF+PLF) versus (TLIF/PLIF+PLF). The estimated pooled SDM showed less blood loss in anterior approach compared to posterior approach (SDM: −0.528, 95% CI: [−0.777, −0.278] [P < .001]), with no difference in operative time and length of hospital stay. Despite both groups showing significant improvement in pain and functional scores at final follow-up, ODI and VAS were not significantly different between groups with ODI (SDM: −0.644, 95% CI: [−1.948, 0.621] [P .311]) and VAS (SDM: 0.113, 95% CI: [−0.173, 0.400] [P .439]). The complication rate for anterior approach was higher than posterior whereas reoperation rate was higher in posterior approach than anterior. Conclusions: No significant difference between anterior and posterior approaches was found in the global assessment of fusion rate and clinical outcomes, despite a higher rate of complications using the anterior approach.

A100: PROs Are More Than a Research Tool: A Prospective Assessment of Patient Understanding and Expectation Management

Ross McEntarfer¹, Natasha Greene¹, Jeremy Shaw¹, Chase Grover¹, William Ryan Spiker¹, Brandon Lawrence¹, Nicholas Spina¹, and Darrel Brodke¹

¹University of Utah, Salt Lake City, UT, USA

Introduction: Patient-reported outcome (PRO) measures such as PROMIS Physical Function (PF) are increasingly being used in spine patient populations to define the course of disease, to understand the effect of treatment, and to improve the value of care. Using individualized PRO data at the point of care with patients is the next step in translating outcome research. However, patient comprehension and preference in receiving and interpreting their outcome data has yet to be assessed. The purpose of this study is to assess the patient’s perception of their own physical function and to define preferences in how their data are presented to better convey appropriate expectations of treatment. Materials and Methods: A unique 15-question survey was developed to assess both a patient’s understanding and preference in receiving their PROMIS PF outcome data (whether graphical, plain language, or score formats). We surveyed 100 consecutive patients presenting to the spine clinic who finished a PROMIS PF questionnaire. Demographic information was gathered from survey and chart review. All findings are presented using descriptive statistics with survey validity and internal reliability using STATA 13. Results: Mean age of patients was 59, with 58% female and 47% having a college degree or higher. Ninety-two percent of patients understood the outcome data when presented in graphical format with no significant difference in preference when given the option of plain language descriptions. Fifty-eight percent of patients thought all or most plain language statements accurately reflected their physical function when translated from their score, while 31% thought some statements were inaccurate. Seventy-nine percent of patients preferred graphical or plain language descriptions when compared to a raw number scores (21%). Those who preferred the graphical and plain language were also more likely to find it useful when making a treatment decision (78%), and only 1 patient found the PROMIS data in any format not helpful. Conclusion: To date no other studies have investigated spine patients’ understanding and preference with individualized PROMIS PF outcome data. These data suggest that patients both understand and find PROMIS PF outcome data useful and would integrate it into their treatment decision making. This has broad applications moving forward in how data can be used to educate and counsel patients about their treatment options and better define realistic expectations after treatment.

OP19: Deformity-Thoracolumbar (Adolescent) 2

A101: The Use of Preoperative Magnetic Resonance Imaging for Patients With Scoliosis at University Hospital Southampton, UK

Andrew Cottam¹, Evan Davies¹, and Sulagna Roy¹

¹University Hospital Southampton, Southampton, UK

Introduction: The purpose of this study was to determine the use of preoperative MRI (magnetic resonance imaging) scans at UHS (University Hospital Southampton) and the prevalence of underlying neurological abnormalities. We also wanted to explore whether there is a link between abnormal MRI scans and intraoperative cord monitoring problems and postoperative complications. The use of preoperative MRI for pediatric scoliosis patients is not standardized, and recommendations vary geographically and in the literature. Patients with undiagnosed intradural abnormalities (eg, tethered cord and synrinx) are thought to be at increased risk of neurological injury during scoliosis correction surgery. The incidence of these conditions in patients with scoliosis varies widely in the literature (5% to 50%). Methods: A retrospective study was conducted on 303 pediatric scoliosis patients who were scheduled for surgery between 2014 and 2017 at UHS. Patient notes and imaging and operation notes were reviewed. Of the 303 patients, 218 patients received a preoperative MRI scan. Intraoperative spinal cord monitoring (SSEPs and MEPs) was reviewed. The data were then statistically analyzed using Fishers exact test, χ² test, and independent t test as appropriate. Results: The cohort was made up of 227 patients with idiopathic scoliosis (209 adolescent and 18 early onset) and 76 patients had a neuromuscular or syndromic diagnosis. In the idiopathic group, the prevalence of neurological abnormalities was 15%. Of those 35 patients with an underlying abnormality, only 14% (5) had cord monitoring problems. In our cohort 3 patients suffered permanent neurological injury after surgery, all had normal preoperative MRI scans. Of the 5 wake up tests performed, only 1 had an abnormal MRI scan. There were no significant differences between those who had abnormal and normal MRIs for gender, age, curve characteristic, type of scoliosis, and cord monitoring issues (P > .05). There was also no significant difference in preoperative Cobb angles between the 2 groups (P = .25). Conclusion: Preoperative abnormal MRI scans are not a good predictor of cord monitoring problems or neurological complications (contrary to most previous reports in the literature), and our patients with intradural abnormalities were not at increased risk of neurological complication. An inference of this may be that routinely scanning all preoperative scoliosis patients confers no patient benefit.

A102: Abnormalities Associated With Congenital Scoliosis in High-Altitude Geographic Regions

Dongpo Hou¹ and Yong Hai¹

¹Beijing Chaoyang Hospital, Capital Medical University of China, Beijing, China

Introduction: Congenital scoliosis (CS) is not uncommon and is usually seen as a sporadic entity, due to anomalous development of the vertebrae including failure of formation and segmentation during embryogenesis including uneven segments (hemi-vertebrae and wedged vertebrae), fused segments (block vertebrae), and problems in midline fusion (butterfly vertebrae). Although the causes of congenital scoliosis have not been fully identified. Previous studies have demonstrated that both genetics and environmental factors are implicated in the development of vertebral abnormalities. This study is to assess the different characteristics of congenital scoliosis between low-altitude geographic regions and high-altitude geographic regions in Chinese population and discuss the role of hypoxia on those differences. Methods: A total of 120 patients with congenital scoliosis who underwent surgical treatment in our hospital between January 2009 and October 2017 were identified. Complete data were reviewed, including medical records, X-ray, CT (computed tomography), and MRI (magnetic resonance imaging) preoperatively. According to the patient’s birthplace, they were divided into low-altitude geographic regions (low group) and high-altitude geographic regions (high group). Characteristics of vertebral deformities, rib deformities, and intraspinal malformations in the 2 groups were analyzed. Results: A total of 397 segments of vertebral deformities were involved in the 2 groups, of which 136 were involved in high group and 261 in low group. The average segments involved were 4.5 and 2.9, respectively, in 2 groups. A total of 63.3% patients in high group have rib deformities, which is significantly higher than that of low group (41.1%); and the proportion of patients with complex rib deformities in high group was also higher than that in low group (57.9% vs 24.3%). The incidence of CS associated with intraspinal malformations in low group was 38.9%, which was similar to those reported previously; however, the incidence of that in high group was 63%, significantly higher than previous reports. Conclusion: Our results suggested that CS patients in high-altitude geographic regions might tend to have higher proportion and more severe of rib deformities, and also be more likely to accompany with intraspinal malformations. So we supposed that hypoxia not only aggravated the proportion and severity of rib deformities but also affected the development of spinal cord in humans.

A103: Patient-Related Outcomes in Adolescents and Young Adults With Idiopathic Scoliosis Undergoing Spine Deformity Surgery: A 2-Year Follow-up Cohort Study Based on a Single-Center Spine Outcomes Registry

Miranda van Hooff¹, Maarten Spruit¹, Raf Mens¹, Philip Horsting¹, Marinus de Kleuver², and Luuk de Klerk¹

¹Sint Maartenskliniek, Nijmegen, Netherlands

²Radboud University Medical Center, Nijmegen, Netherlands

Introduction: International consensus exists to systematically monitor patient-related outcomes (ie, patient-reported outcome measures [PROMs] and clinician-based outcome measures) in deformity surgery for adolescents and young adults (AYA) with idiopathic scoliosis.¹ Current studies are mainly based on radiographic measurement of the curvature.² Evaluation of outcome domains that matter to patients is recommended (eg, functioning and health-related quality of life),³ but often remain underexposed. As continuous outcome monitoring is important, for example, in value-based health care, recently within an AOSpine knowledge forum deformity study consensus was reached to use a core outcome set to evaluate AYA with idiopathic scoliosis undergoing spine deformity surgery.⁴ The purpose of this study is to evaluate the 2-year follow-up using patient-related outcomes, as specified in the core outcome set, in AYA with idiopathic scoliosis undergoing spine deformity surgery. Material and Methods: A single institution consecutive cohort study in the Netherlands. Since March 2014 all patients undergoing scoliosis surgery are systematically monitored over time and registered in an online web-system, which is connected to the patients’ electronic file. Routinely, relevant patient characteristics, radiological, and perioperative parameters are reported. PROMs used were the following: ODIv2.1a (functioning), SRS22r and EQ5D (health-related quality of life), NRS (0-10; back and leg pain). Improvement over time (preoperative, 1-year and 2-year follow-up [fu]) in AYA, aged ≤25years, is evaluated with repeated-measures ANOVA (P < .05). Clinical relevancy of PROMs is determined by published minimal clinical important changes (SRS22r)⁵ and by satisfactory symptom state comparable to healthy persons (ODI ≤ 22).⁶ Clinician-based outcome measures were postoperative complications and revision surgery. The institution’s review board approved the study and an opt-out procedure was applied as ethical approval for this study was not required. Results: Ninety-one patients with AYA idiopathic scoliosis were included for analysis (age 15.4 years [SD 2.6], range 12-25; females [n = 72; 79.1%]), with primary Cobb of 56.1° (SD 9.1). The majority had a Lenke type 1 (t1 n = 42 [46.2%]; t2 n = 16; t3 n = 13; t4 n = 5; t5 n = 6; t6 n = 6; not reported n = 3). The PROMs response rate at 2-year fu was 70.5% (91/129). Preoperatively, the non-responders did not differ from those who completed the PROMs at fu. Except for leg pain (F _[2,89] = 0.84; P = .44), all PROMs showed a similar pattern; significant improvement up to 1-year fu and maintenance of results between 1- and 2-year fu. At 2-year fu 82 (90.1%) reached satisfied symptom state comparable to healthy persons (preoperatively n = 73 [80.2%]) and 68 patients reached relevant improvement on SRS22r self-image (74.4%). Seven complications (7.7%) occurred in 7 patients of whom 1 patient (1.1%) required revision surgery for non-union. Conclusion: After spine deformity surgery AYA with idiopathic scoliosis improve significantly on all outcome domains and reach a relevant improvement in functioning and self-image. The 2-year response rate to the PROMs is acceptable, a high response is recommended to use the results as reliable and valuable information for studies and quality control. As scoliosis surgery is regarded as high complex care, continuous and standardized outcome monitoring is important to improve the quality of spine care, which is the basis for value-based spine care.

References

1. de Kleuver M, Lewis SJ, Germscheid NM, et al. Optimal surgical care for adolescent idiopathic scoliosis: an international consensus. Eur Spine J. 2014;23:2603-2618.

2. Weinstein SL, Dolan LA, Cheng DC, Danielsson A, Morcuende JA. Adolescent idiopathic scoliosis. Lancet. 2008;371:1527-1537.

3. Bagó J, Climent JM, Pérez-Grueso FJ, Pellisé F. Outcome instruments to assess scoliosis surgery. Eur Spine J. 2013;22(suppl 2):S195-S202.

4. de Kleuver M, Faraj SSA, Holewijn RM, et al. Defining a core outcome set for adolescent and young adult patients with a spinal deformity. Acta Orthop. 2017;88:612-618.

5. Bagó J, Pérez-Grueso FJ, Les E, Hernández P, Pellisé F. Minimal important differences of the SRS-22 Patient Questionnaire following surgical treatment of idiopathic scoliosis. Eur Spine J. 2009;18:1898-1904.

6. van Hooff ML, Mannion AF, Staub LP, Ostelo RW, Fairbank JC. Determination of the Oswestry Disability Index score equivalent to a “satisfactory symptom state” in patients undergoing surgery for degenerative disorders of the lumbar spine—a Spine Tango registry-based study. Spine J. 2016;16:1221-1230.

A104: Does Pregnancy Increase Loss of Correction After Posterior Spinal Fusion?

Pawel Grabala¹, Tomasz Guszczyn¹, and Ilkka Helenius²

¹Medical University in Bialystok, Bialystok, Poland

²University of Turku and Turku University Hospital, Turku, Finland

Introduction: Patients and their families frequently asked about long-term outcomes after scoliosis surgery and their ability to get pregnant and future curve progression after delivery. We are not equipped to these questions due to insufficient evidence. We hypothesized that pregnancy has no impact on loss of correction in women who underwent posterior spinal fusion. Materials and Methods: A retrospective study with group comparison. Patients underwent scoliosis surgery in 1999 to 2014. Women were recruited from 4 centers for this study and inclusion criteria were women age 16 to 40 years who underwent surgery for scoliosis in adolescence and subsequently became pregnant (SPG). The study population was compared to other group: women who underwent scoliosis surgery but had not yet became pregnant (SNP). Queries regarding curve before surgery, after surgery, before pregnancy, and up to 1 year after delivery, at final follow-up (FFU) minimum 2 years after surgery. Additional patient-related outcomes measures were performed by SRS-22r. Results: A total of 128 patients were enrolled: PG—62 and NP—66. Average age at the time of scoliosis surgery was 16 years. Mean time of delivery after scoliosis surgery was 5 years (range 2-12, SD 2.80). The mean preoperative Cobb angle of main curvature was 65 (SD 12), after surgery 18 (SD 9), at FFU 20 (SD 8) for NP, and respectively, 67 (SD 11), 17 (SD 9), at FFU 20 (SD 8) for PG. The mean preoperative thoracic kyphosis (T5-T12) was 26.5 (SD 11.9) for PG and 24.7 (SD 14.5) for NP, after surgery—19.2 (SD 9.5) for PG, 18.8 (SD 8.9) for NG, and at FFU, respectively, 20.3 (SD 9) for PG and 21.3 (SD 8.5) for NP. Screw density was 1.6 (SD 0.2). No significant loss of correction at FFU after was noted in both groups. The SRS-22r outcomes scores were 3.88 for PG and 4.02 for NP. Conclusion: The women who have undergone surgery for scoliosis and have had one or more pregnancies do not appear to have curve progression in scoliosis.

A105: Predictors of Extended Length of Hospital Stay in Adolescent Idiopathic Scoliosis Patients Undergoing Posterior Segmental Instrumented Fusion: An Analysis of 407 Surgeries Performed at a Large Academic Center

Assem Sultan¹, Ryan Berger¹, William Cantrell¹, Linsen Samuel¹, Deepak Ramanathan¹, Jessica Churchill¹, Anas Minkara¹, Joshua Golubovsky¹, Salam Bachour¹, Selena Pasadyn¹, Jaret Karnuta¹, Pierre Tamer¹, Thomas Kuivila¹, David Gurd¹, and Ryan Goodwin¹

¹Cleveland Clinic, Cleveland, OH, USA

Introduction: Reducing the length of hospital stay (LOS) and identifying risk factors associated with extended admission have become increasingly relevant to health care policy makers. There is currently limited research identifying risk factors that correlate with extended stay in patients undergoing posterior segmental instrumented fusion (PSIF) for adolescent idiopathic scoliosis (AIS). The aims of this study were to (1) compare patient-specific characteristics among those who had short versus long hospital stays and (2) identify independent risk factors that may correlate with extended LOS in AIS patients who underwent PSIF. Methods: We analyzed prospectively collected data from a single-institute, longitudinally maintained database of pediatric scoliosis patients who underwent surgical correction from January 1, 2009, to December 31, 2017. The database query returned 703 patients. To identify the study cohort, the following inclusion criteria were applied: (1) patients between 10 to 25 years of age, (2) who were diagnosed with adolescent idiopathic scoliosis, and (3) who underwent PSIF of ≥4 levels. Additionally, we excluded (1) patients with neuromuscular, congenital, and juvenile scoliosis; (2) patients who presented for revision surgery at our institution; and (3) patients who had inadequate radiographs. This yielded a study cohort of 407 patients that were included in our final analysis. Median LOS in the whole cohort was 4 days. Patients were divided into 2 groups, long lengths of stay (LLOS), defined as beyond 4 days, and those with short lengths of stay (SLOS), defined as 4 days or less. In both cohorts, patient demographics, comorbidities, preoperative scoliosis curve measurements, surgery-related characteristics, and complications were analyzed. Univariate and multivariate analyses were performed to identify independent risk factors associated with extended LOS. Results: Patients who had extended LOS tended to be female (84.6% vs 75%, P = .01), had more levels fused (9 ± 2 vs 7 ± 2 levels, P < .001), had more major postoperative complications (0.8% vs 7.4%, P = .002), had more blood loss during surgery (723 ± 548 vs 488 ± 341 cc, P < .001), and received less epidural analgesia (EA) for pain control (69% vs 89%, P < .001). Except for higher thoracic kyphosis, LLOS patients did not have worse preoperative radiographic curve parameters. Multivariate logistic analysis identified female gender, having ≥9 ± 2 levels of fusion, operative blood loss, major postoperative complications, lack of EA, and higher thoracic kyphosis as independent risk factors correlating for extended LOS. Conclusion: Independent risk factors identified by this study may be utilized to recognize AIS patients at risk of prolonged hospital stay. Important and clinically relevant factors that were identified as independent predictors of extended LOS are postoperative major complications, operative blood loss, having ≥9 ± 2 levels of fusion, and higher degrees of preoperative thoracic kyphosis. EA using opioid-only medications demonstrated a potentially protective effect. Several other factors such as comorbidities, preoperative curve Cobb’s angle, and the number of osteotomies performed during surgical correction did not show statistical difference among those who had short versus long LOS. Awareness of these factors will help surgeons and hospitals optimize perioperative care for factors that are unmodifiable and employ preventive measures for those that can be avoided or modified.

A106: Rib Distraction Devices: Do They Really Increase Lung Volume in Patients With Thoracic Spine Anomalies?

Nikita Khusainov¹, Sergey Vissarionov¹, and Dmitry Kokushin¹

¹ Department of Spine Pathology and Neurosurgery, Saint-Petersburg, Russian Federation

Introduction: Surgical techniques for the treatment of pediatric patients with multiple congenital spine anomalies vary from rib distraction devices to spinal correction with fusion. Surgery should correct the deformity and improve lung function and at the same time prevent the “domino effect” and development of thoracic insufficiency syndrome. Previously described positive effect of rib distraction devices on chest wall expansion and lung function improvement is still debatable as well as its ability to correct the deformity. We aimed to compare the efficacy of the 2 aforementioned techniques. Material and Methods: Thirteen patients were surgically treated during the period of 2016 to 2018. Patients in group 1 (n = 7) were treated with hemivertebrae excision and spinal correction with fusion, and patients in group 2 (n = 6) were treated with rib distraction devices. X-ray and computed tomography results were assessed preoperative and 1 year postoperative. Amount of scoliotic and kyphotic deformity correction was assessed using Cobb angle measurement. Lung volume on the convex side and total lung volume were measured using computed tomography data in tissue segmentation and separation mode. Results: Mean age at the beginning of treatment for patients in group 1 and group 2 was 3.6 ± 0.98 years and 5 ± 1.4 years, respectively. Before treatment in group 1 mean scoliotic curve was 39.7 ± 25.1° and kyphotic curve was 26.7 ± 4.6°; in group 2, mean scoliotic curve was 69.7 ± 21.7° and kyphotic curve 32.3 ± 13.5°. All patients had multiple thoracic spine anomalies and all patients in group 2 had rib synostosis on the concave side, which was primary indication for rib distraction device implantation. Total lung volume for groups 1 and 2 preoperatively was 972.8 ± 339.1 cc and 990.6 ± 399 cc, respectively. Lung volume on the convex side preoperatively was 465 ± 144.9 cc (47.8% from total volume) for group 1 and 479.7 ± 203.8 cc (48.4% from total volume) for group 2. Postoperatively in group 1 mean scoliotic curve was 14.4 ± 11.9° (P = .01), mean kyphotic curve was 21.7 ± 3.8° (P = .04); in group 2 mean scoliotic curve was 63.8 ± 16.3° (P > .05) and kyphotic curve 29.8 ± 12.2° (P > .05). Total lung volume for groups 1 and 2 postoperatively did not change significantly and was 972.8 ± 346.1 cc and 1237.2 ± 407.1 cc, respectively (P > .05). Lung volume on the convex side postoperatively also did not change and was 460.9 ± 126 cc (48% from total volume; P > .05) for group 1 and 585.6 ± 187.6 cc (48.4% from total volume) for group 2 (P > .05). Conclusion: Surgical treatment with rib distraction devices does not significantly increase lung volume in patients with multiple thoracic spine anomalies. Spinal correction and fusion provide significantly better curve correction and do not negatively alter lung volume and thus should be the method of choice when possible.

OP20: Degenerative-Cervical 3

A107: A Novel Anatomic Landmark to Assess Adequate Decompression in Anterior Cervical Spine Surgery: The Posterior Endplate Valley (PEV)

Ilyas Aleem¹, Jason Alder², Joseph Popper², Brett Freedman², Ahmad Nassr², Mohamad Bydon², and Bradford Currier²

¹University of Michigan, Ann Arbor, MI, USA

²Mayo Clinic, Rochester, MN, USA

Introduction: The cervical pedicle, which is the anatomic landmark defining the boundaries of the foramen, is hidden from view intraoperatively in the anterior cervical approach, potentially leading to incomplete foraminal decompression. An intraoperative landmark that heralds the location of the pedicle and therefore can be relied upon as a guide for decompression has not been previously described. The objective of this radiographic study was (1) to assess the reliability of using the posterior endplate valley (PEV) to predict the cranial-caudal location of the pedicle; (2) to characterize the pertinent surgical anatomy by measuring the interpedicular distance and foraminal height in the subaxial cervical spine; and (3) to assess the impact of age and degenerative disease on the PEV-pedicle relationship, interpedicular distance, and foraminal height. Material and Methods: We retrospectively reviewed cervical computed tomography (CT) images of younger (<50 years) and older (≥50 years) asymptomatic patients presenting to our institution. Patients with previous cervical surgery or trauma of the cervical spine were excluded. Using the coronal reconstructed image taken at the posterior margin of the vertebral body, we constructed a line between the superior aspect of the pedicles and measured the distance from this line to the PEV. We defined measurements as positive if the PEV was superior to the pedicular line and negative if inferior to the line. Interpedicular distance and foraminal height were also measured. Statistical comparisons between groups were made using t tests. Results: Fifty patients with a mean age of 37.2 years and 50 patients with a mean age of 75.6 years were included in the final analysis for a total of 100 patients. The mean distance from the pedicular line to the PEV from C3-7, respectively, was 1.0 ± 0.99 mm, 0.01 ± 0.76 mm, −0.09 ± 0.70 mm, −0.20 ± 0.71 mm, and −0.27 ± 0.79 mm. No significant difference between young and elderly patients was noted (P < .05) with respect to this relationship. Intervertebral foraminal size was significantly greater in younger compared to elderly patients at all levels except C2-3. The mean interpedicular distance of included patients was 23.05 ± 1.76 mm. Conclusion: This study demonstrates, for the first time, that the PEV is an accurate surgical landmark that is consistently at most 1 mm from the superior aspect of the cervical pedicle in the subaxial spine.

A108: Risk Factors of Stand-Alone ACDF Cage Subsidence

Melanie Schindler¹, Anna Rienmüller¹, Yu-Mi ryang², Petra Krepler¹, Reinhard Windhager¹, and Georg Grohs¹

¹Medical University Vienna, Vienna, Austria

²Technical University Munich, Munich, Germany

Introduction: Anterior cervical discectomy and fusion (ACDF) is a well-established method for the treatment of cervical spine pathologies such as foraminal and/or central stenosis. Although this procedure is commonly performed and well recognized, complications such as cage subsidence are frequent and possibly lead to revision surgery. In this retrospective study, we analyzed various possible risk factors leading to cage subsidence after ACDF. Material and Methods: We performed a retrospective analysis of all patients who underwent primary ACDF at our institution between 1997 and 2017. A total of 64 patients (31 females, 33 males) with a mean age of 58.8 ± 12.27 years and 93 cages implanted were included. The mean follow-up was 24 ± 6 months. All preoperative and postoperative as well as follow-up lateral and anteroposterior X-rays of the cervical spine were examined for cage subsidence. A decrease of >3 mm was considered as cage subsidence. Additionally, parameters of the sagittal balance such as focal and global lordosis (C2-C7), sagittal vertical axis, and T1 slope were measured on postoperative lateral X-rays. In a logistic regression analysis, we included age, type of cage (screw-fixation or peek), number of cages, preoperative diagnosis including osteoporosis and diabetes, focal and global lordosis (C2-C7), sagittal vertical axis, and T1 slope as possible influencing factors for cage subsidence. Results: We found an overall rate of stand-alone cage subsidence of 34%. Higher cage subsidence was found in peek implants than in screw fixated implants. The binary logistic regression exhibited a significant correlation between cage subsidence and osteoporosis (odds ratio OR 3.4; P = .045), as well as between focal lordosis and cage subsidence (OR .88; P = .0318), and T1 slope and cage subsidence (OR 1.07; P = .07). The multifactorial analysis only revealed postoperative focal lordosis and T1 slope as significant factor influencing cage subsidence. Conclusion: Preoperatively existing diagnosis of osteoporosis and local parameters of the sagittal balance such as focal lordosis and a higher T1 slope presented as significant factors influencing cages subsidence after stand-alone ACDF. We therefore suggest including those factors in preoperative planning, such as adequate treatment of osteoporosis and choice of implant/technique to restore focal lordosis.

A109: Comparison of the Outcomes of Stand-Alone Cage With Cage-Plate Fixation in 1- and 2-Level ACDF for Treating Cervical Disk Disease

Mohammadreza Etemadifar¹, Ali Andalib¹, and Hossein Shafiei¹

¹Isfahan University of Medical Sciences, Esfahan, Islamic Republic of Iran

Introduction: Anterior cervical discectomy and fusion (ACDF) is an accepted method for treating degenerative cervical spine disease. The purpose of this study was to evaluate whether adding plate to interbody cages changes the outcome of treatment in patients with degenerative cervical spine disease at 1- or 2-level ACDF. Method: In this prospective cross-sectional descriptive study, 80 patients undergoing ACDF were enrolled according to the inclusion and exclusion criteria. Demographic data, before and after surgery, and clinical symptoms were investigated. Data were collected by means of Visual Analogue Scale (VAS) and Neck Disability Index (NDI) questionnaires. The adverse effects and surgical outcomes were evaluated based on Odom’s criteria and patients’ satisfaction. The collected data of the 2 groups were then compared and assessed. Results: There was no significant difference between the groups with regard to gender, age, duration of surgery to visit, surgical level, preoperative and postoperative VAS score, preoperative NDI score, results based on Odom’s criteria, and satisfaction of patients (P > .05). The VAS and NDI scores were significantly reduced in the 4 groups after the operation compared to the preoperative stage. Postoperative NDI scores in the ACDF with cage-stand-alone group at 1 level were significantly lower than other groups (P < .05). Conclusion: Both the methods revealed acceptable outcomes in comparison to the preoperative stage, and despite some minor differences, there are generally no significant differences in outcomes and complications.

A110: Relationship Between Decompression Range and Decompression Limitation of Cervical Laminoplasty

Yinze Diao¹, Yu Sun¹, Shaobo Wang¹, Fengshan Zhang¹, Li Zhang¹, and Shengfa Pan¹

¹Peking University Third Hospital, Beijing, China

Introduction: Posterior cervical spine surgery is an indirect decompression technique that relieves anterior compression by backward shifting of spinal cord, which determines the limitation of decompression and is affected by the range of decompression. However, in previous studies, there is no quantitative analysis of the relationship between the decompression range and limitation. The purpose of this study is to reveal the quantitative relationship between the 2 variables and to provide a basis for the rational selection of decompression range. Material and Methods: A total of 129 cases of open-door laminoplasty were reviewed, which were divided into groups C4-7 (n = 11), C3-6 (n = 61), C3-7 (n = 32), and C2-7 (n = 25). The cervical curvature of each group was measured and compared between groups. Anterior cord distance (ACD) of each level from C2-3 to C6-7 was measured as the decompression limit at each level of each group. According to the relationship between location of a level and decompression range, a given level can be defined as external decompression margin segment (EDMS; located at the decompression margin outside the decompression range, underwent a extramarginal decompression), internal decompression margin segment (IDMS; located at the decompression margin inside the decompression range, underwent a intramarginal decompression), and central decompression segment (CDS; not only the adjacent lamina but also distant lamina was covered by the decompression range, underwent a central decompression) (eg, for decompression range of C4-7, C3-6, C3-7, and C2-7, the level C3-4 is: EDMS, IDMS, IDMS, and CDS, respectively). ACDs at the same level of each group were compared to find the difference of decompression limits between groups. Results: There was no significant difference in cervical curvature between the groups. At level C2-3, the mean value of ACD of groups C4-7, C3-6, C3-7, and C2-7 was 5.49 mm, 5.89 mm, 6.13 mm, and 8.58 mm, respectively, and there were significant differences between C2-7 and the other 3 groups. At level C3-4, ACD was 4.90 mm, 6.58 mm, 6.60 mm, and 8.74 mm, respectively. Groups C4-7 and C2-7 both have significant differences compared to the other 3 groups. At level C4-5, ACD was 6.04 mm, 7.47 mm, 7.59 mm, and 8.42 mm, respectively. There was significant difference between group C4-7 and the other 3 groups. At levels C5-6 and C6-7, there was significant difference between group C3-6 and the other 3 groups. Conclusion: For a given cervical spine level, compared under different decompression range, the relationship among decompression limits is as follows: extramarginal decompression < intramarginal decompression < central decompression, and the differences of decompression limits are significant. Central decompression is the maximum decompression of a given level, and no greater decompression effect will be achieved by expanding the range of decompression. A reasonable range of decompression should be determined based on the relationship between the compression and the effect available with different decompression schemes.

A111: The Effect of Aging on Cervical Sagittal Balance Parameters

Ivan Ye¹, Ray Tang¹, Zoe Cheung¹, Samuel White¹, Jun Kim¹, and Samuel Cho¹

¹Mount Sinai Health System, New York, NY, USA

Introduction: Cervical sagittal balance is essential for maintaining neutral head posture and horizontal gaze. However, degenerative changes associated with increasing age can result in alterations in the cervical spine that lead to sagittal imbalance. Cervical sagittal imbalance has been associated with neck pain, headache, fatigue, and disability. The objective of this study was to examine the natural course of age-related changes in the sagittal profile of the cervical spine by identifying correlations between age and cervical sagittal balance parameters using standard radiographs. Material and Methods: This was a retrospective radiographic study of cervical spine radiographs taken between December 2017 and June 2018 at a single institution. Only radiographs with visible T1 superior endplates were included. Radiographs with cervical instrumentation were excluded. The relationship between age and cervical sagittal balance parameters was determined using Pearson’s correlation coefficient. Cervical sagittal balance parameters that were examined included C1 inclination, C2 slope, C2-C7 Cobb angle, cervical sagittal vertical axis (cSVA), cervical tilt, upper C7 slope, lower C7 slope, T1 slope, and T1 slope minus cervical lordosis (T1S-CL). Patients were then divided into cohorts according to age (Group 1: 18-40 years old; Group 2: 40-65 years old; Group 3: >65 years old). An analysis of variance (ANOVA) was performed to determine the presence of any differences in cervical sagittal balance parameters among the 3 age groups. Tukey adjustments were then used to identify which age group was responsible for the observed difference. Results: A total of 151 patients were included in this study. Age was weakly correlated with C2-C7 Cobb angle (r = 0.231; P = .004), upper C7 slope (r = 0.280; P < .001), lower C7 slope (r = 0.283; P < .001), and T1 slope (r = 0.189; P = .020). The ANOVA found significant differences among the 3 age groups in terms of C2-C7 Cobb angle (Group 1: 5.8 ± 12.9°; Group 2: 7.3 ± 10.2°; Group 3: 14.8 ± 10.5°; P = .002), upper C7 slope (Group 1: 20.9 ± 9.6°; Group 2: 22.2 ± 7.2°; Group 3: 28.9 ± 10.8°; P < .001), lower C7 slope (Group 1: 21.1 ± 9.2°; Group 2: 21.3 ± 7.4°; Group 3: 28.5 ± 10.5°; P < .001), and T1 slope (Group 1: 25.6 ± 9.4°; Group 2: 26.9 ± 7.3°; Group 3: 30.9 ± 10.7° P = .031). Conclusion: In this radiographic study of 151 patients, there were weak associations between age and cervical sagittal balance parameters—namely, C2-C7 Cobb angle, upper C7 slope, lower C7 slope, and T1 slope. C2-C7 Cobb angle, upper C7 slope, lower C7 slope, and T1 slope all demonstrated age-related changes with older patients having higher C2-C7 Cobb angles, as well as higher C7 and T1 slopes. The remaining cervical sagittal balance parameters remained relatively constant with age.

A112: Does Extension Dysfunction Affect Postoperative Loss of Cervical Lordosis in Patients Who Undergo Laminoplasty?

Dong Wuk Son¹, Geun Sung Song¹, Sang Weon Lee¹, Soon Ki Sung¹, Jun Seok Lee¹, and Doo Kyung Son¹

¹Pusan National University Yangsan Hospital, Pusan, Republic of Korea

Introduction: Recent studies have focused on the correlation between loss of cervical lordosis (LCL) after laminoplasty and T1 slope, but it has a controversy. We focused on the role of extension function (EF) in explaining the less LCL in high T1 slope patients after laminoplasty. Material and Methods: We retrospectively analyzed 50 consecutive patients who underwent open-door laminoplasty (>1-year follow-up). EF is defined as extension C2-7 Cobb’s angle (CA) minus neutral C2-7 CA. LCL is defined as follow-up CA minus preoperative CA [CA (FU) − CA (PRE)], and significant kyphotic change was defined as LCL < −10°. Results: The distribution of LCL was −3.70 ± 7.98 and the significant kyphotic change occurred in 20% of the patients (10/50). Correlation analysis revealed that LCL was not related to the CA (PRE) and C7 slope (PRE). EF, C2-7 sagittal vertical axis (PRE), and C2 slope (PRE) were found to be risk factors for LCL by multiple linear regression analysis. The receiver operating characteristic curve analysis revealed that EF could predict the significant kyphotic change well. The cutoff value of EF was 14°. All significant kyphotic change occurred at EF < 14°. On limiting the number of patients with preoperative straight curvature (n = 28), no significant kyphotic change occurred in any patient whose EF was ≥14°. Conclusion: We have identified a new factor, EF, which could predict LCL after laminoplasty. All significant kyphotic changes after laminoplasty occurred particularly when the EF was <14°. In addition, no significant kyphotic change occurred in patients with straight curvature when the EF was ≥14°.

OP21: Novel Technologies 1

A113: Superior Predictability of Incidental Osteoporotic Vertebral Fractures of Opportunistic QCT Compared to Reference DEXA

Yu-Mi Ryang¹, Maximillain Loeffler², Anna Rienmüller³, Bernhard Meyer², and Jan S. Kirschke²

¹Helios Klinikum Berlin-Buch, Berlin, Germany

²Klinikum rechts der Isar, Technical University Munich, Munich, Germany

³University Hospital Vienna, Vienna, Austria

Introduction: Osteoporotic vertebral compression fractures (OF) are typically incidental or occur after low-energy trauma and are associated with a severe health economic burden. Opportunistic quantitative computed tomography (QCT) seems to be a feasible alternative to dual-energy X-ray absorptiometry (DEXA) in osteoporosis screening. The aim of this study was to compare central DEXA with opportunistic QCT of the lumbar spine in the risk assessment and predictability of OF. Material and Methods: Patient records of 326 consecutive patients ≥50 of age years were reviewed for lumbar CT (baseline) and DEXA within a 12-month period (from 2008 to 2017); patients with spinal tumors were excluded. We included 86 patients with follow-up imaging after at least 12 months or who sustained an incidental OF documented in earlier imaging. Fracture status and trabecular bone mineral density (BMD) by asynchronous QCT were retrospectively evaluated in baseline CT and fracture status was reassessed at follow-up. Two Cox proportional-hazards models were used to assess the risk of OF with BMD (QCT) and t-scores (DEXA) each with age and sex as covariates. The area under the ROC curve (AUC) for BMD and t-scores were calculated for the prediction of incidental OF. Results: After a mean follow-up of 26.3 months, 23 patients had a new or progressive OF showing significantly lower mean BMD values than patient without incidental OVCF [osteoporotic vertebral compression fracture] (P < .001). The risk of OVCF increased substantially per SD decrease in BMD (hazard ratio [HR] = 2.79, P < .001). There was no significant difference in t-scores between patients with and without OF in the DEXA group (P < .145); the association with fracture risk was weaker for DEXA t-scores (HR = 1.55, P = .013) with a similarly significant gain in risk per year increase in age (HR = 1.08, P = .007). The predictability of incidental OF was good with BMD (AUC = 0.793, 95% confidence interval [CI] = 0.69-0.897) and nonsignificant for DEXA (AUC = 0.601, CI = 0.47-0.733). Conclusion: We found a higher association and better predictability of incidental OF with BMD measured by opportunistic QCT compared to central DEXA t-scores. Therefore, QCT seems to be a simple and low-cost alternative to the standard method DEXA for osteoporosis screening with a better predictive value for OF.

A114: Proteome Profiling in Human Intervertebral Disc: A Comparison Between Herniated and Healthy Disc Samples

Shanmuganathan Rajasekaran¹, K. S. Srivijay Anand¹, Ajoy Prasad Shetty¹, and Rishi Mugesh Kanna¹

¹Ganga Hospital, Coimbatore, India

Introduction: Intervertebral disc (IVD) pathology is the commonest cause of low back pain and yet its etiopathogenesis remains incompletely understood. It has been probed from different angles of genetic predisposition, mechanical etiology, trauma, and subclinical infection, yet there is no complete understanding. It is important that they are analyzed at the molecular level for a better understanding. This experimental study was designed to understand the proteome of nucleus pulposus (NP) tissue in normal disc samples and appreciate its differential expression in disc herniation (DH). Materials and Methods: Eighteen disc samples in DH group and 4 MRI (magnetic resonance imaging) normal discs in Control group (2 Scoliosis, 2 Organ donors) were harvested under sterile conditions. Protein extraction, precipitation, and LC-MS/MS (liquid chromatography–tandem mass spectrometry) analysis were done using established standard methods. The entire proteome map and alteration in protein expressions were further analyzed using the Database for Annotation, Visualization and Integrated Discovery (DAVID) version6.8. Gene Functional Classification Tool in DAVID database was used for cluster analysis. Results: We identified 2086 proteins in DH and 651 proteins in NG. A total of 1905 (74.5%) of the proteins were exclusive to DH whereas 470 proteins (18.4%) were exclusive to NG and 181 (7.1%) proteins were common. Of the common proteins, 45% belonged to glycoproteins category, followed by extracellular matrix (ECM) (30%), proteoglycans (7%), protease inhibitor (6%), serine protease inhibitor (5%), and leucine-rich repeat proteins (7%). GO analysis revealed the antibacterial humoral response proteins and defense response to Gram-positive bacterium with highly significant P value of 5.31E-05 and 3.77E-04, respectively. Cluster analysis of common proteins revealed 6 clusters. Cluster I consisted cartilage intermediate layer proteins (CILP) and proteoglycan link proteins CILP2, HAPLN1, HAPLN3, and A1BG involved in immune response cartilage scaffolding. Cluster II was serpin protease inhibitors (SERPINs) that play a vital role in inflammatory and coagulation cascades. Cluster III had small leucine-rich proteoglycans (SLRPs), which interact with each other and form the core of ECM. Cluster IV had intermediate filaments (IFs) of varied molecular functions including ageing, cell proliferation, and protein folding. Cluster V had collagen, which is involved in cell adhesion, cross bridging, has cytoprotective effect by counteracting apoptosis, and oxidative damage. Cluster VI had matricellular proteins implicated in complement activation. Interestingly, we identified 2 clusters that were very specific/unique to DH: complementary protein cluster and immunoglobin cluster. Immunoglobulin heavy constant mu (IGHM), immunoglobulin heavy constant gamma 3 (IGHG3), immunoglobulin kappa variable 3-20 (IGKV3 200) and immunoglobulin kappa variable 4-1 (IGKV4-1) were detected only in DH. They represent either an adaptive immune response or antibacterial humoral response and promote leukocyte migration and phagocytosis. Conclusion: Proteomic analysis of human IVD shows a clear evidence of a pro-inflammatory state in DH. The cluster analysis of proteins revealed novel markers having potential for targeted therapies. AIBG, SERPIN D1, SERPINA3 along with ITIH2, which are acute phase proteins, were upregulated in DH indicating an inflammatory response. The complement activation cascades and microbial enrichment pathways in DH raise the suspicion of infection, which warrants further probing.

A115: Evaluation of the Basal Metabolism of Degenerative Lumbar Intervertebral Discs Based on Pfirmann Grade and Presence of Instability

Chaiyapruk Pundee¹, Naomi N. Lee¹, Jacob S. Kramer¹, Aaron M. Stoker¹, Aaron M. Stoker¹, Don K. Moore¹, Theodore J Choma¹, and James L. Cook¹

¹ University of Missouri, Columbia, MO, USA

Introduction: Intervertebral disc (IVD) degeneration is implicated in back pain, a leading cause of spine-related disability. It may also cause segmental instability. Both local and systemic inflammatory processes have been implicated in the pathophysiology of IVD degeneration, though it is unclear if Pfirmann grade severity or presence of instability affects degenerative IVD metabolism. These differences have not previously been investigated. Hence, this study was designed to examine the basal metabolism of degenerative lumbar IVD tissues collected from patients undergoing lumbar microdiscectomy or fusion with the hypothesis that tissues from patients with higher Pfirmann grades and those with radiologic evidence of instability would produce significantly higher levels of degradative enzymes and inflammatory mediators. Material and Methods: Tissue Collection and Culture: With IRB approval (IRB#2 010 692) and informed patient consent, degenerative IVD tissue excised as part of a standard-of-care spinal surgery, which would otherwise be discarded, was obtained from patients (n = 7, mean age 53, 4 female) being treated for symptomatic degenerative lumbar IVD disorders. Pfirmann grading (0-4) of the IVD was determined by evaluation of MRI (magnetic resonance imaging) by a single reviewer (grade 3 n = 3, grade 4 n = 4), and for instability as defined by the presence of a spondylolisthesis on upright lateral radiographs by 2 reviewers (unstable n = 2, stable n = 5). Tissues composed of degenerative nucleus pulposus (NP) and annulus fibrosus (AF) were collected and explants of combined NP and AF (based on gross assessment) were created with a 6 mm diameter biopsy punch. Two explants per disc were cultured for 3 days, after which media were collected for biomarker evaluation. Media Analyses: Media were tested for MMP-1, MMP-2, MMP-3, MMP-7, MMP-8, MMP-9, MMP-13, TIMP-1, TIMP-2, TIMP-3, TIMP-4, GRO-α, MCP-3, PDGF-AA, PDGF-AB/BB, IL-2, IL-4, IL-6, IL-8, MCP-1, MIP-1α, MIP-1β, RANTES, TNF-α, and VEGF using commercially available assays according to the manufacturer’s protocol. Statistical Analysis: Significant differences between groups were determined by t-test or rank sum test based on normality of the data using SPSS with significance set at P ≤ .05. Results: With the small number of samples analyzed, we did not identify any significant differences between basal biomarker production by degenerative IVD tissues of Pfirmann grade 3 versus 4 discs. However, tissues from degenerative IVDs with instability were found to produce significantly higher levels of MMP-8, and significantly lower levels of MMP-1, MMP-2, and MMP-13 compared to tissues from stable degenerative IVDs. Conclusion: The data from this study indicate that there is not a significant difference in the metabolism of IVD tissues with a Pfirmann grade of 3 or 4. However, the presence of instability in the degenerated lumbar IVD resulted in a significant increase in the production of MMP-8, but the production of other degradative enzymes was significantly decreased. Further study is required to determine how observable changes in IVD architecture relates to changes in tissue metabolism, as understanding how gross findings of IVD degeneration relate to its pathophysiology may provide insight into potential therapies for prevention and treatment of radiologic findings, as well as the associated symptoms of, lumbar intervertebral disc degeneration.

A116: Intraoperative CT-Based Navigation Is More Accurate and Safer Way of Lower Thoracic and Lumbar Pedicle Screw Insertion Than Traditional Fluoroscopy Guidance

Pavel Barsa¹, Robert Frohlich¹, and Petr Suchomel¹

¹Regional Hospital of Liberec, Liberec, Czech Republic

Introduction: The aim of this prospective randomized study was to compare accuracy of conventional fluoroscopy guided way of pedicle screw insertion with navigation based on intraoperative computed tomography (CT) imaging. Material and Methods: A group of 200 consecutive patients scheduled for posterior pedicle screw stabilization in the region of T11-S1 was equally distributed into 2 groups: either for conventional fluoroscopy-guided pedicle screw insertion or to the group where intraoperative CT scans obtained by movable 32-slice CT scanner served as the base for trajectory planning and subsequent screw insertion guided by the optoelectronic navigation system. The accuracy of screw insertion had been verified by intra- or postoperative transaxial CT images. Incorrect screw position defined as any breaching of pedicle or vertebral body cortex by the screw distally to the entry point was assessed by the independent radiologist. Intraoperative adverse events and postoperative clinical sequelae related to an incorrect screw insertion as well as size of inserted screws were recorded and results statistically analyzed. Results: In the subgroup of 100 patients where traditional fluoroscopy guidance has been used during 542 pedicle screw insertions, 53 screws (9.78%) met the criteria for incorrect position. The incidence of screw malpositioning in this subgroup was significantly higher than in patients where CT-based navigation has been utilized. In the latter 100 patients subgroup, out of 572 navigated screws only 5 (0.87%) penetrated vertebral cortex distally to the entry point. With the exception of 3 intraoperatively replaced screws in the fluoroscopically guided subgroup, all the cortical breaches in both subgroups were clinically irrelevant without the need for postoperative repositioning. The largest group of screws inserted with the help of CT-based navigation represented screws of 7 mm diameter (47.6%). The size of screws in fluoroscopy guided subgroup was significantly smaller with majority (83.8%) of 6 mm screws. Conclusion: Intraoperative CT scan-based navigation seems to be an effective way of instrumentation guidance in lower thoracic and lumbar pedicle screws. Significantly higher accuracy of screw insertion as well as larger size of implants compared to traditional fluoroscopy guidance documents benefits of intraoperative 3D visualization and confident grasp over surgical anatomy. The precision and safety of implant insertion is the precondition enabling use of this technique also in anatomically more complex segments of the spine.

A117: Fusion Performance of Attrax® Putty Versus Autograft in Instrumented Posterolateral Spinal Fusion: A Randomized Intrapatient Controlled Noninferiority Trial

A. M. Lehr¹, F. C. Oner¹, E. A. Hoebink², D. H. R. Kempen³, J. L. C. van Susante⁴, R. M. Castelein¹, and M. C. Kruyt¹

¹University Medical Center Utrecht, Utrecht, Netherlands

² Amphia Hospital, Breda, Netherlands

³OLVG, Amsterdam, Netherlands

⁴Rijnstate Hospital, Arnhem, Netherlands

Introduction: Spinal fusion is a frequently performed surgical procedure for many spinal conditions. Autologous bone grafting is the gold standard to establish a bony fusion, but this procedure has some drawbacks including graft harvesting morbidity and limited amounts. This has led to development of various alternatives over the past decades including synthetic ceramics. Current third-generation microporous ceramics are promising due to the high bioactivity as shown by ectopic bone induction in many animal models. This study aimed to demonstrate noninferiority of AttraX® Putty (NuVasive Inc, USA), a composite of β-tricalcium phosphate and hydroxyapatite, in comparison to autograft in instrumented posterolateral spinal fusions. Material and Methods: After obtaining medical ethical approval, 100 nontraumatic adult patients indicated for primary instrumented posterolateral fusion between T10 and S1 were included in this multicenter, randomized, intrapatient controlled, noninferiority trial. After instrumentation and preparation for bone grafting, the allocation side of AttraX® Putty was disclosed. The contralateral side of the fusion trajectory was grafted with autologous bone graft. Thereby each patient served as its own control. As primary outcome, the fusion performance of both AttraX® Putty and autograft was assessed at 1-year follow-up on computed tomography (CT) scans. Each segment and side was scored as fused, doubtful fusion, or nonunion by 2 blinded observers independently. The noninferiority margin was set at 15%, with 80% power and α of 0.05. The rate of successful posterolateral fusions was used and tested for noninferiority with McNemar’s test. Secondary outcomes included the rate of (serious) adverse events compared to control populations from literature, Oswestry Disability Index (ODI), EQ-5D-5L, and Visual Analogue Scale (VAS) for back pain. Results: A total of 87 patients were included in the primary analysis. There were 42 males and 45 females with a mean age of 55 (range 33-79) years. A mean of 1.7 (range 1-8) spinal segments were instrumented for fusion. The overall posterolateral fusion rate was 71%. At the AttraX® Putty side 55% of the segments were scored as fused, compared to 52% at the autograft side. After correction for multi-level fusions resulting in a single fusion score per side, the observed difference between the treatment conditions was 2.6% with a 90% confidence interval of −9.1% to +13.7% (McNemar’s test P = .868). None of the reported adverse events could be directly related to the use of AttraX® Putty. The ODI score improved from mean 43% (±17) at baseline to mean 21% (±19) at 1-year follow-up. Conclusion: This randomized intrapatient controlled study including 87 patients demonstrated noninferiority of AttraX® Putty in comparison to autograft in terms of posterolateral fusion performance in instrumented thoracolumbar spinal fusions.

A118: Quantitative Analysis of Indirect Decompression in Extreme Lateral Interbody Fusion and Posterior Spinal Fusion With a Percutaneous Pedicle Screw System for Lumbar Spinal Stenosis

Hiroshi Nomura¹, Yamashita Akihisa², Tetsuya Watanabe², and Kenzo Shirasawa²

¹ Nomura Orthopaedic Clinic, Shimonoseki, Japan

²Shimonoseki City Hospital, Shimonoseki, Japan

Introduction: Recently, extreme lateral interbody fusion (XLIF) has been increasingly utilized to treat lumbar spinal canal stenosis (LSS) with or without spondylolisthesis to lead indirect decompression. However, distinct mechanism of the indirect decompression with XLIF is still unknown. The authors sought to quantify the results of clinical and radiological analyses of XLIF plus percutaneous pedicle screw fixation (PPS) for patients with LSS by focusing on the distinct mechanism of indirect decompression. Methods: Data obtained from a total of 37 patients with 47 surgical sites were retrospectively analyzed. Clinical outcomes for all patients were evaluated using the Japanese Orthopaedic Association (JOA) score and the improvement rate of the JOA score. Preoperative and postoperative MRI (magnetic resonance imaging) and CT (computed tomography) were used to measure the transverse area of both the dural sac (DS area), ligamentous flavum (LF area), and transverse bony area of the spinal canal (SC area) in the axial sections and the length of the intervertebral disc bulge (DB length) in sagittal sections at the individual surgical levels. Then, the rate of change (RC) of the DS area (RC-DS), the RC of the LF area (RC-LF), the RC of the SC area (RC-SC), and the RC of the DB length (RC-DB) from the preoperative period to the postoperative period were calculated. Results: Preoperative clinical symptoms improved significantly after surgery for all patients regardless of whether the RC-DS was large or small. The RC-DS, RC-LF, RC-SC, and RC-DB were approximately 203%, 74%, 111%, and 37%, respectively. Next, we divided all surgical sites into the small expansion group (SE group; RC-DS <150%) and large expansion group (LE group; RC-DS ≥150%) according to the degree of RC-DS and found that the bulging was significantly shorter in the LE group than in the SE group, although there was no difference in the RC-LF between the LE group and SE group. Conclusions: We found that XLIF plus PPS fixation led to excellent clinical improvements in patients with LSS. We suggest that indirect decompression after XLIF is influenced by the degree of reduction in DB rather than by thinning of the LF, although the grade of the indirect decompression was not related to the clinical outcome.

A119: Percutaneous Occipitocervical Fusion: Optimal Choice to Obtain Best Cervical Pedicles Screws Accuracy

Wojciech Kloc¹, Marek Roclawski², Rafal Pankowski², Marek Derenda³, Patryk Kurlandt³, Jakub Wisniewski³, Katarzyna Aleksandrowicz³, and Stanislaw Adamski³

¹University of Warmia and Mazury, Olsztyn, Poland

²Medical University, Gdansk, Poland

³Copernicus Medical Centre, Gdansk, Poland

Introduction: Occipitocervical fusion (OCF) is an effective surgical method to treat various craniovertebral junction (CVJ) pathologies. A rigid fixation achieved from OCF displaces other techniques of CVJ stabilization unfortunately when deep and wide wound procedure is performed. The aim of this study is to share our experience in OCF and lately performed percutaneous OCFs with intraoperative CT (computed tomography)-guided navigation system. Material and Methods: Of 31 patients who underwent OCF, 3 were performed percutaneously. O-arm CT scans were used to illustrate and measure radiologic parameters. Screws were implanted in C1 lateral masses (2), isthmus of C2 (62), and C3 pedicles (62) and assessed according Gertzbein Robbins (GR) classification from A to E. Results: A total 126 screws were implanted; 114 of them was performed in open surgery and 12 percutaneously. Outcome in GR classification for screws implanted in open surgery was as follows: A 58 (50.88%), B 22 (19.3%), C 16 (14.04%), D 10 (8.77%), and E 9 (7.02%), while in percutaneous it was A 11 (91.67%) and B 1 (8.33%). In open surgery one screw was revised. Conclusion: Percutaneous occipitocervical fusion seems to be a good option to achieve desirable effect in cervical pedicle screws implantation. During the procedure whole nuchal muscles are preserved. CT-guided surgery and microscope view are necessary to perform percutaneous OCF.

A120: Therapeutic Potential of a Novel Small Molecule in Modifying Lumbar Disc Degeneration

Yi Sun¹, Kenny Kwan¹, Yuching Hsu¹, Ellen Zhang¹, Kenneth Cheung¹, and Victor Leung¹

¹University of Hong Kong, Hong Kong

Introduction: Low back pain, one of the most debilitating medical conditions worldwide, is attributed to degeneration of intervertebral disc (IVD). IVD degeneration is associated with a reduction of proteoglycans in the nucleus pulposus (NP) and ultimately a replacement by fibrocartilaginous matrix, causing a loss of mechanical strength and shock-absorbing capacity. By a high-throughput screening, we previously identified a novel small molecule, termed as CSG1, that is capable of recovering chondroitin sulfate proteoglycan production in degenerative human NP cells and modifying injury-induced mouse disc degeneration. Yet the efficacy of this lead small molecule in large animal models remains elusive. Goat discs have similar anatomy and geometry to, and therefore resemble the nutritional environment and motion segment mechanics of the human lumbar discs. Moreover, the NP cell composition is comparable to that in human. Lumbar disc degeneration in goats induced by intradiscal injection of chondroitinase ABC (ChABC) features a progressive loss of proteoglycan and recapitulates other hallmarks of human disc degeneration. Here we hypothesize that local delivery of CSG1 can modify induced lumbar disc degeneration. We aim to investigate if CSG1 can ameliorate NP matrix breakdown and disc height loss in a ChABC-mediated goat IVD degeneration. Material and Methods: Surgery protocols were approved by local ethics committee. A pilot study of intradiscal injection of safranin-O and Alcian blue stain was performed to validate the delivery efficiency. After the validation, L2/3 and L4/5 discs of 4 goats (n = 8 discs) were randomized to receive intradiscal injection of chondroitinase ABC (0.5 U/mL, 200 μL) via a standard fluoroscopically guided posterior approach using a 27 G needle. At 1-month post-injection, the small molecule CSG1 (200 nM) or DMSO (0.1%) carrier was delivered into the operated discs in the same approach. Degeneration progression was assessed longitudinally by radiography and disc height index (DHI) was determined. The discs were harvested at 6-month post-treatment for histology. NP matrix changes were examined by FAST staining. Assuming Gaussian distribution, data were analyzed by 2-way ANOVA with Bonferroni posttest, and P values were calculated to determine statistical differences. Results: Pilot study with safranin-O (at L2/3) and Alcian blue (at L4/5) injection showed that the delivery was well confined in the NP and inner annulus regions (Figure 1). ChABC induced 30% loss of DHI by 4 months post-injection (Figure 2). CSG1 injection resulted in a DHI almost comparable to the unoperated discs after 3 months (P = .183; Figure 2), suggesting a preservation of disc space. Moreover, CSG1-injected NP showed minimal matrix changes with preserved proteoglycans content (data not shown). Osteophyte in vertebral bone were observed in one motion segment level after ChABC injection (Goat #1; Figure 2). Conclusion: In this study, we demonstrated the function of CSG1 in preventing the loss of disc space and NP matrix in the degeneration progression. The vertebral bone changes after ChABC injection may imply a function of chondroitin sulfate-proteoglycans in bone metabolism. Our findings validated the therapeutic potential of CSG1 in treating IVD degeneration and lay an important foundation for future clinical trials.

A121: No Improvement in Pedicle Screw Accuracy and Increased Radiation With Intraoperative CT-Based Navigation Compared to Free Hand Technique in Idiopathic Scoliosis Surgery

Wiktor Urbanski¹, Rafal Zaluski¹, and Piotr Morasiewicz¹

¹University Hospital Wroclaw, Wroclaw, Poland

Introduction: Pedicle screw insertion in spinal deformities is demanding and if inappropriately placed may result in complications. Navigation systems based on intraoperatively obtained 3-dimensional images were developed to prevent or minimize pedicle screws misplacements. However, the method requires significant amount of radiation during procedure; moreover, there is a lack of data confirming superiority of navigation above other techniques. The objectives of this study were to compare accuracy of 2 methods of pedicle screws placement: intraoperative 3D image-based navigation and a freehand technique in patients with idiopathic scoliosis, as well as to evaluate the radiation received by patients in both methods. Methods: In 49 patients who underwent posterior spinal fusion with all pedicle screw constructs for idiopathic scoliosis, 835 pedicle screws were inserted by 2 surgeons. Consecutive patients had screws inserted with freehand technique and with use of navigation, resulting in 451 pedicle screws positioned with navigation and 384 freehanded. Two observers not involved in the treatment evaluated the position of the screws. The pedicle breach was assessed on CT (computed tomography) scans in sagittal and horizontal views according to the following grading system: grade 0 no pedicle wall violation, grade 1 perforation ≤2 mm, grade 2 perforation 2 to 4 mm, grade 3 perforation of >4 mm. Grades 0 and 1 were considered as a properly positioned; grades 2 and 3 represented malposition. The radiation doses from o-arm and c-arm used throughout the procedure were collected. Results: No differences were found in accuracy between the freehand and navigated groups, 96.8% freehand and 95.8% in the navigation group, respectively, of properly positioned pedicle screws (P = .518). Grade 3 pedicle screws were observed only in the freehand group and were all located in the upper thoracic spine. Patients undergoing navigated pedicle screw placement received significantly greater mean radiation dose than those whose screws were placed freehand (1070.5 ± 447 mGycm vs 390.5 ± 53 mGycm, P < .001). Conclusions: We did not observe better accuracy of pedicle screw placement with CT-based navigation in patients with moderate idiopathic scoliosis undergoing primary surgery, but the patients underwent significantly greater exposure to radiation. Therefore, caution is essential to balance surgical safety and adverse effects of increased radiation associated with CT-based navigation.

OP22: Trauma-Thoracolumbar 2

A122: International Validation of the AOSpine PROST (Patient-Reported Outcome Spine Trauma)

Said Sadiqi¹, Marcel W. Post², A. Mechteld Lehr¹, F. Cumhur Oner¹, and AOSpine Knowledge Forum Trauma³

¹University Medical Center Utrecht, Utrecht, Netherlands

² Rehabilitation Center “De Hoogstraat,” Utrecht, Netherlands

³AOSpine Knowledge Forum Trauma, AO Foundation, Davos, Switzerland

Introduction: No outcome instrument is specifically designed and validated for traumatic spinal column injury patients, making it is difficult to measure the effect size of various treatment options in a variety of traumatic spinal column injuries. The AOSpine Knowledge Forum Trauma initiated a project to develop the Patient Reported Outcome Spine Trauma (AOSpine PROST), consisting of 19 items. An international multicenter validation study was performed to test the psychometric properties of the Dutch and English language versions. Material and Methods: The study was conducted among 4 level-1 trauma centers from the Netherlands and North America, and consisted of 3 parts: cross-sectional, test-retest, and responsiveness. Besides the AOSpine PROST, patients also filled out the generic outcome measure SF-36 (36-item Short Form questionnaire) for the purpose of criterion validity. Descriptive statistics were used to analyze the characteristics. Content validity was assessed by evaluating the number of inapplicable questions and any missing question. Also floor and ceiling effects were analyzed. Spearman correlation tests were performed to analyze the correlation within AOSpine PROST items, and the correlation to SF-36. Factor analysis was performed to explore any dimensions within the AOSpine PROST. The internal consistency was assessed by calculating Cronbach’s α and item-total correlation coefficients (ITCC). Test-retest reliability was analyzed using intraclass correlation coefficients (ICC). Finally, responsiveness of the Dutch language version was assessed by the changes in prospective scores with effect sizes (ES) and standardized response mean (SRM). Results: From August 2015 until September 2017, a total of 179 Dutch patients and 196 North American patients were included. Good results were obtained for content validity, as general comments were provided by patients when they had to indicate a missing question in AOSpine PROST. No item showed floor effects; ceiling effects were seen for a few items. Good Spearman correlations were seen between the items of the AOSpine PROST, both for the English (0.38-083) and Dutch (02.8-0.79) versions. Compared to the SF-36, the strongest correlation for the English version was seen with the component Physical Role (0.81; P < .001), while the Dutch language most strongly correlated to Physical Functioning (0.79; P < .001). Factor analysis showed 2 possible dimensions (Eigenvalues >1), explaining 76% of the variance. The internal consistency was excellent both for the English version (n = 135, Cronbach’s α 0.97, ITCC 0.50-0.90) and the Dutch version (n = 119, α 0.96, ITCC 0.50-0.86). Also test-retest reliability was excellent for the English (n = 43, ICC = 0.96 [0.93-0.98]) and Dutch language (n = 62, ICC = 0.92 [0.87-0.95]). Concerning responsiveness analysis, very good results were seen when exploring change scores between 2 weeks and 3 months (n = 59; ES 1.81, SRM 2.03, P < .001), which also showed larger ES when analyzing for different patient degrees of recovery. Conclusion: This study showed excellent psychometric properties for both the Dutch and English language versions of the AOSpine PROST. In the future, the tool will be cross-culturally translated and adapted into many languages. Surgeons around the world are encouraged to use the tool in their daily clinical practice as well in research in order to contribute to improvements in the provided care for spine trauma patients.

A123: Evaluation of Spinopelvic Balance in Patients With Thoracolumbar Fractures Treated Surgically and Conservatively: A Cross-Sectional Study

Lucas Miotto¹, Rodrigo Goes¹, Nelson Astur¹, Alberto Gotfryd¹, Maria Fernanda Caffaro¹, and Robert Meves¹

¹Santa Casa de São Paulo, São Paulo, Brazil

Introduction: Thoracolumbar fractures are considered complex lesions and have high prevalence rate mainly in the young and adult population, most often related to high-energy mechanism such as car accident and falls from height. This type of fracture can cause neurological deficit in up to 50% of cases. The region susceptible to fracture because it is a region of transition of the spine in which a rigid dorsal column articulated with the ribs of the thoracic cage becomes the lumbar spine without fixations. Burst fractures, without neurological deficit, can cause imbalance of the sagittal parameters; thus, biomechanical conditions that accelerate degenerative processes generate impairment in the function and life quality of these patients. The objective of the study is to correlate the sagittal and spinopelvic parameters with functional and health-related quality of life scores of patients treated surgically and conservatively. Material and Methods: A cross-sectional study was conducted, which collected epidemiological, clinical, and radiological data from patients who were treated due to thoracolumbar fracture types A3 and A4, without neurological deficit in the Hospital Santa Casa de São Paulo. Applied Scale of Dennis, EVA, and SF-36 and measured regional kyphosis, lumbar lordosis, thoracic kyphosis, pelvic incidence, pelvic tilt, sacral slope, VAS, mismatch (lumbar lordosis − pelvic incidence) and compares between the surgical and nonsurgical groups their outcome. Results: Sample size 49/age 50.8/T10 1 (2%), 0 (0%)/1 (100%), T11 2 (4%), 0 (0%)/2 (100%), T12 16 (32.6%), 8 (50%)/8 (50%), L1 19 (38.7%), 9 (47.3%)/10 (56.7%), L2 10 (20.4%), 7 (70%) 3/(30%). Accident motor vehicle 23 (46.9%), fall of height 26 (53.1%), Denis pain conservative 24, 2.708, 1.0417, P < .05/surgical 25, 3320, 1.0296. SF-36: Pain: Surgical 25, 35 840, 18 8208/conservative 24, 50.667, 24.8188, P < .05.Vitality: Surgical 25, 48 600/conservative 24, 60.625, 22.3272, P < .05. Emotional Limitations: Surgical 25, 5280, 12 3475/conservative 24, 25 000, 39.6485, P < .05. Conclusion: Most patients were considered without deformity in the global sagittal plane, despite having regional deformity at the fractured level, with SVA (sagittal vertical axis) between −4 and 4, as determined by the SRS-Schwab classification of adult spinal deformity. Patients treated conservatively had better functional outcomes than surgical patients with statistical significance in the SF-26 and Denis classification. Conservative treatment proved to be more effective than surgical treatment for long-term thoracolumbar fractures.

A124: Traumatic Cauda Equina Injury in Canada: What Predicts Neurological and Bowel or Bladder Sphincter Outcomes?

Najmedden Attabib^1,2, Colleen O’Connell¹, Dilnur Kurban³, Carly Rivers³, Chris Bailey⁴, Sean Christie¹, Karen Ethans⁵, Heather Flett⁶, Julio Furlan⁷, Eve Tsai⁸, the RHSCIR Network³, and Amanda Carson²

¹Dalhousie University, Halifax, Nova Scotia, Canada

²Saint John Regional Hospital, Saint John, Canada

³The Rick Hansen Institute, Blusson Spinal Cord Centre, Vancouver, British Columbia, Canada

⁴University of Western Ontario, London, Ontario, Canada

⁵University of Manitoba, Winnipeg, Manitoba, Canada

⁶University of Toronto, Toronto, Ontario, Canada

⁷Toronto Western University, Toronto, Ontario, Canada

⁸Ottawa Combined Adult Spinal Surgery Program, The Ottawa Hospital, Ottawa, Ontario, Canada

Introduction: Traumatic cauda equina injury (TCES), usually an outcome of injury from L1 and below, results in deleterious independence impairments both functional (bowel or bladder sphincter dysfunction) and neurological (sensory or motor weaknesses). Improving neurologic and functional dysfunction following TCES is of key importance to affected individuals. The ability to prognosticate outcomes will aid both in communication and rehabilitation goal-planning. This study aims to determine what/if any factors predict neurologic and functional recovery. Material and Methods: Observational study utilizing patients (N = 187) with acute TCES participating in the Rick Hansen Spinal Cord Injury Registry (RHSCIR). At admission and discharge measures of interest were neurological level of injury, lower extremity motor function, voluntary anal contractions (VAC), and change in urinary or bowel function. Data were collected using International Standards for Neurological Classification of Spinal Cord Injury (ISNSCI) examinations, lower extremity motor scores (LEMS), the Functional or Spinal Cord Independence Measure (FIM or SCIM). Demographic variables of interest were collected and analyzed using descriptive statistics. Multiple linear regressions were run with for both bladder and bowel change scores as the dependent measure. Independent measures included age, sex, rehabilitation length of stay (LOS), rehabilitation onset, neurological level, and time from injury to surgery. Results: Participants mean age was 38.4 ± 16.6 years and were predominantly male (76.5%). Mechanisms of injuries were 48.1% fall related, 28.4% transportation related, 14.8% sports related, and 8.8% due to other traumatic causes. The neurological level of injury was separated into 1 of 3 categories with 50.8% at L1, 27.3% at L2, and 21.9% at L3-S3. At admission 47.4% of patients had VAC; 38% who had no VAC at admission recovered function at discharge. Multivariate analysis indicated that age, gender, and rehabilitation LOS were significant independent predictors of bladder function change—with older age, females, and longer rehabilitation LOS being associated with increased bladder score recovery. Gender, time from injury to surgery, rehabilitation onset, and rehabilitation LOS were significant predictors of bowel function change—with females, lower time to surgery (<72 hours), lower rehabilitation onset, and increased rehabilitation LOS being associated with increased bowel score recovery. The regression model for modeling neurologic recovery did not identify any significant predictors. Conclusion: The most novel finding of this study was the impact of time of surgery on bowel function recovery, which has not previously been shown. Additionally, time to surgery and length of stay are variables that can be manipulated by medical professionals. Expanding our knowledge on predicting outcomes in those with TCES presents an opportunity for medical professionals to improve prognosis of bladder, bowel function, and to support rehabilitation goal planning and communication with individuals and their families, especially for those at risk for poor outcomes.

A125: Reliability of the AO Classification of Thoracolumbar Fractures Compared to TLICS and Magerl

Alynson Larocca Kulcheski¹, Marcel Luiz Benato¹, Pedro Grein del Santoro¹, Andre Sebben¹, and Xavier Soler i Graells¹

¹Hospital do Trabalhador, Curitiba, Brazil

Introduction: To test the reliability of the new AO/2013 classification compared with AO/Magerl and TLICS (Thoracolumbar Injury Classification System). Methods: Four spine surgeons retrospectively and blindly evaluated imaging and clinical data from 98 patients with thoracolumbar fractures. Results: Using the κ coefficient, we obtained the best reproducibility for the AO/2013 classification compared to the other two, represented by κ coefficient of 0.690. We could also obtain, with good reproducibility among the evaluators (κ 0.690), the most common subtypes of AO/2013 classification with indication for surgery. Conclusion: We believe that the new AO/2013 classification has proven to be a good communication tool among spine surgeons with good reproducibility, but more studies should be conducted in several centers for the results to be consolidated and the prognosis between the types of injury is better understood.

A126: Analysis of Factors Affecting Postoperative Loss of Reduction in Unstable Thoracolumbar Fractures After Posterior Surgery

Jaewan Soh¹

¹Soonchunhyang University Cheonan Hospital, Cheonan, Republic of Korea

Introduction: Unstable thoracolumbar fractures require surgical intervention to achieve adequate reduction and restore spinal stability. However, loss of reduction can cause kyphotic deformity, which results in several complications. The authors studied the factors influencing the postoperative loss of reduction. Material and Methods: Of 52 patients who were treated by posterior surgery and had completed a minimum follow-up of 1 year, we excluded 9 patients who had multiple-level fractures or underwent combined anterior-posterior surgery, which yielded 41 patients. Postoperative loss of reduction was defined as 30% or more loss of vertebral body height as measured on lateral radiography at 1-year follow-up compared to lateral radiograph taken immediate after surgery. The associations between the patients’ gender, age, level of fracture, injury of posterior column, initial loss of fractured vertebral body height, the load-sharing score, TLICS (Thoracolumbar Injury Classification System) score, number of fixed segments above and below the fracture, type of pedicle screws, the degree of postoperative reduction, the degree of postoperative corrected kyphotic angle, the changes in insertion angle of the upper and lower pedicle screws of the fractured vertebral body at 1-year follow-up, and postoperative loss of reduction were analyzed. SPSS 14.0 was used for analysis of data. Results: Among 41 patients, 7 patients had postoperative loss of reduction. The number of fixed segments above and below the fracture (P = .045), initial loss of fractured vertebral body height (P = .013), and the changes in insertion angles of the lower pedicle screws (P = .007) were significantly associated with postoperative loss of reduction. Conclusion: In patients with posterior surgery for thoracolumbar unstable fractures, loss of reduction occurred when single-segment fixation above and below the fracture was used, 50% or more loss of initial fractured vertebral body height, and 5° or more changes in insertion angles of the lower pedicle screws of the fractured vertebral body. This result indicates that 2-segment pedicle screw fixation is more stable than single-segment fixation when 50% or more loss of vertebral body height at the initial fracture and loss of fracture reduction is predicted when insertion angles of the lower pedicle screw of the fractured vertebral body changes. Further studies are needed with more cases and long-term follow-up of patients.

A127: Treatment of Thoracolumbar Trauma by Short-Segment Percutaneous Transpedicular Screw Instrumentation With Parallel Distraction Using Polyaxial Locking Mechanism (Correction Key): Retrospective Comparative Study

Tsuyoshi Okudaira¹, Hideo Baba¹, Takayuki Yamaguchi¹, Chieko Imai¹, Akira Saho¹, and Hiroaki Konishi¹

¹Nagasaki Rosai Hospital, Sasebo, Japan

Introduction: In an effort to minimize soft tissue injury during spinal surgery, percutaneous transpedicular screws are currently used in the treatment of thoracolumbar fractures, although there are substantial drawbacks. The most cited drawbacks include a reduced angular stability of the polyaxial screws and diminished potential to correct kyphotic deformity. To overcome these drawbacks, polyaxial locking setscrews (Correction Key) were applied to cranialy placed pedicle screws followed by parallel distraction. Material and Methods: Five consecutive patients (3 men and 2 women, average age 49.6 years) with thoracolumbar fractures were enrolled in the study. These patients were treated by short-segment, minimally invasive, percutaneous pedicle screw instrumentation, and parallel distraction (Expedium Verese, DePuy Synthes, henceforth referred to as modified MIS group). The control group was composed of 20 patients (13 men and 7 women, average age 45.4years) who were stabilized using a short-segment transpedicular construct that was performed through a standard midline incision (USS, DePuy Synthes, henceforth referred to as conventional open group). The other control group was composed of 23 patients (16 men and 7 women, average age 55.8 years) who were treated with conventional minimally invasive percutaneous pedicle screw instrumentation (conventional MIS group). The pain profile was assessed by Visual Analogue Scale (VAS). Radiological evaluation consisted of supine admission radiography, CT (computed tomography), MRI (magnetic resonance imaging), and follow-up standing radiographs. Kyphotic deformity and radiographic follow-up were defined by the vertebral body index (VBI: relation between the anterior and posterior wall height of the fractured vertebra), vertebral body angle (VBA: angle defined by the upper and lower endplate of the fractured vertebra), and bisegmental Cobb angle (angle defined by the upper endplate of the first vertebra above the fractured one and by the lower endplate of the first vertebra below the fractured one). Results: The mean surgical duration in the modified MIS group was 51 minutes and was significantly shorter than the open conventional group with the average time of 111 minutes. The modified MIS group had a significantly lower estimated blood loss of 20 g compared with 158 g of the conventional open group. The modified MIS group and conventional MIS group yielded analogous estimated blood loss. Scores on the VAS in patients in the percutaneous screw and parallel distraction group and conventional MIS group during the first 7 postoperative days were significantly lower than those in the USS control group. There was no significant difference between groups in VBI, VBA, and Cobb angle values in the modified minimally invasive technique and conventional open procedures. The modified MIS group and the open conventional group are superior to the conventional MIS group in the reduction of kyphotic deformity. Conclusion: This study preliminarily confirms that the modified percutaneous transpedicular screw technique is radiologically comparable with conventional open procedures. A significant reduction in blood loss, postoperative pain, and surgical time were the primary advantages associated with this modified minimally invasive technique in the short-term period. This modified minimally invasive technique may overcome drawbacks associated with conventional minimally invasive procedures.

A128: Vertebral Body Lavage Reduces Hemodynamic Response to Vertebral Body Augmentation With PMMA

Christoph Albers¹, Lorin Benneker¹, and Sven Hoppe¹

¹University Hospital, Bern, Switzerland

Introduction: Vertebral body lavage (VBL) prior to percutaneous cement augmentation procedures for acute osteoporotic vertebral compression fractures (VCF) reportedly reduces pulmonary fat embolism syndrome by decreasing the amount of bone marrow particles displaced into the circulation in animal models. However, no clinical trial has confirmed these findings in humans. We therefore asked whether VBL affects (1) systemic blood pressure, (2) heart rate, and (3) oxygen saturation in patients undergoing percutaneous cement augmentation procedures for acute VCF. Material and Methods: In a retrospective comparative study, 145 consecutive patients undergoing percutaneous cement augmentation procedures for acute VCF of the thoracolumbar spine (mean age 74 ± 12 [42-96] years; 70% female; 475 levels treated) were included. Patients undergoing VBL prior to cement application were allocated to the “lavage group” (n = 61 patients; 203 spinal levels treated), and patients without prior VBL to the “control group” (n = 84 patients, 271 spinal levels treated). The groups matched in terms of preoperative demographics and procedure-related factors. Mean arterial blood pressure (MAP), heart rate, and oxygen saturation were monitored immediately prior and 3 minutes after cement injection. Logistic regression analysis was performed with a drop of MAP of at least 10 mm Hg (ΔMAP ≥ 10 mm Hg) before and after cement injection as the dependent input variable and demographic, radiographic, and procedural factors as independent variables. Results: MAP decreased by mean 3 ± 7.3 (range 0-30 [confidence interval (CI) 0.5-6.7]) mm Hg before and after cement injection in the “lavage group” and 9 ± 10.5 (range −3 to 35; [CI 7-11]) mm Hg in the “control group” (P < .001). There were no differences in terms of heart rate and oxygen saturation before and after cement application within each group nor between the 2 groups. Multivariate logistic regression analyses revealed the risk of having a decrease of MAP of 10 mm Hg or more to be elevated more than 3-fold if no VBL lavage was performed prior to cement injection (adjusted odds ratio 3.49, P = .03). Conclusion: VBL leads to decreased drop in blood pressure after percutaneous cement augmentation procedures. This effect is likely the result of a decreased amount of bone marrow substance displaced into the circulation ultimately reducing the risk of pulmonary fat embolism.

A129: Posterior Vertebral Column Resection in Management of Severe Posttraumatic Thoracolumbar Kyphosis

Ahmed ElNaggar¹, Shamel ElGawhary¹, and Tarek ElHewala¹

¹Zagazig University, Zagazig, Egypt

Introduction: Thoracolumbar fractures are the commonest spinal fracture, but its treatment remains one of the major controversies. Posttraumatic kyphosis (PTK) is considered the commonest serious complication of poorly managed thoracolumbar fractures that can affect patients’ health-related quality of life. Different corrective osteotomy techniques are described to correct this deformity but with limited correction in sever angular kyphotic deformity. Our hypothesis was that among several osteotomies described to correct kyphosis, vertebral column resection (VCR) provides the highest angle of correction required to manage patients with severe PTK. Our aim is to evaluate the safety and efficacy of posterior VCR in the treatment of patients with severe degrees of posttraumatic thoracolumbar kyphosis and determine the degree of correction in sagittal imbalance and its relationship with functional outcome of the patients. Material and Methods: We evaluate, retrospectively, 12 patients who were suffering from PTK with mean local kyphosis angle (LKA) 64.1 ± 6.3° and were managed with posterior VCR in the period between 2012 and 2015. All patients were assessed clinically using the Visual Analogue Scale (VAS) and Oswestry Disability Index (ODI) and radiologically using LKA, global kyphosis (GK), lumbar lordosis (LL), and sagittal vertical axis (SVA). Results: The mean LKA and SVA score improved from 64.1 ± 6.3° and 52.6 ± 8.3 mm preoperatively to 8.8 ± 3.4° and 13.1 ± 4.8 mm postoperatively, respectively. All were statistically significant. Functionally, the VAS score of back pain significantly reduced from 6.8 ± 0.9 to 1.7 ± 0.8, and the ODI significantly improved from 59.8 ± 7.5 to 11.6 ± 3.4. No major complication occurred. We encountered intraoperative dural tear in 2 patients, superficial infection in 1 patient, deep infection in 1 patient, and temporary paraparesis in 1 patient. Conclusion: Sagittal spinal deformity is one of the major problems of patients with posttraumatic kyphosis, which affects the health-related quality of life. The present study showed that posterior vertebral column resection technique can resolve this deformity with satisfactory outcome. The complications of this procedure are its major drawback even in the hands of expert surgeons. However, with appropriate application and fine technique, PVCR can show satisfactory improvement in deformity correction and functional outcome.

A130: Sagittal Alignment and Mismatch After Short Segment Fixation of Thoracolumbar Fracture: 4-Year Follow-up of Thoracolumbar Fracture Treated by Short Segment With Intermediate Screw Fixation

Tarek ElHewala¹, Amr El-Adawy¹, Shamel ElGawhary¹, and Mohamed Abdeen¹

¹Zagazig University, Zagazig, Egypt

Introduction: Thoracolumbar fractures are commonly managed by posterior pedicle screw fixation. Loss of reduction and posttraumatic kyphosis are serious complications of poorly managed fractures, which can affect the overall sagittal alignment and balance of the spine. Stability of the short segment fixation remains questionable in stabilizing these fractures. Intermediate screw fixed in the fractured level had improved construct stability biomechanically and clinically. The aim of this study is to have a long-term follow-up of patients treated with short segment fixation with intermediate screws (SSFIS) in the management of thoracolumbar burst fractures and identify any changes in the sagittal alignment of these patients compared to the normal population. Material and Methods: Long films of the whole spine of 21 patients with thoracolumbar burst fracture (T11-L2) type A3 and A4 AOSpine classification, who were treated between 2009 and 2014 with short segment posterior pedicle screw fixation with intermediate screws in the fractured vertebra, were reviewed and evaluated for local kyphotic angle (LKA) correction and maintenance, sagittal alignment and balance. Visual Analogue Scale (VAS) for back pain and Oswestry Disability Index (ODI) were also recorded. Results: Average follow-up period of these patients was 4.7 ± 0.9 years. Postoperative correction of the local kyphotic angle (LKA) was maintained with mean loss of correction 1.8 ± 0.6°, which was statistically insignificant. Sagittal parameters and balance (sagittal vertical axis, pelvic incidence (PI), pelvic tilt, lumbar lordosis (LL), and PI-LL mismatch) are within normal ranges. The follow-up VAS for back pain was reduced by average 0.6 with recording of episodes of back pain at the site of fixation with maintenance of ODI scores and patient satisfaction in comparison to the early postoperative results. Conclusion: Short segment fixation with intermediate screws can maintain long-term reduction and fixation of thoracolumbar burst fracture and spinal sagittal alignment and balance with significant patient satisfaction and functional outcome. More multicenter studies needed to confirm these results.

OP23: Degenerative-Cervical 4

A131: Cervical Degenerative Disc Disease and Subclinical Discitis: Cause or Contaminant?

Amit Bhandutia¹, Louis Bivona¹, Alysa Nash¹, Farooq Usmani¹, Jael Camacho¹, Meghan Hughes¹, Daniel Gelb¹, Eugene Koh¹, Kelley Banagan¹, and Steven Ludwig¹

¹University of Maryland Medical Center, Baltimore, MD, USA

Introduction: The presence and potential effect of low-virulent bacteria, Propionobacterium acnes (P acnes), residing within intervertebral discs, remains a controversial topic surrounding degenerative disc disease (DDD). Recent studies have demonstrated a correlation between lumbar DDD and culture-positive disc space infection. This study aims to determine the prevalence of underlying infection in patients undergoing anterior cervical decompression and fusion (ACDF) and includes a corresponding contaminant control. Material and Methods: After institutional review board approval, patients scheduled to undergo elective ACDF were prospectively enrolled between January 2017 and May 2018. During anterior exposure and prior to discectomy, a sample of the longus colli muscle was obtained as a control with an unused, sterile instrument. On exposure of the appropriate level, an unused, sterile instrument was used to obtain a sample of disc tissue. Samples were homogenized, gram stained, and cultured in both aerobic and anaerobic medium for 5 and 14 days, respectively. Results: Of the 96 patients in the study, 41.6% were male with an average age of 53.8 ± 10.9 years and BMI (body mass index) of 29.7 ± 5.9. Control cultures of longus colli muscle were positive in 48 patients (50.0%). Compared to the 20.8% of male patients in negative control (NC) patients, there were significantly more (62.5%) male patients in the positive control (PC) group (P < .0001). Disc cultures were positive in 52 (54.2%) patients and 81 (48.8%) discs. The negative disc (ND) and positive disc (PD) patients were similar with respect to age, BMI, prior spine surgery, diagnosis, symptoms, and smoking status. Compared to the 20.5% of male ND patients, there were significantly more (59.6%) male patients in PDs (P = .0002). Of the 44 ND patients, 73% also had NC. Likewise, of the 52 PDs, 69% also had PCs. McNemar’s test did not show a statistically significant correlation between PDs and CPs patients (P = .45 but the odds ratio was 6). P acnes grew in 85.2% of the discs and 77.1% of the control cultures. P acnes grew alone in majority of the cultures and had concurrent Streptococcus or Staphylococcus growth in a few instances (4 of 37 for CCPs and 12 of 69 for DCPs). Staphylococcus was the second most common organism and grew in 16.7% and 11.1% of PC and PDs, respectively. Conclusion: Our study makes an important contribution to the ongoing debate regarding the infectious etiology of DDD. The 54% infection rate of cervical spine discs correlates with infection rate in the lumbar spine reported by many studies. Our control cultures of longus colli muscle show a strong association with the disc culture: 69% of patients with positive discs also had positive control cultures. P acnes was the primary organism in 77% of control culture and 85.2% of disc cultures. The strong correlation between disc cultures and longus colli cultures suggest possible contagious spread of infection. Our findings show that infection of the disc space is a possible cause of DDD of the cervical spine.

A132: The Impact of Older Age on Functional Recovery After Surgical Decompression for Degenerative Cervical Myelopathy: Results From an International, Multicenter, Prospective Dataset in 757 Patients

Jamie R. F. Wilson¹, Jetan H. Badhiwala¹, Fan Jiang¹, Jeff Wilson², Branko Kopjar³, Alex Vaccaro⁴, and Michael Fehlings¹

¹Toronto Western Hospital, Toronto, Ontario, Canada

²St Michael’s Hospital, Toronto, Ontario, Canada

³University of Washington., Seattle, WA, USA

⁴ Rothman Orthopaedic Institute, Philadelphia, PA, USA

Introduction: Surgical decompression has been shown to improve long-term function, disability, and quality of life (QOL) in degenerative cervical myelopathy (DCM); however, the role of surgery, and the effect on functional and QOL outcomes, in elderly patients with DCM is controversial. Material and Methods: Of 757 patients enrolled in the prospective multicenter AOSpine CSM North America and International studies, 107 patients were identified as elderly (70 years or older) at the time of surgery. Outcomes were assessed at 6, 12, and 24 months and included functional status (mJOA [modified Japanese Orthopaedic Association]) and QOL (SF-36 [36-item Short Form]). In addition to unadjusted univariate analyses, multiple linear regression was used to evaluate the effect of age ≥70 on change in outcome scores adjusting for diabetes, cardiac comorbidities, cigarette smoking, psychiatric disorders, number of operated levels, surgical approach, and baseline mJOA. Results: The baseline mJOA in the elderly group was significantly lower than the younger group (11.0 [95% confidence interval [CI] 10.4-11.5] vs 12.9 [12.7-13.1]; P = <.01). The SF-36 Physical Component Summary (PCS) score was no different (33.5 [31.8-35.1] vs 34.5 [33.8-35.2]; P = .28) and the SF-36 Mental Component Summary (MCS) score was significantly higher in the elderly group (43.0 [40.6-45.5] vs 39.8 [38.7-40.8]; P = .02). The unadjusted change in mJOA scores were similar in both groups at 6 months (elderly 2.30 [1.71-2.88] vs younger 2.21 [2.02–2.40]; P = .75), 12 months (2.79 [2.18-3.41] vs 2.50 [2.29-2.70]; P = .30), and 24 months (2.63 [1.99-3.27] vs 2.71 [2.51-2.92]; P = .77). However, after covariate adjustment, the coefficient for change at 6 months in the elderly group was −0.84 (P = <.01), −0.74 at 12 months (P = <.01), and −1.22 at 24 months (P = <.01). The change in SF-36 physical scores were lower in the elderly group at 6 months (3.14 [1.46-4.81] vs 5.78 [5.07-6.48]; P = <.01), 12 months (5.56 [3.86-7.26] vs 6.41 [5.67-7.16]; P = .40), and 24 months (3.43 [1.37-5.49] vs 6.17 [5.40-6.94]; P = <.01). The coefficient of change when adjusted at 6 months in the elderly group was −3.02 (P = <.01), −1.16 at 12 months (P = .27), and −3.65 at 24 months (P = <.01). The change in SF-36 mental scores were not significantly different at 6 and 12 months (5.23 [2.40-8.06] vs 6.01 [5.05-6.97]; P = .56, 5.96 [3.32-8.61] vs 6.18 [5.22-7.14]; P = .87) but were significantly lower in the elderly group at 24 months (2.59 [−0.028 to 5.47] vs 5.96 [4.97-6.96]; P = .01). The coefficient of change after adjustment at 6 months was −0.97 (P = .50), −1.23 at 12 months (P = .38), and −4.53 at 24 months (P = <.01) in the elderly group. Conclusion: In this combined dataset from 2 large prospectively collected multicenter studies on DCM surgery, the group aged 70 or greater demonstrated significantly worse functional and QOL recovery when compared to the younger cohort after adjusting for the effect of comorbidities, number of operated levels, surgical approach, and baseline mJOA. Elderly patients undergoing surgery for DCM should therefore be counseled appropriately regarding expectations of surgery.

A133: Influence of Uncinate Process Removal on Postoperative Radiologic Outcomes: A Focus on Subsidence and Sagittal Balance in Patients Receiving Single-Level Anterior Cervical Discectomy and Fusion

Sung Hyun Noh¹ and Kyung Hyun Kim¹

¹Gangnam Severance Hospital, Yonsei University College of Medicine, Seoul, Republic of Korea

Objective: When performing cervical foraminal stenosis with cumulative osteophyte or uncovertebral hypertrophy, some surgeons perform concurrent uncinate process resection (UPR) with anterior cervical discectomy and fusion (ACDF). We sought to evaluate the clinical and radiological consequences of UPR during ACDF versus those seen with ACDF without UPR. Methods: We retrospectively reviewed the data from all patients who underwent single-level ACDF with plate and cage between 2011 and 2015. In total, 105 patients participated in 2 years of follow-up. Among them, 37 underwent ACDF with UPR, and 68 underwent ACDF without UPR. We examined patient clinical status by determining Visual Analogue Scale (VAS), Neck Disability Index (NDI), and Odom’s criteria. Moreover, radiographic studies were analyzed retrospectively. Participants were divided into the 2 aforementioned groups. The demographics, age, sex, body mass index (BMI), diabetes mellitus (DM), smoking, and operation level in the 2 groups were compared. Radiographic parameters of disc height, C2-C7 lordosis, T1 slope, C2-C7 sagittal vertical axis (SVA), center of the sellar turcica-C7 SVA (St-SVA), spinocranial angle (SCA), and fusion rate were measured on plain radiographs at preoperation, immediately postoperation, and during the follow-up period (median follow-up duration: 37.7-17.5 months). Results: The improvement in VAS score for arm pain was significantly better in the ACDF with UPR group versus in the ACDF without UPR group immediately postoperation. The fusion rates, C2-C7 lordosis, T1 slope, and C2-C7 SVA after single-level ACDF were not significantly different between the 2 groups. Subsidence occurred in 32 patients (ACDF with UPR: 19 cases [51%] vs ACDF without UPR: 13 cases [19%]; P < .05). Patients who experienced subsidence showed gradual worsening of sagittal radiologic parameters such as C2-C7 SVA, SCA, and St-SVA. Multivariate analysis for risk factors for subsidence showed that preoperative C2-C7 SVA, St-SVA, and SCA were significant factors in the presentation of subsidence after ACDF (P < .05). Conclusions: Subsidence occurred more frequently after ACDF with UPR than after ACDF without UPR. Surgeons should be aware of these results so that unnecessary UPR during ACDF can be avoided.

A134: The Effect of Tandem Lesions in Patients With Cervical Ossification of the Posterior Longitudinal Ligament

Yoshiharu Kawaguchi¹, Masato Nakano¹, Taketoshi Yasuda¹, Shoji Seki¹, Kayo Suzuki¹, Yasuhito Yahara¹, Hiroto Makino¹, Masahiko Kanamori¹, and Tomoatsu Kimura¹

¹University of Toyama, Toyama, Japan

Introduction: Cervical ossification of the posterior longitudinal ligament (OPLL) is associated with spinal ossified lesions. Tandem lesions cause additional neurological deficit, affecting the outcomes of cervical laminoplasty. This study aimed to clarify the effect of tandem lesions on the outcomes of cervical laminoplasty in comparison with data on patients without tandem lesions. Materials and Methods: From April 1981 to October 2015, 201 patients underwent laminoplasty for cervical OPLL; however, 167 patients were followed for >2 years. Twenty-eight patients underwent an operation for tandem lesions. The pathologies of tandem lesions were assessed and divided into 2 groups: the thoracic and thoracolumbar tandem group (T-tandem group: 9 patients) and the lumbar tandem group (L-tandem group: 26 patients). One hundred patients without tandem lesions served as the control group. The maximum Japanese Orthopaedic Association (JOA) score and the most recent score for recovery was compared between the tandem and control groups. Results: The maximum score and recovery rate and the score and recovery rate at the last follow-up in the tandem group were lower than in the control group. There was a significant difference in walking ability between the T-tandem and L-Tandem groups. Three of 9 patients in the T-tandem group used a wheelchair, whereas none used a wheelchair in the L-tandem group at the last follow-up. Conclusions: Neurological recovery in patients with tandem lesions was poorer than in patients without lesions. In particular, a thoracic tandem lesion brought about poor neural recovery; thus, special attention should be paid to cervical OPLL with thoracic lesions.

A135: Clinicoradiological Efficacy of Posterior Occipitocervical Fusion With Plate-Screw-Rod Construct and Allograft in Craniovertebral Instability in 52 Patients

Mihir Upadhyaya¹

¹Bombay Hospital & Research Centre, Mumbai, India

Introduction: Very few studies have evaluated occipitocervical (OC) fusion using plate screw rod construct with allograft for a long-term follow-up with a large number of patients. Hence, we aim to evaluate the clinicoradiological outcomes of patients who had undergone OC fusion using plate screw rod construct with allograft over a minimum 5-year follow-up. Material and Methods: Study Design: Retrospective analysis of 52 prospectively collected data of patients who underwent OC fusion using plate screw rod construct and allograft (n = 59) in a single institute by senior spine surgeon from 2011 to 2015 and analyzed by an independent observer. Institutional review board approval was obtained before conducting the study. Patients’ functional status was evaluated by the Modified McCormick scale and preoperative versus postoperative neurological function was evaluated by Japanese Orthopaedic Association (JOA) scoring system. Fusion assessment was done by both direct and indirect evidence; direct being bony trabeculae at the graft-recipient interface on sagittal computed tomography reconstruction and indirect on lateral cervical radiographs was considered evidence of union. Inclusion criteria: craniovertebral instability with/without basilar invagination; posterior only surgery; reducible C1-C2 instability; minimum 24-month follow up. Exclusion criteria: anterior surgery; revision surgery; sublaminar wiring; follow-up <24 months. Results: The minimum follow-up period was 2 years and final follow-up at 2 years, mean 26 months (range 24-48). JOA scores were 13.83 ± 2.734 before surgery and 15.67 ± 1.854 at the final follow-up. All neurologically intact patients remained same after surgery. Among 26 patients with cervical myelopathy, 24 patients (92.31%) had improvement of the modified JOA scores at last follow-up and improvement by 1 grade in the functional status according to modified McCormick scale in 40 patients (76.92%). No myelopathy or deterioration in any patient was noted. Postoperative X-ray at 6 months did reveal 1 instrument loosening (29.6%). There were 2 cases of vertebral artery (VA) injury during operation. Two cases of superficial infection and one case of dural leak having complication rate of 11.53%. Significant improvement noted in radiological parameters with reduction of basilar invagination ADI (preoperative = 4.2 mm [range 3.8-6.2 mm] vs postoperative = 2.5 mm [range 1.4-4.0 mm]) (P < .085). After surgery, most patients’ posterior OC2 angle entered the normal range as the preoperation angle decreased from 122° to 108° (P < .007). Conclusion: OCF using rod screw construct + allograft is an effective method in treating craniocervical fusion, which provides excellent clinical outcome with biomechanically sound fixation, high fusion rate, and optimal correction of misalignment in the occipital-atlantoaxial region.

A136: Effects of Renin-Angiotensin System Inhibitors on Spinal Fusion: Retrospective Analysis of a Cohort of ACDF Patients

Alexander Perdomo-Pantoja¹, Feras Shamoun¹, Christina Holmes², Wataru Ishida¹, Seba Ramhmdani¹, Ali Bydon¹, Nicholas Theodore¹, L. Lo Sheng-Fu¹, and Timothy F. Witham¹

¹ Johns Hopkins University, Baltimore, MD, USA

²Florida A&M University-Florida State University College of Engineering, Tallahassee, FL, USA

Introduction: A total of 250 000 to 400 000 spinal fusion surgeries are performed per year in the United States.¹ Although the utilization of instrumentation and bone graft alternatives has enhanced fusion rates, pseudarthrosis remains a problem, with incidence rates ranging from 0% to 56%.² Renin-angiotensin system (RAS) blockers are commonly used in the treatment of cardiovascular diseases, such as arterial hypertension, which affect 50% to 80% of spinal fusion patients.³ Recent evidence suggests an association between RAS blockers and bone metabolism, particularly in the context of bone healing.^4-6 This association might play a critical role in spinal fusion patients who also suffer from arterial hypertension. The purpose of this study was to determine whether there is a correlation between the use of antihypertensive medication and outcomes in anterior cervical discectomy and fusion (ACDF) surgery. Materials and Methods: We performed a retrospective chart review of 128 degenerative disc disease patients who underwent ACDF, with 1-year minimum follow-up. Information on demographic data, comorbidities, antihypertensive medication, neurological examination, and fusion status were collected. Patients with incomplete data were excluded. Spinal fusion was evaluated via plain cervical X-ray, resorting to dynamic radiographs in cases of doubt. Results: Of the initial 128 patients, 102 met the inclusion criteria, including 53 (52%) females and 49 (48%) males, with a median age of 53.7 years. Out of the included patients, 14 (13.7%) smokers, 20 (19.6%) diabetic patients, and 46 (45.1%) hypertensive patients were identified. Out of the 46 hypertensive patients, 44 (95.7%) were taking antihypertensive drugs (17 received angiotensin II receptor blockers [ARBs], 14 received angiotensin-converting enzyme inhibitors [ACEIs], and the remaining 13 were taking other medication). In the analysis of fusion rates, patients treated with ARBs exhibited a higher fusion rate compared to those untreated or treated with ACEIs (P = .03 and .005, respectively). Patients treated with ARBs displayed a relative benefit of fusion over untreated patients of 1.289 (95% confidence interval [CI] 1.122, 1.480, P = .03). Smoking presented a negative correlation with spinal fusion compared to nonsmokers (P = .03). Patients treated with ACEIs displayed a lower fusion rate than untreated patients; however, the difference was not significant (P > .05). Interestingly, in the neurological examination analysis, treatment with ACEIs was correlated with lower preoperative modified Japanese Orthopedic Association (mJOA) and Nurick scores compared to untreated patients (P = .01 and <.001). Patient age, diabetic status, hypertension, and other drugs, did not significantly influence fusion rates or neurologic examination scores. Conclusions: Our retrospective study shows that treatment of ACDF patients with ARBs was associated with higher fusion rates than untreated patients or patients treated with ACEIs. Conversely, smoking was related to pseudoarthrosis. We also found that ACDF patients treated with ACEIs exhibited lower preoperative mJOA and Nurick scores compared to untreated patients. Further studies, including a larger population and different spinal levels, are required to confirm that ARB treatment has beneficial effects on spinal fusion.

References

1. Rajaee SS, Bae HW, Kanim LEA, Delamarter RB. Spinal fusion in the United States: analysis of trends from 1998 to 2008. Spine (Phila Pa 1976). 2012;37:67-76.

2. Gruskay JA, Webb ML, Grauer JN. Methods of evaluating lumbar and cervical fusion. Spine J. 2014;14:531-539. doi:10.1016/j.spinee.2013.07.459

3. Rajaee SS, Kanim LEA, Bae HW. National trends in revision spinal fusion in the USA: patient characteristics and complications. Bone Joint J. 2014;96-B:807-816. doi:10.1302/0301-620X.96B6.31 149

4. Yamamoto S, Kido R, Onishi Y, et al. Use of renin-angiotensin system inhibitors is associated with reduction of fracture risk in hemodialysis patients. PLoS One. 2015;10:e0122691. doi:10.1371/journal.pone.0 122 691

5. Bayar A, Turan A, Gülle K, Akpolat M, Turan İ, Turhan E. The effects of the angiotensin converting enzyme inhibitor enalapril and the angiotensin II type 1 receptor blocker losartan on fracture healing in rats. Clin Invest Med. 2015;38:E164-E172.

6. Izu Y, Mizoguchi F, Kawamata A, et al. Angiotensin II type 2 receptor blockade increases bone mass. J Biol Chem. 2009;284:4857-4864. doi:10.1074/jbc.M807610200

A137: Identifying Prognostic Profiles of the Symptoms and Quality of Life in Patients With Cervical Spondylotic Myelopathy Underwent Surgical Treatment: A Cluster-Analytic Approach

Feifei Zhou¹, Yilong Zhang¹, and Yu Sun¹

¹Peking University Third Hospital, Beijing, China

Introduction: Surgical decompression is considered to be the most effective way for cervical spondylotic myelopathy (CSM). Patients’ objective and subjective reported outcome measurements are always evaluated pre- and postoperatively in one specific item (eg, motor function of lower limbs in modified JOA [Japanese Orthopaedic Association] or social function domain in SF-36). However, all the items, no matter objective or subjective method, interactively influence patients’ feedback to the surgical intervention. The aim of this study is to identify profiles of the symptoms and quality of life in patients with CSM after surgery by a cluster analysis. Methods: Cohort study with prospective data collection. Both mJOA score (7 items, Nos. 1 to 7) and SF-36 (8 items, Nos. 8 to 15) were evaluated preoperatively, at 3 months, 1 year, and more than 2 years postoperatively. SPSS17.0 and R 3.3 statistical software were used to analyze all the data by Spearman correlation analysis and hierarchical clustering analysis, while with the Spearman correlation coefficient 0.4 as the threshold, the edge was defined over or equal to 0.4. Results: A total of 232 consecutive patients with CSM treated by surgery were enrolled. Because there was no cervical spondylotic amyotrophy case, so “No. 2: Shoulder and Elbow” in mJOA score was totally deleted. Finally, 14 items were divided into 5 algorithms (Figure 1). Algorithm 1: Two items (role-physical and role-emotional in SF-36) were correlated at each follow-up time while there no correlation between preoperative and postoperative for each item. Algorithm 2: Miscellaneous correlation: scores from these 2 items (motor function of lower extremities in mJOA and physical-function in SF-36) in each time point were correlated with each other. Algorithm 3: The preoperative score was correlated with postoperative score (motor function of upper extremities, sensory of lower extremities, and bladder function in mJOA). The recovery profiles of these 3 items of mJOA were to the great extent decided by the preoperative score. Algorithm 4: The preoperative score (of sensory of upper extremities in mJOA) was not correlated with postoperative score while the postoperative scores were correlated with each other. The recovery profile of sensory of upper extremities was more based on the score at the first follow-up. Algorithm 5: The preoperative score was not correlated with postoperative score while the postoperative scores were correlated with each other (sensory of trunk, body-pain, vitality, social-function, mental health, and general health). Conclusion: Five meaningful algorithms of the objective and subjective outcome assessments in patients with CSM going through surgery intervention were identified. This result may represent a stepping stone toward a more thorough research with larger samples aimed at revealing the prognostic feature of CSM patients.

Figure 1.

Cluster analysis of all outcome evaluation items taking different follow-up time points into account.

A138: Effects of Regional Cervical Sagittal Alignment on Clinical Outcomes After Decompressive Surgery for Cervical Spondylotic Myelopathy

Hui Yu Koh¹, Kenneth Cheung¹, and Kenny Kwan¹

¹The University of Hong Kong, Hong Kong

Introduction: Cervical kyphosis is recognized as an important pathological factor in cervical spondylotic myelopathy (CSM) of myelopathy severity. However, it is not known if cervical sagittal alignment influences clinical outcome after decompression for CSM. This study aims to report on the relationship between regional cervical sagittal alignment and clinical outcomes after surgical decompression in patients with CSM. Methods: Consecutive patients undergoing surgical decompression for CSM with a minimum of 2-year follow-up were analyzed. Regional cervical sagittal parameters (C2-C7 lordosis, C7 slope, T1 tilt and C2-7 sagittal vertical axis [SVA]) were measured, and outcome assessments (Visual Analogue Scale [VAS], Finger Escape Sign [FES], 10 second test [10 s], Nurick score, Modified Barthel Index [MBI], and modified Japanese Orthopaedic Association [mJOA] score) were obtained pre- and postoperatively. Results: Fifty-two patients (34 males and 18 females) were included in the study. The mean age at the time of surgery was 64.5 years (range 37 to 86) and the mean follow-up period was 45.1 months (range 30 to 65). Sixteen had anterior, 35 had posterior, and 1 patient had combined approach. There was no statistical difference in regional cervical sagittal alignment irrespective of surgical approach pre- and postoperatively. Abnormal preoperative C2-7 SVA is associated with a higher C7 slope (P = .007), worse preoperative mJOA score (P = .002), worse MBI (P = .004), but a higher mJOA score recovery rate (P = .044). The mean mJOA scores improved postoperatively (9.95 ± 3.12 to 13.4 ± 2.27, P = .000) and 31 patients reached minimal clinically important difference (MCID) mJOA score improvement (mJOA ≥ 2). Preoperative C2-7 SVA is the only radiological parameter that was significantly associated with postoperative mJOA score improvement (P = .011). Other assessment outcomes were not affected by any of the sagittal parameters. Conclusion: This study looks at the relationship between regional cervical sagittal alignment and clinical outcomes in CSM patients after decompressive surgery. Abnormal preoperative C2-7 SVA was associated with higher C7 slope, worse mJOA score, worse MBI but a better mJOA score recovery. Other regional cervical sagittal parameters and clinical outcomes were not clearly related after CSM decompression surgeries.

A139: Predicting Outcomes After Surgical Decompression for Mild Degenerative Cervical Myelopathy: Moving Beyond the MJOA to Identify Surgical Candidates

Jetan Badhiwala¹, Laureen Hachem¹, Zamir Merali¹, Christopher Witiw¹, Farshad Nassiri¹, Muhammad Akbar¹, Saleh Almenawer², Markus Schomacher³, Jefferson Wilson¹, and Michael Fehlings¹

¹University of Toronto, Toronto, Ontario, Canada

²McMaster University, Hamilton, Ontario, Canada

³Charité University Hospital, Berlin, Germany

Introduction: Patients with mild degenerative cervical myelopathy (DCM) represent a heterogeneous population, and indications for surgical decompression remain controversial. We sought to dissociate patient phenotypes within the broader population of mild DCM associated with degree of impairment in quality of life (QOL) and surgical outcomes. Materials & Methods: This was a post hoc analysis of patients with mild DCM (modified Japanese Orthopaedic Association [mJOA] score 15 to 17) enrolled in the AOSpine CSM North America and CSM International studies. A k-means clustering algorithm (maximum 10 iterations) was applied to baseline SF-36 scores to separate patients into 2 clusters (k = 2) based on pattern and degree of impairment in QOL. Baseline variables (demographic, clinical presentation, and treatment characteristics) and surgical outcomes were compared between clusters. The primary outcome of interest was change in QOL (SF-36) at 1 year postsurgery compared to baseline. Secondary outcomes included change in mJOA, Nurick grade, Neck Disability Index (NDI), and SF-6D scores. A k-nearest neighbors (kNN) algorithm was used to evaluate the ability to classify patients into the 2 clusters by significant baseline clinical variables alone. Results: One-hundred eighty-five patients were eligible. Two groups were generated by k-means clustering over baseline QOL (SF-36) scores. Cluster 1 (“more impaired”) had a greater proportion of females (44% vs 28%, P = .029) and symptoms of neck pain (32% vs 11%, P = .001), gait difficulty (57% vs 40%, P = .025), or weakness (75% vs 59%, P = .041). While baseline mJOA correlated with neither baseline QOL nor outcomes, Cluster 1 was associated with significantly greater improvement in disability (NDI) (P < .01) and QOL (SF-36) (P < .05) scores following surgery. A kNN algorithm could predict cluster classification with 71% accuracy by neck pain, motor symptoms, and gender alone. Conclusions: Using an unsupervised machine learning approach, we identified a more impaired phenotype of patients, characterized by predominance of motor and neck pain symptoms and female gender, in whom surgical intervention resulted in greater gains. Identification of patient-related predictive factors may help guide surgical decision-making and inform treatment paradigms for mild DCM.

OP24: Pain and Clinical Evaluation

A140: Cervical Epidural Steroid Injections: Incidence and Determinants of Subsequent Surgery

John Kleimeyer¹, Jayme Koltsov², and Serena Hu²

¹Emory University, Atlanta, GA, USA

²Stanford University, Redwood City, CA, USA

Introduction: CESIs (cervical epidural steroid injections) are increasingly used in the nonsurgical management of cervical disorders despite narrowing indications. Patients may delay or avoid surgery following CESIs; however the rate and determinants of subsequent surgery remain uncertain. Therefore, we used a retrospective analysis of nationally representative administrative claims data to determine: (1) the proportion of patients having surgery following CESI and (2) the timing and factors associated with subsequent surgery. Material and Methods: Survival analyses were performed using the Truven Health MarketScan® databases from 2007 to 2015. Inclusion criteria were CESI for cervical disc herniation, stenosis, or radiculopathy, age ≥18, and active enrollment for 1 year prior to CESI to screen for exclusions. Patients were followed until they underwent cervical surgery or their enrollment lapsed. Rates of surgery were assessed with Kaplan-Meier survival curves. Factors associated with subsequent surgery were assessed with multivariable Cox proportional hazard models, α = 0.001 due to large sample size. Results: A total of 192 777 CESI patients were included (age 50.9 ± 11.3 years, 55.2% female). Within 6 months of CESI, 11.2% of patients underwent surgery, increasing to 14.5% by 1 year and 22.3% by 5 years. Ages 35 to 54 and male sex were associated with increased likelihood of subsequent surgery. Patients with radiculopathy were less likely to undergo surgery following CESI than stenosis or herniation; those with multiple diagnoses were more likely. Patients with comorbidities including CHF, other cardiac comorbidities, or chronic pain or northeast US region location were less likely. Some 40.2% of patients underwent >1 CESI, 84.6% within 1 year. Additional injections were associated with reduced rates of subsequent surgery. Conclusion: Following CESI, 22.3% of patients underwent surgery within 5 years. Multiple patient-specific risk factors for subsequent surgery were identified, and patients undergoing repeated injections were at lower risk. Determining which patients may progress to surgery will improve resource utilization and inform shared decision-making.

A141: Re-Irradiation With Spine Stereotactic Body Radiotherapy in a Heavily Pretreated Population

Jay Detsky¹, Timothy Nguyen¹, Sten Myrehaug¹, Hany Soliman¹, Chia-Lin Tseng¹, and Arjun Sahgal¹

¹ Sunnybrook Health Sciences Centre, Toronto, Ontario, Canada

Introduction: Spinal metastases develop in 40% of patients diagnosed with cancer. These patients have traditionally received conventional external beam radiotherapy (cEBRT), although the use of upfront high dose per fraction, highly conformal stereotactic body radiotherapy (SBRT) is becoming increasingly common in selected patients. Unfortunately, after cEBRT or SBRT, up to 20% of patients will develop clinical or radiographic progression necessitating further local treatment. Historically, re-irradiation has been limited due to concerns about toxicity to the spinal cord. This study presents outcomes after re-irradiation with spine SBRT in a heavily pre-treated cohort. Material and Methods: A retrospective review of a prospective database was undertaken of consecutive patients treated with spine SBRT following progression after previous cEBRT or SBRT. SBRT was delivered with courses between 2 and 5 fractions at doses from 24 to 35 Gy. The primary outcomes examined were overall survival (OS), local failure (LF), and the rate of subsequent vertebral compression fracture (VCF). Event rates were evaluated using the Kaplan-Meier method with Cox proportional hazards model to assess the hazard ratio (HR) of potential risk factors. Results: A total of 41 patients involving 83 spinal segments were included in this analysis. Prior to SBRT retreatment, 30 patients had undergone one course of cEBRT, 6 with prior SBRT, and 5 with 2 prior courses of cEBRT. The most common primary tumor sites were lung (27%), renal (22%), and breast (12%). Of the 83 spinal segments, 60% had epidural disease (36% high grade and 24% low grade) and 55% had paraspinal extension. The locations of the treated segments were thoracic (41%), lumbar (40%), cervical (11%), and sacral (8%). Median follow-up time was 12.5 months (range: 0.5 to 52.4). The median dose (EQD2) to 90% of the CTV was 28.7 Gy. The median EQD2 maximum dose to the spinal cord PRV was 22.5 Gy. The median OS was 14.1 months (95% confidence interval [CI] = 6.1-20.4). The OS at 1 and 2 years were 59% and 28%, respectively. OS was impacted by the presence of neurological symptoms (hazard ratio [HR] = 4.2) and brain metastases (HR = 2.9). Local failure was observed in 15 of 83 segments with a median time to LF of 8.2 months (range: 0.4-13.8). The rates of LF at 6, 12, and 24 months were 8%, 14%, and 19%, respectively. The primary pattern of failure was the development of new epidural disease. Three patients had subsequent VCF (1 symptomatic, 1 asymptomatic) at a median time after re-irradiation of 8 months. There were no cases of radiation myelopathy. Conclusion: Re-irradiation with spine SBRT is safe and effective even in a heavily pretreated population. Rates of local control and VCF with SBRT re-irradiation are similar to when SBRT is used in the upfront setting. A median OS of 14 months after retreatment highlights the need to optimize local control for recurrent spinal metastases.

A142: Duration of Preoperative Symptoms Does Not Affect Clinical Outcome After Lumbar Arthrodesis

Jannat Khan¹, Bryce Basques¹, Young Soo Hong¹, Jessica Gosse¹, Kevin Alter¹, Landan Banks¹, Ahish Chitneni¹, Matthew Colman¹, and Howard An¹

¹Rush University, Chicago, IL, USA

Introduction: Nonoperative measures are generally employed/exhausted prior to consideration of surgical intervention for degenerative lumbar spine pathology. As such, patients often experience symptoms for various lengths of time prior to surgical referral. It is unclear at what point surgical intervention may become less efficacious at relieving pre-operative symptoms. This study aims to examine the effect of symptom duration on clinical outcomes after posterolateral lumbar fusion for degenerative and isthmic spondylolisthesis. Materials and Methods: After institutional review board approval, a retrospective cohort study was conducted of consecutive patients undergoing primary elective open posterior lumbar spinal fusion at one academic institution from 2014 to 2018. Patients excluded from analysis: <18 years at surgery time, had fusion to treat a lumbar fracture, tumor, or infection. Patient and operative data were collected, and duration of symptoms measured in months and was dichotomized at the 50th percentile (24 months). Patient/operative characteristics were compared between symptom duration groups using χ² analysis and independent sample t-tests for categorical and continuous data. Preoperative and final postoperative Visual Analog Scale (VAS) Back pain, VAS Leg pain, Oswestry Disability Index (ODI) were collected. Preoperative, immediate postoperative, and final radiographs were assessed to measure lumbar lordosis (LL), pelvic tilt, pelvic incidence (PI), and the PI-LL difference was calculated. Rates of dural tear, postoperative complications, postdischarge destination, re-operation, achievement of minimally clinically important difference (MCID) for VAS back, VAS leg, and ODI, and pseudoarthrosis were collected. Binary outcome variables were compared with multivariate logistic regression, and continuous outcome variables using multivariate linear regression. Multivariate analyses controlled for baseline patient and operative characteristics. Results: A total of 650 posterior lumbar fusion patients were identified, 334 patients had preoperative symptoms for 2 or more years. There were no statistically significant differences between groups in terms of age, BMI (body mass index), diabetes mellitus, smoking status, psychiatric condition history, preoperative opiate use, and high ASA (American Anesthesiologists Association) score (3 or higher) (P > .05). There was a higher proportion of female patients among those who had pain for 2 years or more compared to patients who had pain for less than 2 years (P = .042). There was no difference in groups in terms of severity of spondylolisthesis (P > .05). Patients with longer preoperative symptoms had significantly smaller proportion of patients who underwent single-level fusions, and a higher proportion of patients who required 3 or more level fusions (P = .054 and .015, respectively). Both groups had similar changes in lumbar lordosis, pelvic tilt, and PI-LL difference from preoperative to final follow-up (P > .05). Both groups had no differences in preoperative VAS back, VAS leg, and ODI, as had similar improvements in all clinical outcome measures at final follow-up (P > .05). Patients with longer symptom duration had lower proportion of postoperative complication, but similar rates of dural tear, discharge to rehab facility, re-operation, and pseudarthrosis. Conclusions: Patients with pain for ≥2 years showed similar clinical and radiographic improvement after posterolateral lumbar arthrodesis than those with pain <2 years. Patients with long-standing preoperative symptoms for whom posterior lumbar fusion is considered can be counseled that they still have an opportunity to derive substantial clinical benefit from surgery.

A143: Evidence of Impaired Pain Modulation in Adolescents With Idiopathic Scoliosis and Chronic Back Pain

Alisson R. Teles¹, Don Daniel Ocay¹, Abdulaziz Bin Shebreen¹, Andrew Tice¹, Jean A. Ouellet¹, and Catherine E. Ferland¹

¹Shriners Hospitals for Children, Montreal, Quebec, Canada

Introduction: Although 40% of adolescent idiopathic scoliosis (AIS) patients present chronic back pain, the pathophysiology and underlying pain mechanisms remain poorly understood in this population. The aim of this study was to identify the presence of sensitization in nociceptive pathways and to assess the efficacy of the diffuse noxious inhibitory control in patients with AIS presenting chronic back pain. We hypothesized that the development of chronic pain in AIS is associated with alterations in pain modulatory mechanisms in these patients. Material and Methods: Ninety-four patients diagnosed with AIS and chronic back pain underwent a throughout pain assessment using standard and validated quantitative sensory testing (QST) protocol. The measurements included mechanical detection thresholds (MDT), pain pressure threshold (PPT), heat pain threshold (HPT), heat tolerance threshold (HTT), and a conditioned pain modulation (CPM) paradigm. Altogether, these tests measured changes in regulation of the neurophysiology underlying the nociceptive processes based on the patient’s pain perception. Results: Efficient pain inhibitory response was observed in 51.1% of patients, while 21.3% and 27.7% had suboptimal and inefficient CPM, respectively. Temporal summation of pain was observed in 11.7% of patients. Significant correlations were observed between deformity severity and pain pressure thresholds (P = .023) and CPM (P = .017), neuropathic pain scores and pain pressure thresholds (P = .015) and temporal summation of pain (P = .047), and heat temperature threshold and pain intensity (P = .048). Conclusion: Chronic back pain has an important impact in quality of life of adolescents with idiopathic scoliosis. Using a detailed pain assessment, we were able to demonstrate a high prevalence of impaired pain modulation in this group. The association between deformity severity and somatosensory dysfunction may suggest that spinal deformity can be the trigger for abnormal neuroplastic changes in this population contributing to chronic pain syndrome.

A144: Morphological and Biomechanical Approach to Understand Perioperative Back Pain in Patients With Adolescent Idiopathic Scoliosis

Maxime St-Georges¹, Alisson Teles¹, Jean Ouellet¹, and Catherine Ferland¹

¹McGill University, Montréal, Quebec, Canada

Introduction: Adolescent idiopathic scoliosis (AIS) is defined as a 3-dimensional deformity of the spine comprised of a 10° or more lateral curve.¹ It is an anomaly that progresses during the adolescent years, and it affects about 2% of the adolescent population and women are more at risk of a greater degree of curvature.^1-3 An abundance of literature investigates AIS and back pain but very little relationships between morphological aspects of the spine and pain felt is reported.^4,5 Literature also demonstrates that greater imbalance leads to greater pain and some studies have demonstrated certain significant relationships between AIS and imbalance.^2,6,7 Nevertheless, to our knowledge, no studies have explored morphology, pressure mapping, and pain all at once. The aim of this project was to create a clinical 3-dimensional (3D) biomechanical and morphological comprehensive assessment in order to better quantify perioperative back pain in patients with AIS. Materials and Methods: Twenty patients with AIS scheduled to undergo posterior spinal fusion surgery were recruited during the spine clinics at the Shriners Hospitals for Children, Canada. They were asked to take a head-to-toe X-ray in the EOS Imagery machine while wearing sensory insoles and standing upon a Cartesian plane placed at the bottom of said device before surgery, 6 weeks and 6 months postsurgery. Spines were reconstructed in 3D using the sterEOS software to obtain 3D parameters. Pain intensity and location were collected using a self-reported questionnaire involving a grid divided picture of the back, where patients indicated the location of any painful site and its intensity. Regression analyses to assess correlation between variables at the different time points and repeated measures to identify differences in time were completed by using SPSS. Results: Our results demonstrated that pain intensity was minimized after surgery. Patients indicated less spread pain 6 weeks after surgery compared to before. Pain intensity on the specific reported back regions before surgery was less 6 weeks after surgery, especially in the thoracic and thoracolumbar/lumbar regions (P = .085 and P = .004, respectively). Patients with a greater amount of pressure in the back of their feet before surgery were more likely to have greater pain (P = .025). No difference was observed in the pressure measurements taken before surgery and 6 weeks after surgery. The apical vertebral rotation of the thoracolumbar/lumbar region, the sum of torsion of the spine, and the sacral slope were reduced significantly 6 weeks after surgery (P = .034, P = .002, P = .030, respectively). Patients’ results at 6 months are currently being analysed and will be presented at the conference. Conclusion: These preliminary results suggest that spinal surgical correction in adolescents with idiopathic scoliosis causes a normalization of morphological changes that do not affect the biomechanics of the feet 6 weeks after surgery. The surgical correction alleviates pain, especially in the thoracolumbar/lumbar region 6 weeks after surgery. Acknowledgments: We would like to thank the whole clinical team for their precious collaboration, Moticon, The Shriners Hospitals for Children, and the AO Foundation for their financial support in making this project possible.

References

1. Nault ML, Mac-Thiong JM, Roy-Beaudry M, et al. Three-dimensional spinal morphology can differentiate between progressive and nonprogressive patients with adolescent idiopathic scoliosis at the initial presentation. Spine (Phila Pa 1976). 2014;39:E601-E606.

2. Dalleau G, Damavandi M, Leroyer P, Verkindt C, Rivard CH, Allard P. Horizontal body and trunk center of mass offset and standing balance in scoliotic girls. Eur Spine J. 2010;20:123-128.

3. Sullivan TB, Reighard FG, Osborn EJ, Parvaresh KC, Newton PO. Thoracic idiopathic scoliosis severity is highly correlated with 3D measures of thoracic kyphosis. J Bone Joint Surg Am. 2017;99:e54.

4. Smorgick Y, Mirovsky Y, Baker KC, Gelfer Y, Avisar E, Anekstein Y. Predictors of back pain in adolescent idiopathic scoliosis surgical candidates. J Pediatr Orthop. 2013;33:289-292.

5. Sieberg CB, Simons LE, Edelstein MR, et al. Pain prevalence and trajectories following pediatric spinal fusion surgery. J Pain. 2013;14:1694-1702.

6. Emami F, Yoosefinejad AK, Razeghi M. Correlations between core muscle geometry, pain intensity, functional disability and postural balance in patients with nonspecific mechanical low back pain. Med Eng Phys. 2018;60:39-46.

7. Pasha S, Baldwin K. Are we simplifying balance evaluation in adolescent idiopathic scoliosis? Clin Biomech. 2018;51:91-98.

A145: Patient-Reported Outcomes Measurement Information System (PROMIS) Pain Interference (PI) Is Superior to the Likert Pain Scale (LPS) for Pain Assessment in Spine Patients

David Bernstein¹, Matthew St. John¹, and Addisu Mesfin¹

¹University of Rochester Medical Center, Rochester, NY, USA

Introduction: An accurate understanding of a patient’s pain is beneficial in setting expectations and pain management goals. PROMIS is a universal, validated PRO (patient-reported outcome) tool that may capture clinical information more completely than other traditional methods. We aimed to determine whether PROMIS PI (pain interference) or the LPS (Likert Pain Scale) demonstrates a stronger association with physical function as determined by PROMIS Physical Function (PF). Materials and Methods: Spine center patients from February 2015 to November 2017 were asked to complete PROMIS PF and PI domains, as well as to report their pain level on a 0 (no pain) to 10 (worst pain) LPS at each visit. Pearson correlation coefficients were calculated between PROMIS PF and PROMIS PI; PROMIS PI and LPS; and PROMIS PF and LPS. Fisher r-z transformation method was utilized for confidence intervals and to determine significant correlation differences. Results: A total of 21 774 first visit, 11 130 second visit, 6575 third visit, 4202 fourth visit, and 2819 fifth visit patients’ data were recorded. PROMIS PF demonstrated a low significant correlation with the LPS over all visits (r = 0.46-0.49, P < .05). PROMIS PI demonstrated a moderate significant correlation with the LPS over all visits (r = 0.59-0.63, P < .05). PROMIS PF demonstrated a high significant correlation with PROMIS PI over all visits (r = 0.73-0.77, P < .05). Overall, PROMIS PI demonstrated significantly better correlation to self-reported physical function, as determined by PROMIS PF, than the LPS (P < .05). Conclusion: While PROMIS PI and the LPS both demonstrated significant correlation with self-reported physical function, PROMIS PI had a significantly stronger correlation. This suggests PROMIS PI may better capture the impact of pain on patient health than the LPS. This finding can assist surgeons in setting pain level expectations, which may help decrease opioid use and patient dissatisfaction. Furthermore, such information is important as health care shifts to a value-based medicine model with a focus on efficiency, as well as high-quality care at an affordable cost.

A146: PROMIS Pain Interference Natural Language: What Does Your PROMIS Pain Score Mean?

Jeremy Shaw¹, Ross McEntarfer¹, Natasha Greene¹, Angela Presson¹, Chong Zhang¹, Brandon Lawrence¹, William Ryan Spiker¹, Nicholas Spina¹, and Darrel Brodke¹

¹University of Utah, Salt Lake City, UT, USA

Introduction: Patient-reported outcome measures are an important tool for assessing health status across patient populations. Historically, commonly used outcome measures in spine care were narrow in scope and limited by the burden associated with their administration. The Patient-Reported Outcome Measurement Information System (PROMIS) was developed to overcome these limitations and provide a reliable and versatile testing system using item response theory and computerized adaptive testing. While PROMIS has been validated in the lumbar spine population, its use has been limited in part because many clinicians lack a common language understanding of the meaning and the significance of PROMIS scores. The purpose of this study was to develop plain language descriptions to apply to PROMIS Pain Interference (PI) scores. Methods: We retrospectively analyzed prospectively collected PROMIS Bank v1.1 Pain Interference (PROMIS-PI) data for patients presenting for back and/or lower extremity complaints. Patients with missing scores, standard error >0.32, missing injury location, and assessments with <4 questions or more than 12 questions were excluded. PROMIS-PI scores from each encounter were grouped into buckets based on non-overlapping confidence intervals. Common language descriptions based on published questions and answers were ed. Questions with fewer than 25 assessments were excluded from the bucket, as were those where fewer than 15% of assessments completed the question. The clinical tool consisting of three summary statements per bucket was developed based on non-overlapping confidence intervals. Discrepancies were reconciled via a clinical consensus process with 2 fellowship trained orthopedic spine surgeons. Results: In total, 14 823 assessments and 6582 unique patients were included in the analysis. A total of 94.3% of assessments were completed in 4 questions to generate the PROMIS-PI score. The number of assessments and unique patients were normally distributed. “Normal pain” was assigned a PROMIS-PI score of 50, with a higher number indicating increased pain interference. For the study cohort, the mean PI was 63.3 ± 7.4, median was 63 (interquartile range = 58, 67) and the range was 0 to 84. Conclusion: This study developed rational plain language descriptions to enhance both patient and physician understanding of the PROMIS Pain Interference domain. These data have broad utility in both the clinical and research settings. Additional work is required to evaluate patient and clinician perceptions of these novel natural language descriptions.

A147: Trends in Opioid Usage 2 Years Following Thoracolumbar Spine Surgery

Mina Aziz¹, Peter Jarzem¹, Michael Johnson², and Michael Weber¹

¹McGill University, Montreal, Quebec, Canada

²University of Manitoba, Winnipeg, Manitoba, Canada

Introduction: It is estimated that opioid overdoses claim the lives of 116 patients per day in the United States.¹ The underpinnings of this epidemic are multifactorial but as surgeons we have to consider what is our role in this public health crisis. As spine surgeons, we treat many patients with chronic pain. In 2015, Vowles et al estimated that 21% to 29% of patients misused opioids when they were prescribed for management of chronic pain.². A number of medical associations have produced guidelines and statements to address the issue of opioid prescribing in clinical practice. Most recently, the Canadian Orthopedic Association (COA) produced a position statement on “Opioids and Orthopedic Practice,” which was published in June of 2018. In this statement, the COA highlighted many strategies to help practitioners avoid opioid over-prescribing. The COA also stressed the importance of timely access to musculoskeletal care in order address the underlying cause of patients’ pain. The purpose of this study is to evaluate the effect of undergoing thoracolumbar spine surgery on patient’s opioid usage. We hypothesize that patients who undergo thoracolumbar spine surgery to address their underlying spine pathology will reduce their opioid usage and improve their quality of life on patient-reported outcome measures. Material and Methods: We conducted a retrospective review of prospectively collected data within the Canadian Spine Outcome Registry Network (CSORN) for patients who underwent thoracolumbar procedures by the McGill Spine group. All adult patients who underwent thoracolumbar procedures and completed their 2-year follow-up and had complete datasets were included in the analysis. Data was analyzed using the Statistical Analysis System (SAS) software. ANOVA was used to analyze continuous variable and statistical significance was set to P < .05. Results: Out of 46 patients identified, 45 patients met the inclusion criteria. Twenty-nine patients reported that they did not use opioids at the time of the initial assessment while 16 patients reported opioid usage. At 2-year follow-up, 35 patients reported that they do not use opioids. Of the 29 patient who were not using opioids at the time of the initial assessment, one patient reported being on opioids at 2-year follow-up. Of the 16 patients who were using opioids at the time of the initial assessment, 7 patients had completely stopped all opioid usage at the 2-year follow-up representing a 37.5% reduction in the number of patients using opioids. Furthermore, patients who underwent thoracolumbar surgery reported a reduction of 3.6 points, 3.4 points, and 17.8 points on the back-pain scale, leg pain scale, and the Oswestry Disability Index, respectively. Conclusion: Chronic pain and opioid misuse continue to be significant challenges that are affecting the quality of life of many of our patients. While our study is small in size, it clearly highlights the impact of addressing the patients’ underlying pathology on their pain scores, patient reported outcomes and opioid usage. In the future, the results from this study should be compared to those from other Canadian centers.

References

1. Center for Behavioral Health Statistics and Quality. 2016 National Survey on Drug Use and Health: Detailed Tables. Rockville, MD: Substance Abuse and Mental Health Services Administration; 2017.

2. Vowles KE, McEntee ML, Julnes PS, Frohe T, Ney JP, van der Goes DN. Rates of opioid misuse, abuse, and addiction in chronic pain: a systematic review and data synthesis. Pain. 2015;156:569-576. doi:10.1097/01.j.pain.0 000 460 357.01 998.f1

A148: Effect of Psychosocial Risk Factors on Global Treatment Outcome in Nonspecific Low Back Pain Patients Attending Group Physical Therapy

Julia Szita¹, Aron Lazary¹, Peter Pal Varga¹, Katalin Feher¹, and Ivan Lambertus¹

¹National Center for Spinal Disorders, Budapest, Hungary

Introduction: Nonspecific low back pain (NSLBP) can lead to a great societal and individual burden and loss of quality of life. In the development and the prognosis of the disease, psychosocial risk factors play an immense role. The purpose of the current study was to evaluate the characteristics of NSLBP patients and to investigate the effect of group physical therapy treatment on these specific features. Methods: A 3-month-long prospective follow-up study was conducted involving acute and chronic nonspecific low back pain patients (n = 70) referred to group physical therapy at an outpatient clinic. Previously established psychosocial risk factors leading to chronicity and/or poor outcome were collected via a questionnaire booklet. Standardized patient reported outcome measures were used to investigate the individual level of kinesiophobia (Tampa Scale for Kinesiophobia), pain catastrophizing (Pain Catastrophizing Scale), fear avoidance beliefs (Fear Avoidance Beliefs Questionnaire), quality of life (WHOQoL-BREF), pain intensity (VAS for LBP and Leg Pain). Global treatment outcome was measured on a 5-point Likert-type scale and then got dichotomized to good and poor outcome. Statistical analyses were performed to evaluate the differences in psychosocial risk factors between poor and good outcome groups and change of the mean scores following the treatment. Univariate logistic regression was applied to investigate the effect of the variables on global treatment outcome. Results: We found a significant improvement in the good outcome group at 3-month follow-up in the mean scores of Pain Catastrophizing Scale (PCS) subscales magnification, helplessness (P = .043 and .028, respectively) and WHOQoL-BREF physical and psychological subscales (P = .031 and .009, respectively), whereas none of the scores have significantly improved in the poor outcome group. Investigating the characteristics of the poor and good outcome groups, they significantly differed in Tampa Scale for Kinesiophobia (P = .048), Fear Avoidance Beliefs Questionnaire global score, physical activity and work subscale (P = .008, .024, and .046, respectively), and the WHOQoL-BREF physical activity subscale (P = .047) before the treatment. Univariate logistic regression analysis for poor and good outcomes showed that scoring poorly on the psychological subscale of the WHOQoL-BREF questionnaire was predictive for poor outcome at 3 months (P = .003, χ² = 8.745). Conclusion: We have found that NSLBP patients with poor outcome at 3-month follow-up were characterized by a significantly higher level of kinesiophobia, fear avoidance, and low level of quality of life regarding physical activity. In our study, pain catastrophizing (magnification, helplessness) and both of the psychological and physical quality of life significantly improved in the good outcome group after 3 months. As for predicting poor outcome impairment of psychological dimension of quality of life was found to have an unfavorable effect.

OP 25: Deformity-Thoracolumbar (Adult) 2

A149: Comparison of 3 Lumbopelvic Fixation Techniques in Long Fusion to the Sacrum in Osteoporotic Adult Spinal Deformity Patients (>60 Years): Clinical and Radiological Outcomes

Selhan Karadereler¹, Emel Kaya Aumann¹, Sinan Kahraman¹, Isik Karalok¹, Cem Sever¹, Yunus Emre Akman¹, Tunay Sanli¹, Meric Enercan¹, and Azmi Hamzaoglu¹

¹Istanbul Spine Center, Istanbul, Turkey

Introduction: Iliac screw techniques provide additional biomechanical stabilization to S1 in long fusion to the sacrum. Recently S2 alar-iliac screw (S2AI) fixation became more popular than adult spinal deformity (TIS; entry point posterior superior iliac spine) and distal iliac screw (DIS; entry point posterior inferior iliac spine) due to the lower implant profile and need for less extensive soft tissue dissection. S2AI fixes the sacroiliac joint differently from TIS and DIS. This study aimed to compare the clinical and radiologic outcomes between TIS, DIS, and S2AI fixation in osteoporotic adult spinal deformity patients Methods: A total of 158 (103 females, 55 males) adult spinal deformity patients who underwent long fusion to the sacrum with 3 different iliac screw techniques (TIS, DIS, S2AI) were reviewed. Radiologic parameters were compared between preoperatively and follow-up standing X-rays. Hospital charts and follow-up CT (computed tomography) scans were used to compare the screw lengths, diameters, and implant-related complications including S1 and iliac screw loosening, rod breakage, sacroiliac joint dissociation, surgical site infections, and hematoma. ODI (Oswestry Disability Index) scores were compared for clinical assessment Results: Fifty-four patients had TIS, 51 patients had DIS, 53 patients had S2AI fixation. Mean ages were 67, 71, and 68; mean follow-up was 54.2 (24-174) months, respectively. Iliac screw lengths were 85 mm, 90 mm, 105 mm; diameters were 9.0 mm, 8.5 mm, 9.5 mm for TIS, DIS, S2AI, respectively. There were 7 patients (12.9%) in TIS, 6 patients (11.7%) in DIS, and 3 patients(5.6%) in S2AI with implant-related complications. One patient (1.8%) in TIS group and 1 patient (1.9%) in DIS group showed sacroiliac joint dissociation (TIS = 1, DIS = 1, S2AI = 0). Postoperative hematoma was detected in 14 patients (8.8%) (TIS = 4, DIS = 7, S2AI = 3) and surgical site infections developed in 2 patients (1.2%) (TIS = 1, DIS = 1, S2AI = 0). ODI scores improved in all groups. Conclusion: All techniques provided sufficient stability for lumbosacral fusion in adult spinal deformity patients. S2AI technique showed lower implant-related complication rates when compared to both iliac screw techniques. TIS and DIS had higher rates of hematoma and surgical site infections due to extensive soft tissue dissection compared to S2AI. S2AI does not lead to sacroiliac joint dissociation as it provides stability by fixing the sacroiliac joint.

Table

Complication Rates of Lumbopelvic Fixation Techniques.

A150: The Fate of UIV, UIV+1 Compression Fractures After Long Level Spinal Fixation: What Should We Do? Observation Versus Revision surgery

Bang Sang Hahn¹, Kyung Hyun Kim¹, and Yong Eun Cho¹

¹Yonsei Medical College, Severance Hospital, Seoul, Republic of Korea

Introduction: Compression fractures of UIV, UIV+1 after long segment spinal fixation for adult spinal deformity are commonly detected on follow-up X-rays. It is important to recognize early specific characteristics of dangerous compression fractures that need to be surgically corrected in the future, because such conditions can result in severe neurologic deficits and poor clinical outcome. The purpose of this study is to investigate the early radiologic characteristics of UIV, UIV+1 compression fractures after long segment spinal fixation surgery. Material and Methods: A total 27 patients who presented newly developed compression fracture at UIV, UIV+1 after long segment spinal fixation (minimum 5 vertebral bodies, LIV of L5 or distal) were reviewed retrospectively. Patients were divided into 2 groups according to following management: revisional surgical correction (Group A, n = 13) and conservative care (Group B, n = 14). Pre- and postoperative images, and images taken shortly before and after the occurrence of fracture, were evaluated for radiologic characteristics. Results: Mean age was 73.3 ± 5.3 and 72.0 ± 4.5; mean BMD (bone mineral density) was −2.17 ± 0.9 and −1.97 ± 0.8; mean fused vertebral level was 7.3 ± 1.6 and 7.7 ± 2.2 for groups A and B, respectively (P > .05). Both groups showed SVA (sagittal vertical axis) correction after long segment spinal fixation (δSVA, A: 53.7 ± 73.8 mm, B: 51.0 ± 72.1 mm) without significant differences between groups (P > .05). Both groups showed improvement of PI-LL mismatch after long level spine surgery (A: from 34.5 ± 22 to 14.2 ± 13, P < .002; B: from 33.9 ± 24 to 15.5 ± 14, P < .012). In group A (n = 13), 10 patients (76%) ended up presenting weakness of lower extremities and especially 3 of them were paraplegic status before revision surgery. The other 3 patients (23%) in group A were neurologically stable and underwent revision surgery due to severe back or leg pain. There were no neurologic deteriorations in Group B patients. Wedging rate of compressed vertebral body and local kyphotic angle increased in both groups shortly after the occurrence of compression fracture, but without significant differences between groups (P > .05). Shortly after the compression fracture, Group A showed greater decrement of Disc height (P = .001) than Group B. Binary logistic regression indicated decrement of disc height was a risk factor of future revision surgery. Group A also showed more severe degeneration of the disc adjacent to compressed vertebrae, preoperatively (P = .004). Conclusion: UIV, UIV+1 compression fractures with preoperative disc degeneration and greater decrement of disc height at early stage were associated with future neurological deterioration and necessity of revisional corrective surgery.

A151: Decision Analysis in Quest of the Ideal Treatment in Adult Spinal Deformity Revisited: Is Surgery Still the Best at 2 Years and Adjusted MCID Values?

Emre Acaroglu¹, Selcen Yuksel², Sinan Bahadir³, Selim Ayhan¹, Alba Vila-Casademunt⁴, Ibrahim Obeid⁵, Francesco Sanchez Perez-Grueso⁶; European Spine Study Group⁴

¹Ankara Spine Center, Ankara, Turkey

²Yildirim Beyazit University, Ankara, Turkey

³Amasya University, Amasya, Turkey

⁴Hospital Universitari Vall d’Hebron, Barcelona, Spain

⁵ Bordeaux University Hospitals, Bordeaux, France

⁶ Hospital Universitari La Paz, Madrid, Spain

Introduction: Identification of the best treatment modality in adult spinal deformity (ASD) provides a challenge. Surgery (S) has been shown to yield better results compared to non-surgical methods (NS) by several studies using fixed minimum clinically important change (MCID) values for ODI (Oswestry Disability Index) as thresholds for improvement or deterioration. Recent studies, however, suggest that MCID values may vary significantly by the treatment modality (S vs NS). The objective of this study was to analyze the utilities and improvement rates provided by S and NS modalities in ASD at 2 years of follow-up, using ODI as the decisive parameter and treatment-specific MCID values as thresholds of improvement/deterioration. Material and Methods: The overall study population comprises 1452 patients with ASD from a multi-center international ASD database. Average patient age was 59.6 years; 1216 patients were female, 236 male. A total of 746 were S (mostly instrumentation and fusion ± decompression) whereas 706 NS (mostly observation). Of these, 546 S and 414 NS patients with 2-year follow-up were identified. S group was further subcategorized into no complication (N, 1259 patients), minor complications (Min, 103 patients), and major complications (Maj, 90 patients) groups to analyze the effect of complications on results. MCID values for ODI calculated by LCA analysis specific to ASD and S and NS treatments were used (14.31 in overall population, 14.96 in S population, 2.48 in NS population). Based on these, results were categorized as improved (improvement > MCID), unchanged, and deteriorated (worsening > MCID). Utilities, as measures of the disease (and treatment) burden (range 0 =worst to 1 = no burden) were calculated for each population and treatment modality based on VAS (Visual Analogue Scale) mapping. Finally, these data were incorporated in decision trees. Results: For utilities, S provided a higher value (0.583) than NS (0.549); hence, less burden. Utilities in S were sensitive to the presence of treatment complications, being 0.634, 0.564, and 0.497 in N, Min and Maj categories, respectively. In regard to treatment results, at the end of the second year, 38.3% of S patients were improved, 39.2% unchanged, and 22.5% deteriorated, whereas for NS patients, these values were 39.4%, 10.5%, and 50.1%, respectively. Results in S group were sensitive to complications with improvement rates of 40.1%, 39.3%, and 33.3% and deterioration rates of 19.2%, 22.5%, and 29.4% for N, Min, and Maj, respectively. Conclusion: Our results suggest that surgical treatment still has a lesser disease burden and a lesser chance of deterioration, but equal chances for improvement at the end of the second year follow-up. These results are different than similar reports in the past (using fixed MCID values) in that the improvement rates in NS, being reported as very poor previously, have now increased to the level of S treatment; and the effect of complications on results are better delineated. In summary, as surgery appears to be a better modality in the absence of any complications, future efforts need be directed to decreasing the complication rates of surgical treatment for better results in ASD.

A152: Iliac Screws Versus S2 Alar-Iliac Screws for Pelvic Fixation, a Meta-Analysis

Yun-Lin Chen¹, Wei-Hu Ma¹, and Nan-Jian Xu¹

¹ Department of Spine Surgery, Ningbo, China

Introduction: In pelvic fixation, the iliac screws (IS) technique can offer biomechanically stronger constructs than conventional techniques; however, the perioperative complications are problematic. Its disadvantages include screw prominence, wound dehiscence, and postoperative pain secondary to the screws. S2 alar-iliac (S2AI) screws are another technique for pelvic fixation. Several studies showed S2AI had less complication when compared to IS. Methods: Medline, Cochrane, and Embase were searched until December 2017 using the following keywords: iliac screw, s2 alar-iliac screws, and pelvic fixation. Non-RCT (randomized controlled trial) or RCT were included in the analysis. Outcome measures were clinical and radiographic complications. Results: Five retrospective studies involving 498 patients were included. There were 194 iliac screws and 304 s2 alar-iliac screws. The mean patient age ranged from 55.2 to 66.3 years, and the majority was female. S2AI was associated with less operation time (P = .04), less surgical site infection (P < .0001), less implant loosening (P = .002), less reoperation rate (P < .00 001), less wound dehiscence (P = .0008), less implant prominence (P = .0006) when compared to IS. However, there was no significant deference for average blood loss (P = .24), implant breakage (P = .21), and postoperative sacroiliac joint pain (P = .77). Conclusions: S2AI has less clinical and radiographic complications when compared to IS for pelvic fixation.

A153: Core Set of Outcome Measures for Adult Spinal Deformity Surgery: A Global Consensus-Based Approach

Sayf Faraj¹, Miranda van Hooff², Tsjitske Haanstra¹, Anna Wright³, David Polly⁴, Steven Glassman⁵, and Marinus de Kleuver¹

¹Radboud University Medical Center, Nijmegen, Netherlands

²Sint Maartenskliniek, Nijmegen, Netherlands

³Virginia Mason Medical Center, Seattle, WA, USA

⁴University of Minnesota, Minneapolis, MN, USA

⁵ Norton Leatherman Spine Center, Louisville, KY, USA

Introduction: Real-world evidence is playing an increasing role in health care decisions. Real-world data are collected by means of outcome registries, and such outcome registries are most valuable if outcome data are comparable between institutions and include outcomes that cover the full cycle of care. In ASD (adult spinal deformity) surgery, a lack of a standardized systematic approach toward outcome measurement and reporting exists, hindering optimal monitoring and comparison of the quality of care delivered in different settings globally. The aim of this study was to define a standardized set of patient-relevant outcomes, measurement instruments, and contributing risk and case-mix variables for ASD surgery. Methods: A systematic review of the ASD surgical literature was performed with the aim to identify ASD outcomes reported in the scientific literature. This review was used as input for a modified Delphi study. A geographically balanced panel of 25 ASD experts from the Scoliosis Research Society participated in multiple Delphi rounds to agree on a standard set of patient-relevant outcomes most essential to systematically monitor outcomes of ASD surgery. After consensus was reached about a core set of patient-relevant outcomes, consensus was reached on the appropriate combination of measurement instruments to capture the outcomes and contributing risk- and case-mix variables. Six consensus rounds were held; 5 web-based surveys and 1 face to face meeting. A threshold of 75% similar votes was used for consensus. Results: The core set encompasses patient-reported and surgical procedure-related outcomes covering the full cycle of care according to Porter and Teisberg’s 3-tiered hierarchy^1,2 including survival (30-day mortality), degree of health or recovery (aspects of functional status), time to return to recovery (time to return to work and achievement of functional status), disutility of care (operative mortality, return to operating room during hospital stay, surgical site infection, spinal cord injury, and nerve root injury), sustainability of health and nature of recurrences (30-day readmission and revision), and long-term consequences (implant breakage, pseudoarthrosis, progression of deformity, pain, and disability due to complication). Panelists agreed that the ODI v2.1a, SRS-22r, NRS (0-10) back and leg pain, EQ-5D-3L, and CEQ Expectation questionnaires are the most comprehensive and appropriate set of patient-reported measurement instruments that cover aspects of human functioning. Finally, contributing risk and case-mix variables related to demographics, preoperative status, and surgical procedure were agreed upon. Conclusions: The implementation of standardized outcome assessment facilitates comparison across registries and pooling of data, in order to improve the quality of daily clinical practice in ASD surgery and reach new avenues in predictive analytics. This will subsequently allow for global benchmarking of standardized risk-adjusted outcomes and ultimately lead toward value-based care.

References

1. Porter ME, Teisberg EO. Redefining Health Care: Creating Value-Based Competition On Results. 1st ed. Boston, MA: Harvard Business Review Press; 2006.

2. Porter ME. What is value in health care? N Engl J Med. 2010;363:2477-2481. doi:10.1056/NEJMp1011024

A154: Posterior Surgical Treatment for Moderate to Severe Thoracolumbar Focal Kyphosis: Minimum 3-Year Follow-up

Yan Zeng¹, Zhongqiang Chen¹, Xiaochen Qu¹, Weishi Li¹, and Chuiguo Sun¹

¹Peking University Third Hospital, Beijing, China

Introduction: To evaluate the middle-long term of radiological and clinical outcomes of the corrective surgery for patients with moderate to severe focal kyphosis in thoracolumbar spine. Material and Methods: Fifty-seven patients with moderate to severe focal kyphosis of the thoracolumbar spine underwent apical segmental resection osteotomy with dual axial rotation correction (pVCR) at our hospital. There were 30 male and 27 female patients. The mean age was 34.3 years. The kyphosis-level radiographs were obtained from each patient before surgery, immediately after surgery, and at follow-up. Local kyphosis and scoliosis Cobb angles were measured. Full-spine standing radiographs were obtained before surgery and at follow-up, and the spine sagittal and coronal balance were evaluated. The height of patients, the Frankel grading system for neurological functions, the Oswestry Disability Index for life quality, the Visual Analogue Scale score for back pain, and the Patient Satisfactory Index for satisfaction to surgery were applied before surgery and at follow-up. The radiological and clinical outcomes were further analyzed in different subgroups of patients according to etiology, severity of kyphosis, age, level of kyphosis apex, Frankel grade before surgery, and complications. Results: The average follow-up time of patients was 46.1 months. The average kyphosis angle reduced from 94.6° before surgery to 31.0° immediately after surgery, and remained at 34.4° at follow-up. The sagittal balance of the spine, height of patients, Frankel grading, Oswestry Disability Index, and Visual Analogue Scale score improved. The Patient Satisfactory Index (PSI) showed a satisfied rate of 91.2%. The correction rate was significantly higher in patients with kyphosis angle less than 95° and age less than 35 years. The clinical improvement rate was significantly higher in patient with kyphosis apex at lower thoracic spine or thoracolumbar segment (kyphosis apex at T10 or below), Frankel grade E before surgery, and no complication group. The incidence of intraoperative and early stage complications was 38.6%, and the incidence of instrumentation failure was 10.5%. The most severe complication was transient spinal cord injury, and the incidence was 7.0%. All complications got good relief after appropriate intervention. Conclusion: Apical segmental resection osteotomy with dual axial rotation correction is an effective procedure to treat moderate to severe focal kyphosis, and the prevention of serious neurological complications is fundamental to achieve the ideal clinical results.

A155: Anterior Surgery with Hyperlordotic Cages Is Novel for Correction of Adult Spinal Deformity

M. Burhan Janjua¹ and Vincent Arlet²

¹UT Southwestern Medical Center, Dallas, TX, USA

²Department of Orthopedic Surgery, Department of Neurosurgery, Philadelphia, PA, USA

Objective: Traditional correction for flat back syndrome is performed with posterior-based surgery or combined anterior-posterior surgeries or posterior-anterior-posterior surgeries in revision cases. The aim of this study was to evaluate the outcomes from anterior surgery with the use of hyperlordotic cages (HLCs) in patients with flat back syndrome. Materials and Methods: All patients operated for flat back syndrome with or without prior posterior lumbar surgery have been included in the study. Full-spine standing radiographs were obtained. Pre- to postoperative sagittal alignment was analyzed in all cases. Radiographic parameters were analyzed including T1PA, SVA, PT, PI, LL, SS, PI-LL, T4-12TK. Statistical analysis consisted of paired and unpaired t-tests and stepwise multivariate linear regression analysis. Results: Fifty patients (58 years old, 72% female with mean BMI [body mass index] was 28) demonstrated significant radiographic alignment difference in spinopelvic and global parameters from pre- to postoperative standing: LL (−37.04° vs −59.55°, P < .001), SS (35.12 vs 41.13, P < .001), PI-LL (23.55 vs 6.46), T4-12 TK (30.59 vs 41.67), PT (28.22 vs 22.13), SVA in mm (80.94 vs 37.39), T1PA (28.70° vs 18.43°, P < .001). Using linear regression analysis, predicted pre- to postoperative change in standing LL corresponded to pre- to postoperative changes in standing PI-LL mismatch, T1PA, TK, SS, PT, and SVA (R ² = 0.59, 0.38, 0.25, 0.16, 0.12, and 0.17, respectively). Five degrees of pre- to postoperative change in T1PA translate to −4.15° change in LL. Conclusions: Anterior surgery with hyperlordotic cages followed by posterior instrumentation is an efficient way to correct flat back syndrome. HLC is an effective and reliable tool to maximize lumbar lordosis up to 30°, which is equivalent in magnitude to a pedicle subtraction osteotomy, but associated with less blood loss, quicker recovery, lower complication rate, and good surgical outcome.

A156: The Effect of Increased Age on Perioperative Outcomes in Adult Idiopathic Scoliosis

Nathan Lee¹, James Lin¹, Joseph Lombardi¹, Alex Ha¹, Griffin Baum¹, Joseph Osorio¹, Simon Morr¹, Ronald Lehman¹, and Lawrence Lenke¹

¹Columbia University Medical Center, New York, NY, USA

Introduction: The surgical indication for adult idiopathic scoliosis (AdIS) is generally accepted to be curves greater than 50°, as they will continue to progress throughout life. However, the ideal timing to perform deformity correction in AdIS is not clear, especially in the well-functioning middle-aged adult. Surgeons have little objective data when counselling patients on whether to undergo surgery or delay treatment until they are older and more symptomatic. The purpose of this study is to determine the impact of age (>55 vs ≤55 years) on perioperative and short-term outcomes in operative AdIS patients. Material and Methods: Consecutive patients with AdIS undergoing spinal reconstruction by the senior author from 2015 to 2018 were identified. Patients were separated into 2 groups, age >55 years and age ≤55 years, and then compared across various clinical factors. Chi-square and t-tests were used to evaluate for statistical significance. Results: Seventy-one consecutive patients were identified, 36 were >55 and 35 were <55. The average age was 48.6 (±17.1) years. Patients older than 55 had more comorbidities (25.7% vs 8.6% with ASA of 3). Patients older than 55 had longer fusions (15.9 vs 12.4 levels, P < .0001), more fusions to pelvis (97.1% vs 45.7%, P < .0001), greater intraoperative blood loss (1459 vs 1064 mL, P = .010), required more intraoperative transfusions (470 vs 220 mL, P = .005), longer operative times (523 vs 424 minutes, P = .001), and longer hospital stays (6.9 vs 5.5 days, P = .003). No statistical differences were observed in rates of postoperative transfusions, pulmonary embolism, pneumonia, sepsis, postoperative ileus, deep venous thrombosis, and cardiac complications. Conclusion: AdIS patients older than 55 had a greater number of comorbidities than their younger counterparts. Older patients had longer fusion constructs, greater intraoperative blood loss, longer operative times, and longer hospital stays. Inpatient complications are low and similar between groups.

A157: Mixing Metals in Adult Deformity Constructs

Sahitya Denduluri¹, Jayme Kolstov¹, Chason Ziino¹, Nicole Segovia¹, Craig McMains², Jonathan Falakassa³, John Ratliff¹, Ivan Cheng¹, Todd Alamin¹, Serena Hu¹, and Kirkham Wood¹

¹Stanford Medical Center, Palo Alto, CA, USA

²Community Health Network, Indianapolis, IN, USA

³Southern California Orthopedic Institute, Van Nuys, CA, USA

Introduction: Direct contact between orthopedic implants of dissimilar metals has not been historically recommended due to the potential for galvanic corrosion, thereby increasing the risk for metal toxicity, infection, implant weakness, or pain/hypersensitivity to corrosion products. Titanium (Ti) screws are preferred in the spine because of improved MR (magnetic resonance) imaging. While Ti rods can be used, Ti rods are notch sensitive and therefore may not be ideal for long deformity constructs. Cobalt chrome (CoCr) rods are often selected due to assumed compatibility, with stainless steel (SS) rods a less commonly used alternative. Though in vitro and in animal studies have demonstrated the low rate of galvanic corrosion in these mixed-metal constructs, no studies to date have evaluated rod fractures or other implant-related complications in patients. Material and Methods: Patients with adult spinal deformity (ASD) who underwent at least a 5-level thoracic and/or lumbar posterior instrumented fusion between January 2013 and May 2015 were reviewed, excluding those with neuromuscular deformity, tumor, or infection. Implant-related complications were defined as pseudarthrosis, proximal junctional kyphosis (PJK), or hardware failure (rod fracture or screw pullout in the absence of pseudarthrosis or PJK). Results: Of 77 cases reviewed, 28 patients were treated with Ti rods (36%), 41 with SS rods (53%), and 8 with CoCr rods (10%). Ti screws were used in all patients. Median follow-up was between 12 and 20 months for all groups. There were no differences between groups with regard to age, body mass index, or proportion of smokers. All CoCr patients underwent primary surgeries, whereas the SS and Ti groups contained both primary and revision procedures (P = .011). Implant-related and non–implant-related complications did not differ among the 3 rod-screw combinations (P = .122 and .407, respectively). Among the Ti group, there were 12 implant-related complications (43%): 6 pseudoarthroses, 3 PJK, and 3 rod fractures. In the SS group, there were 8 implant-related complications (20%): 5 PJK and 3 pseudoarthroses. There were 2 implant-related complications in CoCr group (25%), both involving pseudoarthrosis. Conclusion: To our knowledge, this is the first study to compare implant-related complications between mixed-metal and same-metal constructs in multi-level posterior fusions. We have shown in this single-institution study that the former is a clinically safe and scientifically valid option for ASD.

OP26: Tumor 3

A158: Association Between Body Fat and Muscle Distribution and Overall Survival in Patients With Bone Metastases Receiving Radiotherapy Treatment: A Prospective Cohort Study

Bart Pielkenrood¹, Paul van Urk², Joanne van der Velden¹, Nicolien Kasperts¹, Joost Verhoeff¹, Gijsbert Bol¹, Helena Verkooijen², and Jorrit-Jan Verlaan²

¹University Medical Center Utrecht, Utrecht, Netherlands

²Diakonessenhuis, Utrecht, Netherlands

Introduction: An increasing number of patients is diagnosed with spinal metastases due to increased overall cancer survival rates. These patients might need radiotherapy with or without surgical stabilization. Patients with a life expectancy of less than 3 months are deemed unfit for surgery. Therefore, accurate estimation of survival is crucial. The aim of this study was to evaluate whether abdominal fat and muscle distribution and muscle attenuance are associated with overall survival. Materials and Methods: Within the UMC Utrecht PRESENT cohort, we identified all 310 patients who received radiotherapy for thoracic or lumbar spinal metastases. Using an in-house developed delineation tool (VolumeTool), a single transverse slice at the L3 vertebra was used to assess the variables of interest, that is, visceral fat area, subcutaneous fat area, total muscle area, and muscle attenuance. Muscle attenuation is defined as the density of the muscle, which decreases when the lipid content is increased. Subsequently the ratio between visceral and subcutaneous fat was calculated. Univariable and multivariable logistic regression analyses were performed to determine the association between our variables of interest and survival at 90 and 365 days, adjusted for potential confounders (age, sex, primary tumor, Karnofsky performance scale, number of bone metastases, non-bone metastases, and neurological symptoms). Results: Median age was 67 years; 63% were male. The most common primary tumors were lung (28%), followed prostate (27%) and breast (18%). Median follow-up was 202 days. Overall survival rates at 90 and 365 days were 71% and 36%, respectively. In the univariable analysis, the subcutaneous fat area, fat ratio, and muscle attenuance were significantly associated with survival at 90 and 365 days. In multivariable analysis, for both 90 and 365 days survival, only muscle attenuance was significantly associated with overall survival (hazard ratio [HR] = 0.89, 95% confidence interval [CI] = 0.85-0.94, and HR = 0.92, 95% CI = 0.89-0.96, respectively). Conclusions: Accurate prediction of survival is crucial for optimal treatment selection of patients with spinal metastases. This study showed that muscle attenuance was significantly and independently associated with overall survival, while body fat and muscle distribution was not. Routinely available computed tomography scans contain useful prognostic information that may contribute to optimized care.

A159: Developing Therapeutic Guidance for Potentially Unstable Spinal Metastases by Evaluating Health-Related Quality of Life Outcomes for Surgery and/or Radiotherapy

Anne Versteeg¹, Arjun Sahgal², Laurence Rhines³, Daniel Sciubba⁴, James Schuster⁵, Michael Weber⁶, Aron Lazary⁷, Stefano Boriani⁸, Chetan Bettegowda⁴, Michael Fehlings⁹, Michelle Clarke¹⁰, Paul Arnold¹¹, Ziya Gokaslan¹², and Charles Fisher¹³

¹University Medical Center Utrecht, Utrecht, Netherlands

²Sunnybrook Odette Cancer Centre and University of Toronto, Toronto, Ontario, Canada

³ The University of Texas MD Anderson Cancer Center, Houston, TX, USA

⁴ Johns Hopkins University, Baltimore, MD, USA

⁵ Hospital of the University of Pennsylvania, Philadelphia, PA, USA

⁶ McGill University and Montreal General Hospital, Montreal, Quebec, Canada

⁷ National Center for Spinal Disorders and Buda Health Center, Budapest, Hungary

⁸ IRCCS Istituto Ortopedico Galeazzi, Milan, Italy

⁹ University of Toronto and Toronto Western Hospital, Toronto, Ontario, Canada

¹⁰Mayo Clinic, Rochester, MN, USA

¹¹The University of Kansas Hospital, Kansas City, KS, USA

¹²Brown University and Rhode Island Hospital and The Miriam Hospital, Providence, RI, USA

¹ University of British Columbia and Vancouver General Hospital, Vancouver, British Columbia, Canada

Introduction: Understanding health-related quality of life (HRQOL) outcomes and adverse effects of different treatment modalities for patients with potentially unstable spinal metastases informs treatment decision-making. The objective of this study was to evaluate HRQOL outcomes at 12 weeks posttreatment in patients with potentially unstable spinal metastases treated with either surgery ± radiotherapy or radiotherapy alone. Materials and Methods: An international multicenter prospective observational cohort study was performed. Demographic, primary tumor, treatment, adverse event, and HRQOL data were collected. The NRS pain score, EQ-5D (3L), SF-36v2, and Spine Oncology Study Group Outcome Questionnaire (SOSGOQ2.0) were used to evaluate HRQOL. A Spinal Instability Neoplastic Score (SINS) score between 7 and 12 defined potentially unstable spinal metastases. HRQOL scores were modeled and compared within and between treatment groups using mixed effect models with adjustment for baseline characteristics. Results: A total of 202 patients with potentially unstable spinal metastases were identified, 120 patients were treated with surgery ± radiotherapy and 82 with radiotherapy alone. Patients who underwent surgery ± radiotherapy more frequently exhibited mechanical pain (P < .001), a lytic tumor (P = .031), a greater median SINS score (10 vs 8 for radiotherapy alone), and worse baseline performance status, HRQOL, and pain scores. The combination of having mechanical pain and a vertebral compression factor was more often observed in patients treated with surgery ± radiotherapy compared to treatment with radiotherapy alone. At 12 weeks posttreatment, surgically treated patients experienced a 2.7-point decrease in adjusted mean NRS pain (95% confidence interval [CI] = −3.9 to −1.5, P < .001) and a 12-point increase in adjusted mean SOSGOQ2.0 (95% CI = 5.1 to 19.0, P < .001) as compared to a 1.4-point decrease in NRS pain (95% CI = −2.9 to 0.0, P = .053), and a 6.2-point increase in SOSGOQ2.0 (95% CI = 2.2 to 14.6, P = .352) in those treated with radiotherapy alone. Conclusion: The significant improvements observed in pain and HRQOL scores after surgery ± radiotherapy, as compared to the moderate improvements after treatment with radiotherapy alone, reflect appropriate surgical treatment selection and provide therapeutic guidance. Surgical consultation for patients with a SINS from 7 to 12 with the presence of mechanical pain and a vertebral compression fracture is recommended. This study was funded by an Orthopaedic Research and Education Foundation grant and by AOSpine International through the AOSpine Knowledge Forum Tumor.

A160: Cement Augmentation Procedures in Metastatic Spinal Disease: Is There Increased Dissemination of Tumor Cells Into the Circulatory System?

Moritz C. Deml¹, Carlo R. Largiadèr², Daniel Schärer², Lorin M. Benneker¹, Christoph E. Albers¹, Sven Hoppe¹, Sebastina F. Bigdon¹, and Helena Milavec¹

¹Bern University Hospital, Bern, Switzerland

²Bern University, Bern, Switzerland

Introduction: With an increasing survival of cancer patients suffering from metastatic spinal disease, surgical procedures such as vertebroplasty (VP) and kyphoplasty (KP) have become more rampant. Iatrogenic spreading of tumor cells during tumor surgery may lead to dissemination of the disease and therefore remain a major concern of tumor surgery. We aim to evaluate the effect of cement augmentation onto systemic tumor cell dissemination and to investigate the kinetics of the appearance of circulating tumor cells (CTCs) during VP/KP. Materials and Methods: We prospectively collected blood samples of 11 consecutive patients suffering from either breast (n = 7, 63.6%), prostate (n = 3, 27.3%) or colon (n = 1, 9.1%) cancer, with spinal involvement undergoing KP (n = 10, 90.9%) or VP (n = 1, 9.1%) at our institution between January 2017 and July 2018. We collected patient data as well as treatment details for primary tumor care. The indication for surgery was either intractable pain or instability determined according to the instability criteria of the “Spinal Instability Neoplastic Score” (SINS) with a threshold value of ≥8. The mean SINS was 12.2 (range 8-15). In all patients, intraoperative transpedicular biopsies were obtained for histologic confirmation of the diagnosis. The mean age was 66.1 (range: 39-77/SD: 10.6) years and 8 patients (72.7%) were female. Blood samples were collected from either a central or peripheral venous catheter (n = 6/5; 54.5/45.5%), before starting the procedure and at 4 different time points during the procedure (1, 5, 10, and 20 minutes after cement application). One patient was excluded from the study due to clotted blood samples leaving 10 patients for the analyses. The samples were analyzed with the RT-PCR-ADNA test (Qiagen), which is a qualitative immunomagnetic-cell-selection-system for enriching circulating tumor cells from full blood. It includes a highly sensitive RT-PCR (reverse transcriptase polymerase chain reaction) for detecting breast, prostate, and colon cancer-specific mRNA markers (Actin: internal PCR control; GA733-2, MUC-1, Her-2: breast cancer; PSMA, PSA, EGFR: prostate cancer; GA733-2, CEA, EGFR: colon cancer). All patients received local radiotherapy after the spinal procedure. Results: All tumor were histopathologically confirmed but one, in which the biopsy for prostate cancer was negative. In this case, no mRNA markers could be detected at any time point. In 9 (100%) out of 9 cases with positive histopathologic proof for the tumor, specific mRNA markers for the individual tumor entity were observed in the blood circulation 5 and 10 minutes after cement application. After 1 and 20 minutes, specific m-RNA markers were detected in 5 (55.6%) patients. Conclusion: Cement augmentation of bony spinal metastases cause CTC release into the circulatory system. The highest concentration of CTCs was measured 5 and 10 minutes after cement application. After 20 minutes, a relative drop in CTC markers was observed and thus we conjecture that the optimal time for blood sampling for detecting CTCs would be between 5 and 10 minutes after cement augmentation. Further investigations on cell viability and prophylactic strategies to reduce the increased release of CTCs are needed to account for these findings.

A161: Development and Validation of Clinical Prediction Models of Survival and Clinical Outcomes for Patients With Metastatic Epidural Spinal Disease: A Systematic Review

Anick Nater¹, Jetan H. Badhiwala¹, James Hong², So Kato³, Melanie Anderson⁴, and Michael Fehlings¹

¹University of Toronto, Toronto, Ontario, Canada

²Toronto Western Research Institute, University Health Network, Toronto, Ontario, Canada

³University of Tokyo, Tokyo, Japan

⁴ University Health Network, Toronto, Ontario, Canada

Introduction: In multivariable prognostic research, the development and external validation are the first phases typically involved toward the establishment of clinical prognostic models (CPMs) in practice. This systematic review aims to identify and assess CPMs created to predict clinical outcomes in patients with metastatic epidural spinal disease (MESD) and their subsequent validation studies. Material and Methods: Three electronic databases were searched (January 1, 1990, to December 31, 2017), without language restriction, to identify studies that addressed the following review question: What are the existing CPMs that have been developed and/or externally validated to predict survival or other clinical outcomes in patients with MESD? Data extraction, reporting, and appraisal of the selected studies were conducted following recommended guidance: CHARMS, TRIPOD, and PROBAST (CRD42017072908). Results: Among 8077 unique full-text articles, 117 were included. Among the 52 articles describing a CPM (CPM creation, n = 44; update of an existing CPM, n = 8), 44 did not include any assessment of model performance (calibration and/or discrimination) while 20 reported the number of outcome events and 7 discussed missing data. Among the 5 articles with the term “external validation” or “external validity” in the title or , missing data, number of outcome events, and both calibration along with discrimination were discussed in 4, 3, and 2 studies, respectively. Conclusion: Since 1990, while over 50 CPMs predicting clinical outcomes in patients with MESD were developed, only 5 studies claimed performing an external validation of any of these tools. The majority of the studies included in this review did not report on key methodological and data analysis elements. Greater rigor in the development and validation of CPMs could promote the establishment of CPMs in clinical practice in this patient population.

A162: Recalcification in Lytic Bone Metastases of the Spine After Radiotherapy

Bart Pielkenrood¹, Thomas Visser², Wouter Foppen¹, Joanne van der Velden¹, Wietse Eppinga¹, Nicolien Kasperts¹, Gijsbert Bol¹, Helena Verkooijen¹, and Jorrit-Jan Verlaan¹

¹University Medical Center Utrecht, Utrecht, Netherlands

²Utrecht University, Utrecht, Netherlands

Background: Due to an increased survival of cancer patients, the incidence of bone metastases has increased. Osteolytic bone metastases may cause fracture, pain, instability, and spinal cord compression. It has been reported that radiotherapy (RT) may induce recalcification in these lesions. It is not completely clear, however, which factors are associated with recalcification. The aim of this study was to investigate the change in bone mineral density in spinal metastases after RT. Materials and Methods: Within the UMC Utrecht PRESENT cohort we identified all patients who received radiotherapy for lytic spinal metastases. Patients were included if computed tomography (CT) scan was available pre- and post-RT. Bone density of the metastases was measured in Hounsfield units (HU). A preset region of interest (ROI) was drawn manually in each metastatic lesion. As a reference, a measurement of bone density in vertebrae without a lesion that did not receive RT was used. Factors tested for association were origin of the primary tumor, RT dose and scheme, and use of bisphosphonates. Change in bone density was analyzed using the paired t-test, differences between the treated lesions and the reference value was analyzed using the Welch t-test. Factors associated with the recalcification were tested with linear regression analysis. Results: A total of 119 patients with a lytic spinal metastasis were identified. Because of unsuitable or missing baseline or follow-up scan, osteosynthesis material in the ROI, or a fracture in the ROI, 86 patients were excluded. In 33 patients, 51 lesions were identified. The median age at baseline was 60 years (interquartile range [IQR]: 54-64), and the median follow-up was 8.2 months (IQR: 3.2-17.8). Of the available follow-up scans within 3 months (n = 21), the difference between baseline and follow-up in the metastatic lesions was 103.0 HU (95% confidence interval [CI]: 55.1-150.8; P < .001). At baseline, the difference in HU between the reference vertebra and the lesion was 172.5 HU (95% CI: 110.1-234.7; P < .001). At 3 months, the difference between the reference vertebrae and the lesion decreased to 61.6 (95% CI: −15.3 to 138.4; P = .178). Taking all first follow-up scans into account, the mean difference in HU in the metastatic lesions between baseline and first follow-up was 39.1 HU (95% CI: −25.5 to 103.7; P = .228). For all first-follow-up scans, the mean difference between the reference value and the lesions declined to 108.3 HU (95% CI: 15.7-200.9; P = .057). When patients used bisphosphonates during the RT, an increased difference in HU was measured compared to patients who were not using bisphosphonates (192.8 vs 53.8 HU, P = .008). Conclusion: In this confined study of 33 patients with lytic spinal metastases, we found that recalcification could be induced by radiotherapy. To confirm the effect of radiotherapy and other factors, a study with an increased sample size should be considered.

A163: A Phase I Trial on the Use of Photodynamic Therapy in Vertebral Metastases

Carl Fisher¹, Zakariya Ali², Arjun Sahgal², Elizabeth David², Edward Chow², Cari Whyne², Shane Burch³, Brian Wilson¹, amd Albert Yee²

¹Princess Margaret Cancer Centre, University Health Network, Toronto, Ontario, Canada

²Sunnybrook Health Sciences Centre, Toronto, Ontario, Canada

⁴University of California–San Francisco, CA, USA

Introduction: The spine is a common site of metastasis. Complications include pathologic fracture, spinal cord compression, and neurological deficits. Vertebroplasty (VP) and balloon kyphoplasty (KP) are minimally invasive stabilization procedures used as a palliative treatment to improve mechanical stability, quality of life, and reduce pain. Photodynamic therapy (PDT) is a tumor-ablative modality that may complement mechanical stability afforded by VP/KP. This first-in-human study evaluates PDT safety when applied in conjunction with VP/KP. Material and Methods: This dose escalation trial involved 1 light-only control group and 4 light-drug doses (50 100 150 200 J; n = 6) delivered at 150 mW from a 690 nm diode laser by 800-micron optical fibers prior to KP/VP. Patients eligible for VP/KP in treating pathologic fracture or at-risk lesions at a single level were recruited. Exclusion criteria included spinal canal compromise or neurologic impairment. PDT is a 2-step binary therapy of systemic drug followed by intravertebral light activation. Light was applied via bone trochar prior to cementation. This study used a benzoporphyrin derivative monoacid (BPD-MA), Verteporfin (Visudyne), as the photosensitizer drug in the therapy. Drug/light safety, neurologic safety, generic (SF-36), and disease-specific outcomes (VAS, EORTC-QLQ-BM22, EORTC-QLQ-C15-PAL) were recorded through 6 weeks. Phototoxicity and the side effects of the BPD-MA were also examined following PDT use. Results: Thirty (10 male, 20 female) patients were treated (13 KP, 17 VP). The average age was 61 and significantly different between genders (male 70 years vs female 57 years; P < .0005). Primary cancer sites were breast (36.7%) and lung (23.3%). All patients completed the study through the 6-week follow-up. No dose group showed significant increases in pain as defined by the EORTC-QLQ questionnaires. The 50 J and 100 J groups showed significant reductions in pain compared to the control (P < .05). The 50 J group had the best response; the group was comprised mostly of lytic tumors, an average power density of 12.1 mW/cm² measured at various distances ranging from 1.2 to 2.4 cm, and 5 out of 6 lesions located from L2-L5. Dosimetry data were collected on 9 patients with 2 patients recording no usable data. Of the remaining 7 patients, fluence values ranged from 0.72 to 33.90 mW/cm in a mixture of lytic and mixed blastic/lytic tumors. No significant correlation was found between fiber distance and fluence (P > .05), and tumor status (lytic vs mixed blastic/lytic: P > .05). In most cases, fluence rates were similar throughout PDT treatment time, indicating a relatively stable treatment. Twelve (40%) patients experienced complications during the study, none of which were attributed to PDT therapy. This included 2 kyphoplasty failures due to progression of disease, 1 case of shingles, 1 ankle fracture, 1 prominent suture, 1 case of constipation due to a lung lesion, 1 case of fatigue, and 5 patients experienced pain that was surgically related or preceded therapy. Conclusion: Vertebral PDT appears safe from pharmaceutical and neurologic perspectives. KP/VP failure rate is broadly in line with reported values and PDT did not compromise efficacy. The 50 J group demonstrated an improved response. Ongoing studies determining safe dose range and subsequent efficacy studies are necessary.

A164: Delay Resulting From Additional Referrals in Patients Requiring Surgical Treatment for Spinal Metastases

Floris van Tol¹, F. C. Oner¹, and J. J. Verlaan¹

¹University Medical Center, Utrecht, Netherlands

Introduction: Symptomatic spinal metastases occur in approximately 10% of all oncological patients. Patients with mechanical instability or neurological deficits often require surgical treatment, while others may be treated conservatively. For timely arrival at the appropriate and definitive caregiver, patients rely on adequate diagnostics and referrals earlier in the referral chain. This study investigates referral patterns in secondary and tertiary care centers in the Netherlands in patients with spinal metastases. Materials and Methods: Data were collected on all patients receiving surgical treatment for symptomatic spinal metastases between March 2009 and June 2017 within our tertiary care center. The number of successively involved health care providers prior to the diagnosis and the destination of referral after the diagnosis were noted separately for secondary and tertiary care centers. Diagnosis delay was defined as the time between first medical consultation and the diagnosis. Results: In total, 223 patients were diagnosed in secondary care centers and 52 patients were diagnosed in tertiary care centers. In secondary care centers, median diagnosis delay was 5 days for patients being diagnosed at the first involved health care provider, 27 days for patients requiring one additional referral, and 81 days for patients requiring 2 additional referrals. In tertiary care centers, these intervals were 1 day, 26.5 days, and 10 days, respectively. In secondary and tertiary care centers, 57.8% and 80% of patients were directly referred to their definitive caregiver after diagnosis, respectively. Conclusion: In both secondary and tertiary care centers, there is an exponential relationship between the number of consecutively involved health care providers and total diagnosis delay. To minimize diagnosis delay, diagnostics should be performed immediately upon suspicion of spinal metastases rather than referring the patient to another specialty. Also, a relatively large proportion of patients was not referred to the appropriate definitive caregiver after diagnosis. Prior to the actual referral, definitive caregivers should be consulted to prevent further delay due to inappropriate referrals.

A165: Chondrosarcoma of the Spine: A Single Tertiary Spinal Oncology Center Experience

Muaaz Tahir¹, Mel Grainger¹, Petr Rehouzek¹, Simon Hughes¹, Alistair Stirling¹, Morgan Jones¹, and Marcin Czyz¹

¹The Royal Orthopaedic Hospital NHS Trust, Birmingham, UK

Introduction: Chondrosarcoma is an aggressive malignant bone neoplasm that accounts for almost 10% of all primary bone tumors. Nearly 10% of all chondrosarcomas occur in the spine. Due to its resistance to both radiation and chemotherapy, the mainstay of treatment is radical surgical excision. However, en bloc excision in the spinal tumors is complex with the literature reporting significant rates of complications and local recurrence in the order of 30% to 40%. Performing this high-risk and technically demanding procedures in dedicated spinal oncology centers may decrease risk of perioperative morbidity and mortality increasing chance for the cure. This study aims to investigate clinical outcomes in a large single-center cohort of patients undergoing treatment for spinal chondrosarcomas. Materials and Methods: We retrospectively reviewed a prospectively collected database containing all spinal chondrosarcoma cases treated at our institution between 2006 and 2016. Patient records including demographics, clinical letters, operative notes, and pathology reports were analyzed in an attempt to correlate treatment strategies with outcomes. Results: Thirty-one cases were identified in total, 27 involving the mobile spine (4 cervival, 14 thoracic, 9 lumbar) and 4 involving the sacrum or pelvis. There were 23 males and 8 females, median age at diagnosis was 49 (range 21-76). Twenty-three patients were treated surgically; 17 had primary en bloc resections whereas the remaining 6 patients underwent intralesional procedures. Of those who underwent en bloc excision only 2 patients had a recurrence (12%) and 2 suffered a postoperative complication, both wound infections (12%). There were no cases of surgically induced neurological deficits, severe medical complications, or perioperative mortality. Two patients suffered surgical wound infection (12%). Three of 31 patients died (1 had undergone en bloc resection), their mean interval to death was 11 months (range 2-26). One- and 5-year survival rates in this series were 94% and 59%, respectively. The en bloc group had a much higher 5-year survival rate of 88%. Conclusion: En bloc resection, with wide or marginal histologic margins, offers potential curative surgery for chondrosarcoma. Intralesional excision almost always results in recurrence; therefore, optimal planning and surgical technique in order to avoid breaching the pseudocapsule is paramount. We recommend surgery take place in designated and specialist spinal oncology units in order to improve outcomes and reduce complication rates.

A166: Update on Custom-Made 3D Printed Titanium Implants for Anterior Column Reconstruction Following En Bloc Resection for Spinal Tumors

Marco Girolami¹, Giovanni Barbanti-Brodano¹, Stefano Bandiera¹, Riccardo Ghermandi¹, Silvia Terzi¹, Gisberto Evangelisti¹, Giuseppe Tedesco¹, Valerio Pipola¹, and Alessandro Gasbarrini¹

¹IRCCS Istituto Ortopedico Rizzoli, Bologna, Italy

Introduction: Reconstruction of the segmental defect after en bloc resection for spinal tumors aims at immediate stability and secondary solid fusion. The present communication provides an update on the results of the ongoing study on use of 3D printed custom-made implants for anterior column reconstruction. Methods: In 17 patients submitted to en bloc resection for spinal tumor between November 2015 and June 2018 at the same institution, anterior column reconstruction was done using 3D printed custom-made implants. Resection was planned according to Enneking and Weinstein-Boriani-Biagini staging systems. Implants were designed according to the preoperative planning of the resection on computed tomography (CT) scan. Results: At an average 18 months follow-up (range 1-28), 1 major mechanical complication occurred and 1 implant was replaced due to recurrence of the disease. Mechanical complication consisted in a massive subsidence into the adjacent vertebral body occurred with development of progressive distal junctional kyphosis. Critical analysis of the construct revealed insufficient posterior instrumentation, but custom-made implants itself did not show postoperative mechanical complications (breakage or migration of the implant). However, leading this to surgical revision of the construct, it was considered a major mechanical complication. The removed implant was processed, and sectioned for histology study that revealed new bone formation into the implant. Conclusions: Custom-made 3D printed titanium implants seems to be a viable option for restoration of the anterior column after en bloc resection for spinal tumor. Longer follow-up will be needed for fusion rates and long-term complication rates.

OP27: Degenerative-Cervical 5

A167: Incidence of Fusion Procedure Following Single-Level Lumbar Decompression

Kevin Wiley¹, Jiandang Zhang², Blake Formanek¹, Zheng Wang², Zorica Buser¹, and Jeffrey Wang¹

¹University of Southern California, Los Angeles, CA, USA

²The General Hospital of PLA, Beijing, China

Introduction: Lumbar decompression without arthrodesis is a common procedure for the treatment of symptomatic degenerative lumbar spinal stenosis. Potential outcomes after decompression can include incomplete relief of symptoms, progression of disease, adjacent segment disease, or instability requiring additional arthrodesis. In this study we used a retrospective analysis of a national insurance billing database to determine the incidence of arthrodesis, revision surgery, after single-level lumbar decompression. Material and Methods: Data were collected using commercially available PearlDiver database (Humana). Patients undergoing laminectomy, laminotomy with facetectomy/foraminotomy, and removal of herniated disc material between January 2007 and December 2012 were included in the study. Current Procedural Terminology codes were used to identify decompression procedures. This cohort of patients was then followed for 1 year, 2 years, and 4 years for revision surgery including spinal fusion. Results: Among 18 323 patients who underwent single-level lumbar decompression, the incidence of fusion surgery was 1.73% (246/14 197) within 1 year, 2.80% (293/10 473) within 2 years, and 6.45% (308/4476) within 4 years of the index procedure. Among the lumbar fusion procedures within 4-year follow-up, non-interbody fusion was performed in 66.6% (205/308) of the cases, and single-level fusion in 55.2% (170/308) of the cases performed. Conclusion: In the current study using a retrospective review from the national commercially available PearlDiver database the incidence of arthrodesis after single-level decompression was 6.45% within 4-year follow-up. This data helps better define the rate of revision surgery following single-level lumbar decompression and provides valuable information to patients with regard to clinical decision making and outcomes.

A168: Subsidence Rates After Lateral Lumbar Interbody Fusion: A Systematic Review

Mohamed Macki¹, Sharath Kumar Anand¹, Paul Park², and Victor Chang¹

¹Henry Ford Hospital, Detroit, MI, USA

²University of Michigan, Ann Arbor, MI, USA

Introduction: The evidence regarding the consequences of subsidence with lateral lumbar interbody fusion (LLIF) has been sparse. The objective of this study is to calculate the incidence of subsidence and reoperation for subsidence following LLIF. A secondary outcome examined the quantitative degree of subsidence by calculating the percent change in the height of the intervertebral space secondary to interbody subsidence at various postoperative follow up times. Materials and Methods: Following the Meta-analysis (and Systematic Review) Of Observational Studies in Epidemiology (MOOSE) Guidelines, a systematic review searched for all cohort studies that focused on subsidence rates after LLIF, including extreme lateral interbody fusions (XLIF) and direct lateral interbody fusion (DLIF). Neoplastic, infectious, and/or metabolic indications for LLIF were similarly excluded as these pathologies may compromise bone quality and, thus, confound the rate of cage subsidence. Corpectomies were removed from the systematic review because (1) indications for removal of vertebral body typically reflect those excluded pathologies and (2) subsidence refers to a different biomechanical process. Results: This systematic review identified a subsidence incidence with LLIF of 10.3% (N = 141/1362 patients in 14 articles) and reoperation rate for subsidence of 2.7% (N = 41/1470 patients in 16 articles). In the secondary outcome measure, the disc height decreased from 5.6% after 3 months, 6.0% after 6 months, 10.2% after 12 months, to 8.9% after 24 months (P < .001). Conclusion: Subsidence after LLIF carries a nonnegligible risk that may be incorporated in surgical consent discussions in selected patients.

A169: Impact of Tobacco Smoking on Outcomes After Posterior Decompression Surgery in Patients With Cervical Spondylotic Myelopathy

Narihito Nagoshi¹, Hitoshi Kono², Osahiko Tsuji³, Ryoma Aoyama⁴, Kanehiro Fujiyoshi⁵, Yuta Shiono⁶, Masayuki Ishikawa⁷, Kenshi Daimon⁸, Naobumi Hosogane⁹, Kota Watanabe¹, Masaya Nakamura¹, Morio Matsumoto¹, Ken Ishii¹⁰, and Junichi Yamane¹

¹Keio University, Tokyo, Japan

²Keiyu Orthopaedic Hospital, Gunma, Japan

³Saitama Medical Center, Saitama, Japan

⁴Tokyo Dental College, Ichikawa General Hospital, Chiba, Japan

⁵Murayama Medical Center, Tokyo, Japan

⁶Nerima General Hospital, Tokyo, Japan

⁷International University of Health and Welfare (IUHW), Mita Hospital, Tokyo, Japan

⁸Ogikubo Hospital, Tokyo, Japan

⁹National Defense Medical College, Tokyo, Japan

¹⁰International University of Health and Welfare School of Medicine, Chiba, Japan

Introduction: Smoking has been associated with poor outcomes in the field of spinal surgery. However, the impact of tobacco smoking on outcomes following posterior decompression surgery has not been fully evaluated in patients with cervical spondylotic myelopathy (CSM). Methods: This is a retrospective multicenter study. Five hundred and eighty-seven patients diagnosed as CSM were enrolled at 17 high-volume institutions in Japan. Patients underwent cervical laminoplasty or laminectomy, and were followed up for at least 1 year after surgery. Outcome measures were preoperative smoking status, perioperative complications, the Japanese Orthopaedic Association scale (JOA), and the Visual Analog Scale (VAS) for neck. Smoking and nonsmoking groups were compared using unpaired t-test for continuous variables or a χ² test for categorical variables. Results: There were 182 (31%) current smokers and 405 (69%) nonsmokers including previous smokers. Smokers were younger than nonsmokers (average 65.1 vs 68.4, P < .01). There were no significant differences in BMI (body mass index), number of operated laminae, operative time, and number of comorbidities; but the estimated blood loss during surgery was significantly higher in the smokers (57.6 mL vs 37.0 mL, P < .01). Regarding postoperative complications, there was no significant difference in the rate of surgical site infection, cerebrospinal fluid leakage, hematoma, and neurological deficit. However, smokers showed a significantly higher risk for delirium (3.3% vs 0.2%, P < .01). Both smokers and nonsmokers had comparable functional recovery in JOA scores (difference 3.1 vs 2.9, P = .33) and neck pain reduction using VAS (difference −1.8 vs −1.4, P = .24) at the final follow-up. Conclusion: This is the largest study analyzing the efficacy and safety of posterior surgical decompression in smokers with CSM. Although the estimated blood loss was larger in smokers, they gained functional restoration and neck pain reduction at the final follow-up. Attention is required, however, on postoperative complications such as delirium. This is an important factor to note when explaining the risks of surgery to patients who smoke.

A170: Outcome Analysis of Single Open-Door Posterior Decompression With Instrumented In Situ Fusion for Cervical OPLL Patients of the K-Line (−)

Yawei Li¹, Guohua Lv¹, and Bing Wang¹

¹The Second Xiangya Hospital of Central South University, Changsha, China

Introduction: Due to continued movement in the extension and flexion direction, cervical laminoplasty has been reported to lead to poor outcomes in K-line (−) ossification of posterior longitudinal ligament (OPLL). We investigated the outcome of single open-door posterior decompression with instrumented in situ fusion (PDF) surgery for patients with K-line (−) OPLL of the cervical spine, who may have a poor surgical prognosis. Materials and Methods: From February 2008 to February 2015, 38 cases including 30 males and 8 females undergoing posterior decompression due to K-line (−) OPLL in our institution. The age ranged from 48 to 76 years, mean 56.3 ± 9.7 years. The OPLL canal occupation ratio ranged from 38.6% to 72.5%, mean 58.1 ± 13.6%. According to surgical procedures, 26 case were in PDF group and 12 case were in LMP group. The data collected from both groups included complications, C2-C7 Cobb angle, and neurologic symptoms evaluated based on the Japanese Orthopaedic Association (JOA) score and JOA score recovery rate. Results: More than 2 years of follow-up was completed in all patients. At final follow-up, the postoperative JOA score was 11.8 ± 1.9 points on average, improving from preoperative 7.7 ± 1.6 points in PDF group (P < .05); and the postoperative JOA score was 9.1 ± 2.1 points, improving from preoperative 7.9 ± 1.5 points in LMP group (P < .05). The mean JOA score recovery rate was 44.2% and 29.5% (P < .05), respectively. In PDF group, the C2-C7 Cobb angle was 4.1 ± 2.0°, similar with preoperative 3.8 ± 1.6° (P > .05). On the contrary, the C2-C7 Cobb angle was −2.1 ± 1.8°, lower than preoperative 3.9 ± 1.2° (P < .05). Ten cases had C5 palsy (PDF: 7; LMP: 3), and 3 cases had wound infection (PDF: 2; LMP: 1). Complication rate was 34.6% and 33.3% (P > .05), respectively. Conclusions: Compared with LMP, PDF without correcting cervical alignment for patients with K-line (−) OPLL showed moderate neurological recovery and improved clinical efficacy.

A171: Postoperative C5 Palsy Is Best Predicted by Use of Intraoperative C5 Laminectomy Among Patients Operated for Degenerative Cervical Myeloradiculopathy

Zach Pennington¹, Daniel Lubelski¹, Ali Ahmed¹, Adam D’Sa¹, Erick Westbroek¹, Matthew Goodwin¹, and Daniel Sciubba¹

¹Johns Hopkins University, Baltimore, MD, USA

Background: Annually, it is estimated that 18 000 persons will undergo posterior cervical spine procedure, for which degenerative myeloradiculopathy is one of the most common operations. Postoperative C5 palsy is one complication of this procedure, occurring in ∼8% of cases and increasing care cost by more than $2000 per patient. Here we present a model for predicting postoperative palsy using clinical presentation, preoperative imaging, and procedural details. Materials and Methods: A cohort of 385 patients undergoing posterior cervical decompression with or without concomitant fusion at the Johns Hopkins Hospital between January 1, 2014, and December 31, 2016, was queried. Data collected included operative details, clinical presentation, and cervical anatomy on preoperative imaging. A multivariate logistic regression was used to identify significant predictors of postoperative palsy, defined as a decrease in deltoid strength of one or more points on manual muscle testing. Results: Of those patients with full data, 11.3% suffered a postoperative C5 palsy. Patients experiencing a postoperative palsy were more likely to report preoperative hand clumsiness, radicular pain the C5 distribution, and undergo a C5 laminectomy intraoperatively. On multivariate analysis, significant predictors of postoperative palsy were revision versus index case and the performance of a C5 palsy. A receiver-operating curve of our model demonstrated a C-statistic of 0.8075. Conclusions: Postoperative C5 palsy is best predicted by an aggregate model incorporating both operative details, clinical presentation, and radiographic parameters. Preoperative C5 radicular pain, prior history of posterior cervical decompression, and performance of a C5 laminectomy were the strongest predictors of postoperative palsy, consistent with the notion that preoperative C5 root irritation and intraoperative C5 root manipulation contribute to the etiology of C5 palsy.

A172: Correlation Study Between Buttoning Sign and Severity in Patients With Degenerative Cervical Myelopathy

Torphong Bunmaprasert¹ and Komkhem Kaewbaudee¹

¹Chiang Mai University, Chiang Mai, Thailand

Introduction: Degenerative cervical myelopathy (DCM) is defined as dysfunction of the spinal cord by extrinsic compression that is caused by degenerative diseases of the cervical spine. Hand dysfunction from DCM strongly affects patient’s daily living. Difficulty buttoning the button is a common main complaint of DCM patients. Purpose is to define the correlation of the time for buttoning the button and severity of DCM patients evaluated by using Modified Japanese Orthopaedic Association Scale (mJOA). Material and Methods: Thirty-two DCM patients were divided into mild, moderate, and severe groups by mJOA score. All patients were asked to perform buttoning the button, and then time taken was recorded. Spearman rank correlation coefficient was used to study the correlation between time for buttoning and severity groups of DCM patients. Results: There were 13 patients in mild group, 13 patients in moderate group, and 6 patients in severe group. The correlation of time taken to perform buttoning the button and mJOA score was moderate negative relationship (r = −0.54, P = .0098). The correlation of time taken to perform buttoning the button and motor dysfunction score of the upper extremities in mJOA score was a strong negative relationship (r = −0.72, P = .0002). Conclusion: For these preliminary results, time to perform buttoning the button significantly has moderate negative correlated with mJOA score but significantly strong negative correlated with motor dysfunction score of the upper extremities in mJOA score.

A173: The Down Questionnaire: A Novel Screening Tool for Cervical Spondylotic Myelopathy

Ifije Ohiorhenuan¹, Kaku Barkoh², Ray Hah¹, and Ian Buchanan¹

¹University of Southern California, Los Angeles, CA, USA

²Texas Orthopaedic and Sports Medicine, Houston, TX, USA

Introduction: Cervical spondylotic myelopathy (CSM) is the most common cause of spinal cord injury in adults over 55. However, since its onset is insidious, accurately diagnosing CSM can be challenging, often requiring referral to a subspecialist as well as advanced imaging. To help identify patients at risk for CSM, this case-control study compared responses to a series of 4 questions (DOWN questionnaire) in myelopathic and non-myelopathic patients. Materials and Methods: Ninety-two patients, 46 with and 46 without myelopathy, were recruited for the study. Each patient answered 4 questions encompassing common symptoms associated with CSM. Responses between patient groups were compared and Cohen’s κ was used to assess for agreement between responses and the diagnosis of myelopathy. Results: We found a sensitivity of 91% and a κ of 0.54 to 3 positive responses and a sensitivity of 72% and a κ of 0.61 to 4 positive responses. Conclusion: Positive responses to 3 or more DOWN questions has moderate agreement with a diagnosis of myelopathy based on history, physical exam, and review of advanced imaging by an orthopedic or neurological surgeon. The DOWN questionnaire is a potentially useful screening tool by primary care and gerontologists to identify patients at risk for CSM.

A174: Recede-Myelopathy: Regeneration in Cervical Degenerative Myelopathy—A Multicenter, Double-Blind, Randomized, Placebo Controlled Trial Assessing the Efficacy of Ibudilast as an Adjuvant Treatment to Decompressive Surgery for Degenerative Cervical Myelopathy

Benjamin Davies¹, Michael Fehlings², Michelle Starkey¹, Sukhvinder Kalsi-Ryan¹, Mark Kotter¹

¹University of Cambridge, Cambridge, UK

²University of Toronto, Toronto, Ontario, Canada

Introduction: Degenerative cervical myelopathy (DCM) affects up to 5% of people over 40. While decompressive surgery can arrest progression and offer meaningful benefit, recovery is incomplete and patients retain lifelong disability, contributing to some of the lowest quality of life scores of all chronic diseases. Preclinical work has demonstrated inhibition of phosphodiesterase 4 can stimulate axonal regeneration and remeylination, via oligodendrocyte progenitor cells. Ibudilast is a well-characterized PDE4 inhibitor in use for the treatment of asthma and cerebrovascular disorders. Materials and Methods: (EduraCT 2017-004 856-41). RECEDE-Myelopathy is a phase III, multicenter, double-blind, randomized, placebo controlled study to assess the efficacy of a 6-month course of Ibudilast as an adjuvant treatment to decompressive surgery for DCM. Patients with moderate DCM (modified Japanese Orthopaedic Association [mJOA] 8- 14) scheduled for their first operation, will be randomized 1:1 to placebo or Ibudilast therapy. Assessments will be conducted at baseline, preoperatively, and 3, 6, and 12 months postoperatively. An internal power calculation after 25 to 80 participants will be performed to confirm the overall sample size required, estimated at 350 to 400. The study is funded by the National Institute for Health Research, United Kingdom. Results: The study will employ dual primary endpoints: the mJOA score and VAS pain at 6 months. Additional secondary endpoints include GRASSP-Cervical Myelopathy, SCIMvs3, 30 m Walk Test, NDI, SF-36, EQ-5D, and QuickDASH. Exploratory end points include respiratory physiology, gait laboratory analysis, and Carer Quality of Life. Conclusions: This will be the first regenerative medicine trial within the field of degenerative cervical myelopathy. Recruitment is due to start January 2018 and additional centers are required.

A175: Preventing C5 Palsy After Cervical Decompression by Nerve Root Untethering and Intraforaminal Ligament Release

Andrew Jack¹, Brooks Osburn¹, Zane Tymchak¹, Wyatt Ramey¹, Robert Hart¹, Rod Oskouian¹, Jens Chapman¹, and Shane Tubbs¹

¹Swedish Neuroscience Institute, Seattle, WA, USA

Introduction: Postoperative C5-palsy (C5P) is a common and disabling complication after cervical decompression. Several theories exist explaining its mechanism, including nerve root tethering upon dorsal spinal cord (SC) migration. We investigate this theory by performing an anatomical study comparing C5 and C6 root translation and tension upon dorsal SC migration before and after root untethering by cutting the cervical intraforaminal ligaments (IFL). To our knowledge, this is the first study to investigate this relationship. Materials and Methods: Six cadaveric prosections were completed. Extradural and intradural nerve roots were exposed via supraclavicular brachial plexus approach and C4-6 corpectomies. Metallic pins were inserted into the C5 and C6 roots and adjacent foraminal tubercle. Translation was measured as the distance between pins upon dorsally displacing the intradural nerve roots and SC 5 mm (average SC translation post-decompression) before and after cutting IFL. Strain was expressed as the percent change in intraforaminal nerve root translation measured relative to intradural translation before and after cutting IFL. Root translation and strain were also measured during shoulder depression to simulate intraoperative positioning and the protective effect of the IFL. Two-way repeated measures ANOVA with post hoc statistical testing was completed comparing translation and strain of the different root levels before and after IFL cutting. Statistical significance was set at .05. Results: Selective IFL cutting increased translation at both C5 and C6. Mean C5 translation was 1.12 mm (±0.75 mm) before and 4.89 mm (±0.97 mm) after IFL cutting. Mean C6 translation was 2.17 mm (±1.18 mm) before and 4.16 mm (±0.48 mm) after IFL cutting. A significant interaction in translation was found between IFL status (cut vs not-cut) and root level, F(1, 5) = 23.89, P = .005. Significantly more root translation was seen if the IFL were cut versus not-cut, F(1, 5) = 47.35, P = .001. No difference was found between root levels, F(1, 5) = 1.18, P = .33. Selective IFL cutting increased translation upon shoulder depression at both C5 and C6. Mean C5 translation was 0.54 mm (±0.51 mm) before and 6.90 mm (±2.64 mm) after IFL cutting. Mean C6 translation was 0.37 mm (±0.34 mm) before and 5.05 mm (±1.30 mm) after IFL cutting. A significant interaction in translation was found between IFL status (cut vs not-cut) and root level upon shoulder depression, F(1, 5) = 6.91, P = .047. Significantly more root translation with shoulder depression was seen if the IFL were cut versus not-cut, F(1, 5) = 29.87, P = .003. A significant difference was also found between root levels, F(1, 5) = 10.47, P = .02. Conclusion: Upper cervical nerve roots are tethered in their respective foramina by the IFL. Cutting these ligaments results in increased root translation suggesting they may be either protective (preventing peripheral nerve strain from being transmitted to the SC when the ligaments are intact) or harmful (by tethering the intraforaminal nerve root and potentially contributing to C5P after decompression) depending on the clinical context. Further investigation examining the role of selective nerve root untethering as a means of treating and/or preventing C5P is required.

OP28: Basic Science

A176: Bioengineered Smart Human Neural Stem Cells to Degrade Scar and Enhance Regeneration in Chronic Spinal Cord Injury

Christopher Ahuja¹, Mohamad Khazaei², Zijian Lou², Priscilla Chan³, Sohanthan Udayashankar², Nayaab Punjani¹, William Luong², Suyue Lyu², Jian Wang², and Michael Fehlings¹

¹University of Toronto, Toronto, Ontario, Canada

²University Health Network, Toronto, Ontario, Canada

Introduction: Human induced pluripotent stem cell-derived neural stem cells (hiPS-NSC) are an exciting regenerative approach for traumatic spinal cord injury (SCI) as NSCs have the capacity to replace neural circuits, remyelinate denuded axons, and provide local trophic support. Unfortunately, most individuals are in the chronic phase of their injury where dense perilesional chondroitin sulfate proteoglycan (CSPG) scarring significantly impairs neurite outgrowth and regenerative cell migration. Scar-modifying enzymes can enhance NSC-mediated recovery; however, nonspecific administration via intrathecal catheters increases the risk of off-target CNS effects. We aimed to generate a novel, genetically engineered line of hiPS-NSCs, termed Spinal Microenvironment Modifying and Regenerative Therapeutic (SMaRT) cells, capable of locally expressing a scar-degrading enzyme to enhance functional recovery without the risk of nonspecific administration. Material and Methods: Using non-viral PiggyBac transposon technology, a proprietary scar-degrading enzyme was genetically integrated into hiPS-NSCs under a constitutive human EF1α promoter and a monoclonal line was generated by fluorescence activated cell sorting. Enzyme expression and activity was extensively characterized in vitro by biochemical assays, slot blot quantification of intact and enzyme-degraded CSPGs, and cell culture assays. To assess efficacy in vivo, T-cell deficient Rowett Nude rats (N = 60) with chronic (8 week) C6-7 clip-contusion injuries were randomized to receive (1) hiPS-NSCs, (2) SMaRT enzyme-expressing hiPS-NSCs, (3) vehicle control, or (4) sham surgery (laminectomy alone). Animals were assessed with weekly neurobehavioral tests including open-field locomotor scoring (BBB), forelimb locomotor assessment scale (FLAS), inclined plane stability, forelimb grip strength, CatWalk digital gait assessment, and tail flick for 32 weeks posttransplant. Results: The scar-degrading ENZYME and fluorescent reporter are robustly expressed by the transgenic SMaRT cells. SMaRT cells retain key human NSC characteristics including the capacity to repeatedly proliferate, spontaneously form neurospheres, and differentiate along all 3 neuroglial lineages. The enzyme expressed by SMaRT cells appropriately degrades CSPGs in vitro and allows hiPS-NSC-derived neurons to extend into CSPG-dense scar-like areas after 7 days in culture. Conditioned SMaRT cell media also rapidly degrades in situ rodent CSPGs in ex vivo injured spinal cord cryosections. While blinded histological analyses are ongoing, an interim analysis of several animals shows that grafted human cells are extending remarkably long (cervicomedullary junction to mid-thoracic) axons through rodent white matter at 8 weeks posttransplant. The graft continues to mature over 32 weeks posttransplant demonstrating more numerous, thinner, and longer processes which stain for mature neuron marker, NF200. Several graft-derived neurons also demonstrate a bulbous morphology and stain for growth associated protein, GAP43, suggestive of active axonal growth cones at 32 weeks posttransplant. Conclusion: This work provides exciting proof-of-concept data that genetically engineered SMaRT cells can degrade CSPGs in vitro and that human NSC grafts can form long axonal processes in the typically inhibitory chronic cervical SCI niche. SMaRT cells hold the potential to be the first effective regenerative therapy for individuals with chronic spinal cord injury where even modest motor recovery can have tremendous implications for quality of life.

A177: The Effects of Bone Morphogenetic Protein 2 (BMP-2) on Transplanted Bone Marrow Cell Survival and Proliferation in a Mouse Model of Spinal Fusion

Alexander Perdomo-Pantoja¹, Wataru Ishida¹, Maritza Taylor¹, Colson Tomberlin¹, L. Lo Sheng-Fu¹, Timothy F. Witham¹, and Christina Holmes²

¹ Johns Hopkins University, Baltimore, MD, USA

²Florida A&M University-Florida State University, Tallahassee, FL, USA

Introduction: Currently, one of the more common bone tissue engineering approaches involves the combined delivery of bone morphogenetic protein 2 (BMP-2) and osteoprogenitor cells. While BMP-2 is well known to promote bone formation in vivo and induce differentiation of mesenchymal stem cells,¹ little is known about whether it affects the proliferation or survival of transplanted cells in tissue engineering applications. We recently developed a mouse model which enables the noninvasive in vivo tracking of transplanted bone marrow cells (BMSCs) over time during spinal fusion. In this study, we use this model to explore the role of BMP-2 on transplanted BMSC survival and proliferation during spinal fusion. Materials and Methods: BMSCs were isolated from the long bones of either FVB/NJ mice or syngeneic transgenic FVB mice which constitutively express the luciferase gene and seeded at a concentration of 5 × 10⁶ cells/sponge onto a collagen sponge (Medtronic) pre-incubated with either 0, 1, or 2 ng of rhBMP-2 (Medtronic). FVB/NJ host mice underwent bilateral posterolateral lumbar spinal fusion surgery and were divided into 4 groups according to the bone graft substitute used: (1) collagen sponge + luciferase-expressing BMSCs (n = 13); (2) collagen sponge + low-dose BMP-2 + luciferase-expressing BMSCs (n = 13); (3) collagen sponge + high-dose BMP-2 + luciferase-expressing BMSCs (n = 13); and (4) collagen sponge seeded with syngeneic BMSCs (n = 6). The group seeded with syngeneic BMSCs was used as a bioluminescence imaging control. Bioluminescence imaging was performed at 3, 6, 9, 12, and 15 days postsurgery and at 4, 6, and 8 weeks postsurgery. Fusion was assessed 8 weeks postsurgery via microCT imaging. Results: All groups transplanted with luciferase-expressing BMSCs showed an increase in bioluminescence flux signal over the first 12 days, suggesting early transplanted cellular proliferation. This was followed by a dramatic decrease in signal from day 15 onwards, indicating cellular death. Between days 6 and 15, the group treated with high concentration BMP-2 exhibited the highest flux signal intensity, followed by the group treated with a low BMP-2 dose, and the group with no rhBMP-2. From week 4 onwards, however, all groups showed similar, dramatically reduced flux signals. At 8 weeks, both groups treated with BMP-2 were found to fuse at similar rates via CT imaging (CT fusion scores: 1.62 ± 0.65 for low dose, 1.69 ± 0.63 for high dose); however, the higher dose yielded a higher fusion mass volume (5.80 ± 1.95 mm³ vs 4.33 ± 1.65 mm³, P = .01). No animals in groups treated with cells alone fused (CT fusion score: 0.00 ± 0.00 for luciferase-expressing BMSCs, 0.00 ± 0.00 for syngeneic BMSCs). Conclusions: BMP-2 appears to increase early transplanted BMSC proliferation in a dose-dependent manner; however, by week 4 postsurgery, this effect decreases and all groups exhibit a dramatic decrease in transplanted cell survival. As has been found in previous studies, transplanted BMSCs alone were unable to achieve successful fusion, while increasing concentrations of BMP-2 yielded higher fusion mass volumes. Immunohistochemistry is currently being performed in order to determine the fate of transplanted cells within the fusion mass.

Reference

1. El Baily I, Jiskoot W, Nejadnik MR. Formulation, delivery and stability of bone morphogenetic proteins for effective bone regeneration. Pharm Res. 2017;34:1152-1170.

A178: Self-Assembling KKQL6KK Peptide Biomaterial Optimizes Human Neural Stem Cell Treatment for Traumatic Spinal Cord Injury

Christopher Ahuja¹, Mohamad Khazaei², Yao Yao², Zijian Lou², Priscilla Chan¹, Nayaab Punjani¹, Sohanthan Udayashankar², William Luong², Suyue Lyu², Jian Wang², and Michael Fehlings¹

¹University of Toronto, Toronto, Ontario, Canada

²University Health Network, Toronto, Ontario, Canada

Introduction: Neural stem cells (NSCs) have emerged as a promising therapeutic strategy for regeneration after traumatic spinal cord injury (SCI). Unfortunately, the hostile postinjury microenvironment is a significant hindrance to regeneration. QL6 (K2QL6K2; Medtronic Inc) is a novel self-assembling peptide biomaterial with unique features including a neutral pH, biodegradability, and an extracellular matrix-like assembled morphology. After acute SCI in animal models it has been shown to reduce inflammation and astrogliotic scarring resulting in significant functional recovery. QL6 also reduces cavity volume and supports the survival of allogeneic mouse NSCs after SCI when used as a co-transplant. However, QL6’s ability to support translationally relevant human-induced pluripotent stem cell-derived NSCs (hiPS-NSCs) has not yet been demonstrated. Material and Methods: hiPS-NSCs were passaged into a pseudo-monolayer on QL6 versus a Geltrex control. The mechanism of adhesion was assessed by 60-minute EDTA assay and quantitative real-time PCR (polymerase chain reaction). NSC survival, proliferation (Ki-67 staining), and neurosphere formation/morphology was extensively characterized in vitro. T-cell deficient RNU rats (N = 70) capable of receiving xenografted cells were given C6-7 clip-contusion injuries, which closely mimic the pathobiology of human SCI. Animals were randomized at 8 weeks post-injury to (1) vehicle treatment, (2) hiPS-NSCs transplant, (3) QL6 transplant, (4) QL6 + hiPS-NSCs co-transplant, or (5) sham surgery (laminectomy alone). All rats receive daily (5 days/week) treadmill training commencing 4 weeks after transplant. Results: hiPS-NSCs proliferated robustly on assembled QL6 versus a geltrex control assessed by Ki67⁺ cell proportion (29% vs 6%, P < .01). EDTA assay showed that hiPS-NSC binding to QL6 is mediated by calcium-independent mechanisms. On qPCR, hiPS-NSCs grown on QL6 versus geltrex control demonstrated downregulation of apoptosis markers after 3 and 7 days and upregulation of pro-neurogenic markers and select calcium-independent cell adhesion molecules. Importantly, QL6 promoted the formation of adherent neurospheres, the native conformation of NSCs, even when applied in a thin 50 µm layer similar in size to microcystic cavitations within the cord. Blinded neurobehavioral assessments of transplanted rats are now ongoing with a 20-week post-injury endpoint. Assessments include BBB open-field locomotor scoring, inclined plane test, Montoya forelimb reaching test, forelimb grip strength, forelimb locomotor assessment scale, Irvine-Beatties-Bresnahan forelimb recovery scale scoring, and CatWalk digital gait assessment. Conclusion: This work provides proof-of-concept data that the uniquely engineered QL6 self-assembling peptide biomaterial can support the growth of translationally relevant human iPS-derived neural stem cells.

A179: Cobb Measurement With a Conventional Convex Echography Probe: A Prospective, Randomized, and Triple-Blinded Study

Joan Ferràs-Tarragó¹, Jorge Mario Morales Valencia¹, Silvia Pérez-Vergara¹, Pedro Rubio-Belmar¹, Jose Luís Bas-Hermida¹, Paloma Bas-Hermida¹, and Teresa Bas-Hermida¹

¹ Orthopaedics and Traumatology, Valencia, Spain

Introduction: Adolescent idiopathic scoliosis is a vertebral 3-dimensional deformity occurring in around 3% of adolescents, with higher prevalence in females. Serial X-rays, which need to be done every 3 months in cases of high risk of progression, imply considerable accumulation of radiation. Due to the age at which this pathology occurs, there is high risk of developing associated conditions, such as breast cancer. Thanks to the advances in ultrasound techniques, deformity measurement systems free from ionizing radiations have been validated, although spinal surgeons do not used them routinely due to the technical difficulties of the protocols and the need of a special software. The aim of our work is to assess the reproducibility and correlation of an ultrasound measuring system based on the positioning of the transverse processes. Material and Methods: Prospective, single-center randomized and triple-blinded study. Two independent researchers trained in ultrasound examined the spinal deformities of 31 children. The measurements were compared against those performed with an X-ray by 3 scoliosis expert surgeon. Statistics were performed by an independent researcher. Parametric methods were used. Results: We found a 95% ([0.91-0.97] P < 2.2e-16) correlation between the degree of scoliosis measured with the proposed ultrasound system and the 30 × 90 X-rays in standing position. There was an intraobserver reliability of 97% (R ² = 0.97, CI 95% = 0.95-0.98; P < 2.2e^-16) and an inter-observer. Conclusion: An approximation of the Cobb angle measure is possible with ultrasound by using the transverse processes as reference. This is a very rapid, simple, and direct system for assessing the principal spinal deformity measure in young people, although it does not allow estimating the associated axial or sagittal rotation.

Figure 1.

A180: Transcriptome High-Throughput Sequencing and Bioinformatics Analysis of the Differentially Expressed Genes for Thoracic Ossification of Ligamentum Flavum Before and After Mechanical Stress

Shanglong Ning¹, Jun Miao¹, Xinlong Ma¹, Zhongqiang Chen², and Michael Janssen³

¹Tianjin Hospital, Tianjin University, Tianjin, China

²Perking University Third Hospital, Beijing, China

³ Center for Spine and Orthopedics, Denver, CO, USA

Introduction: Thoracic ossification of ligament flavum (TOLF) of the spine is characterized by ectopic bone formation in the flavum ligament. Many factors contribute to OLF, including genetic background, dietary habits, metabolic abnormalities, and mechanical stress. Up to now, the molecular mechanism of is not clear yet. Many studies have demonstrated the critical role of mechanical stress in the development of TOLF, and it was reported that mechanical stress induced osteogenic differentiation of ligament cells derived from TOLF. But there is no correlated study on the transcriptome analysis of TOLF to explore the molecular pathogenesis of TOLF systematically and roundly. The purpose of this study was to detect the differentially expressed genes between TOLF and non-TOLF and between TOLF before and after mechanical stress. In addition, we sought to identify a few candidate genes and pathways by using bioinformatics analysis. Material and Methods: Clinically, the patients with non-TOLF, single-level TOLF, and multiple-level TOLF (n = 3 in each group) were included in this study. The primary ligament cells that derived from the 3 groups were separately cultured and induced osteogenesis with cyclic mechanical stress for 24 hours using a device called Flexcell FX-4000. Purified mRNA and cDNA extracted from the samples was subjected to sequencing. NOISeq method was used to statistically identify the differentially expressed genes (DEGs) between the 2 groups. Deep analysis using bioinformatics tools based on DEGs was performed using Gene Ontology (GO) enrichment, Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway enrichment, and protein-protein interaction network analysis. The Classic DEGs were verified using real-time quantitative polymerase chain reaction (real-time qPCR). Results: A total of 671 genes of TOLF group were upregulated and 314 genes were found to be downregulated as compared to the control group. And 341 genes of TOLF after stress induction group were upregulated and 502 genes were found to be downregulated as compared to the group without stress stimulus. In the comparison between TOLF group and non-TOLF group, the top 3 cellular components in GO ontologies analysis were extracellular matrix (ECM) components. GO functions were mainly related to the glycoprotein in the cell membrane and ECM. GO process were related to completing response to stimulus or developmental process. In addition, 22 significant GO terms associated with upregulated genes were found to be closely related to ossification. And in the comparison between TOLF-SI group and TOLF control group, the top GO terms associated with upregulated genes included cell cycle, and so on, and they are related to nuclear chromosome and tubulin binding. Then, extracellular matrix and cell adhesion could be found in the top downregulated GO terms. In the comparison between TOLF group and non-TOLF group, 22 significant KEGG enrichment pathways were found (corrected P < .05), and the top 2 pathways were cell cycle and ECM-receptor interaction, in addition, TGF-β, p53, FoxO, and PI3K-Akt pathway also had a very significant difference. And in the comparison between TOLF-SI group and TOLF control group, 21 significant KEGG enrichment pathways were found (corrected P < .05), and cell cycle and ECM-receptor interaction pathway were also on the top; in addition, p53, focal adhesion, Wnt and PI3K-Akt pathway also had a very significant difference. Some differentially expressed genes were verified with real-time PCR. Conclusion: The gene expression profiling of TOLF group and non-TOLF group revealed differential gene expression. The similar result was achieved in TOLF-SI group and TOLF control group. Global transcriptome analysis revealed TOLF cells expressed more ossification related genes than non-TOLF group, and TOLF ligament cells exhibit an osteoblast phenotype. Some genes including L1RL1, PTHLH, DKK1, BMP6, SPP1, and FGF1 may be closely related to TOLF, and so are some signaling pathway, for example, cell cycle and ECM-receptor interaction, and the network they constituted. Wnt pathway abnormality and local inflammation may be related to disc ossification.

A181: Preclinical Testing of Anti-Inflammatory Compounds Using a Whole Intervertebral Disc Organ Culture Model

Gernot Lang¹, Zhen Li², Yannik Gehlen¹, Fabian Heizmann¹, David Kubosch¹, Norbert Südkamp¹, Mauro Alini², and Sibylle Grad²

¹University Medical Center, Freiburg, Germany

²AO Research Institute Davos, Davos, Switzerland

Introduction: Disease-modifying antirheumatic drugs (DMARDs) have recently been used for the treatment of degenerative disc disease and have been shown to effectively relieve pain symptoms. However, their underlying anti-inflammatory and regenerative activity is poorly explored. The present study aimed to investigate the effects of the tumor necrosis factor–α (TNF-α) inhibitor Etanercept and the selective JAK3-inhibitor Tofacitinib in a degenerative, inflammatory intervertebral disc (IVD) organ culture model. Materials and Methods: Part I. Whole bovine caudal IVDs were cultured within a bioreactor for 4 or 11 days. The control group (PHY) was cultured under physiological loading (0.02-0.2 MPa; 0.2 Hz; 2 h/d) and high glucose (4.5 g/L) medium. In the degenerative group (DEG), detrimental loading (0.32-0.5 MPa; 5.0 Hz; 2 h/d) and low glucose (2.0 g/L) medium was applied to mimic the conditions of abnormal mechanical stress and limited nutrition. TNF-α was injected into the nucleus pulposus (NP) (100 ng/IVD) as proinflammatory trigger. Expression of inflammatory, anabolic and catabolic genes, release of nitric oxide (NO), cytokines and glycosaminoglycan (GAG), and cell viability were analyzed. Part II: Etanercept (3.5 mg/70 µL/IVD) or phosphate buffered saline (PBS) (70 µL/IVD) was injected into the IVD cultured under DEG + TNF-α conditions. Tofacitinib citrate (2.5 µg/mL) or PBS was added daily into the culture medium to simulate a systemic application of the drug. After 4 days, the effect of Etanercept and Tofacitinib on cellular gene expression and molecule release from IVDs was analyzed. One-way ANOVA or Kruskal-Wallis tests were used to determine statistical differences (n = 6-8). Results: Part I: TNF-α injection combined with detrimental loading and low glucose medium significantly upregulated the expression of interleukin (IL)-1β, IL-6, and IL-8 genes in cells of the cultured IVD; moreover, NO and IL-8 release from IVD were increased, indicating a proinflammatory effect (all P < .05 vs PHY). The combined initiators also significantly upregulated matrix metalloproteinase 1 (MMP-1) gene expression, downregulated gene expression of type I collagen in annulus fibrosus (AF) and type II collagen in NP (P < .05 vs PHY). Furthermore, the DEG + TNF-α conditions induced a degrading state, as indicated by markedly elevated GAG release. Under DEG conditions, the cell viability dropped to ∼40% in AF and 60% in NP after 11 days (P < .05 vs PHY). Part II: DEG + TNF-α treatment upregulated the expression of catabolic enzymes MMP-1, MMP-3, proinflammatory cytokines IL-1β, IL-6, IL-8, and cyclooxygenase-2 (COX-2); whereas Etanercept and Tofacitinib partially reduced these effects. Compared with the DEG + TNF-α group, both drug treatments could partially decrease the release of GAG. Etanercept partially attenuated the release of NO, whereas Tofacitinib had no effect on NO release. Conclusions: The combination of detrimental dynamic loading, nutrient deficiency, and intradiscal TNF-α injection synergistically triggered a proinflammatory and degenerative situation within the IVD, demonstrating the potential of our whole organ culture model for preclinical testing of anti-inflammatory or regenerative therapies. Etanercept and Tofacitinib showed the ability to slow down the degenerative response and reduce inflammation in the organ culture model, though combined anti-inflammatory and anabolic treatment may be required to completely halt or reverse the catabolic phenotype and degenerative cascade.

A182: Combined Annulus Fibrosus Repair and Nucleus Pulposus Replacement in an In Vivo Ovine Study

Christoph Wipplinger¹, Stephen R. Sloan², Sertac Kirnaz¹, Rodrigo Navarro-Ramirez¹, Franziska A. Schmidt¹, Antonella Schiavinato³, Lawrence J. Bonassar², and Roger Härtl¹

¹Weill Cornell Medicine, New York, NY, USA

²Cornell University, Ithaca, NY, USA

³Fidia Farmaceutici S.p.A, Abano Terme, Padua, Italy

Introduction: Lumbar disc herniation is a major cause of morbidity leading to enormous costs for western health care systems. Degenerative fissures of the annulus fibrosus (AF) allow the nucleus pulposus (NP) to herniate and impinge nerve roots. The current standard treatment for disc herniations is microdiscectomy in which the herniated part of the NP is removed. However, the AF defect remains untreated and the removed NP is not restored leading to the risk of reherniation and further degeneration of the intervertebral disc (IVD). Our group showed in recent in vivo animal studies successful AF repair using a high-density collagen (HDC) gel. Furthermore, in an ex-vivo study, we could show that combining our AF repair technique with NP replacement using a hyaluronic acid (HA) gel results in greater restoration of native IVD mechanics and hydration. The objective of the current study is to assess the efficacy of this combined approach in an in vivo sheep model. Material and Methods: A total of 8 skeletally mature female Finn sheep were used in this study. We performed a previously described anterolateral, retroperitoneal pre-psoas approach to access the IVDs L1-6. IVDs were randomized into 5 groups: (1) intact, (2) injured via 3 × 10 mm box annulotomy followed by a removal of 200 mg of NP, (3) injured as above followed by HDC gel patch for AF repair, (4) injury followed by injection of a HA gel into the NP, and (5) injury followed by HDC AF repair and NP HA replacement. At 6 weeks postoperatively, sheep were sacrificed and underwent postmortem 3T-MRI scans of the lumbar spine. Disc height index (DHI) analysis and Pfirrmann grading (PG) were performed on each segment using MRI. All animal use followed approved IACUC protocol. Results: Intact control discs were not degenerated and had an average PG of 1 while injured, untreated discs had significant degeneration with an average PG of 3. Discs receiving the combined injection and collagen AF patch individually showed fewer signs of degeneration than injured alone and the combined treatment resulted in the least amount of degeneration with PG not significantly different than the intact controls. DHI confirmed the trends seen in the PG, where injured discs lost 20% of the intact disc height, the individual NP and AF repairs restored 5% to 10% of intact disc height, and the combined repairs preserved 90% of the intact disc height. Conclusion: PG and DHI results demonstrate that individual NP and AF repairs are able to prevent disc degeneration better than no treatment at all; however, the greatest preservation of disc health was seen with combined AF and NP repairs. The collagen AF patch is necessary to prevent the HA NP injection from herniating out of the disc, while the HA NP injection is necessary to restore lost water content following injury. While no changes were seen with NP volume, this is consistent with previous literature.

A183: A Novel, Biological Therapeutic (Ntg-101) for the Treatment Degenerative Disc Disease

Ajay Matta¹, Muhammad Zia Karim¹, Hoda Gerami¹, Peter Jun², Martha Funabashi², Greg Kawchuk², Alyssa Goldstein³, Warren Foltz³, Marshall Sussman⁴, Bjorn Eek⁵, and W Mark Erwin^1,4

¹Notogen Inc, Toronto, Ontario, Canada

²University of Alberta, Edmonton, Alberta, Canada

³University Health Network, Toronto, Ontario, Canada

⁴University of Toronto, Toronto, Ontario, Canada

⁵Private Practice, Long Beach, CA, USA

Introduction: Degenerative disc disease (DDD)–related spinal pain is an enormously expensive malady and a major cause of disability worldwide with no effective treatment available that can promote repair. We have developed a novel biological therapeutic based on the notochordal cell secretome (NTG-101) that includes recombinant human transforming growth factor β1 + connective tissue growth factor (TGFβ1 + CTGF). Here we present extensive in vivo evidence in 2 validated preclinical animal models of DDD and human intervertebral disc (IVD) nucleus pulposus (NP) cells that a single injection of NTG-101 can mitigate the progression of DDD and induce a regenerative effect on the degenerative IVD. Material and Methods: We used image-guided needle puncture disc injury to induce DDD in two pre-clinical animal models of DDD (rat and chondrodystrophic [CD] canine) followed by a single injection of NTG-101 and assessment at the endpoint. We assayed our results using histological, immunohistochemical, genomic, protein detection, imaging (radiographs and MRI), and biomechanics assays for NTG-101 vs placebo. We assessed outcomes 10 weeks postinjury in rats and 18 weeks post injury in CD canines. Assays consisted of histological and immunohistochemical scoring, expression analysis of extracellular matrix (ECM), inflammation, and pain-associated genes/proteins using qRT-PCR, and Western blotting analysis. We also used radiographic and MRI techniques and quantitative biomechanical analyses. Additionally, we assayed the effects of this biological therapeutic on human IVD NP cells in vitro using genomic evaluation of pro inflammatory and pain-related genes as well as pro-anabolic extracellular matrix genes. Results: Both animal models showed increased expression of inflammation- and pain-associated genes and proteins (Cox-2, MMP-13, IL-1β, TNFα, IL-6, and IL-8) using qRT-PCR, immunohistochemistry, and Western blotting after needle puncture injury. We also observed increased PGE2 levels in injured, degenerative IVD NPs (rats and CD canines). However, a single, intradiscal injection of NTG-101 markedly reduced the expression of Cox-2, MMP-13, IL-1β, TNFα, IL-6, and IL-8 in treated IVD-NPs. Furthermore, NTG-101 treated IVD NPs demonstrated a robust increase in the expression of the vital ECM proteins, (aggrecan and collagen type 2) and a marker of stemness (Oct4). Radiographic analysis demonstrated retention of normal disc height in NTG-101 treated discs in comparison with placebo-injected IVDs. MRI and gross pathological analysis showed that placebo injected discs developed significant DDD whereas NTG-101 injected IVDs appeared healthy. Biomechanical analysis demonstrated that NTG-101 injected discs conferred significantly improved viscoelastic properties in flexion and lateral bending. Human NP cells treated with the biological therapeutic reduced gene expression of Cox-2 and MMP-13 and upregulation of aggrecan and collagen 2. Conclusion: A single injection of NTG-101 into the injured/degenerative IVD in both rat-tail and CD canine IVDs mitigates the progression of DDD and induces a regenerative effect resulting in a healthy IVD NP with near normal cellularity and healthy ECM. CTGF + TGF-β1 used in combination, downregulate pro-inflammatory and pain-related genes and upregulate the expression of important extracellular matrix genes indicating that these growth factors in combination have an antidegenerative/pro-anabolic effect upon human NP cells.

A184: Nose to Back: Compatibility of Nasal Chondrocytes With Environmental Conditions Mimicking a Degenerated Intervertebral Disc

Max Gay¹, Arne Mehrkens¹, Michael Rittmann¹, Andrea Barbero¹, Ivan Martin¹, and Stefan Schaeren¹

¹University Hospital Basel, Basel, Switzerland

Introduction: This project aims to determine whether nasal chondrocytes (NC) can be considered as an autologous cell source for cell therapy of disc degeneration by comparing them to mesenchymal stromal cells (MSCs) and articular chondrocytes (ACs), 2 cells sources used in phase 2 clinical trials at the moment. Material and Methods: Cells are cultured in in vitro micromass culture conditions mimicking facets of a degenerated intervertebral disc (IVD), such as hypoxia, low glucose, inflammation, and acidity. The production of the extracellular matrix (ECM) is evaluated using immunohistochemistry, quantitative real-time PCR, and biochemical analysis. Results: Our data demonstrate MSCs, ACs, and NCs have a similar glycosaminoglycan (GAG) and collagen 2 production in response to in vitro conditions simulating singular facets of the IVD environment and the addition of transforming growth factor β1 (TGF-β1). However, NCs synthesis more of these ECM components than MSCs or ACs when cultured in the same conditions without the addition of the growth factor. Furthermore, growth factor primed NCs maintain an exceeding production of GAG and collagen 2 compared with both growth factor primed MSCs and ACs in harsher conditions, which combine different characteristics of the degenerated IVD environment. Interestingly, the indifference of ECM production of NCs in response to inflammation factors cannot be linked to the absence of gene expression of the respective receptors. Moreover, NCs as the only cell sources display gene expression of the transcription factor FoxF1, a marker for nucleus pulposus cells. Conclusion: In summary, NCs are more similar to nucleus pulposus cells than MSCs and ACs, as they can better produce ECM in an in vitro IVD environment and express the nucleus pulposus marker FoxF1. These findings encourage the assessment that employing NCs in a cell therapy treatment of degenerated disc disease could promote new matrix production in the disc, which could inhibit or delay further disc height loss if not even lead to disc height gain.

OP29: Deformity-Thoracolumbar (Adolescent) 3

A185: Back Pain and Outcomes of Pregnancy After Instrumented Spinal Fusion for Adolescent Idiopathic Scoliosis

Pawel Grabala¹, Tomasz Guszczyn¹, and Ilkka Helenius²

¹Medical University in Bialystok, Bialystok, Poland

²University of Turku and Turku University Hospital, Turku, Finland

Introduction: We hypothesized that women undergoing scoliosis surgery who became pregnant would go to full-term, have uncomplicated pregnancies, and similar patient-reported outcome measures than nonpregnant patients and healthy controls. Materials and Methods: We studied 205 females (mean age at surgery 15 years, range, 12-25 years) out of 251 with adolescent idiopathic scoliosis (AIS) undergoing scoliosis surgery during 1998-2015 at several institutions in one country and 173 healthy women, without AIS, parous and nulliparous selected at random. Results: The cohort was divided into 108 women with AIS and no pregnancy (SNP), 97 with AIS and pregnancy (SP), 91 healthy controls and no pregnancy (HNP), and 82 healthy controls with pregnancy (HP) (Table 1). Mean (SD) follow-up was 5.0 (2.5) years in the SNP, 6.0 (2.0) years in the SP, 5.0 (2.5) years in the HNP, and 5.0 (2.5) years in the HP group, respectively (N.S.). The average time from surgery to childbirth was 5 years (range 2-13 years, SD 2.5). Back pain during pregnancy was observed in 48% (47/97) of the SP group and 34% (28/82) of the HP group (p>0.05). Back pain after childbirth was present in 43% of the (42/97) SP group and 42% (34/82) of the HP group (N.S.). Patients fused to L3 or L4 experienced more frequent low back pain during pregnancy (40%, 19/47) than those fused above L3-L1 and L2 (32%, 15/47), T11 and T12 (27%, 13/47), P < .05. Cesarean section was performed for 64% (62/97) and 33% (27/82) in SP and HP groups, respectively (P < .05). As the lowest instrumented vertebra moved caudal (L1, L2, L3, L4), the frequency of cesarean section increased (P < .05, R = 0.8). The majority of cesarean sections in the SP group were performed in patients with a long stabilization to L4 (55%) versus patients fused shorter—above and to L3 (45%) (P < .05). The most common analgesia for cesarean section was spinal (75% and 86% in SP and HP groups, respectively). Patients fused to L4 required general anesthesia significantly more often (7%, 4/53) versus those fused above L4 (4%, 2/44, P < .05). The health-related quality of life after spinal fusion showed general satisfaction and were similar for SNP and SP groups (mean 3.84 and 3.91, respectively, N.S.). No sexual dysfunction was reported. Conclusions: Women with a history of spinal fusion for AIS reported a similar incidence of back pain as healthy pregnant women, but when spinal fused to L3 or L4, the incidence of low back pain was higher. They also required cesarean section more frequently than healthy women. This risk increased to 55% when spinal fusion to L4 was performed. The quality of life and sexual function were at same level in women after scoliosis surgery than in healthy controls.

A186: Long-Term Follow-up of Gillet’s Technique in Lumbar Spondylolysis

Daniela Linhares¹, Pedro Cacho Rodrigues², Manuel Ribeiro da Silva¹, Vitorino Veludo¹, Rui Pinto¹, and Nuno Neves¹

¹Centro Hospitalar São João, Porto, Portugal

²Hospital da Prelada, Porto, Portugal

Introduction: Spondylolysis (SL) is prevalent among young adults and teenagers with low back pain. On failure of conservative treatment, repair procedures seek preservation of spinal mobility. In 1999, Gillet described the pedicle-based V-rod technique for direct pars repair. However, no study presents the long-term follow-up of this procedure. This work aims to describe and analyze the long-term radiological and clinical outcome of V-rod technique in SL treatment. Material and Methods: All patients submitted to surgical correction of lumbar SL with a V-rod system over 2 years were included. Those with adjacent disk degenerative changes and/or high-grade spondylolisthesis were excluded. A retrospective preop clinical and radiological evaluation was performed, along with a prospective assessment on the early (clinical)—up to 1 year—and late (clinical and radiological)—at least 10 years after the initial procedure—postoperation. Clinical evaluation comprised Oswestry DIsability Index (ODI) and visual analogue scale (VAS). Results: A total of 22 patients were included, 50% male, mean age 27.6 ± 11.2 years, 18.8% of smokers. In all, 42.9% patients were overweight/obese and the remaining had a normal body mass index (BMI). In 21, SL was present at L5, and in others at L4. In the preoperative X-ray, 50% had Myerding I spondylolisthesis, similar among gender. A significant decrease on ODI score was observed from pre- to early and late postoperative evaluation (43.5% ± 21.0% to 13.1% ± 11.7% and 20.9% ± 22.1%, P < .001 and P = .04, respectively) but not during postoperative evaluations (P = .27). The same was found in pain-VAS (preoperative 8.1 ± 1.1 vs early postoperative 3.0 ± 2.9 and late postoperative 3.8 ± 3.4, P < .001, P = .001, and P = .699, respectively). Preoperative ODI and VAS were significantly higher in overweight/obese but similar at postoperation. Three patients needed revision fusion surgery due to persistent pain; all had listhesis in the preoperative X-ray. After excluding patients submitted to revision surgery the mean change in disability was of 68.3% ± 30.1% from pre- to early postoperative and of 45.0% ± 58.2% to late postoperative, with the first being significantly higher than 50% (minimal clinically important difference [MCID]); the mean change in VAS was 5.2 ± 2.9 and 5.3 ± 3.2, respectively, both significantly higher than 1.7 (MCID). No significant differences in evolution of ODI and VAS were found among gender, smoking status, or preoperative radiological classifications. In late postoperation, X-ray was obtained for 16 patients, all with no sign of instability in dynamic views or additional listhesis. At a mean follow-up of 687.7weeks a Cox regression showed a survival rate of 81.8% for V-rod. Conclusion: SL reparable with V-rod system affects more commonly young patients, mainly at L5. Surgical treatment with V-rod system is associated with a significant improvement in ODI and VAS and radiologic stability, both in early and late postoperation. Obese/overweight patients benefit equally from the surgery. For the first time, a study with more than 2 years of follow-up reports a benefit that is maintained even 10 years after the initial intervention, associated with a low rate of failure.

A187: Outcome Analysis of Operative Hemivertebrae Correction at Children Depending on the Length of Instrumental Fixation

Egor Filatov¹, Sergey Ryabykh¹, and Dmitriy Savin¹

¹Ilizarov Center, Kurgan, Russian Federation

Introduction: Retrospective outcome analysis of patients’ treatment with congenital spinal anomaly. Purpose: Outcome analysis of operative hemivertebrae correction at children depending on length of instrumental fixation. Material and Methods: A total of 119 patients with spinal deformity on background hemivertebrae were evaluated at the age from 1 to 18 years over a period from 2010 to 2018. All interventions for patients were performed through dorsal pedicle access. Patients were divided to 5 groups according to variant of access, fixation, and vertebrotomy level. Long-term result was studied from 6 months to 8 years. Results: In group I (unilateral access with monosegmental fixation), there were 15 children with average age 48.3 ± 10.0months. Scoliosis before operation was averagely 31.1° ± 1,7°, correction 74.9%. Kyphosis before operation was 29.3° ± 6.0°, correction 84.0%. Blood loss was 213.6 ± 51.0 mL. Operation time was 165 ± 14.1 minutes. Complications: 2—deformity progressions with reoperation. Correction loss averagely was 4.1°. In group II (bilateral access with monosegmental fixation), there were 24 patients with average age of 53.1 ± 7.8. Scoliosis before operation was 32.5° ± 1.7°, correction 83.7%. Kyphosis was 30.9° ± 2.5°, correction 100.6%. Blood loss was 215.4 ± 36.6 mL. Operation time 160.5 ± 14.1 minutes. Complications: 2 fractures of vertebral arch pedicle with prolongation of instrumentation intraoperative, 4 deformity progressions with reoperation, 1 patient wears corset. Correction loss averagely was 1.6°. In group III (bilateral access with trisegmental fixation), there were 29 patients with average age of 78.2 ± 11.3months. Scoliosis was 36.6° ± 2.2°, correction 83.1%. Kyphosis was 37.6° ± 3.9°, correction 86.7%. Blood loss was 342.7 ± 36.2 mL. Operation time was 197.0 ± 15.3 minutes. Complications: 2 transient neurological complications, 4 deformity progressions with reoperation. Correction loss averagely was 0,88°. In group IV (bilateral access with polysegmental fixation), there were 49 children at the average age 112.7 ± 7.8 months. Scoliosis was 40° ± 2.3°, correction 72.5%. Kyphosis was 58.2° ± 7.3°, correction 81.6%. Blood loss was 549.3 ± 60.6 mL. Operation time was 288.8 ± 15.5 minutes. Complications: 4 neurological complications, 1 screws malposition, 4 deformity progressions with reoperation. Correction loss averagely was 1.6°. In group V (application of vertebratomy out of main abnormality area in conjunction with bilateral access with polysegmental fixation), there were 10 children with average age of 97.2 ± 15.5 months. Scoliosis was 41.2° ± 3.7°, correction 71.6%. Kyphosis was not noted. Blood loss was 303 ± 158.4 mL. Operation time was 224 ± 83.6 minutes. Complications: 2 deformity progressions with reoperation. Correction loss was 0.87°. Discussion: Fixation method choice is determined by patient age, deformity size, and concomitant vertebral abnormality. It is reasonable to apply monosegmental bilateral fixation with extirpation of abnormal vertebra up to 5 years. At the age of 5 to 10 years, application of volume vertebrotomy and trisegmental fixation for deformity correction is explained. In children older than 10 years with extended deformities, application of polysegmental fixation is effective. Vertebrotomy application out of apical area for deformity correction in cervicothoracic level and lumbosacral passage shows good result of deformity correction and reduces operation duration and blood loss volume in comparison with classical method.

A188: The Influence of Cervical Spine Shape in Adolescence Idiopathic Scoliosis

Eduardo Moreira Pinto¹, Artur Teixeira¹, Jorge Alves², and António Miranda²

¹Entre Douro e Vouga Hospital Center, Santa Maria Feira, Portugal

²Tâmega e Sousa Hospital Center. AOSpine (Platinum Member), member of North American Spine Society (NASS) and Institut de la Colonne Vertébrale (ICV), Bordeaux, France; Member of the Board of SPOT (Portuguese Society of Orthopaedics and Traumatology), Penafiel, Portugal

Introduction: The human standing posture involves a delicate balance between the spine and pelvis. The thoracic kyphotic angle depends on postural and morphological parameters that will influence the distribution of loads and, consequently, the biomechanical environment of the spine. The discussion by the scientific community regarding the shape of the cervical spine in idiopathic scoliosis has been continuing for several decades. Despite the investigations carried out to understand the relationship between the spinal curves, there are very few studies that have tested a correlation between cervical alignment and the thoracic curvature of the spine divided in to segments. The objective of this study is to evaluate the relationship between cervical alignment (lordotic, kyphotic, or straight), thoracic kyphosis and lumbar lordosis in patients with adolescent idiopathic scoliosis (AIS). Material and Methods: We conducted a retrospective study of 80 patients with AIS. The parameters evaluated were age, sex, scoliotic curvatures differentiated according to Lenke classification, cervical spine alignment, lumbar lordosis, and thoracic sagittal Cobb angle (between T1-T5, T5-T12, and between T1-T12). Results: In patients with AIS, the superior sagittal thoracic Cobb segment, contrary to the inferior, demonstrated a significant positive correlation with cervical spine alignment (P < .05). As there is an increase in superior thoracic Cobb segment, there is an increase in cervical lordosis. Conclusion: In AIS, proximal (T1-T5) and distal (T5-T12) thoracic kyphosis have different contributions on the global thoracic sagittal curvature. Therefore, there appears to be a symbiosis between the cervical spine shape and the sagittal Cobb angle of the proximal thoracic spine because as there is a decrease in lordotic curvature of the cervical spine there is a tendency to decrease the upper thoracic curvature. On the other hand, the distal thoracic segment or lumbar lordosis did not present a significant correlation with the shape of the cervical spine.

A189: Does the Implant Density Change the Correction Rate of the Main and Accompanying Minor or Major Structural and Nonstructural Curves? A Comparison Between Consecutive and Intermittent Pedicle Screw Constructs

Alpaslan Senkoylu¹, Mehmet Cetinkaya¹, Ismail Daldal², Ali Eren³, and Erdem Aktas⁴

¹Medical School of Gazi University, Ankara, Turkey

²Sakarya Training and Research Hospital, Sakarya, Turkey

³Gumushane Kelkit Government Hospital, Gumushane, Turkey

⁴TOBB University, Ankara, Turkey

Introduction: There is no consensus on the optimum pedicle screw density needed for the desired rigidity to be provided up to the solid fusion mass is succeeded in adolescent idiopathic scoliosis (AIS). The aim of the study was to evaluate the clinical outcomes and the coronal correction rate of the main and also the accompanying curves of AIS corrected with pedicle screws inserted bilaterally by posterior-only approach whether consecutively or intermittently including only 2 types of construct designs. Methods: The prospectively collected data of patients who underwent corrective surgery for AIS between January 2010 and December 2015 were reviewed retrospectively. Two groups were constituted according to the pedicle screw construct type whether implanted at all levels (CPSC) or intermittently (IPSC). The preoperative, early postoperative, and 24-month follow up X-rays and SRS-22 scores were reevaluated. The preoperative and postoperative Cobb angles of the main and accompanying curves, the correction rate, and the flexibility of the curves were calculated. Results: There were 60 patients (8 male, 52 female), 30 in each group. The mean age was 14.6 ± 2.5 years. The mean preoperative Cobb angle was 57.03° and 57.46°, the mean postoperative Cobb angle was 14.93° and 14.4°, and the mean correction rate was 76.22% and 75.31% in IPSC and CSPC groups, respectively (P > .05 for all). The preoperative and postoperative accompanying curve magnitudes and correction rates were also similar (P > .05 for all). Those radiographic outcomes were also consistent with clinical self-reported outcomes evaluated by SRS-22 questionnaire (P > .05). Conclusion: Both pedicle screw constructs placed with posterior-only approach whether intermittently or consecutively for correction of adolescent idiopathic scoliosis had satisfactory outcomes following the surgery. These outcomes were confirmed by both SRS-22 scores and the radiographs taken perioperatively and at follow-ups. The IPSC and CSPC groups did not demonstrated any statistically significant change in correction rate of the main and minor or major accompanying structural and nonstructural curves, and also in SRS-22 scores.

A190: Predictors of Postoperative Pain and Satisfaction in Patients With Adolescent Idiopathic Scoliosis

Anastasios Charalampidis¹, Hans Möller¹, and Paul Gerdhem¹

¹Karolinska Institutet and Karolinska University Hospital, Stockholm, Sweden

Introduction: Spinal fusion surgery for the treatment of idiopathic scoliosis can result in postoperative back pain and dissatisfaction. Although in many cases back pain may not be persistent and decreases over time in the postoperative period, in other cases patients will continue to experience pain for a long period after surgery. Therefore, identifying factors contributing to persistent pain following fusion surgery may help optimize the treatment strategies for these patients, resulting in better clinical outcomes. Material and Methods: A total of 389 patients with juvenile or adolescent idiopathic scoliosis, aged 10 through 25 years at surgery, were identified in the Swedish Spine register (Swespine), having patient reported outcome at a mean of 3.5 years. The patients were divided into groups based on postoperative reported visual analogue scale back pain (poorest quartile vs the others) and the satisfaction question in the SRS-22r with answers 1-3 representing dissatisfied patients and 4-5 representing satisfied patients. Predictors of postoperative back pain and dissatisfaction were searched in the preoperative data. The Student t test, the Pearson chi-square test, the Fisher exact test, or logistic regression were used for statistical analysis. Missing values were imputed using SPSS. Results: Patients in the highest quartile of postoperative back pain had significantly higher preoperative back pain and body mass index (BMI), and lower subdomain scores for SRS-22r function, pain, mental health, subtotal score, EQVAS, and EQ-5D (all P ≤ .018) than the others. Postoperatively, patients in the highest quartile of back pain had an increase of their preoperative back pain visual analogue scale (VAS) scores with 16.49 points (95% CI [10.50, 22.48]), and a decrease of EQ-5D index (0.08 [0.01, 0.15]) and SRS-22r scores (0.43 [0.12, 0.75]). In the other quartiles, preoperative back pain VAS scores decreased (12.26 [8.73, 15.80]) and EQ-5D and SRS-22r increased (0.10 [0.07, 0.12] and 0.30 [0.17, 0.43], respectively). Dissatisfied patients had lower preoperative subdomain scores for SRS-22r function, pain, self-image, subtotal score, lower EQVAS, lower score in EQ-5D and longer length of stay after surgery (all P ≤ .023) compared with the other quartiles. Females had a greater risk of developing back pain (P = .021), but gender was not associated with postoperative satisfaction. A multiple logistic regression was performed to ascertain predictors of increased postoperative back pain and dissatisfaction. A high level of preoperative back pain was independently associated with high postoperative back pain (odds ratio 1.03 [1.01, 1.04]). High preoperative SRS-22r pain scores indicating a low level of pain were independently associated with high satisfaction postoperatively (1.75 [1.16, 2.62]). Conclusion: A high level of preoperative pain was associated with high postoperative back pain and dissatisfaction. Patients and caregivers should be aware of the risk of postoperative dissatisfaction and back pain in patients with preoperative back pain.

A191: The Adolescent Idiopathic Scoliosis International Disease Severity Study: Do Operative Curve Magnitude and Complications Vary by Country?

Courtney Toombs¹, Baron Lonner², Akil Fazal³, Oheneba Boachie-Adjei⁴, Tracey Bastrom⁵, Ferran Pellise⁶, Mohamed Ramadan⁷, Wael Koptan⁸, Yasser Qiu⁸, Feng Zhu⁹, Yong Qiu⁹, and Harry Shufflebarger¹⁰

¹Yale New Haven Hospital, New Haven, CT USA

²Mount Sinai Beth Israel Hospital, New York, NY, USA

³Nairobi Spine and Orthopaedic Centre, Nairobi, Kenya

⁴FOCOS Orthopaedic Hospital, Accra, Ghana

⁵Rady Children’s Hospital, San Diego, CA, USA

⁶Hospital Vall d’Hebron, Barcelona, Spain

⁷Tanta University, Tanta, Egypt

⁸Cairo University, Cairo, Egypt

⁹Nanjing University Drum Tower Hospital, Nanjing, China

¹⁰Miami Children’s Hospital, Miami, FL, USA

Background: The prevalence of adolescent idiopathic scoliosis (AIS) in diverse regions of the world has been studied. Access to care varies widely and differences in disease severity and operative treatment outcomes are not well understood. This study aimed to determine variation in disease presentation and operative complications for AIS patients from an international cohort. Methods: This is a retrospective study carried out at 7 surgical centers in the United States (Manhattan and Miami), Ghana, Pakistan, Spain, Egypt, and China. A total of 541 consecutive patients with AIS were evaluated. Preoperative major curve magnitude, operative parameters and complications were compared among sites using ANOVA with post hoc tests and Pearson’s correlation coefficients. Univariate and multivariate forward stepwise binary logistic regressions determined the variables most predictive of complications. Results: Lowest-average-income countries (Ghana, Egypt, and Pakistan) displayed larger curves, more levels fused, longer operating time (OT), and greater estimated blood loss (EBL) than the other countries (P ≤ .001). Increasing curve magnitude was correlated with greater levels fused, longer OT, and greater EBL in all groups (P = .01). In the univariate regression analysis, Cobb magnitude, levels fused, EBL, and OT were associated with complication occurrence. Only OT remained significantly associated with complication occurrence after adjusting for Cobb magnitude, levels fused and site (OR = 1.005, CI 1.001-1.007, P = .003). Complications were greatest in Pakistan and Ghana (21.7% and 13.5%, respectively) and lowest in Miami (6.5%). Conclusions: Larger curve magnitudes in the least-income countries correlated with more levels fused, longer OT, and greater EBL, indicating that increased curve magnitude at surgery could explain the difference in operative morbidity between low- and high-income countries. With operative time as the prevailing predictive factor of complications, we suggest that increased curve magnitude leads to longer OTs and more complications. A lack of access to orthopedic care may be the largest contributor to the postponement of treatment.

A192: Changing Pelvic Anatomical Morphology During Adolescence: A Longitudinal Study

Hongda Bao¹, Zhen Liu¹, Xu Sun¹, Jun Jiang¹, Saihu Mao¹, Bangping Qian¹, Bin Wang¹, Yang Yu¹, Zezhang Zhu¹, and Yong Qiu¹

¹Drum Town Hospital, Nanjing University Medical School, Nanjing, China

Introduction: The importance of pelvis has been well recognized in recent decades. There are several studies showed the same results that pelvic incidence (PI) is growing during adolescence, but few reported the specific increase pattern. The measurement of PI incorporates “sacral plate” from the sacrum and “femoral axis” from the ilia, so the anatomical morphology changes of pelvis and sacrum may influence the PI. The purpose of this study was to evaluate the relationship between PI and bone age using skeletal maturity indicators and to investigate the pattern of changing PI during peak skeleton growth in puberty. We also studied whether the PI is associated with the anatomical morphology changes of pelvis and sacrum. Material and Methods: This retrospective longitudinal study recruited patients with adolescent idiopathic scoliosis (AIS). The inclusion criteria were age between 9 and 18 years with full spine images and having at least 3 follow-ups. The radiological anatomical parameters were measured in each follow-up include PI, pelvic height (PH), pelvic width (PW), sacral width (SW), femoral head-sacrum (FH-S), sacrum-coccyx (S-C) length, sacrum-coccyx (S-C) distance, and sacral curvature ratio. According to the skeletal maturity, all subjects were divided into 3 groups: Low Risser Group (Risser 0 to 1), Moderate Risser Group (Risser 1 to Risser 3), and High Risser group (Risser 3 to 5). From the intra- and interobserver reliability analysis, the intraclass correlation coefficient (ICC) was calculated. Results: A total of 318 patients with AIS were included in our study. The fastest growth velocity of PI is in Risser 1 (female 1.5°/year and male 1.6°/year). The PI also increased rapidly in Risser 0.5 (female 1.2°/year and male 1.5°/year) and Risser 2 (female 1.1°/year and male 1.4°/year). When the Risser grade reaches 3, which was a sign of the end of skeleton growth, the PI still has increasing potential and it increases slower in Risser 4 and Risser 5. There showed significant correlation between PI and age in the 3 groups (P < .05) and the correlation is stronger in Low Risser Group and Moderate Risser Group than that in High Risser Group. There showed significant correlation between PW and PI in all groups while PH and PI showed correlation only in Low Risser Group. There was significant correlation between SW and PI, FH-S and PI, S-C length and PI (P < .05) and the Moderate Risser Group had the strongest correlation. The S-C distance and S-C ratio did not show any correlation with PI (P < .05) expect the male of High Risser group. Conclusion: PI shows and increase during skeleton growth from the age of 9 to 17 years. It has the fastest growth in Risser 1 and retains growth potential in Risser 5. Male adolescents have lager PI. PI growth is correlated with pelvic height, femoral head-sacrum distance, pelvic width, and sacral width.

A193: Liposomal Bupivacaine Use in Pediatric Spinal Deformity Surgery; A Controlled Cohort Study

Assem Sultan¹, Morad Chughtai¹, Reid Chambers¹, Joel Kolmodin¹, John Seif¹, Thomas Kuivila¹, Ryan Goodwin¹, and Robert Tracy Ballock¹

¹Cleveland Clinic, Cleveland, OH, USA

Introduction: The use of liposomal bupivacaine (LB) has recently emerged in multiple orthopedic specialties, including adult spinal surgeries. However, there is a paucity of studies investigating its utility for postoperative pain control in pediatric spine deformity patients undergoing surgical correction. Therefore, the purpose of this study was to compare (1) pain relief efficacy, (2) opioid consumption, (3) length of stay (LOS), (4) discharge disposition, and (5) its safety and complications, in pediatric patients who underwent spinal deformity correction surgery and received LB as a part of their postoperative pain management protocol versus patients who only received a standard pain management protocol. Methods: This study was a retrospective review of a single-institution database of pediatric patients who underwent spinal deformity surgery using a specific inclusion and exclusion criteria. A total of 195 patients that received LB as part of their postoperative pain management regimen were identified and compared to 128 patients who received standard pain management without LB. Pain intensity (using numeric pain scale), opioid consumption, LOS, and discharge disposition were recorded. Potential LB-related complications were reported as frequencies and compared for superiority using Pearson chi-square tests or Fisher exact tests. Noninferiority tests were performed assuming a 10% margin, using the Farrington-Manning score test, with adjusted P values using the Holm-Bonferroni step-down method. The overall distribution of complications was analyzed using multivariate tests based on generalized estimating equations to determine the common and average treatment effects. Odds ratios with 95% confidence intervals from these tests were calculated and presented. The analyses were repeated using just the first surgery per subject, to avoid the potential influence caused by repeat surgeries on the same subject. Results: The LB cohort demonstrated lower pain scores (postoperative day 1 [POD1]; 2.8 ± 2.8 vs 4.6 ± 2.6, POD2; 2.9 ± 2.7 vs 4.4 ± 2.7, P < .0001), lower overall opioid consumption (78.2 vs 129 morphine equivalents [MME]; P = .0001) and consistently from POD0 through 3 (mean differences; 7.47, 9.04, 17.2, 17.3 MME and P < .05), shorter LOS (median = 3 days, interquartile range [IQR] = 3-4 vs 4 days; IQR = 4-6 days, P < .001), and similar discharge-to-home disposition (98% vs 97%). Complications were similar among the cohorts in superiority and 10% noninferiority analyses. Patients in the LB cohort had lower odds for complications using both the common effect and average treatment effect statistical analyses (odds ratios [OR]; 95% confidence intervals [95% CI] = 0.77; 95% CI = 0.64-0.93, P = .009 and 0.67; 95% CI = 0.50-0.90, P = .008). Similar findings were consistently demonstrated in the subanalysis of patients who underwent only primary procedures. Conclusion: The present study reports the use of liposomal bupivacaine in pediatric patients undergoing spinal deformity correction surgery demonstrating its safety and efficacy when added to the postoperative pain management regimen for these patients. Our study remains the first to investigate its use in this population of patients and will serve as an impetus for further research.

OP30: Tumor 4

A194: Current Treatment Strategy for Newly Diagnosed Chordoma of the Mobile Spine and Sacrum: Results of an International Survey

Nicolas Dea¹, Charles Fisher¹, Jeremy Reynolds², Joseph Schwab³, Laurence D. Rhines⁴, Ziya Gokaslan⁵, Chetan Bettegowda⁶, Arjun Sahgal⁷, Aron Lazary⁸, Alessandro Luzzati⁹, Stefano Boriani¹⁰, Alessandro Gasbarini¹¹, Ilya Laufer¹², Raphaele Charest-Morin¹, Feng Wei¹³, William Teixeira¹⁴, Niccole Germscheid¹⁵, Francis Hornicek¹⁶, Thomas DeLaney³, and John Shin³

¹Vancouver General Hospital, Vancouver, British Columbia, Canada

²Oxford University Hospital, Oxford, UK

³Massachusetts General Hospital, Boston, MA, United States

⁴The University of Texas MD Anderson Cancer Center, Houston, TX, USA

⁵The Warren Alpert Medical School of Brown University and Rhode Island Hospital and The Miriam Hospital, Providence, RI, USA

⁶Johns Hopkins University School of Medicine, Baltimore, MD, USA

⁷University of Toronto and Sunnybrook Odette Cancer Centre, Toronto, Ontario, Canada

⁸National Center for Spinal Disorders and Buda Health Center, Budapest, Hungary

⁹ Centro Di Chirurgia Oncologia Ortopedica e Ricostruttiva del Rachide (CCOORR), IRCCS Istituto Ortopedico Galeazzi, Milan, Italy

¹⁰GSpine4 Spine Surgery Division, IRCCS Galeazzi Orthopedic Institute, Milan, Italy

¹¹Rizzoli Institute, Bologna, Italy

¹²Memorial Sloan Kettering Cancer Center, New York, NY, USA

¹³Peking University Third Hospital, HaiDian District, Beijing, China

¹⁴Institute of Cancer of the State of São Paulo, Sao Paulo, Brazil

¹⁵AOSpine International, Davos, Switzerland

¹⁶University of California, Los Angeles, CA, USA

Introduction: To understand the spectrum of current treatment protocols for managing newly diagnosed chordoma of the mobile spine and sacrum. Material and Methods: A survey on the treatment of spinal chordoma was distributed electronically to members of the AOSpine Knowledge Forum Tumor, including neurosurgeons, orthopedic surgeons, and radiation oncologists from North America, South America, Europe, Asia, and Australia. Survey participants were preidentified clinicians from centers with expertise in the treatment of spinal tumors. The data were collected and analyzed using descriptive statistics. Results: A total of 39 of 43 (91%) participants completed the survey. Most (80%) favored en bloc resection without preoperative neoadjuvant radiation therapy (RT) when en bloc resection is feasible with acceptable morbidity. The main area of disagreement was with the role of postoperative RT, where 41% preferred giving RT only if positive margins were achieved and 38% preferred giving RT irrespective of margin status. When en bloc resection would result in significant morbidity, 33% preferred planned intralesional resection followed by RT, and 33% preferred giving neoadjuvant RT prior to surgery. In total, 8 treatment protocols were identified: 3 in whom en bloc resection is feasible with acceptable morbidity and 5 in whom en bloc resection would result in significant morbidity. Conclusion: The results confirm that there is treatment variability across centers worldwide for managing newly diagnosed chordoma of the mobile spine and sacrum. This information will be used to design an international prospective cohort study to determine the most appropriate treatment strategy for patients with spinal chordoma.

A195: Impact of Neurologic Deficits on Surgical Outcomes and HRQoL Following Treatment for Metastatic Epidural Spinal Cord Compression

Ori Barzilai¹, Anne Versteeg², Arjun Sahgal³, Laurence Rhines⁴, Daniel Sciubba⁵, James Schuster⁶, Michael Weber⁷, Peter Pal Varga⁸, Stefano Boriani⁹, Chetan Bettegowda⁵, Micahel Fehlings⁵, Michelle Clarke¹⁰, Paul Arnold¹¹, Ziya Gokaslan¹², Charles Fisher¹³, Ilya Laufer¹; AOSpine Knowledge Forum Tumor

¹Memorial Sloan Kettering Cancer Center, New York, NY, USA

²Utrecht University Medical Center, Utrecht, the Netherlands

³ Sunnybrook Odette Cancer Center and University of Toronto, Toronto, Ontario, Canada

⁴The University of Texas MD Anderson Cancer Center, Houston, TX, USA

⁵Johns Hopkins University School of Medicine, Baltimore, MD, USA

⁶Hospital of the University of Pennsylvania, Philadelphia, PA, USA

⁷McGill University and Montreal General Hospital, Montreal, Quebec, Canada

⁸National Center for Spinal Disorders and Buda Health Center, Budapest, Hungary

⁹GSpine4 Spine Surgery Division, IRCCS Galeazzi Orthopedic Institute, Milan, Italy

¹⁰Mayo Clinic, Rochester, MN, USA

¹¹The University of Kansas Hospital, Kansas City, KS, USA

¹²The Warren Alpert Medical School of Brown University and Rhode Island Hospital and The Miriam Hospital, Providence, RI, USA

¹³University of British Columbia and Vancouver General Hospital, Vancouver, British Columbia, Canada

Introduction: A critical knowledge gap exists regarding the impact of neurologic deficits on surgical outcomes and health-related quality of life (HRQoL) in patients surgically treated for metastatic epidural spinal cord compression (MESCC). This study analyzes the impact of neurological status on functional status, HRQoL measures and postoperative survival. Methods: Patient demographic data, overall survival, ASIA impairment scale, the Spinal Instability Neoplastic Score (SINS), treatment details including perioperative complications, and HRQoL measures, including the Short Form 36 version 2 (SF-36v2) and the Spine Oncology Study Group Outcomes Questionnaire (SOSGOQ2.0) data were extracted from a multicenter international prospective cohort study (EPOSO). Results: A total of 239 patients were included. 158 patients presented with American Spinal Injury Association (ASIA) E, 60 were ASIA D, and 21 ASIA A-C at baseline. Higher SINS scores at baseline correlated significantly (P = ? ) with lower ASIA scores. At 6 weeks post treatment, 113 of 119 patients with ASIA E remained stable, 23 of 44 patients with ASIA D improved to ASIA E, and 6 out of 11 patients with ASIA A-C at baseline improved to ASIA D or E. Patients with baseline ASIA scores of E or D survived significantly (P = .0004) longer than patients with ASIA A-C scores. At 6 weeks follow-up, better ASIA scores were associated with better scores on SF-36v2 and SOSGOQ items, including walking 100 yards (P = .006), leg weakness (P < .001), spine-related pain (P = .048), and bowel and bladder control (P = .005), and at 12 weeks, included walking 100 yards (P = .039), leg weakness (P < .001), activity level (P = .006), and working ability (P = .043). Postoperatively, patients with ASIA A-D were more likely to have urinary tract infections (P < .001) and wound drainage (P = .003) than those with ASIA E. Conclusions: Patients with neurologic deficits due to MESCC have worse HRQoL, decreased overall survival, and increased risk of surgical complications. Nevertheless, surgery can result in stabilization or improvement in neurologic function which translates into better HRQoL. Postoperative care and follow-up are challenging for patients with neurologic deficits as they experience more postoperative complications.

A196: Development of Clinical Prognostic Models for Postoperative Survival and Quality of Life in Patients With Metastatic Epidural Spinal Cord Compression Treated Surgically

Anick Nater¹, Junior Chuang¹, Kuan Liu¹, Nasir A. Quraishi², Dritan Pasku², Jefferson R. Wilson¹, and Michael Fehlings¹

¹University of Toronto, Toronto, Ontario, Canada

²Queen’s Medical Centre, Nottingham University Hospital NHS Trust, Nottingham, UK

Introduction: Surgery is generally considered for patients with metastatic epidural spinal cord compression (MESCC) with at least 3 months of life expectancy. No existing clinical prognostic models (CPMs) are consistently used, and no CPMs predict quality of life (QoL). We aimed to develop the first CPMs of postoperative survival and QoL, using Transparent Reporting of a Multivariable Prediction Model for Individual Prognosis or Diagnosis (TRIPOD) guidelines and based on the largest sample of prospective data of surgical patients with MESCC. Material and Methods: Using TRIPOD guidelines, we created 1-year survival and QoL CPMs using data from 258 patients (AOSpine North America [NA] MESCC study and Nottingham MESCC registry) using a Cox model, logistic regression and manual backward elimination. The outcome measure for CPMs of QoL was the minimal clinical important difference (MCID) in EQ-5D scores. Internal validation involved 200 bootstrap iterations; calibration and discrimination were evaluated. Results: A higher Short Form–36 (SF-36) physical component score (PCS) was associated with longer survival (hazard ratio [HR]: 0.96) whereas primary tumor other than breast, thyroid, and prostate (unfavorable, eg, lung, HR: 2.57; others, HR: 1.20), organ metastasis (HR: 1.51), male (HR: 1.58), and MESCC treated with radiotherapy preoperatively (HR: 1.53) were not (c-statistic: 0.69 [95% CI: 0.64-0.73]). Four factors were associated with the likelihood of achieving a MCID improvement in EQ-5D at 3 months: KPS <70% (OR: 2.50), living in NA (OR: 4.06), SF-36 PCS (OR: 0.95) and mental component (OR: 0.96) (c-statistic: 0.74 [95% CI: 0.68-0.79]). Calibration for both CPMs was very good. Conclusion: We developed and internally validated the first CPMs of survival and QoL at 3 months postoperatively in patients with MESCC using TRIPOD guidelines. A web-based calculator is available (http://spine-met.com) to assist clinical decision making in this complex patient population.

A197: Surgical Treatment of Spinal Metastases: Results of a Large Single-Center Study

Christoph Stihsen¹, Bernhard Springer¹, Philipp Funovics¹, Josef Grohs¹, Petra Krepler¹, and Reinhard Windhager¹

¹Medical University of Vienna, Vienna, Austria

Introduction: Few studies focusing on spinal metastases receiving surgical treatment are currently available. The objective of the current study was to analyze a series of consecutive patients undergoing surgery for spinal metastases in order to identify factors that influence overall survival. Material and Methods: In a retrospective single-center study, 209 patients with spinal metastases who received surgical treatment between 1990 and 2015 were reviewed. The group consisted of 87 female and 122 male patients with a mean age of 60 years (median 61 years, range 17-87 years). The mean overall follow-up of the cohort was 12 months (range 0.1-160 months). The most frequent underlying malignancies were renal cell carcinoma in 65 cases (31%), breast cancer in 35 cases (17%), lung cancer in 33 cases (16%), cancer of the colon in 17 cases (8%), prostate cancer in 14 cases (7%), and thyroid cancer in 8 cases (4%). Results: Overall survival was 70%, 45% and 27% at 6 months, 1 year, and 2 years postsurgery, respectively. On multivariate analysis, breast- and thyroid-cancer metastases showed a significantly improved survival, compared with the other entities (P = .01, P = .02). The presence of visceral metastases proved to have a significant negative effect on overall survival (P < .001). The surgical approach (ventral vs dorsal). did not influence the total survival significantly (P = .21). There could be detected significantly more complications in the ventral approach cohort (P = .03). Conclusion: Here we present the results of one of the largest single institution studies as found in the literature, analyzing the surgical outcome of spinal metastases. The histological entity of the primary tumor proved to be a strong prognostic factor with regard to overall survival. Our analyses showed no difference in survival between dorsal and ventral surgical approach; however, the ventral approach group proved to have significantly more complications. The question arises to what extend such complications affect the quality of life of patients. Further investigations have to be performed.

A198: Molecular Phenotypes, Topical Survival Times and Therapy Options in Spinal Metastatic Breast Cancer

Daniel Adler¹ and Michael Akbar¹

¹Ruprecht-Karls-University Heidelberg, Heidelberg, Germany

Introduction: With 70 000 new patients per year breast cancer is the most frequent tumor entity in women and responsible for 29% of new malignant tumors. In up to 85%, breast cancer develops spinal metastases, which may cause instability with pain, pathological fractures and in up to 10% neurological deficits. A total of 50% of women with breast cancer are forced to change working sphere and 37% cannot work at times or not at all, which poses a serious socioeconomic problem. Around the turn of the millennium, 5 intrinsic subtypes in breast cancer were described, 4 of them clinically relevant (luminal A, luminal B, HER2-enriched, basal like/triple-negative). With regard to expected survival time, substantial differences are described in these phenotypes. Material and Methods: We carried out a search of the topical literature. Results: To adequately illustrate the poor survival time in triple negative and HER-2 enriched phenotype until the year 2015, recommendations were published to modify score systems (revised Tokuhashi- and Tomita score). Highly invasive and risk-afflicted operative techniques should be indicated critically with hormone receptor-negative and in particular with triple negative receptor status. Topical oncological draughts of treatment with breast cancer offer improved therapy options and actually provide clearly longer survival times. With endocrine responsive phenotypes (luminal A and B) median survival time (MST) of 60 months are described (26 months in 2015). After the introduction of new anti-HER-2 therapeutics (Trastuzumab, Pertuzumab, etc) MST for HER-2 enriched subtypes is actually published with 50 months (9 months in 2015). In patients with triple-negative phenotype MST is described with 15 to 19 months (6-9 months in 2015). Conclusions: Prior to elective surgery, the evaluation of the present illness course, the topical metastatic status, and the molecular classification is recommended. With endocrine responsive (luminal A and B) phenotypes actual literature describes a MST of 60 months, with HER-2 enriched types 50 months and with triple negative tumors a MST of 15 to 19 months. Against earlier published results, we explicitly do not recommend a downstaging in the Tokuhashi and Tomita score. Solitaire metastases in principle can be treated independently of molecular phenotype by curative means (MST at least 15-19 months). Biological age, general condition, accompanying illnesses, intrinsic subtype, tumor load (number and extent of metastatic organs) as well as preceding therapy lines influence decision making. Cooperation of spinal surgeons and oncologists is essential to define adequate therapy strategies.

A199: A Comparison of Metastasectomy and Decompression With Additional Stereotactic Radiotherapy in Patients With Spinal Metastasis of Renal Cell Carcinoma

Nikita Zaborovskii¹, Dmitrii Ptashnikov¹, Dmitrii Mikhaylov¹, Oleg Smekalenkov¹, Sergei Masevnin¹, and Anton Denisov¹

¹R. R. Vreden Russian Research Institute of Traumatology and Orthopedics, Saint-Petersburg, Russian Federation

Introduction: Spinal metastasectomy can increase the likelihood of disease-free survival, the benefits of these surgery must be weighed against the risk for increased perioperative morbidity and mortality rates. Stereotactic body radiotherapy (SBRT) is an emerging treatment option for spinal metastases with demonstrated efficacy in the upfront, postoperative, and retreatment settings, as well as for tumor histologies considered radioresistant. The purpose of our research is to compare spinal metastasectomy and palliative spinal decompression with additional SBRT. Material and Methods: A retrospective study included patients undergoing surgery for spinal metastases of renal cell carcinoma (mRCC) operated within 2009 and 2015. Analyzed data consisted of 40 patients (27 men and 13 women with a median age of 52.5 years [95% CI of median 42.5-55] years). There were 2 groups of patients. The metastasectomy group included 24 patients who underwent complete resection of spinal metastasis via total spondylectomy. Decompression-SBRT group consisted of 14 patients, who were treated using spinal palliative decompression and additional SBRT postoperatively. We evaluated overall survival and local tumor control at follow up. The comparisons were done by using Kaplan-Meier survival analysis and log-rank test as well as exact permutation test. Results: Overall median follow-up times were 43 months (1-year survival rate, 95.8%; 3-year survival rate, 78.9%) in the metastasectomy group and 38 months (1-year survival rate, 93.8%; 3-year survival rate, 57%) in the decompression-SBRT group (P = .450). There were 4 (16.7%) cases of local recurrence after metastasectomy and 4 (25%) cases of local recurrence after decompression with additional SBRT. We did not find significant differences between groups (P = .691). Conclusion: Given that patients with spinal metastases often have contraindications for complete metastasectomy, a spinal decompression with additional SBRT may alternative as this procedure provide effective local tumor control.

A200: Zoledronate at Low Doses Maintains Its Antitumor Effect in Prostate Cancer–Induced Bone Metastasis Cells

Elie Akoury¹, Pouyan Ahangar¹, Antone Nour¹, Derek Rosenzweig¹, and Michael Weber¹

¹McGill University, Montreal, Quebec, Canada

Introduction: Up to 80% of patients with primary tumors of the prostate, breast, or lung will develop spine metastases. These bone metastases are the most common cause of cancer-related pain, leading to complications like bone fracture and spinal cord injury, which negatively affect quality of life. Treatment modalities include the use of Zoledronate, a bisphosphonate class drug. Zoledronate has been proven not only to be an effective inhibitor of tumor-induced bone resorption and associated pain, but also to have a direct antitumor activity. Systemic administration at high doses is the current approach to deliver Zoledronate. However, such procedure causes debilitating side effects in patients. Our aim is to test in vitro the effect of low doses of Zoledronate for a long period. Materials and Methods: We used LAPC4, a prostate cancer line derived from lymph node metastasis, and primary cells from patients with prostate-induced bone metastasis. Cells were incubated with Zoledronate at low concentrations (1, 3, and 10 µM) for 7 days in low serum conditions (1% fetal bovine serum) and assessed for proliferation, migration and invasion using alamarblue/Vybrant MTT, Falcon cell culture inserts and Cultrex 3D spheroid cell invasion kits, respectively. Statistical analysis was preformed using 1- or 2-way ANOVA and Tukey post hoc tests at a 95% confidence level. Results: In a dose-dependent manner, we show that Zoledronate significantly decreased cell proliferation mostly at 10 µM in LAPC4 (31% ± 17.7%, P < .001) and the prostate-induced bone metastasis cells (40% ± 25.7%, P = .01). The MTT cell proliferation data mirrored those of the alamarblue assay. Cell migration and 3D spheroid invasion were only decreased following Zoledronate 10 µM treatment in LAPC4 ([22.8% ± 8.1, P = 0.04] [6.9% ± 1.7%, P = .05]) and in the prostate-induced bone metastasis cells ([62.3 ± 23.4%, P = .04] [27.9% ± 20.16%, P = .001]). Conclusions: This project harnesses the potential of using Zoledronate at low doses for long treatment periods while retaining its antitumor effect on prostate-induced metastases in clinically relevant cell populations.

A201: What Is the Role of Surgery for Spinal Metastasis of Lung Cancer? Propensity Score Matched Analysis With Radiotherapy Alone

Hyung-Youl Park¹, Young-Hoon Kim¹, Kee-Yong Ha¹, Sang-Il Kim¹, and Jun-Hyung Cho¹

¹Seoul St. Mary’s Hospital, College of Medicine, The Catholic University of Korea, Seoul, Korea, Seoul, Republic of Korea

Introduction: The incidence of metastatic spine disease is increasing due to improved survival of cancer patients. Despite its frequency, the debate on therapeutic modality still remains unsettled for some limitations such as heterogeneous primary sites and different patient’s condition. In addition, molecular targeted therapy has been introduced and reported to be effective for some cancers. The purpose of this study is to evaluate the efficacies of surgery as well as targeted therapy limited to spinal metastasis of lung adenocarcinoma. Material and Methods: Twenty-two patients treated with surgery and radiotherapy (surgery group) were matched in a 1-to-1 format with 95 patients undergoing radiotherapy (radiotherapy group) by age, sex, revised Tokuhashi score, spinal instability neoplastic score (SINS) using propensity score for lung adenocarcinoma. Performance status as Eastern Cooperative Oncology Group (ECOG) and neurologic status, including Frankel grade and possibility of ambulation, were assessed between the surgery group and radiotherapy group. In subgroup analysis, clinical outcomes and survival were evaluated along with type of surgery, and molecular targeted therapy for mutation of epidermal growth factor receptor (EGFR) or anaplastic lymphoma kinase (ALK). Results: Matched patient demographics of 22 patients were not significantly different between the surgery group and radiotherapy group. However, prevalence of ECOG ≥3 and Frankel grade ≤C was significantly higher in the surgery group (27.3% vs 4.5%, P = .039, respectively). The number of the patients unable to ambulate was also significantly higher in the surgery group (27.3% vs 4.5%, P = .039). Although clinical outcomes were similar after treatment, significant improvements of Frankel grade (22.7% vs 0%, P = .018) and ambulation (18.2% vs 0%, P = .036) were observed only in the surgery group. However, survival was not significantly different between the 2 groups (P = .790). Regarding the type of surgery, no significant differences of clinical outcomes and survival were observed between decompressive surgery and wide surgery. Patients with EGFR or ALK mutation were significantly associated with female gender (58.8% vs 22.2%, P = .014) and less exposure of smoking (2.1 ± 6.0 vs 20.1 ± 22.1 pack-years, P = .002). Although clinical outcomes were not significantly different according to the mutation of EGFR or ALK, survival in patients with mutation was significantly longer than without mutation (21.0 ± 2.8 vs 5.0 ± 1.5 months, P = .042). In addition, significant improvement of ECOG was observed in the mutation group (35.3% vs 7.4%, P = .02). Conclusion: Surgery combined with radiotherapy significantly improved Frankel grade and ambulation compared with radiotherapy alone, while molecular targeted therapy significantly improved the performance status and survival rate. In metastatic lung adenocarcinoma, patients with neurologic deterioration should be considered for surgery combined with radiotherapy, especially in case with mutation of EGFR or ALK.

A202: Survival, Local Control, and Health-Related Quality of Life in Oligometastatic and Polymetastatic Spinal Tumors: A Multicenter, International Study

Ori Barzilai¹, Anne Versteeg², Arjun Sahgal³, Laurence Rhines⁴, Mark Bilsky¹, Daniel Sciubba⁵, James Schuster⁶, Michael Weber⁷, Peter Pal Varga⁸, Stefano Boriani⁹, Chetan Bettegowda⁵, Micahel Fehlings¹⁰, Yoshiya Yamada¹, Michelle Clarke¹¹, Paul Arnold¹², Ziya Gokaslan¹³, Charles Fisher¹⁴, Ilya Laufer¹; the AO Spine Knowledge Forum Tumor

¹Memorial Sloan Kettering Cancer Center, New York, NY, USA

²Utrecht University Medical Center, Utrecht, the Netherlands

³Sunnybrook Odette Cancer Center and University of Toronto, Ontario, Canada

⁴The University of Texas MD Anderson Cancer Center, Houston, TX, USA

⁵Johns Hopkins University School of Medicine, Baltimore, MD, USA

⁶Hospital of the University of Pennsylvania, Philadelphia, PA, USA

⁷McGill University and Montreal General Hospital, Montreal, Quebec, Canada

⁸National Center for Spinal Disorders and Buda Health Center, Budapest, Hungary

⁹IRCCS Galeazzi Orthopedic Institute, Milan, Italy

¹⁰University of Toronto and Toronto Western Hospital, Toronto, Ontario, Canada

¹¹Mayo Clinic, Rochester, MN, USA

¹²The University of Kansas Hospital, Kansas City, KS, United States

¹³The Warren Alpert Medical School of Brown University and Rhode Island Hospital and The Miriam Hospital, Providence, RI, USA

¹⁴University of British Columbia and Vancouver General Hospital, Vancouver, British Columbia, Canada

Background: Treatment of oligometastatic (5 metastases or less) spinal disease has trended toward ablative therapies, yet little is known about the prognosis of patients presenting with oligometastatic spinal disease and the value of this approach. The objective of this study is to compare survival and clinical outcomes of cancer patients with oligometastatic spine disease to those with polymetastatic (more than 5 metastases) disease. Methods: International, multicenter, prospective study. Patients who were admitted to a participating spine center with a diagnosis of spinal metastases, underwent surgical intervention and/or radiation treatment between August 2013 and May 2017 were included. Data collected included demographics, overall survival, local control, treatment information including surgical, radiation and systemic therapy details. Health-related quality of life (HRQoL) measures included EuroQol-5D (EQ-5D-3L), Short Form 36 (SF-36v2), and the Spine Oncology Study Group Outcomes Questionnaire (SOSGOQ). Results: Of 393 patients included, 215 presented with oligometastatic and 178 with polymetastatic disease. A significant survival advantage of 90.1% versus 77.3% at 3 months, and 77.0% versus 65.1% at 6 months from time of treatment was found for patients presenting with oligometastatic disease. Importantly, both groups experienced significant improvement in multiple HRQoL measures at 6 months posttreatment with no differences in these outcome measures between the 2 groups. Conclusions: Treatment of oligometastatic disease offers a significant survival advantage compared to polymetastatic disease, regardless of treatment choice. HRQoL measures improve in both groups, demonstrating a palliative benefit for all treated patients.

OP31: Infections 2

A203: Propionibacterium Acnes: Biofilm Is Present in Nucleus Material From Human Disc Herniations

Christof Birkenmaier¹, Manu N. Capoor², Filip Ruzicka³, Jonathan E. Schmitz⁴, Garth A. James⁵, Tana Machackova⁶, Radim Jancalek³, Martin Smrcka⁷, Radim Lipina⁸, Fahad S. Ahmed⁶, Todd F. Alamin⁹, Neel Anand¹⁰, John C. Baird⁶, Nitin Bhatia¹¹, Sibel Demir-Deviren¹², Robert K. Eastlack¹³, Steve Fisher⁵, Steven R. Garfin¹⁴, Jaspaul S. Gogia¹⁵, Ziya L. Gokaslan¹⁶, Kalvin C. Kuo¹⁷, Yu-Po Lee¹¹, Konstantinos Mavrommatis¹⁸, Elleni Michu⁶, Hana Noskova⁶, Assaf Raz², Jiri Sana⁶, A. Nick Shamie¹⁹, Philip S. Stewart⁵, Jerry L. Stonemetz²⁰, Jeffrey C. Wang²¹, Timothy F. Witham²⁰, Michael F. Coscia²², Vincent A. Fischetti², and Ondrej Slaby⁶

¹Ludwig-Maximilian-University Munich, Munich, Germany

²Rockefeller University, New York, NY, USA

³Masaryk University, St Anne’s Faculty Hospital, Brno, Czech Republic

⁴Vanderbilt University School of Medicine, Nashville, TN, USA

⁵Montana State University, Bozeman, MT, USA

⁶Central European Institute of Technology (CEITEC), Masaryk University, Brno, Czech Republic

⁷University Hospital Brno, Masaryk University, Brno, Czech Republic

⁸University Hospital Ostrava, Ostrava University, Ostrava, Czech Republic

⁹Stanford University Medical Center, Stanford University, Stanford, CA, USA

¹⁰Cedars-Sinai Institute for Spinal Disorders, Los Angeles, CA, USA

¹¹University of California Irvine, School of Medicine, Irvine, CA, USA

¹²UCSF Medical Center, San Francisco, CA, USA

¹³Scripps Clinic Division of Orthopedic Surgery, San Diego, CA, USA

¹⁴University of California San Diego, San Diego, CA, USA

¹⁵Kaiser Permanente San Jose Medical Center, San Jose, CA, USA

¹⁶The Warren Alpert Medical School of Brown University, Rhode Island Hospital, Providence, RI, USA

¹⁷Kaiser Permanente Oakland Medical Center, Oakland, CA, USA

¹⁸Celgene Corporation, Information Knowledge and Utilization, San Francisco, CA, USA

¹⁹David Geffen School of Medicine, University of California, Los Angeles, CA, USA

²⁰The Johns Hopkins Hospital, Baltimore, MD, USA

²¹University Southern California, Los Angeles, CA, USA

²²²⁴OrthoIndy Hospital, Indianapolis, IN, USA

Introduction: A number of studies have found Propionibacterium acnes (P. acnes) in microbiologic cultures from intervertebral disc tissue extracted patients undergoing microdiscectomy in ∼25% of cases, suggesting a possible link between low-grade bacterial infection and disc degeneration. Since P. acnes is also a prominent skin commensal, there has been great difficulty in satisfactorily exclude the possibility that these organisms represent perioperative contamination rather than true infection. This study seeks to validate P. acnes prevalence in cultures from resected disc material, while providing microscopic evidence of P. acnes biofilm. Material and Methods: Nucleus material from 368 patients undergoing microdiscectomy for symptomatic disc herniation were divided into several portions, one being homogenized, subjected to quantitative anaerobic culture and assessed for bacterial growth. A second fragment was frozen for additional analyses. Bacterial colonies were identified by means of matrix-assisted laser desorption/ionization time of flight (MALDI-TOF) mass spectrometry and P. acnes phylotyping was conducted using multiplex polymerase chain reaction (PCR). For a sub-set of specimens, bacteria localization within the disc was assessed utilizing confocal laser scanning microscopy (CLSM) and fluorescent in situ hybridization (FISH). Results: Positive microbial cultures were obtained from 162 discs (44%), including 119 cases (32.3%) with P. acnes. In 89 cases, P. acnes was the only bacterium cultured; in 30 cases, it was isolated in combination with other bacteria (primarily coagulase-negative Staphylococcus spp.) Among positive specimens, the median P. acnes bacterial burden was 350 colony-forming units (CFU)/g (12 to ∼20 000 CFU/g). Thirty-eight P. acnes isolates were subjected to molecular subtyping, identifying 4 of 6 defined phylogroups: IA1, IB, IC, and II. Eight culture-positive specimens were evaluated by fluorescence microscopy and revealed P. acnes in situ. Notably, these bacteria demonstrated a biofilm distribution within the disc matrix. P. acnes bacteria were more prevalent in males than in females (39% vs 23%, P = .0013). Conclusion: This study confirms that P. acnes is highly prevalent in herniated disc tissue. It also provides the first visual evidence of P. acnes biofilms within such specimens. This evidence of P. acnes biofilm demonstrates beyond a reasonable doubt that a true infection exists rather than microbiologic contamination. These findings open the door to speculation whether P. acnes may play a role as a promotor of degenerative disc disease and whether such an infection correlates to a degenerative disc becoming clinically symptomatic.

A204: Improvement in Frankel Scale in Caries Spine Patients With Frankel Scale “A” Undergone Anterior Decompression and Cage With Autologous Bone Graft After a 5-Year Follow-up

Ammar Dogar¹, Haseeb Hussain¹, Amer Aziz¹, Shahzad Javed¹, and Rizwan Akram¹

¹Ghurki Trust Teaching Hospital, Lahore, Pakistan

Introduction: Spinal tuberculous infection is the most common and dangerous form of skeletal tuberculosis. It constitutes one-third to one-half of all bone and joint tuberculosis. It is a result of hematogenous dissemination from primary focus in the lungs, lymph nodes, and so on. According to the current estimates of the WHO, tuberculosis now kills 3 million people a year worldwide. Therefore, early diagnosis and management of spinal tuberculosis has special importance in preventing serious complications. Materials and Methods: It is a prospective case series study. A total of 419 patients either through the outpatient department or emergency department with Frankel Scale “A” from 2003 to 2011 were included. After taking proper history, examination, investigations, and proper consent, all the patients underwent anterior decompression and locally made cage with autologous bone graft. Data was initially entered on preformed proforma and then using SPSS 17.0. After surgery, patients were given brace for 6 months and were followed at 6 weeks, 6 months, and then yearly up to 5 years to assess improvement in Frankel scale. The objective of the study was to determine improvement in Frankel scale after anterior decompression and stabilization with locally made standalone cage and bone graft in caries spine. Result: A total of 419 patients presented with Frankel Scale “A” between 2003 and 2011. Twenty-one (5.01%) patients were lost to follow-up. Out of remaining 398 213 (53.52%) were male and 185 (46.48%) were females. A total of 123 (30.90%) patients were younger than 15 years, 167 (41.96%) were between 16 and 40 years old, and 108 (27.14%) were older than 40 years. All patients were having Frankel scale “A.” After 1 year, only 19 (4.77%) patients showed Frankel scale “E,” while after 3 and 5 years, respectively, 167 (41.96%) and 189 (47.49%) patients showed Frankel scale “E.” Overall, 90 (22.61%) patients showed Frankel scale “D,” 67 (16.83%) showed “C,” 24 (6.03%) showed “B,” while 28 (7.04%) patients showed no improvement after 5-year follow-up. Conclusion: Early diagnosis and early intervention give excellent results. However, the patients with frankel Scale “A” with delayed presentation after anterior decompression, stabilization with cage and bone graft results in significant improvement of Frankel scale. Moreover, there is no need for added instrumentation.

A205: Is Gene Xpert Rifampicin Resistance Really That Expert in Detecting MDR? Drug Sensitivity/Resistance Patterns of Gene Xpert Positive Tuberculous Spondylodiscitis

Mihir Upadhyaya¹

¹Bombay Hospital & Research Centre, Mumbai, Maharashtra, India

Introduction: Xpert MTB/RIF assay is regarded as a great achievement of modern medicine for the rapid diagnosis of multidrug-resistant tuberculosis (MDR-TB). TB presents as a serious public health problem, globally with an estimated 10.4 million new TB cases in 2015. It is a leading cause of deaths from infectious diseases (with worldwide estimated 1.8 million deaths in 2015), large numbers of deaths mainly occurring in low- and middle-income countries. In 2015, 480 000 new MDR-TB cases were estimated to occur worldwide. Worldwide, large percentage of cases of MDR-TB remain undiagnosed. Furthermore, there is an increasing trend of prevalence of MDR-TB in India. The conventional Mycobacterium tuberculosis (MTB) drug susceptibility testing is a gold standard technique but requires long period of time (8 weeks) to give definitive report. Furthermore, it requires more sophisticated and higher bio-safety level laboratory along with the well-trained staff. Therefore, the diagnosis of MDR-TB by using the conventional drug susceptibility testing is not possible in some resource-limited states of India. To address these issues, Xpert MTB/RIF assay was developed, and it was endorsed by World Health Organization (WHO) in December 2010 as an accurate, feasible, rapid, affordable, and near-point-of-care TB diagnostic test for use in resource-limited settings. Hence, we aim to evaluate the drug sensitivity/resistance patterns in Gene X pert + Rifampicin resistance positive (RIF+) patients. Material and Methods: A total of 465 biopsy-proven patients were analyzed for drug sensitivity patterns and specificity as well as detection of multidrug-resistant TB. Study Design: Prospective, independent observer, single institute from 2013 to 2017. Patient Selection: Inclusions: Spondylodiscitis patients with biopsy proven + Gene Xpert RIF+ were included. DST (drug susceptibility testing) patterns available with Gene Xpert showing RIF+. Exclusions: Gene Xpert showing RIF−. DST patterns not available. Data Collection: Drug sensitivity + resistance patterns, mono resistance, MDR, XDR, and TDR Results: Incidence of Gene Xpert positive 80% and Gene Xpert negative 20%. Rifampicin resistance positive 23% and Rifampicin resistance negative 77%. RIF resistance is not a marker of MDR in 23% cases. MDR incidence 13.33% in 0.06% and XDR incidence 1.59%. In first-line drugs, maximum resistance was found toward isoniazid (92%), followed by rifampicin (81%), streptomycin (69%), and pyrazinamide (46%). In second-line drugs, resistance was shown toward ethionamide (35%), followed by ofloxacin (32%), kanamycin (4.5%), amikacin (4.5%), capreomycin (0.9%), and clofazimine (0%). Overall, 5.26% of Gene Xpert positive + RIF + sensitive to all drugs on DST. A total of 2.01% MDR were missed by Gene Xpert (RIF resistance negative but culture MDR +). Conclusion: Gene Xpert RIF +does not mean MDR always, do not empirically treat them without DST. Gene Xpert is a very sensitive test, but false positives are common. Gene Xpert can miss MDR TB patients, hence culture is the only way to recognize them. MDR is a rising bug, do not neglect it; catch it early so we all can be safe in future.

A206: Reducing Surgical Site Infection in Pediatric Scoliosis Surgery: A Multidisciplinary Improvement Program and Prospective 4-Year Audit

Geoffrey Tipper¹, Jonathan Lucas¹, Athar Siddiqui², and Lillian Chiwera²

¹Guy’s and St Thomas’ Hospital Trust, London, UK

²Guy’s Hospital, Great Maze Pond, UK

Introduction: Surgical site infection (SSI) in pediatric patients following spinal corrective surgery is associated with significant morbidity, lengthened hospitalization, and expense. Reported incidences of infection are typically 6% for idiopathic cases and up to 8 times this for neuromuscular diagnoses. In 2013, a cluster of SSIs in our institution resulted in an incidence of 8.6%, a significant increase on baseline. This prompted a multidisciplinary revision of practice and a prospective audit to monitor the results. Material and Methods: A prospective audit was initiated of all patients aged ≤18 years undergoing corrective spinal deformity surgery. A protocol for minimization of SSI was developed, including (1) preoperative skin decontamination with Octinesan, (2) operative site preparation with double-strength chlorhexidine, (3) betadine-soaked swabs applied to wound edges, (4) blood loss minimization through administration of tranexamic acid and Surgiflo, (5) pulsed lavage with 3 L 0.9% saline postinsertion of metalwork, (6) defined protocol of glove changes, (7) gentamicin to cover catheterization, (8) Antibiotic prophylaxis with cefuroxime preoperatively and 48 hours postoperatively, and (9) normothermia. The center employed an infection control nurse practitioner with a specific remit to spinal surgery, overseeing prophylactic measures and collating data as part of a prospective and ongoing audit. Results: Between 01/01/2014 and 01/01/2018 414 operations were performed on 355 patients: male:female = 1:2.2, mean age 13.7 years (range 2.5-17.9 years), 233 idiopathic (mean operated vertebral levels 12), 122 neuromuscular (mean operated vertebral levels 14), 66 growing rod procedures, 41 multistage procedures. The mean number of operated vertebral levels both overall and in infected cases was 13 (range 2-17). There were 9 surgical site infections in eight patients (50% neuromuscular), resulting in an overall SSI per operation of 2.2%. Conclusion: A multidisciplinary approach with standardized measures significantly reduces surgical site infection in pediatric scoliosis surgery.

A207: Revision Spine Surgery in Patients Without Clinical Signs of Infection: How Often Are There Occult Infections in Removed Hardware?

Isador Lieberman¹ and Xiaobang Hu¹

¹Texas Back Institute, Texas Health Presbyterian Hospital, Plano, TX, USA

Introduction: Hardware removal is commonly required during revision spine surgeries. However, the presence and significance of occult infections in removed hardware has not yet been established in those patients who have no clinical signs of infection. The purpose of this study is to examine the incidence of occult infection in a series of revision spine surgeries and to study its correlation with preoperative inflammatory markers. Material and Methods: We retrospectively reviewed all patients who underwent revision spine surgery and hardware removal between 2010 and 2016. Patients who had preoperative clinical signs of infection were excluded. The hardware and surrounding tissue culture results were obtained. The patients’ diagnosis and preoperative inflammatory marker (erythrocyte sedimentation rate [ESR], C-reactve protein [CRP], procalcitonin) levels were recorded. Results: A total of 162 consecutive patients were included in this study. The patients’ mean age was 61 years (range 14-88 years). One hundred and three patients (63.6%) were female. Seventy-two patients (44.4%) had loose hardware and 88 patients (54.3%) had pseudarthrosis. Postoperatively, the hardware and/or surrounding tissue culture was positive in 15 patients (9.3%). The most common identified organisms were Propionibacterium acnes (7/15, 46.7%) and Staphylococcus (6/15, 40.0%). The other identified organisms were Pseudomonas aeruginosa (1/15, 6.7%) and Serratia marcescens (1/15, 6.7%). Only 4 patients with positive cultures had elevated preoperative ESR and CRP levels. Only 2 patients with positive cultures had elevated preoperative procalcitonin levels. There is no correlation between the patients’ preoperative ESR, CRP, procalcitonin levels and positive culture results (P > .05). Conclusion: Our study shows that occult infections are present in 9.3% of patients who underwent revision spine surgery and hardware removal although they did not have clinical signs of infection. Those commonly used preoperative inflammatory markers such as ESR, CRP, and procalcitonin may not be sensitive enough to detect occult infections in these patients.

A208: Canadian National Survey on Surgical Site Infection Prophylactic Strategies in Spine Surgery

Reina Yao¹, Mark Xu², Michael Goytan², and John Street¹

¹University of British Columbia, Vancouver, British Columbia, Canada

²University of Manitoba, Winnipeg, Manitoba, Canada

Introduction: Numerous prophylactic strategies have been proposed to mitigate risk of surgical site infection (SSI). However, no national guidelines exist and practices of Canadian spine surgeons remain unknown. We conducted a Canadian Spine Society (CSS) survey to capture current practices with respect to SSI prophylaxis. Material and Methods: We emailed the CSS membership a 37-item questionnaire developed after a systematic review of the literature. Results: A total of 52 spine surgeons completed the survey, of whom 94% were fellowship trained. With the exception of preoperative antibiotics in instrumented cases, no SSI prevention strategies were universally used by respondents. For all prophylactic strategies, there was greater reported use for instrumented cases as compared with uninstrumented. Strategies used by most respondents were methicillin-resistant Staphylococcus aureus (MRSA) screening (64% in instrumented and 59% in uninstrumented cases), optimization of diabetic blood glucose (69% and 60%), hair clipping from the operative site (86% and 83%), iodine-impregnated drape application in instrumented cases (60%), routine Cell Saver use, saline wound irrigation (97% and 93%), intrawound vancomycin in instrumented cases (66%), and administration of antibiotics before and after surgery. All respondents gave preoperative antibiotics in instrumented cases, with 95% in uninstrumented cases, and almost all used a first-generation cephalosporin. Ninety percent re-dosed intraoperatively every 4 hours. In cases of penicillin allergy, there was equal use of vancomycin and clindamycin. Seventy-eight percent gave postoperative antibiotics in instrumented cases compared with 53% in uninstrumented cases, with 87% of respondents giving 24 hours’ duration or less. Sixty-seven percent of respondents closed with monofilament subcuticular absorbable suture, whereas 43% used staples. Factors considered high risk by most respondents were obesity, diabetes, immunodeficiency, history of infection, revision surgery, tumor cases, instrumentation, and prolonged surgical duration. Conclusion: There is significant variation in the adoption of SSI prophylactic strategies by CSS members. This survey will help direct future development of national guidelines for spine SSI prophylaxis.

A209: Evidence of Bacterial Biofilm Initiation at the Screw-Bone Interface and the Method of Avoiding Such

Aakash Agarwal¹, Jeffrey Wang², Neel Anand³, Chris Karas⁴, Steve Garfin⁵, Boren Lin¹, and Hossein Elgafy¹

¹University of Toledo, Toledo, OH, USA

²University of Southern California Spine Center, Los Angeles, CA, USA

³Cedars Sinai Medical Center, Los Angeles, CA, USA

⁴OhioHealth, Columbus, OH, USA

⁵University of California San Diego, San Diego, CA, USA

Introduction: Postoperative infections occur at the higher end of 25 to 13%, as cited in the literature, and are underestimated due to various reasons in such publications. Furthermore, the widespread and indiscriminate use of local vancomycin powder has raised concerns of near-term antibiotic resistance. Recent evidences also show that local vancomycin powder was associated with a higher prevalence of gram negative and polymicrobial organisms in patients, who ultimately developed postoperative infection. Despite concerns associated with vancomycin application immediately before closure, it is theoretically impossible to irrigate the screw-bone interface postimplantation. Consequently, any contamination of pedicle screw before implantation is permanent, and has the potential to cause deep-bone infection, or hardware loosening due to encapsulation of biofilm between the bone and the screw. Therefore, continued vigilance and effective preventive measures should be undertaken if available. The objective of the current study is to provide well-quantified and effective results for one such measure. Material and Methods: Two groups of sterile prepackaged pedicle screws, one with an intraoperative guard (group 1: an upcoming practice) and the other without such a guard (group 2: current practice), each consisting of 26 samples distributed over 23 independent time-points (spinal fusion surgeries) and 4 independent hospitals, were loaded onto the insertion device by the scrub tech and left on the sterile table. 20 minutes later, the lead surgeon who had just finished preparing the surgical site, touches the pedicle screw. Then instead of implantation it was transferred to a sterile container using fresh clean gloves for bacterial analysis. Results: Spectrophotometry results detected saturated levels of turbidity within 24 to 48 hours in samples from group 2. The samples from group 1 showed no turbidity for the entire duration of the incubation period (14 days). Every plate, from each sample of group 2 had visible colony-forming units (CFU) growth within 24 hours past streaking. The total CFUs ranged from 10⁵ to 10⁷ per sample. The colonies continued to grow until confluency was reached. No CFU growth occurred in plates extracted from group 1 for the duration of the incubation period (7 days). Please note that the incubation period for spectrophotometry was 14 days (or until saturation) and for streaking it was 7 days (or until confluency). Conclusion: The standard unguarded pedicle screws presented bioburden in the range of 10⁵ to 10⁷ CFUs per screw, whereas the guarded pedicle screws showed no bioburden. The results showed high size of effect despite variability (multiple surgeons, staff, hospitals and practices), and was binary in nature. The study proves that it is necessary to shield sterile implants intraoperative, otherwise they become unsterile. The result of this study provides multicenter clinical evidence that standard pedicle screw handling techniques leads to contamination of pedicle screws and thereby the screw-bone interface. This can be avoided by using a sterile intraoperative guard, which can come preinstalled inside the individually packaged sterile screws, thereby shielding the pedicle screws intraoperatively until implantation. This adaptation of an implant guard is achieved without affecting the standard surgical flow or increasing the responsibilities on the surgical staff.

A210: Predicting and Preventing Post Tubercular Cervical Kyphosis in Pediatric and Adolescent Patients

Aashish Ghodke¹, Ram Chaddha², Agnivesh Tikoo², Harshal Bamb¹, and Sagar Kelkar¹

¹Robert Jones and Agnes Hunt Orthopaedic Hospital, Oswestry, Shropshire, UK

²Apollo Hospitals, Navi Mumbai, Maharashtra, India

Introduction: Children have cartilaginous vertebral body end plates, therefore a tuberculous lesion in children leads to extensive destruction. Whenever tuberculosis heals by nonoperative treatment, kyphosis continues to progress with growth. The anterior growth potential of the vertebral body is either destroyed because of disease itself or because of surgical excision of disease focus or may be due to the alteration in the growth resulting from the effect of biomechanical forces on the growth plate of both the fusion mass and the vertebral segments within the kyphotic region. Unabated posterior growth may contribute to increase the kyphotic deformity. There is not enough evidence in literature to predict and prevent post-tubercular cervical kyphosis in pediatric and adolescent patients with or without neurological deficit. If so, then there are no definite criteria to define and differentiate this group of children. In our study, we try to establish a criterion that tries to differentiate these children and the need for a kyphotic deformity prevention surgery in them. Materials and Methods: A total of 34 pediatric and adolescent patients with tuberculous spondylodiscitis of the cervical spine underwent anterior cervical decompression, stabilization, and fusion. Of the 34, 20 were males and 14 were females, and the average age was 16.5 years. All the patients were given 6 weeks of AKT before the surgery and were neurologically intact, had functional neurological deficit (motor power >grade 3), or improved to grade 3 power on AKT. We divided patients based on MRI and X-ray findings into 3 groups. Involvement of 3 contiguous vertebral bodies with 2 intervening discs on X-rays or MRI—group A; involvement of 3 contiguous vertebral bodies with 2 discs on MRI with normal looking X rays—group B; patients with involvement of a single vertebral body with involved posterior elements with a contiguous paradiscal lesion—group C. Exclusion criteria consisted of patients who presented with flaccid quadriplegia, patients who presented with spastic paraplegia with involvement of sphincters, patients with functional power, those who had 2 vertebral bodies and intervening disc involved on X-rays and MRI, those who had a single vertebral body involved on X-rays and MRI, and those who had a paradiscal lesion on X-rays and MRI. Results: Our observations from the study state that paradiscal lesions heal without significant deformity, single vertebrae with a contiguous paradiscal involvement heals without significant deformity, vertebral body involvement with intact posterior elements heals with insignificant kyphosis involvement of all 3 columns heals with significant kyphosis, involvement of 2 level heals with significant kyphosis and involvement of >2 levels leads to an inherently unstable spine. Conclusion: Kyphosis of the cervical spine can be a debilitating condition that leads to a significant neurological dysfunction. Once kyphosis sets in, the axial loads cause further kyphosis, thus initiating a vicious cycle leading to progression of kyphosis. Surgical correction of cervical kyphosis is associated with a high complication rate. Therefore, establishing a criterion that predicts the potential for development of a significant kyphosis, will help us achieve a better functional outcome.

A211: The Impact of Surgical-Site Infection on Functional Recovery, Surgical Outcomes, and Quality of Life After Adult Posterior Thoracolumbar Spinal Surgery

Jennifer Urquhart¹, Parham Rasoulinejad¹, Lori Nutt¹, Alyssa Fleming¹, Linda Kuska¹, Kevin Gurr¹, Fawaz Siddiqi¹, and Christopher Bailey¹

¹Schulich School of Medicine and Dentistry, The University of Western Ontario, Lawson Health Research Institute, and London Health Sciences Center, London, Ontario, Canada

Introduction: The purpose of the present study was to determine if surgical-site infection (SSI) has an effect on functional recovery, surgical outcomes, and quality of life up to 2 years following posterior thoracolumbar spinal surgery. This study is an ambispective cohort study and was performed at a single tertiary spine center. Material and Methods: Retrospective analysis was performed on patients enrolled in a previously completed prospective randomized controlled trial (RCT) that examined postoperative antibiotic use in association with closed suction drain removal. The trial included 552 patients who were randomized to either 24 hours postoperative antibiotics or 72 hours postoperative antibiotics (to ensure 24 hours of antibiotics post-drain removal, as drains were discontinued by 48 hours postoperatively). In the present study, a comparison was made between patients who had an SSI within 1 year of surgery, as defined by the CDC criteria (n = 80), and those who did not have an SSI (n = 472). Patient-rated functional recovery and quality of life measures included the Oswestry Disability Index (ODI), the numeric rating scale for pain, SF-12 summary scores (mental component summary [MCS] and physical component summary [PCS]), quality-adjusted life-years (QALYs) and satisfaction with treatment. Surgical outcome measures included adverse events, readmissions and additional surgery. Outcomes were assessed prior to surgery and at 1.5, 3, 6, 12, and 24 months. The time-weighted average treatment effect was determined using a mixed effects model of longitudinal regression for repeated measures. Comparison was made at the 2-year time-point accounting for significant baseline covariates. Results: The incidence of infection was 14.5% (80/552), and the median time to infection was 15 days. A total of 31 were deep and 49 were superficial infections. Patients that had an infection had a higher body mass index (P < .0001), were more likely to have had preoperative prophylactic vancomycin (P = .048) and be undergoing a revision procedure (P < .001). They also had worse preoperative mental functioning (MCS 43.9 ± 11.0 vs 40.5 ± 12.5, P = .024), more emergency room visits (2.5% vs 11.3%), readmissions (0.0% vs 46.3%), and additional surgery (5.5% vs 38.8%) than patients that did not have an SSI (P < .05 all comparisons). No differences were found between groups with time-weighted average treatment effects for leg pain −0.004 (95% CI, −0.5 to 0.4), back pain −0.2 (95% CI, −0.6 to 0.2), or MCS 1.9 (95% CI, −0.1 to 3.8). However, patients who had an infection had worse ODI −4.0 (95% CI, −7.5 to −0.5, P = .024) and PCS 1.8 (95% CI, 0.1 to 3.5, P = .038) scores. Comparison between the groups at 2 years showed no difference in any health-related outcome measure, QALYs (1.39 ± 0.391 vs 1.37 ± 0.365) or satisfaction with treatment (77.8% vs 68.8%; no infection vs infection, respectively). Conclusion: SSI nearly doubled the readmission and additional surgery rates. Patients with SSI had poorer overall physical function particularly during first 6 months after surgery, representing the delay to recovery associated with the infection. But importantly, the negative impact resolved during the second postoperative year. In this patient cohort, longer-term pain, QALYs, and patient satisfaction were not compromised by an SSI at 2 years after surgery.

OP32: Epidemiology

A212: Patterns of Neurological and Functional Recovery After Complete Traumatic Spinal Cord Injury: Evidence from Group Based Trajectory Modelling in the Nascis 3/Sygen/Nactn/Stascis Studies

Blessing Jaja¹, Jetan Badhiwala², Ralph Schar², Shekar Kurpad³, Robert Grossman⁴, Bizhan Aarabi⁵, Maxwell Boakye⁶, James Harrop⁷, Fred Geisler⁸, Jefferson Wilson², and Michael Fehlings,² on behalf of the North American Clinical Trials Network Collaborators

¹University Health Network, Toronto, Ontario, Canada

²University of Toronto, Toronto, Ontario, Canada

³ Medical School of Wisconsin, Wauwatosa, WI, USA

⁴Weill Cornell Medical College, Houston, TX, USA

⁵University of Maryland School of Medicine, Baltimore, MD, USA

⁶University of Louisville Hospital, Louisville, KY, USA

⁷Thomas Jefferson University, Philadelphia, PA, USA

⁸Independent Neurosurgeon, Chicago, IL, USA

Introduction: Complete spinal cord injury (SCI) is characterized by total disruption of motor and sensory functions at the lowest sacral roots due to a complete conduction block at the segmental level of the injury. Classified as grade A injury by the ISNCSCI system, this injury type is the primary pattern that is seen in majority of SCI, with estimated incidence of 45%. Though it is associated with worse outcomes, complete SCI might reflect a heterogenous entity, including subpopulations with distinct clinicopathologic characteristics and prognosis. Our aim in this study, therefore, was to investigate whether there are subpopulations of complete SCI with homogenous longitudinal trajectories of recovery in neurological function and walking ability. And if so, to identify patient, clinical and treatment characteristics that are predictive of trajectory group membership. Material and Methods: A total of 882 subjects with complete injury (ASIA A) at baseline examination were pooled from 4 prospective longitudinal multicenter cohort studies including: the NASCIS, Sygen, NACTN and STASCIS studies. Group-based trajectory modelling (GBTM) was used to identify groups of subjects following similar distinct courses of motor recovery (assessed using the Lower Extremity Motor Scores; LEMS) and of walking ability (assessed using the SCIM subscore for indoor mobility). Data was obtained at baseline, and at 3-, 6-, and 12-month follow-up examination. Within the GBTM framework, multinomial logit regression was applied to identify predictors of trajectory group membership, with adjustment for fixed effect of study. Results: Three distinct trajectories of motor recovery were identified, including (1) no motor recovery group, with predicted LEMS averaging zero during the 12-month period (72.9%, n = 659); (2) minimal motor recovery group, with predicted mean LEMS of 5/50 at baseline; mean LEMS increased over time to 13/50 at 12 months (18.1%, n = 151); (3) good motor recovery group, with predicted mean LEMS of 8/50 at baseline that improved to 34/50 at 3 months and plateaued at 41/50 from the 6- to 12-month follow-up period (9.1%, n = 77). Injuries to the thoracic and lumbosacral spinal levels were associated with belonging to the minimal recovery and good recovery trajectory groups, in comparison with the no recovery group. In addition, injury from falls was associated with belonging to the good recovery group relative to injury from motor vehicle collisions (relative risk ratio, 3.38; 95% CI, 1.53-7.46). We found no association between age, sex, weight, systolic blood pressure, timing of surgery, or steroid use and trajectory group membership. Comparable trajectory patterns were noted in walking ability. Conclusion: This study has demonstrated that complete SCI consists of latent homogenous subpopulations with potential divergent trajectories of recovery in motor function and walking ability. This finding could have implications for phenotypic characterization and improving knowledge about the natural evolution of complete SCI, which might be useful in designing interventional studies, targeting therapies and predicting differential response to therapy.

A213: Congenital Lumbar Stenosis Prevalence: A New Population-Based Definition

Mauricio Campos¹, Ismael Cañete¹, Pablo Besa¹, María Jesús Lira¹, and Macarena Sepúlveda¹

¹Pontificia Universidad Católica de Chile, Santiago, Chile

Introduction: Congenital lumbar spinal stenosis (CLSS) is poorly defined and therefore its prevalence remains unclear. Its correct and timely diagnosis can prevent complications in younger population. Previous articles have studied prevalences of CLSS in symptomatic or middle-aged population. Classic definitions of CLSS use absolute values to determine stenosis, such as Verbiest et al’s 12-mm cutoff used later by Kalichmann et al. We set out to determine normal values for lumbar spine diameters, and therefore detect outliers that could be considered as CLSS. We then compared the prevalence obtained by this new population-based threshold outliers against absolute diameter values as proposed previously. Material and Methods: Cross-sectional study. We evaluated consecutive abdomen and pelvis CT-scans of patients 40 years old or younger. All images were requested for non-spine-related diagnosis, between 2014 and 2016. We excluded patients with previous spinal surgery, vertebral fractures, vertebral infections, vertebral malformations, spinal tumors, and patients with listhesis. All measurements were made on the axial view using bone window grayscale, by one trained senior orthopedic resident. We preferred this view because it has been proven to have a better agreement than measuring in the sagittal view. The anteroposterior diameter of the spinal canal was measured at the mid-vertebral body level for each lumbar level (from L1 to L5). This method has a reported fair to excellent interobserver reproducibility. Given prior prevalence studies, minimal sample size was calculated at 204 patients. We measured the anteroposterior diameter (APD) of the lumbar spinal canal (LSC). We determined that APD had a normal distribution using the Kolmogorov-Smirnov test. The 2.5% of patients with the smallest LSC was determined to have CLSS, setting the new population-based threshold at 2 standard deviations below mean APD. We then compared the prevalence of CLSS using our new threshold against the Verbiest criteria. A McNemar test was performed considering a 5% significance. Results: A total of 204 patients were studied. New thresholds for each lumbar level were calculated. The CLSS threshold from L1 to L5 were 14.8, 14.1, 13.1, 12.9, and 13.5 mm, respectively. There was no significant association between APD and sex or age. According to our proposed criteria 4 patients presented with one level of CLSS and 3 patients with L1-L3 multilevel CLSS, with an overall 3.44% prevalence of CLSS. Estimated prevalence using the Verbiest criteria was 1.47%, significantly lower than our population-based prevalence (P = .046). Conclusion: There is a difference in CLSS prevalence in our sample using a population-based threshold, compared with prevalence determined by classic absolute numbers. Although this would seem to overestimate classic prevalence, we find it to be methodologically more sound than assuming a fixed LCS size. This approach has the advantage that it could be applied to different populations with diverse ethnicity and therefore, could have more external validity. Future studies should correlate this definition with symptoms.

A214: Radiographic Implications of the Surgical Waiting List for the Treatment of Spinal Deformity

Leonardo Yukio Jorge Asano¹, Marina Rosa Filézio¹, Vinicius Alves De Andrade¹, Andre Evaristo Marcondes Cesar¹, and Luciano Miller Reis Rodrigues¹

¹Faculdade de Medicina do ABC, Santo Andre, Brazil

Introduction: Scoliosis is defined by the Scoliosis Research Society (SRS) as a deviation of the spine in the coronal plane greater than 10°. It is a condition that mainly affects children and adolescents. The risk of progression of the curvature in idiopathic scoliosis depends on several factors, such as long waiting lists. During the waiting period, patient conditions may continue to worsen, leading to aggravated symptoms, and, therefore, leading to a negative impact on the mental health and quality of life of the patients. Material and Methods: Radiographs of 87 patients with scoliosis who had been on the waiting list for surgery for more than 6 months were selected. Two surgeons answered questionnaires analyzing the radiographs when entering the waiting list and the current images of each patient. Results: Data from 87 patients were analyzed. The mean waiting time for surgery was 21.7 months (ranging from 7 to 32 months). The average progression of the Cobb angle in the curvature was 21.1°. Delayed surgery implied changes in surgical planning, such as greater need of instrumentation, osteotomies, and double approach. Conclusion: Long waiting lists have a significant negative impact on surgical morbidity of patients with scoliosis, since they increase the complexity of the surgery. Level of evidence: IV. Type of study: Descriptive study.

A215: The Influence of Preoperative Opioid Use on Hospital Length of Stay and Patient Reported Outcomes Following Elective Spine Surgery

Jean-Christophe Murray¹, Mayilee Canizares², Jennifer D. Power², Anthony V. Perruccio², and Raja Rampersaud², on behalf of CSRON Investigators

¹University Health Network, Toronto, Ontario, Canada

²Krembil Research Institute, University Health Network, Toronto, Ontario, Canada

Introduction: Preoperative opioid use in spine surgery is very common. Limited studies have demonstrated an association of preoperative opioid use with negative surgical outcomes. Our primary objectives were to assess the influence of preoperative opioid use on (1) hospital length of stay (LOS) and (2) patient-reported pain and disability 1 year after elective spine surgery. Material and Methods: Retrospective review of Canadian Spine Outcomes and Research Network (CSORN) Registry data across 8 provinces. Patients (n = 1931) with a primary degenerative diagnosis, hospital LOS, and baseline and 1-year postsurgical patient reported outcomes were included. For objective 2, we examined postoperative patient-reported pain from a visual analog scale (0 = no pain to 10 = unbearable pain) and neck or back disability scores for patients with cervical and thoracolumbar diagnoses, respectively. The main covariate of interest was preoperative opioid use (daily, intermittent, and no use). Other covariates included sociodemographic (eg, age, sex, education, work status), lifestyle (eg, smoking, BMI), and clinical variables (eg, depressive mood, surgery type) as well as baseline pain and disability scores. Multiple linear regression models were used to examine the associations between preoperative opioid use and LOS as well as 1-year postsurgical pain and disability. Results: Daily, intermittent and no preoperative opioid use was reported by 18.3%, 32.4%, and 50.7% of patients, respectively. Independent factors significantly associated with preoperative daily opioid use were thoracolumbar diagnosis (odds ratio [OR] = 2.73), depressed mood (OR = 2.22), 3+ comorbidities (OR = 1.93), disc herniation diagnosis (vs stenosis) (OR = 1.79), positive legal claim (OR = 1.56), in labor force but not working (OR = 1.54), anxious mood (OR = 1.48), degenerative disc disease (OR = 1.41), and being a current smoker (OR = 1.37). Older patients, those exercising regularly, and those currently working were significantly less likely to report daily opioid use. Mean LOS was 4.4, 5.1, and 5.2 days (P = .19) for no, intermittent, and daily preoperative opioid use, respectively. In multivariable linear regression models, preoperative daily opioid use was independently associated with increased LOS (β = 1.40; P < .0001), whereas intermittent use was not (β = 1.16; P < .072). Daily opioid use was independently associated with worse self-reported pain (β = 0.48; P = .003) and disability (β = 4.83; P < .0001) at 1-year postsurgery as compared with patients not using opioids. Intermittent use was associated with worse disability (β = 2.44; P < .042) but not pain. Legal claims (β = 0.55; P = .001), smoking (β = 0.47; P = .007), anxiety (β = 0.85; P = .004), and higher baseline pain (β = 0.24; P < .0001) were associated with worse pain scores at 1 year. Worse disability scores were associated with older age (β = 0.90; P = .029), legal claims (β = 5.93; P < .0001), smoking (β = 2.83; P = .013), 3+ comorbidities (β = 4.85; P = .003), anxiety (β = 2.28; P < .026), and higher baseline Oswestry Disability Index (ODI)/Neck Disability Index (NDI) (β = 0.44; P < .0001). Conclusion: Preoperative opioid use is highly prevalent in patients undergoing elective spine surgery across Canada, with 50% reporting intermittent or daily use. Preoperative use of opioids in this setting is independently associated with greater pain and disability at 1-year postsurgery, with daily user having a greater negative impact on outcome. Furthermore, daily preoperative use is also associated with increased hospital LOS. Future work will focus on examining efficacy of opioids and predictors of changes in opioid use post-surgery and its associated impact on outcomes.

A216: Epidemiological Trends of Degenerative Cervical Myelopathy in the United States, 2002-2014: A National Population-Based Analysis

Jetan Badhiwala¹, Farshad Nassiri¹, Christopher Witiw¹, Saleh Almenawer², James Harrop³, Christopher Shaffrey⁴, Michael Fehlings¹, and Jefferson Wilson¹

¹University of Toronto, Toronto, Ontario, Canada

²McMaster University, Hamilton, Ontario, Canada

³Thomas Jefferson University, Philadelphia, PA, USA

⁴Duke University, Durham, NC, USA

Introduction: Degenerative cervical myelopathy (DCM) is the leading cause of spinal cord injury (SCI) and a major source of disability globally; however, data on the national trends of DCM incidence is limited. To that end, we aimed to evaluate contemporary trends in the epidemiology and health care resource utilization of DCM in the United States using population-based survey data. Furthermore, we sought to compare the incidence and costs of surgically treated DCM to acute traumatic SCI and common surgical diagnoses affecting the central nervous system. Materials and Methods: Patients who underwent surgery for DCM were identified from the 2002-2014 AHRQ HCUP National Inpatient Sample (NIS) datasets using ICD-9-CM codes based on a diagnosis code for cervical spondylosis and/or OPLL (ossification of the posterior longitudinal ligament) plus cervical myelopathy, plus a procedure code for anterior or posterior decompression plus/minus fusion, or disc replacement. Data pertaining to patient-, treatment-, admission-, and hospital-related factors and outcomes was collected. Patient-specific discharge weights were used to obtain national estimates. We calculated standard errors and 95% confidence intervals (CIs) using strata and cluster variables to account for hospital-level clustering of patients and the sampling design. We calculated overall and age- and sex-stratified incidences of surgically-treated DCM per 100 000 persons for each year using population estimates from the U.S. Census Bureau. These were compared with incidences of acute traumatic SCI and three very common neurosurgical procedures: (1) evacuation of subdural hematoma (SDH) (including acute, subacute, and chronic); (2) craniotomy for tumor (including primary and metastatic); and (3) clipping or endovascular coiling of cerebral aneurysm. Hospitalization charges and costs were indexed to year 2008 to adjust for inflation using CPI data from the U.S. Bureau of Labor Statistics. Trends were analyzed by join point regression analysis for incidences to determine the average annual percent change (AAPC), the Cochrane-Armitage trend test for categorical variables, and linear regression for continuous variables. Results: A total of 618 323 patients underwent surgery for DCM in the United States between 2002 and 2014. The overall incidence of surgical DCM rose from 10.3 per 100 000 persons in 2002 to 21.1 per 100 000 persons in 2014, representing an AAPC of 5.9% (95% CI 5.5% to 6.3%). The highest incidence (61.1/100 000) and growth rate (AAPC 7.4%) were seen in the 65- to 84-year-old age group. Age-adjusted incidence and annual growth rates for DCM exceeded those of acute traumatic SCI (5.0/100 000; −0.5%), evacuation of SDH (8.2/100 000; 0.4%), craniotomy for brain tumor (14.2/100 000; 1.1%), and clipping or coiling of cerebral aneurysm (7.2/100 000; 2.4%). Cumulative hospital charges and costs for DCM rose from $0.99 billion and $0.29 billion in 2002 to $5.13 billion and $1.32 billion in 2014, respectively. The proportion of cases for which Medicare was the primary payer increased from 30.5% in 2002 to 43.6% in 2014. Conclusions: DCM is the leading indication for operation on the CNS. The incidence of, and healthcare dollars spent on, surgery for DCM has steadily risen over the past decade likely owing both to an aging population and increasing recognition of the role of surgical management in these patients.

A217: Ninety-Day Readmission Following Multilevel Anterior Cervical Fusion

Andrew Chung¹, Nina Lara¹, Nicolas Rubel¹, Varun Arvind², Michael Chang³, and Samuel Cho²

¹Mayo Clinic Arizona, Phoenix, AZ, USA

²Icahn School of Medicine at Mount Sinai, New York, NY, USA

³Sonoran Spine Center, Phoenix, AZ, USA

Introduction: Hospital readmission is becoming an increasingly important performance metric following surgery as readmissions are costly and have significant clinical consequences. While 30-day readmissions have been studied in the setting of the cervical spine surgery, an adequate characterization of 90-day readmissions has yet to be undertaken. Material and Methods: The Nationwide Readmissions Database (NRD) is a newer national database that tracks approximately 50% of all inpatient hospitalizations. Using unique patient linkage codes, patients can be longitudinally tracked, allowing for the study of longer-term readmission data. In this study, we identified all adult patients (18 years or older) undergoing elective multilevel anterior cervical fusion surgery in 2014. Combined anterior/posterior procedures were excluded to minimize confounding. The sample was then divided into 2 cohorts based on readmission status. Ninety-day readmission rates were calculated. Associations between patient demographic, surgical data, as well as hospital data and 90-day readmissions were established utilizing univariate analyses. Comorbidity burden was calculated using the Elixhauser comorbidity index. Multivariate logistic regression was then performed to determine predictors of readmission. Finally, total hospital costs associated with both the initial hospitalization and readmission (if applicable) were calculated controlling for inflation. All data presented are weighted national estimations. Results: A total of 112 808 patients undergoing multilevel anterior cervical fusion were identified. Mean age of the patients was 55.5 years. There was no difference in age between the 2 cohorts. Average length of stay for the index admission was only slightly greater in the readmitted cohort (0.3-day increase; P = .02). There was no difference in comorbidity burden between the 2 cohorts. The 90-day rate of readmission was 0.8%. The strongest predictors of readmission included initial postoperative transfer to a skilled nursing facility (OR 2.1; P < .001), diabetes with end-organ complications (OR 1.7; P = .04), revision surgery (OR 1.6; P = .03), anemia (OR 1.5; P = .04), and an index diagnosis of myelopathy (OR 1.5; P < .001). Readmissions were costly, with patients who were readmitted incurring costs of $50 279, compared with $16 989 for patients who were not readmitted (P < .001). The most common reason for readmission was hardware failure (35.1%). The most common procedures performed during the readmission event were posterior cervical fusion (51%), anterior cervical discectomy and fusion (ACDF) (24%), revision ACDF (20%). Conclusion: While the incidence of readmission following multilevel anterior cervical fusion appears low, these readmissions are associated with significant patient morbidity and ultimately, hospital costs. As such, familiarity with the factors that increase the likelihood of 90-day readmission will allow for improved outcomes and health care utilization.

A218: Using Machine Learning to Search Electronic Medical Records for Surgical Site Infections Following Orthopedic Surgery

Varun Arvind¹, Brian Cho¹, Aly Valliani¹, Andrew Warburton¹, Daniel Ranti¹, Deepak Kaji¹, Ivan Ye¹, Ray Tang¹, Zoe Cheung², Jun Kim², and Samuel Cho²

¹Icahn School of Medicine at Mount Sinai, New York, NY, USA

²Mount Sinai Hospital, New York, NY, USA

Introduction: Quality improvement studies have provided surgeons with systematic insight on how to make changes to their practice to improve patient outcomes and avoid complications. However, currently most automated methods to identify patient complications for quality improvement studies rely on prelabeled administrative codes sampling data from fixed electronic medical databases. Medical notes contain a rich supply of medical data, yet the format of unstructured text has precluded these data from use in large retrospective analyses. Therefore, the utilization of free-text from electronic medical records using natural-language processing in combination with machine learning can provide an additional tool for patient complication surveillance and identification. The purpose of this study was to develop a machine learning algorithm that can identify postsurgical surgical site infection (SSI) based on unstructured patient notes. Material and Methods: This was a retrospective analysis of medical notes from the Medical Information Mart for Intensive Care (MIMIC III), a large, single-center database comprised of intensive care unit admissions data from 2001 to 2012. A total of 947 patients from the orthopedic service were identified. Patients who experienced SSI were identified using ICD-9 code (99 859). Overall, 15 004 notes were analyzed in total from the 947 patients. Of the 15 004 notes, 1961 notes were from patients positive for SSI and 13 043 notes were negative for SSI defined by ICD-9 code. Area under the receiver operating characteristic curve (ROC), accuracy, sensitivity, and specificity of the model were evaluated to assess performance of the model. Natural language processing was used to analyze unstructured narrative discharge notes and train a support vector machine algorithm to train a classifier to automatically identify patients with SSI based solely on unstructured discharge notes. The data was randomized with 70% used for training and 30% used for testing. Results: The SVM (support vector machine) algorithm identified SSI notes with a sensitivity and specificity of 73% and 85%, respectively. The overall accuracy was 83.4% with area under receiver operating characteristic curve of 0.79. Conclusion: Combining natural language processing with machine learning is a valid approach for automatic identification of SSI using unstructured text from discharge notes. This method provides a robust survey of surgical encounters and may be used for real-time surveillance of postoperative complication risk and for automated identification of patient complications such as SSI for quality improvement studies. Medical notes contain a vast amount of information that is currently difficult to leverage for use in large retrospective quality improvement studies. In this study, we present the use of machine learning and natural language processing to identify associated complications based on analysis of unstructured free-text from medical notes.

A219: Trajectories of Recovery in Motor Function After Incomplete Traumatic Spinal Cord Injury

Blessing N. R. Jaja¹, Fan Jiang¹, Jetan Badhiwala², Ralph Schar², Shekar Kurpad⁷, Robert Grossman⁵, James Harrop⁶, Fred Geisler⁸, Bizhan Aarabi³, Maxwell Boakye⁴, Michael G. Fehlings², and Jefferson Wilson², on behalf of the North American Clinical Trials Network Collaborators

¹University Health Network, Toronto, Ontario, Canada

²University of Toronto, Toronto, Ontario, Canada

³University of Maryland School of Medicine, Baltimore, MD, USA

⁴University of Louisville Hospital, Louisville, KY, USA

⁵Weill Cornell Medical College, Houston, TX, USA

⁶Thomas Jefferson University, Philadelphia, PA, USA

⁷Medical School of Wisconsin, Wauwatosa, WI, USA

⁸Independent Neurosurgeon, Chicago, IL, USA

Introduction: Incomplete spinal cord injury (SCI) reflects a heterogenous group of SCI with variable preservation of sensory and motor function beyond the level of injury. Spontaneous recovery frequently occurs among this group of SCI which might follow hitherto unknown developmental pathways over time that when understood could contribute to unravelling the distinct clinicopathologic subpopulations that have been hypothesized to exist among incomplete SCI. In this study, our objective was to investigate whether latent trajectories exist that might potentially approximate the complex patterns of recovery in motor function among subjects with incomplete SCI; and if so, to determine what clinical and treatment factors are related to trajectory group membership. Material and Methods: We identified 795 subjects with incomplete SCI (ASIA grade B/D injuries) from 3 prospective, multicenter longitudinal studies: the NASCIS, the Sygen, and the NACTN studies. We analyzed the Lower Extremity Motor Scores (LEMS) at 4 time points (baseline, 3-, 6-, and 12-month follow-up) using group-based trajectory modeling (GBTM) to identify distinct subgroups of subjects whose motor recovery follow a similar pattern over the time course. Identified trajectories were related to demographic, clinical and treatment factors using multinomial logistic regression. Using ordinary least square regression, we studied the relationship of the trajectories to 12 months functional outcome assessed according to the Functional Independence Measure (FIM) motor subscore. Adjusted analysis accounted for the effect of study. Results: We identified 3 trajectory patterns: (1) marginal motor recovery—trajectory members (23.7% of studied cohort) had uniformly low predicted mean LEMS (average 3 points out of 50) during the 12 months; (2) progressive motor recovery—this group (41.9%) experienced poor motor function at baseline that improved rapidly during the first 3 months, and less markedly thereafter (mean LEMS = 9 at baseline; 35 at 3 months, and 42 at 12 months; (3) good motor function—this trajectory describes subjects (31.4%) who had good motor function during the 12-month follow-up, exemplified by high mean LEMS = 41 at baseline and 49 at 6 months. Baseline AIS grade was highly associated with trajectory group membership with better AIS grades associated with groups having better prognosis (P < .001). Trajectory group membership was strongly associated with functional outcome. We noted trajectories with better motor recovery also had higher mean score on the FIM motor subscale (marginal recovery, 43.9, 95% CI = 39.7-48.3; progressive recovery, 78.4, 95% CI = 75.6-81.3; good motor function, 88.9, 95% CI = 85.5-91.3). There was no association between trajectory pattern types and age, sex, weight, injury etiology, systolic blood pressure, injury level, time from injury to surgery, and steroid administration. Conclusion: The complex patterns of motor recovery after incomplete SCI could be adequately approximated into 3 longitudinal trajectory pattern types that are strongly predicted by baseline injury severity. This study findings demonstrate the utility of GBTM in assisting with mapping the distinct phenotypes of incomplete SCI with similar natural history with respect to prognosis, which has the potential to further our ability at the classification of this highly heterogenous subset of SCI.

A220: Patients’ Expectations for Spine Surgery: What Are They, Are We Meeting Them, and Do They Correlate to One-Year Postoperative Satisfaction

Mayilee Canizares¹, Andrew Glennie², Kenneth Thomas³, Sean Christie², Eugene Wai⁴, Charles Fisher⁵, Anthony V. Perruccio¹, and Raja Rampersaud¹, on behalf of CSRON Investigators

¹Krembil Research Institute, University Health Network, Toronto, Ontario, Canada

²QEII Health Sciences Centre, Halifax, Nova Scotia, Canada

³Foothills Medical Centre, Calgary, Alberta, Canada

⁴Ottawa Hospital, Ottawa, Ontario, Canada

⁵Blusson Spinal Cord Centre, Vancouver, British Columbia, Canada

Introduction: The objectives of this national study were to examine the most important change patients expected from their spine surgery, the extent to which patient expectations of spine surgery were fulfilled and to examine the association between unmet expectations and postoperative satisfaction after accounting for sociodemographic and clinical characteristics. Material and Methods: Data from the Canadian Spine Outcomes and Research Network (CSORN) Registry with a primary degenerative spinal diagnosis and 1-year follow-up data was used for analysis (n = 1985). Patients indicated on a questionnaire their level of expectation as it related to 7 separate dimensions as: no change (0), somewhat better (1), better (2), and much better (3). Patients also indicated which of the 7 areas was most important to them. For each dimension, patients reported their overall satisfaction with surgery and whether their expectations were fulfilled at 1-year follow-up. Sociodemographic and clinical variables including pain and back/neck disability scores were also collected. Logistic regression models were fit to examine (a) the association between expectations being met and baseline sociodemographic and clinical variables, including pre- and postsurgical changes in pain and disability scores and (b) the association between expectations being met with overall satisfaction with surgery (extremely/somewhat satisfied vs neither/somewhat/extremely dissatisfied) after accounting for sociodemographic, clinical, and patient-reported outcome variables. Results: A total of 29.5% ranked “improve leg/arm pain,” 26.3% ranked “back/neck pain,” 21.4% ranked “general physical capacity,” and 15.3% ranked “independence in everyday activities” as the most important expected change. No differences were found in the ranking of expectations by demographic or clinical variables. Overall, 88.8% of patients reported that their most important expected change was met: 95.8% among those expecting improvement in independence in everyday activities, 94.5% for improvements in sporting activities/recreation, 89.0% and 87.3% for improvements in leg/arm pain and back/neck pain respectively, 85.4% for improvements in general physical capacity, and 82.7% of those expecting improvements in mental well-being. Factors associated with expectations being met were spondylolisthesis diagnosis versus stenosis and improvements in pre- to postsurgical pain and disability scores (P < .0001 for all variables). In addition, about 93% of patients who had their most important expectations met were extremely/somewhat satisfied with their surgery versus 52.2% among those with unmet expectations. Similar trends were seen for each of the expectation dimensions examined. In multivariate analysis, meeting the most important expected change was independently associated with being extremely/somewhat satisfied with the surgery (OR = 4.5, 95% CI = 3.2-6.3). In addition, postoperative pain and disability were significantly associated with satisfaction, such that reporting worse postoperative Oswestry Disability Index (ODI)/Neck Disability IndexNDI and/or pain was associated with not being satisfied with surgery. Conclusion: The most important expected change from spine surgery is highly variable across patients. The vast majority of patients reported their most important surgical expectation as being met. While met/unmet expectations were strongly associated with satisfaction/dissatisfaction, this must be considered against a backdrop of “most important expected change” being highly variable across patients. Preoperative inquiry and education regarding realistic individual patient expectations may be paramount in determining the drivers of individual patient satisfaction with spine surgery.

OP33: Deformity-Thoracolumbar (Adult) 3

A221: Costs and Utility of Postdischarge Acute Inpatient Rehabilitation Following Adult Spinal Deformity Surgery

Alekos Theologis¹, Darryl Lau¹, Adelyn Tsu¹, Vedat Deviren¹, and Christopher Ames¹

¹University of California San Francisco, San Francisco, CA, USA

Introduction: As the United States’ health care system transitions to a value-based approach, defining costs for entire episodes of care is fundamental to determining value. To further the discussion on costs of operative care for adults with spinal deformity (ASD), this study aims to evaluate the costs and functional utility of postdischarge acute inpatient rehabilitation after surgery for ASD. Materials and Methods: Retrospective analysis of ASD who underwent operative intervention (≥7 levels fused) at a single tertiary care referral center and who were discharged postoperatively to one acute inpatient rehabilitation facility. Patient demographics, operative details, and costs were obtained for the index operation’s inpatient encounter. Data for acute rehabilitation included direct costs, length of stay, and patient function, as assessed by the Functional Independence Measure (FIM) instrument (scores 18-30 [total assistance], 31-45 [maximal assistance], 46-63 [moderate assistance], 64-80 [minimal assistance], 81-100 [supervision], 101-125 [modified independence], 126 [complete independence]). Student t tests and chi-square tests were used to compared continuous and categorical variables. Pearson’s tests were used to assess correlations. P < .05 set statistical significance. Results: Ninety-patients (95 episodes of care; average age 70.5 ± 10.6 years; male 31, female 59) were included for analysis. After operations, average inpatient length of stay was 8.2 ± 2.6 days (range, 4-16 days) (Figure 1). Operative details included posterior levels fused 13.6 ± 3.6 (range, 7-21), posterior column osteotomies (per patient) 7 ± 3.7 (range 1-15; n = 49 patients), 42 three-column osteotomies (pedicle subtraction osteotomy [PSO] 35, vertebral column resection [VCR] 7), and 11 interbody fusions (anterior lumbar interbody fusion [ALIF] 2, transforaminal lumbar interbody fusion [TLIF] 9). Average direct costs were $90 738 ± $24 166 (range, $36 521 to $141 090) for index inpatient stays and $38 808 ± $14 752 (range, $8679 to $100 829) for postdischarge acute rehabilitation (Figure 2). Patients spent an average of 11.7 ± 4.0 days (range, 2-28 days) in acute rehabilitation postdischarge (Figure 1). Average direct cost per day in rehabilitation was $5162 ± $2344 (range, $964 to $14 151). Significant improvements in function, as assessed by FIM scores, were observed (admit, 66 ± 14 [range, 15-91] vs discharge, 94 ± 14 [range, 54-116]; % improvement: 29 ± 14) (Table 2). No correlations were found between age, inpatient length of stay, inpatient direct costs and rehabilitation length of stay or direct costs. Patients who achieved minimal clinically important difference (MCID) for FIM (22-point increase; n = 70) had significantly more than 3 COs performed, significantly lower admission rehab FIM scores (65 vs 71), significantly higher discharge FIM scores (98 vs 83), and significantly greater rehabilitation costs ($40 616 vs $33 745) than those who did not achieve MCID (n = 25). Conclusions: Postdischarge inpatient acute rehabilitation following operations for adult spinal deformity is associated with an average direct cost of $38 808 per case. While rehabilitation resulted in significant functional improvements, it came at a very significant economic expense amounting to $3.7 million, which accounted for 30% of the costs for 95 episodes of care.

A222: Preoperative Prediction of Cost and Catastrophic Cost (CC) in Adult Spine Deformity (ASD) Surgery: Feasibility Analysis of Predictive Analytics to Establish 90-Day Bundled Payments

Christopher P. Ames¹, Miquel Serra-Burriel², Justin S. Smith³, Jeffrey L. Gum⁴, Michael P. Kelly⁵, Ferran Pellisé⁶, Ahmet Alanay⁷, Emre R. Acaroglu⁸, Francisco Javier Sánchez Pérez-Grueso⁹, Frank Kleinstück¹⁰, Ibrahim Obeid¹¹, Samrat Yeramaneni¹², Richard Hostin¹³, Corinna Zygourakis¹⁴, Virginie Lafage¹⁵, Frank J. Schwab¹⁵, Christopher I. Shaffrey³, Douglas C. Burton¹⁶, Shay Bess¹⁷, European Spine Study Group ESSG¹⁸, International Spine Study Group ISSG¹⁵

¹University of California San Francisco, San Francisco, CA< USA

²Universitat Pompeu Fabra, Barcelona, Spain

³University of Virginia Medical Center, Charlottesville, VA, USA

⁴Norton Leatherman Spine Center, Louisville, KY, USA

⁵Washington University at St Louis, St Louis, MO, USA

⁶Hospital Vall d’Hebron, Barcelona, Spain

⁷Acibadem University, Istanbul, Turkey

⁸Ankara ARTES Spine Center, Ankara, Turkey

⁹Hospital Universitario La Paz, Madrid, Spain

¹⁰Schulthess Klinik, Zurich, Switzerland

¹¹Bordeaux University Hospital, Bordeaux, France

¹²Baylor Scott & White Health Research Institute, Dallas, TX, USA

¹³Baylor Scott and White, Plano, TX, USA

¹⁴Johns Hopkins Hospital, Baltimore, MD, USA

¹⁵Hospital for Special Surgery, New York, NY, USA

¹⁶University of Kansas Medical Center, Kansas City, KS, USA

¹⁷Presbyterian St. Luke’s/Rocky Mountain Hospital for Children, Denver, CO, USA

¹⁸Vall d’Hebron Institute of Research, Barcelona, Spain

Introduction: There has been increasing interest in cost prediction for bundled payment models and risk sharing initiatives. Catastrophic costs (CC) outliers are typically excluded in bundled payment settings making their preoperative identification critical. Bundled payment models and risk sharing initiatives have been proposed as means of controlling adult spine deformity (ASD) surgery cost, but these approaches require accurate cost prediction. Material and Methods: A predictive analytic model to preoperatively determine direct cost and CC outliers can be developed using a cohort of ASD surgeries with actual direct hospital costs. We modeled the direct cost based on prospective multicenter ASD data. We performed regression models (generalized linear regression and random forest) for direct costs and classification models (random forest) for CC for ASD surgery. The goal of the regression models was to explain the determinants of direct costs (patient, surgical, and contextual factors). The goal of the CC models was to predict which patients would have a direct CC (>$100 000). Results: A total of 210 ASD patients (83% women, 45% revisions) from 4 sites in 4 geographic US areas were included. Cost data were actual direct costs incurred to the hospital. Average index cost per patient was $75 772. A total of 14.8% of patients had a cost above the $100 000 threshold. Direct cost could be predicted preoperatively using random forest models with an accuracy of 72.1% (example case in figure). Of total variance explained, 22.6% was site and surgeon fixed-effects. Top predictors of cost in order were surgeon, number of levels fused, interbody fusion and site. Catastrophic cost was predicted with 90.4% accuracy and 87.7% area under the receiver operating characteristic curve (AUC). In our sample alone, reducing the CC occurrence by one-third the associated savings would be $452 181. Across the United States, these savings would extrapolate to at least $80 million/year. Conclusion: This study demonstrates that direct cost in ASD surgery can be reliably predicted in a preoperative setting. CC outliers can be predicted with >90% accuracy preoperatively. The high degree of cost variance explained by factors such as site and surgeon suggest potential efficiency gains offered by standardization in patient selection and treatment strategies.

A223: Framework for Developing an AOSpine Comprehensive Classification for Adult Spinal Deformity

J. Naresh Babu¹, Miranda L. van Hooff², Kenny Kwan³, Yabin Wu⁴, Kenneth Cheung³, Manabu Ito⁵, Jong-Beom Park⁶, David Polly⁷, and Marinus de Kleuver²

¹Mallika Spine Centre, Guntur, Andhra Pradesh, India

²Radboud University Medical Center, Nijmegen, Netherlands

³The University of Hong Kong, Hong Kong SAR, China

⁴AOSpine International, Davos, Switzerland

⁵National Hospital Organization, Hokkaido Medical Center, Hokkaido, Japan

⁶Uijeongbu St Mary’s Hospital, The Catholic University of Korea College of Medicine, Uijeongbu, Republic of Korea

⁷University of Minnesota, Minneapolis, MN, USA

Introduction: In the aging population adult spinal deformity (ASD) has an increasing burden on society, by causing severe back and leg pain, reducing quality of life, and impairing function in patients with symptomatic deformity.^1-3 ASD is a multifactorial condition comprising a combination of an abnormal curvature of the spine and coexistence of spinal canal stenosis, radiculopathy, and multiple comorbidities. The available ASD classifications are predominantly aimed at classifying the curvature based on radiographic parameters and do not consider associated conditions. As patients with ASD behave differently, this is insufficient for clinical decision making in treatment management. Consequently, a lack of universal consensus on the treatment aspects exists, resulting in controversial treatment outcomes. The purpose of this study is to develop a comprehensive clinical multifactorial classification to match patients with ASD to surgery decision and appropriate surgical management strategy. Materials and Methods: The development of the comprehensive AOASD classification follows the 3-phased mixed methods pathway as proposed by Audigé et al,⁴ which is based on the methodology of developing health measurement scales: (1) development, including pilot agreement study (current study); (2) reliability and accuracy in clinical setting; and (3) prospective validation and association with patient outcomes.⁴ The current study comprises the first (development) phase and consists of 3 parts: part A—systematic review; to identify and summarize purposes, characteristics, usage, and methodological measurement properties of existing ASD classifications; part B—to develop a preliminary ASD classification, based on part A and using the outcome domains that matter to patients and modifying case-mix and risk factors that have been established as part of the SRS (Scoliosis Research Society) Core Outcome Set Scoliosis for ASD. A 3-round modified Delphi study will be performed with an international AOSpine panel of 20 to 30 ASD-experts, covering all continents, to reach formal consensus on a preliminary AOASD classification. Part C consists of international expert consensus on the appropriate use of the preliminary AOASD classification. Patient cases (anonymized case descriptions including full-spine X-rays) are classified by using the preliminary classification and the level of agreement among panelists will be determined (ie, content validity and preliminary quantified reliability concerning the clinical usage of the classification). Results: The proposed methodology systematically addresses the development, reliability testing and validation phases of the development of classification. By incorporating the results of SRS Core Outcome Set Scoliosis for ASD data into the classification, the international expert consensus and global perspective will be achieved. Conclusion: ASD is an increasing multifactorial condition. A comprehensive classification covering clinically relevant factors to match patients to surgery and the appropriate surgical treatment strategy is currently lacking. In this 3-part study, we present the framework for developing a comprehensive multifactorial classification that supports spine surgeons in decision making, ultimately enhancing treatment outcomes in ASD surgery.

References

1. Pellisé F, Vila-Casademunt A, Ferrer M, et al; European Spine Study Group, ESSG. Impact on health related quality of life of adult spinal deformity (ASD) compared with other chronic conditions. Eur Spine J. 2015;24:3-11. doi:10.1007/s00586-014-3542-1

2. Schwab F, Dubey A, Gamez L, et al. Adult scoliosis: prevalence, SF-36, and nutritional parameters in an elderly volunteer population. Spine (Phila Pa 1976). 2005;30:1082-1085.

3. Ames CP, Scheer JK, Lafage V, et al. Adult spinal deformity: epidemiology, health impact, evaluation, and management. Spine Deform. 2016;4:310-322. doi:10.1016/j.jspd.2015.12.009

4. Audigé L, Bhandari M, Hanson B, Kellam J. A concept for the validation of fracture classifications. J Orthop Trauma. 2005;19:401-406.

A224: The Intraoperative Neurological Monitoring for Corrective Surgery of Thoracolumbar Kyphosis: Results and Trouble Treatment

Yan Zeng¹, Zhongqiang Chen¹, and Yue Ma¹

¹Peking University Third Hospital, Beijing, China

Introduction: To analyze the validity of somatosensory-evoked potentials (SEP) and motor-evoked potentials (MEP) in the monitoring of spinal cord function during corrective surgery for thoracolumbar kyphosis and suggest the proper management of abnormal monitoring events. Material and Methods: From June 2008 to February 2017, 196 cases of thoracolumbar kyphosis underwent corrective surgery at our medical center. The SEP + MEP monitoring for neurological function of spinal cord was routinely applied in the surgical procedure. The warning criteria of SEP was over 50% decrease of amplitude or over 10% elongation of latent period, and the warning criteria of MEP was over 80% decrease of amplitude or vanishing of recording data. Prompt and appropriate measures were applied to correct the abnormal change. The postoperative neurological function was observed for the patients who suffered the intra-operative positive monitoring event. The relation of spinal cord injury and abnormal neurological monitoring during surgery was analyzed, and standard management process for positive event was founded. Results: All the patients had the surgery accomplished. The osteotomy techniques were Smith-Petersen osteotomy (SPO) for 66 cases, pedicle subtraction osteotomy (PSO) for 47 cases, and vertebral column resection (VCR) for 83 cases. The average kyphosis correction rate was 69.3%. Among the 196 patients, 11 cases (5.6%) suffered spinal cord injury after surgery, and all of them had intraoperative SEP or MEP abnormal change. In these 11 patients, there were 9 cases of posttuberculous kyphosis and 2 cases of congenital kyphosis. As for the apex level, there were 7 cases above T6, 3 cases in T7-T10 and 1 case below T11. In the patients who had no spinal cord injury, 48.1% of them had SEP warning, and 8.1% had the warning continued until the end of surgery despite of proper management; 18.9% of patients had transient MEP warning, and 2 cases kept the warning until the end of surgery. The upper-middle thoracic spine deformity combined with preoperative neurological compromise had the highest risk of spinal cord injury during surgery. The monitoring physician, anesthetist, and spine surgeon should cooperate during surgery. When the warning is reported, the relevant factors of anesthesia will be considered first, followed by the checking of key steps and high-risk manipulation, such as screws position, osteotomy, correction, and so on. The prompt and proper measurement should be applied to ensure spinal cord safety in maximum extent. Conclusion: Spinal cord safety is the key point during corrective surgery for thoracolumbar kyphosis. Intraoperative SEP + MEP monitoring can report the change of neurological function promptly and effectively. The management of positive monitoring event should follow standard process and execute strictly.

OP34: Trauma-Cervical 3

A225: The Effect of Frailty on Patient Outcome After Traumatic Spinal Cord Injury

Daniel Banaszek¹, Tom Inglis¹, Travis Marion², Eryck Moskven¹, Carly Rivers³, Alana Flexman⁴, Tamir Ailon¹, Nicolas Dea¹, Raphaele Charest-Morin¹, Marcel Dvorak¹, Charles Fisher¹, Brian Kwon¹, Scott Paquette¹, and John Street¹

¹Vancouver General Hospital, Vancouver, British Columbia, Canada

²Thunder Bay Health Sciences Center, Thunder Bay, Ontario,Canada

³Rick Hansen Institute, Vancouver, British Columbia, Canada

⁴Vancouver Coastal Health Research Institute, Vancouver, British Columbia, Canada

Introduction: Frailty, defined as a state of decreased reserve and susceptibility to external stressors, has previously been shown to negatively affect postoperative outcome in an elective spine surgery population. This relationship has not, however, been previously studied in a population of patients with traumatic spinal cord injury. This study sought to determine the effect of frailty on patient outcome after traumatic spinal cord injury (tSCI). Material and Methods: Study Design: Prospective cohort study. Subjects: All patients with tSCI were identified in our prospectively collected database from 2007 to 2016. Baseline demographic and comorbidity data was collected for all subjects, including for all variables of the previously validated modified frailty index (mFI). Outcome Measures: Analysis was conducted to examine correlations between patient age, total motor score (TMS) on admission, and mFI on patient outcome variables including acute length of Stay (LOS), number of adverse events (AEs), and in-hospital mortality. Results: A cohort of 634 patients were identified. Of these, 488 (77%) were male, with an average age of injury of 50.3 ± 19.8 years. Mean time to presentation to definitive care was 8.0 ± 12.0 hours. The most common mechanisms of injury included falls (46.5%), transport (25.2%) and sports injuries (21.1%). On presentation, injury AIS grades were AIS A in 245 (39.1%), AIS B in 52 (8.3%), AIS C in 123 (19.1%), and AIS D in 206 (32.9%). The most common levels of injury included, high cervical (37.4%), low cervical (36.3%), and thoracolumbar 26.3%) regions of the spine. The most common adverse events experienced by patients during their inpatient stay included urinary tract infection (43.1%), pneumonia (42.1%), neuropathic pain (40.9%), cardiac arrest/failure/arrhythmia (22.6%), and delirium (21.5%). Bivariate analysis revealed multiple statistically significant associations. mFI was a strong predictor of increased acute LOS (corr = 0.163; P < .0001), number of AEs (corr = 0.1664; P < .0001), and in-hospital mortality (corr = 0.155; P < .0001). Age at injury was also significantly correlated with acute LOS (corr = 0.0809; P = .0418), number of AEs (corr = 0.0937; P = .0231), and in-hospital mortality (corr = 0.2639; P < .0001). Last, motor score on admission was also predictive of acute LOS (corr = −0.4749; P < .0001), number of AEs (corr = −0.3069; P < .0001), and in-hospital mortality (corr = −0.2249; P < .0001). Subgroup analysis was then performed on patients aged >65 years. Mean frailty index was not predictive of acute LOS (P = .1533), number of AEs (P = .2337), or in-hospital mortality (P = .6593). Age at injury was not predictive of acute LOS (P = .0571); however, it remained significant for number of AEs (P = .0058), and in-hospital mortality (P < .0001). This was also true for motor score on admission, which was predictive of acute LOS (P < .0001), number of AEs (P = .0038), and in-hospital mortality (P < .0001). Conclusion: Age, mFI, and TMS on admission are important determinants of outcome in patients with tSCI. Furthermore, frailty score is predictive of outcome in the general tSCI population, but not in the elderly. This suggests that younger, “frail” individuals have significantly poorer outcomes than young, healthy individuals, however the inter-relationship between advanced age and decreased physiologic reserve is not as clear. Identification of frailty in a younger population as a pre-injury risk factor may be useful for perioperative optimization, risk stratification, and patient counseling.

A226: The Effects of Glycemia in Survival and Neurological Recovery After Traumatic Spinal Cord Injury

Julio Furlan¹, and Michael Bracken²

¹Toronto Rehabilitation Institute, University Health Network and University of Toronto, Toronto, Ontario, Canada

²Yale University, New Haven, CT, USA

Introduction: While experimental and clinical evidence indicates that hyperglycemia worsens neurological outcome after traumatic brain injury, the impact of hyperglycemia in secondary mechanisms of neuronal damage after acute spinal cord injury (SCI) has been little investigated. This study examined the potential association of glycemia in the hyperacute stage postinjury and outcomes after acute traumatic SCI. Material and Methods: This retrospective cohort study includes all patients who were enrolled into the NASCIS-3 trial. Data on the glycemia within 24 hours, at 48 hours and at day 7 after acute SCI were examined as potentially associated with survival within the first year post-SCI as well as neurological recovery at 6 weeks, at 6 months, and at 1 year after SCI. Neurological recovery included the NASCIS motor, sensory, and pain scores. Analyses of the dependent continuous variables (ie, NASCIS motor, sensory, and pain scores) were performed using linear regression analyses adjusted for the major potential confounders. Survival analysis was performed using Kaplan-Meier curve and log-rank test. Results: There were 76 women and 423 men with mean age of 35.7 years (age range from 14 to 92 years) who mostly sustained cervical SCI due to motor vehicle accident followed by falls. On admission, 96.6% of the individuals had hyperglycemia. Glycemia varied from 125 to 533 mg/L among the individuals with hyperglycemia. There was a significant decline in the initial glycemia within 24 hours (188.20 ± 2.29 mg/L) when compared with glycemia at 48 hours (164.44 ± 2.08 mg/L) and at day 7 post-SCI (125.02 ± 2.25 mg/L; P < .0001). The results of the regression analyses revealed that higher glycemia within 24 hours postinjury was associated with lower motor, sensory, and pain scores at 6 weeks and at 6 months, but not at 1 year following SCI (Table 1). Glycemia at 48 hours and at day 7 postinjury was not associated with motor, sensory, and pain scores at 6 weeks, at 6 months or at 1 year following SCI, except for a poorer motor recovery and a greater pain score at 1 year after SCI related to higher glycemia at 48 hours postinjury. Survival analysis (Figure 1) revealed that hyperglycemia within 24 hours and at 48 hours postinjury was not associated mortality within the first year after SCI (P = .3254 and P = .0696, respectively). However, hyperglycemia at day 7 postinjury was associated with greater mortality after SCI. Conclusion: The results of this study suggest that hyperglycemia at day 7 postinjury may be associated with greater mortality within the first year following SCI. Among the survivors, glycemia at 24 hours postinjury was associated with poorer motor and sensory recovery as well as greater pain scores within the first 6 months after SCI. However, glycemia within 24 hours, at 48 hours, and at day 7 postinjury did not adversely affect the individuals’ neurological recovery at 1 year following SCI. Further investigations are needed to clarify the reasons why hyperglycemia may be associated with greater mortality at 1 year after SCI, which could have clinical implications.

Table 1.

Results of the Regression Analyses Adjusted for Neurological Scores at Admission, Age, Sex, Trial Drug protocol, Level of SCI, Glasgow Coma Score, and Serum Creatinine Concentration.

Dependent Variable		R 2	F	P
Motor at 6 weeks	Model	0.603	56.10	<.0001
	Glycemia at 24 hours		6.95	.0088
Sensory at 6 weeks	Model	0.597	53.95	<.0001
	Glycemia at 24 hours		4.00	.0464
Pain score at 6 weeks	Model	0.554	47.24	<.0001
	Glycemia at 24 hours		7.82	.0055
Motor at 6 months	Model	0.524	39.70	<.0001
	Glycemia at 24 hours		4.03	.0455
Sensory at 6 months	Model	0.563	46.24	<.0001
	Glycemia at 24 hours		3.01	.0836
Pain score at 6 months	Model	0.497	36.47	<.0001
	Glycemia at 24 hours		5.39	.0209
Motor 1 year	Model	0.476	31.82	<.0001
	Glycemia at 24 hours		2.98	.0854
Sensory at 1 year	Model	0.525	38.55	<.0001
	Glycemia at 24 hours		1.72	.1910
Pain score at 1 year	Model	0.493	34.6	<.0001
	Glycemia at 24 hours		2.39	.1227
Motor at 6 weeks	Model	0.598	54.75	<.0001
	Glycemia at 48 hours		2.33	.1282
Sensory at 6 weeks	Model	0.592	52.73	<.0001
	Glycemia at 48 hours		0.03	.8731
Pain score at 6 weeks	Model	0.548	45.71	<.0001
	Glycemia at 48 hours		2.96	.0863
Motor at 6 months	Model	0.520	39.17	<.0001
	Glycemia at 48 hours		1.79	.1822
Sensory at 6 months	Model	0.559	45.56	<.0001
	Glycemia at 48 hours		0.01	.9171
Pain score at 6 months	Model	0.494	35.89	<.0001
	Glycemia at 48 hours		1.70	.1929
Motor at 1 year	Model	0.487	33.08	<.0001
	Glycemia at 48 hours		4.10	.0437
Sensory at 1 year	Model	0.556	43.53	<.0001
	Glycemia at 48 hours		0.14	.7092
Pain score at 1 year	Model	0.492	34.45	<.0001
	Glycemia at 48 hours		4.27	.0396
Motor at 6 weeks	Model	0.638	58.83	<.0001
	Glycemia at day 7		2.91	.0892
Sensory at 6 weeks	Model	0.638	58.10	<.0001
	Glycemia at day 7		0.01	.9046
Pain score at 6 weeks	Model	0.583	48.10	<.0001
	Glycemia at day 7		0.02	.8928
Motor at 6 months	Model	0.517	34.97	<.0001
	Glycemia at day 7		2.13	.1458
Sensory at 6 months	Model	0.566	42.38	<.0001
	Glycemia at day 7		0.28	.5962
Pain score at 6 months	Model	0.498	35.20	<.0001
	Glycemia at day 7		0.43	.5128
Motor at 1 year	Model	0.468	28.05	<.0001
	Glycemia at day 7		1.25	.2647
Sensory at 1 year	Model	0.524	35.05	<.0001
	Glycemia at day 7		0.07	.7867
Pain score at 1 year	Model	0.492	31.59	<.0001
	Glycemia at day 7		0	.9824

Figure 1.

Survival analysis using Kaplan-Meier curve with log-rank test (P < .0001).

A227: Management of C2 Odontoid Peg Fragility Fractures: A UK-Wide Survey

Anna L. Watts¹, and James Tomlinson¹

¹Sheffield Teaching Hospital NHS Foundation Trust, Sheffield, UK

Introduction: Patients sustaining fragility fractures of the C2 odontoid peg have significant 30-day mortality rates of 10% rising to 16% at 90 days. Substantial controversy exists regarding optimal management of these fractures. To date there have been no prospective randomized control trials that would serve to inform best practice. We aimed to determine current practice in the management of these fractures throughout the United Kingdom. Material and Methods: A UK-wide, cross-sectional survey was conducted asking 10 questions regarding the initial management, investigation and follow-up of an elderly patient with a type 2 fragility odontoid peg fracture. This was publicized through the British Orthopaedic Association website and sent to all members of the Society of British Neurological Surgeons (SBNS) via email. Results: We received 91 responses, 53% from orthopedic consultants, 33% from neurosurgical consultants, and 12% from senior spine fellows. A total of 87% of respondents chose treatment with a cervical orthosis, with 85% of these opting for a semirigid Aspen or Philadelphia collar compared with 15% opting for soft cervical collar. Other treatments included operative or halo fixation (6%) and treatment with analgesia alone and no orthosis (7%.). Time to first follow-up varied from 1 week to 6 weeks, with some not following up patients at all. Disparity existed between investigations performed at last follow-up with 19% performing anteroposterior and lateral radiographs, 67% lateral flexion/extension radiographs, and 12% computed tomography. If the imaging suggested nonunion, 66% allowed the patient to remove any orthosis and mobilize freely while 28% considered operative fixation. Conclusion: There is widespread variation in the management, investigation, and follow-up of fragility peg fractures. Multiple imaging modalities are used, yet frequently have no impact on subsequent management. The high-risk nature of these injuries suggests a strong case for a randomized trial to treatment methods. We should be aiming to reduce the widespread variation in treatment in this high-risk group of patients.

A228: A Simple Surgical Technique for Easier Anterior Odontoid Screw Fixation

Sherif Elghamry¹, Youssry Elhawary¹, and Mohamed Khattab¹

¹Orthopedic and Spine, Cairo, Egypt

Introduction: Anterior odontoid screw fixation (AOSF) is a reliable osteosynthetic technique that is used for open reduction and fixation of type 2 and rostral type 3 odontoid fractures. The inability to use a drill guide that affixes to the spine, however, may limit the ability to insert a screw with a trajectory that grasps the fracture fragments. The purpose of this study is to describe a simple, safe, and less expensive technique for insertion of the odontoid screw through a minimally invasive anterior approach. Materials and Methods: Using beveled bone marrow biopsy needle (Jamshidi needle) as sleeve for the guide wire insertion, Nine patients (8 males and 1 female) with odontoid fractures were treated with AOSF under fluoroscopic guidance between October 2014 and March 2017. Results: Operations were completed with less technical difficulties. All the screws where placed in good position without intra- or postoperative soft tissue or neuro-vascular complications. Satisfactory clinical results were achieved in all patients. After a mean follow-up of 33.6 months (range 16-45 months), radiographic fusion was documented for 8 of 9 patients (89%). No clinical complications or hardware loosening or breakage occurred. Conclusions: Based on our results, this less invasive AOSF procedure using Jamshidi needle and fluoroscopy is technically easy and associated with good clinical and radiological results.

OP35: Degenerative-Lumbar 3

A229: Surgery for Lumbar Spinal Stenosis in Octogenarians: Is It Worth It?

Sultan Aldebeyan¹, Ahmed Aoude¹, Felipe E. Nares¹, Gregory McIntosh¹, Alex Soroceanu¹, and Kenneth Thomas¹

¹University of Calgary, Calgary, Alberta, Canada

Introduction: Lumbar spinal stenosis, with the consequent neurogenic claudication is a common diagnosis among the elderly. Lumbar decompression is the most common spine surgical procedure performed for such pathology. With the aging of our population, elective spine surgery is becoming more common in the elderly. Even though surgical decompression is an effective treatment, numerous patients are not considered good candidates for surgery due to advanced age or comorbidities. The aim of this study was to identify if age, more specifically being older than 80 years, influences the effect of lumbar decompression surgery on leg pain at the 1-year follow-up. Methods: Prospectively collected CSORN (Canadian Spine Outcomes Research Network) data was used to identify patients with a pathological diagnosis of lumbar stenosis, who underwent surgery and had 1-year postoperative follow-up data. The cohort was divided into 2 groups based on age (<80, >80 years). The 2 groups were compared with regard to baseline patient characteristics and surgical factors using t tests and χ² tests, as appropriate. The impact of age on 1-year improvement in health-related quality of life (HRQL) was assessed using analysis of covariance (ANCOVA), adjusting for significant confounders identified on univariate analysis. Results: One-year HRQL data was available on 1507 patients (<80 years old n = 1414, >80 years old n = 93). The overall median patient age was 61 years (range 18-90 years), and 49% were female patients. Analysis of confounders demonstrated that the older patients were more likely to have a higher ASA (American Spinal Association) score, less likely to smoke, less likely to be receiving insurance payments, had less preoperative narcotic use, and had a lower body mass index (BMI). The 2 groups were similar with regard to gender distribution, exercise status, pre-operative neurological deficit, symptoms duration, and number of patients who had a fusion and prior lumbar surgery. After adjusting for confounders, there was no statistically significant difference in the 1-year improvement on the Oswestry Disability Index (ODI) (−18.90 vs −21.27, P = .798), EQ5D (−0.06 vs 0.01, P = .072), Short Form–12 mental component summary (SF-12 MCS) (0.60 vs 3.54, P = .067), visual analogue scale (VAS) back pain (−3.88 vs −3.45, P = .16) and VAS leg pain (−4.31 vs −3.69, P = .28). The >80-year-old group experienced statistically less improvement on the SF-12 physical component summary (PCS) (5.96 vs 10.78, P = .027) 1 year after surgery, compared with their younger counterparts (although the difference was smaller than the minimal clinically important difference [MCID] for the PCS). Conclusions: In the present study, we showed that patients older than 80 years did just as well clinically as patients younger than 80 years. Although at the 1-year postoperative follow-up, the change in VAS leg pain and SF-12 PCS score showed less improvement in the cohort of patients older than 80 years, these observed differences were not clinically significant. When considering patient reported outcomes, lumbar surgery in those greater than 80 years old may be a reasonable choice.

A230: Factors Associated With 12-Month Return to Work Following Surgery for Grade 1 Degenerative Lumbar Spondylolisthesis: An Analysis of the Quality Outcomes Database

Leslie C. Robinson¹, Erica F. Bisson², Mohamad Bydon³, Andrew K. Chan¹, Panagiotis Kerezoudis³, Steven D. Glassman⁴, Kevin T. Foley⁵, Eric A. Potts⁶, Christopher I. Shaffrey⁷, Mark E. Shaffrey⁷, Domagoj Coric⁸, John J. Knightly⁹, Paul Park¹⁰, Michael Y. Wang¹¹, Kai-Ming Fu¹², Jonathan R. Slotkin¹³, Anthony L. Asher⁸, Michael S. Virk¹², Mohammed A. Alvi³, Jian Guan², Regis W. Haid¹⁴, and Praveen V. Mummaneni¹

¹University of California San Francisco, San Francisco, CA, USA

²University of Utah, Salt Lake City, UT, USA

³Mayo Clinic, Rochester, MN, USA

⁴Norton Leatherman Spine Center, Louisville, KY, USA

⁵University of Tennessee, Semmes-Murphey Neurologic and Spine Institute, Memphis, TN, USA

⁶Indiana University, Goodman Campbell Brain and Spine, Indianapolis, IN, USA

⁷University of Virginia, Charlottesville, VA, USA

⁸Carolinas Healthcare System and Carolina Neurosurgery & Spine Associates, Charlotte, NC, USA

⁹Atlantic Neurosurgical Specialists, Morristown, NJ, USA

¹⁰University of Michigan, Ann Arbor, MI, USA

¹¹University of Miami, Miami, FL, USA

¹²Weill Cornell Medical Center, New York, NY, USA

¹³Geisinger Health, Danville, PA, USA

¹⁴Atlanta Brain and Spine Care, Atlanta, GA, USA

Introduction: Return to work (RTW) has become an increasingly important metric of spinal surgery success. Lumbar spondylolisthesis is a relatively common spinal pathology that may manifest as low back pain and/or leg pain, causing functional disability and hindering patients from working and performing activities of daily living. Currently, there is a lack of nationwide data evaluating RTW in lumbar spondylolisthesis. Material and Methods: We queried the Quality Outcomes Database for employed patients who underwent posterior spinal surgery for Meyerding grade I degenerative lumbar spondylolisthesis at 1 or 2 levels. We compared patients who returned to work within 12 months—and those who did not—with regard to baseline demographics, comorbidities, clinical characteristics, and patient-reported outcomes. A multivariable logistic regression model was fitted for 12-month RTW, adjusting for an array of preoperative variables. Results: A total of 251 patients with available 12-month follow-up data (81%) were analyzed. Median age was 58 years (interquartile range: 51-65 years) and 126 (50.2%) were females. On multivariable analysis, patients who were working at the time of surgery were significantly more likely to RTW compared with those who were on leave before surgery (OR 4.45, 95% CI 1.83-10.8, P = .01). In addition, post-college education (OR 4.75, 95% CI 1.29-17.6, P = .02) was associated with a significantly higher odds of RTW compared with those with high school or less education. Finally, among patients who underwent arthrodesis, 2 levels of fusion was associated with a lower odds of RTW compared with a single level of fusion (OR 0.28, 95% CI 0.12-0.67, P = .04). Model c-statistic was 0.78. Conclusion: Analysis of data from a national spinal registry revealed that patients who were actively working and had college educations were more likely to RTW 12 months following posterior spinal surgery for lumbar spondylolisthesis. The findings of the present study may assist surgeons and patients with setting pragmatic treatment expectations during preoperative discussions.

A231: Are Modic Changes Related to Long-Term Pain and Disability? A Prospective Cohort Study With 13-Year Follow-up

Peter Udby¹, Søren Ohrt-Nissen¹, Michael Ruud Lassen¹, Stig Brorson¹, Mikkel Andersen², and Leah Carreon²

¹Zealand University Hospital, Køge, Denmark

² Middelfart Hospital, Middelfart,Denmark

Introduction: Back pain is a globally leading cause of disability. Studies have shown that Modic changes (MC) correlate more strongly with low back pain (LBP) than disc degeneration alone. However, the long-term consequences in terms of patient-reported outcomes (PROs) have not been reported. This study evaluates if MC is associated with long-term disability and pain. Material and Methods: In 2004-2005, 207 patients with chronic LBP were enrolled in a randomized controlled trial (RCT) comparing cognitive training to physical therapy. Inclusion criteria were age 18 to 60 years and almost daily LBP of ≥4 for more than 4 months in the past year. Patients completed the following questionnaires at baseline and at 13-year follow-up: back and leg pain (0-10), Roland-Morris Disability Questionnaire (RMDQ), Activities of daily living (ADL), inflammatory pain pattern (IPP) and sick leave due to back pain (days past year). Patients were then stratified based on the presence or absence of MC on their initial magnetic resonance imaging (MRI). Results: Of the 204 cases with baseline MRIs, 82 (40%) had MC in a least one lumbar segment. A total of 170 cases (83%) were available for questionnaire follow-up including 67 (39%) with baseline MC. We found no difference in demographics, smoking status, back or leg pain or IPP scores at baseline or at 13-year follow-up. A decrease in pain scores over time was observed in both groups. Disability measured by the RMDQ score was not significantly different at baseline, but at 13-year follow-up the mean score was 9.6 (6.8) and 7.4 (5.6) in the −MC and +MC groups, respectively (P = .024). Measuring disability by ADL, we found the same results, with no difference at baseline and a follow-up score of 35.5% and 27.7% in the −MC and +MC, respectively (P = .028). Sick-leave due to LBP at 13-years follow-up, 22.9 days and 9.0 days in the −MC and +MC, respectively (P = .003). Conclusion: MC was not found to be a predictor for long-term pain, disability, or sick leave. Rather, the study found that LBP patients with MC have significantly less disability and sick leave at long-term follow-up. We encourage further studies to verify these findings.

A232: Effects of Concomitant Decompression Adjacent to a Posterior Lumbar Interbody Fusion Segment on Clinical and Radiologic Outcomes: Comparative Analysis 5 Years After Surgery

Tomiya Matsumoto¹, Shinya Okuda¹, Yukitaka Nagamoto¹, Yoshifumi Takahashi¹, and Motoki Iwasaki¹

¹Department of Orthopaedic Surgery , Sakai, Japan

Introduction: Although lumbar spinal stenosis (LSS) adjacent to degenerative spondylolisthesis (DS) is often observed, surgical treatment for such conditions remains controversial. The purpose of this study was to examine the effects of concomitant decompression adjacent to the posterior lumbar interbody fusion (PLIF) segment on the clinical and radiological outcomes 5 years after surgery. Material and Methods: This was a retrospective 1:1 matched case-control study. Forty-five consecutive patients who had undergone L3/4 posterior decompression with L4/5 PLIF for multilevel LSS with DS were enrolled from 2005 to 2011 (group D). As a control group, 45 age, sex, and preoperative disc height at L3/4-matched patients who had undergone L4/5 PLIF alone for L4/5DS were randomly selected (group A). Disc height, vertebral slippage, range of motion (ROM) of the disc angle, posterior opening angle, segmental lordotic angle, presence of the intradiscal vacuum phenomenon (IVP) at the L3/4 level, and lumbar lordosis at L1-S were measured on lateral radiographs. Radiological parameters were compared between the groups. In terms of clinical outcomes, the Japanese Orthopaedic Association (JOA) score and the requirement for additional L3/4 surgery were evaluated. Results: In terms of pre-/postoperative radiological parameters, group D showed significant differences in disc height (9.4 mm/7.4 mm) and the presence of IVP (7 cases, 15%/22 cases, 49%) (P < .01), while in group A, a significant difference was observed in disc height (9.9 mm/9.0 mm) (P < .05). There were no significant differences in other radiological parameters. In terms of pre-/postoperative radiographic changes between the groups, significant differences were detected regarding disc height narrowing of ≥3 mm (group D 31%, group A 9%) and IVP (group D 33%, group A 11%). The recovery rate of the JOA score was 58% for patients in group D and 61% for those in group A (P = .63). The reoperation rate at the L3/4 level was 2.2% (1 of 45) in group D and 6.7% (3 of 45) in group A. Conclusion: Concomitant decompression adjacent to the PLIF segment accelerated adjacent disc degeneration compared with PLIF alone, but it did not predispose to the development of instability 5 years after surgery. Moreover, the clinical outcomes were not significantly different between group D and group A. Hence, these results suggest that concomitant decompression adjacent to the PLIF segment may be an effective operative method for multilevel stenosis with DS.

OP36: Imaging 1

A233: Do Radiologic Impressions of Cervical Stenosis on Computed Tomography Correlate With Measured Torg-Pavlov Ratios?

John Mangan¹, James McKenzie¹, David Casper¹, Christopher Kepler Kepler¹, and John Koerner²

¹Thomas Jefferson University/Rothman Institute, Philadelphia, PA, USA

²New Jersey Spinal Medicine and Surgery, Maywood, NJ, USA

Introduction: Cervical spine imaging is part of the basic evaluation that occurs for all trauma patients. Typically, patients undergo computed tomography (CT) of the cervical spine to evaluate for fracture or other bony abnormalities. In obtunded patients, it is not possible to perform a reliable physical exam, so a CT scan and the radiologist’s interpretation is crucial to identify spinal disease. Cervical stenosis in the setting of trauma is associated with spinal cord injury. Identification of cervical stenosis on CT is typically determined by the radiologist’s subjective interpretation, which is commonly described as mild, moderate, or severe. The Torg-Pavlov ratio quantitatively assesses for cervical stenosis; a ratio <0.8 is considered indicative of cervical stenosis. This calculation, however, is not routinely preformed when interpreting cervical CT imaging. A Torg-Pavlov ratio of <0.7 has been associated with higher likelihood of having a spinal cord injury after trauma. Material and Methods: A retrospective review was performed of patients that presented to a level 1 academic trauma center who underwent CT scan of the cervical spine without contrast after assault, motor vehicle accident, fall, or other blunt trauma between February and March 2018. All studies were interpreted by an attending musculoskeletal radiologist. Patients found to have fracture, infection or an oncologic process were excluded as were patients that required intubation. All patients with CT studies that did not document the radiologist’s impression of the degree of stenosis in the final read were excluded. The Torg-Pavlov ratio was calculated based on a sagittal CT view of the cervical spine from C3 to C7. Results: Evaluation of 154 patients led to identification of 50 patients with 250 cervical spine levels who met the inclusion criteria the average age of the cohort was 56.7 years (range 19-96 years old). There were 23 males and 27 females. Of the 250 levels analyzed, 131 levels were incorrectly interpreted in the radiology report (52.4%) assuming the Torg-Pavlov definition of stenosis to be the gold standard. A total of 109 levels were labeled as being nonstenotic with a Torg-Pavlov ratio <0.8 (43.6%). Of the levels that were identified as non-stenotic by the radiologist, 55 had a Torg-Pavlov ratio between 0.7 and 0.80, 44 levels had a ratio between 0.6 and 0.70, and 10 levels had a ratio less than <0.60. There were 6 levels incorrectly labeled as stenotic that had Torg-Pavlov radios greater than 0.8. There were 9 levels that were labeled as mild stenosis that were measured as moderate stenosis, 4 levels labeled mild that were measured as severe and 3 levels were labeled as moderate that were measured as severe. Conclusion: This study demonstrates the need for critical evaluation of the degree of cervical stenosis present on a CT scan in order to make appropriate clinical decisions, particularly in obtunded patients. The poor accuracy that subjective interpretation provides is critical for surgeons to understand. Quantitative analysis can provide better assessment of the presence of spinal stenosis when compared with qualitative interpretation.

A234: Multiparametric Quantitative MRI to Measure Tissue Injury in the Spinal Cord: Development of a Diagnostic Model for Myelopathy via Supervised Machine Learning

Muhammad Ali Akbar¹, Jetan Badhiwala¹, Allan Martin¹, Benjamin, De Leener², Julien Cohen-Adad², David Mikulis¹, Adrian Carwley¹, Sukhvinder Kalsi-Ryan¹, and Jefferson Wilson¹

¹University of Toronto, Toronto, Ontario, Canada

²Polytechnique Montreal, Montreal, Quebec, Canada

Introduction: The clinical diagnosis of myelopathy can be challenging as the symptoms and signs can be subjective and diagnostic uncertainty is common. Anatomical magnetic resonance imaging (MRI) showing cord compression has poor specificity. MRI techniques that provide measures of demyelination, axonal injury, and atrophy may provide enhanced accuracy for cases with diagnostic uncertainty. The diagnosis of degenerative cervical myelopathy (DCM) is important because it is a progressive disease and early detection and treatment can avoid permanent disability. We describe a mutliparametric quantitative MRI protocol for microstructure analysis of the spinal cord to determine the precise degree of injury to the spine in the setting of DCM. We used the metrics to develop a diagnostic tool, comparing 5 statistical approaches for classification between healthy subjects and those with DCM. Material and Methods: A total of 35 healthy subjects and 56 patients with DCM were included in the analysis. All subjects were examined clinically followed by MRI scans in a 3-T GE scanner acquiring T2-weighted (T2w) imaging, diffusion tensor imaging (DTI), magnetization transfer (MT), and T2*-weighted (T2*w) imaging covering C1-C7. Image analysis was performed on Spinal Cord Toolbox (SCT) to calculate SC cross-sectional area (CSA), fractional anisotropy (FA), MT ratio (MTR), and T2*w white to gray matter signal intensity ratio (T2*w WM/GM) in the cord and normalized for confounding variables. Statistical analysis was performed with R version 3.3. Diagnostic models were developed using subject characteristics and MRI data using (1) logistic regression (LR), (2) linear discriminant analysis (LDA), (3) principal component analysis followed by logistic regression (PCA-LR), (4) k-nearest neighbors (kNN) with various k values (3, 5, 7), and (5) a support vector machine (SVM) model. Estimates of diagnostic accuracy were reported as corrected area under receiver operating characteristic curves (AUC). Results: All 5 models showed good diagnostic accuracy, with the SVM model showing the highest performance (AUC = 95.6%), outperforming LR (AUC = 93.6%), PCA-LR (AUC = 89.0%), LDA (AUC = 87.9%), and kNN (k = 5, AUC = 84.6%). The SVM model with cost = 100 outperformed other SVM models, which showed AUC ranging from 91.2% to 94.3%. The LR model retained CSAMCL (P = .0007), T2*w WM/GMRostral (P = .04), CSARostral (P = .08), and MTRRostral (P = .39). Conclusion: Multiparametric quantitative MRI techniques can provide immense amounts of data for a single patient. Supervised machine learning algorithms such as SVM allow us as clinicians to harness this information and achieve greater diagnostic accuracy than conventional statistical approaches.

A235: Do Culture Positive Cervical Discs Correlate With Radiographic Changes?

Charles Gordon¹, Robert Morales¹, Alysa Nash¹, Jael Camacho¹, Farooq Usmani¹, Jonathan Hurst¹, Sivacharan Gaddam¹, Eugene Koh¹, Kelley Banagan¹, Daniel Gelb¹, and Steven Ludwig¹

¹University of Maryland Medical Center, Baltimore, MD, USA

Introduction: Degenerative disc disease (DDD) is thought to be associated with infection with low-virulence bacteria. Propionibacterium acnes (P. acnes) is one of the most prevalent organisms identified in disc cultures in patients with lumbar DDD. Animal models of discs inoculated with low virulence bacteria, such as P. acnes, demonstrate the development of degenerative changes compatible with Modic changes (MCs) on magnetic resonance imaging (MRI). There is a lack of studies investigating the correlation between disc infection and the development of MCs in the cervical spine. We hypothesize that patients undergoing elective anterior cervical discectomy and fusion (ACDF) procedures for DDD who have positive disc cultures will have MCs visible on MRI. Material and Methods: Following institutional review board (IRB) approval, a consecutive series of patients undergoing elective ACDF procedures for DDD were enrolled in the study. During surgery, intervertebral disc tissues were taken with a sterile instrument and cultured in aerobic and anaerobic conditions for 14 days. A fellowship-trained neuroradiologist, who was blinded to culture results, reviewed preoperative MRI scans from each patient. Sagittal T1-weighted, T2-weighted, and STIR (short T1/tau inversion recovery) sequences were evaluated for the presence of (1) edema in the cultured discs, (2) type 1 and type 2 MCs in the adjacent endplates, and (3) perivertebral edema. Results: A total 72 patients who underwent ACDF due to cervical DDD were enrolled in our study and 125 discs samples were cultured. Of the discs evaluated, 62 discs had positive culture with 54 of 62 discs positive for P. acnes. Of the culture positive discs, 35.5% (n = 22) had type 1 MCs, 25.8% (n = 16) had type 2 MCs and 53.2% (n = 33) had no MCs. A total of 63 discs were culture negative. Of the culture negative discs, 22.2% (n = 14) had type 1 MCs, 30.2% (n = 19) had type 2 MCs and 58.7% (n = 37) had no MCs. There was no statistically significant association between positive disc culture and MCs on MRI. Disc edema was noted in 11.3% (n = 7) and 11.1% (n = 7) of the culture positive and culture negative discs respectively. Perivertebral edema was appreciated in 7.8% (n = 3) of culture positive and 18.2% (n = 6) of culture negative patients. The rate of perivertebral edema or disc edema was not significantly different between culture positive and negative patients. Conclusion: The presence of low virulence bacteria in the intervertebral discs of patients who underwent anterior cervical discectomy and fusion for cervical degenerative disc disease has no significant association with the development of Modic changes identifiable on MRI. However, given the high culture positive rates, bacterial infection of disc tissue may still play a role in the pathophysiology of degenerative disc disease.

A236: The Role of MRI Nerve Root Sedimentation Sign in the Evaluation of Degenerative and Combined Lumbar Stenosis Surgical Treatment Results

Sergey Lyulin¹ and Konstantin Dyachkov¹

¹Ilizarov Center of Ortopedie, Kurgan, Russian Federation

Introduction: Currently, there is no consensus regarding specific diagnostic criteria for lumbar spinal stenosis (LSS). To define imaging findings of LSS, Barz and colleagues recently described the nerve root sedimentation sign (SedSign). The SedSign was shown to have additional information for decision making in favor of surgery. It may be that the SedSign is related to treatment outcome for individuals with LSS, yet more research is warranted to determine whether or not this is the case. Material and Methods: The study based on analysis of the results of treatment of 50 patients with degenerative and combined lumbar stenosis. The algorithm of examination included magnetic resonance imaging (MRI) in the axial plane in the T2 TSE (turbo spin echo) mode, performed before and 3 months after the surgery. All patients underwent reconstruction of the spinal canal at different levels (L3-5) with instrumentation. The clinical results evaluation were based on visual analogue scale (VAS), Oswestry Disability Index (ODI), Short Form–36 (SF-36), and Frankel scales. Results: Analyzing the results, we used the previously described MRI nerve root SedSign, as an indicator of the adequacy of the treatment performed. SedSign on the level of the reconstruction was negative in 46 cases. MRI nerve root SedSign was positive on the level of the surgery in 4 patients. The neurological symptoms regress we found in 44 patients. All 44 patients had negative MRI nerve root SedSign. Conclusion: MRI nerve root SedSign is the indicator of the adequacy of the treatment performed in patients with degenerative and combined lumbar stenosis.

OP37: Deformity—Cervical

A237: The Characteristics and Management of Pain in Patients With Klippel-Feil Syndrome: Analysis of a Global Patient-Reported Registry

Kishan Patel¹, Hardy Evans², Enrico Tessitore³, Samuel Sommaruga³, Tariq Qureshi¹, Luis Kolb¹, Michael Fehlings⁴, Joseph Cheng², and Aria Nouri²

¹Yale School of Medicine, New Haven, CT, USA

²University of Cincinnati, Cincinnati, OH, USA

³University of Geneva, Geneva, Switzerland

⁴University of Toronto, Toronto, Ontario, Canada

Introduction: Klippel-Feil syndrome (KFS) is characterized by congenital fusion of cervical vertebrae. Because of the rarity of KFS, minimal research has been done to assess the quality and management of pain associated with this disorder. Using a large global database, we report a detailed analysis of the type, location, and treatment of pain in KFS. Material and Methods: Data was obtained from the CoRDS registry. Participants with cervical fusions were categorized into type I, II, or III based on the Samartzis criteria. Participants were asked to rate the effectiveness of pain relief of various treatment modalities, with 0 being no pain relief and 5 being maximal pain relief. One-way analysis of variance (ANOVA) and chi-square tests were conducted with significance considered at P < .05. Results: A total of 75 patients (60 female/14 male/1 unknown) were identified and classified as: type I, n = 21 (28%); type II, n = 15 (20%); type III, n = 39 (52%). A total of 70 participants (93.3%) experienced pain associated with KFS, 2 (2.7%) experienced no pain, and 3 (4.0%) did not respond. The average age of pain onset was 17.5 ± 13.4 years, with gradual onset (38%), gradual onset with sudden worsening (29%), or sudden onset (33%), and the average age of pain worsening was 27.6 ± 15.3 years. Muscle, joint, and nerve pain was primarily located in the shoulders/upper back (76%), neck (72%), back of head (50.7%), middle back (28%), and lower back (28%), and was most commonly characterized as tightness (73%), dull/aching (67%), and tingling/pins and needles (49%). Type III fusions were significantly associated with greater nerve pain (P = .01), headache/migraine pain (P = .02), and joint pain (P = .04). Patients with cervical fusions in the middle region (C2-C6) reported greater muscle, joint, and nerve pain (P = .03), while those with fusions in the lower region (C6-T1) tended to report greater headache/migraine pain (P = .10). Participants rated pain relief from over-the-counter (OTC) oral medications as 2.38 ± 1.16, prescribed oral medications as 3.06 ± 1.06, OTC topical medications as 2.16 ± 1.19, prescribed topical medications as 2.25 ± 0.98, injections as 2.38 ± 0.98, and surgery as 2.64 ± 1.64. Participants that pursued surgical treatment reported significantly more comorbidities (P = .02) and neurologic symptoms (P = .01) than nonsurgically treated participants and were significantly older when pain worsened (P = .04), but there was no difference in levels of muscle, joint, or nerve pain (P = .23), headache/migraine pain (P = .26), total number of cervical fusions (P = .72), location of fusions, or age of pain onset (P = .27). Conclusion: Pain has not been well documented in the literature as one of the symptoms of KFS. In this survey, we found that over 90% of participants experienced pain, primarily in the shoulders, upper back, and neck. Participants with middle cervical fusions or multilevel, contiguous fusions reported greater levels of muscle, joint, and nerve pain. Participants that pursued surgery experienced more comorbidities and neurological symptoms, such as balance problems, gait disturbances, and limited range of neck motion, but did not report more significant pain than nonsurgically treated participants.

A238: Treatment of Mild Cervical Myelopathy: Factors Associated With Decision for Surgical Intervention?

Michael Bond¹, Charles Fisher¹, Jacobs Bradley², Jefferson Wilson³, Christopher Bailey⁴, Sean Christie⁵, Raphael Charest-Morin¹, Paquet Jerome⁶, Michael Johnson⁷, Andrew Nataraj⁸, David Cadotte², Hamilton Hall⁹, Kenneth Thomas¹⁰, Raja Rampersaud⁹, and Nicolas Dea¹

¹Vancouver General Hospital, Vancouver, British Columbia, Canada

²Foothills Medical Centre, University of Calgary, Calgary, Alberta, Canada

³St Michael’s Hospital, University of Toronto, Toronto, Ontario, Canada

⁴Western University, Schulich School of Medicine, London, Ontario, Canada

⁵Dalhousie University, Halifax, Nova Scotia, Canada

⁶Université Laval, Quebec City, Quebec, Canada

⁷University of Manitoba, Winnipeg, Manitoba, Canada

⁸University of Alberta, Edmonton, Alberta, Canada

⁹University of Toronto, Toronto, Ontario, Canada

¹⁰University of Calgary, Calgary, Alberta, Canada

Introduction: Surgical treatment of cervical spondylotic myelopathy (CSM) is associated with improved outcomes in patients across all degrees of disease severity. A trial of supervised rehabilitation is recommended as an alternative to surgical treatment for patients with mild cervical myelopathy (modified Japanese Orthopaedic Association [mJOA] scale score ≥ 15). The purpose of this study was to evaluate which demographic, clinical, or radiographic factors predict selection for surgical intervention in patients with mild cervical myelopathy. Materials and Methods: This is a prospective cohort study of patients with CSM. Patients were recruited from 7 Canadian centers and data was gathered in the Canadian Spine Outcomes Research Network (CSORN) database. Demographic, clinical, radiographic, and health-related quality of life (HRQOL) data was collected on all patients at baseline. Multivariate logistic regression modeling was used to identify predictors of surgical intervention. Results: A total of 360 patients were enrolled, of whom 112 patients were categorized as mild myelopathy. A total of 105 (86.0%) were treated surgically and 17 (14.0%) were treated conservatively. Eighty of those with mild myelopathy (65.5%) were male and the mean age was 54.8 years (SD 12.6). Bivariate analysis revealed no differences between surgical and nonsurgical patients with regard to body mass index (BMI), duration of symptoms, medical comorbidities, and age. mJOA scores were higher in the nonoperative group (16.8, SD 0.99 vs 15.9, SD 0.89, P < .001). There were also differences with respect to Neck Disability Index (NDI), EQ-5D, SF-12 Mental Component and Physical Component Score and PHQ-9 scores, with worse baseline characteristics in those treated operatively (P < .05). There was no difference between groups for any radiographic parameters including spinal cord compression, spine alignment, and cord signal change. Multivariate analysis revealed that lower quality of life scores (EQ5D) was associated with selection for operative management. Conclusion: Patients who were treated surgically for mild cervical myelopathy did not differ from those treated nonoperatively with regard to demographic or radiographic parameters. After adjusted analysis, patients were more likely to have surgical intervention if they had worse quality of life scores as determined by the EQ-5D.

A239: Synthetic Osteoconductive Bone Substitute to Augment Occipitocerivcal Instrumentation in Children

Younus Hanif¹, Dominic Thompson², Steve Tucker², and Mohammed Shahvaiz Younus²

¹Department of Trauma and Orthopaedics , Stoke, UK

²Great Ormond Street Hospital, London, UK

Introduction: In children undergoing occipitocervical fixation (OCF), particularly the very young and those with bone dysplasias, autologous bone graft (iliac crest or rib) may be immature or of poor quality. Furthermore, bone graft donor site morbidity, including pain and local wound-related complications occur in up to 12%. Bone substitutes offer an alternative means of augmenting fixation but there use in a large cohort of children has not been reported. Our aim was to audit the efficacy of bone substitutes in children undergoing instrumented OCF. The primary outcome measure was radiological evidence of stability at last follow up. Material and Methods: A single institution retrospective case note review was performed of children undergoing instrumented OCF with or without decompression. Age at surgery, underlying diagnosis and indications for fixation were recorded. Occipitocervical instability due to trauma and children older than 16 years at the time of surgery were excluded Last available radiology was assessed for evidence of stabilisation and bone formation. Results: A total of 122 children (mean age 8 years) operated between 1993 and 2017 satisfied the inclusion criteria. Underlying diagnoses varies from Down’s syndrome to Moriquos syndrome. Mean follow-up was 8.2 years (range 1-24 years). Six patients had died during follow-up for reasons unrelated to the surgical procedure. Occipital plate, rod and screw constructs were used in all patients. Revision surgery due to instrumentation failure was required in 2/122. Complications were recorded in 6 of 122 patients. By final follow-up, all patients were deemed to have stable fixation. Conclusion: Osteoconductive bone substitutes can be safely used to augment OCF procedures in the pediatric population. Rates of successful fusion and wound related complications are comparable to contemporary series using autologous bone graft.

A240: Demographics, Comorbidities, and Common Procedures in Children With Klippel-Feil Syndrome: Analysis of the Pediatric National Surgical Quality Improvement Program

Hardy Evans¹, Kishan Patel², Abhijith Matur¹, Rafael De La Garza Ramos³, Jonathan Nakhla³, Reza Yassari³, Rani Nasser¹, Charles Stevenson⁴, Francesco Mangano⁴, Aria Nouri¹, and Joseph Cheng¹

¹University of Cincinnati College of Medicine, Cincinnati, OH, USA

²Yale College of Medicine, New Haven, CT, USA

³Albert Einstein College of Medicine, New York, NY, USA

⁴Cincinnati Children’s Hospital, Cincinnati, OH, USA

Introduction: Klippel-Feil syndrome (KFS) is a congenital anomaly that manifests with one or more levels of cervical vertebrae fusion, classically resulting in a combination of a short neck, low posterior hairline and reduced cervical motion. Because of its relatively rare occurrence, studies on KFS have been limited to case reports and series and have not been conducted on a large scale. To address this, we investigated the pediatric NSQIP database to report on the demographics, comorbidities, and common procedures associated with KFS. Material and Methods: Pediatric NSQIP data from 2012 to 2016 were obtained and analyzed. Patients with KFS were identified using ICD-9 code 756.16 and ICD-10 code Q76.1. Pearson chi-square statistics were calculated to determine significant differences in demographics, comorbidities, and presence of surgical complications. Independent-samples t tests were used to determine differences in procedure lengths and time to discharge. All statistical analyses were performed with significance at P < .05. Results: A total of 152 out of 369 176 patients (0.04%) in the database were diagnosed with KFS. Seventy-nine (52%) of KFS patients were male and 73 (48%) were female. There was no statistically significant difference in race or ethnicity between patients with and without KFS. When compared with the non-KFS population, KFS patients had higher rates of pulmonary and airway abnormalities, including structural anomalies (14.5%, P < .001), asthma (13.2%, P < .001), and bronchopulmonary dysplasia/chronic lung disease (7.9%, P = .012). Additionally, 25.6% of KFS patients had one or more cardiac comorbidities (P < .001), including ostium-secundum type atrial septal defects (8.6%, P = .027), ventricular septal defects (7.9%, P < .001), and tetralogy of Fallot (4.6%, P < .001). Similarly, 17.8% of KFS patients had a history of prior cardiac surgery (P < .001). Other common comorbidities included neuromuscular disorders (19.1%, P < .001), renal agenesis/dysgenesis (10.5%, P < .001), and compression of the brainstem (6.6%, P < .001). The most commonly performed procedures were suboccipital craniectomy with laminectomy (n = 15), and laminectomy with release of tethered spinal cord (n = 12). There were no significant differences in procedure length, complications, or time to discharge between the KFS and control populations. Conclusion: Patients with KFS have a higher rate of comorbidities than the general population, particularly those of the cardiovascular, pulmonary, and urogenital systems. Neurosurgical procedures commonly performed on pediatric patients with KFS include suboccipital craniectomy and release of tethered spinal cord. These procedures resulted in similar outcomes to patients without KFS. Further research should include characterization of the adult KFS population, and comparisons of surgical approaches to the spine in patients with KFS.

A241: Feasibility and Accuracy of Atlantoaxial Fluoroscopy-Guided Pedicle Screw Fixation in Patients Younger Than 12 Years: A Radiographic and Clinical Assessment

Yuehui Zhang¹, Jia Song¹, Jing Zhang¹, and Jiang Shao¹

¹Xin Hua Hospital Affiliated to Shanghai Jiao Tong University School of Medicine, Shanghai, China

Introduction: C1-C2 pedicle screw fixation provides excellent bony purchase and avoids neurovascular complications and is a widely accepted treatment method for atlantoaxial dislocation (AAD). There are limited data on the use of this technique for atlantoaxial fusion (AAF) in pediatric patients. This study aimed to investigate the feasibility and accuracy of fluoroscopy-guided atlantoaxial pedicle screw fixation in patients younger than 12 years. Methods: Between 2007 and 2017, 36 consecutive patients younger than 12 years underwent C1-C2 pedicle screw fixation for AAD. Anatomical parameters of the C1 pedicle were measured on preoperative computed tomography (CT). Accuracy of pedicle screw fixation was assessed on postoperative CT and evaluated using the following definitions: type I, screw threads completely within the bone; type II, less than half the diameter of the screw violates the surrounding cortex; and type III, clear violation of transverse foramen or spinal canal. Demographic, surgical, radiation dosing, and clinical data were recorded. Results: The study included 36 patients (21 males and 15 females, average age 6.9 ± 3.2 years) who underwent 144 screw fixations (67 C1 pedicle screws, 68 C2 pedicle screws, 5 C1 lateral mass screws, and 4 C-2 laminar screws) for a variety of pediatric AADs, with 36.5 ± 8.5 months of follow-up at a single institution. Among 135 pedicle screws, 96.3% were deemed “safe” (type I or II) and 80.7% (109/135) of the remaining screws were rated as being ideal (type I); 5 screws (3.7%) were identified as unacceptable (type III). Mean C1 pedicle length, width, height, and angle were 24.5 ± 2.1 mm, 8.4 ± 1.1 mm, 3.5 ± 1.6 mm, and 19° ± 3.2° for all patients, respectively. Average estimated blood loss (EBL) was 92 mL (range, 50-475 mL) and the average radiation dose was 6.2 mGy (in the final 26 cases). There were no neurovascular injuries. All patients showed radiographic stability and symptom resolution. Conclusion: C1-C2 pedicle screw fixation under fluoroscopy is safe and effective for AAD in children younger than 12 years. However, it may be technically challenging due to the special anatomical features of children and should be carried out by an experienced surgeon.

A242: Retrospective Review of Craniovertebral Junction (CVJ) Anomalies: A Clinical Profile and Outcome Analysis of Surgically Treated Patients

Prakash Paudel¹, Nilam Khadka², Prakash Bista², Gopal Raman Sharma¹, and Deepak Shrestha³

¹Nepal Mediciti, Kathmandu, Nepal

²National Academy of Medical Sciences, Kathmandu, Nepal

³Dhulikhel Hospital, Kathmandu, Nepal

Introduction: The craniovertebral junction (CVJ) is a complex anatomic region providing stability and mobility to the most important part of the craniospinal axis. CVJ anomalies can be associated with significant morbidity and mortality. The purpose of this study is to analyze clinical characteristics and outcome after surgical management of patients with CVJ anomalies presented to the Neurosurgery Department, Bir Hospital, Kathmandu Nepal. Material and Methods: A retrospective analysis of 21 patients, managed surgically for craniovertebral instability between 2013 and 2017, was performed. Clinical and radiological reports and operative records were evaluated. Patients were categorized clinically preoperatively and in follow-up according to their Nurick grade. Imaging studies were reviewed for bony and soft tissue details. All patients were managed with posterior surgery: occipitocervical fusion or C1-2 fusion with or without bony decompression. All patients were followed up clinically with repeat Nurick grading. Results: Of the 21 patients, two -thirds were male and one-third were female. Mean age was 40.7 years (16-73 years). The most common causes of CVJ instability were nonunion of old odontoid fracture (38.1%) and os odontoidium (38.1%). A total of 76.2% had intramedullary high signal intensities in T2-weighted magnetic resonance imaging (MRI) while 90% had cervicomedullary compression. Preoperatively, 52.38% had Nurick grading scale of 4 to 5 while 47.62% had 0 to 3 Nurick grade scale. Mean duration of follow up was 20.87 months (6-60 months). Postoperatively, 71.42% improved clinically, 23.80% remained unchanged while one patient deteriorated. Patients with longer duration of symptoms (40.8 vs 28.5 months) and intramedullary high signal intensities on MRI (100% vs 66.67%) trended toward an unfavorable outcome. Preoperative Nurick grade correlated significantly with the postoperative outcome (P = .042). Conclusion: Craniovertebral junction anomalies can be managed with posterior approach alone achieving adequate decompression and stabilization. Early surgical intervention is associated with better clinical outcome.

A243: Outcomes of 3-Column Osteotomy in Cervicothoracic Spine(C7/T1) for Congenital Cervicothoracic Scoliokyphosis in Children: A Report Of 30 Cases in a Single Center

Shengru Wang¹ and Jianguo Zhang¹

¹Peking Union Medical College Hospital, Beijing, China

Introduction: There have been many reports on the treatment of congenital scoliokyphosis. However, the characteristics and treatment of the congenital deformities in the cervicothoracic spine(C7/T1) has not been well described because of the rarity of the condition. This study was conducted to characterize the clinical presentation of congenital deformities in the cervicothoracic spine and report the outcomes of 3-column osteotomy in cervicothoracic spine(C7/T1) for congenital scoliokyphosis. Materials and Methods: Thirty patients (16 male 16, 14 female) averaged 11.0 years with congenital deformities in the cervicothoracic spine (C7/T1) were included in the study. All of them were treated with 3-column osteotomy in cervicothoracic spine (C7/T1): 21 hemivertebra resection, 8 vertebral column resection, and 1 pedicle substraction osteotomy. Eleven patients received osteotomies at C7 and 19 at T1. One-stage anterior and posterior approach was used in 1 patient with C6 resection and posterior-only approach was chosen for other patients. Standing anteroposterior and lateral radiographs were obtained in the postoperation and follow-up. Computed tomography (CT) and MRI scan of the spine, CT angiography (CTA) of bilateral vertebra arteries were taken for every patient. The patients’ charts were reviewed. Results: All patients were followed at least for 2 years. The local deformities of the cervicothracic spine could be classified into 3 types: 19 cases of formation failure, 4 cases of segmentation failure and 7 cases of mixed type. Twenty-eight patients had congenital deformities in other regions of the spine. Seven patients had 9 intraspinal deformities: 4 syringomyelia, 1 diastematomyelia, 2 tethered cord, and 2 meningocele. Preoperative neurological deficits were found in 3 patients. The averaged operation time was 253.8 minutes and the mean blood loss was 960.0 mL. The cervicothoracic curve was corrected from 55.4° to 15.3°. The segmental kyphosis was corrected from 25.4° to 13.1°, and the head tilting improved from 15.2° to 4.8°. Fifteen complications occurred in 12 patients: 6 transient root injuries, 1 transient cord injury, 2 progressions of compensatory curve, 2 implants failures, 2 hemothoraxs, 1 wound delayed union, and 1 atelectasis. Four patients received staged surgical intervention surgical treatment of the compensatory curve. Conclusion: Most congenital cervicothoracic deformities are fixed and have concurrent congenital deformities in other regions of the spine. Patients often have cosmetic problems due to pronounced heading tilting and uneven of the shoulder. Three-column osteotomy is an ideal procedure for this kind of deformities and could provide satisfied correction. However, 3-column osteotomy in cervicothoracic spine (C7/T1) was technical demanding with high risk of root injuries although most of them could be completely relieved during the follow-up. Implants failures may occur in the junction of rods with different diameters. Sufficient evaluation of the bilateral vertebra arteries should be made with CTA before the surgery. One stage or staged surgical treatment may be needed if the compensatory thoracic curve was severe and rigid.

A244: Anterior Correction and Reconstruction for Severe Dystrophic Cervical Kyphotic Deformity

Yifei Gu¹, and Wen Yuan¹

¹Shanghai Changzheng Hospital, Shanghai, China

Introduction: Neurofibromatosis type 1 (NF-1) is a multisystematic disease frequently associated with musculoskeletal abnormalities especially spinal deformities. Although cervical kyphosis due to NF-1 is relatively rare in clinical practice, the risk of severe neurological dysfunction, including myelopathy and life-threatening paralysis is higher than that in patients with thoracolumbar deformities. Only a few of studies have reported the surgical treatment for this challenging condition, and there is a lack of related guidelines for the treatment in the literature. In cases of NF-1-related cervical kyphosis, dystrophic changes, including scalloping of vertebral bodies have been shown to contribute to the pathogenesis and progression of deformity. Ventral compression of spinal cord by the diastrophic vertebral bodies may cause directly neurological damage. Kyphotic changes in cervical alignment may also increase the mechanical tension of spinal cord and cause stretch-mediated myelopathy. Anterior decompression and spinal column reconstruction can be achieved via an anterior only approach with appropriate preoperative managements. In this , we introduce a series of 7 patients with severe cervical kyphotic deformity who were treated via the anterior-only approach for the correction and analysis of the clinical and radiological outcomes. Material and Methods: In this study, we reviewed a series of 7 patients who underwent anterior procedures for correction of NF-1-associated cervical kyphotic deformity. After continuous preoperative skull traction, all patients received anterior corpectomy and fusion (ACCF), anterior discectomy and fusion (ACDF), or combined ACCF and ACDF for surgical correction and reconstruction. Pre- and postoperative local and global Cobb angles, correction rate, sagittal vertical axis (SVA) and T1-slope were assessed by X-ray. Japanese Orthopaedic Association (JOA) score, JOA recovery rate, visual analogue scale (VAS), and neck disability index (NDI) scores were recorded to assess the outcome. Results: Kyphosis was corrected successfully in all patients in terms of local and global Cobb angles (P < .05), with a correction rates of 83.1% (range, 66.0%-115.5%) and 88.6% (range, 61.1%-125.0%), respectively. JOA scores of patients was improved from pre-operative 10.4 (range, 6-14) to postoperative 15.4 (range, 14-17), with a recovery rate as 77.6% (range, 66.7%-100%). NDI scores were reduced from pre-operative 25.1 (range, 13-35) to postoperative 8.7 (range, 5-12). VAS scores were reduced from preoperative 7.0 (range, 4-9) to postoperative 2.3 (range, 1-3). Conclusion: This study has demonstrated that anterior correction and reconstruction is an alternative option for the treatment of NF-1-associated severe cervical kyphosis when deformity is localized, flexible, or fixed.

A245: Releasing Anterior Soft Tissue and Reduction of the Irreducible Atlantoaxial Dislocation Associated With Basilar Invagination Through Single-Stage Posterior Approach

Zan Chen¹ and Wanru Duan¹

¹Xuanwu Hospital, Capital Medical University, Beijing, China

Introduction: The management of irreducible atlantoaxial dislocation (AAD) combined with basilar invagination (BI) is challenging. The key factors hindering reduction include contraction of the anterior soft tissue and deformity of the atlantoaxial facet joint. Traditional posterior approach usually lacks the releasing of anterior soft tissue and results in unsatisfied reduction, while posterior fixation followed by the anterior release approach is risky and leads to higher costs and a longer length of stay in hospital. This study aims to release the anterior soft tissue through posterior route and reduce both atlantoaxial dislocation and basilar invagination in a single-stage operation using specially designed tools. Materials and Methods: Thirty-one irreducible AAD accompanied with BI cases (16 males, 15 females; median age of 42.8 ± 4.41) were retrospectively analyzed. All the cases were with congenital osseous abnormalities. We released anterior tissue through posterior route followed by cage implantation into facet joint and occipital-cervical fixation with cantilever technique. The clinical results were evaluated using Japanese Orthopedic Association (JOA) scale, and the main radiological measurements included anterior atlantodental interval (ADI), the distance of odontoid tip above Chamberlain line, clivus-canal angle (CCA) and the length of syrinx. We used paired t test to compare preoperative and postoperative JOA score and radiological measurements. Results: The mean JOA score increased from 11.65 to 14.48 at the 1-year follow-up. Complete reduction of AAD and BI was achieved in 23 patients (74.2%), and partial reduction was achieved in 8 patients (25.85%). The mean clivus-canal angle showed great improvement, from 113.0° preoperatively to 148.4° postoperatively. Shrinkage of the syrinx was observed in 6 patients in a week postoperation. Nine patients were followed up for 18 months and achieved bone fusion. A total of 22 patients has been followed up for 12 months, and 19 (86.3%) achieved bony fusion. One-side vertebral artery occlusion was diagnosed posteriorly in 2 cases without clinical symptoms. Conclusions: It appears effective and safe to treat irreducible AAD associated with BI using our strategy. Releasing anterior soft tissue from posterior approach can avoid transoral operation. Cage implantation intraarticularly and fixation with cantilever technique achieved complete reduction in most cases.

OP38: Novel Technologies 2

A246: Preliminary Results of the Impact of Telemonitoring on Postoperative Spine Surgery

Asdrubal Falavigna¹, Frederico Arriaga Criscuoli de Farias¹, Manuela Figueiró¹, and Natália Fontana Nicoletti¹

¹University of Caxias do Sul, Caxias do Sul, Brazil

Introduction: Degenerative disk disease (DDD) is the most common etiological factor behind lumbar pain, being associated with traumatic, lifestyle, genetic, and environmental factors. Persistent and severe pain often occurs, along with other sensory and motor deficits. Despite its favorable natural course for most patients, over 20% of patients may require surgical care. Monitoring and early management of mood disorders, overweight, the lack of physical activity, and the complications are necessary to improve postoperative results. Telemonitoring or remote monitoring is an innovative strategy to improve patient care, facilitating patient-physician contact through the use of many technological tools. There has been a growing interest in telemonitoring strategies, because of increasing access to smartphone apps and the internet over the last years. Telemonitoring for postoperative care of DDD patients is still to be consistently explored to benefit patients and physicians. The present study aims to determine any impact of telemonitoring on DDD patient improvement and quality of life. Material and Methods: Prospective cohort study of patients with degenerative lumbar spondylolisthesis or lumbar disc herniation refractory to conservative treatment that requires surgical treatment. The study was approved by the Ethics Committee of Caxias do Sul University, number 2.677.226/83 599 818.8.0000.5341. The participants are trained to use the remote monitoring devices: activity tracker wristband, a digital scale, and a smartphone app connected via Bluetooth. Through these, patients provide data on physical activity, hours of sleep, weight, mood, blood pressure, lab exams, and specific questionnaires for patient-reported outcome measurement (PROMs), such as EQ-5D-5L, EQ-5D-VAS, MSQ, HAD, ODI, Prolo scale, pain scale, and satisfaction index. PROMs were answered by the participants during the preoperative period and postoperative at 30-days, 6 months, and 1 year. Results: It was developed of a secure and interactive system approved by Health National Agency and Brazilian Medicine Federal Council. After that, a web-based “dashboard” for text, voice, and image was constructed to enable communications between the remote center, surgeon, and patient. The multidisciplinary monitoring team to work on the remote station was trained using 6 face-to-face meetings and 3 online meetings for simulations. A preliminary study for refinement of the system based on patient/provider input was performed initially with 35 patients. The adherence and interaction with the platform were 94%. Patients provided all the PROMs data inputs and consistently interacted with monitoring center professionals and the surgeon through the platform. Weight loss was observed in 76% of overweight patients using remote nutritionist support and specific exercises. The quality of life improved at 1 and 6 months after surgery when compared with the baseline. A high satisfaction index was observed due to the facilitation of interaction with the remote station and the surgeon. Conclusion: The platform allows for real-time interaction between included patients and the surgeon, as well as continuous educational measures based on the patient demands. High patient satisfaction index and adherence rates on telemonitoring with relevant changes in weight and quality of life have during the follow-up. Telemonitoring has excellent potential to immediate and accurate identification of postoperative complications.

A247: Carbon Fiber Components in Spine Tumor Surgery: A Single Center Case Series

David Choi¹

¹National Hospital for Neurology and Neurosurgery, London, UK

Introduction: Carbon fiber is a strong, light, and resilient material that has transformed the motor vehicle and sports equipment industries. In the field of spine surgery, composite materials are becoming more frequently used to reconstruct the spine after tumor resection. Vertebral body replacements and pedicle screws and rods are now available in carbon fiber composites, with several advantages over standard titanium components. One significant advantage is the lack of beam perturbation when carbon fiber components are subject to radiation or proton beam therapy, resulting in very little scattering or blocking of the radiation beam, and consequently greater accuracy and dose delivery. We present our clinical experience of carbon fiber spine implants. Material and Methods: Retrospective case series of 32 patients who received carbon fiber composite components after resection of spine tumors. The main indication for use of carbon fiber components was for primary spine tumor surgery, particularly when proton beam therapy is the recommended postoperative treatment. We documented the ease of use/insertion of carbon fiber pedicle screws, reviewed the practicalities of using these new systems, and the postoperative complications. Results: There were no implant-specific or carbon-specific complications. 1 patient required revision surgery for loosening of pedicle screws, 22 months after surgery. One patient required revision surgery for Escherichia coli surgical site infection after surgery, 6 months following proton beam therapy. One patient had transient dysphagia after a 2-level cervical corpectomy. Some of the first-generation carbon fiber instrumentation were considerably more time-consuming to insert compared with standard titanium constructs, and there were limitations in carbon fiber rod length and shape. Conclusion: We describe the benefits, limitations and practicalities of first-generation carbon fiber composite pedicle screws and constructs, used after resection of spine tumors. There are theoretical advantages of carbon fiber implants for radiation and proton beam planning and delivery. Complication rates were similar to standard titanium implants.

A248: Utility of “Tip Protocol” in Complex Spine Surgery: An Effective, Simplified, and Reproducible Rescue Algorithm of Management When Intraoperative Neurophysiological Potentials Fall Down

Juan Pablo Cabrera Cousiño¹, Francisco Luna A.¹, Esteban Torche V.¹, Sebastián Vigueras A.¹, Maximo Torche V.¹, Mario Condal M.¹, Rudy García V.¹, Luis Lastra R.¹, Walter Rivas W.¹, Roberto Vigueras A.¹, and Guillermo Valdés I.¹

¹Hospital Clínico Regional de Concepción, Concepción, Chile

Introduction: Complex spine surgery has risk to develop a new neurological worsening, in spite of knowledge and surgical experience. Intraoperative neurophysiological monitoring (IOM) is a well-known tool used to prevent this worsening. However, when intraoperative potentials fall down, it is not always possible to recover them with proper maneuvers. There are several recommendations and protocols during this scenario, but they are time-consuming, hard to apply, and not always reproducible in many spine centers. Our objective is to show the utility and efficacy of a simplified and reproducible IOM algorithm of management in complex spine surgery. Material and Methods: Prospectively collected and retrospectively reviewed study. Spine surgeries with instrumentation and IOM were analyzed, from February 2016 to July 2018. Our initial warning criteria of IOM is 50% amplitude decrease in TcMEP or SSEP; however, when decrease was 80% or greater, rescue maneuvers start. We defined TIP protocol as follows: T: time and temperature, stopping surgical maneuvers or taking instrumentation out, increase body temperature; I: irrigation with warm solution and papaverine; P: intravenous papaverine and increase blood pressure. Results: During this period, 511 spine surgeries with IOM were performed, potentials of 6 patients were unrecordable (1.2%). Sixteen patients had significant amplitude TcMEP decrease (3.17%): 11 of them had transient changes recovered with TIP protocol application, and 5 had permanent changes. Of these latter 5 patients, 2 developed a new motor worsening and 3 woke up with deficit but recovered during follow-up. Time of potentials recovery since TIP protocol was applied is less than 30 minutes. Appearance of 30% amplitude TcMEP recovery at least when 100% loosening occurred, is highly predictable of a nonclinical change. Conclusion: The TIP protocol has shown to be an effective, simplified, and reproducible algorithm of management to recover intraoperative potentials in any spine center, avoiding clinical new neurological deficit during complex spine surgery. When permanent changes are detected, the probability of clinical impact is higher; however, with application of TIP protocol, some of these patients will have transient deficit.

A249: Radiostereometric Analysis as a Diagnostic for Assessing Vertebral Fusion: A Phantom Study

Trevor Gascoyne¹, Sara Parashin¹, Michael Johnson², and Mohammad Zarrabian²

¹Orthopaedic Innovation Centre, Winnipeg, Manitoba, Canada

²Division of Orthopedic Surgery, Winnipeg, Manitoba, Canada

Introduction: The purpose of this research was to determine the feasibility of radiostereometric analysis (RSA) as a diagnostic tool for assessing nonunion following spinal arthrodesis procedures. Furthermore, to estimate clinical thresholds for precision and accuracy of the proposed method in the cervical and lumbar spine. Material and Methods: A 3-level lumbosacral and a 4-level cervical posterior arthrodesis procedures were performed on an artificial spine model (Sawbones, WA). Using a spring-loaded inserter (RSA Biomedical, Sweden), 8 to 10 RSA markers were placed within each of the L4 and L5 segments in the spinous process (L4 only), lamina, transverse processes, posterior and anterior (down the pedicle) wall of the vertebral body. Eight to 10 markers were placed within the proximal sacrum (S1) at the medial and lateral crests, tuberosity, and within the canal wall. Four to 8 RSA markers were placed into the C3-C6 lateral masses. Titanium screws and rods were applied to the spinal segments. Identical procedures were then performed on a cadaveric spine using similar bead placement and hardware. RSA imaging consisted of 12 double exams (24 exams) of the cervical and lumbar regions for both the Sawbones and cadaveric spine to assess precision of measurement under zero-displacement conditions. The most distal vertebrae were considered the datum against which the movement of all other vertebrae was compared. The artificial spine was then dismantled for accuracy assessment in which the middle vertebrae (L5 and C4-C5) were moved relative to the superior (L4 and C3) and inferior (S1 and C6) vertebrae by known, incremental displacements on an imaging phantom device. Displacements occurred along the superior-inferior, anterior-posterior, and flexion-extension (rotational) axes of motion. RSA images were obtained at each displacement. Image analysis was performed using model-based software (RSACore v3.41, Leiden, Netherlands) to visualize implanted RSA beads in 3-dimensional space. Precision was defined as the 95% confidence interval of error in measuring zero-displacement. Accuracy was defined as the mean difference (with 95% confidence interval) between the known and measured displacement. Results: The rate of RSA bead detection was high with 5 to 8 implanted beads being visible in both the lumbar and cervical regions of the artificial and cadaveric spines. Translational RSA precision for both spines was better than 0.25 and 0.82 mm for the lumbar and cervical regions, respectively. Rotational precision was better than 0.40° and 1.9° for the lumbar and cervical regions, respectively. RSA accuracy for the artificial spine overall demonstrated less than 0.11 mm translational bias (margin < ±0.02 mm) and less than 0.22° rotational bias (margin < ±0.15°). Conclusion: This study demonstrates that RSA achieves sufficient precision and accuracy to detect intervertebral micromotion for the purpose of assessing arthrodesis. Well-dispersed RSA bead placement is critical to achieving sufficient accuracy as well as avoiding occlusion by metal hardware. Cervical bead implantation is particularly sensitive to bead clustering due to small vertebrae size and proximity to critical structures. The results of this work will aid in the development of a clinical study to assess arthrodesis in patients.

A250: Anterior Cervical Ossified Posterior Longitudinal Ligament En Bloc Resection: The Efficacy and Advantages of a Novel Surgical Technique for the Treatment of Cervical Ossification of the Posterior Longitudinal Ligament With Myelopathy

Xiongsheng Chen¹, Shengyuan Zhou¹, Yin Zhao¹, and Bo Yuan¹

¹Changzheng Hospital, Second Military Medical University, Shanghai, China

Introduction: Severe neurologic deficit caused by cervical ossification of the posterior longitudinal ligament (COPLL) requires surgical treatment. However, the surgical strategies to treat cervical myelopathy caused by COPLL remains controversial. The aim of this study was to evaluate the efficacy and advantages of anterior COPLL ligament en bloc resection (ACOE) in the treatment of COPLL. Material and Methods: From January 2014 to June 2016, postoperative outcomes of 43 COPLL patients underwent ACOE were compared with postoperative outcomes of 48 COPLL patients underwent anterior cervical ossified mass of posterior longitudinal ligament piecemeal resection (ACOP) from July 2009 to December 2013. The Japanese Orthopaedic Association (JOA) system was used to evaluate the neurologic outcomes of the 2 groups. Results: At the time of preoperation, 1-year postoperative, and the last follow-up, there was no significant difference in JOA scores between the 2 groups. While within 3 months postoperative, the JOA scores in ACOE group were higher than which in ACOP group. In terms of complications in ACOE group, hematoma was not identified, cerebrospinal fluid leakage (CSFL) in 1 case, C5 root palsy in 1 case, internal fixation loosening in 1 case, hoarseness in 1 case, and Horner syndrome in 1 case; in ACOP group, internal fixation loosening was not identified, CSFLs in 6 cases, C5 root palsy in 2 cases, hoarseness in 4 cases, hematoma in 1 case, and Horner syndrome in 2 cases. All complications were effectively recovered. Conclusion: Comparing with ACOP, the ACOE is more efficient and safer with better neurological improvement in a short time follow-up (<3 months). The incidence of CSFL could be significantly reduced by ACOE in the treatment of COPLL compared with ACOP.

A251: Machine Learning Can Identify Postoperative Complications Following Cervical Spine Surgery Using Medical Notes

Varun Arvind¹, Brian Cho¹, Daniel Ranti¹, Aly Valliani¹, Andrew Warburton¹, Deepak Kaji¹, Ivan Ye¹, Ray Tang¹, Zoe Cheung², Jun Kim², and Samuel Cho²

¹Icahn School of Medicine at Mount Sinai, New York, NY, USA

²Mount Sinai Hospital, New York, NY, USA

Introduction: Quality improvement studies have provided surgeons with systematic insight on how to make changes to their practice to improve patient outcomes and avoid complications. However, currently most automated methods to identify patient complications for quality improvement studies rely on prelabeled administrative codes sampling data from fixed electronic medical databases. Medical notes contain a rich supply of medical data, yet the format of unstructured text has precluded these data from use in large retrospective analyses. Therefore, the utilization of free-text from electronic medical records using natural language processing in combination with machine learning can provide an additional tool for patient complication surveillance and identification. The purpose of this study was to develop a machine learning algorithm that can identify patients with common spine surgery complications based on unstructured patient notes. Material and Methods: This was a retrospective analysis of medical notes from the Medical Information Mart for Intensive Care (MIMIC III), a large, single-center database composed of intensive care unit admissions data from 2001 to 2012. Spinal surgery complications include pseudoarthrosis, spinal cord injury, cardiac, respiratory, vascular, central nervous system, hematoma formation, hemorrhage, infection, cerebrospinal fluid leak, carotid or vertebral artery injury, hoarseness, and dysphagia complications as identified by ICD-9 codes. Area under the receiver operating characteristic curve (ROC) were evaluated to assess performance of the model. Natural language processing was used to analyze unstructured narrative notes including physician, consult, discharge, radiology, nursing, pharmacy, and social work notes. The notes were preprocessed using natural language processing methods and an auto-correct function was performed. A support vector machine (SVM) algorithm was used train a classifier to automatically identify patients with complicated cervical spine surgery based solely on the notes. The data was randomized with 70% used for training and 30% used for testing. The SVM algorithm identified surgical spine complication notes with an area under the receiver operating characteristic curve of 0.72. Results: A total of 435 patients from the orthopedic service were identified as undergoing cervical spine surgery (ICD-9: 8101, 8202, 8103), with 158 patients experiencing spinal surgery complication (36%). A total of 84% of patients were male with 96% of cases being emergent and 4% elective. In all, 86.7% of patients were Medicare insured, 10% privately insured, and the rest insured through government insurance, Medicaid, or self-pay. Conclusion: Combining natural language processing with machine learning is a valid approach for automatic identification of complications following cervical spine surgery using unstructured text from discharge notes. This method provides a robust survey of surgical encounters and describes a method that may be used for real-time surveillance of postoperative complication risk and for automated identification of patient complications for quality improvement studies.

A252: Use of Artificial Intelligence to Improve Surgical Referrals in Degenerative Lumbar Spinal Conditions

Nathan Xie¹, Peter Wilson², and Rajesh Reddy²

¹University of New South Wales, Sydney, New South Wales, Australia

²Prince of Wales Hospital, Sydney, New South Wales, Australia

Introduction: The majority of referrals for degenerative lumbar spinal conditions do not lead to surgical intervention. Stratifying referrals from primary care sources would not only expedite care for patients that may benefit from surgical intervention, but also allow for non-operative treatments/strategies to be implemented for those who are unlikely to benefit. Our study explored the clinical and imaging features associated with progression to surgery and allowed us to develop a prediction tool able to identify patients with a higher chance of surgical intervention being indicated. We developed an artificial intelligence machine learning (ML) model, able to learn from previous data to make accurate predictions regarding unseen data. This prediction tool could be used by secondary or tertiary spine care specialists to more effectively triage patients, as well as help reduce wait-lists in underresourced centers. Material and Methods: We identified 55 potential factors in the literature predicting surgical progression or outcome, including sociodemographic features, clinical findings, and findings on imaging. All patients presenting with an elective lumbar spine complaint between 2013 and 2018 at a single Australian tertiary hospital (n = 326) were then retrospectively reviewed via electronic and paper-based medical records, with data being collected for the potential predictive factors. An artificial neural network (ANN) was constructed in IBM SPSS Modeler v18.0, with the outcome being progression to spinal surgery (yes/no). The neural network was “trained” on the dataset, and then internally validated. To compare it with a traditional statistical model, a logistic regression (LR) model was created from the same data. The preliminary models were evaluated on their accuracy, discrimination (area under receiver operating characteristic [ROC] curve [AUC]), and calibration (Hosmer-Lemeshow test [HLT]) metrics. Results: Ten predictive variables were included as input in the final models. These included clinical features such as leg pain intensity, neurogenic claudication, dermatomal distribution of pain, functional status, previous treatment, as well as imaging findings. The ANN was able to predict surgical progression with 94.2% accuracy. It also exhibited excellent discriminative ability (AUC = 0.90), with good fit of the data (HLT > 0.05). This was superior when compared with the LR model (accuracy = 87.4%, AUC = 0.86, HLT >0.05). Conclusion: Both the neural network and regression models predicted surgical progression with a high degree of accuracy, although ANN was shown to be superior. By demonstrating that the operating patterns of single centers can be predicted with a high degree of accuracy, the potential for more appropriate referrals becomes possible, reducing wait-lists and increasing surgical conversion rates. Further research is required to ascertain external performance of these models, particularly in centers with a high degree of variability in surgical indications.

A253: Clinical Results of a Single Magnetically Controlled Growing Rod (MCGR) With a Contralateral Gliding System and Apical Control for Early-Onset Scoliosis: 3-Year Follow-up

Sebastiaan P. J. Wijdicks¹, Simon T. Skov², René M. Castelein¹, Haisheng Li², Moyo C Kruyt¹, and Cody Bünger²

¹University Medical Center Utrecht, Utrecht, Netherlands

²Aarhus University Hospital, Aarhus, Denmark

Introduction: Different growth friendly systems and implants have been developed to control the scoliosis deformity and allow for continuous spinal growth and thereby support the truncal development. Magnetically controlled growing rod (MCGRs) allow for noninvasive distraction by electromagnetic stimulation without sedation. Disadvantages are the high initial costs, lack of apical control, and high failure rates at the mid-term follow-up. To overcome these, we combined a single concave MCGR with a contralateral sliding rod with apical control. MCGR lengthenings were done every 3 months. The aim of this study was to investigate both the 3D correction, the spinal growth and the complication rate of this new hybrid growing-rod sliding-rod concept. Material and Methods: Eighteen children with early-onset scoliosis with an average 3-year follow-up (range 2.0-3.7) from 2 European spine centers were treated with the single MCGR hybrid concept. Median age at surgery was 9 years. Three-dimensional correction, spinal growth, and balance were measured on standard digital radiographs and adverse events were recorded from patient files. Results: The Cobb angle reduced from mean 59° preoperatively to 30° postoperatively and increased to 36° at latest follow-up. Rotation of the apical vertebra improved from mean 27° to 18° postoperatively but was partially lost to 24° during follow-up. Kyphosis increased 5° at last follow-up. Instrumented spine growth averaged 10 mm/year at last follow-up. There were 8 total complications, with 7 implant-related complications. Three patients required unplanned revision; 2 because of failure of distraction and 1 because of rod breakage. There were no infections. Conclusion: These mid-term results of a single MCGR hybrid construct show adequate initial 3D correction and spinal growth. The implant -related complication rate is relatively low compared with bilateral MCGR. Long-term follow-up is needed to fully investigate final spinal growth and 3D correction.

A254: Double Team for Double Crush! Clinicoradiological Efficacy of “Single Staged Double Team” Surgical Decompression for Tandem Spinal Stenosis in Elderly Population (>65 years)

Zahirabbas Merchant¹, Vishal Kundnani¹, Rahul Prakash¹, Ankit Patel¹, Mihir Upadhyaya¹, Neil Kire¹, and Sanyam Jain¹

¹Bombay Hospital & Medical Research Center, Mumbai, Maharashtra, India

Introduction: Safety and efficacy of single-stage decompression procedures in the elderly population (>65 years) have been under debate. Literature evaluating efficacy and safety of single-staged double team combined surgery in elderly population is scarce. Objective: To evaluate clinical outcome, safety, and efficacy of “single staged double team” combined surgery in cervical + lumbar tandem spinal stenosis (TSS) among the elderly population. Material and Methods: Retrospective, cohort study; independent observer. A total of 88 consecutive patients managed by single stage decompression for TSS (cervical laminectomy + lumbar decompression ± fusion–2 surgical team simultaneously operating). Evaluation: Demographic (age/gender/presentation symptoms/ASA grade), clinical–neurological (visual analogue scale [VAS] pain, Nuricks grade, modified Japanese Orthopaedic Association scale [mJOA], Oswestry Disability Index [ODI]), perioperative (blood loss, surgical time, hospital stay), and radiological parameters (instability/myelomalacia) were documented preoperatively, postoperatively, and at follow-up (2 years; mean 3.5 years). Complications (infection, dural leak, neurological, medical -cardiopulmonary/urinary). Hirabayashi recovery rate (HRR) and Odoms criteria used to ascertain the satisfaction rates. SSP (P < .05), regression analysis established for prognostication. Results: A toal of 88 patients (38 male, 50 female), mean age 69 years (range 65-89 years), mean follow-up 42 months (range 24-58 monnths) presented with neurological claudication = 53, cervical myelopathy = 16, combined = 19; mean delay of presentation 6 months (rang e3-24 months). Simultaneous double team single staged surgery (C3-6 cervical laminectomy (n = 88) + lumbar decompression (L4-5 = 49, L5-S1 = 28, L4-S1 = 11+ fusion = 16). Mean blood loss 300 mL (range 200-600 mL), mean operating time 135 minutes (range 80-250 minutes), mean hospital stay 6 days (range 5-17 days). Significant (P < .05) improvement noted in VAS (Pre 6.7, Post 3.4), ODI (Pre 72, Post 26), mJOA (pre 8.9, post 13.7), Nurick’s grade (Pre/post3.5/1.5)at final follow-up. Average HRR was 54.73 with satisfaction rates (80% excellent/good Odoms criteria) reversible C5 palsy (n = 2), neurological worsening (n = 1), dural tear (n = 1), and superficial infection (n = 1). Cardiopulmonary issues = 1 immediate postoperative managed medically. Two patients died due to unrelated reasons. Predominant lumbar symptoms and shorter duration of symptoms. Ambulatory status with no myelomalacia had better outcome. Longer operative time, higher blood loss, and ASA grade had no correlation with outcome. Conclusion: TSS among the elderly is common and often missed diagnosis with delayed presentation. Single staged double team lumbar + cervical surgery (same anesthesia) is safe and efficacious modality with minimal morbidity and optimal results in elderly population. Increasing age and prolonged duration of symptoms are associated with poor outcome. Preoperative ASA grades, blood loss, or long operative time had no correlation with outcome.

OP39: Minimally Invasive Spine Surgery 3

A255: Retrospective Review of Immediate Restoration of Lordosis in Single-Level Minimally Invasive Transforaminal Lumbar Interbody Fusion (MI-TLIF): A Comparison of Static and Expandable Interbody Cages

Sertac Kirnaz¹, Philip Saville², Avani Vaishnav², Steven McAnany², Rodrigo Navarro-Ramirez¹, Christoph Wipplinger¹, Franziska Schmidt¹, Catherine Himo Gang², Sheeraz Qureshi², and Roger Härtl¹

¹Weill Cornell Medicine, New York, NY, USA

²Hospital for Special Surgery, New York, NY, USA

Introduction: Sagittal alignment is becoming an increasingly important consideration in spine surgery. The literature is conflicted regarding the effect of minimally invasive transforaminal lumbar interbody fusion (MI-TLIF) on sagittal parameters and the role of expandable cage technology. We aimed to compare the lordosis generated by static and expandable cages in order to determine if there is a benefit of expandable cages in MI-TLIF and determine what factors affect segmental lordosis and foraminal decompression. Material and Methods: Preoperative regional lordosis (RL), segmental lordosis (SL), and posterior disc height (PDH) were compared with postoperative values in patients undergoing single-level MI-TLIF using expandable or static interbody cages. Patients were stratified based on preoperative SL: low lordosis (<15°), moderate lordosis (15°-25°), and high lordosis (>25°). Regression analyses were conducted to determine factors associated with postoperative SL and PDH. Results: A total of 171 patients were included; 111 in the static and 60 in the expandable cohorts. Patients with low preoperative lordosis experienced an increase in SL and maintained RL regardless of cage type. Those with moderate to high preoperative lordosis experienced a decrease in SL and RL with the static cage but maintained SL and RL with the expandable cage. Although both cohorts showed an increase in PDH, the increase in the expandable cohort was significantly greater. Preoperative SL was predictive of postoperative SL and preoperative SL, preoperative PDH and cage-type were predictive of postoperative PDH. Cage position was not related to postoperative SL or PDH. Conclusion: Expandable cages showed favorable results in restoring posterior disc height and maintaining lordosis in the immediate postoperative period. Cage position did not impact the restoration of lordosis or posterior disc height. Preoperative segmental lordosis was the most significant predictor of postoperative segmental lordosis. Thus, preoperative radiographic parameters and goals of the surgery should be important considerations in surgical planning.

A256: Risk Factors for Subsidence After Oblique Lumbar Interbody Fusion in Degenerative Lumbar Disorders

Vit Kotheeranurak¹, Jun-Sung Kim², Guang-Xun Lin², and Akaworn Mahatthanatrakul³

¹Queen Savang Vadhana Memorial Hospital, Sriracha, Chonburi, Thailand

²The Catholic University of Korea Seoul St. Mary’s Hospital, Seoul, Republic of Korea

³Naresuan University Hospital, Phisanulok, Thailand

Introduction: Subsidence of polyetheretherketone (PEEK) cage is considered as the most commonly potential perioperative complications after oblique lateral interbody fusion (OLIF). Risk factors of cage subsidence have not been published in the literature. We aimed to report the incidence and potential risk factors of PEEK cage subsidence following OLIF for lumbar degenerative diseases. Material and Methods: Retrospective cohort of 64 consecutive patients (22 males and 42 females, mean age was 66.5 years) who received either single- or multilevel OLIF between 2012 and 2016. The mean follow-up period was 34.2 months. All OLIF procedures were performed by the senior author using the OLIF25 Clydesdale Spinal System (Medtronic Sofamor Danek, Minneapolis, MN). The patients were classified into subsidence or nonsubsidence group. Subsidence of the PEEK cage was defined as any violation of either endplate more than 2 mm from the computed tomography (CT) scan in both sagittal and coronal views. We collected preoperative variables, that is, age, sex, body mass index (BMI), bone mineral density (BMD), smoking status, corticosteroid use, diagnosis, operative level, multifidus cross-sectional area, and multifidus atrophy. Cage-related variables (height and length) were documented. Univariate and multivariate logistic regression analysis were used to analyze the risk factors for subsidence. A chi-square test (χ²) was used, and the adjusted odd ratios (ORs) with 95% confidence intervals (CI) for each variable were expressed. Results: Of the 64 patients, 35 patients (54.7%) met the subsidence criteria. Mean BMI was 25.36 kg/m². Forty-one (64.1%) were male and 23 (35.9%) were female. Twenty patients (31.3%) had BMD below −2.5, 31 patients (48.4%) had BMD between −2.5 and −1.5, and 13 patients (20.3%) had BMD above −1.5. Tobacco use was found in 10 patients (15.6%) and 4 patients (6.3%) used corticosteroid. Diagnoses were degenerative scoliosis (31 patients, 48.4%), degenerative spondylolisthesis (31 patients, 48.4%), isthmic spondylolisthesis (1 patient, 1.6%), and recurrent herniated nucleus pulposus (HNP; 1 patient, 1.6%). Eight (12.5%) of the operations were single level and 56 (87.5%) were multilevel surgery. Mean multifidus cross-sectional area was 778.37 m², 3 patients (2.8%) had no atrophy, 43 patients (40.6%) had mild atrophy, 51 patients (48.1%) had moderate atrophy, and 9 patients (8.5%) had severe atrophy. For the PEEK cage characteristics, 18 levels used 8 mm, 58 levels used 10 mm, and 30 levels used 12 mm cage height. Eight levels used 40 mm, 72 levels used 45 mm, and 26 levels used 50 mm cage length. Multivariate logistic regression analysis showed risk factors contributed to PEEK cage subsidence were; age >60 years (OR = 3.14, P = .02), BMD < −2.5 (OR = 2.47, P = .029), cage height (OR = 1.83, P < .01), and severe multifidus muscle atrophy (OR = 4.97, P = .042). Visual analogue scale scores for back and leg pain reduction at 3 months were less in subsidence group, but not different at 6-month and 1-year follow-up. Oswestry Disability Index scores were not different. Conclusion: Risk factors for subsidence in oblique lateral interbody fusion were age >60 years, BMD < −2.5, cage height, and severe multifidus muscle atrophy. The patients who had subsidence had worse early (3 months) postoperative back and leg pain.

A257: A Prospective Randomized Study Comparing Minimally Invasive TLIF With Navigation Versus Open TLIF for Isthmic Spondylolisthesis

Xiaofeng Lian¹, Jianing Ding¹, and Jianguang Xu¹

¹Shanghai Sixth People’s Hospital, Shanghai, China

Introduction: With the assistance of image-guided intraoperative navigation, real-time and virtual imaging of the screws and their plotted trajectory, area of decompression and interbody space are visible in minimally invasive transforaminal lumbar interbody fusion (MIS-TLIF). However, it has been still uncertain whether this visible minimally invasive procedure brings extra benefit compare with open TLIF. Therefore, in this study we compared the surgical outcomes of MIS-TLIF assisted with intraoperative navigation and open transforaminal lumbar interbody fusion (O-TLIF) for isthmic spondylolisthesis (IS). Material and Methods: A total of 53 patients with IS requiring surgical treatment were randomly assigned to 2 groups treated using MIS-TLIF with navigation (group A, n = 26) and open TLIF (group B, n = 27). Both groups were followed up for an average of 25.6 months (range, 24-36 months). The clinical and radiographic outcomes were compared between the 2 groups. Surgical complications were analyzed in both groups. Results: The average operative time was significantly longer in group A than in group B (P < .001), but blood loss during surgery was significantly less in group A than in group B (P < .001). Spondylolisthesis, disk height, and focal lordosis were significantly improved postoperatively in both groups. Reduction of vertebra slippage was significantly better in group B than in group A (P < .001). Screw accuracy rate was 98.2% in group A, significantly more than in group B (90.4%) (P < .001). There was no obvious difference in clinical outcomes between the 2 groups, as assessed using the VAS, ODI and JOA scores. Hospital stay and immobilization were significantly shorter in group A than in group B. Incident of surgical complications were similar in both groups. Conclusion: For IS, MIS-TLIF with navigation can obtain comparable clinical and radiological outcomes as O-TLIF. As less invasive procedure, MIS-TLIF gets faster recovery after surgery. With navigation assistance, MIS-TLIF obtains more accurate screws placement compared with O-TLIF.

A258: Comparative Study Between Endoscopic Interlaminar Laminectomy and Tubular Laminectomy in the Treatment of Lumbar Spinal Stenosis

Angela Carrascosa Granada¹, Ralf Wagner², Willian Velazquez³, Andres Vargas¹, and Anwar Saab Mazzei¹

¹Department of Neurosurgery, Madrid, Spain

² Ligamenta Spine Center, Germany

³ Department of Neurosurgery, Salta, Argentina

Introduction: Open microscopic laminectomy has been the standard surgical method for degenerative spinal stenosis without instability until nowadays. Minimal invasive spine surgery (MIS) is gaining acceptance because of the minimal muscle damage, short recovery times, reduced blood loss, and optimal clinical results. In terms of MIS there are 2 surgical techniques available for treatment of lumbar spinal stenosis: tubular approach (TA) and endoscopic interlaminar approach (EIA). The canal cross-sectional area (CCSA) is an objective way to measure the severity of central spinal canal compression as well as postoperative decompression. The objective was to investigate the difference in the expansion ratio of CCSA in both techniques and the clinical improvement. Material and Methods: We performed a prospective, randomized, and double-blind trial, with a 1-year recruitment period. Inclusion criteria were neurogenic claudication, failure of conservative treatment in last 3 months, evidence of central lumbar stenosis in magnetic resonance images (MRI). Exclusion criteria were degenerative spondylolisthesis grade II or higher, degenerative scoliosis greater than 30°, significant axial back pain, no spinal surgical history, segmental instability on dynamic radiographs. The visual analog scale (VAS), Oswestry Disability Index (ODI), and Japanese Orthopaedic Association (JOA) were collected preoperative, at 1, 3, and 6 months after surgery in all the patients. The cross-sectional area of the canal was measured both preoperatively and postoperatively at axial cuts of T2-weighted MRI in all subjects. Radiological and clinical results were compared between groups. Results: A total of 19 patients were enrolled in the study. A total of 22 levels were decompressed. According to randomization, 10 patients were assigned to the tubular approach and 9 patients were assigned to the interlaminar endoscopic group. The mean age of the patients enrolled in the study was 76 years and the male/female ratio was 8:11. In the tubular approach, the mean surgical time was 120 minutes and the average estimated blood loss was 101 cm³. In the interlaminar endoscopic approach, the mean surgical time was 114 minutes and the average estimated blood loss was 20 cm³. No differences in surgical time between procedures (P = .967) and we found a significative difference in blood loss (P < .0001). Average hospital stay was 1 day without complications. VAS, ODI, and JOA scores in both group significative improved immediate postoperative and on follow-up (P < .005). However, in the endoscopic group we found a better improve at 6 months in ODI scale (P = .037). The difference in the measurements of the change in the surface area after decompression was found to be statistically significant (P < .05). In the tubular approach, CCSA improved 185%, at the interlaminar approach CCSA improved 242%. We did not find significant differences between the techniques (P = .496). Conclusion: The interlaminar endoscopic approach as a treatment in lumbar spinal stenosis is a safe and effective technique with less invasiveness than the tubular approach, offers a widened channel decompression as the tubular approach and offers a better clinical result at 6 months of follow-up with less intraoperative blood loss.

References

 1. Kamson S, Trescot AM, Sampson PD, Zhang Y. Full-endoscopic assisted lumbar decompressive surgery performed in an outpatient, ambulatory facility: report of 5 years of complications and risk factors. Pain Physician. 2017;20:E221-E231.

 2. Polikandriotis JA, Hudak EM, Perry MW. Minimally invasive surgery through endoscopic laminotomy and foraminotomy for the treatment of lumbar spinal stenosis. J Orthop. 2013;10:13-16.

 3. Skovrlj B, Gilligan J, Cutler HS, Qureshi SA. Minimally invasive procedures on the lumbar spine. World J Clin Cases. 2015;3:1-9.

 4. Wada K, Sairyo K, Sakai T, Yasui N. Minimally invasive endoscopic bilateral decompression with a unilateral approach (endo-BiDUA) for elderly patients with lumbar spinal canal stenosis. Minim Invasive Neurosurg. 2010;53:65-68.

 5. Komp M, Hahn P, Oezdemir S, et al. Bilateral spinal decompression of lumbar central stenosis with the full-endoscopic interlaminar versus microsurgical laminotomy technique: a prospective, randomized, controlled study. Pain Physician. 2015;18:61-70.

 6. Yang B, Chen R, Xie P, Liu B, Dong J, Rong L. Microendoscopic decompression via unilateral approach for lumbar spinal stenosis [in Chinese]. Zhongguo Xiu Fu Chong Jian Wai Ke Za Zhi. 2011;25:1158-1163.

 7. Pao JL, Chen WC, Chen PQ. Clinical outcomes of microendoscopic decompressive laminotomy for degenerative lumbar spinal stenosis. Eur Spine J. 2009;18:672-678.

 8. Celik SE, Celik S, Göksu K, Kara A, Ince I. Microdecompressive laminotomy with a 5-year follow-up period for severe lumbar spinal stenosis. J Spinal Disord Tech. 2010;23:229-235.

 9. Costa F, Sassi M, Cardia A, et al. Degenerative lumbar spinal stenosis: analysis of results in a series of 374 patients treated with unilateral laminotomy for bilateral microdecompression. J Neurosurg Spine. 2007;7:579-586.

10. Cavuşoğlu H, Kaya RA, Türkmenoglu ON, Tuncer C, Colak I, Aydin Y. Midterm outcome after unilateral approach for bilateral decompression of lumbar spinal stenosis: 5-year prospective study. Eur Spine J. 2007;16:2133-2142.

11. Alimi M, Hofstetter CP, Torres-Campa JM, et al. Unilateral tubular approach for bilateral laminotomy: effect on ipsilateral and contralateral buttock and leg pain. Eur Spine J. 2017;26:389-396.

12. Alimi M, Njoku I Jr, Cong GT, et al. Minimally invasive foraminotomy through tubular retractors via a contralateral approach in patients with unilateral radiculopathy. Neurosurgery. 2014;10(suppl 3):436-447.

A259: Ao Consensus On Nomenclature for Working Channel Endoscopic Spinal Procedures

Christoph Hofstetter¹, Vit Kotheeranurak², Saqib Hasan¹, Roger Härt³, and Jin-Sung Kim²

¹The University of Washington Seattle, Seattle, WA, USA

²The Catholic University of Korea, Seoul, Republic of Korea

³Weill Cornell Medical College, New York, NY, USA

Introduction: Working-channel endoscopic spinal procedures have undergone a rapid development during the past decade. Evolution in surgical instruments and innovation in techniques have expanded the types of spinal pathology that can be addressed with endoscopic spine surgery. However, this rapid development has led to a heterogeneous nomenclature and definition of working-channel endoscopic spine procedures. A lack of common language has hampered communication between endoscopic spine surgeons, patients, hospitals, and insurance carriers. The current report summarizes the nomenclature utilized for working-channel endoscopic procedures to treat cervical, thoracic and lumbar disc herniations as well as spinal stenosis. We propose a uniform system of nomenclature consisting of the type of visualization (endoscope), approach corridor (anterior, posterior, interlaminar, transforaminal), segment of the spinal column (cervical, thoracic, lumbar), and traditional name for the surgical procedure. Methods: A literature review was performed for publications pertaining to endoscopic spine surgery of the cervical, thoracic and lumbar spine. Articles were grouped pertaining to the specific approach and segment of the spinal column. Available literature was reviewed and articles with varying nomenclature for a specified endoscopic procedure were identified. The current available terminology for each procedure was reviewed by a panel of spine surgeons experienced in endoscopic spine surgery and a consensus nomenclature for each specific procedure was identified. Results: Among the 3455 articles found with the query “endoscopic spine,” a total of 192 articles were identified with varying terminology for endoscopic spine procedures of the cervical, thoracic, and lumbar spine. For anterior cervical endoscopic procedures, a total of 2 procedural names were identified. The proposed nomenclature agreed upon by the panel was “endoscopic anterior cervical discectomy.” For posterior cervical endoscopic procedures, a total of 6 procedural names were identified. The proposed nomenclature agreed upon by the panel was “endoscopic posterior cervical foraminotomy and discectomy.” For posterior thoracic endoscopic procedures, a total of 7 procedural names were identified. The proposed nomenclature agreed upon by the panel was for 2 procedures; “endoscopic unilateral cervical or thoracic laminotomy for bilateral decompression” and “endoscopic transforaminal thoracic discectomy.” For posterior lumbar endoscopic procedures, a total of 15 procedural names were identified. The proposed nomenclature agreed upon by the panel was for 5 procedures; “endoscopic transforaminal lumbar discectomy,” “endoscopic transforaminal lumbar foraminotomy,” “endoscopic interlaminar lumbar discectomy,” “endoscopic interlaminar lateral recess decompression,” and “endoscopic unilateral laminotomy for bilateral decompression.” Conclusions: It is critical to construct and define a consistent framework within endoscopic spine surgery to allow for more robust research and continued evolution in techniques. Hence, we believe that this framework begins with delineation of a universal nomenclature that can facilitate uniformity in both inter- and intradisciplinary spheres. Perhaps the most crucial function of standardization of nomenclature and techniques is to facilitate learning of these techniques to allow the field of spine surgery to concomitantly evolve with our technological advancements. The development and delineation of a uniform framework will ultimately facilitate a new generation of spine surgeons the ability to offer patients access to more targeted procedures while minimizing the collateral surgical trauma that has been an accepted component of traditional open spine procedures.

A260: A Mobile Outside-In Technique of Transforaminal Lumbar Endoscopy for Lumbar Disc Herniations

Hyeun-Sung Kim¹, Ravindra Singh¹, Sung Ho Choi², Nitin Maruti Adsul³, Seong-Hoon Oh⁴, Il Tae Jang¹, Sung Woon Oh¹, and Jung Hoon Noh¹

¹Nanoori Gangnam Hopsital, Seoul, Republic of Korea

²Nanoori Suwon Hopsital, Suwon, Republic of Korea

³Sir Ganga Ram Hospital, New Delhi, India

⁴Nanoori Incheon Hopsital, Incheon, Republic of Korea

Introduction: Percutaneous endoscopic transforaminal lumbar discectomy (PETLD) has now become a standard of care for the management of lumbar disc disease. There are 2 techniques for the introduction of a working cannula with regard to disc—outside-in and inside-out. The aim of this prospective study is to describe the technical aspects of a novel mobile outside-in method in dealing with different types of disc prolapse. Material and Methods: A total of 184 consecutive patients with unilateral lower limb radiculopathy due to lumbar disc prolapse were operated on with the mobile outside-in technique of PETLD. Their clinical outcomes were evaluated based on the type of disc prolapse they had, a visual analog scale (VAS) leg pain score, the Oswestry Disability Index (ODI), and the Macnab criteria. The completeness of the decompression was documented with a postoperative magnetic resonance imaging. Results: The mean age of the patients was 50 ± 16 years and the male/female ratio was 2:1. The mean follow-up was 19 ± 6 months. A total of 190 lumbar levels were operated on (L1-L2: n = 4, L2-L3: n = 17, L3-L4: n = 27, L4-L5: n = 123, and L5-S1: n = 19). Divided into types, the patient distribution was central: n = 14, paracentral: n = 74, foraminal: n = 28, far lateral: n = 13, superior-migrated: n = 8, inferior migrated: n = 38, and high canal compromise: n = 9. The mean operative time was 35 ± 12 minutes (range 25-56 minutes) and the mean hospital stay was 1.2 ± 0.5 days (range 1-3 days). The VAS score for leg pain improved from 7.5 ± 1 to 1.7 ± 0.9. The ODI improved from 70 ± 8.3 to 23 ± 5. According to the Macnab criteria, 75 patients (40.8%) had excellent results, 104 patients (56.5%) had good results, and 5 patients (2.7%) had fair results. Recurrence (including early and late) was seen in 15 out of the 190 levels that were operated on (7.89%). Conclusion: This article presents a novel outside-in approach that relies on a precise landing within the foramen in a mobile manner and does not solely depend on the enlargement of the foramen. It is more versatile in application and useful in the management of all types of disc prolapse, even in severe canal compromise and high migration.

A261: Endoscopic Electrothermic Procedure of the Sacroiliac Joint (ESIJ): 24-Month Follow-up Studies

Richard Ibrahim¹

¹Schmerztherapiezentrum Riem, Munich, Germany

Introduction: Chronic low back pain (CLBP) that lasts for more than 6 months is estimated to occur in 60% to 80% of the general population in their lifetime and is associated with substantial health care costs. The sacroiliac joint (SIJ) complex is one of the major sources of chronic low back pain, accounting for around 10% to 33% of the total number of CLBP cases. The SIJ complex consists of the joint capsule, synovia, various muscles, and ligamentous structures overlying the joint and neuronal structures that innervate the SIJ. Current interventional and surgical treatment options for SIJ pain perception complex include intraarticular or peri-articular injection, radio frequency ablation (RFA), and SIJ fusion. The described interventional procedures are simple procedures and provides quick pain relief, but the effect is short-lived. In addition, SIJ fusion is an invasive surgical procedure that should be reserved for refractory intractable pain of the SIJ. New endoscopic electrothermic procedures of the SIJ capsula, synovial, ligament, and neuronal structures has been utilized in the treatment of facetogenic CLBP in a few clinical reports with favorable results, but to our knowledge, the efficacy of this technique when applied to SIJ-associated CLBP has not been reported. In this study, we utilized endoscopy for the precise microsurgical and ablation (ESIJ) of the potential pain generators associated with the SIJ and evaluated the clinical efficacy of this new technique. Material and Methods: The institutional review board at our institution approved this study. The medical records of 30 consecutive patients who underwent ESIJ for SIJ arthropathy and pain in CLBP between January 2016 and February 2018 were reviewed. The inclusion criteria for this treatment were as follows: patient with predominant back pain (CLBP) and signs and symptoms of SIJ involvement on physical examination and radiological tests (CT scan or MRI), unresponsive to conservative therapy including oral analgesics and physical or osteopathic therapy, persistent CLBP despite previous lumbosacral surgery or/and interventional pain treatment in 24 months follow-up. SIJ complex pathology as the main cause of CLBP is difficult to diagnose due to overlapping patterns with other sources of CLBP and varying patterns of pain between individuals. Radiographic findings of arthropathy or erosion of the SIJ, while not specific, may suggest SIJ-based pathology in patients with clinical suspicion. While numerous physical examination methods have been suggested, provocative tests were reported to have more reliability in numerous reports. In order to confirm the SIJ pain as the main source of CLBP, 3 separate intra-articular SIJ and medial branch blocks of the lower facet joints (L4-S1) were performed under C-arm control at least 3 weeks before the ESIJ procedure. If patient experienced 50% or higher improvement less than 2 weeks in pain from baseline according to visual analogue scale (VAS) after this block, SIJ complex was considered to be the main pain generator, and ESIJ was scheduled. Patient with blockades, irritation, tumors of the SIJ, and nerve root neuropathy or rheumatic diseases or other severe comorbid diseases and medical conditions were excluded. All patients were followed for a minimum of 24 months after the procedure ESIJ in outpatient pain clinics. ESIJ was performed in the operating theatre room under full anesthesia. Patients were discharged 3 days after the procedure. All patients were followed up at the outpatient clinic at 3, 6, 9, 12, 15, and 24 months after the procedure and annually thereafter by Oswestry Disability Index (ODI) and VAS scores. Results: More than 350 endoscopic procedures of SIJ (ESIJ) were performed from January 2016 till February 2018. 30 patients with SIG arthropathy in CLBP over 6 months were included by precise criteria. At baseline, the VAS was 7.23, ODI 44.8, and age of 56 years (11 male, 19 female). After the ESIJ procedure, the patient had an significant improvement in pain relief (measured by VAS) and functional ability (measured by ODI) at 3 to 24 months: VAS reduction from 7.23 baseline to 1.72 and 2.82 after 21 and 24-months, respectively, significantly by 80% pain relief ODI ability improvement from 44.8 baseline to 22.2 significantly in lumbar function. Conclusion: This new easy-to-perform endoscopic electrothermic procedure of the SIJ (ESIJ) shows good to excellent results in SIJ arthropathy in relation to pain relief and functional capacity in long-term analysis. Further research is needed to confirm these results.

A262: Minimally Invasive Percutaneous Endoscopic Surgery Combined With Posterior Fixation in the Elderly With Spinal Tuberculosis

Nanjian Xu¹ and Weihu Ma¹

¹Sixth Hospital of Ningbo, Ningbo, China

Introduction: The treatment of the elderly with spinal tuberculosis remains a challenge for spinal surgeons because of the multiple existing comorbidities. Conventional open surgery is too invasive for some elderly patients. The diagnostic and therapeutic value of posterolateral endoscopic debridement and irrigation has been proved in some recent studies.The purpose of this study is to evaluate the efficacy of minimally invasive percutaneous endoscopic surgery combined with posterior percutaneous pedicle screw fixation in the elderly with spinal tuberculosis, which may traditionally require open surgery. Material and Methods: From March 2015 to July 2016, 19 elderly patients with diagnosis of spinal tuberculosis were treated with minimally invasive percutaneous endoscopic surgery combined with posterior percutaneous pedicle screw fixation, there were 13 males and 6 females, aged from 60 to 88 years (average, 70.4 ± 6.5 years old). All patients were given appropriate chemotherapy for 3 weeks preoperatively.All patients were followed up with evaluation of the changes of local kyphotic angle.Visual analogue score (VAS) and Oswestry Disability Index (ODI) was used to evaluate the clinical effects. C-reactive protein (CRP) and erythrocyte sedimentation rate (ESR) of every patient preoperatively and postoperatively were analyzed.All the above parameters were compared between preoperation and final follow-up. Results: Surgery was accomplished successfully in all cases.The mean operation time was 162.7 ± 15.3 minutes. The follow-up had proceeded for 18 to 28 months (average, 20.4 ± 4.2 months). Average VAS for pain was 6.3 ± 1.4 before surgery and 1.6 ± 0.7 at the time of final follow-up. The ODI score decreased from 81.4% ± 4.8% to 65.3% ± 5.2% at the time of final follow-up compared with preoperation. Average of CRP was 57.8 ± 9.8 mg/L before surgery and 22.5 ± 4.2 mg/L at final follow-up. The ESR decreased from 76.8 ± 13.6 to 33.2 ± 7.9 mm/h at the time of final follow-up compared with preoperation. There was statistically difference in kyphotic angle at final follow-up compared to preoperatively. Conclusion: Minimally invasive percutaneous endoscopic surgery combined with posterior percutaneous pedicle screw fixation brought immediate pain reduction and good clinical results to the elderly who had comorbid disease and had spinal tuberculosis.

A263: Endoscopic Treatment for Thoracic Cord Compression Due to Ossification of Ligamentum Flavum

Yong-Sang Kim¹

¹Barun Hospital, Jinju-Si, Republic of Korea

Introduction: We described the new technique and reviewed clinical and radiological outcomes of thoracic myelopathy caused by ossification of ligamentum flavum (OLF) treated with minimally invasive decompression using unilateral biportal endoscopy (UBE). Material and Methods: A retrospective study of 33 consecutive patients, 9 males and 24 females was performed decompression using biportal endoscopy. From August 2016 to June 2018, diagnosis in each case was established using computed tomography (CT) and magnetic resonance imaging (MRI). A comparison between the pre- and postoperative neurological status was evaluated using Japanese Orthopedic Association (JOA) scoring system. Pre- and postoperative pain is also compared by visual analogue scale (VAS). Results: The mean follow-up period was 5.2 months (range: 0.5-21.3 months). The average age was 72 years (range: 51-82 years). The mean operation time was 71 minutes (range: 50-100 minutes). The mean preoperative JOA score was 6.2 and postoperative JOA score improved up to 9.2 during the follow-up period. The mean immediate postoperative VAS (4 hours after the operation) was 3.1. The mean blood loss was 61.8 cm³ (range: 50-90 cm³). There was no operation-related complication and transfusion. There was no evidence of postoperative instability or kyphosis in the follow-up examination. Conclusion: UBE for thoracic myelopathy caused by OLF is minimally invasive, a nonfusion technique that provides an effective and safe neural decompression.

OP40: Imaging 2

A264: The Relationship Between L5/S1 Disc Degeneration and Facet Joints: A Kinetic MRI Study

Ian Buchanan¹, Haiyin Lee², Ishan Shah¹, Changqing Li², Zorica Buser¹, and Jeffrey Wang¹

¹University of Southern California, Los Angeles, CA, USA

²Xinqiao Hospital, The Third Military Medical University, Chongqing, China

Introduction: The L5/S1 intervertebral disc (IVD) and vertebrae are unique components of the lumbar spine. As part of the lumbosacral region, the L5/S1 IVD marks the transition from the flexible lumbar segment to the fixed sacrum. The radiological factors associated with L5/S1 IVD degeneration are poorly understood. Assessing intervertebral disc pathology radiologically is important in determining the severity of disease. Not much is known about the association of facet joint status and L5/S1 IVD degeneration. Kinetic magnetic resonance imaging (MRI) is an effective method to evaluate soft tissues and ossified structures together, and measure radiological factors in a weightbearing position. In this study, we aimed to explore the relationship between age, gender, facet joint status, segmental motion, and degeneration of the L5/S1 IVD. Material and Methods: Forty-nine patients exhibiting varying levels of lumbar disc generation were included. Lumbar kinetic MRI images were evaluated in flexion, neutral, and extension positions. Angular motion, translational motion, facet joint angle, facet joint space width, and existence of facet joint fluid were measured. Lumbar disc degeneration was assessed using the Pfirrmann grade. Age, facet joint space width, and facet joint angle were used as factors for stratification. For age, involved patients were stratified into 2 groups using age 50 years as a cutoff, and the risk factors for lumbar disc degeneration were analyzed in each group. Facet joint space width was stratified as <2 and ≥2 mm. Facet joint angle was stratified as <90° and ≥90°. All the data was analyzed using 1-way analysis of variance (ANOVA) and logistic regression was used for final determination. Results: Of the 49 patients included in this study, 26 had mild disc degeneration and 23 had severe disc degeneration. There were no significant differences in gender, age, angular motion, translational motion, and the existence of facet joint fluid between these 2 groups. Logistic regression analysis revealed no relationship between facet joint angle and L5/S1 IVD degeneration (P = .104, OR = 0.928-1.007); however, the average width of facet joint was inversely correlated with L5/S1 IVD degeneration (P = .041, OR = 0.234-0.969). Facet joint space width was an independent risk factor for the degeneration of the L5/S1 IVD in patients older than 50 years. No other risk factors were found within each stratified group. Conclusion: We found that the average width of the facet joint space in the severely degenerated group was significantly narrower than in the mildly degenerative group. This indicates that a narrow facet joint space width may be an independent risk factor for the degeneration of the L5/S1 IVD.

A265: Can Cervical Spine Lateral Radiograph Replace Whole-Spine Lateral Radiograph as an Evaluating Tool for Cervical Sagittal Alignment?

Dong-Ho Lee¹, Dong Gyun Kim², Chang ju Hwang¹, Jaewoo Park¹, and Sehan Park¹

¹Seoul Asan Medical Center, Seoul, Republic of Korea

²Bumin Hospital Haeundae, Busan, Republic of Korea

Introduction: Standing whole-spine lateral radiograph (WLR) is commonly used as a tool for evaluating cervical sagittal balance. On this, however, the T1-slope, which is known to be one of the key factors influencing cervical sagittal balance, is frequently obscured by the shoulder and thoracic trunk. Furthermore, hands-on-clavicle position, which is recommended when patients are taken WLR, would affect the T1-slope, head position, and cervical lordosis. In this study, we investigate the differences in various radiographic parameters measured on standing cervical lateral radiographs (CLR) with arms extended and hands on either side between WLR with hands-on-clavicle position. In addition, we assessed whether CLR could be utilized for the evaluation of cervical sagittal alignment instead of WLR. Material and Methods: We retrospectively reviewed radiographic data of 60 consecutive adult patients (male:female = 38:22, mean age = 55.6 ± 1.3 years) who visited our clinic for neck pain only (without radiculopathy or myelopathy). Every patient was taken standing CLR with arms extended and hands on either side, and then WLR with hands-on-clavicle position. On CLR and WLR, we measured and analyzed the following parameters, respectively: (1) visibility of T1 slope, (2) T1 slope, (3) C7 slope, (4) C0-C2 Cobb’s angle (CA_C0-C2), (5) C2-C7 Cobb’s angle (CA_C2-C7), and (6) cervical sagittal vertical axis (cSVA). Results: Visibility of T1 slope was significantly lower in WLR compared with CLR (28% vs 83%, P = .049). Mean CA_C2-C7 on WLR with hands-on-clavicle position was significantly less lordotic than that on CLR with arms extended (11.2° ± 9.2° vs 14.3° ± 11.3°; P = .01). Mean cSVA was more translated posteriorly on WLR than on CLR (9.9 ± 18.9 vs 15.0 ± 13.4 mm, P = .04). However, there was no significant difference in T1 slope, C7 slope and CA_C0-C2 between CLR and WLR. Conclusion: Whole-spine lateral radiograph is a key image to evaluate global sagittal alignment of entire spinal column. However, hands-on-clavicle position for better visualization of thoracolumbar spine could distort true sagittal alignment of the cervical spine. Furthermore, this position would probably obscure cervico-thoracic junction more on plain radiograph, which causes difficulty in evaluating T1 slope. This study manifested that standing CLR could provide better visualization of the upper endplate of T1 without any difference in T1 slope compared with standing WLR. Therefore, it would not be necessary to take WLR repetitively for the evaluation of cervical sagittal alignment unless there is accompanying lesions in thoracolumbar spine on the initial examination.

A266: Spinal CT to MR Image Synthesis Using Generative Adversarial Networks

In Ho Han¹, Jung Hwan Lee¹, Seongsu Joo², Chung-Bin Jin³, and Hakil Kim³

¹Pusan National University Hospital, Busan, Republic ofKorea

²Team Elysium Inc, Seoul, Republic of Korea

³Inha University, Incheon, Republic of Korea

Introduction: Artificial intelligence (AI) and deep learning has been widely applied to medical image analysis. But, application of deep learning has focused on lesion classification of medical images using convolutional neural network. Recently, GANs (generative adversarial networks), that is a type of deep learning algorithm for generating image, has been introduced, and there were some reports to apply GANs to medical imaging. We applied GANs to generate spinal magnetic resonance (MR) image from spinal computed tomography (CT) scans, and qualitatively and quantitatively evaluated the similarity of synthesized MR images to real MR images. Material and Methods: GANs were trained to transform spinal CT image slices into spinal magnetic resonance T2-weighted axial (MRT2) image slices, combining adversarial loss and voxel-wise loss. The experiments were analyzed using 280 paired of lumbar spinal CT scans and MRT2 images. Then, MRT2 images were synthesized from 15 other spinal CT scans. In order to evaluate the reality of synthetic MR images, 2 radiologists, 2 spine surgeons, and 2 residents blindly classified the real and synthetic MRT2 images. Second, 2 experienced radiologist evaluated subdividing similarity of real and synthetic MRT2 images. In addition, quantitative analysis of synthetic MRT2 images was performed using the mean absolute error (MAE) and peak signal-to-noise ratio (PSNR). Results: The degree of similarity of synthetic MRT2 images evaluated by radiologists was 60%-90%. In blind classification of real MRT2 images, the failure rates were from 0% to 40%. The MAE value of each images were ranged from 13.75 to 34.24 pixels (mean: 21.19), and PSNR of each images were ranged from 61.96 to 68.16 dB (mean: 64.92). Conclusion: This is first study applying GANs to synthesize spinal MR image from CT. In spite of small dataset of 280, synthetic MR images were relatively well implemented. Synthesis of medical imaging using GANs is a new paradigm of AI application in medical imaging. We expect that spinal CT to MR image synthesis using GANs may replace real MR or improve diagnostic usefulness of CT. However, for clinical applications, further studies with various pathologies and large dataset are needed.

A267: AOSpine Global Survey: Trends in Utilization of MRI for Spinal Trauma and Spinal Cord Injury Across AO Regions

Saman Shabani¹, Ha Son Nguyen², Mayank Kaushal¹, Hesham Soliman¹, Bizhan Aarabhi³, Mark Kotter⁴, Michael Fehlings⁵, Brian Kwon⁶, James Harrop⁷, and Shekar Kurpad¹

¹Medical College of Wisconsin, Milwaukee, WI, USA

²California Institute of Neuroscience, Thousand Oaks, CA, USA

³University of Maryland School of Medicine, Baltimore, MD, USA

⁴University of Cambridge, Cambridge, UK

⁵University of Toronto, Toronto, Ontario, Canada

⁶University of British Columbia, Vancouver, British Columbia, Canada

⁷Thomas Jefferson University, Philadelphia, PA, USA

Introduction: Spinal cord injury (SCI) can be a devastating condition that can cause profound neurological deficit. In the past decade, MRI has been used to better evaluate the cord and soft tissues. However, the usage of magnetic resonance imaging (MRI) by surgeons located in different global AO regions is unclear. This AO-funded study looked at various scenarios where clinicians from different AO regions obtained MRI for evaluation of patients with SCI. Material and Methods: A survey was conducted across the 6 global AO regions. The survey included questions on the utilization of MRI in different clinical scenarios, including suspected level of spinal injury, presence and extent of neurological deficits, and findings from other imaging modalities. Responses were tabulated in terms of frequencies and percentages. Pairwise comparison of responses between AO regions were performed. Results: In a neurologically intact patient, the predominant imaging modality for all AO regions was computed tomography (CT) (range: 40.3%-62%); pairwise comparisons revealed only a significant difference between Latin America and Asia Pacific. For a neurologically impaired patient, the predominant choice for all AO regions was CT + MRI (range: 47.9%-66%); pairwise comparisons revealed significant differences involving Latin America versus (Asia Pacific, Europe, and Middle East) and Asia Pacific versus (Latin America and North America). The need for MRI for presurgical planning with CT already obtained was markedly different across AO regions; ranged from 20.8% to 47.2% for “absolutely essential” and 0% to 8.3% for “totally unnecessary.” In a patient with incomplete SCI who presented within 4 hours and had CT, the predominant choices for all AO regions were “take patient to the operating room (OR) right away and forgo MRI” (range: 42.7%-74%) and “obtain MRI as long as you could get patient to the OR [operating room] within 8 hours” (range: 10%-33%). North America respondents were most likely to forgo MRI and proceed to the OR. Similarly, in a patient with complete SCI, the predominant option for all AO regions was taking the patient to the OR right away without waiting for MRI, but the range was lower (28.3%-62%), while the option “obtain MRI, because time to decompression is not important” became more prominent (range: 8%-26.6%). Pairwise comparisons noted significant differences between Middle East and Asia Pacific with both exhibiting differences compare to North America, Latin America, and Europe for complete and incomplete SCI. Middle East and Asia Pacific were the 2 least likely to forgo MRI. Most AO regions obtained postoperative MRI only if there was a new deficit (range: 41%-66%). In Africa, 0% obtained a routine postoperative MRI, while 23.5% never obtained an MRI postoperative even if a new deficit(s) was present. Responses from Europe were significantly different from all other AO regions. Conclusion: Decisions regarding the use for imaging modality across various AO regions appears to be heavily influenced by presence of neurological deficits. Further study is needed to determine whether the accessibility to MRI would change the surgeons’ attitude toward obtaining MRI in patients with SCI.

A268: Tridimensional Reconstruction Using Stereoradiography for Evaluating In-Brace Patients With Adolescent Idiopathic Scoliosis (AIS): A Reproducibility Study

Haidara Almansour¹, Wojciech Pepke¹, and Michael Akbar¹

¹Heidelberg University Hospital, Heidelberg, Germany

Introduction: Bracing constitute the mainstay treatment for mild scoliosis. Tridimensional (3D) reconstruction of the spine using biplanar stereoradiographs is increasingly being utilized to assess the success of bracing. This study aims at determining the inter- and intraobserver reproducibility of the 3D spinal models of in-brace patients with adolescent idiopathic scoliosis (AIS). Material and Methods: Full-spine radiographs from 45 consecutive patients who underwent bracing treatment were retrospectively analyzed. Three-dimensional reconstruction was performed using validated software by 2 well-experienced readers. Subsequently, one of the readers repeated the measurements following a 2-week interval. Both readers were blinded to patients’ data and conducted the 3D reconstruction in a randomized order. The contour of each vertebrae (T1-L5) was identified and adjusted. Rotation of every vertebra was computed in the coronal, sagittal, and axial planes. Sagittal and pelvic parameters were evaluated. All measurements were statistically compared to determine agreement of their measurements of in-brace correction using the intraclass correlation coefficient (ICC; 2-way random-effect model). ICC was considered very good, good, moderate, or poor when ICC was, respectively, greater than 0.91, between 0.71 and 0.91, between 0.51 and 0.70, or less than 0.51. Finally, Bland-Altman plots were constructed for parameters, which revealed lower reliability. Results: Both inter- and intraobserver reliabilities for Cobb angle were excellent (ICC >0.95). Furthermore, ICC was >0.95 for T1/T12 kyphosis, T4/T12 kyphosis, L1/L5 lordosis, L1/S1 lordosis, pelvic incidence, pelvic tilt (PT) and for sacral slope (SS). Similarly, ICC was >0.91 for all sagittal and pelvic parameters. ICC for both coronal and sagittal vertebral rotation was 0.91. ICC ranged between 0.51 and 0.98 for axial vertebral rotation. Conclusion: Tridimensional analysis was reliable for Cobb, sagittal parameters, pelvic parameters, coronal, and sagittal rotation. In other words, bracing does not seem to affect the reproducibility of the aforementioned parameters. However, measuring axial vertebral rotation was more challenging especially in upper and middle thoracic spine (T1-T8). Localization and delineation of the important anatomical landmarks, that is, pedicles and spinous processes in 3D seems to be more complicated evident by the relatively lower reliability. Hence, caution is advised when interpreting axial rotation data in the region of the upper and middle thoracic spine.

A269: Machine Learning Automates Cobb Angle Measurements in Scoliosis

Brian Cho¹, Deepak Kaji¹, Aly Valliani¹, Jun Kim¹, Ivan Ye¹, Ray Tang¹, Amy Ahn¹, Oscar Carrillo¹, Varun Arvind¹, Thomas Kroshus¹, Andrew Warburton¹, Daniel Ranti¹, Zoe Cheung¹, and Samuel Cho¹

¹Icahn School of Medicine at Mount Sinai, New York NY, USA

Introduction: Cobb angle measurement is currently the gold standard for the evaluation and surgical planning of adolescent idiopathic scoliosis (AIS) from radiographs. However, this manual process is both time consuming and rater dependent, warranting the development of automated tools for diagnosis. Previous work in computer-aided measurement of Cobb angles have relied on manual identification of bone landmarks, which prevents robust spinal curvature assessment at scale. Segmentation using neural networks is a promising technique for automatically identifying bony structures in medical images. The purpose of this study was to develop a novel fully automated machine learning pipeline that reliably measures Cobb angles from radiographic images. Material and Methods: We obtained 780 lateral lumbar radiographs from the Department of Orthopedics at Mount Sinai Hospital. The data was randomized with 80% of the data allocated for training and 20% for testing. Radiographs in the training data were horizontally flipped, randomly rotated, and cropped to account for natural variation in patient posture and orientation during routine scans. The final augmented dataset of 15 096 images was used to train a U-Net neural network architecture to automatically segment vertebral bodies. Custom computer vision software was developed to measure the lumbar-lordosis Cobb angle from the resulting segmentation. Cobb angles for 50 randomly selected radiographs from the test data were manually measured twice by an attending spine surgeon, fellow, and chief resident to evaluate the algorithm’s performance. Results: Our algorithm achieved overall median raw angle difference of 2.71° ± 21.23° at an average rate of 0.11 seconds/measurement. The median raw angle difference was 5.14° ± 20.85°, 2.60° ± 4.27°, and 3.70° ± 4.28° for the algorithm, fellow, and chief resident, respectively, relative to the attending. Intraclass correlation coefficient (ICC) among surgeons was 0.956 [95% CI: 0.928, 0.974]. Conclusion: Our work demonstrates the efficacy of machine learning to automatically calculate Cobb angles for the identification of AIS from radiographs. The incorporation of this approach into clinical workflows has the potential to significantly reduce interrater variability by standardizing measurements and decrease costs by reducing physician workload, thereby enhancing the quality of clinical care delivered to patients.

A270: Radiographic Evaluation of Minimally Invasive Instrumentation and Fusion for Treating Unstable Spinal Column Injuries

Daniel Cavanaugh¹, Tristan Weir¹, Farooq Usmani¹, Jael Camacho¹, Kim Joanna¹, Mark Shasti¹, Vishal Khatri¹, Louis Bivona¹, Eugene Koh¹, Kelley Banagan¹, Daniel Gelb¹, and Steven Ludwig¹

¹University of Maryland Medical Center, Baltimore, MD, USA

Introduction: Although open pedicle screw fixation and fusion has been the traditional treatment modality for traumatic thoracolumbar (TL) fractures, they are often associated with many perioperative complications. Recent studies have demonstrated better outcomes with minimally invasive percutaneous techniques, though long-term results are still lacking. Facet fusion has shown to be effective in management of degenerative spondylolisthesis, but the role of facet fusion in percutaneous fixation has not been evaluated. Facet fusion in percutaneous fixation may reduce morbidity and promote long-term construct stability. Our objective is to evaluate maintenance of correction after percutaneous fixation with or without facet fusion and compare outcomes. Material and Methods: A retrospective analysis of was performed to identify TL trauma patients who underwent percutaneous fixation with and without facet fusion between 2006 and 2013. Patients were stratified into 2 major groups: with facet fusion and without facet fusion. To evaluate for progressive kyphosis and loss of correction, the Cobb angles were measured on all initial postoperative radiographs at short- and long-term follow-up. Evidence of >2 mm of radiolucency around any screw was also examined to evaluate for instrumentation loosening and failure. In cases where instrumentation was removed, upright radiographs were examined and a lateral Cobb angle was calculated for comparison. Results: Among the 80 cases studied, 24 patients were with and 56 without facet fusion. The two groups were similar with respect to gender, age, Injury Severity Scores, Glasgow Coma Scale, mechanism of injury, neurological status, and fracture morphology. Long-term data (>1 year) was available for a total of 16 patients; 9 with and 7 without facet fusion. The initial kyphosis angles for patients with facet fusion was 4.1° (SE 2.5°) and those without facet fusion was 10.6° (SE 1.9°) (P = .0082). At short-term follow-up, the angles were 7.7° (SE 3.2°) and 13.4° (SE 2.0°) for those with and without facet fusion (P < .0001). Among the long-term follow-up group, there was kyphosis progression from initial to short- to long-term follow up was not significant in both fusion and nonfusion groups. Screw loosening rate was 20.5% (n = 20) at short-term follow-up and 68.8% (n = 11) in the long-term follow up group. The rate of screw loosening in patients with and without facet fusion was 29.2% and 23.2%, respectively (P = .58). With long-term follow-up, the rate of screw loosening was 77.8% and 57.1% for those with and without facet fusion. Fifteen patients that belonged to the facet fusion group underwent instrumentation removal. Conclusion: This retrospective radiographic review of unstable thoracolumbar fractures that underwent percutaneous pedicle screw fixation with and without facet fusion demonstrated statistically significant, but not clinically significant, difference between groups that underwent facet fusion and those that did not. Our results revealed that radiographic screw loosening is a common finding that seems to occur more frequently in regions of greater spinal mobility. However, this finding did not correlate with loss of correction or progression of deformity. Furthermore, instrumentation loosening was found to increase with longer follow-ups.

A271: The Utility of Postoperative Radiographs Following Surgical Treatment of Traumatic Cervical Spine Injuries

Vishal Khatri¹, Alysa Nash¹, Noah Lessing², Farooq Usmani¹, Jael Camacho¹, Eugene Koh¹, Kelley Banagan¹, Daniel Gelb¹, and Steven Ludwig¹

¹University of Maryland Medical Center, Baltimore, MD, USA

²School of Medicine, University of Maryland, Balltimore, MD, USA

Introduction: Postoperative radiographs are routinely obtained following surgical fixation of traumatic cervical spine injuries. Given that pseudoarthrosis and failure of fixation are rare, there is increased concern about the utility of routinely obtaining imaging following stabilization of cervical spine fractures in the absence of new clinical or neurological findings. Multiple studies have concluded that routine postoperative imaging is not necessary but none have evaluated the trauma population. The purpose of this study is to assess the utility of postoperative radiographs following surgical intervention for cervical spine trauma, using postoperative complications and need for revision surgery as marker for potential imaging benefit. Material and Methods: A retrospective chart review of the patients who underwent surgical treatment of traumatic cervical spine injuries over an eight-year period at a level I trauma center was performed. Patient were included if they had at least 1 level fused due to trauma and were between ages 18 and 100 years. At discharge, postoperative upright anteroposterior and lateral radiographs were obtained and reviewed. Patients who required revision surgery were identified and further analyzed to identify if the results of follow up radiographs led to a revision surgery. Clinical notes were evaluated for (1) new radicular or axial neck pain, (2) dysphagia or dysphonia, (3) surgical site infection, (4) mention of hardware related complication including fracture, loosening or pseudoarthrosis. Results: A total of 295 patients were identified that had a cervical spine injury requiring operative stabilization. There were 177 males and 118 females with a mean age of 57.4 years (range, 18-94 years). The mean follow up period was 284 days. All had routine postoperative upright x-rays and imaging at follow-up visits. Ten patients required subsequent surgeries: 5 for wound complications, 2 for neurological decline requiring decompression, 1 for recurrent subluxation, and 2 for nonunion/malunion. Four other patients had broken/loose hardware that were managed conservatively. Advanced imaging (computed tomography/magnetic resonance imaging) was ordered in 14 patients, 10 of whom required additional surgery. In the cases requiring additional surgery, the postoperative radiographs added no value in guiding treatment in the absence of clinical findings. Every patient that underwent additional surgery had clinical findings (ie, wound breakdown, new-onset radiculopathy or continued pain) that were of concern which prompted the advanced imaging. Cervical radiographs alone added an additional $545 868 of health care cost during the study period. Conclusion: Routine postoperative radiographs in cervical trauma have minimal utility in guiding treatment course following surgical stabilization with instrumented spinal fusion. They also present a significant financial cost to healthcare institutions. Given this is one of the largest cohorts of trauma patient looking at utility of postoperative imaging in cervical trauma patients, we do not recommend the use of routine imaging after cervical spine fusion in the trauma population. Postoperative imaging should be reserved to patients with changes in clinical findings that prompt further evaluation.

A272: Decreased Value of Highly Accurate Fractional Anisotropy by Using 3-Tesla Zoom Diffusion Tensor Imaging After Decompressive Surgery in Patients With Cervical Spondylotic Myelopathy: “Aligned Fibers Effect”

Motoyuki Iwasaki¹ and Motoyuki Iwasaki¹

¹Otaru General Hospital, Otaru, Japan

Introduction: Diffusion tensor imaging (DTI) is widely used to show motor fiber tracts around the space occupying lesion in the brain in clinical scene. Despite similar trail to the spinal lesion, due to the smaller cross-sectional area of spinal cord and surrounding organs including variable magnetic susceptibility, the conventional DTI could not achieve a highly accurate evaluation. To date, most of the prior research showed the preoperative decreased FA values in cervical spondylotic myelopathy (CSM) patients. We used Zoom DTI (3-Tesla, Philips) and compared the acquired fractional anisotropy (FA) values of CSM patients during decompression surgery and assessed the change of FA values. Objective: To assess perioperative change of the FA value in CSM patients and prognostic factor by using Zoom DTI. Methods: Twenty-eight CSM patients were enrolled in this study. The group consisted of 15 male (54%) and 13 female (46%), aged from 42 to 92 years (mean 71.3 ± 12.7 years). All of them underwent decompression surgery. Zoom DTI and Japanese Orthopedic Association Score (JOA) was performed before surgery (FA pre and JOA pre), 1 week (FA 1w and JOA 1w) and 6 months (FA 6m and JOA 6m) after surgery. Twelve normal controls also included in this study to compare the values with CSM patients. The imaging parameters: b value = 600. The region of interest (ROI) was set inside the spinal cord margin to omit cerebrospinal fluid from C3/4 to C6/7 level. Results: The mean FA pre, 1w, and 6m of all patients were 0.569 ± 0.114, 0.485 ± 0.096, and 0.556 ± 0.103, respectively. There were significant differences between FA pre and 1w (P = .0076), and between FA 1w and 6m (P = .0442). JOA 6m (13.0) improved significantly compared with JOA pre (10.1) (P < .0001). Only FA 1w had positive relationship with JOA pre and 6 m significantly (R² = 0.334 and 0.186, P = .0049 and .0448, respectively). Conclusion: Presurgical FA in CSM patients did not differ from normal controls, however, significantly decreased after cervical decompression surgery, and significantly increased 6 months after surgery, which was remarkable in intracordal high signal patients. We hypothesized this phenomenon that postsurgical FA value approximates proper state of the damaged cord and presurgical FA includes masked effect as “aligned fiber effect.” Only FA 1w had positive relationship with JOA pre and 6m significantly, which established postsurgical FA can be more accurate prognostic factor than presurgical FA.

OP41: Trauma

A273: Clinical and Radiological Outcomes of Conservative Treatment for Unilateral Sagitally Split Fractures of C1 Lateral Mass

Jong-Beom Park¹ and Do-Gyun Kim¹

¹Uijeongbu St Mary’s Hospital, The Catholic University of Korea, Uijeongbu, Republic of Korea

Introduction: Unilateral sagitally split fracture (USSF) of C1 lateral mass (LM) is a rare variant type of C1 atlas fracture. The integrity of transverse atlantal ligament (TAL) is a key factor to determine the stability in the case of C1 atlas fracture. To date, definite treatment guideline of USSF of C1 LM has not been established. Moreover, the effect of TAL injury on surgical outcomes is still controversial in USSF of C1 LM. Therefore, we performed the current study to investigate clinical and radiologic outcomes of USSF of C1 LM that have been treated conservatively and suggest appropriate treatment guideline. Material and Methods: Twenty-six consecutive cases of USSF of C1 LM were included from 5 trauma centers of tertiary university hospitals. The fractures associated with other cervical spines, such as C2 and occiput, were excluded from the study. The mean age was 52 years; 16 were male and 10 were female. Sixteen were TAL injury group (9 type I and 7 type II by Dickman’s classification) and 10 were TAL intact group. All cases were treated by conservative measures, including skull traction followed by rigid brace for TAL intact group and halovest for TAL injury group for 12 weeks. The mean follow-up was 16 months (range, 12-47 months). Three spine surgeons measured radiologic parameters on lateral radiograph and open mouth view of cervical spine, 2-dimensional reconstructed computed tomography (CT) scans and magnetic resonance imaging (MRI) of initial and last follow-up: Total LM displacement (LMD), unilateral LMD at fracture side, atlanto-dental interval (ADI), clivus canal angle (CCA), atlanto-occipital joint axis angle (AOJAA), and basion-dens interval (BDI). The averages of three measurements were used as final results. The radiologic outcomes were evaluated by comparing initial presentation and last follow-up in 2 groups. Clinical outcomes were evaluated by visual analogue scale (VAS) and Odom’s criteria. Results: At last follow-up, for TAL intact group, total LMD, unilateral LMD at fracture side, ADI, CCA, AOJAA, and BDI were maintained well compared with initial presentation (1.2 vs 1.2 mm, P = .973) (1.0 vs 1.1 mm, P = .828) (1.5 vs 1.3 mm, P = .162) (154.8° vs 151.5°, P = .105) (105.9° vs 105.3°, P = .800) (4.2 vs 3.7 mm, P = .079). However, for TAL injury group, total LMD, unilateral LMD at fracture side, ADI, CCA, AOJAA, and BDI were worsened compared with initial presentation (5.9 vs 6.7 mm, P < .05) (4.3 vs 4.7 mm, P < .001) (2.0 vs 3.0 mm, P < .001) (155.6° vs 145.2°, P < .001) (107.8° vs 98.3°, P < .001) (4.4 vs 2.6 mm, P < .001). The worsening of total LMD, unilateral LMD at fracture side, ADI, CCA, and AOJAA were more severe in type I TAL injury than type II TAL injury (8.0 vs 5.1 mm, P < .05) (6.2 vs 4.4 mm, P < .01) (3.4 vs 2.5 mm, P < .05) (142.2° vs 149.0°, P < .05) (94.1° vs 103.8°, P < .05). The worsening of BDI was severe in type I TAL injury but statistically not significant (2.3 vs 2.9 mm, P = .486). VAS significantly decreased in TAL intact group (4.7 vs 2.1 points, P < .001) but not in TAL injury group (6.8 vs 4.7 points, P = .435). According to Odom’s criteria, satisfactory outcomes were higher in TAL intact group compared to TAL injury group (80% vs 37.5%, P < .05). Conclusion: Conservative treatment for USSF of C1 LM with TAL injury allows subsidence of occiput into C2 with aggravation of LMD of C1. This causes coronal and sagittal malalignment of occipitocervical junction, resulting in unsatisfactory clinical outcomes. Our results suggest that early surgical stabilization should be considered as choice of treatment for USSF of C1 LM with TAL injury.

A274: Analysis of Acute Secondary Adverse Events Following Traumatic Spinal Cord Injury And Its Relationship to Long-Term Neurological Recovery and Functional Outcomes

Fan Jiang¹, Blessing Jaja¹, Shekar Kurpad², Jetan Badhiwala¹, Bizhan Aarabi³, Robert Grossman⁴, James Harrop⁵, Jim Guest⁶, Ralph Schär⁷, Chris Shaffrey⁸, Max Boakye⁹, Elizabeth Toups⁴, Jefferson Wilson⁷, and Michael Fehlings¹

¹Toronto Western Hospital, University of Toronto, Toronto, Ontario, Canada

²Medical College of Wisconsin, Milwaukee, WI, USA

³University of Maryland, Baltimore, MD, USA

⁴Methodist Hospital, Houston, TX, USA

⁵Thomas Jefferson University Hospital, Philadelphia, PA, USA

⁶University of Miami, Miami, FL, USA

⁷St Michael’s Hospital, University of Toronto, Toronto, Ontario, Canada

⁸University of Virginia, Chalottesville, VA, USA

⁹University of Louisville, Louisville, KY, USA

Introduction: Traumatic spinal cord injury (SCI) can result in devastating consequences and permanent deficits for individual patients. The high rates of secondary adverse events (AEs), reportly varying from 38% to 79%, remains as the primary causes of morbidity and mortality in the affected inviduals. Currently, there is a lack of contemporary, prospective, multicentered studies on the spectrum and incidence of AEs and their relationship to long-term outcomes after SCI. The goal of this study is to address this knowledge gap by undertaking an analysis of the North American Clinical Trials Network (NACTN) prospective registry. Material and Methods: This study included 801 patients enrolled from 2005 to 2017 by eleven university-affiliated North American hospitals whose data is included in the NACTN database. Negative binomial regression and penalized maximum logistic regression models were fitted to study factors associated with higher incidence of AEs and to estimate the likelihood of neurological recovery, defined as AIS improvement ≥1 grade point, and functional outcomes according to the Spinal Cord Independence Measure subscores for respiration, bladder, and bowel sphincter management; and ambulation 6 months postinjury. Results: A total of 2303 AEs were identified in 502 (63%) patients. Of note, 352 (44.4%) experienced more than 1 event during the acute admission. When grouped by organ system, the most common AEs were classified as pulmonary (25.1%); hematological (17.5%); cardiac (14.5%); gastro-urinary (9.6%); skin (7.7%), and neuropsychiatric (6.5%). The most frequent specific AEs were found to be pneumonia (9.5% of all recorded AEs), respiratory failure (9.0%), anemia (8.6%), urinary tract infection (6.0%), pleural effusion (4.0%), bradycardia (4.0%), depression (3.9%), and sacral ulcers (3.1%). After accounting for age, baseline AIS (American Spinal Injury Association Impairment Scale) grade, and neurological level of injury, the group with at least 1 AE were less likely to improve on the AIS grade (OR 0.55, 95% CI 0.32-0.96) and are at higher risk for assisted breathing (OR 6.55, 95% CI 1.17-36.67); dependent ambulation (OR 3.43, 95% CI 1.90-6.19) or impaired bladder (OR 9.63, 95% CI 4.35-13.06) or bowel function (OR 7.86, 95% CI 4.31-14.32) at 6 months postinjury. Conclusion: Analysis of this contemporary multicentered series has shown that despite modern treatment methods, acute secondary AEs remain very common after SCI. Moreover, the development of secondary AEs can affect long-term neurological and functional outcomes. These findings underscore the need for prevention and management of secondary AEs following acute SCI.

A275: Reliability Analysis of the Aospine Crost (Clinician Reported Outcome Spine Trauma)

Said Sadiqi¹, J. J. Verlaan¹, Marcel W. Post², A. Mechteld Lehr¹, F. Cumhur Oner¹; and AOSpine Knowledge Forum Trauma³

¹University Medical Center Utrecht, Utrecht, Netherlands

²Rehabilitation Center ‘De Hoogstraat’, Utrecht, Netherlands

³AO Foundation, Davos, Switzerland

Introduction: Apart from a patient-reported outcome measure, there is also need for a simple, reliable and quick to administer tool for the treating surgeons that formalizes the most relevant clinical and radiological assessment parameters and reflects the spine surgeons’ perspective adequately. Together with independent patient-reported outcome, this tool will provide a holistic view of patients’ health in relation to their spine injury. Therefore, the AOSpine Knowledge Forum Trauma developed the AOSpine CROST (Clinician Reported Outcome Spine Trauma). A reliability study was performed to investigate the properties of the tool. Materials and Methods: The AOSpine CROST was developed on the basis of the results of two separate cross-sectional online surveys and a pilot test during an expert committee meeting. A reliability study was conducted among a selected number of internationally renowned experts on spine trauma. They were provided with 20 clinically representative spine trauma cases. The cases were presented as from the trauma setting until outpatient follow-up at 6 weeks or 3 months posttrauma. Also, all relevant clinical data as well as plain radiographs and computed tomography (CT) or magnetic resonance imaging (MRI) slices were provided. The participants were asked to rate the AOSpine CROST for each case in a first round. Four weeks after completing the first ratings, participants received the same cases in a different sequence to rate AOSpine CROST again. The inter- and intrarater reliabilities were analyzed for each parameter of the tool and the total score per case, both for the short term (≤12 months) and long term (>12 months), using kappa and intraclass correlation coefficients (ICC). Results: Sixteen spine trauma experts from all world regions participated in the study. Two participants did not complete the second assessment, resulting in the reliability analysis being performed for 14 raters. Fair to moderate intrarater agreement was seen for individual parameters at the short term (≤12 months) with kappa ranging from 0.40 (general bone quality) to 0.80 (radiographic sagittal alignment). For the long term (>12 months), the kappa values ranged from 0.31 (wound healing) to 0.67 (radiographic sagittal alignment). The inter-rater agreement analysis for individual items showed less favorable results. Slight to fair agreement was reached, both for the short term (0.18-0.60) and long term (0.16-0.46). For both time points, “spinal column mobility” showed the lowest agreement, while the best agreement was reached for “radiographic sagittal alignment.” Analyzing the total score on inter-rater agreement showed fair agreement for non-surgical cases (ICC = 0.52-0.56) and fair to good agreement for surgical cases (0.54-0.60). Conclusion: With the AOSpine CROST, a new concept of clinician reported outcome measurement in spine trauma is introduced. An initial reliability analysis among a selected number of spine trauma experts showed fair to moderate results. In the next steps, the tool will be further refined and tested in prospective studies. Once further validated, this tool has the potential to be used in daily clinical practice and for research purposes in order to create and contribute to evidence-based and patient-centered care.

A276: Pneumonia, Wound Infection, a nd Sepsis After Traumatic Spinal Cord Injury (SCI): The Relation to Long-Term Neurological and Functional Outcomes

Fan Jiang¹, Blessing Jaja¹, Jetan Badhiwala¹, Ralph Schär², Shekar Kurpad³, Robert Grossman⁴, James Harrop⁵, Jim Guest⁶, Elizabeth Toups⁴, Chris Shaffrey⁷, Bizhan Aarabi⁸, Max Boakye⁹, Michael Fehlings¹, and Jefferson Wilson²

¹Toronto Western Hospital, University of Toronto, Toronto, Ontario, Canada

²St Michael’s Hospital, University of Toronto, Toronto, Ontario, Canada

³Medical College of Wisconsin, Milwaukee, WI, USA

⁴Methodist Hospital, Houston, TX, USA

⁵Thomas Jefferson University Hospital, Philadelphia, PA, USA

⁶University of Miami, Miami, FL, USA

⁷University of Virginia, Chalottesville, VA< USA

⁸University of Maryland, Baltimore, MD, USA

⁹University of Louisville, Louisville, KY, USA

Introduction: Traumatic spinal cord injury (SCI) has profound and devastating consequences for individuals. Aside from the devastating neurological and functional compromise, infections have reported incidence reaching 40% in spinal trauma. Evidence suggests that the development of pneumonia and wound infection in SCI patients can impede neurological and functional recoveries. Using combined multicentered prospectively collected databases of the North American Clinical Trials Network (NACTN) and the Surgical Timing in Acute Spinal Cord Injury Study (STASCIS), this study investigated the incidence of hospital-acquired pneumonia, wound infection, or sepsis (PWS) in patients with SCI. And assesses the relationship of PWS with long-term neurological and functional outcomes. Material and Methods: Using the pooled database of NACTN and STASCIS, patients with a diagnosis of SCI and initial ASIA impairment scale evaluation within 7 days of injury were identified. The combined cohort was separated into 2 categories based on whether they had or not have had PWS. Comparative analyses were conducted with regard to clinical and treatment variables using Fisher’s exact test for categorical and Mann-Whitney test for continuous variables. Statically significant variables were selected and evaluated as potential predictors for PWS and ranked using dominance analysis. Penalized logistic regression model fitted with PWS as the independent variable was used to study the association between PWS and neurological and functional outcomes at 6-month follow up. Results: Total of 1299 patients was identified from the pooled database, 180 (14%) developed PWS during hospitalization. Predictors for PWS were identified and included: male (85.8% vs 76.1%, P = .007), cervical or thoracic neurological level injury, AIS (American Spinal Injury Association Impairment Scale) grade A (60% vs 35.5%, P < .001), and late surgical decompression >24 hours of the injury (49.3% vs 39%, P = .04). AIS grade was ranked as the strongest predictor using dominance analysis. After accounting for confounders, the development of PWS was associated with the increased requirement for support with respiration (odds ratio [OR] 3.91, 95% CI 1.42-10.79) and ambulation (OR 3.94, 95% CI 1.50-10.38) at 6 months. Additionally, the group with PWS showed a trend toward reduced AIS ≥1 grade point improvement after 6-month follow up (OR 0.69, 95% CI: 0.39-1.22); however, the difference did not reach statistical significance. Conclusion: There is increasing evidence that PWS in SCI patients during acute hospitalization can affect long-term functional and neurological recovery. The results of this study emphasize the need for clinical efforts in vigilant post-SCI care, preventive strategies and management algorithms for PWS in this patient population.

A277: Current Clinical Trials on Therapies for Individuals With Acute Traumatic Spinal Cord Injury: An Analysis on Their Generalizability

Julio Furlan¹ and B. Catharine Craven¹

¹University Health Network, University of Toronto, Toronto, Ontario, Canada

Introduction: Aging of the population has modified the worldwide epidemiology of traumatic spinal cord injury (SCI). There has been an escalation of fall-related SCIs among the elderly. This study was undertaken to examine the distribution of age among individuals who have been recruited to clinical trials for treatment of acute traumatic SCI. Material and Methods: This retrospective study includes all clinical trials that were registered at ClinicalTrials.gov since the website was launched in February 2000 until April 2018. Only clinical trials focused on therapies for acute traumatic SCI (up to 3 months after injury) were included. Complied data from the protocols of the captured trials were compared to population-based data on the epidemiology of acute traumatic SCI from the National SCI Statistical Center for 2016. Fisher’s exact test was used for the statistical comparisons. Results: There were 10 clinical trials identified in the scoping review and 44 clinical trials registered in at ClinicalTrials.gov. Elderly individuals (65 years or older at the time of SCI) were under-represented in all older clinical trials (P < .05). Among the clinical trials at ClinicalTrials.gov, only 13 out of 44 (29.5%) studies did not impose an upper limit for age in their recruitment criteria. Of the remaining 31 clinical trials, the most common age limit was 65 years (11 of 31) followed by 60 and 75 years (4 of 31 each), 70 years (3 of 31), 62 and 80 years (2 of 31, for each age limit), and 50, 55, 79, 85, and 100 years (1 of 31, for each age limit). Conclusion: Although most recent studies in the literature have challenged the notion that older individuals have less favorable recovery after SCI, several previous and current clinical trials continues excluding spinal cord injured individuals with older age at the time of injury. In addition to limiting their generalizability, those clinical trials may indirectly endorse the ageistic attitudes in the area of spinal cord trauma. Furthermore, the lack of uniformity in the recruitment criteria among the clinical trials imposes challenges for comparisons of their results.

A278: How to Approach L5? Treatment of Lumbal Plexus Injury After Pelvic Vertical Shear Fractures via Pararectus Approach: A Case Series

Sonja Haeckel¹, Sven Hoppe¹, Johannes Bastian¹, Lorin Benneker¹, and Keel Marius²

¹Inselspital, University of Bern, Bern, Switzerland

²Trauma Centre Hirslanden, Zurich, Switzerland

Introduction: Vertical shear fracture occurs usually after high-energy trauma. This rare injury pattern is associated with a complete vertical dislocation of the pelvic ring.¹ Because of the close anatomical relation to the lumbal plexus in front of the sacrum, pelvic shear fractures can lead to L5 nerve injury. If neurological symptoms persist after reduction and fixation of the fracture (usually done by using a posterior approach), the nerve root must be decompressed. In these cases, it is almost impossible to perform a content decompression from posterior, hence an anterior approach is needed to have a full view of the L5 root. The 2012 report by Keel and colleagues² described pararectus approach is a single incision (parallel to the M. rectus abdominis) intrapelvic approach, which is already more and more used for fixation of high acetabular fractures. Here, our aim is to present 4 cases of L5-nerve root injury after vertical shear fracture, we saw during a short period. For the satisfactory decompression the anterior pararectus approach was used. Material and Methods: Four patients with sacral fractures and/or pelvic vertical shear injuries who underwent an operative treatment and neurolysis via pararectus approach are described: Case 1: Male (18 years), fall from 4 meters, diagnosed with a secondary dislocation of a vertical shear fracture with neurological deficits (toe extensor M2); operation—neurolysis L4 and L5 and S1-screw left (percutaneous). Case 2: Female (29 years), fall from 9 meters, diagnosed with a vertical shear fracture right with incomplete plexus paresis; operation—lumbopelvic stabilization and secondary neurolysis L4-S1, foraminotomy L5 and S1, resection of callus and bone fragments was performed. Case 3: Male (45 years), fall from 30 meters, diagnosed with a vertical shear-fracture with fracture of the left sacrum and neurological deficits because of bony fragments (foot and toe extension paresis); operation—neurolysis L4 and L5, resection of bone fragments was performed. Case 4: Male (49 years), rollover car accident, diagnosed with an unstable pelvic ring fracture (type C) with neurological deficits (foot and toe extensor paresis); operation—decompression and L5 neurolysis. The follow-up period was 7 to 27 months. Results: In all cases, the L5 nerve route could be decompressed. The long-term follow-up showed the following: Case 1: neurological symptoms fully regredient. Case 2: significant improvement of function with persisting paresis for foot and toe extension (M3). Case 3: significant improvements with persisting paresis (toe extension left M3, foot extension M0, foot flexion M4, knee flexion, and extension M4). Case 4: no improvement with complete persisting paresis (foot and toe extension M0). Conclusion: The pararectus approach is a single-incision intrapelvic anatomical approach to the lumbal plexus, which allows complete anterior decompression and therefore an excellent alternative to the ilioinguinal approach. In our case series, we saw a complete regression of the neurological symptoms in 1 of 4 patients while 2 of 4 patients improved with persisting neurological deficits. Only 1 patient did not show any signs of improvement.

References

1. Blum L, Hake ME, Charles R, et al. Vertical shear pelvic injury: evaluation, management, and fixation strategies. Int Orthop. 2018;42:2663-2674. doi:10.1007/s00264-018-3883-1

2. Keel MJ, Ecker TM, Cullmann JL, et al. The pararectus approach for anterior intrapelvic management of acetabular fractures: an anatomical study and clinical evaluation. J Bone Joint Surg Br. 2012;94:405-411.

A279: A 10-Year Analysis of Combined Sternal and Vertebral Injuries in a Level-1 Trauma Center

Dorine S. Klei¹, F. Cumhur Öner¹, Luke P. H. Leenen¹, and Karlijn J. P. van Wessem¹

¹University Medical Center Utrecht, Utrecht, Netherlands

Introduction: Flexion and compression forces to the thorax can cause fractures both in the sternum and vertebral column. Combined sternal and vertebral fractures are less common and could cause severe instability of the trunk. Some authors consider the sternum as the fourth vertebral column. We studied the incidence and treatment of these combined fractures in a level-1 trauma center. Material and Methods: Retrospective analysis was performed on patients admitted to our level-1 trauma center from 2007 until 2016 with a combined sternal and vertebral fracture. Patients who died during hospital stay and patients who were lost to follow-up were excluded. Data are expressed as median (interquartile range [IQR]). Results: In 10 years, 86 patients were admitted with combined sternal and vertebral fractures. Vertebral body fractures were located in 27% in the cervical spine, 51% in the thoracic spine, and 22% in the lumbar spine. Median age was 50 (39-63) years, 63 patients were male (73%), and median Injury Severity Score (ISS) was 22 (14-34). A total of 72% of patients had associated rib fractures and 16% had clavicular fractures. All injuries were caused by blunt mechanism. In 16 (19%) patients, the vertebral fracture was accompanied by spinal cord injury. Patients stayed 16 (8-33) days in hospital and 37 of them (57%) were admitted to the intensive care unit (ICU). These patients stayed 7 (4-14) days on the ventilator and 8 (4-16) days in the ICU. Twenty-nine (34%) patients had surgical stabilization of the spinal fracture. Five (17%) of these patients needed reoperation because of 1 nonunion, 1 secondary dislocation, 2 malpositions of a pedicle screw, and 1 pain due to rods. Five of 57 (9%) nonoperated spinal fractures needed surgery for secondary dislocation of the spinal fracture. There was no difference in treatment failure rates between the originally conservatively treated patients and patients who initially had surgery (9% vs 17%, P = .29). Two patients (2%) had operative fixation of the sternum; one patient underwent primary fixation and one originally conservatively treated sternal fracture needed stabilization because of secondary dislocation. Conclusion: Combined sternal and vertebral fractures are rare and mainly caused by high-energy injuries. Secondary dislocation of vertebral fractures was the main reason for surgery in conservatively treated fractures. However, secondary dislocation and nonunion rates were low both in initially conservatively treated and operatively treated patients.

A280: Assessing Stability of Cervical Spine Fractures Using Adjunctive Dynamic C-Spine X-Ray

Ruben Patrick Thumbadoo¹, Jan Herzog¹, Kenny Kwan², Niv Bhamber¹, Jason Bernard¹, Timothy Bishop¹, Kenneth Cheung², and Darren Lui¹

¹St George’s Hospital, London, UK

²Li Ka Shing Faculty of Medicine, Queen Mary Hospital, Hong Kong SAR, China

Introduction: In the management of a trauma patient with cervical spine (c-spine) injury, the need for accurate diagnostic imaging is key to ensure the correct management. Different classification systems have been developed initially Subaxial Injury Classification system (SLIC) and more recently AO cervical spine fracture classification. The SLIC score helps to determine if surgery is required, in contrast to the AO classification only determines fracture pattern to suggest management. Denoting a stable facet fracture as F1 and F2 for potentially unstable facet. Through a multicenter study, we have identified a group of cases where the use of supine CT alone to classify fractures by either SLIC or AO score may be deficient and the use of dynamic C-spine X-ray can help identify instability. Material and Methods: Systematic review of patients across 1 Hong Kong and 1 UK site, looking at patients with cervical spine injuries initial treated conservatively and who subsequently underwent surgical intervention. In conjunction with systematic review of the literature to identify cervical spine injuries with low SLIC scores and dynamic c-spine X-ray indications. There is one article by Humphrey et al, which describes 4 cases of low-scoring SLIC scores suggesting conservative management was satisfactory. Results: There were 11 cases identified in total. The mean SLIC score predynamic X-ray was 0.63 in contrast to the postdynamic x-ray mean SLIC score of 6. Using paired T test, the P value was .00 000 004, which showed statistical significance. These cases were all classified as F1—stable facet fracture by the AO classification. Conclusion: Computed tomography (CT) can propose instability only by suggesting ligamentous injury. Supine imaging eliminates the gravitational loads normally exerted on the c-spine. SLIC can be used to guide management based on clinical and radiological evidence. We show 11 cases of assumed cervical stability based on CT imaging, who underwent dynamic c-spine X-ray that showed instability and increased their SLICS score above the surgical threshold. Stability of fractures is keen in determining the management. According to AO classification fractures seen on CT were “stable” and therefore F1. Based on supine imaging either suggest ligamentous injury or not, hence either stable or unstable. During the dynamic X-ray, the muscles and ligaments are under strain, therefore instability due to ligament damage can be demonstrated. We therefore suggest a combination of SLIC algorithm and AO classification using clinical symptoms and radiological imaging to classify fracture and correlate clinical symptoms, with clinical persistence of neck pain warranting a Miami J collar and erect or dynamic c-spine X-ray to assess stability.

A281: The First Results of the Functional Approach of Surgical Treatment of the Upper Cervical Spine Trauma

Takhir Tamaev¹, Vadim Manukovskiy¹, Konstantin Tulikov¹, and Valeriy Serikov¹

¹Department of Neurosurgery Emergency Care Institute of Medicine, San, Russian Federation

Introduction: Most of fusion techniques of the upper cervical spine are aimed to repositioning, formation of arthrodesis, which leads to a significant restriction of function in the most mobile sections of the cervical spine and, as a consequence, to a significant decrease of the quality of life of the injured. Material and Methods: Forty patients with C1-C2 fixation by Harms technique, without creating conditions for posterior arthrodesis between the C1-C2 vertebrae were divided into 4 groups: C1 Jefferson type (n = 10), odontoid fractures C2 (type II) (n = 14), odontoid fractures C2 (type III) (n = 13), and combined fractures C1-C2 (n = 3). Fracture consolidation was x-ray controlled after 6 weeks, 3 months, and 6 months from the moment of fixation. After consolidation, implants were removed as soon as possible. In case of consolidation of fractures in terms of more than 6 months, people were excluded from the study groups. This was in spite of the inability to restore function in the spinal-motor segment due to strong arthrodesis. In terms of 3 to 6 months after removal of the implants, the patients underwent functional computed tomography (CT) with maximum possible arbitrary head turns in both directions and maximum angle of deviation of the transverse processes of C1 and C2 vertebrae was determined relative to each other in degrees. Rotation of 25° in each direction was accepted for the normal volume of motion. The increase in the volume of movements in the C1-C2 segment by 5° in each direction, or by 10° in one of the sides, which was 20% of the normal indices of healthy individuals, was taken as a functionally significant result. Results: The first signs of consolidation occurred after 3 months in patients in group C2 (type III). In a period of 3 to 6 months consolidated fractures C1 Jefferson type and odontoid process C2 (type II) but also combined fractures of C1 and C2 vertebrae in 1 of 3 patients. For 6 months, consolidation was detected in 35 (87.5%) patients. Functionally positive results were obtained in 28 patients (80%). In 7 (20%) patients, no recovery was observed in the volume of movements. Of these patient, 2 patients were with Jefferson fracture type, 2 patients were with fracture of the odontoid process C2 (type III), and 3 patients with associated injuries C1-C2 vertebrae. The most pronounced functional results (rotation angle of more than 15° in each direction) were patients with fractures of the odontoid process C2 (type II). The worst (<15°) was obtained in patients with fractures of the Jefferson type and the odontoid process C2 (Type III). In 6 months after the removal of the implants, the amount of movements in average grew by 5° to 7° and reached 20° in each direction in 10 patients. Conclusion: The first results show that the 2-step functional approach to the treatment of the upper cervical spine injury can restore the main volume of movements in the C1-C2 joint in most patients.

OP 42: Minimally Invasive Spine Surgery 4

A282: Vascular Complications Associated With the Minimally Invasive Anterior to the Psoas (ATP) Approach for Lumbar Fusion

Chadi Tannoury¹, Avilash Das², Rahul Bhale², Molly Vora², Kathy Chen², Giuseppe Orlando¹, and Tony Tannoury¹

¹Boston Medical Center, Boston, MA, USA

²Boston University, Boston, MA, USA

Introduction: Anterior access to lumbar fusion is desirable and can be applied to a large spectrum of spinal pathologies. However, reported hazards to major vessels, bowels, and ureter can be catastrophic and often lead to under utilization of the anterior lumbar approach even with the assistance of access surgeon. Vascular injuries have been reported up to 6% during anterior spinal surgeries.^1,2 The transpsoas lateral approach (lateral lumbar [LLIF] and extreme lateral [XLIF] interbody fusion) also failed to eliminate such injuries and does not allow safe access to L5-S1, and at times to L4-5 discs.^3-5 Moreover, significant vascular injuries (2.9%-9.5%) were encountered with the newly described oblique lateral interbody fusion (OLIF).^6,7 We report our experience with the anterolateral “anterior to the psoas” (ATP) approach, that allows access to the spine between T12 to S1 under direct visualization. This study is the largest investigation into the incidence of vascular complications associated with the ATP approach. Methods: A retrospective review of 909 consecutive patients who underwent ATP lumbar spine fusion between 2008 and 2017. All procedures were done by 2 spine surgeons without the need of access surgeon.The spine was generally approached from either lateral sides (if no deformity), and from the concave side for coronal deformities. For patients with a unilateral predominant symptomatic mechanical neural compression, the spine was approached from the contralateral side to achieve direct adequate decompression of the stenosed canal, lateral recess, and/or foramen. Variables collected included vascular injuries identified intraoperatively and at any time point postoperatively. Vascular injury was defined major if involving large vessel (eg, aorta, vena cava, common iliac, external or internal iliac vessels) and minor if involving small vessel (eg, segmental, middle sacral vessels, etc). Results: We identified 909 patients (55.9% females, 44.1% males) who underwent spinal fusion using an ATP approach, 540 patients (59%) included L5-S1 level. The average age of the cohort was 59 years. Indications for fusion included lumbar spondylolisthesis with stenosis (n = 622), adult spine deformity with stenosis (n = 128), adjacent segment disease (n = 141), trauma (n = 7), pseudoarthrosis (n = 5), metastasis (n = 4), and infection (n = 2). A total of 2373 levels were fused, with a mean of 2.25 levels patient.The exposure was carried anterolaterally along the retroperitoneum and the anterior border the psoas. When exposing L5-S1, the L5 segmental vessels were typically ligated and coagulated at least 5 mm away from the common iliac tributary. Only 3 patients (0.3%) experienced minor vascular injuries, all involved segmental vessels. The intraoperative bleedings were successfully controlled by the spine surgeon and ceased using local hemostasis (application of thrombin soaked gelatin sponges [Gelfoam] and oxidized cellulose matrix [Surgicel]) withnofurther postoperative bleedings. Conclusions: The minimally invasive ATP approach to the spine affords a significant reduction in vascular complication (0.3%) compared with anterior lumbar interbody fusio (ALIF; 3%-7%),^1,2 XLIF (0.8%-2%),^3-5 and OLIF (2.9%-9.5%).^6,7 In our experience, ATP lumbar fusion is a promising and safer alternative to ALIF, LLIF, and OLIF, and it provides a full access to the spine between T12-S1.

References

1. Bateman DK, Millhouse PW, Shahi N, et al. Anterior lumbar spine surgery: a systematic review and meta-analysis of associated complications. Spine J. 2015;15:1118-1132.

2. Rajaraman V, Vingan R, Roth P, Heary RF, Conklin L, Jacobs GB. Visceral and vascular complications resulting from anterior lumbar interbody fusion. J Neurosurg. 1999;91(1 suppl):60-64.

3. Hijji FY, Narain AS, Bohl DD, et al. Lateral lumbar interbody fusion: a systematic review of complication rates. Spine J. 2017;17:1412-1419.

4. Phillips FM, Isaacs RE, Rodgers WB, et al. Adult degenerative scoliosis treated with XLIF: clinical and radiographical results of a prospective multicenter study with 24-month follow-up. Spine (Phila Pa 1976). 2013;38:1853-1861.

5. Anand N, Baron EM, Khandehroo B, Kahwaty S. Long-term 2- to 5-year clinical and functional outcomes of minimally invasive surgery for adult scoliosis. Spine (Phila Pa 1976). 2013;38:1566-1575.

6. Woods K, Fonseca A, Miller LE. Two-year outcomes from a single surgeon’s learning curve experience of oblique lateral interbody fusion without intraoperative neuromonitoring. Cureus. 2017;9:e1980.

7. Woods KR, Billys JB, Hynes RA. Technical description of oblique lateral interbody fusion at L1-L5 (OLIF25) and at L5-S1 (OLIF51) and evaluation of complication and fusion rates. Spine J. 2017;17:545-553.

A283: Minimally Invasive Lateral Lumbar Interbody Fusion for Clinical Adjacent Segment Pathology: A Comparative Study with Posterior Lumbar Interbody Fusion

Hyung-Youl Park¹, Young-Hoon Kim¹, Kee-Yong Ha², Sang-Il Kim², and Jun-Hyung Cho²

¹St. Mary’s Hospital, The Catholic University of Korea, Seoul, Republic of Korea

²Seoul, Republic of Korea

Introduction: Minimally invasive techniques have been increasingly applied for spinal surgery. No report has compared lateral lumbar interbody fusion (LLIF) with posterior instrumentation (PI) and conventional posterior lumbar interbody fusion (PLIF) with PI for clinical adjacent segment pathology (ASP). The aim of this study is to evaluate the clinical and radiological efficacies of minimally invasive LLIF for clinical ASP. Material and Methods: Forty patients undergoing LLIF with PI (hybrid surgery) were compared with 40 patients undergoing PLIF with PI (posterior surgery). The radiological outcomes, including indirect decompression and cage subsidence, and clinical outcomes such as the Oswestry Disability Index (ODI) and visual analog scale (VAS) were assessed. Postoperative major complications and reoperations were also compared. Results: Correction of segmental Cobb’s angle and segmental lordosis in the hybrid surgery were significantly higher postoperatively (−2.8° vs −0.9°, P = .012; 7.4° vs 2.5°, P = .009) and at the last follow-up (−2.4° vs −0.5°, P = .026; 4.8° vs 0.8°, P = .016) compared with posterior surgery. Regarding indirect decompression of the LLIF, significant increases in thecal sac (83.4 vs 113.8 mm²) and foraminal height (17.8 vs 20.9 mm) were noted on postoperative MRI. Although intraoperative endplate fracture developed in 17.7% of patients undergoing hybrid surgery, cage subsidence was not significantly different between the 2 groups. Clinical outcomes such as back VAS (4.1 vs 5.6, P = .011) and ODI (48.9 vs 59.6%, P = .007) at postoperative 3-month were significantly better with hybrid surgery. Rates of postoperative major complications and reoperations were not significantly different between the 2 groups. Conclusion: Hybrid surgery for clinical ASP significantly improves segmental coronal and sagittal correction, and indirect decompression leading to better clinical outcomes compared with posterior surgery. LLIF combined with supplemental instrumentation can be an effective surgical option for the clinical ASP.

A284: Effect of Supplemental Fixation on Fusion Rates at One Year Following Lateral Lumbar Interbody Fusion

Timothy O’Connor¹, Asham Khan¹, Joshua Meyers¹, Jeffrey Mullin¹, and John Pollina¹

¹University at Buffalo, Buffalo, NY, United States

Introduction: Lateral lumbar interbody fusion is a minimally invasive technique used to treat degenerative spinal disorders that preserves segmental lordosis, limits subsidence, and generally results in relatively high fusion rates. Achieving complete fusion is a primary goal of the procedure, and prior studies have reported encouraging clinical and radiographic outcomes. The use of supplemental fixation after lateral interbody fusion can help in the fusion process; however, the effects of supplemental fixation after lateral interbody fusion remain understudied. The aim of our study is to determine the difference in fusion rates in patients receiving supplemental fixation compared to fusion rates without supplemental fixation. Material and Methods: After obtaining institutional review board approval, a retrospective chart review was conducted on all patients that underwent a one-level lateral lumbar interbody fusion for the treatment of degenerative disc disease with or without spondylolisthesis between September 2013 and February 2018. The patients were placed into 2 groups: one group receiving supplemental posterior fixation with either unilateral or bilateral pedicle screw placement, and one group without supplemental posterior fixation. Patient demographics, body mass index (BMI), preoperative and postoperative lumbar lordosis, coronal Cobb angle, sacral slope, supplemental fixation, subsidence, and fusion grades were collected. Covariates such as age, gender, BMI, and tobacco use were identified and accounted for in the final statistical analysis if a statistically significant relationship was noted between a covariate and an outcome. Results: There were a total of 83 patients included in our analysis. Male and female patients were equally represented (41 male, 42 female). Mean age was 57.6 ± 10.5 years, mean BMI was 30.6 ± 5.2 kg/m², and 17 patients (20.5%) endorsed tobacco use. Forty-five patients (54.2%) had no supplemental fixation. Of the 38 patients that had posterior fixation, 24 patients had unilateral fixation and 14 patients had bilateral fixation. Age and BMI were not significantly different between the 2 groups. In our recent analysis, mean follow up time in our cohort was 11.6 months. Out of the 45 patients that did not receive supplemental fixation, 42 patients (93.3%) achieved complete fusion, and the remaining 3 patients (6.7%) had a grade II fusion, Of the 38 patients who received supplemental fixation, 37 patients (97.3%) achieved complete fusion. The one patient (2.7%) that had a grade II fusion had unilateral fixation. Overall, 79 out of 83 patients (95.2%) achieved complete fusion with a follow-up time now of 11.6 months, regardless of whether supplemental fixation was used. After adjusting for covariates, there were no statistically significant differences between those 2 groups with regard to subsidence (P = .19) or fusion rates (P = .68). There were no statistically significant differences between the 2 groups when comparing lumbar lordosis (P = 0.67), coronal Cobb angle (P = .11), or sacral slope (P = .48). Conclusions: Supplemental posterior fixation with lateral lumbar interbody fusion is understudied, and long-term outcomes evaluating the utility of supplemental fixation are limited. In our patient population, our study demonstrates that over 95% of patients achieve fusion at 11.6 months, and there is no difference in fusion rates or subsidence with supplemental fixation. Further studies are warranted to evaluate the long-term prospective outcomes of these patients.

A285: Comparative Analysis Between Three Different Lumbar Decompression Techniques (Microscopic, Tubular, and Endoscopic) in Lumbar Canal and Lateral Recess Stenosis: Preliminary Report

Chul-Woo Lee¹, Kang-Jun Yoon¹, and Sang-Soo Ha¹

¹St Peter’s Hospital, Seoul, Republic of Korea

Introduction: The purpose of our study is to compare the results of spinal decompression using the full-endoscopic interlaminar technique, tubular retractor, and a conventional microsurgical laminotomy technique, and evaluate the advantages and clinical feasibility of minimally invasive spinal (MIS) lumbar decompression technique in the lumbar canal and lateral recess stenosis. Material and Methods: The authors retrospectively reviewed clinical and radiological data from 270 patients who received microsurgical (group E: 72 patients), tubular (group T: 34 patients), or full-endoscopic decompression surgery (group E: 164 patients) for their lumbar canal and lateral recess stenosis from June 2016 to August 2017. Clinical (visual analogue scale [VAS], Oswestry Disability Index [ODI], and Mcnab criteria), radiologic (spinal canal diameter, segmental dynamic angle, and disc height) and surgical outcome parameters (creatine phosphokinase [CPK] level, operative time, blood loss, and hospital stay) were evaluated pre- and postoperatively, and compared among the 3 groups by means of statistical analysis. Failed cases and complications were reviewed in all groups. Results: The mean follow-up period was 6.38 months. The overall clinical success rate was 89.4%. All groups showed favorable clinical outcome. The clinical and radiologic results were similar in all groups. Regarding surgical outcome, group E showed longer operation time than groups M and T (group E: 84.17 minutes/level, group M: 52.22 minutes/level, group T: 66.12 minutes/level) (P < .05). However, group E and T showed minimal surgical invasiveness compared with group M. Groups E and T showed less immediate postoperative back pain (VAS) (group E: 3.13, group M: 4.28, group T: 3.54) (P < .05), less increase of serum CPK enzyme (group E: 66.38 IU/L, group M: 120 IU/L, group T: 137.5 IU/L) (P < .05), and shorter hospital stay (group E: 2.12 days, group M: 4.85 days, group T: 2.83 days) (P < .05). The rates of complication and revision were not significantly different among the 3 groups. Conclusion: MIS decompression technique is clinically feasible and safe to treat the lumbar canal and lateral recess stenosis, and it has many surgical advantages such as less muscle trauma, minimal postoperative back pain, and fast recovery of the patient compared with traditional open microscopic technique.

A286: The Effectiveness of Percutaneous Endoscopic Stenosis Lumbar Decompression (PESLD) for Facet Joint Preservation in Degenerative Lumbar Spinal Stenosis

Hyeun-Sung Kim¹, Ravindra Singh¹, Sung Woon Oh¹, Jung Hoon Noh¹, Seong-Hoon Oh², Il Tae Jang¹, and Nitin Maruti Adsul³

¹Nanoori Hospital Gangnam, Seoul, Republic of Korea

²Nanoori Hospital Incheon, Incheon, Republic of Korea

³Sir Ganga Ram Hospital, New Delhi, India

Introduction: Clinical instability is an important cause of low back pain. Facet joints affect the stability of the motion segment of the spine. The important point of the decompression operation in spinal stenosis will be the preserve segmental structures, especially facet joints and ligamentous complex. The purpose of this study is to evaluate the effectiveness of Percutaneous Endoscopic Stenosis Lumbar Decompression (PESLD) for facet joint preservation simultaneously with the sufficient decompression of symptomatic spinal stenosis. Material and Methods: We retrospectively reviewed the patients with degenerative lumbar spinal stenosis under surgical decompression. We divided this patients into 3 groups based on the surgical approach and method. Group A (n = 64, 85 levels) is under PESLD, group B (n = 57, 62 levels) is under unilateral laminotomy for bilateral decompression, group C (n = 37, 48 levels) is under unilateral approach, bilateral splitting ligamentum flavectomy. We measured the spinal canal volume, ipsilateral and contralateral facet joint total volumes, inferior articular process (IAP) volumes, facet joint lengths, respectively, on the axial preoperative and postoperative magnetic resonance imaging (MRI). The level of axial image is discal middle section of the spinal stenosis level on the sagittal image. Analysis of paired t test was used for evaluating of the effectiveness of volume change in each group. Results: There were no statistically significant difference between preoperative and postoperative states concerning facet joint length, IAP volume, and facet joint total volume in groups A and C. However, there was statistically significant difference in the group B. (P < .001). Multivariate analysis revealed that postoperative spinal canal is more extensive in groups A and B than in group C (P = .016). In postoperative ipsilateral and contralateral facet length, group A is longer than the other groups (P < .001). Conclusion: According to this study’s result, we obtained sufficient decompression under the segmental motion preservation by using the PESLD (group A).

A287: Lateral Lumbar Interbody Fusion: Minimum Six Years Follow-Up on a Consecutive Series of Patients Done in a Single-Center Setting

Josip Buric¹, Renato Conti¹, Massimiliano Pulidori¹, and Ilaria Barni¹

¹Spine Surfgeryt, Forte dei Marmi, Italy

Introduction: Lumbar lateral interbody fusion is a relatively novel surgical procedure for minimally invasive interbody fusion of the lumbar spine. We report a minimum of 6 years of follow-up on a consecutive series of patients done at a single center. Materials and Methods: A consecutive series of 59 patients, submitted to extreme lateral interbody fusion (XLIF, Nuvasive), were retrospectively evaluated for clinical results and complications on a minimum of 6 years of follow-up. Mean age at the time of surgery was 60.3 years (minimum 38, maximum 81 years). There were 40 female and 19 male patients. Indication for surgery was degenerative disc disease in 21 cases, degenerative listhesis in 10, postlaminectomy syndrome in 9, lumbar stenosis and adult deformity in 6 each and another 2 for gangliar cyst and lythic listhesis each while 3 cases had proximal junctional syndrome. Thirty-one patients had the procedure done at a single level (of whom 28 at L4-L5), 19 had 2-level pathology and 9 cases had an interbody fusion done on more than 2 levels. Five patients had a stand-alone procedure, another 5 had percutaneous pedicle screw instrumentation while the rest were submitted to posterior open instrumentation and fusion. First patient was operated on May 2009 and the last on November 2012. The follow up was done retrospectively with out-patient or telephone interview. The patients were asked the following questions: on a scale from 0 to 10: How do you judge your pain and function when compared with the preoperative status? Have you had any additional procedures/surgeries for your lower back problems or generally, do you take pain killers for your lower back pain (occasionally or continuously). Moreover, all the patients underwent an X-ray for radiological evaluation of the surgical procedure. Results: On minimum 6 years of follow-up, 8 patients were unreachable and thus, considered lost to follow-up. One patient died from a disease not related to the procedure (heart attack 2 years after the procedure). Accounting for immediate and early postoperative complications there was one case of cage infection treated successfully with antibiotics, 4 patients with surgically related anterior thigh pain, numbness and/or weakness and 1 patient with intraoperative endplate subsidence. In 1 patient, at L4-L5 level, the procedure was aborted due to the impossibility to pass through the lumbar plexus. On long-term follow-up, 7 patients presented with proximal junctional syndrome (PJS) of whom 6 were clinically relevant and one was asymptomatic. Except one, all the patients presenting with PJS had a surgery done at more than 2 levels. Overall, during the follow-up, 11 patients were submitted to further surgeries/procedures: 6 for fusion extension including PJS, 2 for sacro-iliac fusion, 3 for total hip replacement and 2 for facet joint injection. Four patients reported same or worse clinical condition while the rest of available patients reported a 5.9 points mean improvement (59%), on a scale from 0 to 10, when compared to preoperative clinical conditions. More than half of the available patients, 54%, reported taking occasionally pain killers or nonsteroidal anti-inflammatory medication, while 10% reported continuous intake of pain medication. Radiologically, no additional endplate subsidence, cage breakage/displacement, or related pedicle screw loosening was observed when compared with early follow-up images. Radiologically, no case presented with clear signs of pseudo-arthrosis at the level of interbody fusion and in all the patients, bony fusion signs could be seen, although this was not evaluated independently. Conclusions: Lumbar lateral interbody fusion showed, on a long-term follow-up, a clinically useful result. Twenty-two percent of patients showed PJS although that could be more correlated to sagittal imbalance than to the interbody fusion itself. No long-term complications related directly to the surgical procedure or to the cage placement were noted.

A288: Endoscopic Surgery and Dynamic Rehabilitation in Foraminal and Extra-Foraminal Disc Herniation: 888 Cases

Marjory Rué¹ and Jean Destandau¹

¹Endoscopic Spine Center, Bordeaux, France

Introduction: Foraminal and extra-foraminal disc herniations in the lumbar region need wide exposure with agressive muscle retraction to expose foraminal region. To avoid this issue, the device (Endospine, Karl Storz) was designed by Dr Jean Destandau and has been used since 1993. To this type of surgery, we organize a dynamic program of rehabilitation with a early rise, walk, and sportive recovery. We will present the surgical technique and a study of 888 cases. Material and Methods: The device is composed of 3 tubes: one for the endoscope, one for suction canula, and the largest one for classical surgical instruments. The endoscope is a rigid and straight 0° endoscope. From February 2001 to December 2015, 888 patients were operated on with this technique. Represented levels were in 47.1% L4-L5, 24.4% L3-L4, 21.1% L5-S1, 0.06% in L2-L3, and 0.005% in L1-L2. The dynamic program is a immediate rise after surgery, a walking distance, and 2 rehabilitation sessions during the 2 nights long hospitalization. Results: A total of 888 patients were operated upon, 56% on the left side, 44% on the right. A total of 6.5% had a previous surgical history of disc herniation and 3% were a real recurrent disc herniation. In all, 32% was operated under spinal anesthesia, 47% had a motor palsy, and 50% a sensitive loss. A total of 59% of the patients were workers, and 63% returned at work with a mean time of 3 weeks. The complication rate was 0.01% divided in 4 fluid leakages, 5 nerve root lesions, 22 articular process resection, and 7 level mistakes, no infection. Patients were followed during 12 years on average. Satisfaction forms were posted, 58% were answered. A total of 96% of patients were satisfied and 94% estimated to be well informed and 100% walked right after the surgery and went back home in their own cars. Conclusion: The efficiency of this minimal invasive technique is certain. Muscle retraction is minimal, visual field maximal and nerve root decompression optimal. The dynamic rehabilitation is really welcome among patients with a short break in working and sportive activities. The whole process without increasing postoperative complications. We could add a hundred patients until December 2018 to go to a thousand.

A289: Clinical and Radiographic Outcomes of Lateral Lumbar Interbody Fusion for Adjacent Segment Degeneration

Siamak Yasmeh¹, James Bernatz², Eli Garrard², Miranda Bice², and Seth Williams²

¹Los Angeles, CA, USA

²University of Wisconsin, Madison, WI, USA

Introduction: Adjacent segment degeneration requiring surgical treatment after previous lumbar fusion occurs at a rate of ∼16% at 5 years and ∼36% at 10 years. Published literature on lateral lumbar interbody fusion for treatment of adjacent segment degeneration is limited. Material and Methods: We retrospectively reviewed all adult patients at our institution who presented with adjacent segment degeneration and stenosis cephalad to an instrumented lumbar fusion refractory to nonoperative treatment and who subsequently underwent lateral lumbar interbody fusion. Clinical and radiographic outcomes were analyzed and compared between preoperative baseline and postoperative values. Paired comparison of variables at follow-up compared to baseline were conducted with either paired t tests or McNemar’s test for pair proportions. Results: Thirty-six patients with symptomatic adjacent segment degeneration at 46 motion segments were included. All cases had either supplemental posterior instrumentation (n = 29) or lateral plating without posterior instrumentation (n = 7) and were performed by a single surgeon. At final follow-up, there was a decrease in the percentage of patients with back and lower extremity pain from 100% and 97.2%, respectively, to 13.9% and 8.3%, respectively (both P < .001). A total 83.3% of patients had complete relief of both preoperative lower extremity pain and back pain at the time of final follow-up. Oswestry Disability Index scores improved from a mean of 49.3 ± 14.2 preoperatively to 30.3 ± 18.9 postoperatively (P < .001). VR12 physical component scores significantly improved from a mean of 25.6 ± 8.1 preoperatively to 34.0 ± 11.6 (P < .01) at final follow-up. VR12 mental component scores were similar over time with a mean of 45.3 ± 12.5 preoperatively to 48.4 ± 13.2 (P = .307) at final follow up. 16.7% of patients had approach-related hip flexion pain/weakness and/or thigh numbness, of which all but one resolved spontaneously. Compared with preoperative baseline measurements, initial postoperative, and final postoperative radiographic measurements had an increase in segmental lordosis (P < .001 and P < .001, respectively), increase in overall lumbar lordosis (P < .05 and P < .01, respectively), decrease in segmental coronal angulation (P = .63 and P < .01, respectively), and an increase in intervertebral height at the operative levels (P < .001 and P < .001, respectively), but overall coronal alignment was not significantly changed (P = .089 and P = .161, respectively). Conclusion: Lateral lumbar interbody fusion achieves good clinical and radiographic outcomes for the treatment of adjacent segment degeneration after previous lumbar fusion.

A290: Hardware-Related Complications and Reoperation Rates Following Long Spinal Fusion to the Sacrum With Iliac Fixation in Patients Treated With Minimally Invasive Anterolateral Anterior to the Psoas (ATP) Approach and Posterior Percutaneous Screws

Tony Tannoury¹, Giuseppe Orlando¹, Avilash Das², Rahul Bhale², Kathy Chen², Molly Vora², and Chadi Tannoury¹

¹Boston Medical Center, Boston, MA, USA

²Boston University, Boston, MA, USA

Introduction: Adult spinal deformities requiring long fusions, to the lumbar-sacral junction are known for high rate of distal hardware failure due to substantial flexion moments and posterior cantilever forces. The reported incidence of pseudarthrosis, loss of posterior fixation, or painful hardware requiring revision is still high. The minimally invasive ATP technique, through a single muscle splitting exposure, allows for a safe access and substantial anterior column support between T12-S1 thereby reducing the risk of posterior hardware failure, pseudoarthrosis, and reoperation. This study is the first to investigate the incidence of posterior hardware-related complications, following long segment anterior-posterior fusion using anterior to the psoas (ATP) and posterior percutaneous fixation. Material and Methods: This is a retrospective review of 60 patients who underwent long spinal fusion to the sacrum with (53) or without (7) iliac fixation for the management of adult spinal disorders between 2007 and 2016. All included patients had a minimum of 2-year follow-up with an average of 42 months final follow-up. Postoperative complications were also collected, including infections, neurovascular injury, implant fracture, implant displacement, hardware prominence and related pain, pseudoarthrosis, bony fractures, and need for surgical revision. Results: All 60 patients included in the study underwent a combined anterior-posterior spinal fusion to S1 using the ATP approach. Of these patients, 53 received iliac fixation. Pelvic fixation was not used in 7 patients with transitional lumbosacral anatomy. The average age of the cohort (72% females; 28% males) was 68.2 years. Indications for fusion were adult spinal deformity (41%), degenerative disc disease with instability (55%), and adjacent segment disease (3%). The average number of fused levels was 7.5.The upper-instrumented vertebra (UIV) was T4 (n = 2), T5 (n = 1), T9 (n = 3), T10 (n = 22), T11 (n = 19), and T12 (n = 13). Posterior wound complications were noted in 3 patients (5%) who had superficial wound infection requiring surgical debridement. In terms of hardware related complications: 4 patients (6.6%) had screws breakage (at T10, T11, T12, and L1, respectively), 4 patients (6.6%) had rod breakage (at T12-L1, L1-L2, L2-L3, and L3-L4 levels, respectively), 1 patient (1.6%) had iliac screw revision due to violation of the acetabulum, 1 patient (1.6%) required removal of 1 iliac screw due to pain,7 patients (11.6%) had screw lucency, 2 patients (3.3%) developed pseudoarthrosis, and 3 patients (5%) developed proximal junction kyphosis (PJK). Of all patients, only 8 (13.3%) required surgical intervention: 3 (5%) due to infection, 3 (5%) due to HF, and 2 (3.3%) due to iliac screw complication. No major vascular injuries, bowel-visceral injuries, or paralysis were encountered using the ATP approach. Conclusion: Long spinal fusion to the pelvis is a challenging procedure characterized by high risk of hardware failure and revision surgeries. In our experience, the use of ATP approach coupled with posterior percutaneous pedicle and iliac screws had a very protective effect against revision surgery for iliac screw complications. No breakage or mechanical failure of the distal construct (L4-S1) was observed, in addition to a very low pseudoarthrosis rate (3.3%). Furthermore, these benefits were not associated with any significant anterior related morbidities such as visceral-vascular complications.

OP43: Surgical Complications 2

A291: The Safety Profile of Intentional or Iatrogenic Sacrifice of the Artery of Adamkiewicz and Spinal Segmental Arteries in Its Vicinity: A Systematic Review

Terence Tan¹, Joost Rutges², Travis Marion³, Charles Fisher⁴, and Jin Tee¹

¹Alfred Hospital, Melbourne, Victoria, Australia

²Erasmus MC, Rotterdam, Netherlands

³Big Thunder Orthopaedic Associates, Thunder Bay, Ontario, Canada

⁴University of British Columbia and Vancouver General Hospital, Vancouver, British Columbia, Canada

Introduction: The artery of Adamkiewicz (AoA) is regarded as the most important segmental artery supplying the spinal cord. The spinal cord also receives its arterial supply from a highly variable number of spinal segmental arteries (SSA). There is paucity of consensus on (1) whether the AoA can be safely ligated and (2) the number of contiguous SSAs that can be safely ligated without the causation of catastrophic spinal cord ischemia or cord infarct. This systematic review aims to determine the risk of postoperative neurological deficits from intentional or iatrogenic sacrifice of the (1) AoA and (2) spinal segmental arteries in its vicinity. Material and Methods: A systematic review of the spine and vascular surgery literature utilizing 5 online databases, with articles published between January 1, 1950 and April 30, 2018 was carried out in accordance with PRISMA guidelines. Articles were included if an angiographically confirmed AoA was ligated, or if an explicitly stated number or range of SSAs in its vicinity was occluded. Study design, level/laterality of AoA, level/number of SSAs were recorded. Outcomes in terms of the crude rate of postoperative motor neurological deficit with occlusion of the AoA, bilateral SSAs, or unilateral SSAs were analyzed. Results: Ten articles, all retrospective case series, were included in this review. Three studies with 50 unique patients with verified occlusion of the AoA demonstrated a crude cumulative postoperative neurological deficit rate of 4.0%. Occlusion of the AoA together with up to 7 pairs of SSAs have not been reported to cause a postoperative motor neurological deficit. When 1 to 6 pairs (ie, bilateral) of SSAs (not including the AoA) were ligated, the crude postoperative neurological deficit rate was 0.6%, as compared with 5.4% when more than 6 bilateral pairs of SSAs were ligated. For unilateral ligation of SSAs of 2 to 9n levels, the crude rate of postoperative neurological deficit does not exceed 1.3%. Conclusion: The current best evidence indicate that occlusion of (1) the AoA with up to 7 contiguous pairs of SSAs and (2) up to 6 pairs of SSAs (without angiographic knowledge of the AoA’s location), is associated with a negligible to low risk of postoperative neurological deficits.

A292: Effect of Chewing Gum in Mild-Moderate Dysphagia After Anterior Cervical Fusion: A Randomized Controlled Trial

Guoyan Liang¹ and Yunbing Chang¹

¹Guangdong General Hospital, Guangzhou, China

Introduction: Postoperative dysphagia is the most common complication of anterior cervical fusion surgery (ACF). However, there is no specific procedure in enhanced recovery after surgery (ERAS) program for reliving dysphagia after ACF. Previous studies showed that chewing gum helps improving swallow frequency and latency, but its efficacy on dysphagia after ACF is still unknown. In this study, we aim to assess the early effect of chewing gum in relieving mild-moderate dysphagia after ACF. Material and Methods: A single-center randomized controlled trial (RCT) was performed (clinical trial number: NCT03524703). The patients, who were diagnosed with mild or moderate postoperative dysphagia after ACF using Bazaz scale, were randomized into 2 groups. The control group was treated with standard care, and the other group was treated with chewing gum 3 times a day for 15 minutes in addition to standard care. Dysphagia Symptom Questionnaire (DSQ) score and Bazaz scale at 3, 5, 7 days, and 1 month after surgery were achieved to compare the degree of dysphagia between groups. Cervical spine lateral view was assessed for the degree of prevertebral soft-tissue swelling from C2 to C7. Results: A total of 84 patients were enrolled: 42 of them were assigned to the gum-chewing group and others were control group. The 2 groups were nearly identical in terms of their demographic and operative characteristics. The degree of dysphagia relieving at 5 days after surgery proved to be significantly larger in gum chewing group (ΔDSQ = 2.75) than in control group (ΔDSQ = −0.2, P < .05). Similar trend was also found from the results of other timepoints. Prevertebral soft-tissue swelling was slightly lower in the gum-chewing group, but not statistically significant. No complications related to chewing gum was found at the final follow-up. Conclusion: Chewing gum appears to be an effective and safe method for relieving mild to moderate dysphagia early after anterior cervical fusion surgery. Our finding suggested chewing gum may be an important element of ERAS procedure for ACF surgery.

A293: The Relationship Between Intraoperative Neurophysiological Monitoring Alert During Different Surgical Steps and Neurological Recovery: A Long-Term Follow-up Study

Yangliang Huang¹, Yong Hu¹, Martin Chan¹, Yat-Wa Wong¹, Kenneth M. C. Cheung¹, and Kenny Kwan¹

¹The First Affiliated Hospital, Sun Yat-Sen University, Guangzhou, China

Introduction: Transcranial motor potentials (TcMEP) and somatosensory-evoked potentials (SSEP) are utilized as gold standard of multimodal monitoring during spinal surgery; however, a few of neurological injury are not prevented by compensational procedures after intraoperative neurophysiological monitoring (IONM) alert. The aim of this study was to determine whether IONM alert during particular surgical steps affected long-term neurological recovery. Material and Methods: A retrospective review of prospectively collected data in 2876 consecutive spinal surgical patients from July 1998 to July 2018. IONM was used in deformity correction, adult cervical and thoracic myelopathy decompression, and cervical or thoracic fracture. Standard patient demographics, diagnosis, and operative features and IONM data were collected. Significant postoperative neurological deficit was defined as permanent loss of muscle power no less than 1 grade. Incompetent of walking was defined as standing difficulty due to weak muscle power even after a long time of rehabilitation. Screw insertion, reduction of fracture, cement injection and insertion of wire were defined as invasive surgical steps. Decompression, anterior disc release and finding the entering point of pedicle were defined as noninvasive surgical steps. Results: The number of IONM alerts occurred in 24 patients: 12 in scoliosis correction, 4 in kyphosis correction, 6 in adult cervical and thoracic myelopathy decompression, 2 in fractures fixation. 10 patients developed permanent neurological deficit at a mean of 9.8 years of follow-up. These alerts occurred during decompression (n = 3), anterior disc release (n = 1), finding the entering point of T3 pedicle (n = 1), screw insertion (n = 1), reduction of fracture (n = 2), insertion of wire (n = 1), and cement injection (n = 1). Among these patients, 100% of invasive procedure patients developed incompetent of walking, while 80% of non-invasive patients were capable of walking. There was significant difference between these two groups (P < .05). Conclusion: Although IONM alerts can help surgeons detect potential neurological damage, this study showed that IONM alerts during invasive surgical steps within an operation were more likely to result in permanent neurological deficit. Surgeons should maximize the use of IONM during these steps to improve the safety of spinal procedures.

A294: Wear and Corrosion in Cobalt Chrome and Titanium Implants in Scoliosis Correction: An In Vitro Analysis With Validation in Explanted Instrumentation

Oded Hershkovich¹, Edward Bayley¹, and Wai Weng Yoon¹

¹Queen’s Medical Centre, Nottingham, UK

Introduction. Metallosis has come to prominence in arthroplasty surgery following the widespread use of metal-on-metal hip replacements. It is not just the articulating surface but also the head-taper junction (similar to the screw-rod interface in spinal surgery) that has been shown to produce metal debris. This leads to soft tissue reaction and pseudotumour formation, with loosening of implants and large soft-tissue defects. The systemic effects of metallosis are as yet unknown. Spinal arthrodesis may be performed using dissimilar metals to exploit the best properties of each (cobalt chromium [CoCr] and titanium [Ti]); however, this has the potential to cause galvanic corrosion and result in metallosis in the soft tissues. There is debate regarding the clinical relevance of galvanic corrosion and limited data regarding its existence in spinal instrumentation. The aim of this study was to determine whether a galvanic current, and therefore corrosion, exists in in vitro spinal instrumentation models using cobalt chrome with titanium versus titanium only constructs (part 1), and to examine explanted spinal metalwork to look for evidence of galvanic corrosion (part 2). To the best of the authors’ knowledge, the issue of galvanic corrosion in spinal instrumentation has not previously been investigated in this manner. Materials and Methods: Part 1—Six spinal instrumentation in vitro models were set up in 2 groups. The control group comprised of titanium rods and screws (Ti/Ti). The test group comprised cobalt chrome rods with titanium screws (CoCr/Ti). The implants were loaded according to the ASTM F2193 standard specifications, cyclically loaded, and any electrical current generated recorded. Part 2—Explanted spinal metal work from 3 patients with unexplained back pain was examined under a scanning electron microscope (SEM) for evidence of galvanic corrosion. Results: Part 1—In the in vitro spinal instrumentation models, the mean maximal fretting current and the mean amplitude current for CoCr/Ti was 14.31 and 7.93 µA, respectively. The mean maximal fretting current and the mean amplitude current for Ti/Ti was 12.22 and 5.89 µA, respectively. The difference was not statistically significant. Part 2—SEM examination of the explanted CoCr rods demonstrated fretting (mechanical) corrosion and crevices and pitting (galvanic) corrosion. Tissue histological examination demonstrated significant chromium debris. Conclusion: Although not statistically significant, the in vitro model suggested that a galvanic current, and therefore galvanic corrosion, occurred with CoCr/Ti constructs. SEM examination of explanted spinal metalwork demonstrated evidence of mechanical and galvanic corrosion as well as metallosis. As metal-on-metal implants containing CoCr used in other areas of the body have raised health concerns from their metallic debris, the results of this study suggest that further research should be conducted to investigate the parameters affecting galvanic corrosion, and the potential health implications for patients receiving CoCr/Ti spinal implants.

References

 1. Hedequist DJ. Instrumentation and fusion for congenital spine deformities. Spine (Phila Pa 1976). 2009;34:1783-1790. doi:10.1097/BRS.0b013e3181ab62b3

 2. Yeung KW, Wong KH. Biodegradable metallic materials for orthopaedic implantations: a review. Technol Health Care. 2012;20:345-362. doi:10.3233/THC-2012-0685

 3. Fraker AC. Corrosion of metallic implants and prosthetic devices. In: Metals Handbook. Vol. 13, 9th ed. Metals Park, OH: ASM International; 1987:1324-1335.

 4. Collier JP, Surprenant VA, Jesen RE, Mayor MB. Corrosion at the interface of cobalt-alloy heads on titanium-alloy stems. Clin Orthop Relat Res. 1991;(271):305-312.

 5. Crowninsheild RD, Maloney WJ, Wentz DH, Humphrey SM, Blanchard CR. Biomechanics of large femoral heads: what they do and don’t do. Clin Orthop Relat Res. 2004;(429):102-107.

 6. Smith TM, Berend KR, Lombardi AV Jr, Emerson RH Jr, Mallory TH. Metal-on-metal total hip arthroplasty with large heads may prevent early dislocation. Clin Orthop Relat Res. 2005;441:137-142.

 7. Langton DJ, Jameson SS, Joyce TJ, Webb J, Nargol AV. The effect of component size and orientation on the concentrations of metal ions after resurfacing arthroplasty of the hip. J Bone Joint Surg Br. 2008;90:1143-1151. doi:10.1302/0301-620X.90B9.20 785

 8. Haddad FS, Thakrar RR, Hart AJ, et al. Metal-on-metal bearings: the evidence so far. J Bone Joint Surg Br. 2011;93:572-579.

 9. Panagiotidou A, Meswania J, Hua J, Muirhead-Allwood S, Hart A, Blunn G. Enhanced wear and corrosion in modular tapers in total hip replacement is associated with the contact area and surface topography. J Orthop Res. 2013;31:2032-2039. doi:10.1002/jor.22 461

10. Jacobs JJ, Gilbert JL, Urban RM. Corrosion of metal orthopaedic implants. J Bone Joint Surg Am. 1998;80:268-282.

11. Kasai Y, Iida R, Uchida A. Metal concentrations in the serum and hair of patients with titanium alloy spinal implants. Spine (Phila Pa 1976). 2003;28:1320-1326.

12. Kim YJ, Kassab F, Berven SH, et al. Serum levels of nickel and chromium after instrumented posterior spinal arthrodesis. Spine (Phila Pa 1976). 2005;30:923-926.

13. Cundy TP, Antoniou G, Sutherland LM, Freeman BJ, Cundy PJ. Serum titanium, niobium, and aluminum levels after instrumented spinal arthrodesis in children. Spine (Phila Pa 1976). 2013;38:564-570. doi:10.1097/BRS.0b013e3182741961

14. Scharf B, Clement CC, Zolla V, et al. Molecular analysis of chromium and cobalt-related toxicity. Sci Rep. 2014;4:5729. doi:10.1038/srep05729

15. Serhan H, Slivka M, Albert T, Kwak SD. Is galvanic corrosion between titanium alloy and stainless steel spinal implants a clinical concern? Spine J. 2004;4:379-387.

16. Kawalec JS, Brown SA, Payer JH, Merritt K. Mixed-metal fretting corrosion of Ti6A14V and wrought cobalt alloy. J Biomed Mater Res. 1995;29:867-873.

17. Urban RM, Jacobs JJ, Tomlinson MJ, Gavrilovic J, Black J, Peoc’h M. Dissemination of wear particles to the liver, spleen and abdominal lymph nodes of patients with hip or knee replacement. J Bone Joint Surg Am. 2008;82:457-476.

18. Cook S, Asher M, Lai SM, Shobe J. Reoperation after primary instrumentation and fusion from idiopathic scoliosis. Toward defining late operative site pain of unknown cause. Spine (Phila Pa 1976). 2000;25:436-438.

A295: Differences in Presentation and Revision Rates of Unilateral and Bilateral Rod Fractures in Long, Thoracolumbar Fusions for Adult Spinal Deformity

Thamrong Lertudomphonwanit¹, Keith Bridwell², Michael Kelly², Prachya Punyarat³, Alekos Theologis⁴, Brenda Sides², and Munish Gupta²

¹Faculty of Medicine Ramathibodi Hospital, Mahidol University, Bangkok, Thailand

²Washington University in St Louis, St Louis, MO, USA

³Thammasat University, Pathumthani, Thailand

⁴University of California, San Francisco, San Francisco, CA, USA

Introduction: Rod fracture is a frequent occurrence after adult spinal deformity (ASD) surgery. However, as not all rod fractures (RFs) cause clinical concern and the current literature on RFs after ASD surgery is limited. An improved understanding of the sequelae after RF occurrence is valuable for further management. Radiographic identification of a clinically relevant rod fracture or symptomatic pseudarthrosis requiring revision surgery will assist in the clinical management of ASD patients. This study assesses the differences in radiographic findings, clinical outcomes and revision rates between ASD patients who sustained unilateral rod fracture (URF) or bilateral rod fracture (BRF). Material and Methods: We performed a retrospective analysis of adults who underwent posterior spinal instrumented fusion (including sacrum; ≥5 levels) from 2004 to 2014 at a single center. Patients were divided into 2 groups based on number of rod fractures, specifically URF vs BRF. Demographics, operative data, radiographic parameters, clinical outcomes and revision rates were compared between groups at baseline and follow-up. Results: Of 526 patients who met inclusion criteria, 96 (18.3%) developed rod fracture (URF n = 70 [73%]; BRF n = 26 [27%]). The majority of the cohort was female (82.3%), had undergone a previous spine surgery (52.1%) and had an ASA (American Society of Anesthesiologists) score of 2 (79.2%). The average duration of follow-up was 64.1 months (median 61.2 months, range 16-133 months). Average number of levels fused was 14. The majority of operations included two 5.5-mm rods, interbody fusions, pelvic fixation, fresh frozen morselized allograft and rhBMP-2. Preoperative demographics and surgical parameters were similar between the groups. At the time of rod fracture occurrence, BRF patients had more positive sagittal vertical axis (SVA: 7.8 vs 4.3 cm; P = .01), less lumbar lordosis (LL: −34.5° vs −46.9°; P = .001), and greater pelvic incidence minus lumbar lordosis mismatch (PI-LL: 22.5° vs 11.6°; P = .02) than URF patients. The BRF group underwent revision surgery more often than the URF group (84.6% vs 18.6%; P < .0001). The displaced BRF group (a subset of the BRF cohort) had a greater percentage of patients who underwent revision surgery than the nondisplaced URF group (a subset of the URF cohort) (87% vs 5%, P < .0001). The BRF group had significantly less improvement in Oswestry Disability Index (ODI) scores (6.7 vs 15.3; P = .02) and Scoliosis Research Society (SRS) subscore (0.3 vs 0.7; P = .003) at the time of rod fracture than the URF group. At final follow-up (≥1 year after rod fracture), the URF patients who did not undergo revision surgery still restored equivalent sagittal alignment correction (SVA = 4.9 cm, LL = −45.1°, and PI-LL = 13.8°, all Ps >.05) and had no difference in ODI scores (21.7 vs 21.9; P = .96) and SRS subscore (3.8 vs 3.7; P = .58) compared with the time at rod fracture. Conclusion: Rod fractures are not uncommon after ASD operations. Asymptomatic, nondisplaced unilateral rod fracture, in our study, did not jeopardize clinical outcomes or radiographic alignment parameters. Bilateral rod fracture patients exhibited more severe sagittal spinopelvic malalignment, less improvement in clinical outcomes, as measured by ODI and SRS subscore and more often came to revision operations than unilateral rod fracture patients.

A296: Machine Learning Can Identify Postoperative Complications Following Posterior Spinal Fusion Using Medical Notes

Varun Arvind¹, Brian Cho¹, Deepak Kaji¹, Ivan Ye¹, Ray Tang¹, Andrew Warburton¹, Aly Valliani¹, Daniel Ranti¹, Zoe Cheung¹, Jun Kim¹, and Samuel Cho¹

¹Icahn School of Medicine at Mount Sinai, New York, NY, United States

Introduction: Quality improvement studies have provided surgeons with systematic insight on how to make changes to their practice to improve patient outcomes and avoid complications. However, currently most automated methods to identify patient complications for quality improvement studies rely on prelabeled administrative codes sampling data from fixed electronic medical databases. Medical notes contain a rich supply of medical data, yet the format of unstructured text has precluded this data from use in large retrospective analyses. Therefore, the utilization of free-text from electronic medical records using natural language processing in combination with machine learning can provide an additional tool for patient complication surveillance and identification. The purpose of this study was to develop a machine learning algorithm that can identify patients with common spine surgery complications based on unstructured patient notes. Material and Methods: This was a retrospective analysis of medical notes from the Medical Information Mart for Intensive Care (MIMIC III), a large, single-center database comprised of intensive care unit admissions data from 2001 to 2012. Spinal surgery complications include pneumothorax, pseudoarthrosis, spinal cord injury, cardiac, respiratory, vascular, central nervous system, hematoma formation, hemorrhage, infection, and cerebrospinal fluid leak complications as identified by ICD-9 codes. Natural language processing was used to analyze unstructured narrative notes including physician, consult, discharge, radiology, nursing, pharmacy, and social work notes. The notes were preprocessed using natural language processing methods and an auto-correct function was performed. A support vector machine (SVM) algorithm was used train a classifier to automatically identify patients with complicated posterior spinal fusion based solely on the notes. The data was randomized with 70% used for training and 30% used for testing. Results: A total of 571 patients from the orthopedic service were identified as undergoing posterior spinal fusion (ICD-9: 8100, 8105, 8107, 8108), with 219 patients experiencing spinal surgery complication (38%). A total of 63% of patients were male with 83% of cases being emergent and 16% elective. In all, 68% of patients were Medicare insured, 21% privately insured, and the rest insured through government insurance, Medicaid or self-pay. Area under the receiver operating characteristic curve (ROC) were evaluated to assess performance of the model. The SVM algorithm identified spinal complications with an area under ROC of 0.64. Conclusion: Combining natural language processing with machine learning is a valid approach for automatic identification of complications following posterior spinal fusion using unstructured text from discharge notes. This method provides a robust survey of surgical encounters and describes a method that may be used for real-time surveillance of postoperative complication risk and for automated identification of patient complications for quality improvement studies.

A297: The Complication Trends of Spinal Deformity Surgery in Japan: Research by the Japanese Scoliosis Society Morbidity & Mortality Committee

Ryo Sugawara¹, Katsushi Takeshita¹, and Hirokazu Inoue¹

¹Jichi Medical University, Tochigi, Japan

Introduction: The worldwide study focusing on the complication survey of spinal deformity surgery, known as Morbidity & Mortality (M&M) Research, have been performed by Scoliosis Research Society (SRS). In Japan, the Japanese Scoliosis Society (JSS) created a longitudinal complication survey of spinal deformity surgery, and established the Morbidity & Mortality (M&M) Committee in 2012. The purpose of this study was to analyze the results of the complication survey in 2012 and 2014, and to report the recent complication trends of spinal deformity surgery in Japan. Methods: A request to participate in this survey was mailed to all JSS members. A questionnaire, which was produced based on the basic survey items of the M&M research by the SRS in 2012 with an addition of complications and modification for practical use in Japan, were sent by email to members who agreed to cooperate in this survey. The survey items included the total number of patients who underwent spinal deformity surgery and surgical complications within each diagnostic group. Diagnosis was grouped into idiopathic scoliosis, congenital scoliosis, neuromuscular scoliosis, spondylolisthesis, pediatric kyphosis, and adult spinal deformity. Complication was grouped into death, blindness, neurological deficit (motor or sensory deficit), infection, massive bleeding, hematoma, pneumonia, cardiac failure, deep venous thrombosis/pulmonary embolism (DVT/PE), gastrointestinal perforation and instrumentation failure. The M&M research team tabulated total case numbers and calculated the complication rate. Results: The total number of cases were 2906 in 2012 and 2129 in 2014. The overall complication rate increased yearly from 10.4% in 2012 to 15.3% in 2014. There were no differences between 2012 and 2014 in complication rates of death, blindness, infections, massive bleeding, postoperative pneumonia, cardiac failure, DVT/PE, gastrointestinal perforation and instrumentation failure. The rate of neurological deficit, especially in motor deficit, decreased significantly in patients with idiopathic scoliosis, and increased in older patients with adult spinal deformity in 2014. Massive bleeding remained more common in patients with neuromuscular scoliosis and adult spinal deformity in 2014. Instrumentation failure was also more common in patients with adult spinal deformity, especially in patients aged 40 to 65 years. Conclusions: The major complication trends of spinal deformity surgery in Japan were an increasing rate of neurological deficit and instrumentation failure, especially in adult spinal deformity.

A298: Complications of Spine Surgery in “Super Obese” Patients

Gennadiy Katsevman¹, Scott Daffner¹, John France¹, Sanford Emery¹, Shari Cui¹, and Cara Sedney¹

¹West Virginia University, Morgantown, WV, USA

Introduction: “Super obese” patients (defined as body mass index [BMI] ≥50 kg/m²) likely experience different risks in spine surgery when compared with patients of lesser weight. Although several studies have compared obese with nonobese patients, these studies group all obese patients (BMI > 30 kg/m²) together. There is a lack of research specifically regarding spine surgery in patients with class 4 or 5 obesity. The purpose of this study was to evaluate the risks and complications associated with spine surgery on super obese (class 4 or 5 obesity) patients. Materials and Methods: All patients undergoing spine surgery at a single academic medical center with a BMI ≥ 50 kg/m² from 2012 to 2016 were retrospectively reviewed. Based on pre hoc power analysis, a 1:1 ratio of age-matched control patients of BMI < 50 kg/m² who underwent spine surgery during the same time period were selected for comparison. Super obese and control patients were initially identified through review of surgical billing records. Subsequently, a review of the electronic medical record was conducted to obtain demographic and clinical data. Age, gender, BMI, surgical indications, complications, mortality, and procedure types were recorded. The groups were compared with regard to complication rate as the primary outcome and mortality rate as the secondary outcome. Elective and nonelective cases were also considered separately. Categorical variables were compared using chi-square analysis with significance set as P < .05. Results: We identified 63 patients in the super obese group who underwent 86 surgeries and 60 patients (64 surgeries) in the control group. Age and the number of elective and nonelective cases were not significantly different between groups. Mean BMI of the super obese group was 55 kg/m² (range 50-77 kg/m²) compared with 29 kg/m² in the control group (range 20-49 kg/m²). Super obese patients had a higher complication rate (30% [n = 19] vs 10% [n = 6], P = .0055) but similar 30-day mortality rate (5% vs 5%) compared with controls. The most common complication among super obese patients was wound dehiscence/infection (n = 8, 13%); 2 patients’ surgeries were aborted. Only 1 patient in the control group had wound dehiscence. For both groups, the complication rate of nonelective surgery was greater than elective surgery, although not significantly so (super obese P = .1035, control P = .1944). Complication rates for elective surgery were 21% (n = 7) for super obese patients and 4% (n = 1) for controls (P = .0507); complication rates for nonelective procedures were 40% (n = 12) and 14% (n = 5), respectively (P = .015). All mortalities in both groups involved nonelective cases. Conclusions: Super obese patients should be considered separately from other patient populations with regard to spine surgery. The complication rate is significantly higher in this population, particularly for nonelective cases. Super obese patients should be counseled regarding these risks, and steps should be taken to mitigate risks whenever possible in this population.

A299: The Efficacy and Safety of Early VTE Chemoprophylaxis Within 24 Hours of Surgical Intervention After Spinal Cord Injury

Don Young Park¹, Seth Austin Ahlquist¹, Howard Young Park¹, and Nick Arya Shamie¹

¹UCLA Medical Center, Los Angeles, CA, USA

Introduction: Spinal cord injuries (SCIs) occur in approximately 54 individuals per million in the United States each year, resulting in significant disability and mortality. This population is at high risk for secondary complications, especially venothrombotic events (VTE), with an incidence greater than 50%. Current SCI guidelines recommend initiation of VTE chemoprophylaxis within 72 hours of injury. The purpose of this study is to investigate the effectiveness and safety of initiating chemoprophylaxis within 24 hours of surgical intervention for SCI versus after 24 hours. Material and Methods: A total of 97 patients underwent surgical intervention for acute traumatic SCI at a single level-1 trauma center from 2013 to 2018. Data from this cohort was retrospectively reviewed for timing of prophylaxis, VTE rates, and postoperative bleeding complication rates. Other variables obtained included age, gender, race, ASIA (American Spinal Injury Association) score, injury level, surgery performed, type of chemoprophylaxis used, medical complications, surgical complications, hospital/ICU length of stay, and disposition status. Exclusion criteria included age <18 years, hypercoagulable gene mutation, conus medularis/cauda equina syndrome, and those that did not receive anticoagulation or received it after 72 hours. Chi-square test of homogeneity was performed in order to establish significant differences between groups. Results: A total of 69 patients were identified for the study. Patient characteristics included an average age of 52.9 years (range 18-92 years), 81.2% male, 47.8% white, 65.2% cervical and 31.9% thoracic SCIs. There were 40.6% ASIA-A, 17.4% ASIA-B, 24.6% ASIA-C, and 17.4% ASIA-D injuries. A total of 89.8% received unfractionated heparin and 10.2% received low molecular weight heparin. Average hospital and ICU length of stay were 17.0 ± 12.1 and 10.7 ± 10.4 days, respectively. Patients who received chemoprophylaxis within 24 hours of SCI had a lower overall rate of VTE (6.1% vs 35.0%, χ² = 9.6, P = .005) and deep venous thrombosis (DVT; 4.1% vs 30.0%, χ² = 9.3, P = .006) than those who received it 24-72 hours after SCI. The rate of PE was not significantly different between the 2 cohorts (6.1% vs 5.0%, P = 1.0). There were no postoperative bleeding complications in either cohort, but there were 2 cases of wound dehisence in the early chemoprophylaxis group. Conclusion: VTE chemoprophylaxis is effective and associated with lower rates of VTE and DVT when initiated within 24 hours of surgical intervention after SCI. The early chemoprophylaxis strategy is also safe, with no postoperative bleeding complications seen in the early chemoprophylaxis group. This strategy may be translated clinically to other patient populations who are at risk for VTE and undergoing spinal surgery.

OP44: Novel Technologies 3

A300: 2-Year Results of a Randomized Controlled Trial Comparing Minimally Invasive Sacroiliac Joint Fusion Versus Conservative Management for Sacroiliac Joint Pain

Djaya Kools¹, Julius Dengler², Bengt Sturesson³, Robert Pflugmacher⁴, Domenico Prestamburgo⁵, and Alessandro Gasbarrini⁶

¹OLV Hospital Aalst, Aalst, Belgium

²Charité Hospital, Berlin, Germany

³Ängelholm Hospital, Ängelholm, Sweden

⁴University Hospital Bonn, Bonn, Germany

⁵ASST Ovest Milanese, Ospedale di Legnano, Legnano, Italy

⁶Instituto Ortopedico Rizzoli di Bologna, Bologna, Italy

Introduction: The sacroiliac joint can be recognized as the main source of pain in a part of the chronic low back and leg pain population. A previous randomized controlled trial conducted in the United States already showed a superior outcome of minimally invasive sacroiliac joint fusion using triangular titanium implants compared with conservative treatment in chronic sacroiliac joint pain patients. This European randomized controlled trial (RCT) reassessed the effectiveness and safety of minimally invasive sacroiliac joint fusion versus conservative management according to the European guidelines for treatment of pelvic girdle pain. Materials and Methods: The diagnosis of sacroiliac joint pain was confirmed by image guided diagnostic block of the joint with local anesthetics. In this multicenter trial, 103 adults with chronic sacroiliac joint pain were randomly assigned to either minimally invasive sacroiliac joint fusion (n = 52) or conservative management (n = 51). Crossover was allowed from conservative treatment to surgical treatment after 6 months of unsuccessful conservative treatment. Patients were evaluated after 1, 3, 6, 12, and 24 months. The study endpoints included self-rated low back and leg pain (visual analogue scale), disability (Oswestry Disability Index), quality of life (EQ-5D), and pelvic girdle function (active straight leg raising test). In the surgical group, computed tomography (CT) scan was performed directly after surgery and after 12 months. Results: In all, 90% of patients in both groups completed the 2-year follow-up. At 2 years, a statistically significant superior outcome was seen after surgical treatment compared with conservative management for all the study endpoints: back and leg pain, disability, quality of life, and pelvic girdle function. Mean improvement in low back pain was 45 points after surgery versus 11 points after conservative treatment (difference of 34 points, P < .0001). Opioid use decreased after surgery, while no significant change was seen after conservative treatment. Similar improvement after surgery was seen after crossing over from the conservative management group. More severe adverse events were seen in the surgical group, but only 4 were related to the study device or surgery procedure (2 cases of increasing sacroiliac joint pain, 1 case of postoperative hematoma, and 1 case of nerve impingement by an implant). The CT at 12 months showed radiolucencies around 8 implants (4%), but these findings were not associated with clinical symptoms. These results strikingly parallel the outcomes previously reported in the US RCT. Conclusion: Minimally invasive sacroiliac joint fusion using titanium triangular implants was a safe procedure and provided a superior clinical outcome in chronic sacroiliac joint pain patients compared with conservative treatment. It can be regarded as a reasonable treatment option.

A301: Results From a Retrospective Study of a Posterior Dynamic Deformity Correction Device for the Treatment of Adolescent Idiopathic Scoliosis

Yizhar Floman¹, Ron El-Hawary², Baron Lonner³, and Randal Betz¹

¹Department of Orthopaedics , Tel Aviv, Israel

²Department of Surgery, Halifax, Nova Scotia, Canada

³Department of Orthopaedics, New York, NY, USA

Introduction: Nonfusion solutions for addressing adolescent idiopathic scoliosis (AIS) are desirable. This posterior dynamic deformity correction device was developed to address single AIS curves. The objective of this retrospective, multicenter trial was to analyze the results of this new posterior dynamic deformity correction device in reducing and maintaining the correction of AIS. Material and Methods: A retrospective review of prospectively collected data on AIS patients with Lenke 1 and 5 curves who were treated with a posterior dynamic deformity correction device. The concavity of the major curve was instrumented with 2 pedicle screws that were connected to an expandable ratcheting rod with ceramic coating. Each device/screw connection allows 50° of freedom in all directions. Results: Of the 32 patients identified, 1 was lost to follow-up. A total of 31 patients with minimum 2-year follow-up had mean preoperative scoliosis of 49° (32°-74°), which improved to 34° (6°-64°) with correction of 32% (−6% to 81%) at final follow up (P < .01). The sagittal profile was minimally changed toward normal after surgery. Trunk shift was reduced by mean 10 mm (0-13 mm). The mean preoperative SRS-22 score improved from 3.9 ± 0.3 to 4.2 ± 0.4 at 2 years postoperative (P < .01). Subgroup analysis of those patients who had preoperative curves <60° and were flexible to ≤35°, which is the current indication for the device (n = 23), demonstrates that the mean preoperative scoliosis of 45° (32°-55°) improved to 29° (6°-46°) with correction of 37% (15%-81%) at final follow-up (P < .05). Conclusion: This study, with 2 to 4 years of follow-up, showed consistent curve improvement and stabilization in patients with curves smaller than 60° and flexibility to less than 35° on side bending views. These are our current indications for this device in AIS. It lends support to the concept that this posterior dynamic deformity correction device may be a viable alternative to fusion and failed bracing for managing AIS.

A302: Live Machine Learning Centered Architecture System to Optimize Lumbar Stenosis Surgical Outcome

Ghaith Habboub¹, Josh Johnston¹, Rod Nault¹, David Watson¹, Sebastian Salas-Vega¹, Mathew Grabowski¹, Michael Steinmetz¹, and Thomas Mroz¹

¹Cleveland Clinic, Cleveland, OH, USA

Introduction: Patients with neurogenic claudication who undergo lumbar laminectomy have varying degree of success.^1-4 Furthermore, it is difficult to identify patients who would benefit from surgery versus those who would not. Our goal in this study is to optimize patient selection, create a parallel artificial intelligence system that works seamlessly with the surgeon that helps with patient selection, and provide value-based care. Methods: We retrospectively reviewed 4121 cases of patients who underwent lumbar laminectomy at the Cleveland Clinic hospitals between the years 2007 and 2017. Of those, we identified ∼200 to 3300 patients who had enough data for the outcomes. Our objectives includes EuroQol 5-dimensional (EQ-5D), back pain visual analogue scale (VAS), right leg VAS, left leg VAS, patient health questionnaire (PHQ9), pain disability questionnaire (PDQ), readmission, venous thromboembolism, costs, and reimbursements. We used Python for analysis relying on Tensorflow, Keras, XGBoost, and Scikit-Learn for machine learning model creation.⁵ Results: Our baseline success rate, defined by EQ-5D minimal clinically important difference (MCID), was 50%. Using our system, we predict an improvement of at least to 75% in patient selection accuracy (50% improvement). This was validated retrospectively. We have analyzed >120 variables, which included demographics, comorbidities, medications, laboratory, quality, and functional data. The data showed unaware associations between variables, such as laboratory values and indirect-compliance measures, and outcomes. This provided us an opportunity to provide adjustment goals to the modifiable variables for the predicted failed surgical patients to become surgical candidates. We also described how we embedded the system into electronic medical records and daily provider routine (Figures 1 and 2). We created a simulated cost-saving analysis based on the above care path versus the current system. Conclusions: With the rising healthcare expenditures, it is necessary to center patient management on value-based medicine. This system can offer us a 50% improvement in patient selection accuracy and lower costs; hence, value-based medicine. Learning Objectives: By the conclusion of this session, participants should be able to: understand how this system can improve the success rate by 50% for surgical management of patients with neurogenic claudication caused by lumbar stenosis utilizing machine learning and big data, understand how to design a framework to embed complex machine learning system into electronic medical record, and understand how this system can provide value-based care by both optimizing outcome and minimizing costs.

References

1. Weinstein JN, Tosteson TD, Lurie JD, et al; SPORT Investigators. Surgical versus nonsurgical therapy for lumbar spinal stenosis. N Engl J Med. 2008;358:794-810.

2. Malmivaara A, Slätis P, Heliövaara M, et al; Finnish Lumbar Spinal Research Group. Surgical or nonoperative treatment for lumbar spinal stenosis? A randomized controlled trial. Spine (Phila Pa 1976). 2007;32:1-8.

3. Amundsen T, Weber H, Nordal HJ, Magnaes B, Abdelnoor M, Lilleås F. Lumbar spinal stenosis: conservative or surgical management? A prospective 10-year study. Spine (Phila Pa 1976). 2000;25:1424-1436.

4. Zaina F, Tomkins-Lane C, Carragee E, Negrini S. Surgical versus non-surgical treatment for lumbar spinal stenosis. Cochrane Database Syst Rev. 2016;(1):CD010264.

5. Chen T, Guestrin C. XGBoost: reliable large-scale tree boosting system. Paper presented at: The 22nd SIGKDD International Conference on Knowledge Discovery and Data Mining; August 13-17, 2016; San Francisco, CA.

Figure 1. Figure 2.

A303: Machine Learning (Ml) and Time Series Motion Classification: Is it Possible to Write Real-World, Single Sensor Algorithms to Detect the Activities of Daily Living Used Patient-Reported Outcome (PRO) Questionnaires?

Nicholas Benson¹, Andrew Fields¹, Arjun Siby Kurian¹, Newton Metcalf¹, Domagoj Coric², Shane Burch³, and Steven Glassman⁴

¹ Medtronic

²Carolina Neurosurgery & Spine Associates, Charlotte, NC, USA

³University of California, San Francisco, San Francisco, USA

⁴Norton Leatherman Spine Center, Louisville, KY, USA

Introduction: Clinical improvements in function are often measured using survey-based patient-reported outcomes (PROs; eg, Oswestry Disability Index [ODI]) with questions about walking, sitting, standing, traveling (driving) and sleeping (lying down). However, survey data is time consuming for both the patient and clinician. Multiple sensor and rules-based algorithms have shown positive predictive values greater than 85% for ability for walking, jogging, sitting, and lying down in a laboratory setting. The purpose of the current study is to determine positive predictive value of activity in a real-world setting of a single wearable sensor containing an extremely high-end inertial measurement unit (IMU). Material and Methods: Three healthy volunteers wore a single navigation-grade IMU (SBG Systems Ellipse2-N) for 60 minutes in a real-world (outside) environment and asked to sit, stand, lie supine, drive, and walk in an unscripted manner. Data was continuously recorded at 200 Hz using a combinatorial supervised ML approach. A multinomial logistic regression (MLR) was first used to classify posture with results fed to a second layer of support vector machines (SVM). A classifier finalized the results in 1-second boundaries by rolling 5-ms samples into a single vote. A 48-minute holdout data set was reserved for testing. Results: A total of 573 600 five-millisecond data points were analyzed. After the 2 layers (MLR followed by SVM), accuracy was found to be 99.65% across all activities. Positive predictive values were 99.24% for walking, 99.67% for sitting, 99.8% for driving and 100% for both lying and standing. Conclusion: The use of a high-end IMU and ML algorithms in this study to classify a subset of ODI activities in a real-world setting achieved greater accuracy than multiple sensors and rules-based algorithms reported in prior studies. This type of technology offers the potential for measured function to replace or supplement patient-reported function focused on established patient-centric activities routinely considered reflective of health status.

A304: Development of a Human Bone Metastases 3D Microenvironment Model

Antone Nour¹, Elie Akoury¹, Pouyan Ahangar¹, Michael Weber¹, and Derek Rosenzweig¹

¹McGill University, Montreal, Quebec, Canada

Introduction: The bone is a dynamic tissue that is constantly broken down and repaired through a balance between osteoclasts and osteoblast activity. However, when various cancers invade the bone, they can disrupt the physiological balance leading to various deleterious changes in bone structure and function. At the moment, radiotherapy, chemotherapy, and surgical resection are the main clinical approaches to bone metastases. Studying cancer cells molecularly in vitro relies on 2-dimensional (2D) monolayer cultures which in no way represents the physiological tissue microenvironment in vivo. Therefore, preclinical and translational cancer research trends are moving toward organoid studies and 3D biomimetic models to provide more clinically relevance in all aspects. Here, we set out to generate a bioink consisting of alginate, gelatin, nanocrystal hydroxyapatite loaded with primary human osteoblasts to produce a robust 3D bone-like microenvironment in order to study human bone metastasis. Material and Methods: A previously reported 3D hydrogel (1% alginate; 7% gelatin) model for cancer cell migration was modified to incorporate nanocrystal hydroxyapatite, primary human osteoblasts and primary human bone marrow derived stromal cells (7.5 × 10⁵ cells/mL). Primary osteoablasts were isolated from cadaveric vertebral bodies of organ donors, and bone marrow mesenchymal stem cells (MSCs) were purchased from Rooster Bio. The constructs were cultured for 28 days in either normal growth medium (Dulbecco’s modified Eagle medium [DMEM]) or osteogenic medium (OM) with and without 0.5 mg/mL hydroxyapatite (HA). Live/dead assays were performed to quantify viability at 28 days of culture. Fixed samples were cryo-embedded, and 12-µm sections were stained with Alizarin red for calcified matrix deposition, and hematoxylin and eosin to observe cells within the matrix. Results: Live/dead analysis revealed strong primary human osteoblast viability in all conditions after 28 days of culture: 91.3% ± 3.18% for DMEM/HA⁺, 92.5% ± 2.5% for DMEM/HA⁻, 88.6% ± 0.38% for OM/HA⁺, and 85.9 ± 6.2% for OM/HA⁻. Primary human MSCs showed approximately 85.1%, 69.4%, 87.8%, and 90% viability in the same 4 conditions, respectively. Alizarin red staining showed that cells grown in DMEM without HA had the least amount of bone mineralized matrix. The combination found to have the most amount of bone mineralized matrix was OM/HA⁺. Conclusion: Our data indicates that combination of 0.5 mg/mL nano-HA with OM produces a favorable microenvironment for primary human osteoblasts and MSCs to generate a bone-like matrix. Our preliminary work with low-cost bioprinting indicates this bioink is extrudable and will be ideal for screening therapeutics against patient-derived tumor cells cultured within a bone-like microenvironment. Ongoing work is assessing migration of fluorescently labeled tumor cells within this construct in the presence or absence of various chemotherapeutics. Furthermore, ongoing immunostaining for collagen type I, osteopontin and sclerostin will elucidate presence of bone matrix and osteocytes within the model. This work will allow better understanding of interactions between normal osteoblasts, stromal cells osteocytes, and patient-derived bone metastases cells while also being able to perform semi-high-throughput screening for novel therapeutics with a higher degree of clinical relevance.

A305: Automated Performance Feedback System Using Artificial Intelligence in a Virtual Reality Hemilaminectomy Simulation

Nykan Mirchi¹, Vincent Bissonnette^1,2, Ghusn Alsideiri^1,3, M. Recai Yilmaz¹, Alexander Winkler-Schwartz¹, Samaneh Siyar^1,4, Hamed Azarnoush^1,4, Nicole Ledwos¹, Robin Sawaya¹, Fahad E. AlOtaibi⁵, Abdulgadir Bugdadi¹, Khalid Bajunaid⁶, Greg Berry², Jean Ouellet², Bekir Karlik¹, and Rolando F. Del Maestro¹

¹Montreal Neurological Institute and Hospital, McGill University, Montreal, Quebec, Canada

^1,2Montreal General Hospital, McGill University, Montreal, Quebec, Canada

³Sultan Qaboos University Hospital, Muscat, Oman

⁴Amirkabir University of Technology (Tehran Polytechnic), Tehran, Iran

⁵King Fahad Medical City, Riyadh, Saudi Arabia

⁶University of Jeddah, Jeddah, Saudi Arabia

Introduction: As the field of surgical education gravitates from subjective to objective assessment, we are exploring the potential of virtual reality surgical simulators to both augment and facilitate this transition. Virtual reality simulators are particularly useful as they not only simulate the visual component of surgery but can also provide instant haptic feedback to the user. Additionally, they can track all movements and forces pertaining to the simulated instruments, thereby generating an enormous amount of data that can be further analyzed. Employing machine-learning methodology on such a dataset, we can obtain more insight on specific bimanual psychomotor skills which differentiate surgical expertise. One can then combine this information to develop a personalized feedback system for users. In this study, we assess the potential of linking expertise classification with a personalized feedback system using artificial intelligence during a spinal procedure from the NeuroVR simulator platform. Materials and Methods: Nineteen skilled participants (7 staff spine surgeons, 2 fellows and 10 PGY 4-6 residents) and 20 novices (6 PGY 1-3 residents and 14 medical students) took part in this study. Using the NeuroVR simulator, we developed a set of metrics able to differentiate the bimanual psychomotor performance of skilled and novice participants. Participants performed a virtual reality left L3 hemilaminectomy task utilizing a simulated operative burr and suction. Using statistical t tests, and a backward feature selection function in Matlab R2018a, we were able to select metrics capable of differentiating surgical expertise. An effort was made to select metrics that were intuitive from an educational perspective. We then employed the classification application in Matlab to train and validate a machine-learning classifier capable of differentiating expertise level (between expert and novice). We combined this classifier information with a unique educational system able to provide users with automated feedback on their metric performance with regard to expert proficiency performance benchmarks. These benchmarks were calculated as the region within one standard deviation of the mean for each metric for expert participants. Results: Using the NeuroVR dataset, over 400 metrics pertaining to safety, coordination, efficiency and motion, were generated. We reduced this number to 9 by using a combination of statistics (2-sample t test, P < .05) and backward feature selection. Using these nine metrics, we created a machine-learning algorithm (support vector machine) capable of differentiating expertise level with an accuracy of 97.4% (sensitivity 94.7%, specificity 100%). A visual feedback system was developed to provide participants with an immediate visual representation of their standing related to expert proficiency performance benchmarks. Conclusions: The educational system outlined establishes a basis for the potential role of integrating artificial intelligence and virtual reality simulation into surgical educational teaching. Machine learning algorithms are capable of identifying metrics that both differentiate levels of expertise and are teachable. The potential of linking expertise classification, objective feedback based on proficiency benchmarks, and instructor input creates a novel educational tool by integrating these three components into a formable educational paradigm.

A306: A Dynamic Growth Friendly Technique Based on a Spring Distraction System (SDS) for Early Onset Scoliosis

Sebastiaan P. J. Wijdicks¹, Justin V. C. Lemans¹, René M. Castelein¹, and Moyo C. Kruyt¹

¹University Medical Center Utrecht, Utrecht, Netherlands

Introduction: Growth-friendly systems are needed in severe early onset scoliosis (EOS) to stop curve progression while allowing the spine to grow. Distraction-based growth-friendly techniques are stiff and require repetitive lengthening’s with subsequent high stresses on the system and spine. Passive sliding growth friendly techniques on the other hand have less potency for correction and growth. To combine the advantages of each approach, we developed a dynamic growth friendly system that integrates continuous distraction with a sliding rod. The spring distraction system (SDS) uses the continuous distraction force of a compressed spring positioned around a sliding rod. The aim of this study was to investigate the 3D correction, the spinal growth, and the complication rate of our new dynamic growth-friendly technique. Material and Methods: Prospective cohort study of the first 20 patients treated with the SDS with at least 1-year follow-up. Adverse events were recorded and Cobb angle, sagittal alignment and spinal growth were measured on standard digital radiographs. Results: Mean age at surgery was 8.3 years and average follow-up was 1.5 years. The average frontal Cobb angle reduced from 59° to 36° and remained stable at 38° during latest follow-up (P = .13). Kyphosis decreased after surgery from 31° to 28° postoperatively. The kyphosis then increased significantly to an average 40° at latest follow-up (P = .02). Two patients improved from a congenital thoracic lordosis to a normal kyphosis at latest follow-up. The postimplantation T1-S1 spine growth rate averaged 10 mm/year. 4 patients required unplanned surgeries. No SDS failures, rod fractures, or screw pull-outs were observed. Conclusion: The SDS shows good surgical correction without the need for any lengthening procedures. The physiological spinal growth is maintained over time and the system has the potential to correct hypokyphosis. The dynamic spring-based system possibly prevents implant-related stress fracture seen in stiff other growth friendly systems. These first clinical results are promising but obviously long-term results and research are the next step.

A307: Vertebral Body Stenting: Early Results With a New Technique for The Treatment of Vertebral Compression Fractures

Abdulrazzaq Alobaid¹, Abdulaziz Al-Mutair², Islam Elsayed², and Tamer Ramadan¹

¹WARA Hospital, Kuwait

²Alrazi Hospital, Kuwait City, Kuwait

Background: Balloon kyphoplasty is an established minimally invasive procedure to restore vertebral body height, and internally stabilize osteoporotic and/or cancer-related vertebral compression fractures. At the same time, there is a technique-inherent problem with the amount of reduction; when filling the kyphoplasty balloons, an acceptable reduction is achieved in many patients. However, the resulting final reduction after cementing is about 25% lower due to the partial collapse after balloons deflation. To avoid this loss of reduction, a newer alternative technique has been introduced, based on the principles of balloon kyphoplasty and vascular stenting. Study Design: A prospective consecutive cohort study of clinical and radiographic results after vertebral body stenting for treatment of vertebral compression fractures. Objectives: To evaluate the safety and effectiveness of vertebral body stenting as a new technique to treat symptomatic vertebral compression fractures. Methods: A total of 3 3 elderly patients with 52 symptomatic vertebral compression fractures were enrolled in a prospective study of vertebral body stenting. Clinical outcomes were measured pre- and postoperatively using the visual analogue scale (VAS), Oswestry Disability Index (ODI), and ambulatory status (AS). All outcomes were assessed before the procedure, and at 1, 12, 24, and 36 weeks after the procedure. Results: The median VAS scores went from 10.0 preoperatively to 6 at 1 week postoperatively, to 5 at 12 weeks following the procedure and to 2 and 1 at 24 weeks and last follow-up, respectively. The preoperative ODI score was 80 before the operation, improved to 45 at 1 week postoperatively, decreased to 35 at 12 weeks and to 22 and 18 at 24 weeks and last follow up, respectively. The ability to move independently and ease of ambulation significantly improved after the procedure (P < .001). The median kyphosis angle was 15.0° before the procedure and decreased by a median of 4.5° after the operation. Conclusions: Elderly patients with symptomatic vertebral compression fractures had rapid, significant, and sustained improvements in back pain, back function, and quality of life following vertebral body stenting.

A308: Scan & Plan Strategy With BodyTom Portable CT Scanner and Renaissance Surgical Robot

Dmitry Dzukaev¹, Anton Borzenkov¹, Aleksander Pejker¹, Vladimir Gulyy¹, and Andrey Aksenov¹

¹Moscow Spine Center, Moscow, Russian Federation

Introduction: Robotics in spinal surgery has now become routine. However, a standard technique has certain limitations: limited use in the thoracic and cervicothoracic spine, accuracy in patients with osteoporosis and high body mass index (BMI), the difficulty of combining fluoroscopy data and computed tomography (CT) planning. In combination, portable scanner BodyTom and surgical robot Renaissance help bypass these restrictions. For the first time, we proposed and developed the routine protocol of robotic spine surgery “Scan & Plan” that includes intraoperative imaging with BodyTom and transpedicular fixation using Renaissance robot. Material and Methods: The procedure with BodyTom-Renaissance under “Scan & Plan” protocol. The patient is positioned on a radiolucent table in a specific position for the planned procedure. We set a unique target for registration specifically designed to combine with the BodyTom. The patient gets his or her spinal imaging right here, in the operating room, lying down on the operating table, in real time (no risk of positioning bias). Imaging of affected spine is transferred directly to the Renaissance station. A virtual operation plan is created, and surgical robot performs screw implantation. After implantation, we performed control CT. Results: We treated 78 patients, 22 of them with spine injuries, 56 with degenerative spine disc lesions, and 9 patients with spinal anomalies that significantly complicate screw implantation on standard technique, or even make it impossible (posttraumatic thoracic deformation, accompanied by spondyloptosis). For the implementation of this method, we tested and introduced a new target for Renaissance, allowing direct import of BodyTom imaging data and procedure planning with these dataplanning. This protocol was tested in the cervicothoracic, thoracic, lumbar, and lumbosacral spine, as well as in cervical spine, both for traumatic and degenerative spine lesions. Conclusion: This new protocol “Scan & Plan” makes it possible to perform procedures, which up to now seemed as extremely difficult or even impossible, including obese patients with high BMI, procedures in cervicothoracic and thoracic spine, complicated deformities, and congenital anomalies. This protocol allows accurate implantation even in most complex cases, as well as standardize the quality of procedures, regardless of the experience of the operating surgeon. It diminishes the risk of imprecise implant positioning, and so decreases the need for revision surgery.

OP45: Degenerative-Cervical 6

A309: Guiding the Expectations of Mortality in Octogenarians and Nonagenarians in Cervical Spinal Surgery Using the Charlson Comorbidity Index

Adeel Ikram¹, Rajendranadh Bommireddy¹, and Zdenek Klezl¹

¹University Hospital of Derby and Burton, Derby, UK

Introduction: Selecting candidates for cervical spinal surgery is currently based on the clinical need of the patient. Patients must be informed of likely potential complications, to aid the decision-making process whether to proceed with surgical intervention. The Charlson Comorbidity Index (CCI) used in medicine, predicts 10-year survival based on patient comorbidities. With an aging population, our study aims to identify whether this index can be used to help guide patients aged 80 years and older understand the increasing risks of mortality that surgery may pose. Material and Methods: A retrospective analysis was performed in patients aged 80 years and older, who underwent cervical spinal surgery at our institution from 2006 to 2018. Inclusion criteria included patients who underwent surgery for trauma, spinal tumours, and degenerative disease (including myelopathy). Exclusion criteria included patients notes being unavailable and isolated Halo application. The CCI was calculated at preoperative assessment and the relationship to perioperative, 6-month and 2-year mortality, length of stay, approach, instrumentation, and destination of discharge observed. Results: Our study analyzed 86 patients. 13 patients were excluded due to the exclusion criteria. With an average age of 84.1 ± 4 years, 34 were male and 52 female. The average length of stay was 24 ± 25 days with a large variability. Indications for surgery included 15 fractures, 3 spinal tumours, and 68 stenosis and myelopathy. A total of 67 had posterior approach, 16 anterior, and 3 combined. From our data, there were 5 perioperative mortalities, with a 60:40 split between myelopathic and trauma cases, respectively. Of the total patients, 14% suffered mortality within 6 months, and 27% of patients within 2 years. Two patients survived longer then 10 years both with a CCI of 4 with no medical comorbidities. Observing 2-year mortality, based on the CCI; 36% of CCI score 4 patients died, 33% of CCI score 5, 50% CCI score 6, and 80% of CCI score 7 and above (57% still within 2 years of surgery). Statistical analysis using the 1-way analysis of variance (ANOVA) test demonstrated a clinically significant correlation between the CCI and the number of days survived following surgery, f-ratio value 4.012, P = .010. A total of 68 patients went back to their own home (which included the deceased), 5 went to rehabilitation prior to going home, 4 went into a residential home, and 9 to a nursing home, with no statistical difference between CCI and discharge destination P = 0.528. Conclusion: Although the prediction of mortality remains a difficult task, one can infer from our study that patients should be informed that with a greater CCI their risk for mortality increases further should they undergo cervical spinal surgery. This should be reflected in the informed consent. A CCI score of 4 where a patient scores solely for their age with no medical comorbidities would normally demonstrate a 53% 10-year mortality; in our study it is 8%. More comorbidity points lead to steep increase in mortality and we therefore advocate serious discussion with the patients of CCI score 7 or greater regarding their expectations.

A310: Diagnostic Delays in Degenerative Cervical Myelopathy (DCM) Are Associated With Greater Disability and Represent a Health Inequality

Daniel Pope¹, Benjamin Davies¹, Oliver Mowforth¹, and Mark Kotter¹

¹University of Cambridge, Cambridge, UK

Introduction: Patients report significant delays in diagnosis of degerative cervical myelopathy (DCM), which is known to limit surgical outcomes. Current diagnostic delays and their contributory factors are poorly quantified in the literature. This study examines diagnostic delay and explores the socioeconomic factors that contribute and the effects of delays on patient outcomes. Material and Methods: Data was gathered via online survey for 778 patients with self-reported DCM. Missing values analysis was performed to consider bias. Time to diagnosis (0-6 months, 7-12 months, 1-2 years, 2-5 years, >5 years) was compared with outcome metrics (mJOA [modified Japanese Orthopaedic Association] disability score, support dependence for activities of daily living, employment status) using chi-square test and post hoc multiple z tests with Bonferroni correction for significant results. A cumulative odds ordinal logistic regression was run to determine the effects of age, sex, country of residence, ethnicity, household income, and educational status on the delay in diagnosis of DCM. Results: Most patients waited (431, 55.4%) more than a year for a diagnosis, with more than a fifth waiting more than 5 years (Figure 1). Patients waiting longer for a diagnosis had lower mJOA scores (P = .001). Longer delays were also associated with greater dependence on others (P = .040) and greater likelihood of unemployment due to disability (P < .0005). Ethnicity had a statistically significant effect on delay in diagnosis. Black patients (n = 25) faced longer delays than white patients (n = 690) (adjusted odds ratio 3.616, 95% CI 1.066-12.261, χ²(1) = 4.256, P = .039) and Asian patients (n = 18) (adjusted odds ratio 16.247, 95% CI 2.646-99.773, χ²(1) = 9.064, P = .003, Figure 2). Age, sex, educational status, country of residence, and household income had no statistically significant effects on delay in diagnosis. Conclusion: Diagnostic delays are associated with worse outcomes as assessed by clinical severity scores and functional status. Patients who face long delays are more likely to be support dependent and more likely to face employment issues. Black patients disproportionately face longer delays in diagnosis, a finding consistent with the wider health care inequality literature.

A311: Using Machine Learning to Predict Surgical Outcomes in Degenerative Cervical Myelopathy

Omar Khan¹, Jetan Badhiwala², and Michael Fehlings²

¹University of Toronto, Toronto, Ontario, Canada

Introduction: Degenerative cervical myelopathy (DCM) is the most common cause of spinal cord dysfunction worldwide, and is caused by a series of degenerative changes in the cervical spine that cause spinal cord compression. Treatment consists of surgical decompression for patients with moderate to severe myelopathy. However, the management for mild myelopathy is controversial since the response to surgery is more heterogeneous. Here, we used machine learning algorithms to determine which groups of mild myelopathy patients benefitted most from surgery. Materials and Methods: This is a post hoc subgroup analysis of a multicenter prospective cohort study using the AOSpine CSM (cervical spondylotic myelopathy) clinical trials. A total of 193 patients with mild DCM at baseline were enrolled. The outcomes of interest were the Short Form-36 (SF-36) mental component summary (MCS) and physical component summary (PCS), patient-reported quality-of-life outcome scores obtained before and 1 year after surgery. The change from baseline in the MCS and PCS scores at 1 year was computed and dichotomized according to whether the change exceeded the minimal clinically important difference (MCID). Next, the 193-patient dataset was split into training and testing sets (75% training: 25% testing). Four machine learning algorithms (classification trees, support vector machines, random forests, k-nearest neighbor) were trained and optimized using the training set, and their performances evaluated using the testing set. Results: All 4 machine learning algorithms performed well on the testing set, with areas under the receiver operating characteristic curves (AUCs) ranging from 0.72 to 0.80 for predicting changes in the MCS. The performance was similar for predicting changes in the PCS, with AUCs between 0.69 and 0.77 across the four models. The classification trees also provided information about patient phenotypes associated with the best postoperative outcomes. For instance, patients with a high MCS or PCS prior to surgery were less likely to improve in that same category at 1 year. In addition, female patients (particularly older females) with a low initial MCS were less likely to experience significant improvement in mental health than males. Furthermore, patients with a very low PCS were more likely to significantly improve in physical health after surgery, but patients in an intermediate range improved based on multiple factors: younger nonsmokers and individuals with a lower body mass index and no lower limb spasticity showed greater improvement in physical health. Conclusions: Four machine learning algorithms were used to determine factors associated with successful outcomes in patients with mild DCM. These models showed good predictive power and provided information about the phenotypes of mild DCM patients most likely to benefit from surgical intervention. Overall, machine learning provides a reliable framework by which clinicians can potentially determine management strategies for patients with mild DCM.

A312: Herniated Discs at the Cervicothoracic Junction

Kyung-Hyun Kim¹, Bang-Sang Hahn¹, and Sung-Uk KUH¹

¹Yonsei University, Gangnam Severance Hospital, Seoul, Republic of Korea

Introduction: Disc herniations at the cervicothoracic junction (C7-T1 level) are unusual, and there have only been a few studies of patients with herniated C7-T1 discs. In addition, previous studies did not focus on the mechanism and causes of solitary cervicothoracic junction disc herniation. The authors investigated the characteristics, symptom duration, clinical course, and biomechanics of cervicothoracic junction disc herniation by comparing patients with C7-T1 disc herniation (C7-T1 group) with control groups. Material and Methods: Thirty-six patients who underwent solitary C7-T1 single-level disc surgery between 2006 and 2015 were included. For radiographic comparison, patients in a herniated C5-C6 disc group and the healthy control group were cohort matched. Results: In the C7-T1 group, the disc herniation mainly occurred in the foraminal space (P < .0001). The C7-T1 group was significantly associated with a history of trauma (P < .0001). In addition, the cervical vertebral body was more readily observed on plain lateral radiographs in the C7-T1 group (7.36 ± 0.068). Patients in the C5-C6 group tended to have the sternal notch more frequently located above the T2-T3 disc space than other groups (P = .014). Conclusion: C7-T1 disc herniation demonstrates unique characteristics. The disc herniates laterally because of the absence of Luschka joints at this level. Hand motor weakness is common and is due to compression of the C8 root, which is mainly composed of a motor component. A history of trauma is closely related with C7-T1 disc herniation. A lower location of the shoulder may increase the mobility of the cervicothoracic junction, increasing the rate of disc herniation at the junction. Understanding the features of disc herniation at the cervicothoracic junction would be helpful for optimal care.

A313: Analysis of the Results of Simultaneous Surgical Interventions in the Treatment of Patients With Tandem Stenosis of the Cervical and Lumbar Spine

Vadim Byvaltsev¹, Andrei Kalinin¹, Marat Aliev¹, Ivan Stepanov¹, Bakhyt Aglakov¹, Bobur Yussupov¹, and Valerii Shepelev¹

¹Irkutsk State Medical University, Irkutsk - Russian Federation

Introduction: The term “tandem stenosis” represents a combined decrease in the bone boundaries of the vertebral canal in the cervical and lumbar spine. Symptomatic stenosis contributes to the emergence of various clinical symptoms, which greatly complicates the diagnosis and determination of treatment tactics for patients, and encourages the application of clinically unjustified approaches to performing surgical interventions. At the moment, there is no single approach to the implementation of surgical interventions in this group of patients, the issues of the primary and the stage of surgical manipulation, as well as the analysis of combined surgical interventions are not consecrated. The purpose of our study was to retrospectively analyze the results of simultaneous surgical interventions in the treatment of patients with tandem-stenosis of the cervical and lumbar spine. Materials and Methods: The study included 47 patients with symptomatic tandem stenosis of the cervical and lumbar parts who performed simultaneous decompressive-stabilizing interventions: laminoplasty or laminectomy with transarticular stabilization was performed on the cervical spine, laminectomy with an open interbody spinal fusion, and transpedicular stabilization or minimally invasive reconstruction of the vertebral channel from paramedian access with interbody spondylodesis and transcutaneous transpedicular stabilization. The intraoperative characteristics of surgical interventions and the features of the postoperative period, the clinical parameters (the level of the pain syndrome according to the visual analogue scale [VAS]), the quality of life by Neck Disability Index (NDI) and Oswestry Disability Index (ODI) the patient’s satisfaction with the operation performed on the MacNab scale, the presence of complications. Results: The ratio of men and women was 2:1. The median (25%, 75%) of age amounted to 46 (42, 58) years. The average duration of the surgery was 122 (98 143) minutes, median of the total blood loss 150 (110 280) mL. All patients were activated the next day after the operation. After simultaneous decompressive-stabilizing intervention, all patients showed a decrease in the degree of pain syndrome by VAS in cervical spine from 77 (63, 92) mm up to 10 (6, 14) mm (P < .001), in lumbar spine from 79 (68, 91) mm up to 11 (6, 13) mm (P < .001), in upper extremities from 89 (73, 98) mm up to 4 (2, 6) mm (P < .001), in lower extremities from 87 (75, 96) mm up to 5 (3, 7) mm (P < .001). An improvement in the functional state by the NDI from 36 (32, 44) points to 4 (2, 6) points (P < .001) and ODI from 66 (54, 72) points to 10 (4, 12) points (P < .001). Most patients noted the effectiveness of the surgery according to the MacNab scale as excellent and good (84%). A complete interbody spondylodesis was diagnosed in 42 patients (89%) in lumbar spine. Complications were observed in 6.4% of cases. Conclusion: A single-center retrospective study of the results of the effectiveness of simultaneous surgical interventions with the use of specialized microsurgical instruments in the treatment of patients with tandem stenosis of the vertebral canal of the cervical and lumbar spine has confirmed their high clinical and radiological effectiveness, with low risks of postoperative complications.

A314: Anterior Versus Posterior Cervical Decompression and Fusion in Patients With Multi-Level Degenerative Cervical Myelopathy

Jetan Badhiwala¹, Yosef Ellenbogen², Jamie Wilson¹, Farshad Nassiri¹, Christopher Witiw¹, Saleh Almenawer², Jefferson Wilson¹, and Michael Fehlings¹

¹University of Toronto, Toronto, Ontario, Canada

²McMaster University, Hamilton, Ontario, Canada

Introduction: The optimal approach for surgical decompression in degenerative cervical myelopathy (DCM) is unclear, and there is significant variation in practice patterns globally. We sought to compare the inpatient complications and costs of anterior (ACDF) versus posterior cervical decompression and fusion (PCDF) using a national administrative healthcare dataset. Materials and Methods: Patients who underwent multilevel (3+ levels) ACDF or PCDF for DCM were identified from the HCUP National Inpatient Sample (NIS) for years 2004 to 2014 using ICD-9-CM codes. Nonelective admissions and cases of redo/revision surgery were excluded. Propensity score matching (1:1 ratio; nearest-neighbor) was performed with age, sex, number, and type of comorbidities (diabetes, chronic pulmonary disease, hypertension), hospital bed size, and use of intraoperative monitoring as covariates. Outcomes, including hospitalization charges and costs, length of stay (LOS), discharge disposition, and inpatient complications (eg, cardiac complications, pneumonia, pulmonary embolism [PE], deep venous thrombosis [DVT], neurological complications, hardware-related complications, postoperative hematoma, wound complications, cerebrospial fluid [CSF] leak, arterial injury, hoarseness, dysphagia) and mortality, were compared between matched ACDF and PCDF groups. Mean differences (MDs) for continuous outcomes and odds ratios (ORs) for categorical outcomes and their corresponding 95% confidence intervals (CIs) were calculated. Results: A total of 17 805 patients were eligible. Propensity score matching resulted in a cohort of 13 884 patients (n = 6942 ACDF; n = 6942 PCDF). Mean age was 63.0 ± 10.5 years. There were no differences in baseline characteristics between ACDF and PCDF groups. PCDF was associated with greater LOS (MD +1.7 days, P < .01) and less frequent routine discharge home (OR 0.26, P < .01). With regard to complications, the PCDF group had a higher rate of myocardial infarction (OR 1.6, P < .01), PE (OR 2.6, P < .01), DVT (OR 3.7, P < .01), neurological complications (OR 1.7, P < .05), hardware-related complications (OR 2.7, P < .01), wound infection/breakdown (OR 6.8, P < .01), and CSF leak (OR 1.7, P < .05). In contrast, the incidence of postoperative hematoma (OR 0.61, P < .01), hoarseness (OR 0.13, P < .01), and dysphagia (OR 0.20, P < .01) were significantly lower following PCDF. Mortality rate was comparable between ACDF and PCDF (0.29% vs 0.26%, respectively; P = .75). Hospital charges (MD +$26 259, P < .01) and costs (MD +$7 728, P < .01) were significantly higher for PCDF than ACDF. Conclusions: At a national level, we found anterior surgery for multilevel DCM to be associated with greater rates of postoperative hematoma, hoarseness, and dysphagia. However, posterior surgery exhibited a higher rate of general medical complications, length of stay, and in-hospital costs. Though our study does not preclude the need for a randomized controlled trial, it offers valuable insight for surgical decision making in patients with DCM.

OP46: Minimally Invasive Spine Surgery 5

A315: Radiographic Outcomes of Custom-Made Asymmetric Titanium Lordotic Cages in Minimally Invasive Anterior Lumbar Interbody Fusion for Adult Degenerative Scoliosis

A Denisov¹, Dmitriy Mikhaylov¹, Dmitriy Ptashnikov¹, Nikita Zaborovsky¹, Oleg Smekalenkov¹, and Sergey Masevnin¹

¹RSRI n.a. Vreden R.R., Saint Petersburg, Russian Federation

Introduction: It is well known that the usage of minimal invasive surgery (MIS) techniques for adult degenerative scoliosis (ADS) require correct placement of interbody cages. For the purpose of improving segmental and total lordosis, interbody cages were invented with certain lordotic angles that contribute not only to appropriate correction degree but also to indirect decompression by increasing of intervertebral heights. Implementation of 3D printing in current practice has allowed more individualized preoperative planning and implant preparation for achieving definite goals of spinal surgery. In case of ADS, asymmetric lordotic titanium cages could help not only in sagittal but also in frontal balance correction that can improve radiographic outcomes. The goal of this study is to evaluate changes of radiographic parameters in patients with ADS on whom anterior correction with asymmetric titanium lordotic cages was performed. Material and Methods: This is a single-center retrospective study of 47 patients (17 men, 30 women) who underwent oblique and anterior lumbar interbody fusion (OLIF and ALIF) using titanium custom asymmetric lordotic cages for ADS correction. Two groups were formed according to SRS-Schwab classification (2012): The first group consisted of 23 patients with moderate PI-LL mismatch and second group had 24 patients with severe PI-LL mismatch. The period of observation was from January 2016 to May 2018. Standing full spine and conventional radiographs at preoperation and 1, 6, and 12-month follow-up were performed. Observed radiographic parameters: sagittal vertical axis (SVA), lumbar lordosis (LL), pelvic tilt (PT) for sagittal balance and central line with Cobb angle for frontal deformation. Intragroup results were compared using Student t test, and intergroup results were evaluated with Mann-Whitney U test. Results: Thirty-two patients underwent 3-level MIS ALIF, OLIF, and 15 underwent 2-level MIS OLIF. The mean age was 61 years, and 64% of the patients were women. In both groups, frontal deformation correction was achieved. Statistical difference between pre- and postoperative radiographic parameters was found in first (P _SVA = .019, P _PT = .0017, P _LL = .023, P < .05) and the second (P _SVA = .042, P _PT = .036, P _LL = .043, P < .05) group. Comparing results entre groups we found that mean postoperative values of SVA, PT, LL were significantly higher in the first group (mean SVA₁ = 25 ± 8 mm, mean SVA₂ = 87 ± 19 mm, P = .018; mean PT₁ = 14° ± 3°, mean PT₂ = 28° ± 4°, P = .032; mean LL₁ = 47° ± 2° mean LL₂ = 31° ± 1°, P = .024, P < .05). In addition, 72% of patients of the second group subsequently required posterior transpedicular fixation and Smith-Peterson osteotomy (SPO) for achieving full sagittal correction. In the first group, only 11% patients required second-stage surgery. Conclusion: Titanium custom asymmetric lordotic cages provided significant increase in sagittal balance correction among patients with moderate deformity (first group), instead of the second group where the majority of patients with severe deformations for reasons of not achieving significant improvement of correction required additional surgery.

A316: Minimally Invasive Antero-Lateral Extracavitary Corpectomy Approach For Thoracolumbar Spinal Diseases in a Colombian Hospital: Cases Series

Alberto Daza¹, David Solano¹, Jorge Torres¹, William Riveros¹, and Marco Garcia¹

¹Hospital Universitario Mayor, Bogota DC, Colombia

Introduction: The treatment of thoracolumbar spine pathologies affecting the vertebral body, including vertebral tumors, spondylodiscitis, and fractures have been traditionally surgically treated by anterior approaches with significant morbidity associated to large thoracotomy wound, entrance into the pleural space with the need for a chest tube, intercostal neuralgia associated with retractor pressure among others. Posterior approaches such as the posterior vertebral body resection require large incisions, are technically demanding, and are very destabilizing due to the posterior element resection required to achieve corpectomy. Minimally invasive thoracoscopic approaches have been introduced but they require a long learning curve and specialized instruments. We present our experience with the minimally invasive anterolateral extracavitary corpectomy (MI-ALEC) approach as described by Uribe et al. This approach minimizes surgical trauma by entering the extracavitary space between the parietal pleura and the internal thoracic fascia avoiding entrance in to the pleural space; it makes possible to communicate the thoracic extracavitary space with the retropleural space without trauma to the diaphragm. Purpose of the Study: The aim of this study is to describe the experience and present a cases series of a neurosurgery department in Colombia with the MI-ALEC approach in patients with thoracolumbar vertebral body disorders. Materials and Methods: Fourteen patients underwent a single-level MI-ALEC and 4 patients underwent a 2-level MI-ALEC and cage reconstruction (T5-L2) for vertebral body tumor (n = 8), unstable vertebral body fracture (n = 7), and instability due to spondylodiscitis (n = 3). The main complaint was axial pain (n = 14); however, 4 patients presented neurological deficit. All procedures were practiced by a senior surgeon TM. Ten procedures were performed for thoracic spine and seven for lumbar spine. The thoracolumbar junction was involved in 1 case. Results: The mean surgical duration was 336 minutes and the average blood loss was 821 mL. All patients had improved or unaltered neurological status. On the visual analogue scale (VAS), the mean local pain was 4 immediately after surgery and 1.8 at 6 months postoperatively. No intercostal neuralgia was encountered. Construct stability was found in all patients at 6 months after surgery. There were 2 pleural tears that were of no consequence and 1 dural tear. One patient died 2 weeks after surgery due to pulmonary embolism. No operative site infections occurred. Conclusion: The MI-ALEC approach is a useful alternative for the treatment of vertebral body pathology requiring corpectomy and/or anterior spinal canal decompression. It can be achieved with low morbidity, low postoperative pain, and favorable aesthetic results. The initial learning curve is a challenge, but it is rapidly overcome especially for surgeons familiar with minimally invasive retractors and techniques.

References

 1. Uribe J, Dakwar E, Le T, Christian G, Serrano S, Smith W. Minimally invasive surgery treatment for thoracic spine tumor removal: a mini-open, lateral approach. Spine (Phila Pa 1976). 2010;35(26 suppl):S347-S354.

 2. Uribe JS, Dakwar E, Cardona RF, Vale FL. Minimally invasive lateral retropleural thoracolumbar approach: cadaveric feasibility study and report of 4 clinical cases. Neurosurgery. 2011;68(1 suppl operative):32-39.

 3. Park MS, Deukmedjian AR, Uribe JS. Minimally invasive anterolateral corpectomy for spinal tumors. Neurosurg Clin N Am. 2014;25:317-325.

 4. Payer M, Sottas C. Mini-open anterior approach for corpectomy in the thoracolumbar spine. Surg Neurol. 2008;69:25-32.

 5. Ballard J, Carlson G, Chen J, White J. Anterior thoracolumbar spine exposure: critical review and analysis. Ann Vasc Surg. 2014;28:465-469.

 6. Bateman D, Millhouse P, Shahi N, et al. Anterior lumbar spine surgery: a systematic review and meta-analysis of associated complications. Spine J. 2015;15:1118-1132.

 7. D’Aliberti G, Talamonti G, Villa F, et al. Anterior approach to thoracic and lumbar spine lesions: results in 145 consecutive cases. J Neurosurg Spine. 2008;9:466-482.

 8. Holland C, Bass D, Gary M, Howard B, Refai D. Thoracic lateral extracavitary corpectomy for anterior column reconstruction with expandable and static titanium cages: clinical outcomes and surgical considerations in a consecutive case series. Clin Neurol Neurosurg. 2015;129:37-43.

 9. Quraishi NA, Konig M, Booker SJ, et al. Access related complications in anterior lumbar surgery performed by spinal surgeons. Eur Spine J. 2013;22(suppl 1):S16-S20.

10. Lubelski D, Abdullah KG, Steinmetz MP, et al. Lateral extracavitary, costotransversectomy, and transthoraric thoracotomy approaches to the thoracic spine review of techniques and complications. J Spinal Disord Tech. 2013;26:222-232.

11. Foreman PM, Naftel RP, Moore TA 2nd, Hadley MN. The lateral extracavitary approach to the thoracolumbar spine: a case series and systematic review. J Neurosurg Spine. 2015;24:570-579.

12. Khoo LT, Smith ZA, Asgarzadie F, et al. Minimally invasive extracavitary approach for thoracic disectomy and interbody fusion: 1-year clinical and radiographic outcomes in 13 patients compared with a cohort of traditional anterior transthoracic approaches. J Neurosurg Spine. 2011;14:250-260.

A317: Extra-Transversalis Approach of Oblique Lumbar Interbody Fusion

Hitoshi Hayase¹, Takahisa Kaneko¹, Tomohiro Murakami¹, Hiroshi Nakagawa², Manabu Sumiyoshi², and Koji Saito²

¹Kojinkai Social Medical Corporation, Hokkaido Ohno Memorial Hospital, Sapporo, Japan

²Kojinkai Social Medical Corporation, Kushiro Kojinkai Memorial Hospital, Kushiro, Japan

Introduction: Lateral lumbar interbody fusion (LLIF) is a procedure that can perform anterior vertebral body fixation in the lumbar spine in a minimally invasive and short time. To avoid ureter or intestinal complications in oblique lumbar interbody fusion (OLIF), it is important to understand anatomic knowledge of retroperitonium space, especially transversalis fascia. Material and Methods: In our facility, we had performed 76 cases of OLIF for 2 years and analyzed intraoperative findings and surgical videos of 10 cases performed extra-transversalis fascia approach of OLIF, including the thoracolumbar lesion. Results: In the OLIF approach, after the skin incision, the external abdominal oblique muscle, the internal abdominal oblique muscle, and the transverse abdominal muscle along the muscle fibers are split. The thin layer, the transversalis fascia over the retroperitoneal fat tissue was observed. The transverse fascia was attached to the tip of 10th rib. It is easy to access the transversalis fascia to split the tip of 10th rib cartilage. The transversalis fascia was detached from the inner layer of the abdominal transverse muscle gently. The transversalis fascia was merged with anterior layer of thoracolumbar (lumbodorsal) fascia, and covered the quadratus lumborum muscle and the iliopsoas muscle surface. It is possible to reach the target of intervertebral space without exposing ureter and intestine, and scattering the pararenal fat. Intraoperative image was presented. Conclusion: At the actual surgery of OLIF, the transversalis fascia under the 3 layers of abdominal muscles was incised, via fat tissue of posterior pararenal space, surgeon’s finger reached the target disc after peeling off the transversalis fascia on the lateral side of psoas muscle. It is important for the safety of OLIF approach, especially multilevel OLIF, to understand the anatomy of the transversalis fascia.

A318: Oblique Lumbar Interbody Fusion for Adjacent Segment Disease

Javier Quillo-Olvera¹, and Jin-Sung Kim²

¹Minimally Invasive Spine Surgery Center, Queretaro, Mexico

²Seoul St Mary’s Hospital, The Catholic University of Korea, Seoul, Republic of Korea

Introduction: Adjacent segment disease (ASD) is a pathology found in patients who have undergone lumbar fusion procedures. Several options have been reported for treating this pathology. This study aims to present the clinical and radiological outcomes of patients treated by oblique lumbar interbody fusion (OLIF) and posterior fixation for ASD. Material and Methods: Medical records from 13 patients underwent OLIF for ASD were reviewed retrospectively in a single institution. Clinical outcomes were assessed with preoperative and the postoperative scores at 3, 6, 9, 12, and 24 months of the visual analog scale (VAS) for back and leg pain. The Oswestry Disability Index (ODI) was evaluated in the preoperative and postoperative. MacNab criteria were asked in the last routine visit at the 12-month. The radiological outcome was assessed with standing radiographs and computed tomography (CT) scan images preoperatively and postoperatively. The parameters evaluated on these imaging studies were the segmental lordosis (SL), lumbar lordosis (LL), sacral slope (SS), pelvic tilt (PT), pelvic incidence (PI), and lastly, PI-LL mismatch, and fusion status. Statistical analysis of changes over time of the clinical evaluations and comparison of the radiological outcomes from preoperative to postoperative also was analyzed. Results: Thirteen patients were included in the study, and all sample completed the follow-up. Seventeen levels were fused by OLIF and fixed using one of the following techniques: open transpedicular fixation, percutaneous transpedicular fixation, and cortical fixation. The mean age was 67.1 years. The mean follow-up was 25.2 months. All clinical parameters improved at the last follow-up compared with preoperative values. In the last follow-up, 12 patients reported excellent outcomes while one patient reported a good outcome defined by the MacNab criteria. Radiological parameters such as intervertebral height, segmental lordosis, lumbar lordosis, and spinopelvic parameters improved after surgery with statistical significance in the comparison. Conclusion: Sufficient indirect decompression, restoration of sagittal alignment and prevention of kyphosis, low risk of lumbar plexus injury, and low cage subsidence were seen in our study. Thus, OLIF in patients with ASD seems to provide better clinical and radiological outcomes at short-term follow-up.

A319: Full-Endoscopic Approach for Removal of the Spontaneous Spinal Epidural Hematoma

Chien-Min Chen¹

¹Changhua Christian Hospital, Changhua, Taiwan, Province of China

Introduction: Spinal epidural hematoma (SEH) is estimated to occur in 1 per 1 000 000 patients per year and can be classified as idiopathic, spontaneous, or secondary. Although the etiology of spontaneous SEH is poorly understood, it has been shown to be associated with minor trauma, chiropractic manipulation, rheumatoid arthritis, and cervical spondylitis. SEHs of the spine can compress surrounding structures and are of clinical importance when they affect the spinal cord or nerve roots. Chronic SEH is very uncommon and the clinical features may be similar to those of a herniated nucleus pulposus or lumbar spinal stenosis. Surgical treatment may be necessary if medical treatment fails. Full endoscopic discectomy performed via a transforaminal and interlaminar approach has been shown to be an effective technique for patients with herniated nucleus pulposus. Material and Methods: Here, we present cases of spontaneous seh in the lumbar spine that was removed using full-endoscopic transforaminal and interlaminar approach under local anesthesia. we evaluated pre- and postoperative visual analogue scale (VAS). Patient satisfaction was also record. Results: The clinical outcomes were good. Conclusion: Percutaneous endoscopic removal of lumbar spinal epidural hematoma maybe an alternative.

A320: Comparison of Clinical Outome and Spinopelvic Parameters Modifications Between Minimally Invasive and Conventional Open Posterior Lumbar Interbody Fusion for Treatment of High-Grade L5-S1 Isthmic Spondylolisthesis

Massimo Miscusi¹, Alessandro Ramieri², Stefano Forcato³, Sokol Trungu³, Alessandro Pesce¹, Pietro Familiari¹, Giuseppe Costanzo², and Antonino Raco¹

¹Sant’Andrea Hospital, Sapienza University of Rome, Rome, Italy

²Sapienza University of Rome, Rome, Italy

³Cardinal Panico Hospital, Tricase, Italy

Introduction: According to Meyerding scale, high-grade isthmic spondilolisthesis are those with more than 50% of slippage. Surgery is indicated when low back pain episodes becomes frequent and refractory to medical and physical management and radicular symptoms due to progressive foramina stenosis. The demonstration of segmental instability at dynamic X-rays is also an indication for surgical treatment. According to literature, in case of high-degree slippage an adequately sized and interbody cage is a key factor to allow satisfactory restoration of segmental lordosis. Our aim was to compare clinical and radiological outcomes of minimally invasive and open surgery for treatment of high-grade isthmic L5 spondylolisthesis. Methods: Two homogeneous groups of 14 patients with painful high-grade isthmic L5 spondylolisthesis (grade III or IV) were respectively operated by posterior minimally invasive interbody fusion (MISS group) and by conventional open interbody fusion (open group). MISS was based on bilateral Wiltse approach, opening of foramina, roots neurolysis, reduction by percutaneous pedicle screwing, including L4, and posterior lumbar interbody fusions (PLIFs). Lumbosacral deformities were classified according to the Spinal Deformity Group. Clinical evaluations were performed applying back and leg VAS, Oswestry Disability Index (ODI), and Short Form–36 (SF-36). Spinopelvic parameters compared between the 2 groups before and after operation were slipping (%), lumbosacral angle (LSA), slip angle, thoracic kyphosis (TK), lumbar lordosis (LL), sagittal vertical axis (SVA), sacral slope (SS), pelvic tilt (PT), and pelvic incidence (PI). Pearson correlation analysis of changes between pre- and postoperation was done (P < .05). Results: Mean age was 24 (range 17-35) years, with female:male ratio 2.5:1. Seven deformities were type 7 and 21 were type 8. Primary healing was obtained in all patients. We had 2 major complications, related to unilateral L5 nerve root palsy, completely recovered at 1 year in open surgery group. Blood loss was greater in the open group (P < .05), as was the postoperative hospital stay (8 ± 2 vs 5 ± 1 days) (P = .02). At follow-up (mean 18 months; range 12-24 months), improvement of slippage was 65% ± 5% in the MISS group, and 62% ± 3% in the open group, without significant differences (P = .07). Back and leg pain VAS scores, ODI, and SF-36 were significantly improved after surgery (P = .037), more significantly in MISS group at the discharge, but without difference between the 2 groups (P > .05) at 3-month follow-up. Modifications of pelvic measurements were recorded in both groups without differences: LSA and slip angle improved in all cases (P < .05), while PT decreased below 25° in 9 (32%) cases (5 open and 4 MISS). Sagittal alignment on SVA was constantly observed. Conclusion: Both surgical procedures seem to obtain good results in terms of late clinical improvement and restoration of spina-pelvic balance, but MISS group showed significant better results in term of blood loss and length of bed rest. MISS should be considered for treatment of high-grade isthmus spondylolisthesis, although attention must be paid to a technical procedure, which requires a careful learning curve.

OP47: Deformity-Thoracolumbar (Adult) 4

A321: The Effect of Surgical Alignment on Dynamic Balance and Neuromuscular Activity in Adult Degenerative Scoliosis Patients Compared With Controls

Ram Haddas¹ and Isador Lieberman¹

¹Texas Back Institute, Plano, TX, USA

Introduction: Patients with adult degenerative scoliosis (ADS) exhibit increased sway within their cone of economy (COE). Increased sway expends more energy while standing, when compared with healthy controls. Spinal realignment surgery has been shown to improve sagittal vertial axis (SVA) and balance. The purpose of this study is to quantify the extent of change in sway associated with maintaining a balanced posture within the COE, in a group of ADS patients’ pre- and postsurgery and compare them with matched nonscoliotic controls. Methods: Thirty-three patients with ADS performed a series of Romberg’s balance tests a week before and 3 months after surgery. Twenty controls performed the same test on one occasion. Sway calculation was based on previously published balance work using a custom software algorithm. Outcome measures were center of mass (COM) and head sway in the sagittal and coronal planes, total sway amount, and spine and lower extremity neuromuscular energy expenditure. Results: When comparing the displacement, ADS patients presented with less COM (Pre: 31.88 cm vs Post: 25.71 cm, P = .042) and head (Pre: 39.85 cm vs Post: 28.11 cm, P = .048) overall displacements in comparison with their presurgery displacement. When comparing the sway, patients with ADS presented less COM (Pre: 1.73 cm vs Post: 1.12 cm, P = .043) and head (Pre: 1.29 cm vs Post: 0.77 cm, P = .050) sagittal sway ROM postsurgical intervention in comparison with their presurgery sway. We found that patients with ADS demonstrated statistically significantly less right multifidus muscle (Pre: 46.26% of sustained maximal voluntary contraction [sMVC] vs Post: 31.15% of sMVC, P = .035) but more rectus femoris muscle (Pre: 39.05% of sMVC vs Post: 47.98% of sMVC, P = .026) activity during 1-minute standing in comparison to their presurgery muscular activity. Although the sway was reduced after surgery, when we compared the displacement of postsurgical ADS patients with controls they demonstrated more COM (Post: 25.71 cm vs control: 19.26 cm, P = .002) and head (Post: 28.11 cm vs control: 19.93 cm, P = .012) overall displacements. Even when muscular activity was reduced after surgery, we found that these scoliosis patients still expended statistically significantly more muscle activity to maintain static standing, as manifest by increased muscle activity in their right erctor spinea (Post: 38.59% of sMVC vs control: 29.40% of sMVC, P = .046), left glutes maximus (Post: 35.48% of sMVC vs control: 28.37% of sMVC, P = .044), and bilateral tibilais anterior (Right: Post: 43.18% of sMVC vs control: 32.49% of sMVC, P = .037; Left: Post: 40.57% of sMVC vs control: 31.97% of sMVC, P = .032) muscles during 1-minute standing in comparison to nonscoliotic controls. Discussion and Conclusion: Surgical realignment reduced the amount of sway, reduced the center of mass displacement, and reduced spine and lower extremity energy expenditure in ADS’ patients. In symptomatic preoperative ADS patients, sagittal sway, lumbar spine, and lower extremity neuromuscular activity, were all increased in comparison with a nonscoliotic control. Although surgical realignment improved ADS functional parameters during a dynamic balance test, these parameters approached but did not fully achieve the non-scoliotic control parameters when measured 3 months after surgery.

A322: Impact of Lumbar Lordosis Correction on Cervical Sagittal Alignment in Adult Spinal Deformity: Is Postoperative Cervical Kyphosis Permanent?

Jung-Hee Lee¹, Key-Young Lee¹, Won-Ju Shin¹, Sang-Kyu Im¹, and Dong-Gune Chang²

¹Kyung Hee University Medical Center, Seoul, Republic of Korea

²Inje University, Sanggye Paik Hospital, Seoul, Republic of Korea

Introduction: Several studies have been reported with regard to cervical kyphosis after correction of adult spinal deformity, resulting from an effort to maintain upright alignment and horizontal gaze. However, there was a lack of evidence for postoperative cervical kyphosis after deformity correction. The aim of this study is to evaluate changes of postoperative cervical kyphosis during follow-up periods and to verify differences of cervical kyphosis according to the amount of correction in lumbar lordosis or sagittal balance. Material and Methods: A retrospective chart review was performed in our institution. Total 122 patients (female = 102, mean age = 69.2 years) with degenerative lumbar kyphosis who underwent thoracolumbar deformity correction with a minimum of 2-year follow-up were analyzed. Cervical, thoracic and lumbar lordosis or kyphosis were measured pre- and postoperatively. To evaluate influencing factors for change of cervical alignment, the patients were divided according to postoperative lumbar lordosis and sagittal alignment. The patients were categorized into 3 groups by postoperative pelvic incidence–lumbar lordosis (PI-LL < −10°: n = 50, −10° < PI-LL < 10°: n = 43, PI-LL > 10°: n = 29) and C7 plumb line (C7PL <0 mm: n = 29, 0 < C7PL < 50 mm: n = 60, C7PL > 50 mm: n = 33). Cervical alignment was compared among each group. All parameters were measured at preoperative, immediately postoperative, and last follow-up periods. Correlation analyses among all parameters were also performed. Results: Mean preoperative LL (+1°) significantly improved immediately after surgery (–59°) and was maintained at the last follow-up (–54°). Mean preoperative C7PL (+167 mm) significantly improved immediately after surgery (+2 mm) and was maintained until the last follow-up (+36 mm). Preoperative cervical lordosis and thoracic kyphosis changed, respectively, from –21° and 3° to –10° and 21° immediately after surgery and slightly changed at the last follow-up (–16° and 25°). At the last follow-up, mean cervical lordosis showed no significant differences according to postoperative PI-LL (<–10°: –15.6° ± 13.5°, –10° to 10°: –14.6° ± 12.1°, >10°: –18.7° ± 11.8°) (P = .375), but there were significant differences for cervical lordosis according to postoperative C7PL (<0 mm: –15.0° ± 12.8°, 0-50 mm: –15.0° ± 11.1°, >50 mm: –24.8° ± 13.6°) (P = .035). The patients with sagittal imbalance (C7PL > 50 mm) showed significantly higher cervical lordosis than other patients. In correlation analysis, cervical lordosis was not correlated with thoracic kyphosis immediately after the surgeries, but with thoracic kyphosis at the last follow-up. Conclusion: Cervical kyphosis seems to deteriorate temporarily immediately after correction of adult thoracolumbar spinal deformity but improves during the follow-up period. Particularly, postoperative cervical kyphosis was influenced by postoperative sagittal vertical axis, but not by amount of lumbar lordosis correction.

A323: The L3l5 Segment in the Coronal (Anteroposterior) Plane Gives Information on Sagittal Balance: New Parameters, Based on Pathogenesis, of Adult Scoliosis Patients

Maximo-Alberto Diez-Ulloa¹, Lucia Domingo-Rodriguez², Alba Pereiros-Parada², Maria 0tero-Fernandez¹, Carlos Pernal-Duran¹, Eva Diez-Sanchidrian², and Blanca Otero-Torron²

¹Hospitalary Universitary Complex of Santiago de Compostela, Santiago de Compostela, Spain

²University of Santiago de Compostela, Santiago de Compostela, Spain

Introduction: Adult scoliosis has been classified into de novo degenerative, progressive from idiopathic adolescent and secondary. Differentiation between de novo degenerative and idiopathic adolescent, based only on radiological parameters, has yielded substantial intraobserver agreement, but just fair interobserver agreement; although when such agreement was reasonable, it seemed that de novo degenerative patients presented sagittal imbalance. Considering the pathogenesis, perhaps some new parameter could be defined to help improve this interobserver agreement. Degenerative changes leading to de novo degenerative scoliosis cause deviation from midline in the lower lumbar spine, not necessarily accompanied by vertebral tilting (apart from rotatory subluxations lateral olisthesis) specially at either or both L3L4 and L4L5 discs; whereas in idiopathic lumbar/thoracolumbar curves lower lumbar vertebra (usually L2 to L4) usually is centered above sacrum, but with tilted plates. Hypothesis: Because of pathogenesis, distance from the central sacral line (CSL) to the center of the disc and vertebral endplate tilting in the coronal plane (anteroposterior [AP] view) differ between de novo and progressive idiopathic scoliotic patients.Secondary hypothesis: Probably, pathogenesis will cause differences in compensatory adaptative changes, so different sagittal parameters should be expected between both groups. Material and Methods: A series of 64 patients operated for adult scoliosis by the same surgeon was reviewed. Variables: age (at surgery), gender, height, weight, parameters of thoracic, lumbar and lumbosacral curves (limits, Cobb angle, apex), T1 coronal and sagittal balance, L1 coronal and sagittal balance, spinopelvic parameters (pelvic incidence [PI], sacral slope [SS], and pelvic tilt [PT]), and then: (a) distance from CSL to center of L3L4 and L4L5 discs and (b) coronal tilt of inferior L3 endplate, superior and inferior L4 endplates, and L5 superior endplate.Three independent observers gathered the data, which were analyzed using SPSS 22.0. Results: Distance L3L4 (DL3L4) to CSL showed a bimodal (valley 15 mm): mean 17.86 mm, median 14.15 mm.There was a correlation DL3L4 with: PT P = .009; PI P = .025; lumbar Cobb P < .001.When stratified according to the bimodal distribution (DL3L4 <15 vs ≤15 mm): (a) PT 22.35° ± 11.74° (mean ± SD) vs 30.05° ± 14.61° (P = .024); (b) PI 55.40° ± 13.2° vs 62.07° ± 18.3° (nonsignificant differences), and (c) lumbar Cobb angle 22.59° ± 12.65° vs 38.81° ± 17.55° (P < .001), no correlation either with apex or lower limit vertebra. Linear regression PT-DL3L4: P = .001; R2 0.134 Conclusion: Distance of L3L4 disc to CSL seems to provide a means to differentiate 2 groups in adult scoliosis. Cut point: 15 mm. But linear regression explains only 13%.

A324: Functional Outcome 5 Years After Circumferential Fusion for Low-Grade Isthmic Spondylolisthesis (LGIS)

Philip Horsting¹, Malou te Molder¹, Marinus de Kleuver², and Miranda van Hooff¹

¹Sint Maartenskliniek, Nijmegen, Netherlands

³Radboud University Medical Center, Nijmegen, Netherlands

Introduction: Functional outcomes after surgery for low-grade isthmic spondylolisthesis (LGIS) is reported limitedly for circumferential approach procedures. Aim of this study were to evaluate the functional outcome over time and to describe adverse events that occurred. Surgical treatment goals are restoration of stability, reduction of slippage, restoration of disc height and angulation, correction of spinopelvic parameters toward neutral values, decompression of nerve roots, and improve functional status. We hypothesized (1) a significant improvement on clinical outcome at 2-year follow-up and maintenance of improved results at 5-year follow-up. Materials and Methods: A total of 43 patients enrolled with LGIS at L5-S1. Each patient received a posterior decompression and pedicle screw instrumentation and an instrumented anterior lumbar interbody fusion in 2 separate procedures. Outcome measures were registered preoperative and at 1-, 2-, and 5-year follow-up. Primary outcome was functional status (Oswestry Disability Index [ODI] [0-100]). Secondary outcomes: Quality of Life (Short Form–36 [SF-36] subdomains and scales [0-100]; EQ5D-3L utility [−0.225 to 1.000; EuroQual visual analogue scale [EQ-VAS] [0-100]), pain intensity (VAS back and leg pain), and number of adverse events that occurred. The patient characteristics, functional outcomes were described (means [SD]) and the adverse events in numbers. To evaluate the study aim a repeated-measures analysis of variance was used to indicate differences over time. Results: Of the 43 patients 39 currently completed 5-year follow-up assessment (20 female), mean age 39.1 (7.8) years with a mean body mass index (BMI) 25.7 (2.3) kg/m²; n = 37; the remaining 4 patients did not significantly differ on baseline (age, BMI, ODI). Functional results at preoperative, at 12 months and 24 months postoperative and at 5 years postoperative assessment were, respectively, for ODI: 35.0 (14.4); 10.0 (15.0); 9.5 (15.3); 21.5 (16.6); for SF-36 physical component summary (PCS): 42.0 (12.7); 70.5 (21.0); 71.5 (21.5); 59.0 (22.3); for SF-36 mental component summary (MCS): 64.0 (16.6); 81 (19.4); 81.5 (19.5); 66.0 (20.2); for EQ-5D: 0.56 (0.29); 0.84 (0.28); 0.83 (0.240); 0.80 (0.13); for VAS (back pain): 61.3 (25.9); 25.9 (26.1); 27.9 (24.1); 37.7 (29.5). The surgery for patients with LGIS elicited statistically significant differences in mean ODI over time, F _{(3, 111)} = 23.57, P < .001. Adverse events: at 2-year follow-up in 14 of 43 events occurred. In 5 patients the events were device-related, 2 of them were reoperated upon, and 1 of these 2 patients received spinal cord stimulation between 2- and 5-year follow-up assessment. At 5-year follow-up, 14 of 39 patients reported to have persistent complaints, 6 of 39 visited a medical specialist during the previous year and: 5 of 14 were ongoing events from 2-year follow-up and 2 of 14 participated in a 2-week intensive combined physical and psychological pain management program. Conclusion: Circumferential-instrumented fusion for LGIS via combined approach improves functional outcome at short- and midterm follow-up. After careful patient selection and unsuccessful conservative management surgical treatment is a suitable alternative with favorable outcome. Complications remain an area of concern.

A325: Variation of MCID by Age, Gender, Baseline Scores, and Change of Direction

Sinan Bahadir¹, Selcen Yüksel², Selim Ayhan³, Alba Vila-Casademunt⁴, Ibrahim Obeid⁵, Francesco Sanchez Perez-Grueso⁶, Emre Acaroglu⁷; European Spine Study Group⁴

¹Amasya University, Amasya, Turkey

²Yildirim Beyazit University, Ankara, Turkey

³ARTES Spine Center, Ankara, Turkey

⁴Hospital Universitari Vall d’Hebron, Barcelona, Spain

⁵Bordeaux University Hospital, Bordeaux, France

⁶Hospital Universitari La Paz, Madrid, Spain

⁷Ankara Spine Center, Ankara, Turkey

Introduction: Minimal clinically important difference (MCID), an important concept to evaluate effectiveness of treatments, does not have a definitive calculation method as of now. Furthermore, MCID value for a specific health-related quality of life (HRQoL) measure shows variation based on the calculation method as well as pathology, baseline scores, comorbidities, and treatment modalities. We aimed to analyze effect of age, gender, and baseline scores on MCID values as well as MCID values for impairment for several HRQoL measures in an adult spinal deformity (ASD) population. Material and Methods: This is a prospective cohort study. Patient population consisted of surgical and nonsurgical patients from a multicenter ASD database who completed pretreatment and 1-year follow-up outcome measures. Core Outcome Measures Index (COMI), Oswestry Disability Index (ODI), Short Form–36 Physical Component Summary (SF-36 PCS), SF-36 Mental Component Summary (MCS), Scoliosis Research Society (SRS)-22R, and an anchor question were used as outcome measures in the study. MCID values for each HRQoL measure were calculated by an anchor-based method by using latent class analysis (LCA) and minimum detectable changes (MDCs) by a distribution-based method. Results: Study showed that baseline scores effect MCID value. Patients with baseline ODI score <20, 20-40, and >40 had MCID value of 2.24, 11.35, and 26.57, respectively. Similarly, patients with baseline COMI score <2.75, 2.8-5.4 and >5.4 had MCID threshold of 0.59, 1.38, and 3.67, respectively. Also, we found that MCIDs for impairment were lower than those for improvement in all HRQoL measures used in the study. Overall MCID thresholds for impairment and improvement were 0.27 and 2.62 for COMI, 2.23 and 14.31 for ODI, and 0.01 and 0.71 for SRS-22. Apart from SF-36 MCS where younger patients had lower MCID threshold (3.50 vs 7.99), MCID values were not affected by age. For those younger and older than 36 years, MCID thresholds were 2.64 and 2.68 for COMI, 15.29 and 17 for ODI, and 0.74 and 0.75 for SRS-22. Finally, gender difference did not have effect in MCID value for any measure. Conclusion: The findings of this study demonstrates that MCID values change by baseline scores, direction of change (improvement/impairment) and to some extent by age, but not by gender. The available literature also supports our findings regarding variability of MCID by several factors. We believe that there is no generic universal MCID value for a given HRQoL that encompasses whole population of a given pathology. MCID, at its current state, should be considered as a concept rather than fixed values.

A326: Effect of Tobacco Smoking on Implant Failure Rate and Risk of Intraoperative Bleeding: Analysis Of 270 Patients From the Prospective, Multicenter Scoli-Risk-1 Study of Complex Adult Spinal Deformity Surgery

Jamie R. F. Wilson¹, Fan Jiang¹, Benny T. Dahl², Justin Smith³, Kenneth M. C. Cheung⁴, Leah Y. Careeon⁵, Christopher I. Shaffrey⁶, Stephen Lewis¹, Lawrence G. Lenke⁷, and Michael Fehlings¹

¹Toronto Western Hospital, Toronto, Ontario, Canada

²Texas Children’s Hospital, Baylor College of Medicine, Houston, TX, USA

³University of Virginia, Charlottesville, VA, USA

⁴University of Hong Kong, Hong Kong SAR, China

⁵Norton Leatherman Spine Centre, Louisville, KY, USA

⁶Duke University, Durham, NC, USA

⁷The Spine Hospital, Columbia University Medical Center, New York, NY, USA

Introduction: Tobacco smoking has been identified in retrospective studies as a predictor for medical complications after adult deformity surgery. However, the effect of smoking on intraoperative complications, implant failure, and other adverse events (AEs) has not been adequately described in prospective studies of adult complex deformity surgery. The goal of this study is to analyze the impact of smoking on rates of postoperative AEs in patients undergoing high-risk adult spine deformity surgery. Material and Methods: The Scoli-RISK-1 study enrolled 272 patients who had undergone complex adult spinal deformity surgery from 15 centers, with a minimum 2-year follow-up. The outcomes and incidence of AEs in patients with a history of tobacco smoking (n = 26) were compared with the nonsmoking patients (n = 244) using univariable analysis. Multivariable regression analysis was used to adjust for the effect of patient demographics, complexity of surgery and other confounders. Results: No difference was observed in the number of levels or complexity of surgery in both cohorts. In the univariable analysis, the rates of implant failure were almost double (n = 7; 26.9%; odds ratio (OR) 2.28; 95% confidence interval (CI) [0.75-6.18]) of that observed in the nonsmoking group (n = 34; 13.9%; P = .088), but this was not statistically significant. Surgery-related excessive bleeding was however significantly higher in the smoking group (n = 5 vs n = 9; 19.2% vs 3.7%; OR 6.22; 95% CI [1.48-22.75]; P = .006). Wound infection rates were similar in the smoking group (n = 3 vs n = 17; 11.5% vs 7.0%; OR 1.74; 95% CI [0.30-6.70], P = .422), as were rates of respiratory complications (n = 1 vs n = 13; 3.8% vs 5.3%; OR 0.71; 95% CI [0.02-5.13], P = 1.000). In the multivariable analysis, the smoking group demonstrated a higher incidence of surgery related adverse events over the 2-year follow-up period (n = 13 vs n = 95; 50.0% vs 38.9%; OR 2.12; 95% CI [0.88-5.09]) (P = .094), but this was not statistically significant. Conclusion: In this large, multicenter prospectively collected study of adult spinal deformity, a history of tobacco smoking significantly increased the risk of excessive intraoperative bleeding compared with the nonsmoking group. Moreover, the rate of implant failure was higher in the smoking group, as was the rate of all surgery-related adverse events over 2 years postsurgery, but these differences were just shy of achieving statistical significance. The authors accept that this subanalysis was likely underpowered to measure the true effects of smoking, but nonetheless, we advise that patients who do smoke undertake an active smoking cessation program prior to undergoing complex adult spinal deformity surgery.

OP48: Novel Technologies 4

A327: Protective Effect of Stimulating EMG on Sacrum 2 Nerve Root in Selective Posterior Rhizotomy for Spastic Cerebral Palsy

Zhiming Cui¹ and Guofeng Bao¹

¹The Second Affiliated Hospital of Nantong University, Nantong, China

Introduction: To evaluate the clinical efficacy of intraoperative stimulating electromyography (SEMG) monitoring in the treatment of spastic cerebral palsy in children with selective posterior rhizotomy (SPR) and its protective effect on S2 nerve roots. Material and Methods: A prospective study of 32 children with spastic cerebral palsy who underwent SPR (L5, S1, and S2 nerve roots) was performed according to the randomized principle: 11 patients in the control group, in this group intraoperative electromyography of the lower extremities did not receive anal sphincter test; 21 patients in the study group underwent intraoperative electromyography and anal sphincter examination. After 1 month of treatment, the gait improvement, the active and passive activities of the knee and ankle joint, the improvement of the muscle tension of the gastrocnemius and the biceps muscle, and the classification of the gross motor function (GMFCS classification) were compared between the 2 groups. The diagnostic value of SEMG monitoring in the assessment of neurological dysfunction in children with SPR for spastic cerebral palsy was analyzed. The difference in the incidence of adverse reactions was compared between the 2 groups. Results: The gait function improvement rate, knee flexion reduction rate, and sputum correction rate were significantly higher in the study group than in the control group (P < .05). The muscle tension and GMFCS scores of the gastrocnemius and biceps muscles in the study group were significantly lower than those in the control group (P < .05). According to the intraoperative SEMG warning criteria, the sensitivity of the SEMG synchronized anal sphincter to monitor neurological dysfunction in children with spastic cerebral palsy was 62.4%, specificity was 42.1%, and accuracy was 85.5%. Recurrence of knee flexion hernia, transient limb dysfunction, laborious numbness in the saddle area, and urinary retention were significantly lower in the study group than in the control group (P < .05). Conclusion: The use of SEMG synchronous anal sphincter monitoring in children with SPR for spastic cerebral palsy can effectively improve the symptoms of patients caused by S2 nerve root spasm, avoid the secondary injury of S2 nerve root excessive amputation, and improve the safety of surgery.

A328: Validity and Reliability Assessment of a Novel Laser-Guided Digital Goniometer

Saimurooban Muralidaran¹, Peter Beshara², Louise Nettleton², Ashley Wilson-Smith¹, Matthew Pelletier², Tian Wang², and William Robert Walsh²

¹The University of New South Wales, Sydney, New South Wales, Australia

²Prince of Wales Hospital, Randwick, Sydney, New South Wales, Australia

Introduction: Adequate cervical and lumbar spine range of motion (ROM) is essential for several activities of daily living. These activities are reported to require 20% to 40% of maximum available cervical ROM and 40%-60% of maximum available lumbar ROM.¹ As such, accurate assessment of cervical and lumbar spine ROM is important to determine an individual’s functional status as well as to evaluate their progress during rehabilitation. While the universal goniometer is still considered the gold standard, studies have shown poor inter-rater reliabilities for cervical and lumbar ROM.^2,3 Hence, the purpose of this study was to determine the concurrent validity and interrater reliability of the HALO, compared to the universal goniometer. Material and Methods: The HALO (model HG1, HALO Medical Devices, Australia) is a laser-guided, 3-axis goniometer and was used for all range of motion measurements in the present study. The device was zeroed on the starting position for each movement, allowing the range of motion to be determined as the difference between the initial and final positions. A standard 12-inch, double arm plastic goniometer (BASELINE, USA) was used for all goniometric measurements. Fifty asymptomatic participants were recruited and randomized into four groups to complete assessments with the universal goniometer and the HALO. Each assessment consisted of either rater A or B asking the patient to complete their full active ROM for the following movements: cervical flexion, cervical extension, cervical lateral flexion (right), cervical rotation (right), lumbar flexion and lumbar lateral flexion (right). All measurements were taken in accordance with the guidelines set out by the American Academy of Orthopaedic Surgeons. Raters A and B were both fully qualified physiotherapists with 10 and 3 years of clinical experience, respectively. Rater A has had 15 hours of experience using the HALO, while rater B had a single training session with the device. SPSS was used for all statistical analysis. Results: For cervical ROM, the universal goniometer had interrater reliabilities, assessed by intraclass correlation coefficients model (2,1), of 0.380, 0.819, 0.510, and 0.225 for flexion, extension, lateral flexion, and rotation, respectively. Meanwhile, the HALO had inter-rater reliabilities of 0.477, 0.780, 0.718, and 0.551 for the same 4 motions. For lumbar ROM, the universal goniometer had inter-rater reliabilities of 0.654 and 0.573 for flexion and lateral flexion, respectively, while the HALO had interrater reliabilities of 0.670 and 0.440 for the same movements. The mean differences between the universal goniometer and HALO ranged from 12° to 15° for cervical flexion, 5° to 7° for cervical extension, 1° to 12° for cervical rotation and a maximum of 2° for cervical lateral flexion, lumbar flexion and lateral flexion. These differences were statistically significant (P < .05) for assessments of cervical flexion, extension and rotation. Conclusion: The HALO can be used interchangeably with the universal goniometer for most ROM measurements of the cervical and lumbar spine and has been shown to have comparable interrater reliabilities. Further research, including replicating studies, is required to confirm the applicability of the HALO to measure cervical and lumbar ROM.

References

1. Cobian DG, Daehn NS, Anderson PA, Heiderscheit BC. Active cervical and lumbar range of motion during performance of activities of daily living in healthy young adults. Spine (Phila Pa 1976). 2013;38:1754-1763.

2. Pile KD, Laurent MR, Salmond CE, Best MJ, Pyle EA, Moloney RO. Clinical assessment of ankylosing spondylitis: a study of observer variation in spinal measurements. Br J Rheumatol. 1991;30:29-34.

3. Nitschke JE, Nattrass CL, Disler PB, Chou MJ, Ooi KT. Reliability of the American Medical Association guides’ model for measuring spinal range of motion. Its implication for whole-person impairment rating. Spine (Phila Pa 1976). 1999;24:262-268.

A329: A Hybrid Technique for Severe Rigid Congenital Scoliosis: Posterior Osteotomy With Short Segmental Fusion and Dual Growing Rod Technique

Jianguo Zhang¹ and Shengru Wang¹

¹Peking Union Medical College Hospital, Beijing, China

Introduction: As some types of congenital scoliosis consist of long curve including sharp deformities producing enormous asymmetric growth at apex, they cannot be well corrected with osteotomy and short segmental fusion. Growing rod technique may be the best option. However, enormous asymmetric growth potential due to the sharp deformities at apex will increase risk of complications, especially implants failures. This study was conducted to evaluate the clinical outcomes of the hybrid technique of one stage posterior osteotomy with short segmental fusion and dual growing rod technique for severe rigid congenital scoliosis. Materials and Methods: Eleven patients (2 males, 9 females) undergoing this hybrid technique for severe rigid congenital scoliosis were retrospectively reviewed. The patients’ charts were reviewed. The analysis included age at initial surgery and the latest follow-up, number and frequency of lengthening, and complications. Radiographic evaluation included measured changes in scoliosis Cobb angle, thoracic kyphosis, lumber lordosis, trunk shift, length of T1 CS1 and instrumentation. Results: The mean follow-up was 83.3 (49-142) months. The mean age at the initial is 6.1 (4-10) years. The averaged fused segments was 2.6 per patient. And the averaged lengthenings were 6.3 per patient. The mean scoliosis improved from 80.3° to 31.2° after initial surgery and was 32.8° at the latest follow-up. The average T1-S1 length was of 1.22 cm per year. The space availability for lung (SAL) increased from 0.86 to 0.96. Conclusions: Osteotomy with short fusion could help improve the correction of the growing rod and eliminate the large asymmetric growth potential around the apex, with little influence to the length of the spine. Dual growing rod technique could maintain correction achieved at initial surgery while allowing spinal growth to continue. This hybrid technique may be an option for young patients who present sharp deformities with large asymmetric growth potential in a long congenital spinal deformity.

A330: Therapeutic Feasibility of Full Endoscopic Deocmpression in One- to Three-Level Lumbar Canal Stenosis via a Single Skin Port Using a New Spinal Endoscopic System

KangTaek Lim¹, Soo Beom Kim¹, Hyung Suk Kim¹, Jin Soo Park¹, and Chun-Kun Park¹

¹ Good Doctor Teun Teun Hospital, Anyang, Korea

Background: Lately, endoscopic spinal procedure, which had been used mainly discectomy via intervertebral foramen, has expanded its realm into surgical treatment of spinal stenosis via trans- and/or interlaminar approach. Now, a new spinal endoscopic system, percutaneous stenoscopic lumbar decompression (PSLD) has been developed, which was designed and optimized for a surgeon to apply to decompression surgery of the spinal stenosis. The system is composed of 8.4 mm of the working sleeve and 5.7 mm of the working channel in diameter, and 12° of the field of view. This system needs objective assessment for its feasibility, technical applicability, and usefulness. Methods: The study was designed to elucidate the clinical feasibility of PSLD. The authors retrospectively conducted the clinical study in 450 consecutive cases of degenerative lumbar stenosis treated with PSLD between April 1, 2016 and Decmber 31, 2016. The surgical level was 483 in number and included 408 cases of 1 level operation, 33 of 2 levels, and 9 of 3 levels through a single skin entry (incision). The patients composed of 254 males and 196 females. All patients underwent pre- and postoperative magnetic resonance imaging (MRI) to measure the degree of widened vertebral canal and were followed up by 12 months to observe changes of visual analogue scale (VAS) score and disability with Oswestry Disability Index (ODI). Results: Postoperative MRI revealed that PSLD enlarged stenotic vertebral by mean 52.0% of preoperative one at the index segment (P < .001), and demonstrated that damage in the normal soft tissues, including muscles and the extent of removed normal bony structures appeared minimal. The mean improvements of VAS score and ODI were 4.0 (P < .001) and 40% (P < .001), respectively. Mean duration of operatng times was 32 minutes for bilateral decompression of 1 level, 65 minutes of 2 levels and 115minutes of 3 levels, mean hospital stay was 1.2 days. Conclusion: The study demonstrates that full endoscopic decompressing procedures using PSLD have therapeutic feasibility in the surgical treatment of lumbar stenosis. PSLD achieved both postoperative improvement and minimal damage in paraspinal soft tissue including muscles and bony tissues, which were confirmed by MRI and measurement of postoperative enlarged area. It has been proved that mono-portal PSLD has a potential of managing both mono- and multilevel spinal stenosis. The PSLD results appeared compatible with that of surgical treatments, and could be a good alternative to them in surgical treatment of lumbar stenosis.

A331: Tele-Rehabilitation Program After Surgery for Uncomplicated Lumbar Discectomy: Randomized Control Trial Methodology

Uladzislau Ulasavets¹, Ewelina Grzywna¹, Jaroslaw Polak¹, Pawel Adamkiewicz², Katarzyna Jasinska¹, Irena Roterman-Konieczna¹, and Marek Moskala¹

¹Jagiellonian University Medical College, Kraków, Poland

²Polskie Centrum Rehabilitacji Funkcjonalnej Votum, Kraków, Poland

Introduction: Tele-rehabilitation (TR) trials for patients after lumbar disc herniation surgery has been poorly studied. We aim to create the theoretical assumptions of a TR program and design a randomized controlled trial (RCT) evaluating the influence of TR on the results of treatment in patients after lumbar discectomy. Material and Methods: The data of adult patients who underwent standard rehabilitation after lumbar disc herniation surgery was analyzed. We included existing questionnaires and scale to measure the outcome. The feasibility of project has been discussed with potential partners from Neurosurgery, Physiotherapy and Bioinformatics and Telemedicine Departments. Results: The model of 2 arm RCT was created. Patients aged 18 to 60 years after lumbar discectomy at single level are proposed to be included into the study. Exclusion criteria includes among others: chronic conditions affecting spine, trauma, spine instability, surgery complications. Estimated number of patients required for trial and evaluation methods will be granted for creating a potential publicly available platform for TR. Baseline measurements include demographic data, neurological symptoms, preoperative treatment, physical activity and surgery characteristic. Randomization was performed in 2 groups: TR group and standard care group. TR group will include face-to-face physiotherapy and TR program available via the Internet. Potential postintervention measurements cover Oswestry Disability Index, visual analogue scale, Short Form–36, and Beck Depression Inventory. Conclusion: We believe that application of early TR to postoperative management of patients after lumbar discectomy may improve the overall outcome not only in developed countries but also in regions around the world with a paucity of socioeconomic resources where professional rehabilitative care does not exist.

A332: A Novel Virtual Reality Simulator for Spinal Surgery Assessment

Nicole Ledwos¹, Nykan Mirchi¹, Vincent Bissonnette¹, Anika Sawni¹, Alexander Winkler-Schwartz¹, M. Recai Yilmaz¹, and Rolando Del Maestro¹

¹Montreal Neurological Institute and Hospital, McGill University, Montreal, Quebce, Canada

Introduction: Cervical spine disorders encompass one of the most common medical problems worldwide. In the United States, the incidence for cervical spine degeneration, across all age groups, has shown an increasing trend over the past 10 years. As a result, operations such as the anterior cervical discectomy and fusion (ACDF) have become frequently performed; making it an important procedure in resident education. Current residency programs in spine surgery have increased emphasis on patient safety and started shifting toward competency-based curriculums. Residents often lack exposure in spine surgical procedures since it is a shared field between neurosurgery and orthopedics. Virtual reality simulators can supplement these training programs by providing opportunities for residents to practice technical skills in risk-free environments. Virtual reality platforms in spine surgery have been slow to develop with current models simulating relatively simple procedures. In this communication, we present a virtual reality simulated C4-C5 anterior cervical discectomy and fusion carried out on the Sim-Ortho platform. We outline the Sim-Ortho platform and demonstrate how the data can be used to investigate differences in surgical skill. Material and Methods: The Sim-Ortho platform is a voxel-based virtual reality simulator that uses 3D technology for a multisensory experience. The haptic touch feedback simulates real life by giving users varying force and resistance depending on the tool being used and the anatomical tissue being interacted with. The simulated ACDF is segmented into 4 consecutive steps: (a) cutting the disc, (b) removal of the disc, (c) osteophyte removal, and (d) posterior longitudinal ligament removal. Users can select from a scalpel, bone curette, pituitary rongeur, nerve hook or burr for each step based on personal preference. Data is recorded by the platform in real-time during the duration of the simulated procedure. The voxel-based platform allows for new methods of data analysis and different visualizations of the data. Results: A numeric file and voxel file containing data on the simulated anatomical structures and tools, are exported for each user and analyzed in MATLAB. Analysis involved comparisons between varying levels of expertise for a multitude of metrics. For example, in a preliminary investigation, we converted voxel files to a 3D matrix of density values to visualize removal of the cervical disc. In addition, we calculated the amount of disc voxels removed by the bone curette and pituitary rongeur to illustrate potential differences in tool preference and pace of disc removal. Conclusion: A simulated C4-C5 ACDF utilizing the Sim-Ortho simulator provides a platform to assess the understanding, assessment, and training of this procedure. This study represents the first-step in analyzing surgical techniques associated with virtual reality ACDF. The importance of this study lies in the potential validation and future introduction of virtual reality spine simulators in neurosurgery and orthopedic curriculums to compliment the current educational scheme.

OP49: Best Paper Session

A333: The Safety and Efficacy of Riluzole as a Neuroprotective Therapy in Patients Undergoing Surgery for Cervical Spondylotic Myelopathy: Results of the CSM-Protect Trial

Michael Fehlings¹, Jetan Badhiwala¹, Henry Ahn², Francis Farhadi³, Christopher Shaffrey⁴, Ahmad Nassr⁵, Praveen Mummaneni⁶, Paul Arnold⁷, W. Bradley Jacobs⁸, K. Daniel Riew⁹, Michael Kelly¹⁰, Darrel Brodke¹¹, Alexander Vaccaro¹², Alan Hilibrand¹², James Wilson¹³, James Harrop¹², S. Tim Yoon¹⁴, Kee Kim¹⁵, Daryl Fourney¹⁶, Carlo Santaguida¹⁷, and Branko Kopjar¹⁸

¹University of Toronto, Toronto, Ontario, Canada

²St Michael’s Hospital, Toronto, Ontario, Canada

³Ohio State University, Columbus, OH, USA

⁴University of Virginia, Charlottesville, VA, USA

⁵Mayo Clinic, Rochester, MN, USA

⁶University of California, San Francisco, San Francisco, CA, USA

⁷Kansas University Medical Center, Kansas City, KS, USA

⁸University of Calgary Spine Program, Calgary, Alberta, Canada

⁹Columbia University, New York, NY, USA

¹⁰Washington University in St Louis, St Louis, MO, USA

¹¹University of Utah, Salt Lake City, UT, USA

¹²Thomas Jefferson University/Rothman Institute, Philadelphia, PA, USA

¹³Louisiana State University, New Orleans, LA, USA

¹⁴Emory University, Atlanta, GA, USA

¹⁵UC Davis Spine Center, Sacramento, CA, USA

¹⁶University of Saskatchewan Royal University Hospital, Saskatoon, Saskatchewan, Canada

¹⁷McGill University Health Centre, Montreal, Quebec, Canada

¹⁸University of Washington, Seattle, WA, USA

Introduction: Cervical spondylotic myelopathy (CSM) is the most common cause of spinal cord dysfunction globally. Surgical decompression has been shown to improve clinical outcomes. Nonetheless, many patients suffer from significant residual disability. Based on robust preclinical evidence from animal models of CSM, we hypothesized that the sodium/glutamate blocking drug riluzole may enhance neurological recovery and clinical outcomes following surgical decompression for CSM. Material and Methods: In this phase III multicenter, double-blinded, placebo-controlled randomized trial (ClinicalTrials.gov NCT01257828), patients with moderately severe CSM (modified Japanese Orthopaedic Association [mJOA] score 8 to 14) were randomized to receive riluzole (50 mg orally every 12 hours for 14 days before surgery and continued for 28 days after surgery) or a similar regimen of placebo. All patients received surgical decompression. Patients with a history of prior cervical spine surgery were excluded. Subjects, physicians, and evaluators remained blinded to treatment allocation throughout randomization and follow-up. The primary outcome was the change in mJOA score from baseline to 6 months postoperatively. Secondary outcomes included measures of function (Nurick grade), disability (Neck Disability Index [NDI]), health-related quality of life (QOL) (Short Form–36 [SF-36]), health utilities (EQ-5D), neurological function (American Spinal Injury Association [ASIA] motor and sensory scores), grip strength, and pain (visual analogue scale [VAS]). Outcomes were evaluated at enrollment, preoperative hospital admission, 35-day, 6-month, and 1-year follow-ups. Outcomes were analyzed by mixed-model for repeated measures; follow-up specific least squares means were compared between riluzole and placebo groups. Missing observations were multiply imputed. Results: A total of 300 patients were randomized and 290 underwent decompression surgery. Of these, 141 patients received riluzole and 149 placebo. Mean age was 58.0 ± 10.1 years. There were 129 (44.5%) females. The follow-up rate was 90% at 6 months. Subjects in both trial arms improved with regard to all endpoints for functional status, disability, QOL, neurological function, grip strength, and pain. There was no difference between riluzole and placebo groups in change in mJOA score at 6 months (2.45 and 2.82, respectively; P = .16), and similarly at 12 months (2.49 vs 2.83; P = .16). Likewise, the extent of improvement in Nurick grade, NDI, SF-36, EQ-5D, ASIA motor and sensory scores, pain at arm and grip strength were comparable between the treatment groups at 6- and 12-month follow-up. In unadjusted analysis, patients treated with riluzole showed a significantly greater reduction in neck pain (VAS) at 35 days than those who received placebo; this effect was maintained at 6 months and 1 year (P < .05). Conclusion: Adjuvant treatment with riluzole for 6 weeks perioperatively does not enhance functional and neurological recovery of surgical decompression in the setting of CSM. The potential effects of riluzole in reducing pain in surgically treated CSM patients merit further study.

A334: Burnout and Quality of Life Among Spine Surgeons: Results of a Worldwide Survey

Alisson R. Teles¹, Fahad Abduljabbar², Jean A. Ouellet¹, Catherine E. Ferland¹, Chung C. Wong³, Giuseppe Barbagallo⁴, Darrel Brodke⁵, Abdulaziz Al-Mutair⁶, and Asdrubal Falavigna⁷

¹McGill Scoliosis and Spine Group, McGill University, Montreal, Quebec, Canada

²King Abdulaziz University, Jeddah, Saudi Arabia

³Sarawak General Hospital, Jalan Tun Ahmad Zaidi Adruce, Kuching, Malaysia

⁴Policlinico “G. Rodolico” University Hospital, Catania, Italy

⁵University of Utah, Salt Lake City, UT, USA

⁶Alrazi Hospital, Ministry of Health/Kuwait Institute for Medical Specialization, Adailiya, Kuwait

⁷Universidade de Caxias do Sul, Caxias do Sul, Brazil

Introduction: Burnout is a syndrome characterized by emotional exhaustion, depersonalization, and decreased sense of accomplishment that leads to decreased effectiveness at work. To date, there is a lack of information on the prevalence of burnout among spine surgeons worldwide and the risk factors associated with this condition. Our objective was to evaluate the prevalence of burnout, to assess the personal and professional characteristics associated with burnout, and to evaluate its relationship with quality of life in spine surgeons. Material and Methods: An electronic survey with members of AOSpine was performed in May 2018. The survey evaluated demographic variables, practice characteristics, burnout, and quality of life. English, Spanish, and Portuguese versions of the Maslach Burnout Questionnaire and EQ5D were used to evaluate burnout and quality of life, respectively. Univariate and multiple regressions were used to identify factors associated with higher burnout scores and quality of life. Results: A total of 818 surgeons from 87 countries answered the survey. The majority of participants were from Latin America (LA = 61.6%) followed by Europe (E = 16.6%), Asia (A = 11.1%), North America (NA = 3.9%), Africa (3.4%), Middle East (ME = 2.7%), and Oceania (0.6%). Majority of respondents were male (93.4%), married (76.3%), had at least 1 child (74%), orthopedic surgeons (62.2%), with more than 10 years in practice (45.2%), working between 40 and 60 hours per week (54.4%), and treating both adult and pediatric spinal pathologies (53.6%). The prevalence of burnout was 30.6%. High levels of emotional fatigue and depersonalization were observed in 18.1% and 23.2%, respectively; and low levels of personal fulfilment in 21%. Independent risk factors for burnout were currently being fellow/resident (adjusted odds ratio [AOR] = 2.16, P = .003), work more than 60 hours per week (AOR = 1.46, P = .032), and not practicing in Latin America (A-OR: NA = 2.79, E = 1.62, ME = 1.45, A = 1.78). Higher levels of emotional fatigue (R = −0.35, P < .0001), depersonalization (R = −0.21, P < .0001), and lower levels of personal fulfilment (R = 0.22, P < .0001) were associated with lower scores of quality of life in EQ5D. Conclusion: Burnout is a common condition among spine surgeons worldwide. The overall prevalence was 30.6%, being lower in Latin America (26.2%) and higher in North America (48%). There is a significant association between burnout scores and decreased general quality of life. These results highlight the need for development of interventional programs to better identify, prevent and manage this condition among practicing spine surgeons.

A335: The Trans-Sap Approach: A Novel Technique for Endoscopic Decompression

Saqib Hasan¹, Brie White-Dzuro¹, Ralf Wagner², and Christoph Hofstetter¹

¹The University of Washington, Seattle, WA, USA

²Ligamenta Wirbelsäulenzentrum, Frankfurt, Germany

Introduction: Endoscopic approaches to decompress the lumbar spine have been demonstrated to allow for similar clinical outcomes compared to traditional open and tubular techniques with favorable risk profiles, faster recovery, and high patient satisfaction. Transforaminal approaches to the lumbar spine are well established and are typically performed via the medial aspect of the foraminal annular window (Kambin’s triangle). The traditional transforaminal approach allows for access to the ventral portion of the lateral recess ventral to the traversing nerve root. Here, we describe a modified endoscopic transforaminal procedure that utilizes a safe approach corridor via the superior articular process (SAP) of the index level and provides circumferential access to the lateral recess. Methods: Twenty-two patients were identified in our prospective database of endoscopic spine procedures performed at 2 institutions who had undergone a trans-SAP endoscopic approach for decompression of lumbar spine pathology. Patients included in the current cohort had failed conservative therapy including at least 6 weeks of physiotherapy, nonsteroidal anti-inflammatory drugs (NSAIDs) or epidural steroid injections and were excluded if they had symptomatic central canal stenosis, symptomatic contralateral lateral recess stenosis, dynamic instability, or a sequestered disc fragment. The endoscopic trans-SAP was performed by 2 senior surgeons with experience in endoscopic spine surgery. Briefly, the trans-SAP technique is performed using an oblique posterolateral trajectory targeting the lateral aspect of the SAP of the pathologic level. A Jamshidi needle is advanced into the SAP toward the mid-pedicle line. We then use sequential reaming of the SAP to create a safe working corridor to access the lateral recess and foramen of the index level. Bipolar electrocautery, an articulating drill, pituitary and kerrison rongeurs are subsequently utilized to decompress the neural elements. Outcomes were measured using visual analogue scale (VAS) and Oswestry DIsability Index (ODI) scores at 2 weeks, 3 months, and at last follow-up. Continuous variables are shown as means ± standard error of the mean (SEM). Repeated measurements, including VAS and ODI, were compared using a 1-way analysis of variance and paired-sample t test with SPSS 24 for Mac. Results: The current study includes 22 patients with unilateral symptomatic stenosis with an average follow-up period of 9.8 months ± 38.6 days. Of the 22 patients studied, there were 15 males and 7 females, average age was 60.7 ± 2.8 years. Average estimated blood loss (EBL) per level was 10.8 ± 1.6 mL and average surgical time per level was 80.6 ± 8.2 minutes. The most common surgical level was L4/5 (56.5%) and L3/4 (21.7%). Operative details are listed in Table 1. Patients experienced significant improvement of the VAS for ipsilateral leg pain when comparing pre-operative values to all postoperative time points (preoperatively 6.8 ± 0.44 vs 2 weeks postoperatively 4.1 ± 0.73, 3 months postoperatively 2.9 ± 0.57 and at last follow-up 3.1 ± 0.57, P < .001, Figure 1). A similar degree of improvement was found when comparing VAS back pain scores preoperatively to all postoperative time points (preoperatively 5.5 ± 0.61 vs 2 weeks postoperatively 3.4 ± 0.58, 3 months postoperatively 2.6 ± 0.45, and at last follow-up 3.8 ± 0.52, P < .001). Prior to surgery patients were moderately disabled as indicated by an ODI of 59 ± 2.7, at last follow-up there was significant improvement in ODI scores (42.6 ± 2.7, P < .001, Figure 1). The percentage of patients reaching minimal clinically important differece (MCID) for VAS leg pain was approximately 74%, while the percentage of patients reaching MCID for ODI was 71%. The complication profile was favorable with no dural tears and no postoperative iatrogenic motor or sensory deficits. A total of 2 patients required revision surgeries and were converted to fusion after persistent leg and back pain. Conclusions: The trans-SAP approach is a novel endoscopic technique that utilizes the superior articular process for docking of the endoscope and provides a safe and reliable anatomic landmark for decompression lumbar spine pathology. Docking of the endoscope onto bone rather than docking into Kambin’s triangle in the traditional trans-foraminal approach may minimize the risk of iatrogenic injury to the exiting nerve root in particular in the case of degenerative facet hypertrophy. Our patient cohort demonstrates significant improvement in pain and functional outcome scores with low complication rates.

Figure 1.

Graphs depicting clinical outcomes following endoscopic decompression via the trans-SAP (superior articular process) approach. (A) Reduction of the visual analog score for ipsilateral leg pain are recorded; (B) patients experience alleviation of their disability as measured by the Oswestry Disability Index.

A336: Predictors of Functional Recovery in Traumatic Central Cord Syndrome

Hetshree Joshi¹, Jetan Badhiwala¹, Jefferson Wilson², Robert Grossman³, Bizhan Aarabi⁴, Michael Bracken⁵, and Michael Fehlings¹

¹Toronto Western Hospital, University of Toronto, Toronto, Ontario, Canada

²St Michael’s Hospital, University of Toronto, Toronto, Ontario, Canada

³Houston Methodist Hospital, Houston, TX, USA

⁴University of Maryland, Baltimore, MD, USA

⁵ Yale Center for Perinatal, Pediatric and Environmental Epidemiology, New Haven, CT, USA

Introduction: Traumatic central cord syndrome (TCCS) refers to traumatic injury of the cervical spinal cord resulting in greater impairment in upper compared to lower extremities, bladder dysfunction, and a varying degree of sensory loss below the level of the injury.¹ In elderly patients, TCCS typically results from a hyperextension injury in the setting of a stenotic spondylotic canal.¹ With the growing elderly population, TCCS is soon expected to become the most common type of traumatic spinal cord injury. Therefore, identifying effective management strategies to improve functional outcomes is a priority. With advances in surgical care and technology, there has been a shift in treatment paradigms for TCCS toward more proactive strategies involving surgical decompression.² Nonetheless, the question of timing of surgery remains unanswered.^3-5 Materials and Methods: Data was merged from 3 large, multicenter, international databases, including (1) NASCIS, (2) NACTN, and (3) STASCIS. Patients 16 years of age or older with a diagnosis of TCCS (defined as American Spinal Injury Association [ASIA] lower extremity motor score (LEMS) > ASIA upper extremity motor score (UEMS) were identified. The primary outcome was change in Functional Independence Measure (FIM) motor subscore at 1 year. Regression techniques were used to evaluate possible predictors of FIM motor subscore, including age, sex, comorbidities, baseline neurological status, time to surgery, and surgical approach. Results: A total 170 patients with TCCS with data on 1-year follow up functional outcomes met the eligibility criteria. Baseline characteristics including age, weight, height, and gender were measured. The most common mechanism of injury was fall (79/170; 46.47%), followed by followed by motor vehicle accident (59/170; 34.71%). Initial analysis with functional outcomes showed only 19 patients were completely functionally independent in daily living activities based on the motor subscale of FIM. Univariate analysis assessing improvement in FIM motor subscore showed significance in baseline AMS (ASIA Motor Score) (P < .001), presence of fracture (P = .012), time to surgery (P = .045), initial AIS (American Spinal Injusry Association Impairment Scale) grade (P = .001), and initial GCS (Glasgow Coma Scale) score (P = .021). Using this, we generated a multivariate regression model with factors including age, baseline AMS, time to surgery, and presence of fracture. This model showed a significant positive correlation of baseline AMS (P = .001) and negative correlation of time to surgery (P = .039) on improvement in functional motor outcomes in patients with TCCS. Age, sex, weight, height, systolic blood pressure, steroid use, and type of injury did not show any significance in predicting 1-year FIM motor scores. Conclusion: Our findings demonstrate significant associations of initial AMS, presence of fracture/dislocation, and time to surgery with functional motor recovery. Earlier surgical decompression in the setting of TCCS may result in improved functional outcomes.

Acknowledgments

We would like to acknowledge the Fehlings Lab, University of Toronto, University Health Network, and the Neurosurgery Research and Education Foundation.

References

1. Brooks NP. Central cord syndrome. Neurosurg Clin N Am. 2017;28:41-47.

2. Jin W, Sun X, Shen K, et al. Recurrent neurological deterioration after conservative treatment for acute traumatic central cord syndrome without bony injury: seventeen operative case reports. J Neurotrauma. 2017;34:3051-3057.

3. Lenehan B, Fisher CG, Vaccaro A, Fehlings M, Aarabi B, Dvorak MF. The urgency of surgical decompression in acute central cord injuries with spondylosis and without instability. Spine (Phila Pa 1976). 2010;35(21 suppl):S180-S186.

4. Molliqaj G, Payer M, Schaller K, Tessitore E. Acute traumatic central cord syndrome: a comprehensive review. Neurochirurgie. 2014;60:5-11.

5. Anderson KK, Tetreault L, Shamji MF, et al. Optimal timing of surgical decompression for acute traumatic central cord syndrome: a systematic review of the literature. Neurosurgery. 2015;77(suppl 4):S15-S32.

A337: Characteristics of Upper Limb Impairment Related to Degenerative Cervical Myelopathy: Development of a Sensitive Hand Assessment (GRASSP Version Myelopathy)

Sukhvinder Kalsi-Ryan¹, Eric Massicotte², Allan Martin², and Michael Fehlings²

¹University of Toronto, Toronto, Ontario, Canada

²Toronto Western Hospital, Toronto, Ontario, Canada

Introduction: Degenerative cervical myelopathy (DCM) involves spinal cord compression, which causes neurological decline. Neurological impairment in DCM is variable and can involve complex upper limb dysfunction, including loss of manual dexterity, hyper-reflexia, focal weakness, and sensory impairment. The modified Japanese Orthopedic Association (mJOA) score, relies on the patients’ subjective perceptions, while objective measures such as strength and sensory testing do not capture subtle changes in dexterity and function. Objective: (1) Characterize hand function in the DCM population; (2) develop and validate the GRASSP Version Myelopathy (GRASSP-M), a clinical assessment tool that quantifies upper limb impairment in all severities of DCM. Materials and Methods: A single-center, observational, cross-sectional study. A total of 148 patients with DCM (categorized into mild, moderate, and severe based on mJOA grade) and 21 healthy subjects were enrolled. A complete neurological exam, the mJOA, the QuickDASH, grip dynamometry, and the GRASSP-M were administered. The sample was stratified according to DCM severity using the mJOA. Patients were then subjected to a battery of clinical assessments, including the GRASSP-M. Results: Strength, sensation, and manual dexterity significantly declined with increasing DCM severity (P ≤ .05). Impairment in hand dexterity showed better discrimination between mild, moderate, and severe DCM categories than strength or sensation. In addition, the GRASSP-M was found to be both a reliable (intraclass correlation coefficient [ICC] > 0.75 for intra- and interrater reliability) and valid (with both concurrent and construct validity) tool. Conclusion: These results show that patients’ subjective reporting of their functional status, especially in the mild DCM category, may underrepresent the extent of functional impairment. The GRASSP-M is an objective tool designed to fully characterize patients’ functional impairment related to the upper limb, which proves useful to diagnose and quantify mild dysfunction, monitor patients for deterioration, and help determine when patients should be treated surgically.

A338: The Fate of High Pelvic Tilt With Good Alignment (False Type 2) in Adult Spinal Deformity Surgery Fused to the Sacrum

Thamrong Lertudomphonwanit¹, Alekos Theologis², Munish Gupta³, Lawrence Lenke⁴, Keith Bridwell³, and Michael Kelly³

¹Ramathibodi Hospital, Mahidol University, Bangkok, Thailand

²University of California, San Francisco, San Francisco, CA, USA

³Washington University in St Louis, St Louis, MO, USA

⁴Columbia University Medical Center, New York, NY, USA

Introduction: Recent works suggest that postoperative high pelvic tilt (PT) with good overall alignment (the Roussouly “false” type 2) is associated with increased rates of proximal junction kyphosis (PJK) and pseudarthrosis. The Global Alignment and Proportion (GAP) Score has proposed regional alignment goals. This study aims to investigate the durability of adult spinal deformity (ASD) surgeries fixed with persistent lumbopelvic parameter mismatch; high PT in patients with high pelvic incidence (PI). We hypothesized that patients with persistent lumbopelvic malalignment after ASD surgery will have higher pseudarthrosis and PJK rates than historical norms for “well-aligned” patients. Materials and Methods: Review of adults who underwent posterior spinal instrumented fusions (including sacrum; ≥5 levels) for spinal deformity from 2004 to 2014 at a single center. Inclusion criteria: >2 years follow-up, PI > 55°, postoperative sagittal vertical axis (SVA) < 10 cm, and high postoperative PT (calculated as PT > 0.41 * PI – 9 (ideal pelvic alignment based on GAP Score]). Demographics, radiographic parameters at baseline and at follow-up, and surgical techniques were analyzed. Mechanical failures, including PJK (Glattes criteria), and pseudarthrosis were determined. Paired Student t tests were used for intragroup comparisons for continuous data. Results: Eighty-three patients (male 10; female 73; average age 61 ± 10 years; range, 29-78 years) were included for analysis. Average follow-up was 54 ± 27 months (range, 24-154 months). Average preoperative deformity parameters included: C7 SVA 8.7 ± 6.9 cm, lumbar lordosis (LL) −30.6° ± 19.7°, PI-LL 34.1° ± 20.3°, PT 35.5° ± 10.3°, maximum Cobb angle 43.3° ± 24.7°, central sacral vertical line (CSVL) 3.1 ± 2.6 cm. Surgical intervention was performed most commonly in the revision setting (58%) via a posterior-only approach (88%) and included interbody fusions (57%) and bone morphogenetic protein–2 (BMP2) (100%). Average number of levels instrumented/fused was 13.3 ± 4.1 (upper thoracic upper instrumented vertebrae [UIV] [T1-5]-43; middle thoracic UIV [T6-T9]-8; lower thoracic UIV [T10-L2]-32). At early postoperative period, there were significant improvements in all sagittal and coronal alignment parameters compared with preoperative values (C7SVA = 3.4 ± 3.8 cm, LL = −44.1° ± 12°, PI-LL = 20.4° ± 13.3°, PT = 32.3° ± 6.6°, maximum Cobb angle = 22.2° ± 17.8°, CSVL = 2.4 ± 1.8 cm; all P < .05). At final follow-up, significant improvements still remained in all sagittal and coronal alignment parameters (all P < .05), although average PI-LL (24.1° ± 13.5°) and PT (33.2° ± 8.3°) remained undercorrected. Proximal junctional angle between UIV and UIV+2 significantly increased after operation (preoperative period: 0.7° ± 8.4°; final follow-up 13.2° ± 9.1°; P < .01). The rate of radiographic PJK was 48.2%. The rate of revision operation for PJK was 2.4%. The rate of instrumentation failure was 25.3%. Revision for pseudarthrosis was 9.6%. Conclusion: In this single-center experience, well-balanced patients with high pelvic incidence fixed with persistent lumbopelvic parameter mismatch and high PT (false Roussouly type 2) had rates PJK and pseudarthrosis consistent with historical norms for “well-aligned” patients. High postoperative pelvic tilt may be durable in some cases of ASD surgery, questioning proposed lumbopelvic alignment goals. Further investigations into the durability of high PT in patients with ASD are needed.