Faculty Bibliography

Introduction/UNASSIGNED:Early Onset and Adolescent Idiopathic Scoliosis, relatively common diagnoses (∼3% general population), have been associated with developmental dysplasia of the hip (DDH); a more rare spectrum of anomalies related to the abnormal development of acetabulum, proximal femur, and hip joint. To the best of our knowledge, no high powered investigations have been performed in an attempt to assess incidence and associated risks of DDH in scoliosis patients. Methods/UNASSIGNED:The KID database was queried for ICD-9 codes from 2003 to 2012 pertaining to EOS (Congenital and Idiopathic <10y/o) and AIS patients. Descriptive analysis assessed patient demographics and yearly trends in hip dysplasia rates. EOS and AIS patients with hip dysplasia were isolated, and incidence of hospital admissions for associated anomalies (osteonecrosis, osteoarthritis, recurrent hip dislocation, hip ankylosis) and hip arthroplasty (total + partial) were investigated. Univariate analysis of hip pathology determined significant predictors of hip arthroplasty. Binary logistic regression analysis was used to determine the relationship between these predictors. Results/UNASSIGNED:111,827 scoliosis patients (EOS: 25,747; AIS: 77,183) were included. AIS patients were older (15.2 vs 4.3), more female (64.2% vs 52.1%), had a higher CCI (0.84 vs 0.64), and less racially diverse (all p < 0.001). The incidence of hip dysplasia was 1.4% for AIS patients and 3.9% for EOS patients (p < 0.001). Of the AIS (n = 1073) and EOS (n = 1005) patients with hip dysplasia, 0.3% (p > 0.05 between groups) developed hip osteonecrosis, 0% of patients were coded as having a hip labral tear, hip ankylosis, and 0.6% (EOS: 0.2%; AIS: 0.9%, p = 0.025) developed hip osteoarthritis. AIS patients were more likely to have recurrent hip dislocations (35.4% vs 17.0%, p < 0.001), and both groups had similar primary hip arthroplasty rates (6.7% vs 5.4%, p = 0.118) and revision hip arthroplasty rates (0% vs 0.4%, p = 0.053). Hip osteoarthritis (OR: 13.43[5.21-34.66], p=<0.001) and older age (OR: 1.039[1.007-1.073], p = 0.017) were the only significant predictors of hip arthroplasty (p=<.001). Conclusions/UNASSIGNED:The incidence of hip dysplasia in EOS and AIS populations is higher than that of the general population. The rate of DDH was 3.9% and 1.8% for EOS and AIS, respectively. While the incidence of DDH is higher, associated anomalies of osteoarthritis, osteonecrosis, labral tears, and ankylosis appear to be a minimal risk for AIS and EOS patients with Hip Dysplasia.

Study Design/UNASSIGNED:Prospective cohort study. Objectives/UNASSIGNED:To describe the rate of short-term complications following the posterior use of recombinant human bone morphogenetic protein-2 (rhBMP-2) in cervical deformity (CD) surgery. Methods/UNASSIGNED:CD patients from 2013 to 2015 were enrolled in a prospective, multicenter database. Patients were divided into those receiving rhBMP-2 (BMP) and no rhBMP-2 (NOBMP). The relationship between BMP use, demographic variables surgical variables, radiographic parameters and complications was evaluated. Results/UNASSIGNED:= 0.09, 0.08, 0.06) between the use of BMP and complications (major or operative). Conclusions/UNASSIGNED:BMP use was not directly associated with an increased incidence of early complications in this prospective cohort of operative adult CD patients. Its use was associated with increased number of levels instrumented and fused.

BACKGROUND CONTEXT/BACKGROUND:Cost-utility analysis, a special case of cost-effectiveness analysis, estimates the ratio between the cost of an intervention to the benefit it produces in number of quality-adjusted life years. Cervical deformity correction has not been evaluated in terms of cost-utility and in the context of value-based health care. Our objective, therefore, was to determine the cost-utility ratio of cervical deformity correction. STUDY DESIGN/METHODS:This is a retrospective review of a prospective, multicenter cervical deformity database. Patients with 1-year follow-up after surgical correction for cervical deformity were included. Cervical deformity was defined as the presence of at least one of the following: kyphosis (C2-C7 Cobb angle >10°), cervical scoliosis (coronal Cobb angle >10°), positive cervical sagittal malalignment (C2-C7 sagittal vertical axis >4 cm or T1-C6 >10°), or horizontal gaze impairment (chin-brow vertical angle >25°). Quality-adjusted life years were calculated by both EuroQol 5D (EQ5D) quality of life and Neck Disability Index (NDI) mapped to short form six dimensions (SF6D) index. Costs were assigned using Medicare 1-year average reimbursement for: 9+ level posterior fusions (PF), 4-8 level PF, 4-8 level PF with anterior fusion (AF), 2-3 level PF with AF, 4-8 level AF, and 4-8 level posterior refusion. Reoperations and deaths were added to cost and subtracted from utility, respectively. Quality-adjusted life year per dollar spent was calculated using standardized methodology at 1-year time point and subsequent time points relying on maintenance of 1-year utility. RESULTS:Eighty-four patients (average age: 61.2 years, 60% female, body mass index [BMI]: 30.1) were analyzed after cervical deformity correction (average levels fused: 7.2, osteotomy used: 50%). Costs associated with index procedures were 9+ level PF ($76,617), 4-8 level PF ($40,596), 4-8 level PF with AF ($67,098), 4-8 level AF ($31,392), and 4-8 level posterior refusion ($35,371). Average 1-year reimbursement of surgery was $55,097 at 1 year with eight revisions and three deaths accounted for. Cost per quality-adjusted life year (QALY) gained to 1-year follow-up was $646,958 by EQ5D and $477,316 by NDI SF6D. If 1-year benefit is sustained, upper threshold of cost-effectiveness is reached 3-4.5 years after intervention. CONCLUSIONS:Medicare 1-year average reimbursement compared with 1-year QALYdescribed $646,958 by EQ5D and $477,316 by NDI SF6D. Cervical deformity surgeries reach accepted cost-effectiveness thresholds when benefit is sustained 3-4.5 years. Longer follow-up is needed for a more definitive cost-analysis, but these data are an important first step in justifying cost-utility ratio for cervical deformity correction.

Optimizing functional outcomes and disability status are essential for effective surgical treatment of cervical spine disorders. Mental impairment is common among patients with cervical spine complaints; yet little is known about the impact of baseline mental status with respect to overall patient-reported outcomes. This was a retrospective analysis of patients with cervical spondylosis with myelopathy(CM) or radiculopathy(CR: cervical disc herniation, stenosis, or spondylosis without myelopathy) at 2-year follow-ups. Patients were assessed for several health-related quality of life HRQOL) measures at baseline and 24-months post-operatively: Neck Disability Index (NDI), Visual Analog Scale(VAS), Short Form-36(SF) Physical(PCS) and Mental(MCS) Components. Patients were dichotomized by MCS score: LOW-MCS(SF-MCS < 40th percentile) vs. HIGH-MCS(SF-MCS > 60th percentile). Independent and paired t-tests compared improvement in each group for HIGH-MCS and LOW-MCS cohorts. 375 patients were analyzed(65.4yrs, 67.6%F). LOW-MCS radiculopathy patients showed significant improvement in NDI, VAS Neck and Arm Pain(p < 0.05). HIGH-MCS radiculopathy patients showed greater improvement in NDI score, VAS Neck and Arm Pain, and improvement in PCS(all p < 0.05). Comparing baseline and 2-year follow-up, LOW-MCS CM patients showed significant improvement in PCS, NDI, VAS Neck and Arm Pain(p < 0.05). HIGH-MCS myelopathy patients group showed marked improvement in NDI scores, VAS Neck and Arm Pain(p < 0.05). LOW-MCS CR patients were more likely to be less satisfied 2-years post-op(p < 0.001). Postoperative CR patients with lower baseline mental status saw less improvement and significantly worse outcomes than patients with higher baseline mental status. Improving baseline mental health may improve post-operative recovery. Implementing additional screening and care can optimize functional outcomes and disability status for patients with CR.

BACKGROUND:Cardiac anomalies are prevalent in patients with bony spinal anomalies. Prior studies evaluating incidences of bony congenital anomalies of the spine are limited. The Kids' Inpatient Database (KID) yields national discharge estimates of rare pediatric conditions like congenital disorders. This study utilized cluster analysis to study patterns of concurrent vertebral anomalies, anal atresia, cardiac malformations, trachea-esophageal fistula, renal dysplasia, and limb anomalies (VACTERL anomalies) co-occurring in patients with spinal congenital anomalies. METHODS:Retrospective review of KID 2003-2012. KID-supplied hospital- and year-adjusted weights allowed for incidence assessment of bony spinal anomalies and cardiac, gastrointestinal, urinary anomalies of VACTERL. K-means clustering assessed relationships between most frequent anomalies within bony spinal anomaly discharges; k set to n - 1(n = first incidence of significant drop/little gain in sum of square errors within clusters). RESULTS:There were 12,039,432 KID patients 0-20 years. Incidence per 100,000 discharges: 2.5 congenital fusion of spine, 10.4 hemivertebra, 7.0 missing vertebra. The most common anomalies co-occurring with bony vertebral malformations were atrial septal defect (ASD 12.3%), large intestinal atresia (LIA 11.8%), and patent ductus arteriosus (PDA 10.4%). Top congenital cardiac anomalies in vertebral anomaly patients were ASD, PDA, and ventricular septal defect (VSD); all three anomalies co-occur at 6.6% rate in this vertebral anomaly population. Cluster analysis revealed that of bony anomaly discharges, 55.9% of those with PDA had ASD, 34.2% with VSD had PDA, 22.9% with LIA had ASD, 37.2% with ureter obstruction had LIA, and 35.5% with renal dysplasia had LIA. CONCLUSIONS:In vertebral anomaly patients, the most common co-occurring congenital anomalies were cardiac, renal, and gastrointestinal. Top congenital cardiac anomalies in vertebral anomaly patients were ASD, PDA, and VSD. VACTERL patients with vertebral anomalies commonly presented alongside cardiac and renal anomalies.

STUDY DESIGN/METHODS:Retrospective review. OBJECTIVE:Determine whether alignment or myelopathy improvement drives patient outcomes after cervical deformity (CD) corrective surgery. SUMMARY OF BACKGROUND DATA/BACKGROUND:CD correction involves radiographic malalignment correction and procedures to improve motor function and pain. It is unknown whether alignment or myelopathy improvement drives patient outcomes. METHODS:Inclusion: Patients with CD with baseline/1-year radiographic and outcome scores. Cervical alignment improvement was defined by improvement in Ames CD modifiers. modified Japanese Orthopaedic Association (mJOA) improvement was defined as mild [15-17], moderate [12-14], severe [<12]. Patient groups included those who only improved in alignment, those who only improved in mJOA, those who improved in both, and those who did not improve. Changes in quality-of-life scores (neck disability index [NDI], EuroQuol-5 dimensions [EQ-5D], mJOA) were evaluated between groups. RESULTS:A total of 70 patients (62 yr, 51% F) were included. Overall preoperative mJOA score was 13.04 ± 2.35. At baseline, 21 (30%) patients had mild myelopathy, 33 (47%) moderate, and 16 (23%) severe. Out of 70 patients 30 (44%) improved in mJOA and 13 (18.6%) met 1-year mJOA minimal clinically important difference. Distribution of improvement groups: 16/70 (23%) alignment-only improvement, 13 (19%) myelopathy-only improvement, 18 (26%) alignment and myelopathy improvement, and 23 (33%) no improvement. EQ-5D improved in 11 of 16 (69%) alignment-only patients, 11 of 18 (61%) myelopathy/alignment improvement, 13 of 13 (100%) myelopathy-only, and 10 of 23 (44%) no myelopathy/alignment improvement. There were no differences in decompression, baseline alignment, mJOA, EQ-5D, or NDI between groups. Patients who improved only in myelopathy showed significant differences in baseline-1Y EQ-5D (baseline: 0.74, 1 yr:0.83, P < 0.001). One-year C2-S1 sagittal vertical axis (SVA; mJOA r = -0.424, P = 0.002; EQ-5D r = -0.261, P = 0.050; NDI r = 0.321, P = 0.015) and C7-S1 SVA (mJOA r = -0.494, P < 0.001; EQ-5D r = -0.284, P = 0.031; NDI r = 0.334, P = 0.010) were correlated with improvement in health-related qualities of life. CONCLUSION/CONCLUSIONS:After CD-corrective surgery, improvements in myelopathy symptoms and functional score were associated with superior 1-year patient-reported outcomes. Although there were no relationships between cervical-specific sagittal parameters and patient outcomes, global parameters of C2-S1 SVA and C7-S1 SVA showed significant correlations with overall 1-year mJOA, EQ-5D, and NDI. These results highlight myelopathy improvement as a key driver of patient-reported outcomes, and confirm the importance of sagittal alignment in patients with CD. LEVEL OF EVIDENCE/METHODS:3.

STUDY DESIGN/METHODS:Retrospective radiographic review. OBJECTIVES/OBJECTIVE:To understand the effect of variability in sacral endplate selection in transitional lumbosacral vertebrae (TLSV) and its impact on pelvic, regional, and global spinal alignment parameters. BACKGROUND:TLSV can have the characteristics of both lumbar and sacral vertebrae. Difficulties in identification of the S1 endplate may come from nomenclature, number of lumbar vertebrae, sacra, and morphology and may influence the interpretation and consistency of spinal alignment parameters. METHODS:Patients with TLSV were identified and radiographic measurements including pelvic incidence (PI), pelvic tilt (PT), sacral slope (SS), lumbar lordosis (LL), sagittal vertical axis (SVA), T1-pelvic angle (TPA), pelvic incidence-lumbar lordosis (PI-LL) mismatch, thoracic kyphosis (TK), and spinal inclination (T1SPi) were obtained. Radiographic measurements were performed twice with the sacral endplate at the cephalad and caudal options. Paired t tests assessed the difference between different selection groups. RESULTS:Of 1,869 patients, 70 (3.7%) were found to have TLSV on radiographic imaging. Fifty-eight (82.9%) had lumbarized sacral segments whereas 12 (17.1%) had sacralized lumbar segments. T1-SPi (mean: -1.77°) and TK (mean: 34.86°) did not vary from altering sacral endplate selection. Selection of the caudal TLSV as the sacral endplate resulted in an increase in all pelvic parameters (PI: 66.8° vs. 44.3°, PT: 25.1° vs. 12.7°, and SS: 41.6° vs. 31.6°), regional lumbar parameters (LL: -54.1° vs. 44.0°, PI-LL: 12.7° vs. 0.3°), and global parameters (SVA: 46.1 mm vs. 28.3 mm, TPA: 23.3° vs. 10.8°) as compared to selecting the cephalad TLSV. All mean differences between radiographic parameters were found to be statistically significant (p < .001). CONCLUSIONS:Variation in sacral endplate selection in TLSV significantly affects spinal alignment parameter measurements. A standardized method for measuring TLSV is needed to reduce measurement error and ultimately allow more accurate understanding of alignment targets in patients with TLSV. LEVEL OF EVIDENCE/METHODS:Level III.

Introduction: Cervical deformity (CD) correction has becomeincreasingly complex and challenging. Osteotomies are commonlyperformed to correct sagittal malalignment, however the risks andbenefits of performing a major osteotomy for cervical deformitycorrection have been understudied. The purpose of this study was toinvestigate the risks and benefits of performing a major osteotomy forCD correction.Methods: Retrospective review of a multicenter prospective CDdatabase. CD was defined as at least one of the following: C2-C7Cobb [10, CL [10, cSVA [4 cm, CBVA [25. Patientsstratified based on having a major osteotomy (MAJ-pedicle subtraction osteotomy or vertebral column resection) or minor (MIN).Propensity score matching (PSM) was performed controlling forbaseline cSVA and T1S. Flexibility of the deformity was assessedusing C2-C7 lordosis and T1S change greater than 10 betweenflexion and extension. Independent t-tests and Chi Squared tests wereused to assess differences between MAJ and MIN.Results: 89 CD patients were included (62 years, 65%F). 19 (21.3%)CD patients underwent a MAJ osteotomy. MAJ and MIN had nodifferences in any baseline radiographic parameters, with the exception of cSVA (MAJ: 59.3 mm, MIN: 41.9 mm, p = 0.007). AfterPSM for cSVA, 38 patients were included (60 years, 60%F). 19(21.3%) CD patients underwent a MAJ osteotomy (14 pedicle subtraction osteotomy, 5 vertebral column resection). MAJ patientsunderwent more invasive surgeries, with more levels fused (10.6 vs7.1, p <0.001) and blood loss (1442 cc vs 802 cc, p = 0.036),despite similar operative time and intra-and post-operative complication rates as MIN patients. At 3 M post-op, MAJ and MIN patientshad similar NDI, mJOA, and EQ5D scores, however by 6M and 1Ypost-op MAJ patients reached MCID for NDI less than MIN patients(10.5 vs 57.9%, p = 0.003). Comparing patients with fixed versusnon-fixed CL, MAJ patients with non-fixed lordosis trended towardsimprovement in NDI (p = 0.30) but also trended towards highercomplication (78 vs 43%, p = 0.182) and reoperation rates (44 vs 0%,p = 0.069) than fixed deformities. Rigid deformities trended towardsimprovement in TS-CL (43% improve vs 33%, p = 0.54) and cSVA(14 vs 0%, p = 0.49) for MAJ patients and lower complication rate(MIN most commonly had DJK and reoperation) (43 vs 100%,p = 0.09).Conclusions: Cervical deformity patients who underwent a majorosteotomy had similar clinical outcomes at 3-months but worseclinical outcomes at 6-months and 1-year, assessed by NDI and EQ-5D, as compared with patients with minor osteotomies, in partbecause patients undergoing major osteotomies have more severedeformities and have more prolonged recovery kinetics. Patients withflexible curves showed similar alignment and clinical outcomes butincreased complication risk when undergoing a major osteotomy.Contrarily, patients with rigid deformities who underwent a majorosteotomy trended towards radiographic and clinical improvementand lower rates of DJK and reoperation

Background: Recent studies suggest non-routine discharge, includingdischarge to inpatient rehab and skilled nursing facilities, is associatedwith increased cost of care. Given the rising prevalence of cervicaldeformity (CD)-corrective surgery and the necessity of value-basedhealthcare, it is important to identify indicators for non-routine discharge in surgical CD patients.Study Design: Retrospective review of prospective, multicenter CDdatabase.Methods: Included: Surgical CD patients (C2-C7 Cobb [10,CL [10, cSVA [4 cm, or CBVA [25) [18 years with discharge and baseline (BL) radiographic data. Non-routine dischargedefined: inpatient rehab or skilled nursing facility. ConditionalInference Decision Trees identified predictors of non-routine discharge, and cut-off points at which predictors have a global effect.A Conditional Variable Importance Table used non-replacementsampling set of 3000 Conditional Inference trees to identify influential patient/surgical factors. Binary logistic regression indicated effectsize of influential factors at significant cut-off points. Means comparison testing assessed the relationship between non-routinedischarge and reop/HRQL outcomes.Results: Included: 138 patients (61 +/- 10 years, 63%F) undergoingCD-corrective surgery (8.2 +/- 4.6 levels; 49% posterior-onlyapproach, 16% anterior-only, 35% combined). 29% of patientsexperienced non-routine discharge (21% inpatient rehab, 8% SNF).BL cervical and upper-cervical malalignment was the strongest predictor of non-routine discharge: [1] C1 slope [14 (OR:8.4 [95%CI:3.1-22.7]), [3] C2 slope [57 (OR: 7.0 [2.6-18.3]), [4] TSCL [57 (OR: 5.9 [2.2-15.9]), [14] C0 slope[-0.66 (OR: 4.2[1.9-9.3]), [15] cSVA [40 mm (OR: 4.6 [2.0-10.9]), [18] McGregor's slope [1.9 (OR: 4.1 [1.7-9.9]). Patient-related predictors ofnon-routine discharge were [2] BL gait impairment (OR: 5.29[2.3-12.4]), [8] age [59 years (OR: 4.3 [1.6-11.1]), [10] apex of CDprimary driver [C7 (OR: 3.9[1.8-8.6]), and [13] admission tosurgical ICU (OR: 5.4 [1.9-14.8]). Experiencing 2 or more complications was predictive of non-routine discharge (OR: 4.2 [1.9-9.2]),but the only specific complications predictive of non-routone discharge were EBL [900 cc (OR: 3.6 [1.7-7.7]) and presence of anyneuro complication (OR: 2.8 [1.8-8.4]). The only surgical predictor ofnon-home discharge was [12] fusion[8 levels (OR: 4.0 [1.8-9.0]).LOS [6 days was also predictive of non-routone discharge (OR: 4.0[1.8-8.9]). There was no relationship between non-routine dischargeand reop within 3 months (P = 0.249), 6 months (P = 0.793), or1 year (P = 0.814) of index procedure. Despite no differences in BLEQ-5D (P = 0.946), non-routine patients had inferior 1-year postopEQ-5D scores (non-routine: 0.75, home: 0.79, P = 0.044).Conclusions: Preop cervical malalignment was a top predictor ofnon-routine discharge in surgical CD patients. Age, driver of deformity, and [8 level fusion also predicted non-routine discharge, andshould be taken into account to improve resource allocation andpatient counseling

Summary: In cervical deformity surgery, failure to correct cSVA andC2 slope (C2S) is associated with poor clinical outcomes. Currentsurgical planning and intraoperative measurements are limited tocervical lordosis (CL) correction. By predicting T1 Slope (T1S)change from baseline to 3 month postop and adding a correctionfactor for the change in distal junctional kyphosis angle (DJKA),cSVA and C2S can be predicted more accurately.Hypothesis: While correction of CL is commonly used to predictpost-op alignment, post-op cSVA, and C2S can be predicted betterusing additional variables.Design: Prospective cohort study.Introduction: Cervical malalignment is associated with severe disability. Currently, surgical planning and intraop measurements ofcorrection are limited to CL change. We aim to develop a predictivemodel for postop cSVA C2S and T1S using more than CL change.Methods: A prospective database of operative CD patients wasanalyzed. Inclusion criteria were cervical kyphosis [10, cervicalscoliosis [10, cSVA [4 cm or CBVA [25. The patients wererandomly filtered to include 66.7% of the cohort for model development. Predictive models were developed to estimate post-op T1S,cSVA, and C2S using linear regression. The new predictive equationswere validated in the remaining 33.3% of the cohort.Results: 153 patients with CD met inclusion criteria. T1S changedsignificantly (32.4-35.2, p =.05) from baseline to 3M follow-up.The mean DJKA change was-6.59. 101 patients were included inmodel development. To predict post-opT1S, CLchange and preT1Sexplained 62.4% of the variability of data (R2 = 0.624). By includingDJKA, R2 improved to 0.724. When predicting postop cSVA,CLchange and preop cSVA accounted for 57.2% of variability(R2 = 0.572). With change in DJKA, the R2 improved to 0.661. Themodel was optimized with the change in T1S (R2 = 0.777). Pre-opC2S and CLchange lead to poor predictability in post-op C2S(R2 = 0.348). Using DJKA change, the R2 improved to 0.550. Byincluding DJKA and T1S change, the model was optimized(R2 = 0.926).The predictive equations were applied to the remaining 52 patientsusing the mean DJKA (-6.59) for the DJKA variable and the T1Schange calculated for the predictive model. Predicted postop alignments correlated to postop T1S, cSVA, and C2S (R = 0.712,R = 0.736, and R = 0.584 respectively, p <0.01).Conclusions: Realignment in CD surgery is critical to obtain optimaloutcomes. Current surgical planning of CD does not accurately predict postop cSVA and C2S. A reliable predictive model is presentedfor cSVA and C2S using changes in T1S and DJKA