Oral Posters: Values and Outcomes in Spine Surgery

Presented by: J. Elysée - View Audio/Video Presentation (Members Only)


R. Lafage(1), J. Elysée(1), B. Liabaud(1), R. Hart(2), J. Smith(3), H.J. Kim(1), P. Passias(4), C. Ames(5), G. Mundis(6), D. Burton(7), S. Bess(4), E. Klineberg(8), F. Schwab(1), V. Lafage(1), International Spine Study Group (ISSG)

(1) Hospital for Special Surgery, Spine Service, New York, NY, United States
(2) Oregon Health Sciences University, Department of Orthopaedic Surgery, Portland, OR, United States
(3) University of Virginia Medical Center, Department of Neurosurgery, Charlottesville, VA, United States
(4) NYU Langone Medical Center, Spine Division, Department of Orthopaedics, New York, NY, United States
(5) San Francisco Medical Center, University of California, Department of Neurosurgery, San Francisco, NY, United States
(6) Scripps Clinic Torrey Pines, La Jolla, CA, United States
(7) University of Kansas Medical Center, Department of Orthopaedic Surgery, Kansas City, KS, United States
(8) University of California, Davis, Department of Orthopedic Surgery, Sacramento, CA, United States


Introduction: Surgical correction of regional malalignment is a mainstay in ASD reconstruction, but its impact on the unfused spinal segments (reciprocal changes) is difficult to anticipate. In addition, focal change at the UIV (PJK) can occur and influence final alignment. This study aims to analyze the relationship between thoracic reciprocal changes and the correction of thoracolumbar deformity in order to be able to predict reciprocal changes in the unfused thoracic spine as well as the UIV level following ASD surgery.

Methods: Inclusion criteria were patients with upper-most instrumented vertebra (UIV) between T6-L1 and fusion to the sacrum/pelvis with minimum 2-year follow-up. Radiographic parameters included common spino-pelvic parameters (PT, PI-LL, TK, and SVA) and fused/unfused thoracic kyphosis measurements (TKfused/TKunfused). Correlation analyses and multi-linear regressions were performed using changes in the fused thoracolumbar spine to predict reciprocal changes in the unfused thoracic spine. Sub analysis on prediction of change at the UIV junction was carried out and subsequent PJK developed post operatively was predicted using a similar approach. Predictive models were developed using a random selection of75% of the population and validated with the remaining 25%.

Results: 128 patients were included (mean=63yo, 75% F). All sagittal parameters improved from pre- to post-op. ΔTK correlated significantly with ΔPI-LL with and without control for ΔTKfused (r=0.566 and r=0.591). Regression through the origin between ΔTKunfused using ΔTKfused and ΔPI-LL had an r2 of 0.841. Validation of the prediction on the remaining 25% of patients revealed a mean error of -2.3˚ and an RMSE of 9˚. Similarly, prediction of change at the UIV junction revealed an r2 of 0.769 with a mean error of -1.8˚ and RMSE of 11˚. Prediction of PJK post-operatively using baseline alignment and prediction of the UIV change showed an overall accuracy of 68%, with a precision of 82%.

Conclusion: Pre-op planning for ASD surgery is complex; changes occurring in the unfused segments of the spine may affect the final global alignment. In this study, significant correlations between changes in TK and PI-LL were noted. Furthermore, predictive models for change on the unfused TK and at the UIV level demonstrated good results (r2>0.760). These findings can be helpful in anticipating reciprocal changes in an effort to reduce suboptimal post-op alignment and focal deformity (PJK).