Full-body radiographic analysis of postoperative deviations from age-adjusted alignment goals in adult spinal deformity correction and related compensatory recruitment

P. G. Passias
C. M. Jalai
B. G. Diebo
D. L. Cruz
G. W. Poorman
A. J. Buckland
L. M. Day
S. R. Horn
B. Liabaud
R. Lafage
A. Soroceanu
J. F. Baker
S. McClelland
J. H. Oren, Zucker School of Medicine at Hofstra/Northwell
T. J. Errico
F. J. Schwab
V. Lafage

Abstract

© 2019 International Society for the Advancement of Spine Surgery. Background: Full-body stereographs for adult spinal deformity (ASD) have enhanced global deformity and lower-limb compensation associations. The advent of age-adjusted goals for classic ASD parameters (sagittal vertical axis, pelvic tilt, spino-pelvic mismatch [PI-LL]) has enabled individualized evaluation of successful versus failed realignment, though these remain to be radiographically assessed postoperatively. This study analyzes pre- and postoperative sagittal alignment to quantify patient-specific correction against age-adjusted goals, and presents differences in compensation in patients whose postoperative profile deviates from targets. Methods: Single-center retrospective review of ASD patients ≥18 years with biplanar full-body stereographic xrays. Inclusion: ≥ 4 levels fused, complete baseline and early (≤6-month) follow-up imaging. Correction groups generated at postoperative visit for actual alignment compared to age-adjusted ideal values for pelvic tilt, PI-LL, and sagittal vertical axis derived from clinically relevant formulas. Patients that matched exact ± 10-year threshold for ageadjusted targets were compared to unmatched cases (undercorrected or overcorrected). Comparison of spinal alignment and compensatory mechanisms (thoracic kyphosis, hip extension, knee flexion, ankle flexion, pelvic shift) across correction groups were performed with ANOVA and paired t tests. Results: The sagittal vertical axis, pelvic tilt, and PI-LL of 122 patients improved at early postoperative visits (P <.001). Of lower-extremity parameters, knee flexion and pelvic shift improved (P <.001), but hip extension and ankle flexion were similar (P >.170); global sagittal angle decreased overall, reflecting global postoperative correction (8.38 versus 4.48, P <.001). Rates of undercorrection to age-adjusted targets for each spino-pelvic parameter were 30.3% (sagittal vertical axis), 41.0% (pelvic tilt), and 43.6% (PI-LL). Compared to matched/overcorrections, undercorrections recruited increased posterior pelvic shift to compensate (P <.001); knee flexion was recruited in undercorrections for sagittal vertical axis and pelvic tilt; thoracic hypokyphosis was observed in PI-LL undercorrections. All undercorrected groups displayed consequentially larger global sagittal angle (P <.001). Conclusions: Global alignment cohort improvements were observed, and when comparing actual to age-adjusted alignment, undercorrections recruited pelvic and lower-limb flexion to compensate.