Low bone density at the upper instrumented vertebra, not the residual fractional curve, predicts mechanical failure in adult spinal deformity: a ten-year single-center series.

BACKGROUND: Adult spinal deformity (ASD) is a heterogeneous, three-dimensional condition. In coronal malalignment, the lumbosacral fractional curve has been recognized as an essential driver of both pathology and cosmetic deformity. While various factors have been identified in association with postoperative complications, it remains unclear whether the magnitude of the fractional curve is an independent predictor of the risk of mechanical failure. The two objectives of this study were (1) to determine whether the magnitude of the fractional curve independently predicts mechanical failure, and (2) to determine whether the magnitude of the fractional curve influences lower lumbar lordosis correction. METHODS: We conducted a single-center retrospective study of patients who underwent ASD surgery between 2016 and 2025 (mean follow-up: 21.0 ± 12.5 months). The primary outcomes were radiographic parameters, mechanical complications, and patient-reported outcome measures. Bone mineral density and preoperative comorbidity burden were also assessed as potential predictors of surgical outcomes. RESULTS: Of the 210 patients who underwent ASD correction surgery, the mean age was 66.9 ± 8.9 years, 140 (66.7%) were female, and mechanical failure was observed in 46 cases (21.9%). A multivariate logistic regression identified preoperative Hounsfield unit at the upper instrumented vertebra (UIV) as an independent predictor of decreased odds of mechanical failure (OR 0.99, 95% CI 0.98-0.99). A greater fractional curve was significantly associated with greater correction of lower lumbar lordosis (β = 0.20, p = 0.008) while it was not a significant predictor of mechanical failure. CONCLUSION: A greater preoperative fractional curve was associated with increased correction of lower lumbar lordosis and a tendency toward L4-S1 overcorrection, particularly for fractional curves >20°. The magnitude of the fractional curve was not an independent predictor of mechanical failure. Lower preoperative Hounsfield units at the upper instrumented vertebra were independently associated with increased risk of mechanical failure. These findings suggest that the fractional curve should be used to guide the lower lumbar correction strategy rather than mechanical risk stratification, while bone quality at the UIV warrants routine preoperative assessment to reduce the risk of postoperative mechanical failure.