Surgical correction of Adult Spinal Deformity in accordance to the Roussouly classification: effect on postoperative mechanical complications.

STUDY DESIGN: Single-center, retrospective. OBJECTIVES: To assess the restoration of ideal sagittal spine shape in accordance to the Roussouly classification and the effect on postoperative mechanical complications. Surgical correction of Adult Spinal Deformity is both challenging and complex. The risk of postoperative complications is considerable, especially mechanical complications requiring revision surgery. Attention has been directed toward defining alignment targets in attempts to minimize these risks, and the Roussouly classification has been proposed as a potential surgical aim. METHODS: All patients undergoing ASD surgery from 2013-2016 were included at a single, quaternary institute. Successful restoration of Roussouly spine shape was retrospectively assessed, and patients were classified as either "restored" or "non-restored". Cumulative incidence of revision surgery due to mechanical failure was estimated using the Aalen-Johansen estimator, with death as the competing risk. A multivariable proportional odds model was used to estimate the effect of the Roussouly algorithm on revision surgery due to mechanical failure. RESULTS: We identified a complete and consecutive cohort of 233 patients who were followed for a mean period of 36 (+/- 14) months. The 2-year cumulative incidence of revision surgery was 28%. Comparing the "restored" to the "non-restored" group, the overall revision rates were high in both groups. However, when adjusting for known cofounders in a multivariable proportional odds analysis, there was an almost fivefold increased odds of revision due to mechanical failure in the "non-restored" group (p = 0.036). CONCLUSION: Surgical correction of ASD in accordance to the ideal Roussouly spine shape was correlated to a marked and significant decrease in risk of revision surgery due to mechanical failure. Nonetheless, the overall revision risk was elevated in both groups. LEVEL OF EVIDENCE: Therapeutic III.