Limited fusion strategy for congenital scoliosis: is it truly one and done?

Abstract

To investigate the rate and risk factors for reoperation following a short-segment fusion strategy for congenital scoliosis (CS). CS patients treated with posterior fusion were identified and stratified by history of revision surgery. Demographic and pre- and postoperative radiographic data were compared to determine the risk factors for reoperation. Thirty-five patients (mean age 5.0 years; mean follow-up 7.6 years) underwent an average of 3.5 levels fused. Index procedures were performed in the thoracic (49%), thoracolumbar (29%), lumbar (17%), and lumbosacral (6%) regions. Eighteen (51%) patients underwent reoperation at a mean of 5.2 years. Preoperative curve magnitude was significantly lower in patients who did not require reoperation (34.2° vs 66.7°, p < 0.001), with preoperative curves ≥ 45° having a 94% reoperation rate compared to 6% in curves < 45°. Preoperative curve magnitude strongly predicted reoperation (AUC 0.99, 95% CI 0.97-1.00), with 45° as the optimal cutoff (94.4% sensitivity, 94.1% specificity). Larger pre- and postoperative compensatory curves and constructs that did not span both end vertebrae were also risk factors for reoperation (p < 0.05). In multivariate analysis, only greater preoperative curve magnitude retained significance for reoperation (p = 0.028). Although short-segment fusion for CS resulted in a 51% reoperation rate, stabilization for 6.2 years with this staged strategy allowed for growth and delayed additional surgery to a later, optimal age. Risk factors for reoperation included preoperative curves ≥ 45° and constructs failing to span both end vertebrae. III.