Selective versus nonselective fusion for idiopathic scoliosis: does lumbosacral takeoff angle change?

Abstract

Retrospective review of a prospective, multicentered database. To determine the relationship between preoperative lumbosacral takeoff angle (LSTOA) and postoperative thoracolumbar/lumbar Cobb angle (TL/L Cobb angle) in patients undergoing selective thoracic fusions Selective fusion of the thoracic curve can improve the lumbar curve inpatients with idiopathic thoracic scoliosis and a compensatory lumbar curve. Predicting improvement is controversial and determining whether to perform a selective fusion or nonselective fusion can be difficult. Patients had undergone either nonselective or selective spinal fusion for adolescent or juvenile idiopathic scoliosis (Lenke 1B/3B/1C/3C). Outcome measures were: coronal and sagittal thoracic Cobb angle, TL/L Cobb angles, lumbar apical vertebral translation, LSTOA and coronal decompensation. Analyses compared relationships between preoperative and postoperative radiographic measures. Positive, significant correlations were found between preoperative LSTOA and preoperative TL/L Cobb angle in the nonselective (r=0.7; P<0.001) and selective (r=0.5; P<0.001) fusion groups. Mean two-year postoperative coronal TL/L Cobb angles were significantly improved in nonselective and selective fusion groups (32° and 20°, respectively, P<0.001). In the nonselective fusion group, LSTOA significantly decreased by 11° (P<0.001), and in the selective group, the LSTOA had a modest but significant decrease of 2° (P<0.001). The nonselective fusion also resulted in more lordosis between T10 and L2 (7.5° of lordosis) than the selective approach (2.7° kyphosis, P<0.001). For both groups, upper thoracic kyphosis increased after surgery (P<0.001, P<0.001). For nonselective fusions, regression modeling predicted TL/L Cobb angle at two-year follow-up based on preoperative TL/L Cobb angle and preoperative LSTOA (r=0.4, P<0.001). Collectively, these data demonstrate the preoperative TL/L Cobb angle and LSTOA can be useful predictors of postoperative TL/L Cobb angle after a selective instrumented fusion. Analyses of distal fixation levels demonstrated that to appreciably change the LSTOA using a posterior instrumented fusion, the distal level of fixation must be beyond the lumbar apex.