The selection of fusion levels using torsional correction techniques in the surgical treatment of idiopathic scoliosis

Study Design. This is a retrospective, consecutive case series, with the index patient included. Objectives. To evaluate the evolution and effectiveness of instrumentation techniques designed to untwist the scoliosis deformity. Summary of Background Data. Three-dimensional studies of the idiopathic scoliosis deformity are consistent with the theory that the deformity or deformities evolve as an imperfect torsion or torsions. Methods. From 1989 through 1995, 102 consecutive patients (84 females, 18 males) underwent surgery with increasing emphasis on torsional correction. One hundred patients (98%), with an average age of 14.3 years (range, 10.5-20.8 years), were observed for an average of 40 months (range, 24-81 months). The upper instrumented vertebra evolved to be the centered vertebra. The lower instrumented vertebra was chosen based on its ability to become horizontal on contralateral bend radiographs and was termed the caudal foundation vertebra. Because these techniques evolved over the first 3 years of the study period, a split analysis was performed to evaluate improvements in correction and correction maintenance over the course of the study. Results. The average Cobb angle was 59 degrees before surgery, 18 degrees after surgery (69% correction), and 22 degrees (63% correction) at latest follow-up. A comparison of the first half of the series with the second half showed no significant demographic differences. Curve correction was significantly improved for King-Moe IIB (thoracolumbar-lumbar curve only), King-Moe III, and King-Moe V curve types in the second half of the series. In the last 4 years, curve correction at latest follow-up for King-Moe IIB curves was 61% for the thoracic curve and 65% for the thoracolumbar-lumbar curve. King-Moe III curves had a 68% correction, and King-Moe V curves had a 50% high thoracic and a 72% thoracic curve correction. Thoracolumbar, lumbar, and King-Moe I curves averaged 81% correction of the thoracolumbar-lumbar curve. The angle of thoracic curve inclination improvement at 1 year was maintained at latest follow-up. Conclusions. This method of selecting instrumentation levels while using torsional correction techniques is safe and reliable. The results were improved with the evolution of these techniques and appear to provide improved correction and correction maintenance compared with that of historical controls.