The Lumbosacral Takeoff Angle Can Be Used to Predict the Postoperative Lumbar Cobb Angle Following Selective Thoracic Fusion in Patients with Adolescent Idiopathic Scoliosis

Background: Selective fusion of double curves in patients with scoliosis is considered to spare fusion levels. In 2011, we studied the lumbosacral takeoff angle, defined as the angle between the center-sacral vertical line and a line through the centra of S1, L5, and L4. The lumbosacral takeoff angle was shown to moderately correlate with the lumbar Cobb angle, and a predictive equation was developed to predict the lumbar Cobb angle after selective fusions. The purposes of the present study were to validate that equation in a separate cohort and to assess differences in outcomes following selective and nonselective fusion. Methods: Patients with Lenke 1B, 1C, 3B, or 3C curve patterns undergoing fusion (both selective and nonselective) with pedicle screw constructs and a minimum of 2 years of follow-up were included. Selective fusion was defined as a lowest level of fixation cephalad to or at the apex of the lumbar curve. To validate the previously derived equation, we used this data set and analysis of variance to check for differences between the actual and calculated postoperative lumbar Cobb angles. Pearson correlation, multiple linear regression, and t tests were used to explore relationships and differences between the selective and nonselective fusion groups. Results: The mean calculated postoperative lumbar Cobb angle (and standard deviation) (22.35 degrees +/- 3.82 degrees) was not significantly different from the actual postoperative lumbar Cobb angle (21.08 degrees +/- 7.75 degrees), with an average model error of -1.268 degrees (95% confidence interval, -2.649 degrees to 0.112 degrees). The preoperative lumbar Cobb angle was larger in patients with deformities that were chosen for nonselective fusion (50.2 degrees versus 38.9 degrees; p < 0.001). Performing selective fusion resulted in a 3.5 degrees correction of the lumbosacral takeoff angle (p < 0.001), whereas nonselective fusion resulted in a 9.3 degrees correction (p < 0.001). Conclusions: The lumbosacral takeoff angle can be used to predict the residual lumbar Cobb angle and may be used by surgeons to aid in the decision between selective and nonselective fusion. The change in the lumbosacral takeoff angle following selective fusion is small. Improvement in the lumbosacral takeoff angle and coronal balance is greater in association with nonselective fusion.