Abstract
Flattening of the preimplantation rod contour in the sagittal plane influences thoracic kyphosis (TK) restoration in adolescent idiopathic scoliosis (AIS) surgery. The effects of multilevel facetectomy and screw density on postoperative changes in spinal rod contour have not been documented. This study aimed to evaluate the effects of multilevel facetectomy and screw density on changes in spinal rod contour from before implantation to after surgical correction of thoracic curves in patients with AIS prospectively. The concave and convex rod shapes from patients with thoracic AIS (n = 49) were traced prior to insertion. Postoperative sagittal rod shape was determined by computed tomography. The angle of intersection of the tangents to the rod end points was measured. Multiple stepwise linear regression analysis was used to identify variables independently predictive of change in rod contour (Δθ). Average Δθ at the concave and convex side were 13.6° ± 7.5° and 4.3° ± 4.8°, respectively. The Δθ at the concave side was significantly greater than that of the convex side (P < 0.0001) and significantly correlated with Risser sign (P = 0.032), the preoperative main thoracic Cobb angle (P = 0.031), the preoperative TK angle (P = 0.012), and the number of facetectomy levels (P = 0.007). Furthermore, a Δθ at the concave side ≥14° significantly correlated with the postoperative TK angle (P = 0.003), the number of facetectomy levels (P = 0.021), and screw density at the concave side (P = 0.008). Rod deformation at the concave side suggests that corrective forces acting on that side are greater than on the convex side. Multilevel facetectomy and/or screw density at the concave side have positive effects on reducing the rod deformation that can lead to a loss of TK angle postoperatively.