Correlation of Scoliosis Curve Correction With the Number and Type of Fixation Anchors

Abstract

Clinical and radiologic assessment derived from a prospective multicenter data base of adolescent idiopathic scoliosis (AIS) patients. We investigated if "implant density" or the number of screws correlated with the major curve (thoracic or lumbar) correction at 2 years in patients with AIS. We also investigated the effect of implant density on the change in sagittal contour before surgery to after surgery. Controversy exists regarding number and type of spinal anchors and the number of implant sites used that result in improved correction in AIS. A prospective database of patients with AIS treated by posterior instrumentation between 1995 and 2004 was analyzed. The major curve correction expressed as % correction (from preoperative to 2 years postoperative) was correlated with the percentage of implants relative to the number of available implant sites within the measured Cobb angle. Correlation of % correction to the number of hooks, wires, and screws was also performed. We also analyzed the change in sagittal contour T2-T12, T5-T12, and T10-L2 before surgery and after surgery. This absolute change was then correlated with implant density, as was the number of hooks, wires, and screws. There were 292 patients included with all 6 Lenke curve types represented (250 with major thoracic curves and 42 with major lumbar curves). The overall % coronal Cobb correction was 64% (range: 11%-98%). The implant density within the major curve averaged 61% (range: 6%-100%). There was a significant correlation between implant density and % curve correction (r = 0.31, P < 0.001). The number of each implant type (hooks, wires, and screws) in the construct did not correlate with the % correction; however, the average % correction of the major curve was greater when the Cobb levels were instrumented only with screws (64%) compared to hooks alone (55%), P < 0.01. The greatest % correction 78% was achieved when bilateral segmental screws were used (100% screw density). The higher the implant density within the major thoracic curve, the greater the postoperative loss of kyphosis at T2-T12 (r = -0.13, P < 0.01) and T5-T12 (r = -0.16, P < 0.001). At T10-L2, increasing screw implant density correlated with decreasing kyphosis (r = -0.40, P < 0.001), whereas increasing hook implant density correlated with increasing kyphosis (r = 0.33, P < 0.001). Major curve correction at 2 years correlates most with the implant density that is correction increases with the number of implants used within the measured Cobb levels. Although the absolute number of screws used did not correlate with correction, there was an advantage in lumbar and thoracic curves to using screws compared to hooks. Sagittal contour in the thoracic spine became less kyphotic than the higher the implant density.