Vertebral body tethering for Lenke 1A curves: the lumbar modifier predicts less optimal outcomes.

Abstract

The addition of the L4 "AR" and "AL" lumbar modifier for Lenke 1A idiopathic scoliosis (IS) has been shown to direct treatment in posterior spinal fusion; however, its utility in vertebral body tethering (VBT) has yet to be evaluated. A review of a prospective, multicenter database for VBT in IS was performed for patients with Lenke 1A deformities and a minimum of 2 years follow-up. Patients were categorized by their lumbar modifier (AR vs AL). Less optimal VBT outcome (LOVO) was defined as a final coronal curve > 35°, lumbar adding-on, or revision surgery for deformity progression or adding-on. Ninety-nine patients met inclusion criteria (81% female, mean 12.6 years), with 55.6% being AL curves. Overall, there were 23 instances of tether breakage (23.3%) and 20 instances of LOVO (20.2%). There was a higher rate of LOVO in AR curves (31.8% vs 10.9%, P = 0.01). Patients with LOVO had greater preoperative deformity, greater apical translation, larger coronal deformity on first erect radiographs, and less coronal deformity correction. Failure to correct the deformity < 30° on first erect was associated with LOVO, as was LIV selection short of the last touch vertebra (TV). Independent risk factors for LOVO included AR curves (OR 3.4; P = 0.04) and first erect curve magnitudes > 30 degrees (OR 6.0; P = 0.002). There is a 20.2% rate of less optimal VBT following VBT for Lenke 1A curves. AR curves are independently predictive of less optimal outcomes following VBT and require close attention to LIV selection. Surgeons should consider achieving an initial coronal correction < 30 degrees and extending the LIV to at least the TV to minimize the risk of LOVO.