Using Side-bending Radiographs to Determine the Distal Fusion Level in Patients With Single Thoracic Idiopathic Scoliosis Undergoing Posterior Correction With Pedicle Screws

Abstract

Prospective. To evaluate a strategy to determine the distal fusion level in posterior pedicle screw correction of single thoracic idiopathic scoliosis. No standard method for selecting the lowest instrumented vertebra (LIV) for the correction of thoracic adolescent idiopathic scoliosis with posterior all-pedicle screw instrumentations exists. Thirty-eight patients with single right thoracic (Lenke 1A) adolescent idiopathic scoliosis undergoing posterior pedicle screw fixation were studied. The LIV was determined using guidelines based on preoperative side-bending radiographs. In brief, (1) the whole thoracic Cobb curve should be included in the fusion mass, and the LIV should not be superior to the lower-end vertebra of the Cobb measurement. (2) On the right side-bending radiographs, the LIV should be derotated to neutral in skeletally immature (Risser 0 to 3) patients and the disc immediately below the LIV must open on the left side by at least 5 degrees. (3) On the left side-bending radiographs, the disc immediately below the LIV must be open on the right side by at least 0 degree. The first segment meeting the criteria when proceeding from the lower-end vertebra caudally is chosen as the LIV. Outcomes were based on the standing radiographs. Minimum follow-up was 2 years. The mean preoperative thoracic curve was 48.4±9.2 degrees and 12.6±6.1 degrees at final follow-up, resulting in a mean correction of 74.7%±8.5%. The mean preoperative compensatory lumbar curve of 23.7±7.5 degrees was 6.3±4.8 degrees at final follow-up. A change in lumbar lordosis from -41.2±11.9 degrees preoperatively to -38.2±9.9 degrees at final follow-up occurred. All patients achieved coronal balance and no decompensation or adding-on phenomenon was observed. Compared with the recommended fusion end by the Harrington stable zone method, 86.9% patients were saved 1 or more motion segment. The method described was effective in obtaining satisfactory curve correction, adequate trunk balance, and preservation of motion segments.