Risk factors of postoperative coronal malalignment following long-segment spinal fusion surgery in which multilevel lateral lumbar interbody fusion was used for degenerative lumbar kyphoscoliosis.

Abstract

In patients with adult spinal deformity, especially degenerative lumbar kyphoscoliosis (DLKS), preoperative sagittal malalignment and coronal malalignment (CM) often coexist. Lateral lumbar interbody fusion (LLIF) has recently been widely chosen for DLKS treatment due to its minimal invasiveness and excellent sagittal alignment correction. However, postoperative CM may remain or occur due to an oblique takeoff phenomenon following multilevel LLIF, resulting in poor clinical outcomes. The authors investigated the risk factors for postoperative CM after long-segment fusion corrective surgery in which multilevel LLIF was used in patients with DLKS. Fifty-four consecutive patients with DLKS, main Cobb angle ≥ 20°, and lumbar lordosis ≤ 20° who underwent corrective spinal fusion surgery, including extreme lateral interbody fusion at ≥ 3 segments, were included at the authors' institute between April 2014 and October 2019. Patients who underwent suitable 3-column osteotomy, classified as grade 3-6 per the Scoliosis Research Society-Schwab criteria, were excluded. Patients were divided into CM and non-CM groups based on postoperative CM evaluated using standard standing-position radiographs obtained 2 years postoperatively. CM was defined as an absolute C7-CSVL (deviation of C7 plumb line off central sacral vertical line; calculated by defining the convex side of the CSVL as positive numerical values) value of ≥ 3.0 cm. Patient demographics and preoperative sagittal alignment parameters were evaluated. The following variables were measured to assess coronal alignment: main Cobb angle; preoperative C7-CSVL; amount of lateral listhesis; L4, L5, and sacral coronal tilt angles; coronal vertebral deformity angles; and coronal spine rigidity. Regarding risk factors for postoperative CM, patient characteristics, preoperative sagittal parameters, and coronal parameters did not significantly differ between the 2 groups, except for preoperative C7-CSVL (p = 0.016). Multivariate logistic regression analysis revealed that preoperative C7-CSVL (+1 cm; OR 1.23, 95% CI 1.05-1.50; p = 0.007) was a significant predictor of postoperative CM. Receiver operating characteristic curve analysis demonstrated that the cutoff value for preoperative C7-CSVL was +0.3 cm, the sensitivity was 85.7%, the specificity was 60.6%, and the area under the curve was 0.70. In corrective fusion surgery for DLKS in which multilevel LLIF was used, the occurrence of postoperative CM was associated with preoperative C7-CSVL. According to the C7-CSVL, which was evaluated by defining the convex side of the CSVL as positive numerical values and the concave side as negative numerical values, the CM group had a significantly higher value of preoperative C7-CSVL than did the non-CM group. Alternative corrective fusion methods, other than multiple LLIFs, may be considered in DLKS cases with a C7-CSVL of +0.3 cm or greater.