Selection of proximal fusion level for degenerative scoliosis and the entailing proximal-related late complications.

OBJECTIVE To determine the optimal selection of proximal fusion level for degenerative scoliosis (DS) and investigate the long-term proximal-related complications. METHODS Profiles of 95 consecutive patients with DS who underwent posterior long instrumented fusion were analyzed retrospectively. Perioperative parameters were reviewed stratified into 3 groups according to the relationship between the upper instrumented vertebrae (UIV), horizontal vertebrae (HV) and upper end vertebrae (UEV), namely HV Group (UIV = HV or above), HV-UEV Group (UIV = between HV and UEV) and UEV Group (UIV = UEV or below) in coronal plane and 3 groups in sagittal plane according to segment levels. Clinical and radiographic parameters were studied statistically. RESULTS Average follow-up was 7.8 years (range, 5-13 years). HV Group showed a significant increase in operative time, intraoperative blood loss, postoperative in-bed time, and hospital stays. UEV Group showed greater average Cobb angle, AVT and DW. UEV Group showed significant increase in operative time, blood loss, postoperative in-bed time, and inpatient stay. All three groups showed significant improvement in ODI compared to baseline, while there was no significant difference in LL between three groups. Proximal-related late complications included recurrent junctional scoliosis in 4 cases and junctional kyphosis in 4 cases. CONCLUSIONS Recurrent junctional scoliosis developed more commonly when the fusion was at or below the UEV, and fusion at L1 or L2 showed the highest incidence of junctional kyphosis. Long instrumented fusion to T11 or T12 appeared to be a reasonable alternative when the UIV was above UEV in DS.