Preoperative pelvic obliquity: possible relation to postoperative coronal decompensation in thoracolumbar/lumbar adolescent idiopathic scoliosis.

OBJECTIVE Pelvic obliquity is frequently observed in patients with adolescent idiopathic scoliosis with thoracolumbar/lumbar (TL/L) curve. This study aimed to assess pelvic obliquity changes and their effects on clinical outcomes of posterior fusion surgery. METHODS Data in 80 patients (69 with type 5C and 11 with type 6C adolescent idiopathic scoliosis) who underwent posterior fusion surgery were retrospectively analyzed. Pelvic obliquity was defined as an absolute pelvic obliquity angle (POA) value of ≥ 3°. The patients were divided into groups according to preoperative pelvic obliquity. Moreover, patients with preoperative pelvic obliquity were divided based on POA change from preoperative values versus 2 years postoperatively. Patients were divided based on the presence of selective or nonselective TL/L fusion. Radiographic parameters and clinical outcomes were compared between these groups. RESULTS Among 80 patients, 41 (51%) showed preoperative pelvic obliquity, and its direction was upward to the right for all cases. Coronal decompensation 2 years postoperatively was significantly elevated in patients with preoperative pelvic obliquity (p < 0.05). Thirty-two patients (40%) displayed pelvic obliquity 2 years postoperatively. Among 41 patients with preoperative pelvic obliquity, 22 patients (54%) were in the group with a decrease in POA, and 19 were in the group with no decrease. The group with no decrease in POA showed significant TL/L curve progression throughout the postoperative follow-up period. The patients with nonselective fusion showed a significantly lower incidence of pelvic obliquity at 2 years postoperatively. CONCLUSIONS Postoperative coronal decompensation more frequently occurred in patients with preoperative pelvic obliquity than in those without pelvic obliquity preoperatively. In addition, postoperative pelvic obliquity changes may be related to residual lumbar curve behavior.