Factors related to postoperative coronal imbalance in adult lumbar scoliosis.

Abstract

The object of this study was to analyze the prevalence of postoperative coronal imbalance (CIB) and related factors in patients with adult lumbar scoliosis. This was a retrospective single-center study of data from patients with adult spinal deformity (ASD) who had undergone corrective surgery performed by a single surgeon between 2009 and 2017. The inclusion criteria were as follows: 1) age at surgery > 40 years, 2) Cobb angles of the thoracolumbar/lumbar (TL/L) curve > 40°, 3) upper instrumented vertebra of T9 or T10, 4) lowest instrumented vertebra of L5 or the pelvis, and 5) minimum 2-year follow-up period. Radiographic parameters were measured before surgery, 2 weeks after surgery, and at the latest follow-up. Curve flexibility was also assessed using side bending radiographs. Clinical outcomes were evaluated using the 22-Item Scoliosis Research Society Outcomes Questionnaire (SRS-22) and the SF-36. CIB was considered to have occurred if the C7 plumbline was more than 2.5 cm lateral to the central sacral vertical line (i.e., coronal vertical axis [CVA] > 2.5 cm) at the final follow-up. Parameters between the patients with (CIB group) and without (coronal balance [CB] group) CIB were compared, and factors related to CIB were evaluated. From among 66 consecutively treated ASD patients, a total of 37 patients (mean age at surgery 66.3 years, average follow-up 63 months) met the study inclusion criteria. CIB was found in 6 patients at the final follow-up (16.2%), and the CVA of all patients in the CIB group shifted to the convex side of the TL/L curve. A comparative analysis between the CB and CIB groups, respectively, at the final follow-up indicated the following factors were related to CIB: lumbosacral (LS) curve, 11.0°/16.5° (p = 0.02); LS correction rate (CR), 61%/47% (p = 0.02); and CR ratio (LS vs TL/L), 0.93/0.67 (p = 0.0002). Regarding clinical outcomes, the satisfaction domain of the SRS-22 (CB 4.4 vs CIB 3.5) showed a significant difference between the CIB and CB groups (p = 0.02), and patients in the CB group tended to score better on the pain domain (CB 4.3 vs CIB 3.7), but the difference was not significant (p = 0.06). Postoperative CIB negatively impacted patients' HRQOL. An imbalanced correction ratio between the TL/L and LS curves may cause postoperative CIB. Therefore, adequate correction of the LS curve may prevent postoperative CIB.