Characteristics and standards of severe sagittal imbalance in adult patients with spinal deformities: a retrospective analysis.

ABSTRACT

OBJECTIVE: To analyze the characteristics of "severe" dynamic sagittal imbalance (DSI) in patients with adult spinal deformity (ASD) and establish criteria for them. METHODS: We retrospectively analyzed 102 patients with ASD presenting four cardinal signs of lumbar degenerative kyphosis. All patients underwent deformity corrective surgery and were divided into three groups according to the diagnostic criteria based on the Oswestry disability index and dynamic features (△Timewalk time until C7 sagittal vertical axis [C7SVA] reaches ≥ 20 cm after the start of walking) of sagittal imbalance. The paravertebral back muscles were analyzed and compared using T2-weighted axial imaging. We performed a statistically time-dependent spinopelvic sagittal parameter analysis of full standing lateral lumbar radiographs. Lumbar flexibility was analyzed using dynamic lateral lumbar radiography. RESULTS: The patients were classified into the mild (△Timewalk ≥ 180 s, 35 patients), moderate (180 s > △Timewalk ≥ 30 s, 38 patients), and severe (△Timewalk < 30 s, 29 patients) groups. The back muscles in the severe group exhibited a significantly higher signal intensity (533.4 ± 237.5, p < 0.05) and larger area of fat infiltration (35.2 ± 5.4, p < 0.05) than those in the mild (223.8 ± 67.6/22.9 ± 11.9) and moderate groups (294.4 ± 214.7/21.6 ± 10.6). The analysis of lumbar flexibility revealed significantly lower values in the severe group (5.8° ± 2.5°, p < 0.05) than in the mild and moderate groups (14.2° ± 12.4° and 11.4° ± 8.7°, respectively). The severe group had significantly lower lumbar lordosis (LL, 25.1° ± 22.7°, p < 0.05) and Pelvic incidence-LL mismatch (PI-LL, 81.5° ± 26.6°, p < 0.001) than those of the mild (8.2° ± 16.3°/58.7° ± 18.8°) and moderate (14.3° ± 28.6°/66.8° ± 13.4°) groups. On receiver operating characteristic curve analysis, PI-LL was statistically significant, with an area under the curve of 0.810 (95% confidence interval) when the baseline was set at 75.3°. The severe group had more postoperative complications than the other groups. CONCLUSIONS: Our results suggest the following criteria for severe DSI C7SVA > 20 cm within 30 s of walking or standing, a rigid lumbar curve < 10° on dynamic lateral radiographs, and a PI-LL mismatch > 75.3°.