INTRODUCTION
Adolescent idiopathic scoliosis (AIS) is a multifactorial three-dimensional spinal deformity in youth. Although pelvic obliquity is a risk factor for postoperative coronal decompensation in corrective surgery in AIS, especially Lenke 5C cases, methods of measuring pelvic obliquity remain controversial. Previously, we reported that in adolescent idiopathic scoliosis (AIS) Lenke type 5C patients, there is sacral tilt due to asymmetry of the sacral vertebrae, resulting in a lumbosacral deformity. We suggested that the sacrum should be used as a pelvic parameter rather than the iliac bone. However, there have been no report of sacral vertebral asymmetry in other Lenke types. In this study, we investigated the sacral morphology for each AIS type and examined appropriate sacral tilt parameters using multiplanar reconstructed computer tomography (MPR-CT) images.
METHODS
A total of 128 patients with AIS who underwent corrective surgery between August 2013 and September 2023 at our institution were included. Main curve Cobb angle, C7 plumb line to central sacral vertical line (C7PL-CSVL) and leg length discrepancy (LLD) were measured on standing X-rays. S1 and S2 angles, defined as the upper and lower endplates of the S1 and S2 vertebrae, respectively, were measured in the coronal plane on MPR-CT. These parameters were compared for each Lenke type to investigate the pattern of sacral deformity.
RESULTS
The subjects included 6 males and 122 females, with an average age of 16.9 years. The Lenke classification was as follows: Type 1/2/3/4/5/6 had 68 cases/19 cases/6 cases/2 cases/28 cases/5 cases, respectively. When comparing S1 angles among Type 1/2/5, they were 2.1±1.7°, 2.3±1.9°, and 4.5±3.4°, respectively, and there was a significant difference, particularly with Type 5 (p<0.001). However, there was no significant difference in S2 angles, which were 1.3±1.1°, 1.4±1.1°, and 1.6±1.5°, respectively. When comparing the non-thoracolumbar main curve group (Type 1/2/3) and the thoracolumbar main curve group (Type 4/5/6), S1 angles were 2.1±1.7° and 4.6±3.2°, and S2 angles were 1.4±1.1° and 1.5±1.4°, respectively. There was a significant difference observed only in S1 angles (p<0.001).
DISCUSSION
Sacral tilt due to sacral asymmetry was observed in AIS Lenke 5C, which has a predominantly thoracolumbar curve. However, the deformity was more prominent at the S1 vertebra rather than the S2 vertebra. Many previous reports have defined sacral tilt as the line connecting the base of the S1 superior articular processes, but there may be a need to consider that sacral tilt measured by the S1 superior articular processes is underestimated compared to the actual sacral tilt. When discussing the impact of sacral tilt on the lower lumbar vertebrae, sacral tilt should be measured using the inclination of the upper endplate of the S1 vertebra.