Complementary value of transverse plane descriptors in the SRS-Lenke-Aubin 3D classification for adolescent idiopathic scoliosis.

purposeTo assess the complementary value of transverse plane descriptors (orientation of the regional planes of deformation (ORPD) and local apical vertebral rotations (AVR)) integrated into the new modular three-tiered, four-modifier SRS-Lenke-Aubin 3D classification, relative to conventional 2D radiographic parameters and current Lenke 2D classification in adolescent idiopathic scoliosis (AIS).

methodsTransverse plane deformities of 285 surgically treated AIS cases reconstructed in 3D were quantified using ORPD and AVR, independently assessed for the proximal thoracic (PT), main thoracic (MT), and thoracolumbar/lumbar (TL/L) regions. Correlation analyses evaluated relationships between standard 2D parameters (Cobb angles, thoracic kyphosis (TK), lumbar lordosis (LL)) and transverse plane indices (ORPD, AVR). The distribution of ORPD and AVR subclasses was examined, as well as the associations between conventional Lenke lumbar and thoracic sagittal profile modifiers, and their corresponding 3D transverse plane modifiers. Complementary analyses also included 3D displacement of the apex relative to the end-vertebrae line (DAEVL).

resultsNearly all ORPD-AVR subclass combinations were observed across regions, confirming the system's ability to capture diverse deformity patterns. ORPD and AVR were independent in PT and MT but correlated in TL/L (r = 0.69). Cobb angle correlated moderately with ORPD in MT (r = 0.43) and strongly in TL/L (r = 0.67), while correlations with AVR were moderate in MT (r = 0.50) and TL/L (r = 0.59). TK correlated negatively with MT ORPD (r = -0.58), whereas LL showed no association with TL/L ORPD. DAEVL correlated strongly with Cobb across all regions but only weakly to moderately with ORPD. Associations between Lenke 2D modifiers and ORPD were strong in TL/L (V = 0.59) and moderate in MT (V = 0.37). Multivariate models showed that Cobb and TK explained ~ 44% of MT ORPD variance, while > 55% of ORPD variability across regions remained unexplained by 2D parameters.

conclusionsORPD and AVR provide complementary, region-specific 3D information beyond conventional 2D measures and Lenke modifiers. Their integration into the SRS-Lenke-Aubin 3D classification enhances dimensional completeness while preserving usability, laying the groundwork for future outcome-based evaluations.