Three-dimensional reconstructions of the rib cage in adolescent idiopathic scoliosis: What determines gibbosity?

Introduction Subjects with adolescent idiopathic scoliosis (AIS) often seek treatment with the aim of correcting the gibbosity of the rib cage. Current surgical treatment for AIS mainly focuses on the correction of the spinal deformity. It is not yet known how the gibbosity is related to the spinal deformity. The aim was to quantify the relationship between the gibbosity and the 3D deformity of the scoliotic spine. Material and methods Seventy-four subjects with AIS (62F, age: 15 ± 2 years, Cobb: 32° [20–50°]) age-matched to 44 controls (24F) underwent full-body biplanar X-rays with 3D reconstructions of the spine and the rib cage. Parameters were collected for the rib cage: maximum thickness, maximum width, gibbosity, rib cage volume, volumetric spinal penetration index. Sagittal spino-pelvic parameters (PI, PT, SS, TK, LL) were assessed as well as the frontal Cobb angle, and the axial rotation of the apical vertebra (AVR). A subset of the initial sample consisting of 30 subjects with AIS age-matched to 10 control subjects were considered for the repeatability assessment and had their rib cages reconstructed 3 times by 6 trained operators. The reliability was assessed on this subset using the Intraclass Correlation Coefficients (ICC). The 95% confidence interval (CI), including both intra and inter-observer variability, was calculated for each parameter (confidence in the estimate  Results The reliability analysis of the 3D rib cage parameters showed an ICC > 0.98. The 95% CI of the gibbosity varied between 2.2° and 3.2°. The absolute value of gibbosity was significantly larger in scoliotics (6° ± 4° vs. 3° ± 2°, P P r  = 0.41) and AVR ( r  = 0.6). Gibbosity was not associated to any of the sagittal spino-pelvic parameters. However, the SMLR model revealed that AVR was a significant determinant factor of gibbosity ( R 2  = 0.36, P Discussion 3D reconstructions of the rib cage, based on biplanar X-rays, were shown to be repeatable in both control and AIS subjects. This is the first study to investigate the relationship between the deformity of the rib cage and the one from the spine in subjects with AIS. AVR was shown to be the most determinant factor of gibbosity. Other unexplored factors, probably related to the mechanical properties of the ribs and soft tissues, could also be implicated in gibbosity generation.