STATISTICAL SHAPE MODELLING TO COMPARE SPINOPELVIC STRUCTURES WITH HIP IMPINGEMENT AND DYSPLASIA

Femoroacetabular impingement (FAI) and developmental dysplasia of the hip (DDH) are associated with adverse hip loading leading to early osteoarthritis. FAI can be subclassified into cam-type (aspherical head-neck deformity) and pincer-type (acetabular over-coverage or retroversion) while DDH is characterized by a shallow acetabular undercoverage. Spinopelvic parameters and femoral neck angles can affect hip mobility and symptoms, however, additional features related to FAI and DDH have never been comprehensively studied. The purpose was to examine how anatomical variations could best predict differences between range of motion in cam FAI, retroversion FAI, and DDH populations. Fifty-two cadaveric hips (n = 52; m:f = 28:24, age = 44 ± 11 yrs, BMI = 25 ± 6 kg/m2; approved ethics #MEC-AA-13-032) were obtained from a tissue bank and computed tomography (CT) scanned (Somatom Perspective; SIEMENS). Hip specimens were sub-grouped if they indicated a cam morphology (CAM; 3:00 alpha angle > 50.5°, 1:30 alpha > 60°), acetabular retroversion (RETRO; crossover sign, posterior wall sign, retroversion index), dysplasia (DDH; lateral centre-edge angle 10°, femoro-epiphyseal acetabular roof index > 2°), or healthy control (CON; no morphology). Each specimen's 3D models were reconstructed from their CT data using segmentation software (3D Slicer, The Slicer Community); and imported into statistical shape modelling software (ShapeWorks; University of Utah). Resultant mean shape models of the femur, pelvis, and sacrum were generated using principal component analysis and each pathological group was compared directly to the control group to highlight specific anatomical differences. For range of motion, each cadaveric specimen mounted onto a 6-DOF robot (TX90, Stäubli) that captured internal-external rotations in five sagittal positions: full extension, neutral 0o, flexion 30o, flexion 60o, and flexion 90o. One-way ANOVA examined range of motion differences between groups (CI = 95%); and were compared to the statistical shape models to associated hip morphology with mobility. From the statistical shape models, CAM showed greater acetabular coverage and RETRO showed greater femoral neck width compared to CON, respectively (Figure 1). Interestingly, the CAM sacrum showed an increased inclination of the sacral promontory compared to CON (Figure 2). Furthermore, the CON sacral alar width was significantly greater compared to all other pathological groups. From the range of motion, the RETRO and DDH groups showed an increased range of motion in full extension and flexion positions compared to CON, respectively. No difference was observed between the FAI groups (CAM and RETRO) in any of the flexion positions. The most important finding was that the CAM group indicated a prominently inclined sacral promontory associated with non-restricted hip mobility. An increased sacral promontory may affect pelvic incidence and sagittal moment arms that would typically serve as a protective factor in the development of symptoms and complications. Furthermore, a shorter sacral alar width in the pathological groups (CAM, RETRO, DDH) may play a factor in decreased trunk stability and development of symptoms and complications. The characteristic femoral and acetabular morphologies alone may not be the primary indicators of early clinical signs, restrictive symptomatic hip impingement, or dysplastic instability. For any figures or tables, please contact the authors directly.