Neuromuscular acetabular dysplasia is believed to be posterior in children with cerebral palsy (CP). While an anteroposterior (AP) pelvic radiograph is the current standard-of-care for CP hip surveillance, the 3-dimensional (3D) nature of the acetabulum makes accurate assessment difficult using 2-dimensional imaging (2D). Over a 12-year period at a single tertiary care institution, a consecutive sample of 285 children with CP were retrospectively identified across Gross Motor Function Classification System Level (GMFCS) levels (age: 3 to 22 y, average: 9.3±3.5 years, 43% females). A cohort of age-matched and sex-matched children (n=285) without CP were used for comparison. Using a validated automatic segmentation and anatomy measurement program (Virtual Hip, Musculoskeletal Digital Innovation and Informatics Program, Boston's Children Hospital, Boston, MA), 3D hip models from 3-dimensional computed tomography (3D-CT) scans were developed to automatically measure acetabular index (AI) or acetabular angle (AA), and acetabular rim lateral extension (ARLE) from posterior (9 o'clock) to anterior (3 o'clock). AI/AA was also measured on corresponding radiographs. Mixed linear models were used to compare measurements around the clockface and between 3D-CT and radiographs (for AI and AA only). Linear regression investigated age-related changes. The acetabulum is significantly underdeveloped in both the posterior and anterior regions relative to the superior region in skeletally mature and immature patients with CP, as demonstrated by increasing AI (P<0.05) and AA (P<0.01)-and decreasing ARLE (P<0.01)-at nearly all positions of the acetabulum clockface relative to 12 o'clock. Greater acetabular dysplasia correlated with increasing GMFCS level compared with our control cohort (P<0.01). Finally, singular 2D radiographic measurements of AI and AA were only accurate for the most superior 12 o'clock position but significantly underestimated acetabular deficiency everywhere else (P<0.05). Severity of neuromuscular acetabular dysplasia is underestimated using plain radiographs. Using 3D-CT, acetabular deficiency is appreciated globally and worsens with increasing GMFCS level. Measuring the 3D ARLE for preoperative planning in neuromuscular acetabular dysplasia can provide more information than plain radiographic measurements. Level III-case control or retrospective comparative study.
Nguyen et al. (Wed,) studied this question.