Acetabular Version Increases During Adolescence Secondary to Reduced Anterior Femoral Head Coverage

Background Acetabular version influences joint mechanics and the risk of impingement. Cross-sectional studies have reported an increase in acetabular version during adolescence; however, to our knowledge no longitudinal study has assessed version or how the change in version occurs. Knowing this would be important because characterizing the normal developmental process of the acetabulum would allow for easier recognition of a morphologic abnormality. Questions/purposes To determine (1) how acetabular version changes during adolescence, (2) calculate how acetabular coverage of the femoral head changed during this period, and (3) to identify whether demographic factors or hip ROM are associated with acetabular development. Methods This retrospective analysis of data from a longitudinal study included 17 volunteers (34 hips) with a mean (+/- SD) age of 11 +/- 2 years; seven were male and 10 were female. The participants underwent a clinical examination of BMI and ROM and MRIs of both hips at recruitment and at follow-up (6 +/- 2 years). MR images were assessed to determine maturation of the triradiate cartilage complex, acetabular version, and degree of the anterior, posterior, and superior acetabular sector angles (reflecting degree of femoral head coverage provided by the acetabulum anteriorly, posteriorly and superiorly respectively). An orthopaedic fellow (GG) and a senior orthopaedic resident (PJ) performed all readings in consensus; 20 scans were re-analyzed for intraobserver reliability. Thereafter, a musculoskeletal radiologist (KR) repeated measurements in 10 scans to test interobserver reliability. The intra- and interobserver interclass correlation coefficients for absolute agreement were 0.85 (95% CI 0.76 to 0.91; p < 0.001) and 0.77 (95% CI 0.70 to 0.84), respectively. All volunteers underwent a clinical examination by a senior orthopaedic resident (PJ) to assess their range of internal rotation (in 90 degrees of flexion) in the supine and prone positions using a goniometer. We tested investigated whether the change in anteversion and sector angles differed between genders and whether the changes were correlated with BMI or ROM using Pearson's coefficient. The triradiate cartilage complex was open (Grade I) at baseline and closed (Grade III) at follow-up in all hips. Results The acetabular anteversion increased, moving caudally further away from the roof at both timepoints. The mean (range) anteversion angle increased from 7 degrees +/- 4 degrees (0 to 18) at baseline to 12 degrees +/- 4 degrees (5 to 22) at the follow-up examination (p < 0.001). The mean (range) anterior sector angle decreased from 72 degrees +/- 8 degrees (57 to 87) at baseline to 65 degrees +/- 8 degrees (50 to 81) at the final follow-up (p = 0.002). The mean (range) posterior (98 degrees +/- 5 degrees [86 to 111] versus 97 degrees +/- 5 degrees [89 to 109]; p = 0.8) and superior (121 degrees +/- 4 degrees [114 to 129] to 124 degrees +/- 5 degrees [111 to 134]; p = 0.07) sector angles remained unchanged. The change in the anterior sector angle correlated with the change in version (rho = 0.5; p = 0.02). The change in version was not associated with any of the tested patient factors (BMI, ROM). Conclusions With skeletal maturity, acetabular version increases, especially rostrally. This increase is associated with, and is likely a result of, a reduced anterior acetabular sector angle (that is, less coverage anteriorly, while the degree of coverage posteriorly remained the same). Thus, in patients were the normal developmental process is disturbed, a rim-trim might be an appropriate surgical solution, since the degree of posterior coverage is sufficient and no reorientation osteotomy would be necessary. However, further study on patients with retroversion (of various degrees) is necessary to characterize these observations further. The changes in version were not associated with any of the tested patient factors; however, further study with greater power is needed.