How useful are indirect radiographic measurements of hip instability in borderline hip dysplasia? An MRI-based comparison to a healthy cohort

Abstract

Abstract Purpose Symptomatic hips with borderline hip dysplasia (BHD) morphology pose a challenge in differentiating stable from unstable hips. The current study aims to compare indirect radiographic signs of instability in a symptomatic BHD population to those in a healthy cohort. Methods The study group consisted of patients with a lateral centre–edge angle (LCEA) with values 18° ≤ LCEA < 25° who underwent corrective periacetabular osteotomy (PAO) and reported an improvement in patient-reported outcome measures (PROMs). The comparison group consisted of a healthy cohort of athletes who did not complain of any hip-related symptoms and who had normal values of their hip morphological parameters (LCEA, acetabular index (AI°), alpha angle (α°), femoral version, acetabular version). Indirect signs of instability consisting of the femoro-epiphyseal acetabular roof index (FEAR), iliocapsularis-to-rectus-femoris (IC/RF) ratio and labral dimensions (height-to-length ratio) were assessed in both groups. Partial Pearson correlation, logistic multiple regression analysis and Receiver-Operating Characteristic (ROC) curve analysis were performed to determine correlations, as well as the sensitivity and specificity of these signs to differentiate between healthy hips and BHD. Results On binary logistic multiple regression analysis, the FEAR Index was the only independent predictor to differentiate between BHD and healthy hips (p < 0.001). The IC/RF ratio did not achieve significance. The calculated area under the curve (AUC) was 0.93 (0.87 – 0.99, CI 95%, p < 0.001) for the FEAR Index and 0.81 (0.70 – 0.92, CI 95%, p < 0.001) for the height-length ratio. Using the predefined cut-off values (dysplastic—FEAR Index ≥ 5° or labral height-to-length ratio ≤ 0.5), 27% sensitivity/100% specificity and 20% sensitivity/ 100% specificity, were achieved. ROC analysis provided the following new thresholds: FEAR Index ≥ -5° (73% sensitivity/97% specificity); labral height-to-length ratio ≤ 0.8 (70% sensitivity, 79% specificity). Conclusion In our cohort, the FEAR index was an independent parameter that could differentiate between borderline dysplastic and asymptomatic hips. The previously published values for both the FEAR index and labral hypertrophy ratio had a poor sensitivity in differentiating symptomatic unstable BHD from healthy hips. The cut-off values of ≥ -5° (FEAR index) and ≤ 0.8 (labral height-to-length ratio) provided acceptable sensitivity and specificity when comparing to morphological healthy hips.