Evaluating Stability in Borderline Hip Dysplasia using The Femoro-Epiphyseal Acetabular Roof Index

Classifying hips with mild structural deformity along the spectrum of femoroacteabular impingement (FAI) and developmental dysplasia of the hip (DDH) is currently confusing and controversial. Current radiographic measures may inaccurately predict stability in borderline or transitional hips. Identification of a radiographic marker associated with hip stability would help guide proper diagnosis and management. A recently proposed radiographic measure, the Femoro-Epiphyseal Acetabular Roof (FEAR) Index, is derived from Pauwels’ and Maquet’s theory that the upper femoral epiphyseal plate will become oriented perpendicular to the joint reactive forces in accordance with the Heuter-Volkman principle. We sought to answer the following questions: 1. Is confirmed childhood developmental dysplasia associated with a positive FEAR index? and 2. Does the FEAR index correspond with clinical diagnosis and surgical treatment in cohorts of symptomatic borderline and non-borderline hips diagnosed and treated for both dysplasia and FAI?

Following IRB approval, 375 hips in 209 patients clinically diagnosed with DDH (n=262), FAI (n=106), or mixed morphology (n=6) from 2007 to 2017 were retrospectively collected. Hips clinically diagnosed with DDH were considered unstable, while hips clinically diagnosed with FAI were considered stable. Standardized standing AP pelvis radiographs were analyzed and compared to a cohort of asymptomatic controls (n=125 hips, 3:1 ratio) collected from our institution’s trauma database. Treatments included conservative management, periacetabular osteotomy for unstable hips, or open/arthroscopic surgery for femoroacteabular impingement. FEAR index, lateral center edge angle (LCEA), anterior center edge angle (ACEA), Tonnis angle, and Shenton’s line were measured and compared to clinical diagnosis and treatment. Patients were then grouped by age (childhood 15 years old). The FEAR index was then compared among the three age subgroups for DDH, FAI, and control hips using a two-way ANOVA, thus determining if the FEAR index’s correlation with diagnosis persisted across age ranges. The FEAR index was measured in all symptomatic hips and control hips and compared amongst clinical diagnoses and treatment received using a one-way ANOVA. This analysis was then repeated in a subgroup of borderline hips (defined by an LCEA of 20-25 degrees) drawn from DDH, FAI, and control hips. Receiver operator curves (ROC) were used to determine the diagnostic ability of the FEAR index for diagnosis of DDH versus FAI and to determine a cutoff value to optimize an accurate diagnosis.

The FEAR index was found to vary based on age (with children < 10 having a higher FEAR index, P < 0.001). The FEAR index was also found to vary based on diagnosis (P < 0.001). The relationship between the FEAR index and diagnosis remained consistent within each age group (P=0.149).

The FEAR index was found to be significantly higher in clinically unstable DDH hips (n=262; mean, 3.62 ± 9.8; P < 0.001) and significantly lower in both asymptomatic controls (n=125; mean, -13.08 ± 7.8; P < 0.001) and hips diagnosed with FAI (n=106; mean, -10.99 ± 9.9; P < 0.001). Of those that were clinically diagnosed as an unstable DDH hip, 183 had a positive FEAR index (70%) and 79 had a negative FEAR index (30%). Clinically unstable DDH hips with negative FEAR index were more likely to be managed nonoperative (75%) than operative (25%) (P < 0.001). A lower FEAR index was also found to correlate with open/arthroscopic FAI procedures in the FAI hips (P=0.043), while the LCEA did not vary significantly between operative and nonoperative FAI hips (P=0.186). Using a receiver operator characteristics curve (ROC), the optimal FEAR index cutoff to predict a clinically unstable hip was greater than 0°, (Sensitivity 92.0%, Specificity 93.5%), and less than -4 degrees to predict a clinically stable hip (Sensitivity 81.7%, Specificity 83.2%).

In the radiographically defined borderline dysplastic group (defined by an LCEA of 20-25 degrees; n=169), the FEAR index was significantly higher in the clinically diagnosed unstable DDH hips (n=92; mean, -0.9 ± 5.5; P < 0.001) and significantly lower in both the asymptomatic controls (n=33; mean, -10.2 ± 6.1; P < 0.001) and in the clinically diagnosed stable FAI borderline hips (n=44; mean, -10.0 ± 7.5; P < 0.001). Additionally, borderline hips with a positive FEAR index were more likely to be clinically diagnosed as an unstable hip compared to asymptomatic controls or stable FAI hips (P < 0.001). ROC calculated from the borderline group demonstrated an identical optimal FEAR index cutoff to predict a clinically unstable hip was greater than 0°, (Sensitivity 96.0%, Specificity 97.4%), and less than -4 degrees to predict a clinically stable hip (Sensitivity 75.6%, Specificity 80.0%).

The FEAR index was found to have a developmental basis as it correlates with diagnosis of DDH and FAI within each age category (15 years). The FEAR index is as a valid radiographic measure in the diagnosis of symptomatic hips as unstable (DDH) or stable (FAI) and may be used to guide surgical treatment in borderline hips. The FEAR index, not the LCEA, demonstrated significance in appropriate management of borderline dysplastic and FAI hips. Lastly, a FEAR index greater than 0 degrees correlated strongly with an unstable hip indicative of a higher probability of hip dysplasia.

The author(s) has no commercial interests to disclose.