Can the Femoro-Epiphyseal Acetabular Roof (FEAR) Index Be Used to Distinguish Dysplasia from Impingement?

BACKGROUND: Classifying hips with structural deformity on the spectrum from impingement to dysplasia is often subjective and frequently inexact. Currently used radiographic measures may inaccurately predict a hip's morphological stability in borderline hips. A recently described radiographic measure, the Femoro-Epiphyseal Acetabular Roof (FEAR) index, has demonstrated an ability to predict stability in the borderline hip. This measure is attractive to clinicians because procedures can be used on the basis of a hip's pathomechanics. This study was designed to further validate and characterize the FEAR index in a skeletally immature population, in hips with dysplasia/femoroacetabular impingement (FAI), and in asymptomatic hips. QUESTIONS/PURPOSES: (1) What are the characteristics of the FEAR index in children and how does the index change with skeletal maturation? (2) How does the FEAR index correlate with clinical diagnosis and surgical treatment in a large cohort of symptomatic hips and asymptomatic controls? (3) How does the FEAR index correlate with clinical diagnosis in the borderline hip (lateral center-edge angle [LCEA] 20°-25°) group?
METHODS: A total of 220 participants with symptomatic investigational hips with a clinical diagnosis of dysplasia or FAI between January 2008 and January 2018 were retrospectively collected from the senior author's practice. Investigational hips were excluded if they had any femoral head abnormalities preventing LCEA measurement (for example, Perthes disease), Tönnis osteoarthritis grade greater than 1, prior hip surgery, or prior femoral osteotomy. In the 220 participants, 395 hips met inclusion criteria. Once exclusion criteria were applied, 15 hips were excluded due to prior hip surgery or prior femoral osteotomy, and 12 hips were excluded due to femoral head deformity. A single hip was then randomly selected from each participant, resulting in 206 investigational hips with a mean age of 13 ± 3 years. Between January 2017 and December 2017, 70 asymptomatic control participants were retrospectively collected from the senior author's institutional trauma database. Control hips were included if the AP pelvis film had the coccyx centered over the pubic symphysis and within 1 to 3 cm of the superior aspect of the symphysis. Control hips were excluded if there was any fracture to the pelvis or ipsilateral femur or the participant had prior hip/pelvis surgery. After exclusion criteria were applied, 16 hips were excluded due to fracture. One hip was then randomly selected from each participant, resulting in 65 control hips with a mean age of 16 ± 8 years. Standardized standing AP pelvis radiographs were used to measure the FEAR index, LCEA, and Tönnis angle in the investigational cohort. Standardized false-profile radiographs were used to measure the anterior center-edge angle (ACEA) in the investigational cohort. Two blinded investigators measured the FEAR index with an intraclass correlation coefficient of 0.92 [95% CI 0.84 to 0.96]. Question 1 was answered by comparing the above radiographic measures in age subgroups (childhood: younger than 10 years; adolescence: 10 to 14 years old; maturity: older than 14 years) of dysplastic, FAI, and control hips. Question 2 was answered by comparing the radiographic measures in all dysplastic, FAI, control hips, and a subgroup of operatively or nonoperatively managed dysplasia and FAI hips. Question 3 was answered by comparing the radiographic measures in borderline (LCEA 20°-25°) dysplastic, FAI, and control hips.
RESULTS: The FEAR index was lower in older dysplastic of hips (younger than 10 years, 6° ± 9°; 10 to 14 years, 4° ± 10°; older than 14 years, 5° ± 9°; p < 0.001) and control hips (younger than 10 years, -6° ± 5°; 10 to 14 years, -15° ± 4°; older than 14 years, -16° ± 7°; p < 0.001). The diagnosis and age groups were independently correlated with the FEAR index (p < 0.001). The relationship between the FEAR index and diagnosis remained consistent in each age group (p = 0.11). The FEAR index was higher in all dysplastic hips (mean 5° ± 10°) than in asymptomatic controls (mean -13° ± 7°; p < 0.001) and FAI hips (mean -10° ± 11°; p < 0.001). Using -1.3° as a cutoff for FAI/control hips and dysplastic hips, 81% (112 of 139) of hips with values below this threshold were FAI/control, and 89% (117 of 132) of hips with values above -1.3° were dysplastic. The receiver operator characteristics area under the curve (ROC-AUC) was 0.91. Similarly, the FEAR index was higher in borderline dysplastic hips than in both asymptomatic borderline controls (p < 0.001) and borderline FAI hips (p < 0.001). Eighty-nine percent (33 of 37) of hips with values below this threshold were FAI/control, and 90% (37 of 41) of hips with values above -1.3° were dysplastic. The ROC-AUC for borderline hips was 0.86.
CONCLUSION: The FEAR index was associated with the diagnosis of hip dysplasia and FAI in a patient cohort with a wide age range and with varying degrees of acetabular deformity. Specifically, a FEAR index greater than -1.3° is associated with a dysplastic hip and a FEAR index less than -1.3° is associated with a hip displaying FAI. Using this reliable, developmentally based radiographic measure may help hip preservation surgeons establish a correct diagnosis and more appropriately guide treatment. LEVEL OF EVIDENCE LEVEL: III, diagnostic study.