Automated radiographic hip morphology measurements: An open-access method

The different subtypes of cam morphology as defined by statistical shape modeling and their relationship with the development of hip osteoarthritis: A nationwide prospective cohort study (CHECK) with 10 years follow-up

OBJECTIVE

To determine if subtypes of cam morphology on anteroposterior radiographs exist using statistical shape modeling (SSM), and to assess their association with incident radiographic hip osteoarthritis (RHOA) within 10 years.

DESIGN

The nationwide prospective Cohort Hip and Cohort Knee (CHECK) study included 1002 participants aged 45-65 years with 10-year follow-up. Subtypes of cam morphology were defined as SSM-based shape variations of femoral head-neck junction that are associated with baseline cam morphology (alpha angle ≥60°). The association between each subtype in hips free of osteoarthritis at baseline (Kellgren & Lawrence (KL) grade <2) and incident RHOA (KL grade≥2, or a total hip replacement) was estimated using logistic regression at 10-year follow-up and stratified by sex.

RESULTS

In sex-combined group, but also for males and females separately, cam morphology subtypes were captured in modes 1, 3, 4, and 5 with odds ratios (ORs) ranging from 0.39 (0.27-0.58) to 2.25 (1.64-3.10). For sex-combined group, only mode 3, a flattened head-neck junction, was associated with incident RHOA (OR:1.14, 1.02-1.27). Males' modes 1 and 3 and females' modes 3 and 4 were associated with RHOA. Notably, the female mode 4, a slightly flattened neck but with subtle curvature, was significantly protective for RHOA (OR:0.88, 0.80-0.98).

CONCLUSIONS

We identified four distinct morphological subtypes of cam morphology defined by alpha angle. Only some subtypes were found acting as risk factors for RHOA at 10-year follow-up, which differed between males and females. This highlights the need to study cam morphology beyond the alpha angle alone.