Is There an Association Between Borderline-to-mild Dysplasia and Hip Osteoarthritis? Analysis of CT Osteoabsorptiometry.

BACKGROUND

The definitive treatment of borderline-to-mild dysplasia remains controversial. A more comprehensive understanding of the etiology of osteoarthritis (OA) and clarification of any possible association between borderline-to-mild dysplasia and the pathogenesis of OA are essential.

QUESTIONS/PURPOSES: (1) Does the distribution of acetabular subchondral bone density increase according to dysplasia severity? (2) Is there an association between borderline-to-mild dysplasia and OA pathogenesis?

METHODS

We evaluated bilateral hips of patients with developmental dysplasia of the hip who underwent eccentric rotational acetabular osteotomy (ERAO) for inclusion in the dysplasia group and contralateral hips of patients with unilateral idiopathic osteonecrosis of the femoral head (ONFH) who underwent curved intertrochanteric varus osteotomy (CVO) for the control group. ERAO was performed in 46 patients and CVO was performed in 32 patients between January 2013 and August 2016 at our institution. All patients underwent bilateral hip CT. The study included 55 hips categorized according to dysplasia severity: (1) borderline-mild, 19 hips (15° ≤ lateral center-edge angle [LCEA] < 25°); (2) moderate, 20 hips (5° ≤ LCEA < 15°); (3) severe, 16 hips (LCEA < 5°); and (4) control, 15 hips. Thirty-seven dysplastic hips (age < 15 or > 50 years old, prior hip surgery, subluxation, aspherical femoral head, cam deformity, and radiographic OA) and 17 control hips (age < 15 or > 50 years old, bilateral ONFH, LCEA < 25° or ≥ 35°, cam deformity, and radiographic OA) were excluded. CT-osteoabsorptiometry (OAM) predicts physiologic biomechanical conditions in joints by evaluating subchondral bone density. We evaluated the distribution of subchondral bone densities in the acetabulum with CT-OAM, dividing the stress distribution map into six segments: anteromedial, anterolateral, centromedial, centrolateral, posteromedial, and posterolateral. We calculated the percentage of high-density area, which was defined as the upper 30% of Hounsfield units values in each region and compared least square means difference estimated by the random intercept model among the four groups.

RESULTS

In all regions, the percentage of high-density area did not differ between the borderline-mild group and the control (eg, anterolateral, 16.2 ± 5.6 [95% CI, 13.4 to 18.9] versus 15.5 ± 5.7 [95% CI, 12.4 to 18.5, p = 0.984]; centrolateral, 39.1 ± 5.7 [95% CI, 36.4 to 41.8] versus 39.5 ± 4.7 [95% CI, 36.6 to 42.5, p = 0.995]; posterolateral, 10.9 ± 5.2 [95% CI, 8.0 to 13.8] versus 15.1 ± 6.8 [95% CI, 11.7 to 18.5, p = 0.389]). In the anterolateral region, a smaller percentage of high-density area was observed in the borderline-mild group than in both the moderate group (16.2 ± 5.6 [95% CI, 13.4-18.9] versus 28.2 ± 5.1 [95% CI, 25.5-30.9], p < 0.001) and the severe group (16.2 ± 5.6 [95% CI, 13.4-18.9] versus 22.2 ± 6.8 [95% CI, 19.2-25.2, p = 0.026).

CONCLUSIONS

Our results suggest that the cumulative hip stress distribution in borderline-to-mild dysplasia was not concentrated on the lateral side of the acetabulum, unlike severe dysplasia.

CLINICAL RELEVANCE

Based on the stress distribution pattern, our results may suggest that there is no association between borderline-to-mild dysplasia and the pathogenesis of OA. Further studies are needed to evaluate the association between borderline-to-mild dysplasia and instability of the hip.