The Correlation Between Coxa Magna and Final Outcome in Legg-Calve-Perthes Disease

Abstract

This study evaluated the final outcome of coxa magna that developed as a sequela of Legg-Calve-Perthes disease. The final outcomes at skeletal maturity were assessed by the Stulberg classification in 85 children with unilateral Perthes disease. Among them, 21 children had a bilateral arthrogram at the active stage of the disease, and the arthrogram measurements were compared with those measured at disease healing and at skeletal maturity. Coxa magna was observed in 53% (45/85), with a mean increase in 20.0 +/- 7.2%. These coxa magna and resulting acetabular deformities occurred in the early stage of the disease. In 68 hips with mild (1-9% increase) or moderate (10-19%) coxa magna, the final results were Stulberg I or II in 57 hips and III in 11. In 17 hips with severe coxa magna (> or =20%), the results were I or II in one hip, III in nine, and IV in seven. This means that 41% of the hips with severe coxa magna might have osteoarthritis later in life.

Secondary hip dysplasia increases risk for early coxarthritis after Legg-Calve-Perthes disease. A study of 255 hips

“…Our study showed 72 hips (51% of the affected group) with radii of the contour of the articular sphere in the frontal plane larger than 2.65 cm, and this correlates with the previous report of "coxa magna" being present in a group of 85 hips after LCPD in 53%, when measuring the widest diameter of the femoral head on AP radiograms at skeletal maturity and considering the 10% difference between affected and unaffected sides as a threshold value (Rowe et al 2005).…”

“…Furthermore, within the LCPD group, the Stulberg Class III/IV/V subgroup had higher prevalences of both THA and hip osteoarthritis compared to the Stulberg Class I/II subgroup (P ¼ 0.04 and P ¼ 0.03, respectively) (Froberg et al 2011). Rowe et al (2005) showed that larger heads resulted in skeletal maturity with worse Stulberg class and larger possibility of having osteoarthritis later in life. These results agree with our findings that larger articular surfaces express more unfavourable stress distribution when compared to normal sized ones (Table 2), and with previous results (Kocjancic et al 2014) showing statistically significant correlation between stress gradient index and radiological Stulberg score.…”

“…It should be born in mind that the articular surface is an abstract feature that is subject to the weight bearing and, therefore, outlines a particular region of the femoral head and the acetabulum. However, the deformation of the articular surface corresponds to the deformed appearance of the femoral head, considered by some authors who previously analysed development of hips after LCPD (Rowe et al 2005;Shah et al 2013).…”

“…Patients with most severe residual deformity (class V -"aspherical incongruency" representing an aspherical head with usually spherical acetabulum) developed severe osteoarthritis before 50 years of age, patients with moderate residual deformity (class III/IV -"aspherical congruency") developed mild to moderate osteoarthritis, and patients with less severe or no deformity (class I/II "spherical congruency") had no osteoarthritis (Stulberg et al 1981). A long-term (37-58 years) outcome analysis of 135 non-operatively treated patients (156 hips) after LCPD compared to 312 control hips, showed history of THA in 13% of patients versus none in the controls (P ¼ 0.001) and hip osteoarthritis in 7% of patients versus 1% of the controls (P ¼ 0.04) (Rowe et al 2005). Furthermore, within the LCPD group, the Stulberg Class III/IV/V subgroup had higher prevalences of both THA and hip osteoarthritis compared to the Stulberg Class I/II subgroup (P ¼ 0.04 and P ¼ 0.03, respectively) (Froberg et al 2011).…”

The deformity index as a predictor of final radiological outcome in Perthes’ disease

“…[1][2][3][4] The grading system of choice for determining the long-term clinical outcome is that of Stulberg et al,4 which is recorded at skeletal maturity. While the lateral pillar classification is strongly predictive of outcome 5,6 it does not necessarily predict the outcome after surgery since it is usually assigned when treatment begins.…”