MR imaging as a supplement to traditional decision-making in the treatment of LCP disease.

Background: Heretofore, the general concept in treating Legg-Calve-Perthes (LCP) disease has been containment of the diseased femoral head into the acetabulum. However, surgery or bracing for containment of a deformed femoral head without accurate information on its dynamic relationship with the hip may aggravate hip congruity and lead to impingement between the femoral head and the acetabulum. We used magnetic resonance imaging on an outpatient clinic basis to evaluate the relationship between the deformed femoral head and the acetabulum in moderate-to-severe LCP disease, and applied these findings to management.

Methods: For 103 moderately and severely affected LCP patients (mean age 7.5 y), we made a total of 151 range of motion-magnetic resonance imagings (termed range of motion as each patient was scanned in 5 positions: neutral, abduction, abduction-internal rotation, abduction-internal rotation-flexion, and adduction). For each position, we calculated epiphyseal extrusion index (EEI), head coverage (HC), and medial gap ratio (MGR), and looked for differences between parameter values in neutral and the other positions. Disease severity was noted for each patient according to 3 classification systems (lateral pillar, Catterall, and Salter-Thompson), and differences in parameter values were examined for the various severity grades. The position of greatest congruity, and adjacent soft tissue changes, were also noted. Stulberg results were obtained for 54 patients who had reached skeletal maturity.

Results: For moderately affected (lateral pillar-B) patients, all 3 parameters (EEI, HC, and MGR) improved on abduction, supporting traditional containment theory. For severely affected (lateral pillar-C) patients, EEI and HC improved on abduction, but MGR did not, indicating hinge abduction by the deformed femoral head. The results do not seem to be greatly affected by 1 of the 3 classification systems which we use. In these patients, congruency was improved in adduction, and was aided by the surrounding soft tissues. Our pillar-B patients were treated conservatively and had mostly Stulberg I and II outcomes. Both conservative and operative treatment of our pillar-C patients resulted in mostly Stulberg III outcomes.

Conclusions: For moderately affected patients, we support traditional treatment aimed at containment of the diseased femoral head into the acetabulum. For severely affected patients who show improved congruency in adduction, a valgus femoral osteotomy, aimed at achieving stable congruency rather than containment, may be used as a primary treatment to minimize acetabulofemoral impingement.

Level Of Evidence: Therapeutic study, level II.