Incidence of osteonecrosis associated with corticosteroid therapy among different underlying diseases: prospective MRI study.

The purpose of this study was to clarify the incidence of (CS)-associated osteonecrosis among different underlying diseases and to evaluate the risk factors for steroid-associated osteonecrosis in a prospective MRI study. We prospectively used MRI to study 337 eligible underlying disease patients requiring CS therapy and succeeded in examining 1199 joints (hips and knees) in 302 patients with MRI for at least 1 year starting immediately after the onset of CS therapy (1-year follow-up rate of 90%). The underlying diseases included SLE in 687 joints (173 patients) and a variety of other rheumatological disorders in 512 joints (129 patients). The incidence of osteonecrosis was significantly higher in SLE patients than in non-SLE patients (37 vs 21%, P = 0.001). Logistic regression analysis revealed that adolescent and adult patients had a significantly higher risk of osteonecrosis compared with paediatric patients [odds ratio (OR) = 13.2], that high daily CS dosage (>40 mg/day) entailed a significantly higher risk of osteonecrosis compared with the dosage of <40 mg/day (OR = 4.2), that SLE patients had a significantly higher risk of osteonecrosis compared with non-SLE patients (OR = 2.6) and that male patients had a significantly higher risk of osteonecrosis compared with female patients (OR = 1.6). These findings suggest that the incidence of CS-associated osteonecrosis varies among different underlying diseases.