Background/Purpose: Systemic lupus erythematosus patients, particularly those who received corticosteroids are at high risk of avascular necrosis (AVN). A past meta-analysis identified other risk factors including African-American ethnicity, active disease, Raynaud’s, arthritis, cyclophosphamide use, and IgM anticardiolipin antibodies. In this analysis, we determined the predictors of AVN in a longitudinal lupus cohort in which Caucasian and African-American ethnicities were well represented.

Methods: We analyzed data from 1986 until April 2019. AVN was recorded using the Systemic Lupus Erythematosus International Collaborating Clinics/ American College of Rheumatology (SLICC/ACR) Damage Index. Cox regression was used to identify variables that were associated with time to AVN. First, we assessed the univariate relationship between each variable and time to AVN. Those variables with significant association with time to AVN were then entered into the multivariable model; the ones that remained significant were retained in the final model. Variables that were highly colinear with highest prednisone were not included in the same model. Hazard ratios (HR) and 95% confidence intervals were reported.

Results: This analysis included 2477 patients who were 92.2% female, 39.1% African American, and 52.8% Caucasian. Non- Caucasian patients, male gender, obesity and history of vasculitis conferred a higher risk of AVN. Patients who received prednisone doses more than 60 mg had 10 times the risk of having an AVN, regardless of the duration of treatment, compared to patient who did not receive any prednisone. Patient who were on prednisone 40 mg daily had 6 times the risk of having AVN if the duration of treatment was more than 1 month compared to those who did not receive any prednisone (Table 1). The association between AVN and antiphospholipid antibodies (lupus anticoagulant) was only significant in the univariate analysis.

Conclusion: Prednisone dose remains the most important predictor of AVN. At doses below 60 mg, AVN occurrence is dependent on the duration of highest prednisone dose. However, for doses above 60 mg, the risk is independent of the duration. The major reduction in AVN incidence in later decades likely reflects lower use of such high prednisone doses. Particular attention should be paid to avoidance of prednisone at 60 mg doses in African-Americans, given the doubling of risk of AVN as compared to Caucasians.

ACR Abstract- AVN and SLE_MP_Table1

ACR Meeting Abstracts - https://acrabstracts.org/abstract/risk-factors-for-avascular-necrosis-in-sle-a-multivariate-model/