Background Corticosteroid (CS) is effective for controlling inflammatory reactions in SLE, but may have been known to increase risk of avascular necrosis (AVN) in patients with Systemic lupus erythematosus (SLE). This population-based study aims to investigate the association between cumulative dose of CS and the incidence of AVN in newly diagnosed SLE patients.

Methods We conducted a nested case-control study using Korean National Health Insurance claims database from 2002 to 2018. Cohort of incident diagnosis of SLE (ICD10, M32) with rare intractable disease code (V136) were identified during 2007~2018. Cases, who were diagnosed for any reason with AVN (M87.0x, M87.1x, M87.3x, M87.8x, M87.9x, M90.5x (ICD-10)), were matched up to 10 controls based on gender, age (±5yrs), and date of SLE diagnosis (±6 months). Conditional logistic regression analysis was performed to calculate adjusted odds ratios (ORs) and 95% confidence intervals (CIs) of AVN occurrence associated with cumulative dose of CS [prednisolone-equivalent dose (PED)] from the date of SLE diagnosis to the index date, with adjusting for hypertension, dyslipidemia, immunosuppressants and so forth. Cumulative CS dose was categorized into 3 groups (0≤ < 4g, 4≤ <8g and 8g≤) based on distribution in the study population.

Results Among the 12,375 SLE patients who were newly diagnosed during the inclusion period, a total of 358 AVN patients were identified. Controls without AVN (N=3,276) were identified. In terms of cumulative CS dose, used after first diagnosis of SLE, compared with cumulative CS dose less than 4g, ORs associated with 4≤ <8g and 8g≤ group were 2.48 (95% CI 1.70 – 3.64) and 6.14 (95% CI 4.20 – 8.93), respectively.

Conclusions The trend of association between higher cumulative CS dose and risk of AVN may provide evidence for monitoring patients with continuous use of CS for SLE management, especially in aspect of cumulative dose of CS.