Time trend and risk factors of avascular bone necrosis in patients with systemic lupus erythematosus

Abstract

Objectives The objective of this paper is to study the time trend and risk factors of avascular bone necrosis (AVN) in patients with systemic lupus erythematosus (SLE). Methods Between 1999 and 2014, patients who fulfilled the ACR criteria for SLE and developed symptomatic AVN were identified from our cohort database and compared with those without AVN, matched for age, sex and SLE duration. The standardized incidence ratios (SIRs) of AVN in different SLE age groups were calculated from data derived from our hospital registry and population census. Risk factors for AVN were studied by logistic regression, adjusted by a propensity score for ever use of high-dose glucocorticoids (GCs). Results Fifty-five SLE patients with AVN and 220 SLE patients without AVN were studied. There were 104 AVN sites involved, with the hips being most commonly affected (82%). The point prevalence of AVN in our SLE cohort was 7.4%. The SIRs of AVN in our SLE patients were 131 (86.6–199; p < 0.001) and 56.0 (34.3–91.4; p < 0.001), respectively, in the periods 1995–2004 and 2005–2014. In both decades, the age-stratified SIR was highest in the youngest age group (<19 years). AVN patients were more likely to be treated with GCs and had received a significantly higher cumulative dose of prednisolone since SLE diagnosis (16.5 vs 10.7 grams; p = 0.001). The SLE damage score (excluding AVN) was also significantly higher in AVN than non-AVN patients (2.5 vs 0.4; p < 0.001). Logistic regression revealed that preceding septic arthritis of the involved joint (odds ratio (OR) 17.7 (1.5–205); p = 0.02), cushingoid body habitus (OR 2.4 (1.1–5.2); p = 0.04), LDL cholesterol level (OR 1.4 (1.0–1.9); p = 0.04), maximum daily dose of prednisolone (OR 6.4 (1.2–33.3); p = 0.03) and cumulative dose of prednisolone received in the first six months of the first lupus flare (OR 1.3 (1.0–1.8); p = 0.046) were independently associated with AVN. Conclusions AVN is prevalent in SLE, particularly in younger patients. The use of GCs remains the strongest independent risk factor. A trend of reduction in the SIR of AVN in our SLE patients is observed over the past two decades.