No Effect of Low-Dose Glucocorticoid Maintenance Therapy on Damage in SLE Patients in Prolonged Remission: A Propensity Score Analysis of the Longitudinal Lupus-Cruces-Bordeaux Inception Cohort

Abstract

Background/Objectives: Prolonged remission on low-dose glucocorticoids (GC) is a main goal in patients with systemic lupus erythematosus (SLE). The aim of this study is to assess whether GC ≤ 5 mg/d increases the risk of damage accrual in patients with SLE in prolonged remission. Methods: Observational study of routine clinical care data of the inception Lupus Cruces-Bordeaux cohort. Only patients in DORIS remission during five consecutive yearly visits were included. The endpoint was damage accrual during the 5-year follow-up, either global or specific damage: GC-induced, cardiovascular (CV), lupus and other. Patients no longer on GC therapy by year 5 (GC5-Off) were compared with those who continued GC therapy (GC5-On). Comparisons were made by Cox and Poisson regressions, which were adjusted with propensity score (PE) in order to control for confounding by indication. Results: 132 patients were included, 56 in the GC5-On and 76 in the GC5-Off groups. All patients were on GC ≤ 5 mg/d for the whole follow-up, the mean prednisone dose in the GC5-On group being 2.96 mg/d during the whole study period and 2.6 mg/d during the 5th year. Fourteen patients (10.6%) accrued damage. More patients in the GC5-On group accrued global damage, 16% vs. 7% in the GC5-Off group, p = 0.08, mainly at CV domains (7% vs. 1%, respectively, p = 0.16). In the PS-adjusted Cox and Poisson regressions, the GC5-On group was not significantly associated with global (p = 0.39) or CV damage accrual (p = 0.62), nor with the absolute (p = 0.40) or CV-restricted final SDI scores (p = 0.63). The C-index of the propensity score model was 0.79. Conclusions: Maintaining doses of prednisone < 5 mg/d in lupus patients in prolonged remission is not associated with an increased risk of damage accrual.