Rheumatol Ther. 2026 Apr 28. doi: 10.1007/s40744-026-00857-y. Online ahead of print.
ABSTRACT
INTRODUCTION: Despite the European Alliance of Associations for Rheumatology (EULAR) recommendation to minimize glucocorticoid (GC) use in systemic lupus erythematosus (SLE), prospective data quantifying toxicity across low-dose ranges are lacking. This study aimed to assess toxicity using the GC toxicity index (GTI) in SLE patients and compare toxicity profiles among dose-defined subgroups.
METHODS: Patients from the STAR cohort (May 2023-May 2024) were prospectively followed up for 1 year. Stratified by average daily prednisone (PDN) dose, toxicity was assessed using GTI comprising the aggregate improvement score (AIS) and the cumulative worsening score (CWS) at baseline and 1 year. Three pre-planned stepwise comparisons used dose thresholds of 7.5 mg, 5 mg, and 2.5 mg, with a Bonferroni-corrected significance level of P < 0.0167 (α = 0.05/3). Quantile regression evaluated the association between average daily PDN dose and CWS/AIS.
RESULTS: Of 302 patients, the PDN ≤ 7.5 mg/day group (n = 223) showed statistically lower median CWS [0 (IQR 0-19) vs. 48 (IQR 19-84), P < 0.001] and AIS [0 (IQR – 19-10) vs. 40 (IQR 9-74), P < 0.001] compared to the PDN > 7.5 mg/day group (n = 79). Within the low-dose group, patients with 5 < PDN ≤ 7.5 mg/day (n = 52) exhibited higher median CWS [10.5 (IQR 0-29) vs. 0 (IQR 0-19), P = 0.002] and wider AIS interquartile range [0 (IQR – 18.75-29) vs. 0 (IQR – 20-0), P = 0.010] than the PDN ≤ 5 mg/day subgroup (n = 171). No significant differences in CWS or AIS were observed between the PDN ≤ 2.5 mg/day (n = 90) and 2.5 < PDN ≤ 5 mg/day (n = 81) subgroups. Quantile regression indicated that each 1 mg/day increase in PDN dose raised median CWS by 3.33 points and median AIS by 3.42 points.
CONCLUSIONS: To our knowledge, this study provided the first prospective and quantitative evidence using the GTI to demonstrate that PDN dose reduction to ≤ 5 mg/day was linked to reduced toxicity. Moreover, we found that no dose was entirely safe, which strongly supported the EULAR strategy of rigorous GC minimization.
PMID:42050254 | DOI:10.1007/s40744-026-00857-y
