Background/Purpose: While both metformin (MET) and SGLT2 inhibitors (SGLT2i) have potential immunomodulatory effects in Systemic Lupus Erythematosus (SLE), their comparative efficacy in glucocorticoid (GC)-sparing remains unclear. This study aims to compare GC discontinuation and flare suppression between MET and SGLT2i in SLE patients.

Methods: This is a retrospective cohort study using TriNetX’s Global collaborative network. We defined a cohort of SLE patients (ICD-10 M32) prescribed hydroxychloroquine and with anti-dsDNA antibody testing. Within this cohort, patients initiating glucocorticoid treatment (prednisolone or prednisone) were then categorized into two distinct groups for analysis: the MET group, comprising those subsequently prescribed metformin, and the SGLT2i group, comprising those subsequently prescribed a SGLT2i after glucocorticoid exposure. The primary endpoint was GC discontinuation (defined as no GC prescriptions) within a 3-12 month period following the index prescription of MET or an SGLT2i. Primary outcome sensitivity analyses focused on GC tablets ≤5 mg, with additional analyses in the glucocorticoid ≤5 mg tablet group when MET or SGLT2 inhibitors were subsequently prescribed. The secondary endpoint, assessed in patients from both groups who had discontinued GC, was GC reinitiation within a 3-month to 3-year period. The primary endpoint was compared between groups using the Chi-squared test. For the secondary endpoint, comparisons were made after propensity score matching (age, sex, race, BMI, anti-dsDNA antibody, complements, MMF, Belimumab, Anifrolumab, Diabetes mellitus, HbA1c, Heart failure, End-stage renal disease, Chronic kidney disease, unspecified) utilizing the TriNetX platform.

Results: MET was initiated in 1228 patients, with 89.5% being women, and the mean age was 56 ± 14 years. SGLT2i was initiated in 718 patients, with 82.4% being women, and the mean age was 55±17 years. The duration of SLE disease before the first prescription of MET and SGLT2i was 1827 ± 1386 and 2593 ± 1608 days, respectively. Mycophenolic acid use was 71% in the SGLT2i group and 53% in the MET group. Baseline HbA1c levels were numerically higher in the MET group (6.3 ± 1.4%) than in the SGLT2i group (5.8 ± 1.13%). Among these patients, GC was discontinued in 445 (36.2%) patients in the MET group and in 488 (68.4%) patients in the SGLT2i group. GC discontinuation was significantly more frequent in the SGLT2i group (p < 0.001; Odds Ratio [OR] 3.85, 95% CI 3.23-4.60). The sensitivity analyses corroborated the primary results. After propensity score matching of patients who had discontinued GC, 248 patients remained in each group for the secondary endpoint analysis. In this propensity score-matched cohort, the GC reinitiation rate within 3 months to 3 years after GC discontinuation was significantly lower in the SGLT2i group (10.1%) compared to the MET group (23.0%) (p < 0.0001; OR 3.87, 95% CI 1.60-4.43).

Conclusion: In SLE patients treated with GC, SGLT2i use was associated with a significantly higher rate of GC discontinuation and a lower rate of subsequent GC reinitiation compared to MET. These findings suggest that SGLT2 inhibitors may offer a superior GC-sparing effect in SLE patients.

« Back to ACR Convergence 2025

ACR Meeting Abstracts - https://acrabstracts.org/abstract/glucocorticoid-sparing-effects-and-flare-suppression-metformin-versus-sglt2-inhibitors-in-systemic-lupus-erythematosus/