6G-05 - Timing of SGLT2 inhibitors Initiation on Clinical Complications in Diabetic Metabolic Dysfunction-Associated Steatotic Liver Disease: A Target Trial Emulation Study With a Clone-Censor-Weight Approach

Background: Recent studies have shown clinical benefits of sodium-glucose cotransporter 2 (SGLT2) inhibitors in diabetic metabolic dysfunction-associated steatotic liver disease (MASLD). However, the optimal timing for initiating SGLT2 inhibitors in diabetic MASLD remains unclear.

Objectives: To emulate a target trial to evaluate the impact of SGLT2 inhibitor initiation timing following MASLD diagnosis in patients with preexisting type 2 diabetes.

Methods: Using a linkage of Kangbuk Samsung Health Study (>250,000 participants) and Korea’s nationwide insurance claims data (2013–2022), we included patients diagnosed with MASLD and preexisting type 2 diabetes between 2014 and 2021. MASLD was identified by the presence of fatty liver on ultrasonography, with the first diagnosis date defined as cohort entry. The primary outcomes included hepatic (hepatic decompensation, hepatocellular carcinoma), cardiovascular (myocardial infarction, stroke, heart failure), and renal complications (dialysis, end-stage renal disease). Three cohorts were established for each outcome, excluding patients with the respective outcome at baseline. Using the clone-censor-weight approach, clones were replicated and assigned to one of three strategies at cohort entry: initiation of SGLT2 inhibitors within 6 months, 1 year, or no treatment during the 3-year follow-up. Clones were censored if they deviated from their assigned regimen, and inverse probability of censoring weights were predicted from a pooled multivariable logistic regression model to account for informative censoring. Kaplan-Meier methods were used to calculate the weighted 3-year absolute risks of each outcome, and risk differences (RD) and risk ratios (RR) were calculated with non-initiators as reference. The 95% confidence intervals (CIs) were obtained from 300 non-parametric bootstraps.

Results: A total of 7,947 diabetic MASLD patients (mean age 53.7 years, 67.0% male) were identified. Across three cohorts, 2.4% initiated within 6 months, 4.3% within 1 year, and 81.8% were non-initiators. Early initiation within 6 months was associated with lower risks for hepatic, cardiovascular, and renal events, with RD of -0.21% (95% CI -0.30% to -0.14%), -0.63% (-0.76% to -0.55%), and -0.40% (-0.52% to -0.29%), respectively. The 3-year RRs of each outcome were 0.74 (0.52 to 0.99), 0.33 (0.12 to 0.68) and 0.20 (0.11 to 0.30), respectively. For initiation within 1 year, the 3-year RDs were -0.13% (95% CI -0.67% to 0.40%), -0.04% (-0.71% to 0.26%), and -0.32% (-0.47% to -0.22%), corresponding to RRs of 0.84 (0.40 to 1.15), 0.96 (0.44 to 1.50) and 0.35 (0.20 to 0.48), respectively.

Conclusions: Early initiation of SGLT2 inhibitors may be beneficial in lowering risks of complications in patients with diabetic MASLD.