Cardiorenal outcomes with ertugliflozin by baseline metformin use: post-hoc analyses of the VERTIS CV trial

Abstract Introduction We analysed data from the VERTIS CV trial that investigated the CV and kidney safety and efficacy of the sodium-glucose cotransporter 2 (SGLT2) inhibitor ertugliflozin (ERTU) vs placebo (PBO) to assess the impact of metformin (MET) use at baseline (BL). These analyses are timely because the recent ESC guidelines recommendation to use SGLT2 inhibitor or GLP-1 RAs as initial glucose-lowering therapy in patients with type 2 diabetes (T2D) with or at high risk for atherosclerotic cardiovascular disease (ASCVD) has been questioned because outcome trials of these drug classes included a large proportion of patients with MET as background therapy, yet MET was not used at BL in approximately 25% of patients in each trial. Purpose These analyses determined cardiorenal endpoints of VERTIS CV according to use of BL MET to assess for evidence of treatment effect modification for ERTU by BL MET use, adjusting for the probability (propensity) of BL MET use. Methods VERTIS CV was an international, double-blind, PBO-controlled trial of 2 doses of ERTU (5 mg; 15 mg) vs PBO in patients with T2DM and ASCVD. As prospectively planned, the 2 ERTU dose groups were combined for all analyses vs PBO. Differences in risk of CV and kidney outcomes between ERTU and PBO across subgroups by BL MET use were conducted using Cox proportional hazards model along with propensity adjustment using inverse probability for treatment weighting to account for differences in patient mix between those with and without BL MET influenced by individual BL characteristics and risk factors. Treatment (categorical), BL MET use (categorical) and the interaction term between treatment and BL MET use subgroup (no or yes) were used in each model to assess effect modification by BL MET use. Hazard ratio and 95% CI are presented along with Pinteraction for evaluation of treatment effect modification by BL MET use. Results In VERTIS CV, 8246 patients were randomised to ERTU 5 mg, 15 mg or PBO. Of these, 6286 (76%) patients used MET (alone or with other glucose-lowering agents [GLA]) at BL. Differences in BL characteristics by BL MET use subgroup (no or yes) included a higher mean UACR (204.4 vs 129.8 mg/g), more patients with eGFR <60 mL/min/1.73 m2 (34.8% vs 17.9%), more patients on a single GLA (76.9% vs 18.3%), higher insulin use (67.6% vs 40.9%), lower sulphonylurea use (32.2% vs 43.8%) and a slightly longer disease duration (14.4 vs 12.5 years) in the subgroup without vs with BL MET, respectively. No significant differences in the relative risk for cardiorenal outcomes were observed with or without BL MET use (Figure; all Pinteraction values >0.05). Conclusions In VERTIS CV, there was no evidence for effect modification by BL MET use on the effects of ERTU on cardiorenal outcomes in patients with T2D and ASCVD. Funding Acknowledgement Type of funding sources: Other. Main funding source(s): Sponsored by Merck Sharp & Dohme Corp., a subsidiary of Merck & Co., Inc., Kenilworth, NJ, USA, and Pfizer Inc., New York, NY, USA.