Mineralocorticoid receptor antagonists with sodium–glucose co-transporter-2 inhibitors in heart failure: a meta-analysis

Abstract Background and Aims To investigate the cardiovascular effects of sodium–glucose co-transporter-2 inhibitors (SGLT2i) with concomitant mineralocorticoid receptor antagonist (MRA) use in heart failure (HF) regardless of ejection fraction (EF) and explore the risk of MRA-associated adverse events in individuals randomized to SGLT2i vs. placebo. Methods PubMed/MEDLINE, Web of Science, Embase, and clinical trial registries were searched for randomized controlled trials/post-hoc analyses evaluating SGLT2i in HF with or without MRA use (PROSPERO: CRD42023397129). The main outcomes were composite of first hospitalization or urgent visit for HF/cardiovascular death (HHF/CVD), HHF, and CVD. Others were all-cause mortality, composite renal and safety outcomes. Hazard ratios (HR)/risk ratios were extracted. Fixed-effects meta-analyses and subgroup analyses were performed. Results Five eligible studies were included, pooling data from 21 947 people with HF (type 2 diabetes mellitus, n = 10 805). Compared to placebo, randomization to SGLT2i showed a similar reduction in HHF/CVD and HHF in people who were or were not using MRAs [HHF/CVD: hazard ratio (HR) 0.75; 95% confidence interval (CI) 0.68–0.81 vs. HR 0.79; 95% CI 0.72–0.86; P-interaction = .43; HHF: HR 0.74; 95% CI 0.67–0.83 vs. HR 0.71; 95% CI 0.63–0.80; P-interaction = .53], with a suggestion of greater relative reduction in CVD in chronic HF people randomized to SGLT2i and using MRAs irrespective of EF (HR 0.81; 95% CI 0.72–0.91 vs. HR 0.98; 95% CI 0.86–1.13; P-interaction = .034). SGLT2i reduced all-cause mortality (P-interaction = .27) and adverse renal endpoints regardless of MRA use (P-interaction = .73) despite a higher risk of volume depletion with concomitant MRAs (P-interaction = .082). SGLT2i attenuated the risk of mild hyperkalaemia (P-interaction < .001) and severe hyperkalaemia (P-interaction = .051) associated with MRA use. Conclusions MRAs did not influence SGLT2i effects on the composite of HHF/CVD, HHF or all-cause mortality; however, findings hinted at a more pronounced relative reduction in CVD in chronic HF patients regardless of EF who were randomized to SGLT2i and receiving an MRA compared to those randomized to SGLT2i and not receiving MRAs. SGLT2i attenuated the risk of MRA-associated treatment-emergent hyperkalaemia. These findings warrant further validation in well-designed randomized controlled trials.