Effects of empagliflozin in different phases of diabetes mellitus-related cardiomyopathy: a prospective observational study

Background Diabetes mellitus-related cardiomyopathy (DMCMP), defined as left ventricular (LV) dysfunction caused by hyperglycemia in the absence of coronary artery disease, leads to heart failure (HF). Previous studies have shown that treatment with sodium-glucose co-transporter 2 inhibitor (SGLT2i) reduces the risk of exacerbation of HF. The beneficial effects of SGLT2i on HF depend not only on indirect actions such as osmotic diuresis but also on direct actions on the myocardium, leading to improvements in LV function. However, it remains unclear whether SGLT2i treatment is equally effective in any phase of DMCMP. The aim of this observational study was to compare the efficacy of SGLT2i treatment on LV dysfunction between early and advanced DMCMP. Methods Thirty-five symptomatic non-ischemic HF patients with LV ejection fraction > 40% and type 2 diabetes mellitus (T2DM) treated with empagliflozin (EMPA group) and 20 controls treated without SGLT2i were enrolled. According to the myocardial extracellular volume fraction (ECV), a reliable marker of cardiac fibrosis quantified by cardiac magnetic resonance, the EMPA group was further divided into early DMCMP (n = 16, ECV ≤ 30%) and advanced DMCMP (n = 19, ECV > 30%) groups and followed up prospectively. Echocardiography was performed at baseline and after 12 months. LV function assessed as LV global longitudinal strain (LVGLS) and the ratio of early diastolic mitral inflow velocity to early diastolic mitral annular velocity (E/e′) were compared. Results ECV was strongly correlated with T2DM duration (r2 = 0.65, p p = 0.005, and 4.6 ± 1.5% (early DMCMP) vs. 1.6 ± 3.3% (advanced DMCMP), p = 0.003) and E/e′ (ΔE/e′: − 1.5 ± 4.7 vs. − 0.3 ± 3.0, p = 0.253, and − 3.4 ± 5.5 vs. − 0.1 ± 3.5, p = 0.043). Conclusions The positive effects of empagliflozin on LV dysfunction were more remarkable in early than in advanced DMCMP. Early intervention of SGLT2i for DMCMP may be preferable.