Empagliflozin and Left Ventricular Remodeling in People Without Diabetes: Primary Results of the EMPA-HEART 2 CardioLink-7 Randomized Clinical Trial

Background: Sodium-glucose cotransporter 2 inhibitors have been demonstrated to promote reverse cardiac remodeling in people with diabetes or heart failure. Although it has been theorized that sodium-glucose cotransporter 2 inhibitors might afford similar benefits in people without diabetes or prevalent heart failure, this has not been evaluated. We sought to determine whether sodium-glucose cotransporter 2 inhibition with empagliflozin leads to a decrease in left ventricular (LV) mass in people without type 2 diabetes or significant heart failure. Methods: Between April 2021 and January 2022, 169 individuals, 40 to 80 years of age, without diabetes but with risk factors for adverse cardiac remodeling were randomly assigned to empagliflozin (10 mg/d; n=85) or placebo (n=84) for 6 months. The primary outcome was the 6-month change in LV mass indexed (LVMi) to baseline body surface area as measured by cardiac magnetic resonance imaging. Other measures included 6-month changes in LV end-diastolic and LV end-systolic volumes indexed to baseline body surface area and LV ejection fraction. Results: Among the 169 participants (141 men [83%]; mean age, 59.310.5 years), baseline LVMi was 63.217.9 g/m(2) and 63.8 +/- 14.0 g/m(2) for the empagliflozin- and placebo-assigned groups, respectively. The difference (95% CI) in LVMi at 6 months in the empagliflozin group versus placebo group adjusted for baseline LVMi was -0.30 g/m(2) (-2.1 to 1.5 g/m(2); P=0.74). Median baseline (interquartile range) NT-proBNP (N-terminal-pro B-type natriuretic peptide) was 51 pg/mL (20-105 pg/mL) and 55 pg/mL (21-132 pg/mL) for the empagliflozin- and placebo-assigned groups, respectively. The 6-month treatment effect of empagliflozin versus placebo (95% CI) on blood pressure and NT-proBNP (adjusted for baseline values) were -1.3 mmHg (-5.2 to 2.6 mmHg; P=0.52), 0.69 mmHg (-1.9 to 3.3 mmHg; P=0.60), and -6.1 pg/mL (-37.0 to 24.8 pg/mL; P=0.70) for systolic blood pressure, diastolic blood pressure, and NT-proBNP, respectively. No clinically meaningful between-group differences in LV volumes (diastolic and systolic indexed to baseline body surface area) or ejection fraction were observed. No difference in adverse events was noted between the groups. Conclusions: Among people with neither diabetes nor significant heart failure but with risk factors for adverse cardiac remodeling, sodium-glucose cotransporter 2 inhibition with empagliflozin did not result in a meaningful reduction in LVMi after 6 months.