Cardiometabolic outcomes with dapagliflozin after myocardial infarction by baseline ejection fraction: DAPA-MI

Abstract

Aims: In the randomized DAPA-MI clinical trial, 10 mg of dapagliflozin once daily improved cardiometabolic outcomes versus placebo after acute myocardial infarction (MI) in patients without established diabetes or heart failure (HF). We assessed associations between baseline left ventricular ejection fraction (LVEF) and cardiometabolic outcomes in DAPA-MI.

Methods: The primary outcome, assessed using the win ratio method, was the hierarchical composite of death, hospitalization for HF, non-fatal MI, atrial fibrillation/flutter, Type 2 diabetes, New York Heart Association classification at last visit and body weight decrease of ≥5% from baseline to last visit. For the present analysis, patients were categorized using LVEF at randomization (<50% or ≥50%).

Results: Of the DAPA-MI participants with available LVEF data who received ≥1 dose of study drug (n = 3751), 2913 (77.7%) had LVEF <50% and 838 (22.3%) had LVEF ≥50%. The primary hierarchical composite outcome resulted in a win ratio favouring dapagliflozin of 1.38 (95% CI: 1.21, 1.57; P < 0.001) in patients with LVEF <50% and 1.32 (1.00, 1.73; P = 0.048) in patients with LVEF ≥ 50% (P interaction = 0.76). In a sensitivity analysis excluding patients with LVEF <30%, the primary hierarchical composite outcome resulted in a win ratio favouring dapagliflozin of 1.40 (95% CI: 1.22, 1.61; P < 0.001). There were no significant interactions between baseline LVEF and any secondary outcomes.

Conclusions: Regardless of baseline LVEF, dapagliflozin resulted in significant cardiometabolic benefits versus placebo, although there was no impact on the composite of cardiovascular death or hospitalization for HF.

Keywords: dapagliflozin; heart failure; left ventricular ejection fraction; myocardial infarction; sodium–glucose cotransporter‐2 inhibitors.

Figure 1

Primary and key secondary hierarchical composite outcomes according to LVEF baseline, assessed by the win ratio method. Arrow indicates order of endpoint hierarchy. Estimates include the components on the y‐axis. Percentages are per cent comparisons resulting in a win for dapagliflozin 10 mg, a tie, or a win for placebo. Percentages may not add up to 100% owing to rounding. The components in hierarchical order are: (1) death; (2) hospitalization for heart failure; (3) non‐fatal MI; (4) atrial fibrillation/flutter; (5) new diagnosis of Type 2 diabetes; (6) NYHA class; and (7) weight decrease ≥5%. CI, confidence interval; LVEF, left ventricular ejection fraction; MI, myocardial infarction; NYHA, New York Heart Association.

Figure 2

Ordinal logistic regression analysis of NYHA class during study follow‐up according to LVEF at baseline. OR estimates are for dapagliflozin 10 mg versus placebo for the odds of having a higher NYHA class value (indicating more severe HF symptoms and limitations) at a given time point. ^aOR estimates for patients with baseline LVEF ≥50% after 22 months were unavailable owing to the low number of observations. CI, confidence interval; HF, heart failure; LVEF, left ventricular ejection fraction; NYHA, New York Heart Association; OR, odds ratio.

Figure 3

Estimated mean body weight change with dapagliflozin versus placebo during follow‐up, by LVEF at baseline. Error bars represent 95% CIs. P values represent the statistical significance of the effect of dapagliflozin on body weight versus placebo. CI, confidence interval; LVEF, left ventricular ejection fraction.