Impact of Glycaemic Status on the Cardiac Effects of Empagliflozin when Initiated Immediately after Myocardial Infarction: A Post-Hoc Analysis of the EMMY Trial.

Besides a significant reduction in major adverse cardiovascular events, the EMPA-REG outcome trial has shown a reduced risk for cardiovascular death and hospitalizations for heart failure compared with placebo in people with type 2 diabetes (T2D) treated with the sodium-glucose cotransporter 2 inhibitor (SGLT2i) empagliflozin.1 The EMPEROR trials showed a similar risk reduction for cardiovascular death or hospitalization for heart failure in patients with chronic heart failure, both with preserved and reduced ejection fraction, when treated with empagliflozin, irrespective of the glycaemic status of the study participants.2, 3 Furthermore, the empagliflozin in acute myocardial infarction (EMMY) trial showed beneficial effects of empagliflozin treatment on the N-terminal prohormone of brain natriuretic peptide (NT-proBNP) as well as functional and structural cardiac parameters when treatment was initiated immediately after an acute myocardial infarction (AMI).4 As previous data suggest that dysglycaemia is very common in people with coronary artery disease or MI,5, 6 this post-hoc analysis of the EMMY trial aims to investigate whether the observed beneficial cardiac effects depend on the baseline glycaemic status of the participants. This is a post hoc subgroup analysis of the EMMY trial. Details of the EMMY study design and the primary outcome have been published previously.4, 7 Briefly, the EMMY trial was a multicentre, randomized (1:1 ratio), double-blind and placebo-controlled trial, which investigated the effect of empagliflozin (10 mg once daily), administered for 26 weeks, on heart failure biomarkers and cardiac function metrices when initiated immediately (within 72 h) after AMI. The enrolled participants who underwent percutaneous coronary intervention after an AMI, were 18-80 years of age, haemodynamically stable and had a blood pressure >110/70 mmHg. Patients with other types of diabetes than T2D, a blood pH value of <7.32, haemodynamic instability, acute urinary tract or genital infections, on current SGLT2i treatment, or having received SGLT2i treatment within 4 weeks before enrolment were excluded from the trial. The EMMY trial was approved by the ethics committee of the Medical University of Graz, Austria (EK 29-179 ex16/17, EudraCT 2016-004591-22) and registered at ClinicalTrials.gov (NCT03087773). The trial performance conformed to the 1964 Declaration of Helsinki and adhered to the guidelines of Good Clinical Practice (ICH GCP E6). In this prespecified secondary analysis, we investigated the impact of glycaemic status on treatment effects and potential differences in trajectories of NT-proBNP levels as well as structural and functional cardiac parameters following AMI. Three categories of glycaemic status were compared: normoglycaemia, prediabetes and T2D. Participants were classified as normoglycaemic if the glycated haemoglobin (HbA1c) levels during admission for the AMI were <5.7% (<39 mmol/mol). Prediabetes was defined as an HbA1c of 5.7%-6.4% (39-47 mmol/mol) without glucose-lowering medication. Patients were classified as having T2D in case of previous glucose-lowering medication and/or an HbA1c of >6.4% (>47 mmol/mol) at enrolment. Continuous variables were reported as mean ± standard deviation or median with interquartile range (IQR) and categorical variables as frequency with percentage. The impact of the glycaemic status on the empagliflozin treatment effect on cardiac markers was assessed using multiple linear regression models adjusting for traditional risk factors, including age, sex and baseline NT-proBNP levels. For each cardiac marker the modification effect of glycaemic status was investigated by adding an interaction term between treatment and glycaemic status into the multiple linear regression model. All statistical analyses were performed using R software version 4.1.0. (https://www.r-project.org). Values for p < .05 were considered statistically significant according to glycaemic status. In 456 of the overall 476 participants randomized in the EMMY trial, HbA1c levels were available during the hospitalization for AMI; 63 (13.8%) had established T2D, 158 (34.7%) prediabetes and 235 (51.5%) were considered normoglycaemic, respectively. The percentage of people with obesity increased gradually from the normoglycaemic to the T2D category, as did hypertension. HbA1c levels were 5.4% (IQR 5.3%-5.5%) in the normoglycaemic group, 5.9% (IQR 5.7-6.0%) in those with prediabetes and 6.8% (6.4%-8.3%) in the T2D group, respectively. NT-proBNP levels and left ventricular ejection fraction did not differ between the three groups. Table 1 provides the full set of baseline characteristics. The EMMY trial reported for the overall study population, which meant NT-proBNP over 26 weeks was 15% (95% confidence interval −4.4% to −23.6%) lower with empagliflozin compared with placebo after adjusting for baseline NT-proBNP, sex and diabetes status in the main analysis. The impact of empagliflozin on the changes in NT-proBNP levels from baseline to week 26 was not found to be dependent of glycaemic status at baseline (pinteraction .133 and .582 for the comparison of prediabetes vs. normoglycaemic and T2D vs. normoglycaemia, respectively). Table 2 provides details of the analysis of the secondary echocardiography-based study outcomes for interaction with the three glycaemic groups. Only for the left ventricular ejection fraction, a significant interaction effect (pinteraction = .032) is found when comparing prediabetes with normoglycaemia suggesting a less pronounced effect in people with prediabetes. No such interaction is observed for people with T2D compared with those with normoglycaemia. Within the EMMY trial, impaired glucose metabolism is present in 48.5% of patients admitted to the hospital for AMI. However, the current subgroup analysis shows similar effects of empagliflozin treatment in all glycaemic groups on NT-proBNP levels as well as structural and functional parameters of cardiac function, when the treatment is initiated immediately after an AMI. A number of reasons for these beneficial findings were discussed previously, including inflammation.8 However, in the EMMY trial we observed a significant reduction in inflammatory markers after the AMI but without an impact of the empagliflozin treatment.9 While there is one significant interaction for left ventricular ejection fraction between the prediabetes and normoglycaemia group, suggesting a more pronounced improvement in the latter group with empagliflozin treatment, we would refrain from overemphasizing this finding with a pinteraction of .032. As we performed 10 interaction analyses and none of the other parameters showed a similar signal, this observed interaction might be even by chance. Our results, however, are in line with previous analyses showing similar beneficial effects of empagliflozin across the spectrum of glycaemia in people with heart failure and reduced3 or preserved10 ejection fraction. The EMMY study was no cardiovascular outcome trial and used the heart failure surrogate marker NT-proBNP as the primary outcome. The recently published DAPA-MI trial investigating the SGLT2i dapagliflozin in patients with MI, showed an improvement in cardiometabolic outcomes but without improvement in the composite of heart failure hospitalizations or cardiovascular death compared with placebo.11 In the EMMY trial, the SGLT2i treatment with empagliflozin was initiated in all participants within 72 h after the acute event, while the DAPA-MI trial allowed for up to 10 days after a preceding AMI. The cardiovascular outcome trial of empagliflozin in people with AMI is still ongoing (NCT04509674).12 Another limitation of the trial is that we did not perform oral glucose tolerance tests; hence, the prevalence of people with prediabetes or diabetes based on post-challenge hyperglycaemia might be even higher than the one observed. In this prespecified subgroup analysis of the EMMY trial, empagliflozin significantly improved NT-proBNP and echocardiographic structural and functional parameters largely irrespective of the baseline glycaemic status of the participants. HS and DvL designed the study; AO and F Aziz analysed the data; F Aberer, TP, NJT, PNP, HD, MB, EK and HB conducted the trial and/or collected the data; and CS wrote the first draft of the manuscript. All authors read and revised the manuscript carefully. We would like to thank all the participants of the EMMY trial. The EMMY trial was funded by an investigator-initiated study grant to HS. HS is on the speakers' bureau and/or the advisory board of Amgen, Amarin, Bayer, Boehringer Ingelheim, Daiichi Sankyo, Eli Lilly, K:businesscom, NovoNordisk and Sanofi. All other authors have no conflicts of interest to declare. The peer review history for this article is available at https://www.webofscience.com/api/gateway/wos/peer-review/10.1111/dom.15477. Data is available upon reasonable request to the principal investigators.