New England Journal of Medicine APRIL 06, 2024

Empagliflozin after Acute Myocardial Infarction (EMPACT-MI)

Javed Butler, W. Schuyler Jones, Jacob A. Udell, et al.

Bottom Line

In patients at high risk for heart failure following an acute myocardial infarction, the addition of empagliflozin to standard of care did not significantly reduce the composite primary endpoint of first heart failure hospitalization or all-cause mortality.

Key Findings

1. The primary composite outcome (time to first hospitalization for heart failure or all-cause mortality) occurred in 8.2% of the empagliflozin group versus 9.1% in the placebo group, with incidence rates of 5.9 and 6.6 events per 100 patient-years, respectively (hazard ratio 0.90, 95% CI 0.76-1.06, p=0.21).
2. Empagliflozin therapy resulted in a statistically significant 23% relative risk reduction in the time to first heart failure hospitalization (2.6 vs. 3.4 events per 100 patient-years; hazard ratio 0.77, 95% CI 0.60-0.98, p=0.031).
3. A significant reduction in total (first and recurrent) heart failure hospitalizations was observed with empagliflozin compared to placebo (incidence rate ratio 0.67, 95% CI 0.51-0.89, p=0.006).
4. There was no significant difference in all-cause mortality between the two study arms (5.2% vs. 5.5%; hazard ratio 0.96, 95% CI 0.78-1.19).

Study Design

Design
RCT
Double-Blind
Sample
6,522
Patients
Duration
17.9 mo
Median
Setting
Multicenter, international
Population Adults hospitalized for acute myocardial infarction within 14 days of admission, with signs/symptoms of congestion or newly reduced LVEF <45% and additional risk factors for heart failure
Intervention Empagliflozin 10 mg daily added to standard of care
Comparator Placebo added to standard of care
Outcome Composite of time to first hospitalization for heart failure or all-cause mortality

Study Limitations

The study failed to meet its primary composite endpoint of time to first heart failure hospitalization or all-cause mortality.
The trial was conducted during the COVID-19 pandemic and regional geopolitical conflicts, which may have influenced healthcare utilization patterns and outpatient heart failure detection.
The primary endpoint was time-to-first-event, which may not fully capture the burden of recurrent heart failure events, though exploratory analyses of total events were favorable.
The study population was selected based on high-risk criteria for heart failure post-MI, potentially limiting the generalizability of these specific findings to lower-risk post-MI patients.

Clinical Significance

While the primary composite endpoint was not met, the significant reduction in both first and recurrent heart failure hospitalizations suggests that SGLT2 inhibitors may provide mechanistic benefit in preventing heart failure progression in high-risk patients post-MI, independent of traditional diabetes or heart failure status, warranting careful consideration in clinical decision-making.

Historical Context

Sodium-glucose cotransporter-2 (SGLT2) inhibitors have demonstrated robust cardiovascular and renal benefits in patients with established heart failure and chronic kidney disease. EMPACT-MI was designed to investigate whether early initiation of this class could prevent the development of heart failure following an acute myocardial infarction, a period of high vulnerability for incident heart failure, following previous null or inconclusive results in similar post-MI trials like DAPA-MI.

Guided Discussion

High-yield insights from every perspective

Med Student
Medical Student

SGLT2 inhibitors like empagliflozin are known for their 'metabolic' and 'hemodynamic' effects. In the context of an acute myocardial infarction (MI), which specific physiological mechanism is hypothesized to prevent the progression to heart failure, and why might this be independent of glucose-lowering effects?

Key Response

SGLT2 inhibitors promote osmotic diuresis and natriuresis, which reduces preload and ventricular wall tension. More importantly, they shift myocardial fuel metabolism from glucose to ketone bodies (which are more oxygen-efficient) and reduce myocardial sodium-hydrogen exchange. These mechanisms help mitigate adverse ventricular remodeling and sympathetic overactivity immediately following an ischemic injury, regardless of whether the patient has diabetes.

Resident
Resident

The EMPACT-MI trial enrolled patients with 'high-risk' myocardial infarction. Based on the study's inclusion criteria and outcomes, which specific clinical markers should a resident look for to identify patients who might still derive the most benefit from early SGLT2 inhibitor initiation, despite the neutral primary composite endpoint?

Key Response

While the primary composite endpoint was not significant, the trial showed a nominal reduction in heart failure hospitalizations (a 23% relative risk reduction). Residents should focus on patients with new-onset heart failure, a left ventricular ejection fraction <45%, or those requiring intravenous diuretics during the index admission, as these 'high-risk' features align with the population where the trend toward benefit was observed.

Fellow
Fellow

Compare the primary endpoint results of EMPACT-MI with the DAPA-MI trial. How do the differences in trial design (standard RCT vs. registry-based RCT) and endpoint selection (time-to-first-event vs. win-ratio) influence your interpretation of SGLT2 inhibitor efficacy in the post-MI setting?

Key Response

EMPACT-MI used a traditional time-to-first-event analysis for a hard composite (HFH or death) and was neutral (p=0.21). DAPA-MI used a 'win-ratio' for a broader cardiometabolic composite and was nominally 'positive' but did not show a reduction in hard clinical events like CV death. Together, they suggest that while SGLT2 inhibitors have clear benefits in chronic HF, their 'disease-modifying' effect on mortality immediately post-MI is less potent than established therapies like ACE inhibitors or Beta-blockers.

Attending
Attending

EMPACT-MI demonstrated a reduction in heart failure hospitalizations but no impact on all-cause mortality. In a value-based care model, does this evidence justify the routine addition of SGLT2 inhibitors to the standard 'post-MI cocktail' (DAPT, high-intensity statin, ACEi/ARB, Beta-blocker, MRA) during the index hospitalization?

Key Response

This is a point of clinical debate. The 'neutral' result for mortality makes it difficult to mandate SGLT2 inhibitors for every post-MI patient. However, reducing hospitalizations has significant quality-of-life and cost-saving implications. An attending should emphasize that while not 'practice-changing' for mortality, early initiation is safe and potentially beneficial for preventing the transition from 'at-risk' to 'clinical heart failure,' particularly in patients already having signs of congestion.

Scholarly Review

Critical appraisal through the lens of expert reviewers and guideline development

PhD
PhD

In EMPACT-MI, the primary endpoint was a composite of death or HF hospitalization. Discuss the statistical implications of the 'competing risk' of death and the 'dilution' effect of a neutral mortality component on a potentially sensitive morbidity component in a time-to-first-event analysis.

Key Response

In time-to-first-event models, if the drug has a strong effect on morbidity (HFH) but no effect on mortality, the composite hazard ratio is 'pulled' toward the null by the neutral mortality data. Furthermore, since patients who die cannot subsequently have an HF hospitalization, mortality acts as a competing risk. The PhD perspective would argue that a win-ratio or a joint model might have more accurately captured the benefit seen in the HFH component which was nominally significant when analyzed in isolation.

Journal Editor
Journal Editor

As a reviewer, if the authors of EMPACT-MI had emphasized the nominal p-value of 0.008 for heart failure hospitalization in the abstract despite the non-significant primary endpoint (p=0.21), what specific editorial concerns would you raise regarding 'multiplicity' and 'hierarchical testing'?

Key Response

A seasoned editor would flag this as 'cherry-picking.' Because the primary endpoint failed to reach statistical significance, according to the pre-specified hierarchical testing procedure, all subsequent p-values for secondary endpoints must be considered exploratory or hypothesis-generating only. Over-emphasizing a secondary benefit in the face of a neutral primary endpoint can lead to 'spin' and misrepresent the trial’s failure to meet its main objective.

Guideline Committee
Guideline Committee

Current AHA/ACC guidelines give a Class 1 recommendation for SGLT2 inhibitors in patients with HFrEF. Based on the EMPACT-MI results, is there sufficient evidence to elevate SGLT2 inhibitors to a Class 1 or 2a recommendation for the prevention of heart failure in post-MI patients without pre-existing HFrEF?

Key Response

The committee would likely maintain a conservative stance. Since EMPACT-MI failed its primary endpoint, a Class 1 recommendation for 'universal' post-MI use is unsupported. However, given the consistent signal for reduced HF hospitalizations across EMPACT-MI and DAPA-MI, a Class 2b (may be considered) recommendation might be appropriate for high-risk patients (e.g., LVEF <45% or congestion) to prevent HF progression, acknowledging the evidence is less robust than for established HFrEF.

Clinical Landscape

Noteworthy Related Trials

2015

EMPA-REG OUTCOME

n = 7,020 · NEJM

Tested

Empagliflozin

Population

T2DM patients with high cardiovascular risk

Comparator

Placebo

Endpoint

3-point MACE

Key result: Empagliflozin significantly reduced the rate of cardiovascular death, nonfatal myocardial infarction, or nonfatal stroke compared to placebo.
2020

EMPEROR-Reduced

n = 3,730 · NEJM

Tested

Empagliflozin

Population

Patients with chronic heart failure and reduced ejection fraction

Comparator

Placebo

Endpoint

Cardiovascular death or hospitalization for heart failure

Key result: Empagliflozin significantly reduced the risk of cardiovascular death or hospitalization for heart failure regardless of diabetes status.
2021

DAPA-MI

n = 4,017 · Lancet

Tested

Dapagliflozin

Population

Patients with acute myocardial infarction

Comparator

Placebo

Endpoint

Hierarchical composite of death, hospitalization for heart failure, and other clinical events

Key result: Dapagliflozin did not significantly improve the hierarchical primary outcome compared to placebo in patients after acute myocardial infarction.

Tailored to your role

Want this tailored to you?

Add your specialty or training stage to get role-specific takeaways and more questions.

Personalize this analysis