Potential Dapagliflozin Benefit Post-MI Is Not a 'Mandate'

PHILADELPHIA — Giving the SGLT-2 inhibitor dapagliflozin (Farxiga) to patients with acute myocardial infarction (MI) and impaired left ventricular systolic function but no diabetes or chronic heart failure significantly improved a composite of cardiovascular outcomes, a European registry-based randomized trial suggests.

Dr Stefan James presenting at AHA23 in Philadelphia

In presenting these results from the DAPA-MI trial, Stefan James, MD, of Uppsala University in Sweden, noted that patients randomly assigned to dapagliflozin 10 mg along with the standard of care had improved outcomes based on a composite of seven primary endpoints, which the trial described as the hierarchical "win ratio" composite outcomes, compared with patients randomized to placebo plus standard of care.

"The 'win ratio' tells us that there's a 34% higher likelihood of patients having a better cardiometabolic outcome with dapagliflozin vs placebo in terms of the seven components," James said in an interview. The win ratio was achieved in 32.9% of dapagliflozin patients vs 24.6% of placebo (P < .001).

James presented the results here today at the American Heart Association Scientific Sessions 2023, and they were published online simultaneously in NEJM Evidence.

Lower-Risk Patients 

DAPA-MI enrolled 4017 patients from the SWEDEHEART and Myocardial Ischemia National Audit Project registries in Sweden and the United Kingdom, randomly assigning patients to dapagliflozin 10 mg or placebo along with guideline-directed therapy for both groups.

Eligible patients were hemodynamically stable, had an acute MI within 10 days of enrollment, and impaired left ventricular systolic function or a Q-wave MI. Exclusion criteria included history of either type 1 or 2 diabetes, chronic heart failure, poor kidney function, or current treatment with an SGLT-2 inhibitor. Baseline demographic characteristics were similar between trial arms.

The hierarchical seven primary endpoints were:

• Death, with cardiovascular death ranked first followed by noncardiovascular death

• Hospitalization due to heart failure, with adjudicated first followed by investigator-reported HF

• Nonfatal myocardial infarction

• Atrial fibrillation/flutter event

• New diagnosis of type 2 diabetes

• New York Heart Association functional class at the last visit

• Drop in body weight of at least 5% at the last visit

The key secondary endpoint, James said, was the primary outcome minus the body weight component, with time to first occurrence of hospitalization for HF or cardiovascular death.

When the seventh factor, body weight decrease, was removed, the differential narrowed: 20.3% vs 16.9% (P = .015). When two or more variables were removed from the composite, the differences were not statistically significant.

For 11 secondary and exploratory outcomes, ranging from CV death or hospitalization for HF to all-cause hospitalization, the outcomes were similar in both the dapagliflozin and placebo groups across the board.

However, the dapagliflozin patients had about half the rate of developing diabetes compared with the placebo group: 2.1 % vs 3.9%.

Dr Stefan James

The trial initially used the composite of CV death and hospitalization for HF as the primary endpoint, but switched to the seven-item composite endpoint in February because the number of primary composite outcomes was substantially lower than anticipated, James said.

He acknowledged the study was underpowered for the low-risk population it enrolled. "But if you extended the trial to a larger population and enriched it with a higher-risk population you would probably see an effect," he said.

"The cardiometabolic benefit was consistent across all prespecified subgroups and there were no new safety concerns," James told the attendees. "Clinical event rates were low with no significant difference between randomized groups."

Not a Ringing Endorsement

But for invited discussant Stephen D. Wiviott, MD, a cardiologist at Brigham and Women's Hospital and Harvard Medical School in Boston, the DAPA-MI trial result isn't quite a ringing endorsement of SGLT-2 inhibition in these patients.

Dr Stephen Wiviott

"From my perspective, DAPA-MI does not suggest a new mandate to expand SGLT-2 inhibition to an isolated MI population without other SGLT-2 inhibitor indications," Wiviott told attendees. "But it does support the safety of its use among patients with acute coronary syndromes."

However, "these results do not indicate a lack of clinical benefit in patients with prior MI and any of those previously identified conditions — a history of diabetes, coronary heart failure or chronic kidney disease — where SGLT-2 inhibition remains a pillar of guideline-directed medical therapy," Wiviott said.

In an interview, Wiviott described the trial design as a "hybrid" in that it used a registry but then added, in his words, "some of the bells and whistles that we have with normal cardiovascular clinical trials." He further explained: "This is a nice combination of those two things, where they use that as part of the endpoint for the trial but they're able to add in some of the pieces that you would in a regular registration pathway trial."

The trial design could serve as a model for future pragmatic therapeutic trials in acute MI, he said, but he acknowledged that DAPA-MI was underpowered to discern many key outcomes.

"They anticipated they were going to have a rate of around 11% of events so they needed to enroll about 6000 people, but somewhere in the middle of the trial they saw the rate was 2.5%, not 11%, so they had to completely change the trial," he said of the DAPA-MI investigators.

But an appropriately powered study of SGLT-2 inhibition in this population would need about 28,000 patients, he said. "This would be an enormous trial to actually clinically power, so in my sense it's not going to happen," Wiviott said.

The DAPA-MI trial was sponsored by AstraZeneca. James disclosed relationships with AstraZeneca, Janssen, and Amgen. Wiviott disclosed relationships with Amgen, AstraZeneca, Janssen, Merck, Pfizer, Icon Clinical, Novo Nordisk, and Varian.

This article originally appeared on MDedge.com, part of the Medscape Professional Network.