What is DAPA-HF?

Approximately 63 million people live with heart failure (HF) worldwide. In heart failure, the ability of the heart to fill and empty is compromised, causing shortness of breath, fatigue, swelling in the legs, and difficulty with activities of daily living.

DAPA-HF, a landmark clinical trial in heart failure carried out by a team led by Professor John McMurray, showed that taking dapagliflozin, a drug originally designed to treat diabetes, helped to reduce morbidity and mortality in individuals with heart failure with reduced ejection fraction, regardless of whether they had diabetes or not. Subsequently, the DELIVER trial showed that the benefits of dapagliflozin extend to patients with mildly reduced and preserved ejection fraction, i.e. this treatment is of benefit in all patients with heart failure.


DAPA-HF is a global trial, consisting of 4744 patients across 20 countries spanning Western and Eastern Europe, The Americas, and Asia-Pacific.

Individuals were randomised to one of two groups: dapagliflozin 10mg, or placebo, in addition to standard therapies.

Individuals were eligible to participate if they had symptomatic heart failure, and:

  • Left ventricular ejection fraction ≤40%
  • Elevated NT-proBNP:
    • ≥400 pg/ml, if hospitalised for heart failure in past 12 months.
    • ≥900 pg/ml, if atrial fibrillation/flutter present.
    • ≥600 pg/ml, otherwise.
  • eGFR ≥30 ml/min/1.73m2
  • Systolic blood pressure ≥95mmHg, and did not have symptoms of hypotension
  • Did not have type 1 diabetes

What did DAPA-HF show?

Dapagliflozin reduced death from cardiovascular causes by 18%, and reduced heart failure events (hospitalisation or visits requiring IV therapy) by 30%, compared to placebo, by 2 years. Dapagliflozin also reduced all-cause death by 17% by 2 years, compared to placebo.

The number needed to treat to prevent either 1 cardiovascular death or 1 heart failure event is 21 patients.

Dapagliflozin was well tolerated, and the rate of discontinuation due to adverse events was low (5%) and no different to placebo.

For further details, please see the full manuscript here.