STEP-HFpEF: Semaglutide for HFpEF

Background — HFpEF and obesity

Heart failure with preserved ejection fraction (HFpEF, EF ≥50%) accounts for approximately half of all heart failure cases. Unlike HFrEF (reduced EF), HFpEF has had few effective pharmacologic therapies until recently. SGLT2 inhibitors (per EMPEROR-Preserved, DELIVER) showed cardiovascular benefit; sacubitril-valsartan showed mixed results.

Obesity is a major contributor to HFpEF — approximately 80% of HFpEF patients are overweight or obese. The obesity-related HFpEF phenotype involves increased plasma volume, pericardial restraint, pulmonary hypertension, and metabolic inflammation. Weight loss in the 5% to 10% range improves HFpEF symptoms in lifestyle and bariatric surgery trials².

Trial design

STEP-HFpEF (NCT04788511) enrolled 529 adults with HFpEF (LVEF ≥45%), NYHA functional class II-IV symptoms, BMI ≥30 kg/m², KCCQ-CSS <90, and 6-minute walk distance ≥100 meters. Participants with type 2 diabetes were excluded (T2D was subsequently studied in STEP-HFpEF-DM).

Randomization was 1:1 to semaglutide 2.4 mg weekly (titrated over 16 weeks) or placebo. Co-primary endpoints were change from baseline in KCCQ-CSS and percentage change in body weight at week 52. Secondary endpoints included 6-minute walk distance, NT-proBNP, CRP, and a hierarchical composite of death, heart failure events, and changes in KCCQ-CSS and 6MWT.

Primary outcomes

Mean change in KCCQ-CSS from baseline to week 52 was +16.6 points with semaglutide versus +8.7 points with placebo (treatment difference 7.8 points; 95% CI 4.8-10.9; P<0.001)¹. A KCCQ-CSS improvement of ≥5 points is considered clinically meaningful; the magnitude here is well above that threshold.

Mean weight loss was 13.3% with semaglutide versus 2.6% with placebo (treatment difference 10.7%; P<0.001). The magnitude of weight loss is similar to STEP-1 in non-cardiac populations.

Six-minute walk distance increased by 21.5 meters with semaglutide versus 1.2 meters with placebo. NT-proBNP decreased by 20% with semaglutide versus 6% with placebo. CRP decreased by 43% versus 7%, suggesting substantial anti-inflammatory effects.

Hierarchical composite outcome

A pre-specified hierarchical composite of (1) all-cause death, (2) heart failure events, (3) change in KCCQ-CSS, and (4) change in 6-minute walk distance, analyzed via the win ratio method, favored semaglutide with a win ratio of 1.72 (95% CI 1.37-2.15)¹. The benefit was driven primarily by symptom and exercise improvements rather than hard endpoint reduction, though heart failure events occurred in fewer semaglutide participants (1.1% vs 4.7%).

STEP-HFpEF-DM (T2D extension)

STEP-HFpEF-DM evaluated the same intervention in adults with HFpEF, obesity, AND type 2 diabetes (n=616). Results published in 2024 mirrored the original STEP-HFpEF findings: KCCQ-CSS improvement of 13.7 points with semaglutide versus 6.4 with placebo (difference 7.3; P<0.001)³.

Combined STEP-HFpEF and STEP-HFpEF-DM data establish semaglutide as evidence-based therapy for the obesity-related HFpEF phenotype regardless of diabetes status.

Mechanism in HFpEF

Multiple mechanisms likely contribute. Weight loss reduces plasma volume, pericardial restraint, and pulmonary congestion. Anti-inflammatory effects (large CRP decrease) may reduce myocardial inflammation. Improved insulin sensitivity may reduce diabetic cardiomyopathy contribution. Direct GLP-1 receptor effects on cardiac tissue (receptors present on atrial and ventricular myocytes) may contribute, though the magnitude is debated⁴.

Notably, semaglutide did NOT improve LVEF or diastolic function parameters substantially — the symptom benefit appears to come more from systemic/peripheral changes than from intrinsic cardiac remodeling. This contrasts with sacubitril-valsartan and possibly SGLT2 inhibitors, which appear to affect cardiac remodeling more directly.

Clinical implications

STEP-HFpEF places semaglutide in the HFpEF guideline-directed therapy conversation. The 2024 AHA/ACC heart failure guideline focused update mentions semaglutide as a consideration in obese HFpEF patients. Direct comparison with SGLT2 inhibitors (the established HFpEF therapy) has not been performed; the two classes likely have complementary mechanisms.

For patients with HFpEF and obesity, the choice between semaglutide and SGLT2 inhibitor (or both) depends on additional indications (diabetes, CKD, established CVD), cost, and tolerability. Many cardiologists now use both classes together in this phenotype.

Related editorial coverage

• SELECT trial: CV outcomes in obesity
• Semaglutide peptide profile

Frequently asked questions

Does semaglutide help heart failure?

STEP-HFpEF showed that semaglutide 2.4 mg weekly improved symptoms and exercise capacity in adults with HFpEF (heart failure with preserved EF) and obesity. The KCCQ-CSS improved by 7.8 points more with semaglutide than placebo¹. This effect appears specific to the obesity-related HFpEF phenotype; benefit in HFrEF (reduced EF) is unproven.

Should everyone with heart failure be on semaglutide?

No. STEP-HFpEF specifically enrolled patients with HFpEF (preserved EF ≥45%) and obesity. Patients with HFrEF, non-obese patients, or those without symptomatic HF were not studied. Semaglutide for heart failure should be considered in the obesity-HFpEF phenotype specifically.

Can I combine semaglutide with an SGLT2 inhibitor for heart failure?

Yes, and this is increasingly common in obese HFpEF patients. SGLT2 inhibitors (empagliflozin per EMPEROR-Preserved, dapagliflozin per DELIVER) have demonstrated cardiovascular benefit in HFpEF. Combining the two classes is supported by mechanism and increasingly by retrospective data.

What is HFpEF and is it the same as 'heart failure'?

HFpEF (heart failure with preserved ejection fraction) is a form of heart failure in which the heart's pumping function is normal (EF ≥50%) but the heart cannot adequately relax to fill or pump blood. Symptoms (shortness of breath, fluid retention, exercise intolerance) overlap with HFrEF (reduced EF), but the underlying pathophysiology and treatment differ substantially.