Summary: In a pooled secondary analysis of the STEP-HFpEF and STEP-HFpEF DM randomised trials (n=1,145; 45% with a history of atrial fibrillation), once-weekly semaglutide 2.4 mg was associated with larger gains in heart-failure symptoms in participants with atrial fibrillation than without (KCCQ-CSS 11.5 vs 4.3 points; P interaction = 0.001). Reductions in weight, C-reactive protein and NT-proBNP were observed irrespective of atrial fibrillation status.
PICO Summary
| Element | Detail |
|---|---|
| Population | 1,145 adults with obesity-related HFpEF (LVEF ≥45%, BMI ≥30 kg/m2, KCCQ-CSS <90), pooled from two international double-blind RCTs; 518 (45%) had a history of atrial fibrillation (40% paroxysmal, 24% persistent, 35% permanent) and 627 (55%) did not. |
| Intervention | Once-weekly subcutaneous semaglutide 2.4 mg for 52 weeks, randomised 1:1 (pooled semaglutide arm across both trials). |
| Comparison | Matching placebo, once weekly for 52 weeks (pooled placebo arm); subgroups defined by investigator-reported baseline atrial fibrillation history. |
| Outcome | KCCQ-CSS improvement (semaglutide vs placebo): 11.5 points (95% CI 8.3 to 14.8) with atrial fibrillation vs 4.3 points (95% CI 1.3 to 7.2) without; P interaction = 0.001. Hierarchical composite win ratio: 2.25 (95% CI 1.79 to 2.83) with atrial fibrillation vs 1.30 (95% CI 1.06 to 1.59) without; P interaction < 0.001. C-reactive protein, NT-proBNP and body weight were reduced regardless of atrial fibrillation status (all P interaction not significant). Serious adverse events and serious cardiac disorders were fewer with semaglutide than placebo across subgroups. No absolute risk reduction or NNT was reported for this analysis. |
Semaglutide in obesity-related HFpEF, by atrial fibrillation
Pooled RCT analysis · obesity-related HFpEF · 52 weeks
Once-weekly semaglutide 2.4 mg improved HFpEF symptoms more in patients with baseline atrial fibrillation than without, with a higher hierarchical composite win ratio. The contrast is a subgroup interaction, not a randomised comparison.
Expert Commentary
This is a secondary, pooled analysis of two randomised trials, and atrial fibrillation status was examined as an effect modifier rather than a randomised comparison, so the subgroup contrast should be read as associational rather than causal. Within that frame, the signal is consistent: semaglutide-treated participants gained more symptomatic benefit when atrial fibrillation was present, and the interaction term for the patient-reported and composite endpoints reached statistical significance. Reductions in weight, C-reactive protein and NT-proBNP appeared uniform across subgroups, which is reassuring for the proposed weight-loss and anti-inflammatory mechanism. The principal limitation is that baseline atrial fibrillation was investigator-reported and not adjudicated, and that participants with atrial fibrillation differed systematically at baseline (older, more often male, higher NT-proBNP, more advanced symptoms), so residual confounding of the apparent modifier effect cannot be excluded. The trials were funded by the manufacturer of semaglutide, and several authors are company employees, which warrants the usual caution around an industry-sponsored subgroup finding. Can I use this with my patients? Yes, in a measured way: for an ambulatory adult with obesity-related HFpEF and coexisting atrial fibrillation who is already a candidate for semaglutide, these data support an expectation of meaningful symptom relief, while recognising the analysis does not establish atrial fibrillation as a treatment-selection criterion. A prospectively powered, adjudicated evaluation would help confirm whether atrial fibrillation truly identifies higher responders.
References
Verma S, Butler J, Borlaug BA, et al. Atrial Fibrillation and Semaglutide Effects in Obesity-Related Heart Failure With Preserved Ejection Fraction: STEP-HFpEF Program. J Am Coll Cardiol. 2024;84(17):1603-1614. doi:10.1016/j.jacc.2024.08.023
