GLP-1 for HFpEF: STEP-HFpEF + SUMMIT Tirzepatide Trial Results

Heart failure with preserved ejection fraction (HFpEF) was, for most of the last two decades, the heart-failure phenotype with almost nothing that worked. The SGLT2 inhibitors changed the first part of that picture (EMPEROR-Preserved^[6] and DELIVER^[7]). STEP-HFpEF (Kosiborod 2023 NEJM^[1]) and SUMMIT (Packer 2025 NEJM^[3]) added a second pillar: meaningful symptom improvement and weight loss with GLP-1 receptor agonists in the obesity-driven HFpEF phenotype. This article walks through what each trial actually showed, what the secondary analyses added, and how the practical HFpEF + GLP-1 + SGLT2 stack now reads.

The honest summary

• STEP-HFpEF was positive on the KCCQ-CSS. Semaglutide 2.4 mg weekly produced a +16.6 point KCCQ-CSS change at week 52 vs +8.7 on placebo — a +7.8 point treatment effect that is roughly double the minimal-clinically-important-difference threshold (Kosiborod 2023^[1]).
• SUMMIT extended the signal to tirzepatide and to a hard endpoint. Tirzepatide reduced the composite of cardiovascular death or worsening heart failure events and improved KCCQ-CSS by ~6.9 points vs placebo, with ~13–14% weight loss (Packer 2025^[3]).
• STEP-HFpEF DM replicated the benefit in diabetics. The diabetic subset trial confirmed the KCCQ-CSS and weight benefits hold in HFpEF patients with type 2 diabetes (Kosiborod 2024^[2]).
• Imaging supports a mechanism beyond weight loss. The SUMMIT cardiac MRI substudy (Kramer 2025^[5]) showed reduced LV mass and paracardiac adipose tissue — the kind of structural change that plausibly drives the symptom response.
• Quadruple therapy is the new HFpEF target. ARNI or ACEI plus loop diuretic plus SGLT2 inhibitor plus GLP-1 receptor agonist is the cocktail the obesity-HFpEF patient is steadily moving toward.

HFpEF: the obesity-driven phenotype

HFpEF is defined by symptomatic heart failure with an ejection fraction at or above 50%. Roughly 80% of HFpEF patients are overweight or obese, and a large subset belong to what Obokata, Reddy, and Borlaug have characterized as the “obese HFpEF phenotype” — a syndrome driven by visceral and epicardial adiposity, inflammation, and impaired left-ventricular diastolic compliance (Koepp 2020^[8]). Epicardial adipose tissue physically compresses the heart and produces inflammatory cytokines that stiffen the myocardium; weight loss reduces that fat depot directly. The Reddy 2020 work on diuretic responsiveness^[9] further showed that the obese HFpEF phenotype has a paradoxical worsening of renal function during aggressive decongestion — an argument for upstream weight reduction rather than ever-more diuretic escalation.

STEP-HFpEF: the semaglutide trial

STEP-HFpEF (Kosiborod 2023 NEJM^[1]) randomized 529 adults with HFpEF (EF ≥ 45%) and a BMI ≥ 30 kg/m² to semaglutide 2.4 mg weekly or placebo for 52 weeks. The dual primary endpoints were change in KCCQ-CSS and change in body weight. Key results:

• KCCQ-CSS: +16.6 points semaglutide vs +8.7 placebo — a +7.8 point treatment effect.
• Body weight: −13.3% semaglutide vs −2.6% placebo — a −10.7 percentage point treatment effect.
• 6-minute walk distance: +21.5 m semaglutide vs +1.2 m placebo — a +20.3 m treatment effect.
• NT-proBNP: significantly lower in the semaglutide arm at week 52.
• CRP: significantly reduced — a marker that the anti-inflammatory mechanism is engaged.

The +7.8 KCCQ-CSS treatment effect is roughly twice the minimal-clinically-important-difference threshold of about 5 points used in the HFpEF literature, and is larger than the KCCQ effects from SGLT2 inhibitors in the same population. The 6-minute walk distance gain is also clinically meaningful and unusual — most HFpEF trials struggle to move that endpoint at all.

STEP-HFpEF DM: the diabetic subset

STEP-HFpEF DM (Kosiborod 2024 NEJM^[2]) randomized 616 adults with HFpEF, BMI ≥ 30, and type 2 diabetes to semaglutide 2.4 mg weekly or placebo. The directional results were similar to STEP-HFpEF: meaningful KCCQ-CSS improvement, meaningful weight loss, and improvement in 6-minute walk distance, although the magnitudes were slightly smaller than in the non-diabetic STEP-HFpEF. The trial removed the obvious objection that semaglutide's glucose-lowering effect might be the only mechanism — the benefit persists in patients already on diabetes therapy.

SUMMIT: the tirzepatide trial

SUMMIT (Packer 2025 NEJM^[3]) was the larger, event-driven trial: 731 adults with HFpEF (EF ≥ 50%) and BMI ≥ 30 randomized to tirzepatide (max 15 mg weekly) or placebo. SUMMIT used a hierarchical composite primary endpoint of cardiovascular death or worsening heart-failure events plus KCCQ-CSS change. Results:

• Primary composite: tirzepatide significantly reduced the rate of CV death or worsening HF events vs placebo.
• KCCQ-CSS: approximately +6.9 point treatment effect vs placebo.
• Body weight: −13.9% tirzepatide vs −2.2% placebo.
• 6-minute walk distance: +18.3 m improvement vs placebo.
• CRP: significantly reduced, replicating the STEP-HFpEF inflammation signal.

The Zile 2025 Circulation secondary analysis^[4] unpacked the clinical-trajectory data and confirmed the benefit was consistent across subgroups, with the largest effects in patients with the highest baseline NT-proBNP and the highest baseline body weight. The Kramer 2025 JACC CMR substudy^[5] added the imaging mechanism: reduced left-ventricular mass and reduced paracardiac adipose tissue on tirzepatide — a structural correlate to the symptom and event improvements.

Where SGLT2 inhibitors sit in this picture

EMPEROR-Preserved (Anker 2021 NEJM^[6]) randomized 5,988 HFpEF patients to empagliflozin 10 mg daily or placebo and reduced the composite of CV death or HF hospitalization by 21%. DELIVER (Solomon 2022 NEJM^[7]) replicated that pattern with dapagliflozin 10 mg in 6,263 patients with mildly reduced or preserved EF, with an 18% reduction in the primary composite. KCCQ-CSS effects were approximately +1 to +2 points in both trials — smaller in magnitude than STEP-HFpEF or SUMMIT, but with a much larger evidence base for hard endpoints (and lower cost, no injection).

The clinical interpretation is not GLP-1 versus SGLT2. The mechanisms are complementary: SGLT2 inhibitors lower intravascular volume, reduce preload, and improve cardiac energetics; GLP-1 receptor agonists drive substantial weight loss and reduce the inflammatory and structural drivers of the obesity-HFpEF phenotype. The right question for the obese HFpEF patient is which gets started first — not which to pick.

Magnitude: KCCQ-CSS treatment effect across HFpEF interventions

The practical HFpEF + GLP-1 stacking protocol

1. Baseline workup. Echocardiogram confirming EF ≥ 50%, NT-proBNP, basic metabolic panel, KCCQ-CSS, 6-minute walk distance, body weight, and BP. Document baseline loop-diuretic dose if any.
2. Start on guideline-directed medical therapy first. ARNI (sacubitril/valsartan) or ACEI, plus an SGLT2 inhibitor (empagliflozin or dapagliflozin), plus a loop diuretic if congested. This is the GDMT floor on which the GLP-1 is layered.
3. Add the GLP-1. Semaglutide 2.4 mg or tirzepatide titrated to the maximum tolerated dose, using the standard slow titration schedule. There is no documented interaction with ARNI, ACEI, or SGLT2.
4. Monitor BP and diuretic dose at each titration. Weight loss plus SGLT2 plus GLP-1 can drive a meaningful BP drop. Plan for a 10–20 mmHg systolic decline over 6 months and adjust antihypertensives accordingly. Loop diuretic may need to come down as epicardial and visceral adiposity regresses.
5. Track NT-proBNP every 3 months and KCCQ-CSS plus 6-minute walk every 6 months. Improvement in NT-proBNP from baseline is the most useful interim biomarker; meaningful KCCQ-CSS improvement should appear by month 6 if the protocol is going to work.
6. Watch for lean-mass loss and falls risk. HFpEF patients skew older. Combine the GLP-1 with the protein and resistance-training protocol used in any GLP-1 patient at sarcopenia risk; document grip strength at baseline and at 6 months.

Insurance, coverage, and the cardiology-vs-obesity-medicine route

Neither semaglutide 2.4 mg (Wegovy) nor tirzepatide (Zepbound) carries an FDA indication for HFpEF as of this writing. Wegovy does carry the post-SELECT cardiovascular-event-reduction indication for adults with established CVD and overweight or obesity, which captures many HFpEF patients with prior MI or revascularization. The SELECT subgroup analysis by baseline HbA1c (Lingvay 2024^[10]) showed the CV benefit was consistent regardless of glycemic status. Zepbound carries the SURMOUNT-OSA-derived obstructive-sleep-apnea indication, which many HFpEF patients also meet.

Practically, the prior-authorization route runs either through the existing FDA indication (obesity for both, CVD for Wegovy, OSA for Zepbound) rather than HFpEF itself. Cardiologists, obesity-medicine specialists, and primary-care prescribers are all writing these prescriptions today, and the inflection point from the SUMMIT NEJM publication is that HFpEF clinics are increasingly initiating a GLP-1 in-house rather than referring out.

Risks and what to watch for

The major HFpEF-specific risks of adding a GLP-1 are dehydration and lean-mass loss. Both semaglutide and tirzepatide can cause GI side effects that reduce fluid intake, and combining a GLP-1 with an SGLT2 inhibitor produces dual diuresis. The protocol above explicitly contemplates a downward adjustment of the loop diuretic at each titration step. Heart-rate impacts are minimal with both semaglutide and tirzepatide at standard doses. Hypoglycemia risk is low without background sulfonylurea or insulin therapy. Atrial fibrillation anticoagulation interactions are addressed separately in our GLP-1 anticoagulant interactions article.

Related research

• STEP-HFpEF semaglutide trial deep-dive — the trial protocol and pre-specified endpoints in full
• GLP-1 for HFpEF patient guide — the patient-facing version of this protocol
• Stacking GLP-1 with SGLT2 inhibitors — the mechanistic case for Jardiance or Farxiga on top of semaglutide or tirzepatide
• GLP-1 anticoagulant interactions — the AF + HF stacking question
• GLP-1 muscle-loss prevention protocol — the protein and resistance-training half of the stack for older HFpEF patients

Important disclaimer. This article is educational and does not constitute medical advice. HFpEF management requires coordinated care between cardiology and obesity medicine. Diuretic adjustments, BP medication changes, and GLP-1 initiation in a heart-failure patient should be supervised by a qualified clinician. The trials cited here excluded patients with very advanced HF (NYHA class IV) and patients on insulin pumps, and the protocol above is not a substitute for individualized assessment. PMIDs were verified live against the PubMed E-utilities API on 2026-05-29.

Last verified: 2026-05-29. Next review: every 12 months, or sooner if a new prospective HFpEF + GLP-1 trial or a head-to-head GLP-1 vs SGLT2 HFpEF outcomes trial is published.