SURPASS-CVOT Heart Failure Analysis

An exploratory analysis of SURPASS-CVOT found lower point estimates for several cardiovascular and heart-failure outcomes with tirzepatide than with dulaglutide. This analysis came from a randomized trial, but it was not a separate trial designed to prove that tirzepatide works better specifically in people with heart failure. Every participant had type 2 diabetes and established atherosclerotic cardiovascular disease. A history of heart failure was recorded for 20.3% of the trial population^[1].

Among participants with a history of heart failure, the hazard ratio for all-cause death or a heart-failure hospitalization or urgent visit was 0.83 for tirzepatide versus dulaglutide (95% confidence interval 0.70 to 0.99). The corresponding estimate among those without a heart-failure history was 0.88 (0.78 to 1.00)^[2]. These are relative time-to-event estimates. The abstract does not report absolute event counts for either subgroup, so it cannot show how many events were avoided per 100 people.

How SURPASS-CVOT defined the comparison

The parent SURPASS-CVOT study was a phase 3, double-blind, active-comparator trial. It randomized 13,299 people at 640 sites in 30 countries to weekly tirzepatide or dulaglutide. After 134 ineligible participants were excluded, the modified intention-to-treat analysis included 13,165 people. Treatment lasted about four years on average^[3]. The registry identifies Eli Lilly and Company as the sponsor^[4].

Dulaglutide was an active comparator with established cardiovascular benefit, not a placebo. The parent trial's primary outcome combined cardiovascular death, myocardial infarction, and stroke, a measure called MACE-3. Events occurred in 12.2% of the tirzepatide group and 13.1% of the dulaglutide group. The hazard ratio was 0.92 (95.3% confidence interval 0.83 to 1.01). Tirzepatide met the prespecified noninferiority standard, but did not meet the superiority threshold^[3].

The new paper re-examined outcomes according to whether participants had a history of heart failure. That history marked a higher-risk group: those participants had more coronary artery disease and prior myocardial infarction, and event categories occurred more often. The abstract does not define heart-failure subtype or report ejection fraction. Readers should not treat this as a trial confined to heart failure with preserved or reduced ejection fraction.

What the heart-failure subgroup analysis found

For cardiovascular death or a heart-failure hospitalization or urgent visit, the hazard ratio was 0.85 (0.70 to 1.03) among participants with prior heart failure. It was 0.96 (0.83 to 1.12) among those without that history^[2]. Both confidence intervals crossed 1.0. The estimates are compatible with a benefit, but they also leave uncertainty about whether the observed differences reflect a real treatment effect.

MACE-3 results were similar across the two subgroups. The hazard ratio was 0.90 (0.76 to 1.08) with a heart-failure history and 0.93 (0.83 to 1.04) without one. For MACE-4, which added coronary revascularization to cardiovascular death, myocardial infarction, and stroke, the estimates were 0.81 (0.69 to 0.95) and 0.91 (0.83 to 1.00), respectively^[1]. The MACE-4 interval in the prior-heart-failure subgroup did not cross 1.0, but that does not turn an exploratory secondary comparison into a new primary trial result.

Patient-reported APPADL and EQ-5D-5L visual analogue scores improved and were higher with tirzepatide in both heart-failure-history groups^[2]. The abstract does not provide the absolute score changes or the amount of missing questionnaire data. It therefore supports the direction of the group comparison, but not a claim about how large or noticeable the difference would be for an individual.

Why the findings remain exploratory

Randomization strengthens comparisons between tirzepatide and dulaglutide within the parent trial. It does not make every later subgroup analysis confirmatory. Once investigators split a trial by baseline history and inspect several composite outcomes, the number of comparisons grows. Some confidence intervals can exclude 1.0 by chance. The paper itself calls this heart-failure analysis exploratory, while the parent trial established only overall noninferiority for MACE-3.

The two subgroup point estimates also should not be compared by eye. A hazard ratio of 0.83 in one group and 0.88 in another does not establish that prior heart failure changes the treatment response. That question requires a formal interaction analysis with enough statistical power. The available abstract does not report an interaction result, and it gives no subgroup-specific absolute risks.

A separate post hoc SURPASS-CVOT analysis examined a broader six-component cardiovascular and kidney composite across the full modified intention-to-treat population. It reported an event in 23.7% of tirzepatide-treated participants and 27.4% of dulaglutide-treated participants, with a hazard ratio of 0.84^[3]. That result provides context for the wider trial, but it cannot supply the missing heart-failure subgroup event counts in the newer paper.

What this means for patients

The useful reading is narrow. In people who already had type 2 diabetes and atherosclerotic cardiovascular disease, outcomes generally favoured tirzepatide or were similar to dulaglutide whether or not heart failure appeared in the medical history. The result does not establish benefit in people with heart failure who lack diabetes or atherosclerotic disease. It also does not show that one treatment is the right choice for a particular patient. Our tirzepatide evidence overview provides the broader trial context.

Dulaglutide remains an important comparator because its cardiovascular benefit was already established. The result is therefore different from a placebo-controlled finding. Treatment choice also involves glucose control, adverse effects, contraindications, access, and personal priorities. None of those decisions can be made from this subgroup analysis alone.

What we do not yet know

The abstract leaves several details unresolved. It does not report absolute subgroup event rates, interaction tests, heart-failure subtype, ejection fraction, or changes in background heart-failure treatment. Full reporting is needed to judge whether follow-up and missing data differed by subgroup. The analysis also cannot tell us whether the observed pattern applies beyond the trial's population of adults with both type 2 diabetes and established atherosclerotic cardiovascular disease.

The parent trial reported more gastrointestinal adverse events with tirzepatide, although overall adverse-event incidence appeared similar between groups^[3]. The new abstract does not provide safety results by heart-failure history. A favourable relative estimate for a composite outcome should not be read as a complete benefit-risk assessment or a reason to change treatment without the clinician managing both conditions.

Medical disclaimer: This article is for educational and journalistic purposes only and does not constitute medical advice. Peptides discussed may be classified as prescription medicines or research chemicals depending on your jurisdiction. Always consult a qualified healthcare professional before using any peptide product. PeptideMethods.com does not sell, distribute, or facilitate the sale of any peptide product.