SURPASS-CVOT: What Tirzepatide's Cardiovascular Trial Did and Did Not Prove

SURPASS-CVOT randomised 13,299 adults with type 2 diabetes and established atherosclerotic cardiovascular disease to tirzepatide or dulaglutide, following them for a median of four years. Tirzepatide was noninferior for the composite of cardiovascular death, myocardial infarction or stroke (12.2% versus 13.1%, p=0.003 for noninferiority) but did not meet superiority (p=0.09).

Key takeaways

Published in the New England Journal of Medicine in December 2025 (doi:10.1056/NEJMoa2505928).
13,299 participants, median four-year follow-up — one of the largest incretin outcome trials run.
Primary composite: 12.2% on tirzepatide versus 13.1% on dulaglutide.
Noninferiority was met (p=0.003). Superiority was NOT met (p=0.09).
An active-comparator design is a harder test than placebo control, which strengthens the result.

Key facts
Publication	N Engl J Med 2025;393:2409-2420
Identifiers	NCT04255433 · PMID 41406444 · doi:10.1056/NEJMoa2505928
Design	Randomised, double-blind, active-comparator, event-driven
Participants	13,299
Median follow-up	4 years
Result	Noninferior (p=0.003); superiority not met (p=0.09)
Funding	Eli Lilly and Company

Why is an active comparator a harder test?

Most cardiovascular outcome trials compare a drug against placebo. That answers whether the drug is better than nothing — a real question, but a modest bar when the comparator group receives no active therapy for the mechanism under study.

SURPASS-CVOT compared tirzepatide against dulaglutide, a GLP-1 receptor agonist that already carries its own cardiovascular indication. Beating nothing is one thing; matching or beating an established effective therapy is considerably harder. Noninferiority in that setting is a meaningful finding, because the comparator was not a straw man.

This also explains why the superiority result reads as it does. Against placebo, a drug this effective on weight and glycaemia would very likely have shown a large separation. Against an active agent already reducing events, the remaining headroom is small.

What is the difference between noninferior and superior?

Noninferiority means the trial ruled out the possibility that tirzepatide is meaningfully worse than the comparator, within a pre-specified margin. Superiority means the trial demonstrated it is actually better. These are different claims and they require different statistical thresholds set before any data is seen.

SURPASS-CVOT met the first and not the second. The numerical difference favoured tirzepatide — 12.2% versus 13.1% — but at p=0.09 that difference is within the range chance can produce. Reporting it as proof of superiority overstates what was found.

The honest reading is that tirzepatide's cardiovascular safety in this high-risk population is now established at a standard placebo trials cannot reach, and that it performs at least as well as an agent with an existing cardiovascular indication. That is a substantial result. It is not the same as proving it protects hearts better than the alternative.

What did the secondary findings show?

Reported secondary and exploratory analyses described larger reductions in blood pressure and lipids than the comparator, a mortality signal favouring tirzepatide, and subsequent analyses reporting reductions in major kidney events with slower eGFR decline. Secondary endpoints carry less inferential weight than the pre-specified primary, and they are best read as consistent supporting detail rather than as independent proof.

The renal findings in particular have attracted attention because dedicated renal outcome evidence for tirzepatide has been thinner than for some other agents. They remain secondary analyses of a trial designed around cardiovascular events.

What does this mean if you are taking it?

For someone with type 2 diabetes and established cardiovascular disease, the trial removes a genuine uncertainty: there is now large, long, randomised evidence that tirzepatide does not carry a cardiovascular penalty relative to an established comparator. Before December 2025 that had to be inferred.

For someone taking tirzepatide for weight management without diabetes or established cardiovascular disease, the trial does not describe your population and should not be quoted as if it does. Trial results apply to the people actually enrolled.

How does this fit the wider incretin evidence?

Cardiovascular outcome trials exist because surrogate endpoints have misled before. Drugs have improved HbA1c or lipid panels while failing to reduce — occasionally while increasing — the events patients actually care about. Regulators require outcome data precisely because the correlation between marker and outcome is not dependable.

Read alongside the rest of the programme, SURPASS-CVOT completes a picture rather than standing alone. SURMOUNT-1 and SURMOUNT-2 established weight effect by population. SURMOUNT-4 established what withdrawal costs. SURMOUNT-5 established the head-to-head position against semaglutide. SURPASS-CVOT establishes that the cardiovascular profile in a high-risk diabetes population holds up against an active comparator.

What remains genuinely open is long-horizon data. A median four years is substantial by trial standards and short by the standard of a chronic treatment someone may take for decades. Extension studies and post-marketing surveillance will carry that question, and they are methodologically weaker than the randomised evidence above.

Event rates for the pre-specified primary composite. The difference met noninferiority (p=0.003) but not superiority (p=0.09).

View chart data as a table

SURPASS-CVOT primary composite outcome at median 4 years
Group	Tirzepatide	Dulaglutide
Cardiovascular death, MI or stroke	12.2%	13.1%

Common SURPASS-CVOT claims against what was demonstrated
Claim you may see	Supported?	What the trial actually showed
Tirzepatide is cardiovascularly safe in high-risk T2D	Yes	Noninferior to an active comparator over a median 4 years
Tirzepatide beats dulaglutide for heart protection	No	Superiority was not met (p=0.09)
Tirzepatide cuts cardiovascular death	Not established as primary	Mortality signals came from secondary analysis
Applies to weight-management users without diabetes	No	Population was T2D with established ASCVD
Protects the kidneys	Secondary only	Reported in subsequent analyses, not the primary endpoint

Frequently asked questions

Did SURPASS-CVOT show tirzepatide protects the heart?

It showed noninferiority to dulaglutide, an agent with an existing cardiovascular indication. Superiority was not met (p=0.09).

How large was the trial?

13,299 participants with a median follow-up of about four years.

Why does the active comparator matter?

Because beating an established effective therapy is a much harder test than beating placebo.

Does this apply to Zepbound for weight loss?

No. The population was adults with type 2 diabetes and established atherosclerotic cardiovascular disease.

Who funded it?

Eli Lilly and Company, disclosed in the publication.

Sources

Every clinical claim above links to a primary source: an FDA record, a peer-reviewed publication with DOI or PMID, or a ClinicalTrials.gov registration. Where a figure could not be verified against a primary source, it is labelled rather than asserted.

Disclosure. GLP·Agonists earns Independence commissions when readers sign up through partner links, and NexLife is a current provider (no financial relationship). Outbound partner links are marked rel="nofollow noopener". Rankings follow our published methodology. See conflicts of interest. This article is information, not medical advice.