SURPASS-CVOT: Tirzepatide Cardiovascular Outcomes (2026)

SURPASS-CVOT is the trial that finally answered whether tirzepatide — the dual GIP/GLP-1 drug sold as Mounjaro and Zepbound — protects the heart in people who already have cardiovascular disease. The headline, published in the New England Journal of Medicine on December 18, 2025: among 13,299 adults with type 2 diabetes and established atherosclerotic disease, the rate of major adverse cardiovascular events (cardiovascular death, heart attack, or stroke) was 12.2% on tirzepatide versus 13.1% on dulaglutide, a hazard ratio of 0.92 (95.3% CI 0.83–1.01). Tirzepatide met its goal of non-inferiority (P=0.003) but missed statistical superiority (P=0.09) (Nicholls 2025 ^[1]). The crucial twist that makes this trial unusual — and easy to misread — is that it was not compared against placebo. It was tested head-to-head against dulaglutide (Trulicity), a GLP-1 drug already proven to cut cardiovascular events. This article walks through exactly what that design means, every verified number, and why "non-inferior, not superior" is a more reassuring result than it sounds. For the placebo-controlled GLP-1 heart trial, see the SELECT trial of semaglutide.

The honest summary

• Primary result: non-inferior, not superior. Three-point MACE (CV death, MI, stroke) hit 12.2% on tirzepatide vs 13.1% on dulaglutide — HR 0.92 (95.3% CI 0.83–1.01); P=0.003 for non-inferiority, P=0.09 for superiority (Nicholls 2025 ^[1]). Tirzepatide was clearly not worse than an established heart-protective drug; it did not formally beat it.
• The comparator was an active drug, not placebo. Dulaglutide had already reduced cardiovascular events in REWIND, so SURPASS-CVOT set a high bar by design — any benefit had to show on top of a drug that already works (Nicholls 2024, design paper ^[2]).
• Every individual component pointed the right way. Cardiovascular death, MI, and stroke each numerically favored tirzepatide, though none was a formal superiority claim once the primary missed.
• A broader cardiorenal composite did separate the drugs. A pre-specified/post hoc 6-component endpoint (adding heart-failure hospitalization, revascularization, and kidney events) occurred in 23.7% vs 27.4% — HR 0.84 (95% CI 0.79–0.90; P<0.001) favoring tirzepatide (Nissen 2026, JAMA Cardiol ^[3]).
• Kidney events fell 23%. The pre-specified kidney composite occurred in 6.0% vs 7.6% — HR 0.77 (95% CI 0.68–0.88; P=0.0002) (Zoungas 2026, Lancet Diabetes Endocrinol ^[4]).
• Metabolic edge was large. Tirzepatide cut HbA1c by ~1.66% vs 0.88% and body weight by ~11.6% vs 4.5% — far more than dulaglutide (Nicholls 2025 ^[1]; tctmd 2025 ^[5]).
• An imputed-placebo analysis suggests a real benefit. Bridging through REWIND, tirzepatide vs an inferred placebo was tied to lower MACE (HR 0.72) and all-cause death (HR 0.61) — exploratory, not a head-to-head result (Sattar 2026 ^[6]).

Why this trial was designed against dulaglutide, not placebo

Almost every cardiovascular outcomes trial (CVOT) you read about — SELECT for semaglutide, LEADER for liraglutide, REWIND for dulaglutide itself — compares the drug against a placebo on top of standard care. SURPASS-CVOT did something different and, ethically, more demanding: it compared tirzepatide against dulaglutide 1.5 mg (Trulicity), a GLP-1 receptor agonist that the REWIND trial had already shown reduces major cardiovascular events in people with type 2 diabetes. Tirzepatide was dosed up to 15 mg weekly; both were given as once-weekly subcutaneous injections (Nicholls 2025 ^[1]; Nicholls 2024 ^[2]).

Why not placebo? Because by the time SURPASS-CVOT launched (first patients randomized in 2020), it was widely considered unethical to give a placebo to high-risk diabetes patients with established cardiovascular disease when a proven cardioprotective drug like dulaglutide existed. Giving half of a 13,000-patient trial an inert injection for four years, when an effective option was on the shelf, would have withheld a known benefit. So Eli Lilly — which funded the trial — chose an active comparator instead (Nicholls 2024 ^[2]).

Who was in the trial

SURPASS-CVOT enrolled adults aged 40 or older with type 2 diabetes and established atherosclerotic cardiovascular disease — a secondary-prevention population, meaning everyone already had heart disease, not just risk factors. A total of 13,299 patients were randomized 1:1; 134 were later excluded for not meeting entry criteria, leaving a modified intention-to-treat population of 13,165 (6,586 on tirzepatide, 6,579 on dulaglutide). They were enrolled at 640 sites across 30 countries between May 2020 and June 2022 (Nicholls 2025 ^[1]; Zoungas 2026 ^[4]).

The population was meaningfully sick at baseline: mean age 64.1 years, 29.0% women, mean body-mass index 32.6, mean HbA1c 8.4%, and mean diabetes duration 14.7 years. Roughly a third had microalbuminuria and about 22% had reduced kidney function (eGFR under 60), which is why the kidney sub-analysis below carries weight (Nicholls 2025 ^[1]; Zoungas 2026 ^[4]). The median follow-up was about 4 years — long enough to accumulate the events needed for a CVOT.

The primary MACE result — a detailed walkthrough

The primary endpoint was three-point MACE: a composite of death from cardiovascular causes, nonfatal or fatal myocardial infarction, or nonfatal or fatal stroke, counting the first such event per patient. A primary event occurred in 801 patients (12.2%) on tirzepatide and 862 patients (13.1%) on dulaglutide, giving a hazard ratio of 0.92 (95.3% confidence interval 0.83 to 1.01). The non-inferiority test cleared its margin with P=0.003; the superiority test returned P=0.09 (Nicholls 2025 ^[1]).

Translate that into plain terms. Over four years, the absolute MACE rate was about 0.9 percentage points lower on tirzepatide — an 8% relative reduction that did not reach the statistical threshold for declaring superiority. But because the comparator was a drug that already lowers cardiovascular risk, "matching dulaglutide" is itself a meaningful, reassuring outcome: it tells clinicians that switching a patient to tirzepatide for its larger weight and glucose effects does not cost them cardiovascular protection.

The individual components all leaned in tirzepatide's favor — cardiovascular death, myocardial infarction, and stroke each occurred numerically less often on tirzepatide than dulaglutide. However, once the primary composite missed formal superiority, these component differences are best read as consistent directional support, not independent proof; in a hierarchical testing scheme, you cannot claim significance on downstream endpoints after the top of the hierarchy fails to cross the superiority line. The takeaway is consistency, not a string of positive p-values (Nicholls 2025 ^[1]).

Where tirzepatide DID pull ahead — the broader composites

The three-point MACE story is not the whole picture. A post hoc analysis published in JAMA Cardiology in 2026 examined a broader six-component cardiorenal composite — all-cause mortality, MI, stroke, coronary revascularization, hospitalization for heart failure, and a composite of adverse kidney outcomes. On this wider endpoint the drugs separated clearly: events occurred in 1,559 tirzepatide patients (23.7%) versus 1,803 dulaglutide patients (27.4%), HR 0.84 (95% CI 0.79–0.90; P<0.001). Narrower five- and four-component versions (dropping kidney and then heart-failure endpoints) returned the same HR of about 0.86 (Nissen 2026 ^[3]).

The expanded MACE-style endpoint that adds coronary revascularization also favored tirzepatide, with a hazard ratio in the 0.88 range (Nicholls 2025 ^[1]; tctmd 2025 ^[5]). The pattern across these wider composites is consistent: as you broaden the net to capture heart-failure, revascularization, and kidney events, tirzepatide's advantage over dulaglutide becomes statistically detectable even though the narrow three-point MACE did not formally cross the superiority line.

All-cause mortality and the imputed-placebo question

Because SURPASS-CVOT had no placebo arm, it cannot directly tell you how much tirzepatide reduces death or MACE compared with no GLP-1 therapy at all — only how it stacks up against dulaglutide. To estimate the "missing" placebo comparison, the investigators ran a pre-specified imputed-placebo analysis: they multiplied the tirzepatide-vs-dulaglutide hazard ratio from SURPASS-CVOT by the dulaglutide-vs-placebo hazard ratio from the earlier REWIND trial, using propensity-score matching to align the two populations (Sattar 2026 ^[6]).

That indirect comparison suggested tirzepatide versus an inferred placebo was associated with lower three-point MACE (HR 0.72, 95% CI 0.55–0.94) and notably lower all-cause mortality (HR 0.61, 95% CI 0.45–0.82), plus fewer cardiovascular-death-or-heart-failure events (HR 0.70). These are exploratory, model-based estimates — not a randomized head-to-head against placebo — so they should be read as supportive context rather than proof. But they line tirzepatide up with the magnitude of benefit seen in placebo-controlled GLP-1 CVOTs (Sattar 2026 ^[6]). For the directly randomized placebo comparison in a different population, see SELECT.

Metabolic and safety differences

On the metabolic endpoints there was no contest. Tirzepatide produced substantially larger reductions in HbA1c (about −1.66% vs −0.88%) and body weight (about −11.6% vs −4.5%) than dulaglutide, consistent with tirzepatide's far larger effect on glucose and weight in the SURPASS and SURMOUNT diabetes/obesity programs (Nicholls 2025 ^[1]; tctmd 2025 ^[5]). For the head-to-head weight comparison against semaglutide, see tirzepatide vs semaglutide.

On safety, the overall adverse-event picture was broadly similar between groups, with the expected GLP-1/GIP signature: gastrointestinal adverse events (nausea, vomiting, diarrhea) were more common with tirzepatide — about 42.5% vs 35.9% in the cardiorenal analysis — and discontinuations for adverse events ran somewhat higher on tirzepatide (roughly 13.2% vs 10.1%) (Nissen 2026 ^[3]; tctmd 2025 ^[5]). None of this was unexpected given tirzepatide's stronger appetite and gut effects. For the cardiac-rhythm question specifically, see GLP-1 drugs and heart rate.

What this changes — and what it doesn't

• Tirzepatide is cardiovascularly safe in established heart disease. The core regulatory question for any diabetes drug — does it raise cardiovascular risk? — is answered with a firm no. It matched a proven cardioprotective comparator (Nicholls 2025 ^[1]).
• "No superiority" is not "no benefit." The trial was built to beat an active drug, not placebo. Across broader cardiorenal composites and the imputed-placebo analysis, the direction is consistently favorable (Nissen 2026 ^[3]; Sattar 2026 ^[6]).
• Kidney protection is a real, pre-specified win. A 23% reduction in major kidney events strengthens the case for tirzepatide in patients with diabetic kidney disease (Zoungas 2026 ^[4]).
• Choosing tirzepatide over an older GLP-1 doesn't cost heart protection. For a patient who needs the larger glucose and weight effects, SURPASS-CVOT removes the worry that the trade is cardiovascular safety.
• It does not prove tirzepatide beats semaglutide for the heart. SURPASS-CVOT compared tirzepatide with dulaglutide, not semaglutide; cross-trial comparisons to SELECT are not apples-to-apples (different drugs, populations, and comparators).

Bottom line

SURPASS-CVOT showed that tirzepatide is non-inferior — but not formally superior — to dulaglutide for three-point MACE in people with type 2 diabetes and established cardiovascular disease (12.2% vs 13.1%; HR 0.92, 95.3% CI 0.83–1.01; P=0.003 non-inferiority, P=0.09 superiority) (Nicholls 2025 ^[1]). Because the comparator was an active, already-proven cardioprotective drug rather than placebo, matching it is a reassuring result, not a disappointing one — and tirzepatide pulled clearly ahead on a broader cardiorenal composite (HR 0.84), on major kidney events (HR 0.77), and on glucose and weight, with an imputed-placebo analysis suggesting a substantial mortality benefit versus no treatment (Nissen 2026 ^[3]; Zoungas 2026 ^[4]; Sattar 2026 ^[6]). For patients who need tirzepatide's stronger metabolic effects, the trial confirms they don't have to give up cardiovascular protection to get them.

This article is educational and is not medical advice. Every trial result above is sourced to the peer-reviewed SURPASS-CVOT publications and analyses indexed in PubMed, verified against the live PubMed database before publication. Treatment decisions for diabetes or cardiovascular disease should be made with your prescriber.