Oral orforglipron clears its cardiovascular-safety bar against insulin in diabetes

Orforglipron, an oral non-peptide drug that mimics the gut hormone GLP-1, did not raise the rate of major cardiac events compared with insulin in people with type 2 diabetes at high cardiovascular risk, according to the phase 3 ACHIEVE-4 trial [s1]. The result clears a regulatory safety bar rather than demonstrating a heart benefit: the trial was designed to show the pill was not worse than insulin glargine, and that is what it showed [s1].

The distinction matters because the injected peptide GLP-1 drugs semaglutide and dulaglutide have shown cardiovascular benefit in dedicated outcome trials, but whether a small-molecule pill from the same class carries the same safety profile had not been tested [s1]. ACHIEVE-4 was funded by Eli Lilly and Company, which is developing orforglipron, and readers should weigh the finding with that sponsorship in mind [s1].

What the trial did

ACHIEVE-4 was an event-driven, open-label trial run at 317 sites across 16 countries and territories [s1]. It enrolled adults with type 2 diabetes, a body-mass index of 25 kg/m² or more and glycated haemoglobin (HbA1c) between 7.0% and 10.5% (53–91 mmol/mol), all of whom had established cardiovascular or chronic kidney disease [s1]. Participants were randomly assigned 1:1 to the maximum tolerated dose of oral orforglipron (up to a 36 mg capsule) or to titrated injectable insulin glargine, each once daily [s1][s2].

Between 1 May 2023 and 5 September 2024, 2749 participants were randomised — 1371 to orforglipron and 1378 to insulin glargine [s1]. Of those randomised, 1032 (38%) were female and 1717 (62.5%) male; their mean age was 63.1 years, mean HbA1c 8.2% and mean BMI 33 kg/m²; 2362 (85.9%) had established cardiovascular disease and 1033 (37.6%) had chronic kidney disease [s1]. That high burden of existing disease is what made the trial event-driven: it ran until enough cardiac events had accrued to judge the comparison, rather than for a fixed time [s1]. The primary endpoint was time to a four-component major adverse cardiovascular event (MACE-4): cardiovascular death, non-fatal myocardial infarction, non-fatal stroke, or hospitalisation for unstable angina [s1]. Non-inferiority would be declared only if the upper edge of the 95% confidence interval for the hazard ratio stayed below 1.8 [s1].

What it found

Over a median of two years, a primary MACE-4 event occurred in 57 (4.2%) of 1358 orforglipron participants and 67 (5.0%) of 1343 on insulin glargine, giving a hazard ratio of 0.84 (95% confidence interval 0.59–1.20) and meeting the non-inferiority threshold (p<0.0001 for non-inferiority) [s1]. The point estimate sits below 1, but the confidence interval runs from a possible 41% reduction to a 20% increase, so the trial cannot claim the pill prevents cardiac events — only that it did not make them more common at the margin tested [s1].

The safety texture differed sharply between the two drugs. Gastrointestinal adverse events — the signature of the GLP-1 class — were reported in 851 (62.1%) of 1371 orforglipron participants versus 193 (14.2%) of 1355 on insulin, and were the most common reason for stopping the pill [s1]. In the other direction, clinically significant or severe hypoglycaemia (glucose below 3 mmol/L, or 54 mg/dL) occurred in 93 (6.8%) of orforglipron participants against 260 (19.2%) of those on insulin [s1]. Of 62 deaths in the trial, 19 (1.4%) were in the orforglipron group and 43 (3.2%) in the insulin group; all but one were judged unrelated to treatment [s1].

How to read it

An active-comparator, non-inferiority design against insulin is a pragmatic way to show a new glucose-lowering drug is cardiovascularly safe, and that is the claim ACHIEVE-4 supports [s1]. It is not evidence that orforglipron protects the heart the way the injectable peptides do; a separate, longer superiority trial would be needed for that, and the two-year follow-up here is short for cardiovascular outcomes [s1]. The open-label design — patients and clinicians knew which drug was being used — is a further limit, though hard endpoints like death and myocardial infarction are less vulnerable to that than softer measures [s1].

For context, orforglipron's attraction is that it is a pill rather than an injection, which could widen access if regulators accept the safety package; the trade-off visible here is the high rate of gastrointestinal side-effects that drove discontinuations [s1]. The lower hypoglycaemia rate versus insulin is a genuine point in the pill's favour for this population [s1].

What to watch

The open questions are whether a dedicated outcomes trial will show cardiovascular benefit rather than mere safety, how the pill performs head-to-head against injectable GLP-1 drugs, and how the gastrointestinal tolerability looks beyond two years [s1]. This article describes research and is not medical advice; decisions about diabetes treatment belong with treating clinicians.

Sources

• Cardiovascular safety of orforglipron versus insulin glargine (ACHIEVE-4) — The Lancet, 30 September 2026
• ACHIEVE-4 trial registration, NCT05803421 — ClinicalTrials.gov