Retatrutide TRIUMPH-2 in The Lancet: What the Safety and Tolerability Data Show

What Did the TRIUMPH-2 Lancet Paper Publish?

Eli Lilly announced the detailed TRIUMPH-2 results on 29 September 2026, with the data presented at the 62nd Annual Meeting of the European Association for the Study of Diabetes (EASD) in Milan and published simultaneously in The Lancet as “Retatrutide in adults with obesity and type 2 diabetes (TRIUMPH-2): a double-blind, parallel-group, randomised, placebo-controlled, phase 3 trial.”[1][2] Our overview of the Lancet paper covers design and endpoints; our earlier EASD preview flagged that the full dataset would add safety and subgroup detail beyond the 23 July topline release. This article covers that safety and tolerability layer, which is the part of a Phase 3 paper most useful to researchers modelling the pharmacology of retatrutide (LY3437943).

Method note: the full text of the paper was not accessible to us at the time of writing, so the adverse-event figures below are taken from third-party reporting on the publication (Medscape) and Lilly's release. Readers should confirm figures against the primary paper.

How Was TRIUMPH-2 Designed?

A three-arm design against placebo matters for safety interpretation: it lets reviewers ask whether adverse-event rates rise across arms in a monotonic way, which is the pattern one would expect if an event is driven by exposure to the compound. As the next section shows, some events followed that pattern and one did not.

What Were the Adverse-Event Discontinuation Rates?

As reported, discontinuation due to adverse events was 3.8% in the 4 mg arm, 11.6% in the 9 mg arm and 7.7% in the 12 mg arm, compared with 4.9% for placebo.[3] The non-monotonic result — the highest rate in the middle arm — is a useful reminder that discontinuation reflects a mix of causes (event type, titration scheme, site practice, participant factors) and not exposure alone. Whether the paper offers an explanation is something to check in the full text and its supplementary tables.

It also helps to keep populations apart: these figures come from a type 2 diabetes population, and Lilly's other Phase 3 readouts, such as TRIUMPH-1, enrolled different participants and should not be compared figure-for-figure.

Which Adverse Events Were Most Frequent?

Gastrointestinal events were the most frequently reported and were more common in the retatrutide arms than with placebo, consistent with the incretin class.[3]

• Diarrhoea: 27%, 34% and 34% (4, 9 and 12 mg arms) versus 13% for placebo.
• Nausea: 14%, 21% and 28% versus 8% for placebo.
• Hypotension: 1–6% across arms versus under 1% for placebo.
• Severe hypoglycaemia: no episodes reported.

Nausea rose steadily across arms, while diarrhoea plateaued between the 9 and 12 mg arms. For researchers comparing multi-agonist mechanisms, our mechanism guide explains why a GIP/GLP-1/glucagon receptor profile might produce a different event pattern from a GLP-1-only compound; the trial data alone cannot isolate the contribution of each receptor.

What Did TRIUMPH-2 Report on Dysesthesia?

Dysesthesia — an abnormal skin sensation often described as burning — and related events were reported in 4%, 6% and 7% of the 4, 9 and 12 mg arms, versus 1% for placebo. Events were generally mild and rarely led to discontinuation.[3] Unlike discontinuation, this signal rose in step with the arm, which is the pattern that makes it a priority for mechanistic follow-up. Commentary quoted in the reporting suggested it may relate to sensory neural signalling rather than weight change or nutritional factors, and that the phenomenon has been seen with other GLP-1-based agents; that is commentary, not a finding of the trial. Our dysesthesia explainer covers the earlier TRIUMPH-1 reporting and open questions.

What Are the Limits of These Numbers?

Three cautions apply. First, adverse-event percentages from a single trial in a defined population do not generalise to other populations or formulations. Second, the figures here are secondary reporting and should be checked against the paper's tables, which also define how events were coded and grouped. Third, retatrutide remains an investigational compound: Lilly's stated plan is a Biologics License Application to the FDA by Q1 2027, and it holds no marketing authorisation from the FDA, MHRA or any regulator.[1] See our FDA approval timeline guide for the regulatory picture. This is reporting on licensed-pharma development; it says nothing about research-grade material.

What Should UK Researchers Take From This?

The publication changes nothing about retatrutide's UK status: it is an unlicensed, investigational compound, and nothing in this article is a therapeutic or health claim. Velox Peptides supplies retatrutide strictly as an HPLC-verified in vitro research reagent, with no health or therapeutic claims made for it.

This compound is supplied as a research reagent only. It is not a medicine and has not been evaluated by the MHRA or FDA for use in our products. Not for human or veterinary use. See our Research Use Policy and MHRA Statement.

Frequently Asked Questions