Bajaj HS, Welch M, Shah P and colleagues published TRANSCEND-T2D-1 in The Lancet on 6 June 2026, simultaneously with a symposium at the American Diabetes Association Scientific Sessions in New Orleans.[1][2] The paper is the first peer-reviewed Phase 3 type 2 diabetes readout for retatrutide (LY3437943), an investigational once-weekly peptide that agonises the receptors for glucose-dependent insulinotropic polypeptide (GIP), glucagon-like peptide-1 (GLP-1) and glucagon. Under the treatment-regimen estimand, mean HbA1c fell by 1.94 percentage points and mean body weight by 15.3% at the 12 mg assigned target dose, versus 0.81 points and 2.6% with placebo, at week 40.[1]

These are trial results in a defined research population. They are not a licensed indication. Retatrutide is not authorised by the MHRA. It is not a UK licensed medicine, and nothing in this article is consumption, dosing or self-administration advice.

1. What was published

TRANSCEND-T2D-1 (NCT06354660) is a completed 40-week, multicentre, randomised, double-blind, placebo-controlled Phase 3 trial funded by Eli Lilly and Company.[1][3] It is the first of three global registrational TRANSCEND-T2D studies. Unlike the later TRIUMPH-2 and TRIUMPH-3 obesity readouts already covered on this site, which remain sponsor-disclosed topline figures, TRANSCEND-T2D-1 is now available as a full Lancet paper with both estimands, safety tables and a stated limitation on generalisability.[1][4]

A March 2026 company release had already disclosed topline superiority on the primary and key secondary endpoints.[5] The Lancet publication and ADA presentation supply the peer-reviewed treatment-regimen numbers, confidence intervals, completion rates and the investigators' own caveats.

2. Trial design

The trial ran at 48 sites in the United States, Mexico and India.[1][2] Adults aged 18 years or older were eligible if type 2 diabetes remained inadequately controlled on diet and exercise alone, with HbA1c between 7.0% and 9.5% (53–80 mmol/mol) and body-mass index of at least 23 kg/m². Participants were not to have taken antihyperglycaemic medicines for at least 90 days before visit 1 and were to be naïve to insulin other than gestational use.[1][5]

Of 930 people assessed for eligibility, 537 were randomly assigned 1:1:1:1 to once-weekly subcutaneous retatrutide at assigned target doses of 4 mg (n = 134), 9 mg (n = 133) or 12 mg (n = 136), or placebo (n = 134).[1] Protocol titration started at 2 mg once weekly and increased every four weeks to the assigned target (one step to 4 mg; steps at 2 mg, 4 mg and 6 mg to 9 mg; steps at 2 mg, 4 mg, 6 mg and 9 mg to 12 mg).[5] Those titration steps are protocol facts from a registered trial. They are not instructions for use.

Baseline means were age 48.8 years (SD 12.1), HbA1c 7.9% (SD 1.1), diabetes duration 2.5 years (SD 4.4), BMI 35.8 kg/m² (SD 7.0) and body weight 96.9 kg. Fifty-five per cent of participants were female. Most (85%) had no prior antihyperglycaemic medicine use.[1][6] Four hundred and ninety participants (91%) completed the treatment period on study drug and 504 (94%) completed the study. Thirty-seven (9%) assigned to retatrutide and ten (7%) assigned to placebo discontinued the study intervention early.[1]

The primary endpoint was change in HbA1c from baseline to week 40. A key secondary endpoint was percentage change in body weight over the same interval.[1]

3. Glycaemic results

Under the treatment-regimen estimand — the estimated average effect regardless of adherence or later antihyperglycaemic rescue — mean change in HbA1c at week 40 was:[1][6]

  • retatrutide 4 mg: −1.69% (SE 0.11)
  • retatrutide 9 mg: −1.86% (SE 0.10)
  • retatrutide 12 mg: −1.94% (SE 0.08)
  • placebo: −0.81% (SE 0.12)

Estimated treatment differences versus placebo were −0.88 percentage points (95% CI −1.18 to −0.59) at 4 mg, −1.04 (−1.32 to −0.76) at 9 mg, and −1.12 (−1.39 to −0.85) at 12 mg; all p < 0.0001.[1][2]

An HbA1c below 7.0% (53 mmol/mol) was reached by 82–89% of participants assigned to retatrutide versus 52% assigned to placebo. An HbA1c of 6.5% (48 mmol/mol) or lower was reached by 75–83% versus 40%. An HbA1c below 5.7% (39 mmol/mol) was reached by 35% (4 mg; not significant versus placebo), 40% (9 mg) and 37% (12 mg), versus 14% with placebo.[6]

The investigators state that the glycaemic changes reflected improvements in both fasting and postprandial hyperglycaemia.[6] No participant in any group reported severe hypoglycaemia.[1]

Lilly's March 2026 topline, which used the efficacy estimand (the effect had all randomised participants remained on assigned intervention without initiating additional antihyperglycaemic medicines), reported mean HbA1c reductions of 1.7%, 2.0% and 1.9% at 4 mg, 9 mg and 12 mg versus 0.8% with placebo.[5] Those figures are slightly more favourable at the two higher doses than the treatment-regimen numbers now in The Lancet. Both estimands met the pre-specified superiority claim. They answer different questions, and they should not be pooled.

4. Body weight and cardiometabolic markers

Under the same treatment-regimen estimand, mean percentage change in body weight from the 96.9 kg baseline was −11.5% (SE 0.7) at 4 mg, −13.9% (SE 0.8) at 9 mg and −15.3% (SE 0.8) at 12 mg, versus −2.6% (SE 0.5) with placebo.[1] The estimated treatment difference versus placebo at 12 mg was −12.7 percentage points (95% CI −14.4 to −11.0; p < 0.0001), corresponding to a mean 15.1 kg reduction at that assigned dose.[2] Weight change had not plateaued by week 40 in the 9 mg and 12 mg groups.[1][6]

The efficacy-estimand weight figures in the March topline were higher at the two larger assigned doses: −11.5% (−11.1 kg) at 4 mg, −15.5% (−15.1 kg) at 9 mg and −16.8% (−16.6 kg) at 12 mg, versus −2.5% with placebo.[5] Again, that estimand assumes continued assigned intervention. The Lancet treatment-regimen figures are the more conservative, and the more relevant, published result.

Associated marker changes at 40 weeks included reductions of up to 39.6% in triglycerides, 19.8% in non-HDL cholesterol, 6.4 mmHg in systolic blood pressure and 12.4 cm in waist circumference.[7] These are within-trial biomarker shifts. They are not evidence that retatrutide prevents myocardial infarction, stroke or cardiovascular death. TRANSCEND-T2D-1 was not a cardiovascular outcomes trial.

The 15.3% mean reduction at 12 mg sits below the 20.8% reported at the same assigned dose and a longer (80-week) time point in TRIUMPH-2, which enrolled adults with type 2 diabetes and obesity or overweight on background glucose-lowering therapy, and well below the 28.3% reported in TRIUMPH-1, which enrolled adults with obesity or overweight without diabetes.[4][8] Cross-trial comparisons are limited by duration, background therapy, baseline weight and estimand. The pattern is, however, consistent with the established incretin literature: percent weight change in type 2 diabetes is typically smaller than in non-diabetic obesity cohorts. TRANSCEND-T2D-1 was placebo-controlled. It does not rank retatrutide against tirzepatide or semaglutide.

5. Safety

The investigators describe an adverse-event profile consistent with molecules that have GLP-1 agonist activity.[1] The most frequent events were gastrointestinal, generally mild to moderate, and occurred mainly during titration.[5][6] In the March disclosure, nausea occurred in 16.4%, 19.5% and 26.5% at 4 mg, 9 mg and 12 mg versus 3.7% with placebo; diarrhoea in 18.7%, 26.3% and 22.8% versus 4.5%; and vomiting in 15.7%, 15.0% and 17.6% versus 2.2%.[5] Dysesthesia was reported in 4.5%, 2.3% and 4.4% versus 0.0%; events were generally mild and most resolved during treatment.[5]

Discontinuation because of adverse events was 2.2% (4 mg), 4.5% (9 mg) and 5.1% (12 mg) versus 0.0% with placebo.[1][5] Two deaths occurred, both in the 4 mg group; both were judged unrelated to study drug.[1][2]

At the ADA briefing, investigators and discussants also flagged urinary-tract infection as a finding not previously emphasised in the incretin class: 0.7–2.9% with retatrutide versus 0.0% with placebo in TRANSCEND-T2D-1, mostly in women. A hydration-related mechanism was offered as a hypothesis, not a demonstrated cause.[7] That signal needs confirmation in the remaining TRANSCEND and TRIUMPH programme. It is not, on present evidence, a characterised risk.

6. How to read the estimands

Two estimands appear across the public record, and they are easy to conflate.

The treatment-regimen estimand estimates the average effect of assignment, regardless of whether participants stayed on the study intervention or started additional antihyperglycaemic medicines. It is the set of HbA1c and weight figures in the Lancet findings.[1][5]

The efficacy estimand estimates the effect had all randomised participants remained on assigned intervention for 40 weeks without initiating additional antihyperglycaemic medicines for more than 14 days. It is the set used for the headline 2.0% HbA1c and 16.8% weight figures in Lilly's March release.[5]

Neither estimand converts a research finding into a licensed use. Neither supports comparison with SURPASS, SURMOUNT, STEP or SUSTAIN without a dedicated active-comparator trial. TRANSCEND-T2D-2, which the authors note is underway, is designed as an 80-week comparison with semaglutide.[2] Those results are not yet reported.

7. UK regulatory position

Retatrutide has no UK marketing authorisation. It is not listed in the British National Formulary. It is not the subject of a published NICE technology appraisal. The MHRA has authorised other incretin-class medicines — including injectable and, as of June 2026, oral semaglutide for specified weight-management criteria, and tirzepatide (Mounjaro) — as prescription-only medicines under the Human Medicines Regulations 2012.[9] Those licences do not extend to retatrutide.

An investigational peptide offered for sale as “research retatrutide” is not the licensed product under study in TRANSCEND-T2D-1. It has not been assessed by the MHRA for quality, safety or efficacy. This publication does not change that position. Under the Medicines Act 1968 and the Human Medicines Regulations 2012, a substance presented for the treatment or prevention of disease is a medicinal product and requires a marketing authorisation unless a narrow exemption applies. Research-use-only supply remains a labelling and presentation question, not a therapeutic pathway.

Lilly has previously stated that retatrutide cannot be legally sold or marketed for human use while it remains investigational, and has said it intends a US Biologics License Application after completing the chemistry, manufacturing and controls package — a first-quarter 2027 intention disclosed with the later TRIUMPH-2 and TRIUMPH-3 topline.[4] That is a US-jurisdiction filing plan. It is not an MHRA decision and not a NICE appraisal.

8. What these data do not show

TRANSCEND-T2D-1 enrolled a largely medicine-naïve cohort with short diabetes duration. The authors themselves state that this “reduces the generalisability of the study findings to the broader population of people with type 2 diabetes who might be on prescription antihyperglycaemic medications or insulin.”[6] Results in people already on metformin, SGLT2 inhibitors, insulin or other incretin-class medicines are the subject of other TRANSCEND trials, not this one.

The trial does not show that retatrutide treats, cures or prevents type 2 diabetes, obesity, cardiovascular disease or any other condition. It reports 40-week changes in HbA1c, body weight and selected biomarkers against placebo in a research population. Weight change had not plateaued; longer follow-up is required to describe the trajectory. There is no active comparator in this dataset. Cardiovascular outcome data are absent. Durability after stopping assigned treatment is not reported here.

Unlicensed products sold to UK buyers under the name retatrutide are not equivalent to the investigational medicinal product used in TRANSCEND-T2D-1. Identity, purity, sterility and dose accuracy of those products are outside the scope of this trial and have not been established by the MHRA.

This article is AI-researched and editorially reviewed. It is provided for research and educational purposes only and is not medical advice. Research peptides are not licensed for human consumption in the UK.