How to Read the Retatrutide vs Semaglutide Endpoint Hierarchy in TRANSCEND-T2D-2

What Is an Endpoint Hierarchy and Why Does It Matter?

When Eli Lilly announced topline results from TRANSCEND-T2D-2 on 28 October 2026, the headline claims were that retatrutide (LY3437943) lowered A1C more than semaglutide 2 mg and produced roughly double the body-weight change at 80 weeks.[1] Our earlier reports cover the topline numbers and the trial design. This guide asks a narrower question that matters to anyone reading the eventual paper: which of those claims are statistically protected, and which are only descriptive?

A trial that tests many endpoints and doses would, left uncorrected, produce false positives by chance. Sponsors therefore fix a testing order in advance. An endpoint is tested only if the one before it succeeded, and the overall type I error rate is held at the nominal level, conventionally 5%. Claims outside this order may be real, but they carry no formal error-rate control. This is general trial methodology, not a finding of the trial itself.

What Did the Topline Say About Non-Inferiority and Superiority?

According to trade-press summaries of the release, TRANSCEND-T2D-2 was a randomised, open-label Phase 3 trial enrolling 1,298 adults with type 2 diabetes inadequately controlled on metformin with or without an SGLT2 inhibitor, with a mean baseline A1C of 8.3% and a primary endpoint of A1C change at week 80.[2][3] The registered record is NCT06260722.[4]

Reports describe the primary objective as non-inferiority of retatrutide 12 mg to semaglutide 2 mg on A1C, with superiority also reported for that dose (P < 0.001, controlled for type I error).[2] Both retatrutide doses were reported at about a 2.2 percentage-point A1C reduction against about 1.8 for semaglutide.[3] Body-weight change was reported at about 19.5% to 20.9% for retatrutide against 9.4% for semaglutide.[3]

Does Similar A1C Across Two Doses Mean Both Were Superior?

No, and this is the most common misreading of the topline. A pharmacist-written summary cautions that two retatrutide arms with the same numerical A1C change does not establish that both doses cleared the controlled superiority test.[3] Statistical significance depends on the confidence interval around each comparison and on where each dose sits in the testing order, not on the point estimate alone. Until the full tables appear, the defensible reading is that superiority was reported for the 12 mg dose, and that dose-level claims for 9 mg are unconfirmed.

The same discipline applies to the weight endpoint. Reports state both retatrutide doses were superior for body-weight change under family-wise error control.[3] That is a stronger statement than a simple difference in means, but it still comes from a press release rather than a published analysis plan.

How Does the Open-Label Design Affect Interpretation?

In an open-label trial, participants and investigators know the assigned treatment. Laboratory endpoints such as A1C are objective and relatively robust to this, whereas patient-reported symptoms and the decision to discontinue are more exposed to expectation effects. Lilly reported discontinuation due to adverse events of 4.8% (9 mg) and 7.9% (12 mg) with retatrutide against 2.7% with semaglutide 2 mg.[3] A topline percentage cannot show which events drove those figures; earlier adverse-event reporting is summarised in our retatrutide side-effects literature guide.

What Should Researchers Check When the Full Paper Appears?

Five items are worth checking. First, the pre-specified testing order and whether the 9 mg dose was included in it. Second, the estimand: a treatment-regimen estimand (all randomised participants, regardless of adherence) usually yields smaller effects than an efficacy estimand (on treatment). Third, handling of missing data after discontinuation. Fourth, the confidence intervals, not only P values. Fifth, whether the registry record matches the reported primary endpoint, a standard check against outcome switching.[4]

Lilly has said detailed results will be presented at a future medical meeting and published in a peer-reviewed journal.[1] For how an earlier retatrutide readout moved from press release to journal, see the TRIUMPH-2 Lancet publication.

Is Retatrutide Approved?

No. Retatrutide is investigational, with no marketing authorisation from the FDA, MHRA or any regulator; Lilly has described a planned US submission for the first quarter of 2027.[5] Nothing in this guide describes a use of any compound.

How Does Licensed Pharma Development Differ From Research-Use Compounds?

TRANSCEND-T2D-2 is a regulated clinical programme: a licensed pharmaceutical sponsor, a registered protocol, ethics oversight and medical supervision of participants. Its results do not transfer to a laboratory reagent. Velox Peptides supplies retatrutide solely as an HPLC-verified in vitro research reagent and reports trial news only as third-party information for researchers. For specifications see the research-grade retatrutide page, and for receptor pharmacology see the mechanism of action guide.

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

Frequently Asked Questions