Lilly’s 28 October announcement reports greater blood-sugar reduction and weight loss with investigational retatrutide than with injectable Ozempic 2 mg in an 80-week trial of adults with type 2 diabetes. The result is a specific head-to-head comparison—not a comparison with every semaglutide treatment, and not an approval or prescribing recommendation. Sponsor’s topline announcement
The announcement also describes the conditions behind its estimates and higher adverse-event discontinuation rates in the retatrutide groups. Those details belong beside the headline weight-loss result.
Who was compared, and with what?
TRANSCEND-T2D-2 is a randomized, open-label phase 3 trial. Lilly describes 1,298 adults whose type 2 diabetes was inadequately controlled with metformin, with or without an SGLT2 inhibitor. The trial compared two retatrutide groups with Ozempic 2 mg, all as once-weekly injections.
“Open-label” means treatment assignment was not concealed in the way it would be in a blinded comparison. The population also matters: these were adults with type 2 diabetes receiving background treatment, not a general sample of everyone seeking weight loss.
Ozempic is the named comparator. The study should not be retitled as a comparison with Wegovy, a different semaglutide product and treatment context. Our oral Wegovy explainer addresses that separate product question.
Read the outcome and its units together
Lilly reports the following changes at 80 weeks. The doses identify the research groups; they are not instructions for use.
| Sponsor-reported measure | Retatrutide 9 mg | Retatrutide 12 mg | Ozempic 2 mg |
|---|---|---|---|
| Average HbA1c change from a baseline of 8.3% | −2.2 percentage points | −2.2 percentage points | −1.8 percentage points |
| Average body-weight change | −19.5% | −20.9% | −9.4% |
| Stopped treatment because of adverse events | 4.8% | 7.9% | 2.7% |
The reported HbA1c changes differ by approximately 0.4 percentage points, using the rounded figures. The body-weight reductions differ by 10.1 and 11.5 percentage points from the comparator. Those are differences between group averages, not promises about an individual’s result. Results, safety summary and statistical footnotes
The statistical footnotes also differ: Lilly describes the 9 mg HbA1c superiority comparison as nominally significant, while the 12 mg superiority comparison was controlled for multiple testing. A single statement that every comparison had the same statistical treatment would lose that distinction.
The estimate assumes a particular treatment course
The release presents an efficacy estimand: an estimate of what would happen if all randomized participants remained on their assigned intervention through 80 weeks, allowing interruptions or modifications, without starting additional blood-sugar medication for more than 14 days.
That is a defined research question. It is different from simply asking what happened to everyone regardless of stopping treatment or adding another medicine. The reported discontinuations show why the distinction is material: tolerability and persistence are part of the practical treatment picture, even when an efficacy estimate describes a hypothetical continued-treatment course.
The full report will be needed to examine the analyses and uncertainty in greater detail. The announcement says detailed results will be presented at a future meeting and published in a peer-reviewed journal; this account does not treat that future report as already inspected.
A trial result and medicine availability are separate decisions
The FDA says retatrutide is not a component of an FDA-approved medicine, has not been found safe and effective for any condition, and cannot be used in compounding under US federal law. Products marketed online under its name therefore cannot borrow approval from an approved GLP-1 medicine. FDA’s account of unapproved GLP-1 products
The meaningful development is a sponsor-reported comparison with a specified treatment, population and analysis. Approval, a full published appraisal and the choice of treatment for an individual remain different questions.