GLP-1 drugs

Lilly Wins the First Head-to-Head Against Ozempic

The result moves the competition from weight loss to glycemic control, and it leaves the filing clock as the next thing to watch.

By James T. · Updated October 28, 2026 · 8 min read

Scroll down for medical verification information & sources.

Eli Lilly reported on October 28 that its experimental drug retatrutide beat Novo Nordisk's Ozempic on blood sugar and weight in a Phase 3 trial of adults with type 2 diabetes. The study, TRANSCEND-T2D-2, is the first Phase 3 head-to-head comparison between retatrutide and another incretin drug. Lilly said the trial met its primary endpoint.

The topline numbers are strong. At 80 weeks, retatrutide lowered A1C by 2.2 percentage points at both the 9 mg and 12 mg doses, against 1.8 points for semaglutide 2 mg, the highest approved dose of Ozempic for diabetes. Weight fell 19.5 percent at 9 mg and 20.9 percent at 12 mg, against 9.4 percent for semaglutide. The company said both retatrutide comparisons were statistically significant.

The trial enrolled 1,298 adults whose diabetes was not controlled on metformin, with or without an SGLT2 inhibitor. Mean baseline A1C was 8.3 percent, mean weight was 225.8 pounds, and mean body mass index was 36.8. Patients were randomized 2:2:1 to weekly retatrutide 9 mg, retatrutide 12 mg, or semaglutide 2 mg for 80 weeks.

The trial was built to test noninferiority

The primary endpoint was not superiority. It was noninferiority of retatrutide 12 mg against semaglutide on A1C reduction. That is the question a developer asks when it needs to show its drug is at least as good as the incumbent before it argues about anything else. Retatrutide 12 mg met that endpoint and then showed superiority, with a p-value below 0.001.

The 9 mg dose also met noninferiority as a key secondary endpoint. Its superiority finding was nominal, which means it fell outside the trial's family-wise error control. In plain terms, the 9 mg result is suggestive rather than confirmatory. Lilly will lead with the 12 mg arm, and the 9 mg arm will serve as a dose-response argument.

That structure matters for how the result should be read. The trial was designed to show that retatrutide is not worse than Ozempic. It succeeded at that, and it also produced a superiority result at the higher dose. The difference in A1C, 0.4 percentage points, is modest in absolute terms, but it was enough to satisfy both the noninferiority test and the superiority test at 12 mg.

The numbers separate the two drugs

The weight difference is the more dramatic part of the result. A 20.9 percent reduction at 12 mg is more than double the 9.4 percent seen with semaglutide. In pounds, the 12 mg arm lost an average of 47.4 pounds against 20.9 pounds for semaglutide. The 9 mg arm lost 44.3 pounds.

Retatrutide also produced greater improvements in non-HDL cholesterol, triglycerides, and systolic blood pressure. Those are secondary endpoints, and secondary endpoints in an open-label trial carry less weight than the primary result. They do suggest that the drug's glucagon component is doing something beyond weight loss, which is the mechanism Lilly has been selling since the first Phase 2 data.

The trial was powered for the A1C comparison, and the weight result is a secondary endpoint that the company reported with the same precision. That ordering matters for how regulators will read the data. A weight advantage that is consistent across doses and large in magnitude is persuasive, but it is not the endpoint the trial was designed to prove, and the Food and Drug Administration will look at the primary analysis first.

The A1C result is the one that matters for a diabetes label. Weight loss is what the market pays for, but glycemic control is what the indication requires. Lilly now has a Phase 3 result showing its drug beat the standard of care on the endpoint regulators care about most, in a population that was not achieving control on first-line therapy.

Weight loss is the headline, A1C is the claim

The obesity market has been organized around weight for several years. Patients ask for the drug that produces the largest percentage loss, and payers ask whether the loss is worth the price. A 20 percent reduction moves retatrutide into a different category from the first-generation incretins, and it matches the range that Lilly's own tirzepatide has produced in obesity trials.

The diabetes market is organized differently. A1C is the measure that determines whether a drug gets approved and what label it receives. A 0.4 point advantage over semaglutide is not a cure, but it is a meaningful difference in a disease where many patients cycle through therapies without reaching target. Lilly can now argue that retatrutide is both a better weight-loss drug and a better glucose-lowering drug than the incumbent.

That combination is what makes the result commercially important. A drug that only matched semaglutide on A1C but doubled the weight loss would still be a strong obesity product. A drug that beats it on both measures can be positioned as the next standard of care, with a price to match.

The open-label design limits what the trial can prove

TRANSCEND-T2D-2 was open-label. Patients and investigators knew which drug each participant received. That design is common in head-to-head trials of injectable drugs, because blinding a weekly injection with different devices and titration schedules is difficult. It is also a limitation that regulators and payers will weigh.

A1C and weight are objective measurements, so the open-label design is less of a problem for the primary endpoint. Tolerability is different. Patients who know they are receiving a new drug may report side effects differently, and investigators who know the assignment may manage them differently. The discontinuation rates are therefore not a clean comparison, even though they are the best data available.

The results also use the efficacy estimand, which estimates the treatment effect as if every patient had stayed on the assigned drug. That is a legitimate way to describe how well a drug works when it is taken, but it is not the same as the effect in the population that starts treatment. The trial's full data, including the treatment-policy estimand, will be presented at a medical meeting and published in a journal.

Tolerability is the price of the extra receptor

Retatrutide is a triple agonist. It activates the GLP-1, GIP, and glucagon receptors at once. Semaglutide activates only GLP-1, and tirzepatide activates GLP-1 and GIP. The extra glucagon activity is the source of the additional weight loss, and it is also the likely source of the additional side effects.

Gastrointestinal events were the most common adverse events in the trial. Diarrhea, nausea, constipation, vomiting, and dyspepsia were reported more often with retatrutide, and most were mild or moderate and resolved while treatment continued. Discontinuation due to adverse events was 4.8 percent at 9 mg, 7.9 percent at 12 mg, and 2.7 percent with semaglutide.

The pattern is familiar from the class. GLP-1 receptor agonists slow gastric emptying, which produces nausea and vomiting, and the effect is strongest during dose escalation. Clinicians manage it by starting low and titrating slowly. Retatrutide's trial used a titration schedule, and Lilly said most events were mild or moderate. Whether the schedule is enough to keep patients on the 12 mg dose outside a trial is a question the open-label data cannot answer.

That last comparison is the one payers will notice. The 12 mg dose, which produced the best efficacy, also produced nearly three times the discontinuation rate of the comparator. A drug that is more effective on paper but harder to stay on may produce less benefit in practice. Lilly will need to show that the tolerability profile is manageable with titration and support, not only that the efficacy is superior.

Kenneth Custer, president of Lilly Cardiometabolic Health, said retatrutide could shift the treatment paradigm for people living with type 2 diabetes. That is the company's claim, and the trial supports it better than any previous data point. It does not answer whether patients will stay on the drug long enough to realize the benefit.

The next question is the filing

Lilly has not published a submission timeline with the topline results. The company said detailed data will come at a future medical meeting and in a peer-reviewed journal. Until those data are available, the trial's effect size and its safety profile rest on a company summary.

The regulatory path will depend on what the full dataset shows. The Food and Drug Administration will look at the primary endpoint, the safety database, and the durability of the effect. It will also look at whether the open-label design and the efficacy estimand are adequate to support the claims Lilly wants on the label. A superiority claim over semaglutide is a higher bar than a noninferiority claim, and the 12 mg arm has the data to support it.

The two indications are also regulated differently. A diabetes approval would position retatrutide against Ozempic and the other GLP-1 drugs on A1C and cardiovascular safety. An obesity approval would position it against Wegovy and Zepbound on percentage weight loss, where the bar is higher and payer scrutiny is more intense. The trial supports the diabetes filing first, and the weight data will travel with it.

The commercial path depends on price and access. Analysts already expect retatrutide to be a large product. Leerink's David Risinger wrote that the results strengthen the case for the drug's long-term differentiation and justify a branded premium before semaglutide's patent expires in 2032. He projects $15 billion in annual sales by then, above the $10 billion consensus.

That forecast assumes the drug is approved, reimbursed, and tolerated. The trial answers the efficacy question. The open-label design and the discontinuation gap leave the real-world question open, and the filing will be the next signal of whether Lilly can convert a Phase 3 win into a label that payers will accept.

For now, the result changes the terms of the competition. The first head-to-head trial between two incretin drugs produced a clear winner on the endpoints that matter, with a tolerability cost that the company will have to manage. The next round of data will show whether the advantage holds when the trial is not open-label and when patients are not selected for their willingness to stay on treatment.

Primary sources

Eli Lilly, topline results from TRANSCEND-T2D-2, October 28, 2026. investor.lilly.com
Healio, report on the head-to-head results, October 28, 2026. healio.com
AJMC, analysis of the TRANSCEND-T2D-2 data, October 28, 2026. ajmc.com
BioPharma Dive, report on the trial and analyst reaction, October 28, 2026. biopharmadive.com

Written by James T.

Do you disagree with any of the medical statements in this article? Email [email protected] with your feedback.

This is general information, not medical advice. Talk to your care team before making any change to your treatment.

Medicare

Medicare's AI Prior Authorization Tied to a 15 Percent Drop in Injections

A JAMA study of the WISeR pilot in six states found epidural steroid injections fell about 15 percent and other targeted services about 20 percent, while the authors caution the data cannot show whether the care was necessary.

By James T.·October 28, 2026·8 min read

Health IT

The Federal Health IT Checklist Loses 14 Criteria

The HTI-5 final rule removes 14 of 60 certification criteria, drops the AI model card requirement, and projects $1.55 billion in savings.

By James T.·October 28, 2026·8 min read

Health fraud

The $2 Billion Health Fraud Case Began With the Billing Outliers

Fifteen defendants, six medical professionals, and schemes that paid for codes rather than care. The cases show the payment system's own data becoming the detection system.

By James T.·October 23, 2026·8 min read

Drug Safety

The Compounded GLP-1 Market Outlived the Shortage That Created It

A small tirzepatide recall points to a larger problem: the compounding market that grew during the GLP-1 shortage is still selling, even though the shortage that justified it has largely ended.

By James T.·October 21, 2026·8 min read

Enforcement

A Shuttered Ambulance Company Still Owes for Its Old Data

Massachusetts reached a $1 million settlement with Transformative Healthcare over alleged ambulance upcoding and a 2023 breach that exposed the data of 750,116 residents, years after the company stopped operating.

By James T.·October 20, 2026·8 min read

gene therapy

Hearing Loss Gene Therapy's Hardest Hurdle Is Now the Price Tag

Sensorion's SENS-601 is cleared for trials in France against the most common genetic cause of deafness. The biology is arriving on schedule; the payment models are not.

By James T.·October 9, 2026·8 min read

Medicare

The First Check From Medicare's Never-Used $2 Billion Fund Is $90

The administration is sending one-time payments of $90 to more than 20 million Medicare enrollees out of a fund Congress created to improve the fee-for-service program, and the statute describes a different kind of spending than the checks going out this month.

By James T.·October 3, 2026·8 min read

Health Policy

A Unanimous Senate Tells HHS to Bring HIPAA's Security Rule Up to Date

The Senate passed the Health Care Cybersecurity and Resilience Act by unanimous consent, directing HHS to set minimum cybersecurity practices for hospitals and clinics through the HIPAA Security Rule, with the grants it creates left to a later spending vote.

By James T.·October 2, 2026·9 min read

Data privacy

Italy Fined IQVIA Seven Million Euros Over a Database It Called Anonymous

Italy's privacy regulator fined IQVIA's Italian unit seven million euros over a health database covering a million patients that the company had treated as anonymous, the second European authority to reject that claim about the same company's data.

By James T.·October 2, 2026·8 min read