UBT251 phase 2 diabetes data reach Cell Metabolism as China phase 3 runs

Peer reviewed UBT251 phase 2 results in type 2 diabetes are now listed as in press at Cell Metabolism. The paper by Zhou and colleagues reports that the once weekly triple agonist lowered HbA1c by up to 2.16 percentage points and body weight by up to 9.8 percent at week 24. Those figures match the topline that Novo Nordisk and The United Laboratories released in March 2026. The molecule did not wait for the journal: two phase 3 diabetes trials in China began dosing in September.

What the UBT251 phase 2 trial tested

The trial randomised 211 Chinese adults with type 2 diabetes whose blood sugar was managed by diet and exercise alone or together with metformin. It was double blind and carried two comparators: placebo, and semaglutide at a standard diabetes dose. Three dose levels of UBT251 were given as weekly injections for 24 weeks, and the primary endpoint was the change in HbA1c from baseline. At entry, average HbA1c was 8.12 percent, average body weight was 80.1 kilograms and average body mass index was 29.1. The registry record, NCT07163624, lists the study as completed in November 2025 with 211 participants, which is consistent with the paper. The sponsor named there is The United Bio-Technology (Hengqin), the Chinese developer of the molecule.

UBT251 is a single peptide built to switch on three receptors at once: GLP-1, GIP and glucagon. That places it in the same triple agonist class as retatrutide from Eli Lilly, which is further along in late stage testing. The developer is a wholly owned unit of The United Laboratories International, a group listed in Hong Kong. In March 2025 Novo Nordisk took an exclusive licence to develop, make and sell the drug everywhere except mainland China, Hong Kong, Macau and Taiwan, where the Chinese company keeps the rights. The two firms now run separate programmes for their own territories, which is why the molecule sits among the incretin molecules we track with two development owners and two sets of trials.

HbA1c and weight results against semaglutide

According to the March joint release, the best UBT251 arm cut HbA1c by 2.16 percentage points from baseline at week 24, against 1.77 points for semaglutide and 0.66 points for placebo. Body weight fell by up to 9.8 percent, compared with 4.8 percent on semaglutide and 1.4 percent on placebo. The weight figures are estimates under the efficacy estimand, which assumes that every participant stayed on the assigned treatment. The companies also reported gains against placebo in waist size, blood pressure and blood lipids. They described the safety and tolerability profile as consistent with what other trials of triple agonists have shown, without giving rates of nausea, vomiting or treatment stops in that release.

Body weight fell by up to 9.8 percent, compared with 4.8 percent on semaglutide and 1.4 percent on placebo.

Two points limit how far these numbers go. First, the semaglutide comparison is numeric: neither the March release nor the paper summary we could read says whether the gap was formally tested, and 24 weeks is a short window for a weight comparison. Second, the Cell Metabolism full text sits behind the publisher's access wall, so we could not check the adverse event tables ourselves. Those tables are what clinicians will read first for any new triple agonist, because the glucagon arm of the mechanism is the part with the least long term human data. A peer reviewed paper does mean that the topline survived editorial and referee scrutiny before publication, which a press release alone cannot offer.

Phase 3 trials in China and Novo Nordisk plans

The Chinese side has already moved the molecule into phase 3. ClinicalTrials.gov lists UNIGUIDE-1 (NCT07659574), a triple masked, placebo controlled trial in 360 adults whose diabetes is not controlled by diet and exercise, and UNIGUIDE-2 (NCT07653477), an open label trial in 956 adults on metformin, alone or with a sulfonylurea or an SGLT2 inhibitor, that uses semaglutide as the comparator. Both records give an actual start date of 7 September 2026, and both measure the change in HbA1c at week 40. UNIGUIDE-1 has an estimated primary completion in September 2027 and UNIGUIDE-2 in February 2028. United Biotechnology sponsors both studies, in line with its rights to the Chinese market.

Novo Nordisk is a step behind with its own work. Its registry entries show a phase 2 dose study in type 2 diabetes with about 300 planned participants that began in June 2026, and an obesity study with 333 planned participants that began in February. In March, Novo said topline results from its global obesity trial were expected in 2027. A second UBT251 phase 2 paper, in adults with overweight or obesity, appeared online in Cell Metabolism on 8 October, according to its PubMed record. Readers can see where the molecule sits next to its rivals on our pipeline board, which shows each stage move once we have confirmed it against a registry or a company filing.

What the paper changes for the triple agonist

On our board the publication does not move UBT251. The molecule entered phase 3 in China in September, so this readout confirms the evidence behind that step rather than being a step of its own. We rate the signal as up because the full paper repeats the March topline without any visible change in direction, and because the semaglutide arm gives readers a fixed reference point inside the same trial. The size of the HbA1c fall is large for 24 weeks in a group that started at 8.12 percent. Yet cross trial comparisons with retatrutide or other triple agonists would be unsafe, since populations, doses, durations and estimands all differ between those programmes.

Several questions stay open. The paper covers one population, Chinese adults whose diabetes was managed by diet or metformin, and only 24 weeks of treatment. It does not show how UBT251 behaves over a year or more, in people with heavier body weight, or against tirzepatide. The phase 3 trials in China will answer part of that by 2028, and the Novo Nordisk studies will show whether the results hold in more diverse global populations. Nothing here is a guide to treatment. UBT251 is not approved anywhere, and our research disclaimer page explains why the stories on this site are research news only and never medical advice for any reader.

Conclusion

The UBT251 phase 2 paper turns a March press release into peer reviewed evidence: up to 2.16 points off HbA1c and up to 9.8 percent off body weight at 24 weeks, ahead of semaglutide on both counts in numeric terms. Safety detail remains the gap we could not check. The next checkpoints are the UNIGUIDE phase 3 readouts in China and Novo Nordisk's global phase 2 data, expected from 2027.