A phase 3b trial published online on 14 November 2026 found that two lower doses of cagrilintide plus semaglutide, a combination called CagriSema, reduced body weight more than placebo over 68 weeks. The comparisons below come from the published abstract. Novo Nordisk funded the trial. ClinicalTrials.gov still describes CagriSema as a medicine doctors cannot yet prescribe.

REDEFINE 9, registered as NCT06388187, was run at 24 sites in Canada, France, Germany, the United Kingdom and the United States. Adults could take part if their body mass index was at least 30, or at least 27 with an obesity-related complication, and if they did not have diabetes. From 24 June to 30 October 2024, investigators screened 333 people, excluded 24, and randomly assigned 309. Of those, 104 were assigned to 1.7 mg of cagrilintide plus 1.7 mg of semaglutide, 103 to 1.0 mg of each drug, and 102 to placebo. Injections were given under the skin once a week. The 1.7 mg dose started at 0.25 mg of each drug and was increased over 12 weeks. The 1.0 mg dose started at the same point and was increased over 8 weeks. Everyone was offered nutrition and physical activity counseling through week 74.

The mean age was 48.0 years, with a standard deviation of 13.6. The mean body mass index was 39.1, with a standard deviation of 6.8. Women made up 239 of the 309 participants, or 77 percent. The abstract says 291 people, 94 percent, completed the study, and 254, 82 percent, were still taking their assigned treatment at week 68.

On the efficacy estimand, mean body weight at week 68 changed by -20.7 percent with 1.7 mg of each drug (standard error 0.9), -21.0 percent with 1.0 mg of each (standard error 1.0), and -2.0 percent with placebo (standard error 1.0). Against placebo, the estimated treatment differences were -18.8 percentage points (95 percent confidence interval -21.5 to -16.0) and -19.0 percentage points (-21.8 to -16.2). Both comparisons had a p value below 0.0001. The abstract also gives estimated shares who lost at least 5 percent of their starting weight: 95.1 percent (standard error 2.7) on the 1.7 mg dose, 96.3 percent (standard error 2.2) on the 1.0 mg dose, and 33.4 percent (standard error 5.6) on placebo.

Those weight changes are the efficacy estimand. The abstract does not give a second average that keeps every assigned person in the calculation, including people who stopped the injections. The authors write that different dose levels might leave more room to adjust treatment. Each dose is reported against placebo. The abstract does not include a statistical comparison of the two doses with each other.

Safety was counted for the 308 people who received at least one dose. Adverse events were reported in 102 of 103 people on 1.7 mg of each drug, 98 of 103 on 1.0 mg of each, and 93 of 102 on placebo: 99 percent, 95 percent and 91 percent. The abstract says most were mild or moderate and gastrointestinal. No deaths were reported. It does not give separate counts for nausea, serious adverse events, or people who stopped because of side effects. One person assigned to the 1.7 mg dose is outside the safety count of 103.

The registry lists the trial as completed, with a primary completion date of 22 April 2026, and names Novo Nordisk A/S as the lead sponsor. It records actual enrollment as 300. The paper randomly assigned 309. The registry page does not explain that gap. Nothing in the abstract says these doses are approved or available as routine care.