Cagrilintide: Most of What Circulates Under Its Name Belongs to CagriSema

01What it is

Cagrilintide is a lipidated, disulphide-bridged 37-residue amylin analogue, developed by Novo Nordisk as NNC0174-0833 and called AM833 in the literature. It was not built directly on human amylin. The Nature Communications cryo-EM paper states that pramlintide is the backbone peptide used to generate it, and pramlintide is Symlin, approved by FDA in 2005 under NDA 021332 and dosed three times daily with meals. That approval and that safety record stay with pramlintide.

What Novo added was a C20 fatty diacid on the N-terminus, which makes the peptide bind albumin and last a week. The same April 2025 paper solved structures of cagrilintide bound to all three amylin receptors, AMY1R, AMY2R and AMY3R, plus the calcitonin receptor alone, so it is non-selective across the family. (PMID 40204768) That the weight loss runs through those receptors has been shown only in mice, where knockouts lacking AMY1R and AMY3R lost none at all. (PMID 40609154)

02What the trials found

The first in human trial, NCT02958085, randomised 96 people with overweight or obesity between cagrilintide under the skin and placebo for eight weeks, reaching 800 micrograms per day, and weight loss ran 6 to 8 percent against placebo. The weekly schedule comes out of how long it lingers: a half life of approximately 180 hours, in Novo's own words, supporting once weekly dosing.

Then the only dedicated monotherapy dose-finding trial. Lau and colleagues, Lancet 2021, NCT03856047, funded by Novo: 706 adults for 26 weeks, once-weekly subcutaneous cagrilintide at 0.3, 0.6, 1.2, 2.4 or 4.5 mg, about 100 people per arm, against 99 on once-daily subcutaneous liraglutide 3.0 mg and 101 on placebo. The trial product estimand asks what happened to people who stayed on the drug as assigned rather than to everybody randomised. Weight change at week 26 on the trial product estimand came out minus 6.0, 6.8, 9.1, 9.7 and 10.8 percent up the ladder, against minus 9.0 percent on liraglutide and minus 3.0 percent on placebo. (PMID 34798060) Read that against the comparator rather than the placebo: liraglutide is another injection for the same job, so the trial was asking whether cagrilintide beat a working drug and not a dummy. Only the top dose beat liraglutide, by 1.8 percentage points (95% CI minus 3.5 to minus 0.2, p=0.03), while the two lowest doses separated significantly in the other direction, losing to liraglutide by 3.0 and 2.2 points. None of it was adjusted for multiplicity, the correction you make when a trial tests five doses at once and one of them will flatter itself by chance.

REDEFINE 1, NCT05567796, randomised 3,417 adults without type 2 diabetes 21:3:3:7 across CagriSema, semaglutide, subcutaneous cagrilintide 2.4 mg alone (302 people) and placebo (705) for 68 weeks, the single agent escalated 0.25, 0.5, 1.0, 1.7 then 2.4 mg over 16 weeks and held there. Garvey presented that arm as a post hoc analysis at EASD in September 2025: minus 11.8 percent against minus 2.3 percent on placebo, treatment difference minus 9.5 percentage points, p below 0.0001. A post hoc analysis is one nobody planned, and this one came off a conference podium rather than a paper. 31.6 percent of that arm lost at least 15 percent of body weight against 4.7 percent on placebo, and 15 percent is the highest threshold that presentation reports for the single agent.

A second phase 3, REIMAGINE 2, NCT06065540, carries a subcutaneous cagrilintide 2.4 mg arm of 152 among 2,713 people with type 2 diabetes on metformin, and at 68 weeks Novo's release of 7 June 2026 gives that arm as HbA1c minus 0.80 percentage points and weight minus 8.4 percent, against minus 1.75 and minus 10.2 percent on semaglutide 2.4 mg and minus 1.91 and minus 14.2 percent on CagriSema. (PMID 42251859) I read the abstract and the release, not the paper.

03The numbers that are CagriSema's

REDEFINE 1's headline for the combination is minus 20.4 percent against minus 3.0 percent on placebo, with 40.4 percent of participants losing a quarter or more of their body weight. (PMID 40544433) Set the two arms against each other over the same 68 weeks under the trial product estimand, where CagriSema comes out at minus 22.7 percent and cagrilintide on its own at minus 11.8, and the combination is about double the single agent. Semaglutide is doing most of that work.

FDA's warning letter to Prime Sciences of Scottsdale, MARCS-CMS 721805, dated 31 March 2026, reproduces the vendor's cagrilintide product page, and that page recites REDEFINE 1's 40.4 percent under the abbreviation Cagri while selling the single agent, whose own highest reported threshold is the 15 percent above.

Cagrilintide is not approved by FDA, alone or in combination, and I found no approval anywhere else. A Drugs@FDA query for cagrilintide and for cagrisema returns NOT_FOUND, and the New Drug Application Novo filed on 18 December 2025 is for CagriSema, not for the single agent. Novo's own accounting tells you the rest: in the Q2 2026 half-year report filed with the SEC as a Form 6-K, the string CagriSema appears 22 times and cagrilintide appears exactly once, in the milestones table, as a phase 3 initiation for a high dose. That report is written for people deciding whether to buy the company, which is what makes the word count worth reading.

04How it compares to what is sold beside it

None of it transfers. A row does not become cagrilintide's evidence because a shop puts them on the same page, and cagrilintide is the only one of the four whose entire human record was paid for by a single company.

05Where the 2.4 mg came from

Reported as convention, because that is all it is. What is sold are 5 mg and 10 mg vials labelled research chemical, not for human consumption. NTN Performance publishes 0.6 to 4.5 mg weekly titrated every two weeks, which comes off the phase 2 trial. Peptides.org publishes REDEFINE 1's ladder copied intact, 0.25 mg through to 2.4 mg from week 17. Neither was invented, which is unusual for a peptide convention.

CagriSema is a fixed-dose pen and 2.4 mg is semaglutide's dose, so the maintenance target people copy for the single agent was set by the other half of a combination the single agent is not in. The phase 2 trial, the only dose-finding trial the single agent has had, found 4.5 mg better than 2.4 mg, minus 10.8 against minus 9.7 percent. A thorough QT study is the standard check on whether a drug disturbs the heart's electrical timing. 4.5 mg weekly is also the highest dose ever given to a human in a published trial, there and in a thorough QT study that found no clinically relevant QTcF prolongation. (PMID 39279639)

The dose that will settle it is not public. Novo announced RENEW on 16 September 2025, a dedicated phase 3 monotherapy programme, and RENEW 1 (NCT07220642) and RENEW 2 (NCT07220759) both started on 5 November 2025, 64 weeks of once-weekly subcutaneous cagrilintide against matched placebo. I searched both registrations for a milligram string: RENEW 1 contains none, and RENEW 2's only figure is an unrelated 54 mg. I could not find the dose. That is a gap in what has been published rather than a claim that no dose exists, but every titration chart in circulation may already be aimed at a superseded number.

06What could go wrong

Start with the conference presentation of the single agent arm, the one place a reader sees the two drugs measured the same way by the same people. Nausea ran 23.8 percent on cagrilintide against 43.7 on semaglutide and 55.0 on the combination. Then the direction reverses. Administration site reactions ran 17.5 percent on cagrilintide against 2.6 percent on semaglutide and 12.2 on the combination. Easier on the stomach, roughly seven times more likely to leave a mark where the needle went in.

Then the finding nobody puts on a product page. Novo's own phase 2 protocol, dated 2 October 2018, records that reproductive toxicity studies in rats and mice showed major foetal malformations at what the sponsor itself calls human-relevant exposures. Rabbits were unaffected despite significantly higher exposure. The mode of action had not been identified, and women of childbearing potential were excluded from the phase 2 trial as a result.

The same trial found anti-cagrilintide antibodies in 46.3 to 73.2 percent of participants by week 26, most of them cross-reactive with the person's own native amylin, and I found no follow-up past 26 weeks.

07My take

The work to do with this compound is arithmetic. Cagrilintide alone has been through real randomised trials with real placebo arms and it does something: minus 11.8 percent over 68 weeks at 2.4 mg, minus 10.8 percent over 26 weeks at 4.5 mg, and less than half of semaglutide's effect on HbA1c in phase 3 diabetes. A genuine drug, and also roughly half the figure people quote when they buy it.

What stays with me is the 2018 protocol. Malformations in two rodent species at exposures the sponsor itself called human-relevant, no mechanism, and eight years on I found no published resolution. It sits in an appendix where almost nobody will read it, while a percentage belonging to a different product sits on the front of a shop page.

If you have seen cagrilintide sold, what percentage was printed on the page, and does it match REDEFINE 1's monotherapy arm at minus 11.8 percent or the combination at minus 20.4?

Frequently asked

Is cagrilintide FDA approved?+

No. Cagrilintide is not approved by the FDA alone or in combination, and I found no approval anywhere else.

How much weight did people lose on cagrilintide alone?+

Less than the figure usually printed beside its name. In REDEFINE 1, the single agent arm of 302 adults without type 2 diabetes lost 11.8 percent of body weight over 68 weeks against 2.3 percent on placebo, presented by Garvey as a post hoc analysis at a conference in September 2025.

Is cagrilintide the same thing as CagriSema?+

No. CagriSema is a fixed dose pen holding cagrilintide and semaglutide together, and most of the numbers that circulate under the cagrilintide name were earned by that pen.

Where does the 2.4 mg cagrilintide figure come from?+

It is convention rather than evidence about the single agent, reported here as what circulates and not as a recommendation. 2.4 mg is semaglutide's dose inside the fixed combination pen, so the maintenance target people copy for cagrilintide alone was set by the other half of a product the single agent is not in.

What are the side effects of cagrilintide?+

Nausea ran 23.8 percent on cagrilintide against 43.7 on semaglutide and 55.0 on the combination.

Is cagrilintide safe in pregnancy?+

Novo Nordisk's own phase 2 protocol, dated 2 October 2018, records that reproductive toxicity studies in rats and mice showed major foetal malformations at what the sponsor itself calls human relevant exposures. Rats and mice yes, rabbits no, humans unknown, and I found no published follow-up resolving it.

---

This content is for educational and informational purposes only and is not medical advice. Peptides discussed are research compounds and may not be approved for human use. Nothing here should be used to diagnose, treat, cure, or prevent any disease. Always consult a qualified healthcare professional before starting any peptide, supplement, or protocol. Individual responses vary. Do not self-administer compounds without proper medical supervision.