Cagrilintide: The Amylin Analogue, and What Its Published Trials Actually Report
Cagrilintide is an engineered amylin analogue with a genuinely interesting design story: native amylin forms amyloid fibrils so readily that a long-acting version was considered impractical for years. Here is how that was solved, and what the published trials report — including the one that missed its endpoint.
- Cagrilintide is a 37-residue analogue of human amylin (IAPP) with six substitutions and an N-terminal C20 fatty-diacid acylation, giving a reported half-life of roughly 174–186 hours versus 20–45 minutes for the short-acting analogue pramlintide.
- Its design solves a specific chemistry problem: native amylin is strongly amyloidogenic, and the substitutions create a helix-stabilising Glu14–Arg17 salt bridge plus β-sheet-breaking prolines to resist fibril formation.
- It acts at the calcitonin receptor and all three amylin receptors — a different receptor family from GLP-1 and GIP agonists, which is why the published work studies it alongside incretin agonists rather than against them.
- In REDEFINE 1 (NEJM 2025, n=3,417), cagrilintide alone produced −11.5% body weight at 68 weeks against −3.0% for placebo; co-administered with semaglutide the figure was −20.4% on the treatment-policy estimand.
- In REDEFINE 4 (announced February 2026, n=809), cagrilintide plus semaglutide did NOT meet non-inferiority against tirzepatide 15 mg at 84 weeks — 23.0% versus 25.5%.
- Neither cagrilintide nor the CagriSema combination is approved in Canada, the United States or the EU. A US filing was submitted in December 2025 with a decision expected in Q4 2026.
Why amylin took so long
Amylin is not an obscure molecule. It is a 37-amino-acid hormone released from pancreatic beta cells at the same time as insulin, characterised decades ago, and its biology has been of interest since. The obstacle was never the biology. It was the chemistry.
Human amylin — also called islet amyloid polypeptide, or IAPP — is one of the more amyloidogenic peptides known. Its natural tendency is to convert from an α-helical conformation into β-sheet, and β-sheets stack into fibrils. The published discovery work on cagrilintide opens by naming this directly: the high propensity toward amyloid fibril formation is what made amylin a challenging drug-design effort.
That is a hard constraint for anything meant to last. A molecule intended to persist for days has to resist a conformational collapse it is intrinsically prone to, for days.
What was actually changed
Cagrilintide keeps amylin's overall architecture: 37 residues, the Cys2–Cys7 disulfide bridge, the amidated C-terminus. Novo Nordisk describes it as retaining roughly 84% homology to the native hormone. Six positions differ — 14E, 17R, 25P, 28P, 29P and 37P — and the N-terminus carries a C20 fatty diacid attached through a γ-glutamate linker.
Those changes are not arbitrary, and they divide into two strategies.
**Helix stabilisation.** The N14E and V17R substitutions place a negatively charged glutamate and a positively charged arginine three positions apart, where they form an intramolecular salt bridge. A salt bridge across one turn of an α-helix stabilises that helix — and a molecule held in helix is a molecule not available to seed β-sheet.
**β-sheet disruption.** The prolines at 25, 28 and 29 are the same substitutions found in pramlintide, the earlier short-acting amylin analogue. Proline cannot participate in a β-sheet hydrogen-bonding pattern in the ordinary way, so proline substitutions in a fibril-prone region act as structural breakers. Cagrilintide adds a fourth at position 37.
**Protraction.** Separately from the anti-fibril work, the C20 diacid binds albumin, which is what converts a peptide with a half-life measured in minutes into one measured in days. The published figures make the contrast plain: pramlintide is reported at 20–45 minutes; a 2026 clinical pharmacokinetics study reports cagrilintide's geometric mean half-life at 174–186 hours in people with normal organ function, with renal and hepatic impairment not meaningfully altering it.
Why does this matter for handling the reagent?
Because the failure mode is inherited. Cagrilintide resists fibrillation — it does not abolish the pathway. Shaking, foaming and freeze–thaw all push an acylated peptide toward aggregation, and for an amylin analogue that aggregation has somewhere particularly unhelpful to go. Once fibrils nucleate they do not redissolve on rewarming. Handling guidance for this compound is less a courtesy than a condition of the material staying what the certificate says it is.
What the trials report
Phase 2: cagrilintide on its own
The dose-finding trial (*Lancet* 2021;398(10317):2160–2172) randomised 706 participants across cagrilintide 0.3, 0.6, 1.2, 2.4 and 4.5 mg once weekly, against both placebo and liraglutide 3.0 mg daily, over 26 weeks.
Body-weight change at week 26 ran from −6.0% to −10.8% dose-dependently, against −3.0% for placebo. The top dose beat the active comparator: −10.8% for cagrilintide 4.5 mg against −9.0% for liraglutide 3.0 mg (p=0.03).
One finding from that trial is rarely repeated and deserves to be: anti-drug antibodies appeared in 46.3% to 73.2% of participants, though characterised as mostly non-neutralising. For an analogue of an endogenous human hormone, immunogenicity is a real question rather than a footnote.
Phase 3: REDEFINE 1
*NEJM* 2025;393(7):635–647. 3,417 participants across 22 countries, 68 weeks, four arms: cagrilintide plus semaglutide (the CagriSema combination), semaglutide alone, cagrilintide alone, and placebo.
On the treatment-policy estimand — the one closest to intention-to-treat — the combination produced −20.4% against −3.0% for placebo. The monotherapy arms are the more interesting numbers for anyone studying cagrilintide itself: **cagrilintide alone −11.5%**, semaglutide alone −14.9%.
A note on which number you are reading
Coverage of this compound routinely quotes 22.7% for REDEFINE 1. That figure is real, but it is the *trial-product* estimand, which models idealised adherence. The *treatment-policy* estimand published in the same paper is 20.4%. Both are legitimate analyses answering different questions. Quoting one as though it were the other is the most common error in reporting on this drug, and the same split explains why REDEFINE 2 is variously cited at 13.7% or 15.7%.
Phase 3: REDEFINE 2
*NEJM* 2025;393(7):648–659. 1,206 participants with overweight or obesity and type 2 diabetes, 68 weeks. Body weight −13.7% against −3.4% for placebo, with 73.5% reaching HbA1c ≤6.5% against 15.9%. Gastrointestinal adverse events occurred in 72.5% of the treatment group, characterised as mostly transient and mild to moderate.
Phase 3: REDEFINE 4 — the trial that missed
This one is frequently left out, and leaving it out misrepresents the picture.
REDEFINE 4 was an open-label head-to-head against tirzepatide 15 mg: 809 participants, 84 weeks. Novo Nordisk announced the result on 23 February 2026. The primary endpoint was non-inferiority of the cagrilintide–semaglutide combination to tirzepatide, and **it was not met**. On the efficacy estimand the combination reached 23.0% against tirzepatide's 25.5%; on the treatment-regimen estimand, 20.2% against 23.6%.
Any account of this compound that presents it as straightforwardly the most effective option available is contradicted by the sponsor's own announcement.
Where it is not approved
Neither cagrilintide as a single agent nor the CagriSema combination is approved in Canada, the United States or the European Union.
A US new drug application for the combination was submitted on 18 December 2025 on the strength of REDEFINE 1 and 2, with Novo Nordisk's own Q2 2026 investor materials indicating a decision expected in Q4 2026. The EU regulatory pathway is described by the company as pending REDEFINE 3, the cardiovascular outcomes trial, which is expected to read out in the second half of 2027. We found no public record of a Canadian new drug submission.
Cagrilintide monotherapy remains in phase 3 development following the REDEFINE 1 monotherapy result.
What this material is, and what it is not
Everything above is a summary of published clinical findings. It is reporting, not instruction.
The cagrilintide supplied by Popular Peptides is a laboratory reagent for in vitro research. It is not a pharmaceutical, it is not an unapproved-drug workaround, and it is not equivalent to a medicine that has been through pharmacopoeial manufacture, pharmacist dispensing and clinical oversight. No cagrilintide product has completed that path in any jurisdiction. Nothing in this article should be read as suggesting anyone administer this compound to themselves or to anyone else.
What the data does not say
The published trials study a specific molecule at specific doses in monitored clinical populations under defined protocols. They do not establish anything about material obtained outside that setting, they do not speak to long-term outcomes beyond their own follow-up windows, and — in REDEFINE 4's case — they do not establish superiority to an existing approved comparator. The immunogenicity signal from the phase 2 work remains a live question in a molecule this closely related to an endogenous hormone.
---
Popular Peptides supplies Cagrilintide (10 mg) for in vitro laboratory research use, shipped domestically from Vancouver, BC with a third-party certificate of analysis verifiable by lot number. It is not sold for human or veterinary consumption.
Cagrilintide is supplied for in vitro laboratory research use only. It is not approved by Health Canada for therapeutic use, is not a treatment for any disease, and is not intended for human or veterinary consumption. 18+.
Every batch we sell is independently third-party tested, with the full Certificate of Analysis published and checkable by lot number.