Cagrilintide and amylin: the other satiety hormone

The other satiety hormone

When peptides for obesity come up, the conversation usually starts and ends with GLP-1. Amylin is the other piece. It is a 37-amino-acid peptide that pancreatic β-cells release together with insulin after a meal¹¹². It slows gastric emptying, suppresses postprandial glucagon secretion and promotes meal termination through central mechanisms¹².

Its receptor is unusual. There is no single "amylin receptor": the calcitonin receptor combines with receptor activity-modifying proteins (RAMPs) to form several subtypes. Amylin acts mainly in the circumventricular organs of the central nervous system and interacts with other metabolic hormones such as cholecystokinin, leptin and estradiol¹. This biology differs from that of GLP-1, which is why investigators study it as a complementary pathway.

Pramlintide: the proof of concept

Human amylin tends to form amyloid fibrils, which makes it hard to use as a drug⁷. The answer was pramlintide, a stable analog the FDA approved on 16 March 2005 as a new molecular entity (Symlin)². Its indication was as an adjunct to mealtime insulin in type 1 and type 2 diabetes when desired control is not reached³.

• Type 1 diabetes (651 patients, 52 weeks): HbA1c reduction of 0.29 to 0.34 points versus 0.04 with placebo, and −0.4 kg versus +0.8 kg⁴.
• Insulin-treated type 2 diabetes (656 patients, 52 weeks): HbA1c −0.62 points with 120 µg twice daily and −1.4 kg versus +0.7 kg with placebo⁵.
• Obesity without diabetes (411 study subjects): at 12 months (146 evaluable), placebo-corrected loss was 5.6% and 6.8% with 120 µg three times daily and 360 µg twice daily, and 40–43% of those two arms reached at least 10% loss versus 12% with placebo⁶.

The limit was practical: its short half-life required three daily injections⁷. Today Drugs@FDA lists all its presentations as discontinued². But it showed that activating the amylin pathway is associated with lower weight.

How cagrilintide was built

Cagrilintide was designed to solve that problem. It is a stable, lipidated amylin analog designed for once-weekly administration⁷, and it activates both amylin and calcitonin receptors (a dual agonist, or DACRA)¹². In a phase 1b trial its half-life was 159 to 195 hours, roughly 7 days⁸. The cagrilintide page summarizes its structure and physicochemical data.

Cagrilintide alone: phase 2

The reference trial for the molecule alone is the phase 2 study by Lau and colleagues, published in The Lancet in 2021⁹. It randomized 906 adults without diabetes: 706 to five once-weekly cagrilintide doses (0.3 to 4.5 mg; 100–102 per group), 99 to daily liraglutide 3.0 mg and 101 to placebo, for 26 weeks.

With the trial product estimand, every dose was associated with greater reduction than placebo: 6.0% to 10.8% (6.4 to 11.5 kg) versus 3.0% (3.3 kg). The 4.5 mg dose outperformed liraglutide: 10.8% versus 9.0% (difference 1.8%, p = 0.03). The treatment policy estimand gave similar results⁹. Permanent discontinuation was 10% and was similar across groups⁹.

In type 2 diabetes (92 patients, 32 weeks), cagrilintide alone was associated with an HbA1c reduction of 0.9 points and weight reduction of 8.1%; the combination with semaglutide reached −2.2 points and −15.6%¹³.

CagriSema: the combination in phase 3

The first hint of synergy came from phase 1b: with semaglutide 2.4 mg, cagrilintide 4.5 mg was associated with −15.4% versus −8.0% with semaglutide plus placebo at 20 weeks, in 96 participants⁸. Phase 3 confirmed it at scale.

• REDEFINE 1 (3417 adults without diabetes, 68 weeks): −20.4% with the combination versus −3.0% with placebo (treatment policy); gastrointestinal events in 79.6% versus 39.9%, mostly transient and mild to moderate¹⁰.
• REDEFINE 2 (1206 adults with type 2 diabetes, 68 weeks): −13.7% versus −3.4%; 73.5% reached an HbA1c of 6.5% or less, versus 15.9%¹¹.

REDEFINE 1 also included a cagrilintide-alone arm (302 participants)¹⁰. A 2026 review summarizing that trial describes roughly 11–12% weight loss at 68 weeks with monotherapy¹². For how to read estimands and compare with other next-generation drugs, see After semaglutide.

The new wave of the amylin class

• Amycretin (now zenagamtide): a single molecule that agonizes both GLP-1 and amylin receptors¹⁵. In phase 1b/2a (125 participants), the 60 mg subcutaneous dose was associated with −24.3% versus −1.1% at 36 weeks¹⁴.
• Petrelintide: a long-acting human amylin analog. In the phase 2 ZUPREME 1 trial (485 treated participants), it was associated with −7.9% to −9.8% versus −1.7% at 28 weeks (efficacy estimand), with up to −10.7% at 42 weeks; nausea in 20% versus 6% and vomiting in 3% versus 6%¹⁶.
• A 2026 review also lists eloralintide, MET-233i, ABBV-295 and AZD6234 among amylin analogs in clinical development¹².

Where the evidence ends

The signal is consistent, but it should be read precisely. Gastrointestinal events are common: in phase 2 they affected 41–63% with cagrilintide versus 32% with placebo, mainly nausea⁹. Pramlintide with insulin carries a boxed warning for severe hypoglycemia, particularly in type 1 diabetes³. In the amycretin study a large number of participants withdrew¹⁴.

Long-term data on cagrilintide alone come from a secondary arm of REDEFINE 1, not from a dedicated phase 3 trial¹⁰¹². And every cagrilintide, amycretin and petrelintide trial cited here was funded by its manufacturer^8910111416. For the laboratory and the research protocol, the next question is not whether the pathway works, but which receptor profile (selective amylin or dual amylin-calcitonin) offers the best balance of efficacy and tolerability¹².