Amylin Analogs in 2026: What the Weight-Loss Trials Actually Showed

Evidence review

Amylin analogs went from a footnote to one of the most closely watched classes in metabolic medicine in about two years. Between mid-2025 and early 2026 four separate programmes reported phase 2 or phase 3 results, one of them a head-to-head that did not go the way its sponsor hoped. This is what those trials actually reported – the numbers, the durations, and the caveats that get lost when the headlines are summarised.

Amylin analogs: a short primer before the numbers

Amylin is a 37-amino-acid hormone co-secreted with insulin from pancreatic beta cells. It slows gastric emptying, suppresses glucagon and reduces food intake through brainstem receptors. If you want the physiology in full, start with our explainer on what amylin is and come back.

The one piece of chemistry that matters here is why the natural hormone could never be the drug. Human amylin is aggressively amyloidogenic, and efforts to develop an amylin-based therapeutic stalled over dosage frequency, safety and formulation. Rat amylin does not form amyloid, because it carries prolines at positions 25, 28 and 29 where the human sequence has alanine and two serines – and proline is energetically very unfavourable inside the beta-sheet that amyloid needs. The first approved analog simply borrowed those three residues.

Pramlintide set a modest baseline

Pramlintide received initial US approval in 2005 as an adjunct to mealtime insulin for people with type 1 or type 2 diabetes not at target despite optimal insulin therapy. Across six-month placebo-controlled trials its placebo-subtracted HbA1c effect ran between roughly 0.25 and 0.34 percentage points – and in one dose-titration study, where insulin was actively adjusted, the placebo-subtracted difference was effectively zero. Weight change was on the order of 1 to 2 kg.

It also carries a boxed warning for severe hypoglycaemia, particularly in type 1 diabetes, with events occurring within three hours of an injection, and the label requires mealtime insulin doses to be cut by 50 percent at initiation. That risk belongs to the insulin combination, not to amylin agonism on its own – a distinction that matters when reading the obesity programmes below.

The other constraint was pharmacology: a half-life of about 48 minutes in healthy individuals, which forces mealtime dosing. Everything that came next is, in one sense, an exercise in fixing that number. Our guide to peptide half-life covers why.

Cagrilintide: the once-weekly proof of concept

The 2021 phase 2 dose-finding trial published in The Lancet randomised 706 participants to once-weekly cagrilintide at 0.3 to 4.5 mg, alongside 99 on liraglutide 3.0 mg and 101 on placebo, over 26 weeks. Mean weight loss on the trial-product estimand ran from 6.0 to 10.8 percent across doses, against 3.0 percent on placebo. At 4.5 mg it beat liraglutide 3.0 mg – 10.8 percent versus 9.0 percent, a 1.8-point difference at p=0.03.

Gastrointestinal events and administration-site reactions were the most frequent adverse events, with GI events in 41 to 63 percent of cagrilintide participants against 32 percent on placebo, primarily nausea.

CagriSema: 22.7 percent, then a miss

Combining cagrilintide 2.4 mg with semaglutide 2.4 mg produced the class’s headline numbers. REDEFINE 1, a 68-week phase 3 trial in 3,417 adults with obesity or overweight plus at least one comorbidity and without type 2 diabetes, reported 22.7 percent mean weight reduction against 2.3 percent on placebo on the trial-product estimand. On the treatment-policy estimand the figures were 20.4 percent and 3.0 percent. Just over 40 percent of participants lost at least 25 percent of body weight, and 50.7 percent of those with obesity at baseline reached a BMI below 30. REDEFINE 2, in 1,206 adults with type 2 diabetes, reported 15.7 percent against 3.1 percent. Both were published in the New England Journal of Medicine in June 2025.

The prespecified body-composition substudy is the part worth dwelling on, because it is the mechanistic argument for the whole combination strategy.

Then came REDEFINE 4. In this open-label 84-week phase 3 trial, 809 adults with obesity and at least one comorbidity received either CagriSema or tirzepatide 15 mg. CagriSema produced 23.0 percent weight loss against tirzepatide’s 25.5 percent on the efficacy estimand, and 20.2 against 23.6 percent on the treatment-regimen estimand. The sponsor’s own announcement in February 2026 stated the trial “did not achieve its primary endpoint of demonstrating non-inferiority on weight loss for CagriSema compared to tirzepatide after 84 weeks.”

Petrelintide: the tolerability result

ZUPREME-1, a 42-week randomised, double-blind, placebo-controlled phase 2 dose-finding trial in 493 participants with a mean baseline weight of 107 kg, reported topline results in March 2026. Once-weekly petrelintide met its primary endpoint in all five treatment arms, with up to 10.7 percent mean weight loss at week 42 against 1.7 percent on placebo.

The efficacy number is mid-range for the class. The tolerability number is not. At the maximally effective dose there were no cases of vomiting and no treatment discontinuations for gastrointestinal adverse events. Discontinuation for any adverse event was 4.8 percent, against 4.9 percent on placebo, and vomiting across all petrelintide arms was reported as lower than on placebo. The molecule is described as engineered for chemical and physical stability with no fibrillation around neutral pH – which is what allows co-formulation with other peptides. Phase 3 was endorsed in April 2026 for initiation in the second half of that year.

Eloralintide: selectivity as a design goal

Eloralintide is a selective amylin receptor agonist, and the distinction is real: amylin receptors are assemblies of the calcitonin receptor with one of three receptor activity-modifying proteins, so a molecule can be more or less selective for those complexes over the calcitonin receptor itself. Its 48-week phase 2 in 263 adults, reported in November 2025 and published simultaneously in The Lancet, gave mean weight change of 9.5 percent at 1 mg rising to 20.1 percent at 9 mg, against 0.4 percent on placebo. All arms met the primary endpoint.

The reported adverse events were mild-to-moderate gastrointestinal symptoms and fatigue, more frequent at higher doses, with incidence lower under slower dose escalation and similar to placebo in the 1 mg and 3 mg arms – the same dose-escalation pattern seen throughout this class.

Side effects reported across the class

• Nausea is the signature effect. 48 percent in type 1 and 28 percent in type 2 pramlintide trials against 17 and 12 percent on placebo; 20 to 47 percent for cagrilintide monotherapy; 55 percent against 12.6 percent in REDEFINE 1. Approved labelling notes incidence is highest at initiation and decreases with time in most patients.
• Vomiting and constipation scale with the combination products – 26.1 percent and 30.7 percent respectively in REDEFINE 1 – while petrelintide reported vomiting below placebo across all arms.
• Injection-site reactions were among the most frequent events in the cagrilintide phase 2 trial, and appear in pramlintide’s postmarketing section rather than its controlled-trial tables.
• Hypoglycaemia is a boxed-warning risk specifically when an amylin analog is combined with mealtime insulin.
• Discontinuation for adverse events ran 6.0 percent against 3.7 percent in REDEFINE 1, 8.4 percent against 3.0 percent in REDEFINE 2, and 4.8 percent against 4.9 percent for petrelintide.

For the neighbouring class, see our summary of GLP-1 side effects and the side-effect profile reported for amylin peptides specifically.

What is still unsettled

Three things, and it is worth being explicit about them. First, almost none of these results are head-to-head; the one trial that was did not meet its endpoint. Comparing a 26-week phase 2 to a 68-week phase 3 across different populations and estimands is not a ranking. Second, the tolerability advantage is the most interesting claim in the class and it rests largely on a single phase 2 topline that has not yet been published in full. Third, regulatory status is moving: submissions, decisions and phase 3 initiations described here were accurate at the dates cited and may have changed since. Our note on why peptide trials are small covers why early readouts deserve caution generally.

Frequently asked questions

Are any amylin analogs approved for weight loss?

As of the sources cited here, no. Pramlintide is approved as an adjunct to mealtime insulin for glucose control. CagriSema was under FDA review with a decision anticipated late in 2026, and petrelintide and eloralintide were in or entering phase 3.

Why combine an amylin analog with a GLP-1 drug at all?

Two stated reasons. One is body composition: the REDEFINE 1 substudy showed fat mass falling far more than lean soft tissue, addressing a concern raised about lean-mass loss on incretin therapy alone. The other is tolerability – the hope that a second mechanism adds effect without adding gastrointestinal burden.

Does CagriSema losing to tirzepatide mean amylin does not work?

No. REDEFINE 4 tested one fixed-dose combination against one incretin at one dose over 84 weeks and reported 23.0 against 25.5 percent. It failed a non-inferiority margin, which is a statistical result about that comparison, not a verdict on the mechanism.

Why do nausea rates vary so much between these trials?

Dose, escalation speed and the presence of a GLP-1 partner all move it. Slower titration lowered adverse-event incidence in the eloralintide trial, pramlintide labelling notes nausea decreases with time, and the monotherapy amylin programmes report lower gastrointestinal burden than the combinations.

References

Informational only – not medical advice · 21+. VialHelp is an educational resource and does not sell or recommend peptides. Trial results are reported as published, not as guidance for personal use. Regulatory and development status was accurate at the source dates cited and may have changed.