Cagrilintide Dosage in Units: Reconstitution Chart, Half-Life & the 2026 Phase 3 Record

Cagrilintide at a glance

What it is

Cagrilintide (AM833), a lipidated long-acting synthetic analog of the pancreatic hormone amylin, developed by Novo Nordisk and described in J Med Chem 2021 as compound 23 of that program.

Doses actually used in trials

0.16, 0.30, 0.60, 1.2, 2.4 and 4.5 mg once weekly in phase 1b dose-ranging; 1.0 mg and 2.4 mg once weekly are the two doses carried into the 2026 phase 3 program.

Route and frequency

Subcutaneous, once weekly, co-escalated in roughly 4-week steps toward the 2.4 mg target dose in the phase 3 trials.

Plasma half-life

159-195 hours (about 6.6-8.1 days) across 0.16-4.5 mg, with median tmax 24-72 hours. Measured as an exploratory endpoint in Enebo et al., Lancet 2021 (PMID 33894838) - a real source, which most pages quoting this figure do not give.

What it adds to semaglutide, on glycaemia

0.16 percentage points of HbA1c. In REIMAGINE 2 (n=2,713, 68 weeks) CagriSema 2.4/2.4 gave -1.91 pp against -1.75 pp for semaglutide 2.4 mg alone: statistically significant (p=0.0035), clinically small.

What it adds on bodyweight

Considerably more. In REDEFINE 5 (n=331, 68 weeks) CagriSema reached -18.4% bodyweight against -11.9% for semaglutide 2.4 mg alone, a difference of 6.5 percentage points (p<0.0001).

Regulatory status

Not approved. As of 2026-08-13 a search of the FDA's Drugs@FDA database returns no approved product containing cagrilintide, alone or as CagriSema. Research use only.

Renal or hepatic impairment

Two dedicated PK studies found no clinically relevant difference in exposure across normal, mild, moderate and severe impairment, concluding dose adjustment is not warranted for those populations.

Reported ranges from research/community — examples, not recommendations.

What it is / mechanism

Cagrilintide is a long-acting analog of amylin, the 37-amino-acid hormone co-secreted with insulin by pancreatic beta cells. Native amylin has a strong tendency to form amyloid fibrils, which is the reason a stable therapeutic analog was difficult to make: the medicinal-chemistry paper describing cagrilintide's development opens on exactly that problem, and notes that the existing amylin analog pramlintide requires three injections a day because of its short half-life. Cagrilintide's answer is lipidation — a fatty-acid side chain that promotes reversible albumin binding and slows clearance enough to support once-weekly dosing. Pharmacologically it is a non-selective amylin-receptor agonist. Amylin receptors are the calcitonin receptor complexed with receptor-activity-modifying proteins (RAMP1 and RAMP3), expressed in appetite-regulating regions of the hindbrain. Activation slows gastric emptying and increases satiety signalling, reducing food intake. The reason cagrilintide is almost always discussed alongside semaglutide is that these are different receptors on a shared behavioural endpoint. GLP-1 receptor agonism and amylin receptor agonism both suppress appetite through partly separate circuits, so the combination was expected to be additive rather than redundant. The 2026 phase 3 program is what tested that expectation, and the answer turns out to depend on which endpoint you ask about.

Researched effects

Across the trial program, cagrilintide's measurable contribution is a bodyweight contribution. That statement is more specific than it sounds, and the evidence for it comes from setting two 2026 phase 3 results side by side. REIMAGINE 2 is the largest and most informative trial in the program, and the only one with a cagrilintide monotherapy arm running alongside semaglutide monotherapy. It randomised 2,713 people with type 2 diabetes across 30 countries for 68 weeks, and its primary endpoint was HbA1c with CagriSema 2.4/2.4 against semaglutide 2.4 mg alone. The combination won — and the margin was 0.16 percentage points (-1.91 vs -1.75; 95% CI -0.27 to -0.05; p=0.0035). That is a real result, replicated at scale, and it is also a sixth of a percentage point. On glycaemic control specifically, adding an amylin analog to semaglutide does very little. On weight, the same combination behaves differently. REDEFINE 5 randomised 331 people with obesity in Japan and Taiwan to CagriSema or semaglutide, both escalated to 2.4 mg, for 68 weeks: bodyweight fell 18.4% with the combination against 11.9% with semaglutide alone, a treatment difference of 6.5 percentage points (95% CI -8.4 to -4.6; p<0.0001). The phase 1b work pointed the same way earlier, with 17.1% weight reduction at cagrilintide 2.4 mg plus semaglutide against 9.8% for pooled placebo plus semaglutide at week 20. These are two different trials in two different populations, so this is a cross-trial comparison and not a within-trial dissociation - it should be read as a pattern, not a controlled contrast. But the pattern is consistent and it is the useful thing to know about this compound: cagrilintide is an appetite and bodyweight lever that adds little to glucose control on top of an already effective GLP-1 agonist. As monotherapy, a 2026 meta-analysis of four randomised trials covering 5,425 participants put cagrilintide alone at roughly -6.08% bodyweight (-5.89 kg) versus placebo.

Dosage — reported ranges (overview)

The amounts below are the doses used in published trials, reported for reference. They are not a recommendation, and nothing here is a personal dose. The published dose set is unusually well defined for a compound in this library. Phase 1b dose-ranging used 0.16, 0.30, 0.60, 1.2, 2.4 and 4.5 mg once weekly, co-escalated with semaglutide in 4-week intervals to the target dose over 16 weeks, then held at target for 4 weeks. The 2026 phase 3 program narrowed to two dose levels carried through as fixed-dose combinations: 1.0 mg and 2.4 mg once weekly, each paired with the matching semaglutide dose, again reached by stepwise escalation rather than started at target. Two features of that schedule are worth reading off the pharmacology. Dosing is weekly because the half-life is 159-195 hours - a week is roughly one half-life, which is what produces a workable steady state. And escalation is stepwise because the adverse events across every trial in the program are predominantly gastrointestinal and dose-related; the step-up exists to manage tolerability, not to build up an effect. The highest dose ever taken into a trial is 4.5 mg weekly, in phase 1b. It did not outperform 2.4 mg on weight at week 20 (-15.4% versus -17.1%), which is part of why 2.4 mg is the dose the program carried forward.

A printable protocol sheet with a reconstitution reference and an injection log comes with All-Access Lifetime.

Reconstitution — bac-water math

This section is concentration arithmetic. It is not a dose, and the fact that a number can be calculated does not make it a recommendation. Cagrilintide is dosed in milligrams, so the concentration is simply the vial's total mg divided by the volume of bacteriostatic water added, in mg/mL. On a U-100 insulin syringe, 100 units is 1 mL, so units to draw = dose (mg) / concentration (mg/mL) x 100. Worked example on a 10 mg vial: adding 2 mL of bacteriostatic water gives 10 / 2 = 5 mg/mL. A 2.4 mg reference amount is 2.4 / 5 = 0.48 mL, which is 48 units on a U-100 syringe. A 1.0 mg amount is 20 units. One practical limit is worth stating rather than hiding, because it decides how much water to use. A single 1 mL U-100 syringe holds 100 units. On a 5 mg vial reconstituted with 3 mL, the concentration falls to 1.67 mg/mL and a 2.4 mg amount would require 144 units - more than one syringe. That row is deliberately absent from the table below. On a 5 mg vial with 2 mL, 2.4 mg is 96 units: it fits, but it fills essentially the whole barrel. The 4.5 mg dose used in phase 1b is 90 units at 5 mg/mL and does not fit at all at lower concentrations. Every row below was recomputed independently and every draw falls within a single syringe. Add water slowly down the vial wall and swirl gently; do not shake.

This is concentration math, not a dose recommendation.

Injection / administration basics

In the trials cagrilintide was given as a once-weekly subcutaneous injection. In research handling it is described drawn with a small-gauge insulin syringe to the calculated unit mark, with sites rotated, and the reconstituted solution kept refrigerated between weekly draws - which a long half-life makes practical, since one vial covers several weeks at the doses studied. Foaming and vigorous agitation are avoided during preparation. This describes reported handling in a research setting and is not instruction for personal use.

Half-life & frequency rationale

Cagrilintide's plasma half-life is 159-195 hours - roughly 6.6 to 8.1 days - across the 0.16 to 4.5 mg dose range, with a median time to maximum concentration of 24-72 hours. That figure appears on a great many pages about this compound without a source attached, so it is worth naming one. It comes from Enebo et al., Lancet 2021;397(10286):1736-1748 (PMID 33894838, NCT03600480), a randomised phase 1b multiple-ascending-dose trial in 96 participants, where half-life was an exploratory pharmacokinetic endpoint measured from the last dose at week 19 through week 20. Exposure was dose-proportional (AUC0-168h from 926 to 24,271 nmol*h/L, Cmax from 6.14 to 170 nmol/L) and did not affect semaglutide's exposure or elimination. We checked this figure against the primary source rather than carrying it forward, and it holds. For context, semaglutide measured in the same study had a half-life of 145-165 hours - the two are close, which is one reason a fixed-dose weekly combination is straightforward to formulate. And the contrast with the earlier amylin analog is the point of the whole molecule: pramlintide requires three injections a day because of its short half-life, which is precisely the problem cagrilintide's lipidation was designed to solve. A half-life of about a week means steady state is reached over roughly four to five weeks of weekly dosing, which is also the window over which the trials escalated the dose.

Stacking — overview

The pairing that matters is with semaglutide, and it is not really a stack: CagriSema is a fixed-dose combination developed and tested as a single product, with five phase 3 trials behind it. The rationale is receptor-level - amylin receptors and GLP-1 receptors are separate targets converging on appetite - and the trials confirm the combination outperforms semaglutide alone on bodyweight while adding little on glycaemic control. Pairings with tirzepatide or retatrutide are described in community contexts on the same additive-mechanism reasoning. No trial has tested cagrilintide with any incretin agonist other than semaglutide, so those combinations have neither efficacy nor safety data behind them, and the tolerability arithmetic is unfavourable on its face: the adverse events of both classes are gastrointestinal and dose-related.

For how combinations are grouped by research context, the named blends, and why a pre-mixed blend vial cannot be calculated from its total milligrams, see the peptide stacks guide.

Storage & handling

• Lyophilized vials are described as stored refrigerated at about 2-8 C and protected from light, with short-term room-temperature stability of the sealed powder before use.
• After reconstitution, refrigerate at about 2-8 C, protect from light, do not freeze and do not shake. Because dosing is weekly, a reconstituted vial is typically held for several weeks - which makes bacteriostatic water rather than plain sterile water the relevant diluent, since the benzyl alcohol is what supports repeated entry of a multi-dose vial.
• Amylin's native tendency toward fibril formation is the reason the analog exists at all; avoid agitation and temperature cycling, which are the conditions that promote aggregation in peptides of this class.

Evidence File

The Cagrilintide Evidence File: Fifty-Three Records, Two Results Tables, One Head-to-Head

Cagrilintide has a larger clinical record than almost anything else in this library, and a surprising share of it has never been written up in a journal. This module works from the trial registry rather than from the papers: every record filed under the compound's several different names, what each one set out to measure, which ones have posted numbers, and where the only public result for a finished trial is a corporate press release rather than a paper or a registry table. It traces the most widely repeated figures back to the specific document each came from, reconciles the registry against the journals where the two disagree, and states each absence together with the exact searches that established it.

• 53 registry records under 4 names, 9 invisible to a plain name search
• 2 of 53 records carry a posted results table
• 3 finished tirzepatide comparisons, 1 result, and it is a press release
• 7 circulating figures traced back to their origin documents

Sourcing File

The Cagrilintide Sourcing File: A Two-Dalton Question and an Ambiguous Name

This module does not quote a reference database at cagrilintide; it rebuilds the identity block from the sequence and checks whether the published numbers agree. They mostly do, with one exception found inside a primary chemistry record itself. The result is a set of expected values a certificate of analysis can be read against - the exact mass a mass spectrometer should report, the ion series it will actually produce, the difference that distinguishes correctly folded material from a cheaper alternative, and the arithmetic connecting the milligram number printed on a vial label to the milligrams of peptide inside it.

• Formula C194H312N54O59S2 rebuilt from the sequence to an exact match
• 2.0157 Da separates the closed disulfide from the open dithiol
• Net peptide content spans 90.6% to 98.7% across plausible salt forms
• One development code resolves to an antibiotic in the chemistry database

Benefit & Outcome Review

The Cagrilintide Benefit and Outcome Review: Graded by Estimand, Population and Comparator

Most compounds in this library are graded by asking whether the evidence is human or animal. Cagrilintide inverts that: the human record is large and the preclinical record barely bears on any question a reader would ask. So the grading here turns on four other things - which statistical estimand a number came from, which population produced it, what it was compared against, and whether the trial that would settle it has reported. Ten claims are worked through on that basis, including the ones that fail, the ones that are untested in both directions, and the proportions of trial participants for whom the drug did not do what the headline says.

• 10 claims graded by estimand, population and comparator
• About 1 in 4 did not reach 5% loss at the dose carried forward
• 57.3% of the combination group were at target dose at week 68
• Fatigue risk ratio 3.2, or 92 additional cases per 1,000

Guide FAQ

Quick answers about guide scope, access, and educational use context.