CagriSema REDEFINE-1 results explained vs the next-gen GLP-1 family

What CagriSema is, in one paragraph

CagriSema is a once-weekly combination of cagrilintide, a long-acting analog of the hormone amylin, plus semaglutide, the GLP-1 drug marketed as Wegovy and Ozempic. It targets two fullness pathways at once, amylin in the brainstem and GLP-1 across the gut and brain, for more weight loss than either drug alone. It is investigational, not yet FDA-approved.

Cagrilintide is an investigational peptide from Novo Nordisk that imitates amylin, a hormone your pancreas releases alongside insulin when you eat to signal fullness. Semaglutide is the GLP-1 (glucagon-like peptide-1, a gut hormone that lowers appetite and slows digestion) agonist marketed as Wegovy and Ozempic. CagriSema combines both in one once-weekly injection. The bet: two different fullness signals beat one. The phase 3 trials tested that.

The peer-reviewed papers for the two pivotal trials, REDEFINE-1 (obesity without diabetes) and REDEFINE-2 (obesity with type 2 diabetes), came out in the New England Journal of Medicine in 2025 [1][2]. The results were not what the market had priced in, and most of the headlines that followed conflated the press release with the peer-reviewed paper. This post walks what the actual papers say, in plain English, and puts CagriSema next to its closest competitors: semaglutide, tirzepatide, retatrutide, and survodutide.

What REDEFINE-1 actually showed

REDEFINE-1 randomized 3,417 adults with obesity, without type 2 diabetes, to CagriSema 2.4 mg / 2.4 mg, cagrilintide alone, semaglutide alone, or placebo for 68 weeks. The combination produced the largest mean weight loss, beating both single drugs and placebo, with more participants hitting the ≥5%, ≥20%, ≥25% and ≥30% thresholds than placebo. The benefit over semaglutide alone was modest.

REDEFINE-1 enrolled 3,417 adults with obesity or overweight without type 2 diabetes and randomized them across four arms over 68 weeks [1]. One arm got CagriSema at the trial dose of 2.4 milligrams of cagrilintide and 2.4 milligrams of semaglutide, co-formulated in a single weekly injection. The other three arms got cagrilintide alone at 2.4 milligrams, semaglutide alone at 2.4 milligrams, or placebo. All four groups also received lifestyle intervention alongside the injection [1].

The CagriSema arm delivered the largest mean weight loss of the four, exceeding both cagrilintide and semaglutide as standalone drugs. The responder distribution, the share of participants who hit the ≥5%, ≥20%, ≥25%, and ≥30% weight-loss thresholds, was higher on CagriSema than on placebo at every one of those thresholds [1]. That is the clinically important number, because mean weight loss can hide a wide spread: the mean is what the middle of the curve does, but a real person is somewhere on the curve, not at the mean.

The cardiometabolic secondaries, blood pressure, lipids, waist circumference, inflammatory markers, all moved in the expected direction for the weight lost. The dedicated blood-pressure substudy (Verma 2026) confirmed meaningful systolic and diastolic reductions [3]. Gastrointestinal events, nausea, vomiting, diarrhea, constipation, were the dominant tolerability issue and were titration-dependent, the familiar GLP-1 class pattern with the amylin arm added on top rather than smoothed out.

Why the press release crashed Novo Nordisk's stock and the paper did not

On December 20, 2024 Novo Nordisk released a topline of about 22.7% mean weight loss at 68 weeks. That was below what investors had expected, and the stock fell sharply. The peer-reviewed Garvey 2025 NEJM paper landed months later with similar headline numbers but full context. The lesson: cite the published paper, never the press release.

On December 20, 2024, Novo Nordisk announced topline REDEFINE-1 results in a press release ahead of any journal publication. The topline number, 22.7% mean weight loss at 68 weeks if all participants had adhered to treatment (what Novo Nordisk called the trial product estimand) [15], was below where most analysts had set their target. The comparison investors were running in their heads was unfair on multiple axes: tirzepatide SURMOUNT-1 reported 20.9% at 72 weeks at the 15 mg dose under its treatment-regimen estimand [4] and the retatrutide phase 2 trial reported about 24.2% at 48 weeks [5]. On a quick read, that put CagriSema only a couple of points past an approved drug and short of a phase 2 number.

The stock fell sharply, one of the largest single-day declines for the company in years. Several months later the peer-reviewed paper arrived in NEJM and confirmed the topline directionally while giving the full estimand definitions, statistical analysis plan, secondary endpoints, and adverse-event detail. The canonical numbers for any educator or clinical write-up should come from the Garvey 2025 NEJM paper [1], not the press release. This is a general rule that applies to every late-stage trial in this space: a topline press release is a marketing artifact written for shareholders; the peer-reviewed paper is the document the field judges the drug on.

What is the amylin arm actually doing

Cagrilintide acts on the amylin receptor complex, a hybrid receptor made from the calcitonin receptor plus an accessory protein called RAMP. One of its primary binding sites is the area postrema, a small brainstem region that samples circulating fullness signals, among others. Cagrilintide there reduces meal size, slows gastric emptying, and suppresses postprandial glucagon, effects mechanistically distinct from GLP-1's action.

The reason CagriSema is not just "more semaglutide" is that cagrilintide targets a different receptor system. Amylin is a 37-amino-acid hormone your pancreas releases alongside insulin after a meal. It signals fullness by acting on the amylin receptor complex, a combination of the calcitonin receptor plus a small accessory protein called RAMP (receptor activity-modifying protein). One of the complex's primary binding sites is the area postrema, a small brainstem region with a leaky blood-brain barrier that samples circulating signals, including fullness signals [6].

Three effects matter. First, cagrilintide reduces meal size by engaging the area postrema satiety circuit. Second, it slows gastric emptying through the vagus nerve (the main nerve that connects your brain to your gut and heart). Third, it suppresses postprandial glucagon (the hormone that raises blood sugar between meals), improving postprandial glucose. Boyle and colleagues describe these as amylin's own physiological actions, signalled through the calcitonin-receptor-plus-RAMP complex and the caudal hindbrain [7]. Because that is a different receptor system from the one semaglutide acts on, the combination is designed to add a second lever rather than a bigger dose of the same one.

The press-release vs paper number, decoded

Trial weight loss is reported under different statistical conventions called estimands. The treatment-policy estimand counts everyone, even those who stopped early. The efficacy estimand counts only those who stayed on drug. The two can differ by 1-3 percentage points. When you read a trial number, look up which estimand it came from before comparing across trials.

One more nuance worth getting right. Modern obesity trials report two main weight loss numbers because of estimands, statistical conventions for what question the trial is actually answering. The FDA's guidance on estimands is the ICH E9(R1) addendum, dated May 2021 [17]. The treatment-policy estimand answers "what happens to everyone we randomized to this drug, regardless of whether they stopped taking it." The efficacy estimand answers "what happens to people who actually stayed on the drug through the trial."

The two numbers differ because some participants stop early and the trial has to decide how to handle them. The efficacy estimand is usually 1 to 3 percentage points higher. Cross-trial comparisons go wrong when one trial is quoted under treatment policy and another under efficacy. Always check which one the number came from before claiming one drug beats another.

This programme makes that unusually easy to get wrong, because Novo Nordisk has used three different labels across the trials quoted in this post: treatment policy in the REDEFINE-1 paper [1], trial product in the REDEFINE-1 topline announcement [15], and efficacy in the REDEFINE 4 release [18]. Each source defines its own, and the definitions are not identical, so the safe habit is to carry every number with the label its own source gave it rather than translating between them.

How CagriSema slots against the rest of the next-gen family

CagriSema, tirzepatide, retatrutide, and survodutide all pair GLP-1 with a second hormonal arm: amylin, GIP, both GIP and glucagon, and glucagon respectively. Two head-to-head trials against tirzepatide have reported. At full doses in obesity, REDEFINE 4 put CagriSema at 23.0% against 25.5% and it missed non-inferiority; at lower doses in type 2 diabetes, REIMAGINE 5 favoured CagriSema.

The next-generation obesity drugs are all variations on the same idea: take a GLP-1 agonist and pair it with a second hormonal arm that hits a different lever. Tirzepatide adds GIP (glucose-dependent insulinotropic peptide, an incretin that helps fat tissue store fat more efficiently and improves insulin sensitivity). Survodutide adds glucagon receptor agonism, which raises energy expenditure and burns liver fat, the same lever that makes survodutide the standout in liver-disease trials. Retatrutide adds both GIP and glucagon to GLP-1, the so-called triple agonist. CagriSema adds amylin (central satiety, gastric-emptying delay, and glucagon suppression).

Mean weight loss in the published phase 3 trials clusters in a tight band. Semaglutide STEP-1 reported about 14.9% at 68 weeks [8]. Tirzepatide SURMOUNT-1 reported 20.9% at 72 weeks at the 15 mg dose, treatment-regimen estimand [4]. CagriSema REDEFINE-1 reported 20.4% at 68 weeks under the treatment-policy estimand in the NEJM paper [1], and 22.7% under the trial product estimand in the topline announcement [15]. Retatrutide TRIUMPH-1 topline announced about 28.3% at the 12 milligram dose at 80 weeks under its efficacy estimand, and 25.0% under its treatment-regimen estimand [9]. The trial was published in the NEJM in September 2026, reporting the treatment-regimen 25.0% [24]; what the data mean in practice is covered in our retatrutide TRIUMPH-1 results breakdown. Survodutide's peer-reviewed phase 2 dose-finding trial reported about 14.9% at 46 weeks [13], and its phase 3 SYNCHRONIZE-1 trial, which randomized 725 adults to 76 weeks of treatment [14], reported 13.0% at the 6.0 mg dose under the treatment-regimen estimand in its NEJM paper [23]. These are different durations, different populations and, as the previous section explains, different estimands, so the numbers are not strictly comparable.

One of those comparisons no longer has to be made across trials. REDEFINE 4 put CagriSema directly against tirzepatide 15 mg, and it is registered as NCT06131437 [19]. It enrolled 809 adults with obesity and one or more comorbidities, with a mean baseline body weight of 114.2 kilograms, 84 weeks of treatment, and an open-label design, meaning participants and investigators both knew which drug was being given [18]. Novo Nordisk reported the result on February 23, 2026. Under the efficacy estimand, CagriSema produced 23.0% mean weight loss against tirzepatide's 25.5% [18]. Under the treatment-regimen estimand, 20.2% against 23.6% [18]. The trial's primary endpoint was to show that CagriSema was non-inferior to tirzepatide, meaning not meaningfully worse, and the trial did not achieve it [18].

That is a miss against a bar Novo Nordisk set for itself before the trial ran, not a photo finish, and the point estimates favour tirzepatide under both estimands. Two caveats cut the other way and are worth holding. The release gives no confidence interval or p-value for the comparison itself, so the size of the gap is less certain than its direction. And an open-label trial is the weaker design here: when everybody knows which drug they are on, expectations can move behaviour and reporting in ways a blind trial prevents. The comparisons that remain fully cross-trial are the rest of them, including every one in this post between CagriSema, retatrutide and survodutide. TRIUMPH-5, the retatrutide against tirzepatide head-to-head, had not reported as of this writing [20].

A second CagriSema against tirzepatide trial reported on September 21, 2026, and at lower doses it points the other way [21]. REIMAGINE 5 (NCT06534411) compared CagriSema 1.0/1.0 mg with tirzepatide 5 mg in 1,023 adults with type 2 diabetes treated with metformin, an SGLT2 inhibitor or both, again open-label [22]. At week 60, on the efficacy estimand, CagriSema produced 12.4% weight loss against 9.1%, which Novo Nordisk reports as superior, and an HbA1c reduction of 1.71% against 1.67%, which it reports as non-inferior [21]. The two trials test different doses in different populations. At full doses in obesity, tirzepatide came out ahead; against tirzepatide 5 mg, well below its 15 mg top dose, CagriSema did [18] [21].

Use the projection tool below to put your own starting weight against any of the five. Every efficacy figure in it comes from that drug's published phase 3 paper, on the estimand the paper leads with: treatment-policy or treatment-regimen, both of which count everyone randomized whether or not they stayed on the drug. It turns the mean into pounds for your starting weight and shows the share of participants who reached each weight-loss threshold, where the paper's abstract reports it. The trials still differ in length, dose, and population, so the five cards are a cross-trial comparison, not a ranking. The GLP-1 comparison tool shows the approved GLP-1 drugs and retatrutide in a side-by-side feature table; this tool is the personalized projection.

What CagriSema does not yet have

CagriSema has phase 3 efficacy in obesity and type 2 diabetes, but its published papers do not yet cover cardiovascular outcomes, MASH liver disease, kidney outcomes, or anything past 68 weeks. Its full-dose head-to-head trial against tirzepatide missed its endpoint. Semaglutide has cardiovascular outcomes data and a MASH approval; tirzepatide is approved for diabetes, weight management, and sleep apnea.

The CagriSema evidence base is narrower than the broader question implies. For a side-by-side feature comparison of the approved GLP-1 agents that CagriSema is being measured against, the tirzepatide beyond the dosing chart piece covers muscle preservation, the plateau problem, and the off-ramp question, considerations that will apply equally to CagriSema if it reaches approval. The published phase 3 program covers obesity (REDEFINE-1), obesity plus T2D (REDEFINE-2), and an East Asian regional confirmatory trial (REDEFINE-5 in Japan and Taiwan) [10]. None of the major hard endpoints sit in the dataset yet. The cardiovascular outcomes trial (CVOT, a trial that asks whether the drug reduces actual heart attacks, strokes, and cardiovascular deaths over years of use) is REDEFINE 3, which enrolled 7,101 participants and is still running, with completion estimated for 2027 [16]. Nothing in the three phase 3 papers cited here addresses MASH (metabolic-dysfunction-associated steatohepatitis, a liver disease that follows fat accumulation and inflammation in the liver), kidney outcomes, or durability past their 68-week endpoint [1] [2] [10].

Semaglutide already has the cardiovascular data: SELECT showed a 20% reduction in major adverse cardiovascular events in adults with obesity and established cardiovascular disease but without diabetes [11]. It also has a MASH approval. Tirzepatide is approved for type 2 diabetes, chronic weight management, and obstructive sleep apnea (OSA) in adults with obesity; its MASH evidence is still phase 2. CagriSema has none of those yet. The working assumption is that adding cagrilintide does not undo semaglutide's CV benefit, reasonable but unproven until REDEFINE 3 reads out. One gap on this list has closed since this post first ran, and it did not close in CagriSema's favour: head-to-head data against tirzepatide now exist, and the full-dose trial missed its non-inferiority endpoint [18].

Realistic interpretation, in three sentences

CagriSema is a strong weekly weight-loss drug with mechanistic novelty in the amylin arm and a peer-reviewed phase 3 evidence base in NEJM. It is not best-in-class: at full doses, the head-to-head trial put it behind tirzepatide, missing non-inferiority. It is not approved anywhere, though a new drug application is under FDA review.

The honest framing is that CagriSema is a good drug that came second. The amylin arm is a real mechanistic story with preclinical and clinical support, and the combination delivers meaningful additional weight loss on top of semaglutide alone. But when the two drugs were put in the same trial at full doses, tirzepatide produced more weight loss and CagriSema missed the bar it had set for itself [18]. An earlier version of this post argued for competitive parity on the grounds that no head-to-head trial existed. Two do now. In the full-dose trial, REDEFINE 4, CagriSema missed its primary endpoint against tirzepatide [18]. Its win in REIMAGINE 5 came against tirzepatide 5 mg [21], a third of the 15 mg top dose tested in SURMOUNT-1 [4].

What survives is narrower and more useful. CagriSema is a genuinely effective once-weekly injection built on a mechanism no approved obesity drug currently uses, and a second, different satiety lever is a real addition to a class that has otherwise been variations on GLP-1. What it is not is the front-runner. Tirzepatide outperformed it directly at full doses [18], and retatrutide is sitting on higher phase 3 figures [9]. On the evidence as it stands, this field has a leader, and CagriSema is not it.

For a complete beginner who wants to understand the broader GLP-1 landscape before committing to any one drug, our GLP-1 comparison tool puts semaglutide, tirzepatide, liraglutide, and retatrutide side by side with evidence tiers and cost data. And for anyone trying to understand where investigational compounds like CagriSema sit in the FDA approval pipeline, the FDA category-2 reclassification guide explains the regulatory framework that determines what can be compounded and what cannot. The semaglutide course walks the GLP-1 backbone every drug on this list is built on. The tirzepatide course, retatrutide course, and survodutide course cover the other second-arm strategies. The cagrilintide course covers the amylin arm itself, from amylin biology to the CagriSema rationale.