On this page

CagriSema combines cagrilintide, an amylin analogue, with semaglutide, a GLP-1 receptor agonist. The published REDEFINE 1 and REDEFINE 2 trials tested the combination in different populations. Their results help explain its evidence base, but they are not interchangeable estimates for every person with obesity.

In the papers’ treatment-policy analyses, estimated average body-weight change at 68 weeks was −20.4% versus −3.0% with placebo in REDEFINE 1 and −13.7% versus −3.4% in REDEFINE 2. The first trial excluded diabetes; the second enrolled adults with type 2 diabetes. These are separate randomized comparisons with placebo, not a randomized comparison between those populations.

Sources checked: 25 October 2026. This reference explains the two 2025 primary publications. For other programmes and later results, use the obesity-drug clinical-trial results tracker. A trial result, a medicine’s registration and PBS subsidy are separate records.

REDEFINE 1 versus REDEFINE 2: the published results

Selected 68-week results from two separate phase 3a studies; primary weight estimates use the treatment-policy estimand
Evidence question REDEFINE 1 REDEFINE 2
Population Overweight with a related complication, or obesity; without diabetes. Overweight or obesity with type 2 diabetes.
Participants randomized 3,417 across four groups. 1,206 across two groups.
Design Double-blind; combination, each component alone and placebo. Double-blind; combination versus placebo.
Mean weight change: combination / placebo −20.4% / −3.0%. −13.7% / −3.4%.
Reported difference versus placebo −17.3 percentage points; 95% confidence interval −18.1 to −16.6. −10.4 percentage points; 95% confidence interval −11.2 to −9.5.
Registry identifier NCT05567796. NCT05394519.

Sources: Garvey and colleagues, REDEFINE 1 and Davies and colleagues, REDEFINE 2. The paper-reported differences are retained rather than recalculated from rounded displayed means. Registry links identify the studies; the numerical results above come from the papers.

Both studies included lifestyle intervention. Placebo here does not mean that participants received no background support.

What does the treatment-policy estimand mean?

An estimand defines the precise question an analysis is designed to answer, including how events such as stopping the assigned treatment are handled. Both papers describe their primary effect estimates as treatment-policy estimates, consistent with the intention-to-treat principle.

The practical reading rule is to keep the analysis label attached to its percentage. An estimate about assignment to a treatment strategy and an estimate under assumptions of continued treatment answer different questions. Picking the largest number from one analysis and comparing it with another study’s different analysis can make a headline misleading.

A trial average also describes a group, not a promised outcome. It does not tell a reader the result for a particular individual, the full spread of responses, or what would happen outside the trial’s eligibility criteria and follow-up. For a broader framework, read how to read a weight-loss drug clinical trial.

Why the two trials cannot be directly subtracted

Subtracting 13.7 from 20.4 does not establish the causal effect of having diabetes on weight loss. Participants were randomized within each study; they were not randomized to have or not have diabetes across the two studies.

The populations, background clinical circumstances and study questions differ. A sound comparison starts with the randomized contrast each protocol actually tested. These data support two population-specific findings, while a direct between-population claim would require a design and analysis suited to that question.

What adverse events did the studies report?

Reported gastrointestinal adverse events; these are trial-group event proportions, not individual risk predictions
Study CagriSema group Placebo group Source context
REDEFINE 1 79.6%. 39.9%. The report describes these events as mainly transient and mild to moderate.
REDEFINE 2 72.5%. 34.4%. The report describes most as transient and mild or moderate.

These selected figures do not constitute the whole safety assessment. Severity, serious events, discontinuations, the observation period and less common outcomes matter too. A short results table cannot establish that a combination is suitable for an individual or settle longer-term safety questions.

Why CagriSema is different from semaglutide alone

Semaglutide supplies the GLP-1 component; cagrilintide supplies a different amylin-related mechanism. A result for their combination should be attributed to the tested combination. It should not be relabelled as evidence for cagrilintide alone, semaglutide alone or another amylin compound.

Our amylin versus GLP-1 explainer maps the pathways and distinguishes CagriSema from petrelintide and eloralintide. The GLP-1 foundations guide explains the difference between a natural hormone and a medicine acting at its receptor.

What do these results mean for Australia?

A positive trial does not itself establish Australian registration or PBS subsidy. The TGA’s Wegovy registration record identifies semaglutide and that product’s indication; it is not registration evidence for a different combination.

Keep the evidence chain explicit: the paper addresses the studied intervention and endpoints; the TGA record addresses a specific registered product and indication; PBS records address subsidy under their own criteria. Read the Australian GLP-1 registration and PBS guide for that distinction. This article is an evidence reference, not a claim about current CagriSema availability in every jurisdiction.

Primary sources and related evidence

Citation record

How to cite this article

CagriSema Trial Results: What REDEFINE 1 and REDEFINE 2 Actually Show. Elite Peps Australia. Published 25 October 2026; updated 25 October 2026. https://elitepepsaustralia.com/cagrisema-redefine-1-redefine-2-trial-results/

Source checks: Consult the primary studies, trial records and regulator documents linked in this article for current evidence and status. Publication and update dates do not by themselves indicate an evidence review.