CompoundBrief

Research literature

Retatrutide Trial Results Explained: What the Phase 2 and Phase 3 Data Show

Published October 2026 · 9 min read

Scope: Retatrutide is an investigational compound that is not approved by the FDA or any other regulator. This article summarizes what published and announced clinical trials report, so that the numbers can be read critically. It is not medical advice, it is not a dosing or use guide, and nothing here describes how any individual should use any compound.

Few investigational molecules have produced as many headline figures as retatrutide, and most of those figures are quoted without the design details that give them meaning. A single number such as "28.3%" depends on who was enrolled, how long the trial ran, which dose arm it came from and how the result was calculated. This article walks through the main trials, what each one measured, and the reading rules that keep the numbers from being misread.

The short version

TrialPopulationLengthHeadline result
Phase 2 obesity (NEJM, 2023)338 adults, no diabetes48 weeksUp to 24.2% mean weight reduction (12 mg arm)
TRIUMPH-1 (phase 3, announced 2026)2,339 adults, no diabetes80 weeksUp to 28.3% mean weight reduction (12 mg arm)
TRANSCEND-T2D-1 (phase 3, topline)537 adults with type 2 diabetes40 weeksHbA1c down 1.7 to 2.0 points; weight down 11.5% to 16.8%

What retatrutide is

Retatrutide is a synthetic peptide engineered to act at three receptors at once: the glucagon-like peptide-1 (GLP-1) receptor, the glucose-dependent insulinotropic polypeptide (GIP) receptor and the glucagon receptor. Sponsors describe it as a single-molecule "triple agonist". It is given as a once-weekly subcutaneous injection in the trials.

The GLP-1 and GIP components are shared with earlier incretin-based drugs, which are associated with reduced appetite and improved glucose handling. The glucagon receptor is the distinguishing target. Sponsor and trial literature describes glucagon receptor activity as being of interest for its effects on hepatic fat metabolism (and, in preclinical work, energy expenditure), and the liver-fat results below are why that angle gets attention. How much each receptor contributes in people is still an active research question, which is why head-to-head and mechanism studies matter.

Phase 2: the NEJM trial

The phase 2 trial, published in the New England Journal of Medicine in 2023 (Jastreboff et al.), enrolled 338 adults with a BMI of 30 or higher, or 27 to under 30 with at least one weight-related condition. Mean baseline BMI was 37.3 and mean weight about 107.7 kg. Participants were randomized to weekly retatrutide at 1, 4, 8 or 12 mg, or placebo, with the higher doses reached by stepwise escalation. Some 4 mg and 8 mg groups used different starting doses, which is why the paper reports them as combined groups.

ArmMean change at 24 weeksMean change at 48 weeks
1 mg−7.2%−8.7%
4 mg (combined)−12.9%−17.1%
8 mg (combined)−17.3%−22.8%
12 mg−17.5%−24.2%
Placebo−1.6%−2.1%

Three details in this table matter more than the top number. First, the arms are clearly dose-ordered, which is what a real pharmacological effect looks like. Second, the 12 mg curve was still falling when the trial stopped: the authors wrote that participants continued to lose weight until treatment ended at week 48 and that a plateau had not yet been reached, so 48 weeks is a snapshot, not an endpoint. Third, every participant in the 8 mg and 12 mg groups lost at least 5% of baseline weight, and 26% of the 12 mg group lost 30% or more.

Safety findings from the paper were typical of the incretin class: gastrointestinal events were the most common and mostly occurred during dose escalation. Heart rate rose in a dose-dependent way up to week 24 and then declined, with increases the authors described as similar to those reported for GLP-1 receptor agonists. Skin sensitivity events (cutaneous hyperesthesia) were reported in 7% of retatrutide-treated participants and 1% of those on placebo, and were mild to moderate. Discontinuation because of adverse events ranged from 6% to 16% across retatrutide arms, against none on placebo.

Phase 3: TRIUMPH-1

TRIUMPH-1 is the first of the pivotal obesity trials to report. According to the sponsor's announcement, it randomized 2,339 adults with obesity or overweight and at least one weight-related condition, excluding people with diabetes, 1:1:1:1 to retatrutide 4 mg, 9 mg, 12 mg or placebo. The primary phase ran 80 weeks, with a pre-specified extension to 104 weeks. Mean baseline weight was 112.7 kg and mean BMI 40.0, a heavier population than the phase 2 trial. Every retatrutide arm began at 2 mg and moved up in four-week steps to its target dose, so the 12 mg arm had several months of escalation before it reached the top dose.

ArmEfficacy estimand, 80 weeksTreatment-regimen estimand, 80 weeks
4 mg−19.0%−17.6%
9 mg−25.9%−23.7%
12 mg−28.3%−25.0%
Placebo−2.2%−3.9%

In the 12 mg arm, the sponsor reported that 62.5% of participants lost at least 25% of body weight and 45.3% lost at least 30%, on the efficacy estimand. In a pre-specified extension of 532 participants with a baseline BMI of 35 or higher, the sponsor reported a mean reduction of up to 30.3% at week 104 in the group that had been on 12 mg and went to the maximum tolerated dose.

Reading the numbers: two estimands

The table above shows the same trial producing two different answers, and neither is wrong. They answer different questions.

The efficacy estimand asks what the drug does when participants stay on it as assigned. It is the figure used in most headlines, and it is the larger one. The treatment-regimen estimand asks what happens across everyone randomized regardless of whether they stayed on treatment, including people who stopped early or started other weight-management treatments. It is smaller, and it is closer to what a population would experience in practice. The placebo figure differs between the two analyses as well (−2.2% against −3.9%).

When two articles quote different numbers for the same trial, the estimand is the first thing to check. When a number is quoted with no estimand at all, treat it as the efficacy figure.

Why cross-trial comparisons mislead

It is tempting to line up 24.2% (phase 2, 48 weeks) against 28.3% (phase 3, 80 weeks) and call the difference a drug effect. It isn't a clean comparison. The trials differ in length, in baseline BMI, in dose-escalation schedules and in analysis method. A longer trial with a heavier population and a different escalation schedule will produce a different number even if the underlying pharmacology is identical. The same caution applies to comparing any compound's trial against another's. Only a trial that randomizes the two to the same population and measures the same endpoint answers that question.

Type 2 diabetes trials

The sponsor's first phase 3 diabetes trial, TRANSCEND-T2D-1, enrolled 537 adults with type 2 diabetes, baseline HbA1c of 7.0% to 9.5% and a mean diabetes duration of about 2.5 years. Participants were randomized to 4, 9 or 12 mg of retatrutide or placebo for 40 weeks, escalating from 2 mg in four-week steps. Topline results reported HbA1c reductions of 1.7 to 2.0 percentage points across doses against 0.8 on placebo, and weight reductions of 11.5% to 16.8% against 2.5% on placebo. Topline figures like these come from a press announcement, and the underlying analysis choices are only fully visible once the data are published.

Weight reduction in people with type 2 diabetes tends to be smaller than in people without it across the incretin class, which is one reason the diabetes and obesity trials are reported separately and shouldn't be averaged together. The earlier phase 2 diabetes trial (281 participants) compared retatrutide against dulaglutide and placebo and was presented at the American Diabetes Association meeting in 2023.

The liver-fat substudy

A substudy of 98 participants from the phase 2 obesity trial, with metabolic-dysfunction-associated steatotic liver disease (MASLD), measured liver fat by MRI. Presented in 2023, it reported mean relative reductions in liver fat at 24 weeks of 42.9% (1 mg), 57.0% (4 mg), 81.4% (8 mg) and 82.4% (12 mg), against a 0.3% increase on placebo. Investigators also reported that most participants at the two highest doses reached a liver-fat fraction of 5% or less, which is the usual cutoff for "normal". It is a small substudy and an early signal, and liver fat tracks weight loss closely, so separating a direct hepatic effect from the effect of weight loss takes dedicated study designs.

Safety signals reported in TRIUMPH-1

Reported adverse events were dose-related and mostly gastrointestinal. Numbers below are the sponsor's, for the 4, 9 and 12 mg arms against placebo.

Event4 / 9 / 12 mgPlacebo
Nausea28.6% / 38.4% / 42.4%14.8%
Diarrhea25.2% / 34.1% / 32.0%13.5%
Vomiting10.6% / 22.8% / 25.3%4.8%
Dysesthesia5.1% / 12.3% / 12.5%0.9%
Discontinued for adverse events4.1% / 6.9% / 11.3%4.9%

Dysesthesia, an abnormal skin sensation, is the event that stands out because it was reported in roughly 12% of the two higher-dose arms against under 1% on placebo. It was reported as mild to moderate, and the sponsor stated that most cases resolved during treatment. The 12 mg discontinuation rate is more than double the placebo rate, which is a reminder that the top-line efficacy figure comes with a tolerability cost that grows with dose.

What isn't known yet

TRIUMPH-1 is, for now, a sponsor announcement and conference presentation, and peer-reviewed publication is still to come. The knee osteoarthritis and sleep apnea sub-trials have not reported. TRIUMPH-2 (obesity with type 2 diabetes) and TRIUMPH-3 (obesity with cardiovascular disease) are expected later in 2026. Long-term safety, durability after treatment stops, body-composition effects and cardiovascular outcomes are open questions that these trials can't fully settle. Figures in this article may be revised as the full papers appear.

A note on reference materials

Because retatrutide has not been approved anywhere, laboratories that work with it do so as a research material, and the quality questions are the same ones that apply to any synthetic peptide. Identity should be confirmed by mass spectrometry, purity by HPLC, and quantity by a content test, and a report should name the lab and the lot. The sections of the testing guide and the COA primer cover what each of those results does and doesn't establish, and why purity isn't content explains the most commonly misread line.

Sources

← Back to all articles