Retatrutide in 2026: what five phase 3 readouts show—and what remains unknown
A source-graded update to the existing research dossier, bringing together one peer-reviewed phase 3 publication, four sponsor- or conference-disclosed phase 3 readouts, new peer-reviewed human analyses, and the limits that matter most.
This article adds to the existing evidence dashboard. Earlier phase 1 and phase 2 findings remain available there and have not been removed or rewritten.
RetaGrade evidence intelligence
Source status stays visible from finding to finding.
Peer-reviewed result
Full journal publication with methods and a PubMed record. This is the strongest result tier used in the update.
Sponsor or conference disclosure
A primary company or congress source, but not yet a complete peer-reviewed results paper. These findings remain preliminary.
Directness matters
A biomarker, post-hoc analysis, design paper, or study of another molecule cannot establish a retatrutide treatment effect.
Five phase 3 readouts
The development program is now broader than the earlier dashboard snapshot
TRANSCEND-T2D-1 is peer-reviewed. TRIUMPH-1 through TRIUMPH-4 are shown from primary sponsor or congress disclosures and remain explicitly separated from the published result.
TRANSCEND-T2D-1
Type 2 diabetes
-15.3%
12 mg · week 40 · mean weight change
The peer-reviewed phase 3 publication also reported an HbA1c change of −1.94 percentage points at 12 mg.
TRIUMPH-1
Obesity without diabetes
-28.3%
12 mg · week 80 · mean weight change
A 12 mg extension subset was reported at −30.3% at week 104. That subset is not the full randomized population.
TRIUMPH-2
Obesity plus type 2 diabetes
-20.8%
12 mg · week 80 · mean weight change
The disclosure reported HbA1c reductions up to −1.6 percentage points across retatrutide arms.
TRIUMPH-3
Obesity plus cardiovascular disease
-22.6%
12 mg · week 80 · mean weight change
The disclosure also reported secondary cardiometabolic biomarkers, including hsCRP, lipids, blood pressure, and waist circumference.
TRIUMPH-4
Obesity plus knee osteoarthritis
-28.7%
12 mg · week 68 · mean weight change
WOMAC pain changed by −4.4 points at 12 mg and −4.5 points at 9 mg in the sponsor disclosure.
Highest-dose mean weight change reported in each readout
The bars make the program easier to scan; they do not make the studies directly comparable.
TRANSCEND-T2D-1
T2D · 12 mg · 40 weeks
-15.3%
TRIUMPH-1
No T2D · 12 mg · 80 weeks
-28.3%
TRIUMPH-2
T2D · 12 mg · 80 weeks
-20.8%
TRIUMPH-3
CVD · 12 mg · 80 weeks
-22.6%
TRIUMPH-4
Knee OA · 12 mg · 68 weeks
-28.7%
Interpretation: These are different populations, durations, estimands, and source states. This is not a head-to-head ranking and should not be used to infer comparative superiority.
Cardiometabolic signals
TRIUMPH-3 disclosed large secondary biomarker changes
At the highest dose, the sponsor reported changes in inflammatory and lipid biomarkers alongside weight, waist, and blood-pressure outcomes. They are useful signals, not proof of disease modification.
High-sensitivity C-reactive protein
hsCRP · secondary inflammation biomarker
-51.2%
Triglycerides
Secondary lipid biomarker
-37%
Non-HDL cholesterol
Secondary lipid biomarker
-16.5%
Do not translate hsCRP into an inflammation-treatment claim. hsCRP is a nonspecific biomarker. A reduction does not demonstrate that retatrutide treats endometriosis, an autoimmune disorder, or any inflammatory disease.
−9.3
mmHg systolic blood-pressure change
Highest-dose result disclosed in TRIUMPH-3; not a cardiovascular-outcomes result.
−19.0 cm
Waist-circumference change
Highest-dose result disclosed in TRIUMPH-3.
New peer-reviewed human analyses
Four publications add detail beyond body weight alone
These publications analyze existing phase 2 participants or samples. They expand the mechanistic and surrogate-endpoint picture, but none is a new hard-outcomes trial.
Kidney parameters and albuminuria
A post-hoc phase 2 analysis reported UACR changes of −37.0% versus placebo in the 12 mg T2D group and −28.0% and −31.5% in the 8 mg and 12 mg obesity groups. Most participants began with normal albuminuria, so absolute changes were modest; this does not establish kidney protection.
Appetite and eating behavior
In a prespecified exploratory analysis of 275 phase 2 T2D participants, doses of at least 4 mg were associated with lower hunger and food-consumption measures. Correlations with weight change were modest, not proof of a single causal pathway.
ANGPTL3/8 and lipid biology
A post-hoc clinical analysis paired with hepatocyte experiments found ANGPTL3/8 changes that paralleled triglyceride and LDL changes. It supports a mechanism hypothesis, not a clinical-outcomes conclusion.
Lipid and metabolomic signatures
An exploratory post-hoc analysis described metabolic clusters related to fat oxidation and insulin resistance. The findings are hypothesis-generating and require prospective validation.
Phase 3 safety refresh
The newer disclosures keep gastrointestinal events and discontinuation visible
These are trial-specific ranges across disclosed retatrutide arms. They are displayed separately rather than pooled, because populations, exposure, titration, and reporting differ.
Disclosed · not peer-reviewed
TRIUMPH-1
- Nausea
- 28.6–42.4%
- Diarrhea
- 25.2–34.1%
- Vomiting
- 10.6–25.3%
- Dysesthesia
- 5.1–12.5%
- Urinary-tract infection
- 7.5–8.8%
- Discontinued because of an adverse event
- 4.1–11.3%
Disclosed · not peer-reviewed
TRIUMPH-2
- Diarrhea
- 27.4–33.6%
- Nausea
- 13.7–28.0%
- Vomiting
- 5.5–15.7%
- Dysesthesia
- 4.5–7.3%
- Urinary-tract infection
- 3.8–8.0%
- Discontinued because of an adverse event
- 3.8–11.6%
Disclosed · not peer-reviewed
TRIUMPH-3
- Diarrhea
- 24.4–30.1%
- Nausea
- 21.7–22.4%
- Dysesthesia
- 6.4%
- Urinary-tract infection
- 6.1–7.0%
- Discontinued because of an adverse event
- 9.8–13.5%
How to read these ranges: They are not pooled incidence estimates and should not be compared as if the trials were randomized against each other. Only events and values directly disclosed in the cited primary sources are repeated here. A reported urinary-tract infection rate does not establish causality.
Inflammation, endometriosis, and autoimmunity
The new evidence sharpens the boundary; it does not fill these gaps
The most responsible update is to show exactly where the retatrutide evidence ends and where adjacent hypotheses begin.
Biomarker only
Inflammation
TRIUMPH-3 disclosed a 51.2% hsCRP reduction at the highest dose. hsCRP is nonspecific and secondary here; there is no direct trial showing treatment of an inflammatory condition.
No direct retatrutide evidence
Endometriosis
The June 2026 paper located in the update concerns tirzepatide, not retatrutide, and is a review and mechanistic hypothesis. Its authors state that direct validation is absent. It should not be presented as a retatrutide result.
Open the adjacent 2026 paperNo dedicated subgroup
Autoimmune conditions
No retatrutide efficacy analysis was found for a defined autoimmune-disease subgroup. The phase 2 obesity protocol also excluded significant active autoimmune disease requiring, or likely to require, systemic glucocorticoids, limiting generalizability.
Review the phase 2 protocol dossierTRIUMPH-3 does not establish cardiovascular risk reduction
The disclosed MACE-5 and MACE-3 confidence intervals both cross 1.0. These data are inconclusive and do not replace a dedicated cardiovascular-outcomes trial.
MACE-5
HR 0.82
95% CI 0.55–1.22
MACE-3
HR 1.12
95% CI 0.64–1.96
Development status verified July 30, 2026
Retatrutide remains investigational and is not FDA-approved. Lilly has said it plans a biologics license application in the first quarter of 2027; a filing plan is not approval.
Two additional phase 1 records were verified: NCT06982846 completed with 80 participants and NCT06982859 is active, not recruiting with an estimated 95 participants. Neither registry record has posted results, so they add development context—not efficacy findings.
Source ledger
Every new claim points to its source and source type
Primary publications, trial registries, and official sponsor disclosures are kept distinct. The list also includes design papers and the adjacent endometriosis article so their limits remain visible.