# Retatrutide: an atlas of the evidence Evidence reviewed: 5 September 2026. Investigational · not approved. A once-weekly investigational medicine, studied for obesity and diabetes. Explore its effects on weight, liver fat, blood sugar, pain and sleep, with the evidence and its limits. ## Weight Weight loss is the most extensively studied effect. The size of the change depends on the population, dose, follow-up and how the analysis handles people who stop treatment. ## Liver MRI measurements show substantial reductions in liver fat. The fraction reaching a low liver-fat threshold answers a different question from whether inflammation, scarring or future liver failure improve. ## Blood sugar HbA1c reflects average blood glucose over the preceding months. A fall of two percentage points is substantial, but reaching a normal value while taking a medicine does not establish drug-free diabetes remission. ## Knee & movement People with obesity and knee osteoarthritis reported less pain and better function. WOMAC measures symptoms and daily activities. It does not measure cartilage regrowth. ## Sleep & breathing The sleep-apnea basket within TRIUMPH-1 measured breathing disturbances overnight. A lower event rate is meaningful, but residual disease and the need for other treatment must be judged individually. ## Heart & vessels Blood pressure, lipids and some inflammatory markers improved. The available cardiovascular-event analyses do not yet establish prevention of heart attacks, stroke or death. Read the event results alongside the biomarkers. ## Kidneys Urinary albumin and estimated filtration moved favorably in exploratory analyses. These findings justify dedicated kidney trials; they do not yet prove prevention of kidney failure. ## Fat & lean mass Body-composition measurements distinguish fat from lean tissue. Lean mass includes water and organs as well as muscle. Weight loss should not be described as exclusively fat loss. ## Appetite Questionnaires capture changes in hunger and eating behavior. These participant-reported results add context to the scale, but correlations with weight loss cannot identify the direction of causation. ## Brain & cognition Animal and laboratory research explores effects beyond the established trial endpoints. A biological possibility is worth understanding without treating it as a demonstrated benefit in people. ## Other early research Laboratory and animal studies explore liver injury and obesity-associated tumors. These findings suggest research directions. They do not establish cancer prevention, cancer treatment or reversal of human liver scarring. ## Three receptors. One engineered peptide. Retatrutide is one molecule with activity at three hormone receptors. It is not a mixture of three medicines, and “GLP-3” is not a fourth hormone. ### GLP-1R: Glucose & appetite GLP-1 receptor signaling contributes to glucose-dependent insulin secretion and appetite regulation. ### GIPR: A second incretin signal GIP receptor activation adds another glucose-dependent insulin signal. Its contribution depends on the wider metabolic context. ### GCGR: The glucagon question Glucagon signaling affects liver fuel handling. Increased energy expenditure helps explain the findings in mice; its exact contribution to human weight loss is still being investigated. Structural work shows how retatrutide accommodates all three receptors. Its chemical modifications and fatty-acid attachment help make a long-acting peptide. Receptor potency measured in a laboratory is not a percentage of the clinical benefit. [Li et al. · Cell Discovery · 2024](https://www.nature.com/articles/s41421-024-00700-0) [Coskun et al. · Cell Metabolism · 2022](https://pubmed.ncbi.nlm.nih.gov/35985340/) [Urva et al. · The Lancet · 2022](https://pubmed.ncbi.nlm.nih.gov/36354040/) [Aronne et al. · SURMOUNT-5 · NEJM · 2025](https://doi.org/10.1056/NEJMoa2416394) [FDA · Unapproved GLP-1 drugs](https://www.fda.gov/drugs/drug-alerts-and-statements/fdas-concerns-unapproved-glp-1-drugs-used-weight-loss) [Lilly · TRIUMPH-2 / TRIUMPH-3 · 23 Jul 2026](https://investor.lilly.com/news-releases/news-release-details/lillys-triple-agonist-retatrutide-successful-two-additional) [Lilly · TRIUMPH-1 / TRANSCEND-T2D-1 · 6 Jun 2026](https://www.prnewswire.com/news-releases/lillys-triple-agonist-retatrutide-drove-substantial-improvements-in-weight-a1c-knee-osteoarthritis-pain-and-obstructive-sleep-apnea-demonstrating-its-remarkable-potential-to-treat-obesity-and-its-complications-302793169.html) ## A percentage needs its denominator. ### Which analysis? An efficacy estimand asks about a scenario in which participants remain on treatment, under the trial’s specified assumptions. A treatment-regimen estimand includes the effects of stopping treatment. Neither is a promise of what one person will experience. ### Which dose and population? The largest weight, pain and sleep effects need not come from the same dose. A trial in people with diabetes cannot be treated as the same experiment as one without diabetes. ### Which kind of benefit? A biomarker can improve without proving fewer future clinical events. Less liver fat does not establish fibrosis reversal. Less knee pain does not demonstrate restored cartilage. ### How much missing information? Small substudies, absent scans, multiple exploratory endpoints and selective extension cohorts deserve visible context. Repeated publications from one trial do not constitute independent replication. ## Related medicines. Different questions. SURMOUNT-5 directly compared tirzepatide with semaglutide in adults with obesity without diabetes. Retatrutide was not an arm. TRIUMPH-5 is the dedicated retatrutide–tirzepatide comparison; its status belongs in the study library. Placing percentages from different trials on a podium would conceal differences in participants, duration and analysis. ## The benefits come with a tolerability cost. ### Common symptoms Nausea, diarrhea, constipation and vomiting recur across studies. Their frequency varies by dose and trial. Treatment discontinuation is a practical part of the result. ### Altered skin sensation Dysesthesia, an unpleasant or unusual skin sensation, appears in phase 3 reports. It deserves attention alongside the better-known gastrointestinal effects. ### Heart rate and serious events Early studies reported dose-related heart-rate increases. Serious adverse events need their trial-specific counts and context; a study can be too small or short to detect rare harms. ### Lean tissue and durability Lean tissue can be lost during weight reduction. Longer follow-up and post-treatment data are needed to understand sustained benefit, function and what happens after stopping. Retatrutide remains investigational at this review date. The FDA warns about unapproved products sold as retatrutide. Study doses describe research arms, not instructions for self-treatment. There is no verified retail price or guaranteed approval date. ## The study library ### Phase 2: obesity without diabetes (2023) Peer-reviewed. Adults with obesity, or overweight and a weight-related condition; no diabetes.. 48 weeks. This randomized trial established the large weight-loss signal that led to phase 3. Weight continued to fall through the final visit. Waist circumference also decreased. Gastrointestinal symptoms were common, especially during dose escalation. Heart rate increased in a dose-dependent pattern, peaked around week 24, and subsequently declined. - **Body weight change**: −8.7% . Body weight change Placebo: -2.1 %. 1 mg; 48 weeks. Model-estimated change from baseline; efficacy analysis.. Confidence interval not given in the cited summary. - **Body weight change**: −17.1% . Body weight change Placebo: -2.1 %. 4 mg pooled; 48 weeks. Model-estimated change from baseline; efficacy analysis.. Confidence interval not given in the cited summary. - **Body weight change**: −22.8% . Body weight change Placebo: -2.1 %. 8 mg pooled; 48 weeks. Model-estimated change from baseline; efficacy analysis.. Confidence interval not given in the cited summary. - **Body weight change**: −24.2% . Body weight change Placebo: -2.1 %. 12 mg; 48 weeks. Model-estimated change from baseline; efficacy analysis.. Confidence interval not given in the cited summary. - **Body weight change**: −17.5% . Body weight change Placebo: -1.6 %. 12 mg; 24 weeks. Model-estimated change from baseline; efficacy analysis.. Confidence interval not given in the cited summary. - **Waist circumference change**: −19.6 cm. Waist circumference change Placebo: -2.6 cm. 12 mg; 48 weeks. Model-estimated change from baseline; efficacy analysis.. Confidence interval not given in the cited summary. - **Participants losing at least 15% of body weight**: 83% . Participants losing at least 15% of body weight Placebo: 2 %. 12 mg; 48 weeks. Model-estimated change from baseline; efficacy analysis.. Confidence interval not given in the cited summary. What this cannot tell us: A small phase 2 trial cannot establish long-term safety or superiority over drugs tested in other populations. The weight results are group averages; there was no semaglutide or tirzepatide arm. Funding: Funded by Eli Lilly and Company; sponsor employees contributed to the research. [Primary source](https://www.nejm.org/doi/full/10.1056/NEJMoa2301972) ### Phase 2: type 2 diabetes (2023) Peer-reviewed. Adults with type 2 diabetes and HbA1c 7.0–10.5%, managed with diet/exercise or metformin.. 36 weeks. Retatrutide lowered HbA1c and weight in a trial that included placebo and dulaglutide 1.5 mg. Higher doses produced larger effects. The primary glucose endpoint was measured at 24 weeks, while the weight endpoint shown here was measured at 36 weeks. No severe hypoglycemia or deaths were reported. - **HbA1c change**: −2.02 pp. HbA1c change Placebo: -0.01 pp. 12 mg; 24 weeks. Model-estimated change from baseline; efficacy analysis.. SE 0.11 percentage points; p<0.0001 versus placebo. - **HbA1c change versus dulaglutide**: −2.02 pp. HbA1c change versus dulaglutide Dulaglutide 1.5 mg: -1.41 pp. 12 mg; 24 weeks. Model-estimated change from baseline; efficacy analysis.. p=0.0002 for the comparison. - **Body weight change**: −16.94% . Body weight change Placebo: -3 %. 12 mg; 36 weeks. Model-estimated change from baseline; efficacy analysis.. SE 1.30%; p<0.0001 versus placebo. - **Body weight change versus dulaglutide**: −16.94% . Body weight change versus dulaglutide Dulaglutide 1.5 mg: -2.02 %. 12 mg; 36 weeks. Model-estimated change from baseline; efficacy analysis.. Confidence interval not given in the cited summary. What this cannot tell us: Efficacy analyses included 275 participants after six inadvertent enrollments were excluded. Dulaglutide 1.5 mg does not represent every dose or every incretin drug. Gastrointestinal events occurred in 35% across retatrutide groups versus 13% with placebo. Funding: Funded by Eli Lilly and Company; sponsor employees contributed to the research. [Primary source](https://pubmed.ncbi.nlm.nih.gov/37385280/) ### Liver fat: the MRI substudy (2024) Peer-reviewed. Participants from the phase 2 obesity trial with at least 10% liver fat on MRI.. 48 weeks. Liver fat fell rapidly, with most of the reduction achieved by week 24. Visceral and abdominal subcutaneous fat also decreased. Insulin-resistance markers improved. This is a secondary report from the obesity trial, not another independent trial. Liver-fat normalization describes an imaging threshold and does not establish reversal of fibrosis or cure of MASH. - **Relative liver-fat change**: −82.4% . Relative liver-fat change Placebo: 0.3 %. 12 mg; 24 weeks. Model-estimated change from baseline; efficacy analysis.. Placebo-adjusted difference −82.7 percentage points, 95% CI −95.2 to −70.2. - **Relative liver-fat change**: −86% . Relative liver-fat change Placebo: -4.6 %. 12 mg; 48 weeks. Model-estimated change from baseline; efficacy analysis.. Confidence interval not given in the cited summary. - **Participants with liver fat below 5%**: 86% . Participants with liver fat below 5% Placebo: 0 %. 12 mg; 24 weeks. Exploratory binary MRI endpoint; missing values multiply imputed.. Confidence interval not given in the cited summary. - **Participants with liver fat below 5%**: 93% . Participants with liver fat below 5% 12 mg; 48 weeks. Exploratory model estimate with multiple imputation, not a raw proportion of completed scans.. Only 9 of 18 participants in this dose group had week-48 MRI data; no multiplicity adjustment. - **Visceral abdominal fat volume change**: −48.3% . Visceral abdominal fat volume change Placebo: 2.5 %. Highest reported effect; 48 weeks. Model-estimated change from baseline; efficacy analysis.. SE 4.5%; exploratory MRI endpoint. - **Subcutaneous abdominal fat volume change**: −43.5% . Subcutaneous abdominal fat volume change Placebo: -0.1 %. Highest reported effect; 48 weeks. Model-estimated change from baseline; efficacy analysis.. SE 5.0%; exploratory MRI endpoint. - **Insulin-based HOMA2-IR change**: −69.3% . An indirect measure of insulin resistance; improvement does not establish prevention of diabetes. Highest reported effect; 48 weeks. Model-estimated change from baseline; efficacy analysis.. Confidence interval not given in the cited summary. - **Liver volume decreased**: . Dose-responsive reduction versus placebo; imaging finding. 1–12 mg; 24 and 48 weeks. Model-estimated change from baseline; efficacy analysis.. Confidence interval not given in the cited summary. - **Cytokeratin-18 decreased at selected doses**: . K-18 decreased versus placebo at 8 mg at week 24 and at 8 and 12 mg at week 48. This biomarker finding does not establish histological improvement. 8 mg at week 24; 8 and 12 mg at week 48; 24 and 48 weeks. Exploratory biomarker analysis, Table 2 and Figure 4a.. p<0.05 versus placebo; no multiplicity adjustment and no biopsy endpoint. - **Pro-C3 decreased at selected doses**: . Pro-C3 decreased versus placebo at 4, 8 and 12 mg at week 24 and at 1, 4 and 8 mg at week 48. This is not proof that liver fibrosis regressed. 4, 8, 12 mg at week 24; 1, 4, 8 mg at week 48; 24 and 48 weeks. Exploratory biomarker analysis, Table 2 and Figure 4b.. p≤0.001 at week 24 and p≤0.01 at week 48 versus placebo; no multiplicity adjustment. What this cannot tell us: Only 43.9% had week-48 MRI data. Binary outcomes used multiple imputation assuming missingness at random. No biopsy endpoint; ALT, AST, FIB-4 and ELF did not consistently improve versus placebo. Funding: Funded by Eli Lilly and Company; sponsor employees contributed to the research. [Primary source](https://www.nature.com/articles/s41591-024-03018-2) ### TRIUMPH-4: weight and knee pain (2025) Sponsor announcement. Adults with overweight or obesity and knee osteoarthritis, without diabetes; 84% had BMI ≥35.. 68 weeks. The first phase 3 announcement paired substantial weight loss with less knee pain and better physical function. Both doses met the co-primary endpoints. The greatest pain reduction occurred at 9 mg, while the greatest weight loss occurred at 12 mg. The results describe symptoms and function; they do not demonstrate cartilage regrowth. - **Body weight change**: −26.4% . Body weight change Placebo: -2.1 %. 9 mg; 68 weeks. Efficacy estimand: estimates continued assigned treatment without prohibited rescue interventions.. Confidence interval not given in the cited summary. - **Body weight change**: −28.7% . Body weight change Placebo: -2.1 %. 12 mg; 68 weeks. Efficacy estimand: estimates continued assigned treatment without prohibited rescue interventions.. Confidence interval not given in the cited summary. - **Body weight change including discontinuation**: −23.7% . Body weight change including discontinuation Placebo: -4.6 %. 12 mg; 68 weeks. Treatment-regimen estimand: includes the effect of treatment discontinuation and relevant rescue treatment.. Confidence interval not given in the cited summary. - **WOMAC pain score change**: −4.5 points / 10. The post-hoc relative change was −75.8%; placebo −40.3%. Placebo: -2.4 points / 10. 9 mg; 68 weeks. Efficacy estimand: estimates continued assigned treatment without prohibited rescue interventions.. Confidence interval not given in the cited summary. - **WOMAC pain score change**: −4.4 points / 10. The post-hoc relative change was −74.3%; this is the dose associated with 28.7% weight loss. Placebo: -2.4 points / 10. 12 mg; 68 weeks. Efficacy estimand: estimates continued assigned treatment without prohibited rescue interventions.. Confidence interval not given in the cited summary. - **WOMAC physical function score change**: −4.2 points / 10. WOMAC physical function score change Placebo: -2.1 points / 10. 12 mg; 68 weeks. Efficacy estimand: estimates continued assigned treatment without prohibited rescue interventions.. Confidence interval not given in the cited summary. - **Participants reporting no knee pain**: 14.1% . Participants reporting no knee pain Placebo: 4.2 %. 9 mg; 68 weeks. Post-hoc observed efficacy data.. Confidence interval not given in the cited summary. - **Systolic blood pressure change**: −14 mmHg. Systolic blood pressure change 12 mg; 68 weeks. Efficacy estimand: estimates continued assigned treatment without prohibited rescue interventions.. Additional endpoint not controlled for multiplicity; placebo value not given in announcement. What this cannot tell us: Sponsor topline report. Relative WOMAC changes and pain-free proportions were post-hoc. At 12 mg, adverse-event discontinuation was 18.2% versus 4.0% with placebo; dysesthesia occurred in 20.9% versus 0.7%. Funding: Funded by Eli Lilly and Company; sponsor employees contributed to the research. [Primary source](https://investor.lilly.com/news-releases/news-release-details/lillys-triple-agonist-retatrutide-delivered-weight-loss-average) ### TRIUMPH-1: obesity and its complications (2026) Conference results. Adults with obesity or overweight and a related condition, without diabetes; knee basket n=574, sleep-apnea basket n=243.. 80 weeks. The main trial found substantial weight and waist reductions. Nested trials also found less knee pain and fewer breathing interruptions during sleep. Patient-reported physical and psychosocial quality of life improved. A selected extension followed 532 participants through 104 weeks. These conference findings expand the evidence beyond weight alone. - **Body weight change**: −19% . Body weight change Placebo: -2.2 %. 4 mg; 80 weeks. Efficacy estimand: estimates continued assigned treatment without prohibited rescue interventions.. Confidence interval not given in the cited summary. - **Body weight change**: −25.9% . Body weight change Placebo: -2.2 %. 9 mg; 80 weeks. Efficacy estimand: estimates continued assigned treatment without prohibited rescue interventions.. Confidence interval not given in the cited summary. - **Body weight change**: −28.3% . Body weight change Placebo: -2.2 %. 12 mg; 80 weeks. Efficacy estimand: estimates continued assigned treatment without prohibited rescue interventions.. Confidence interval not given in the cited summary. - **Body weight change including discontinuation**: −25% . Body weight change including discontinuation Placebo: -3.9 %. 12 mg; 80 weeks. Treatment-regimen estimand: includes the effect of treatment discontinuation and relevant rescue treatment.. Confidence interval not given in the cited summary. - **Waist circumference change**: −24.1 cm. Waist circumference change Placebo: -3.6 cm. 12 mg; 80 weeks. Efficacy estimand: estimates continued assigned treatment without prohibited rescue interventions.. Confidence interval not given in the cited summary. - **Participants losing at least 30% of body weight**: 45.3% . Participants losing at least 30% of body weight Placebo: 0.5 %. 12 mg; 80 weeks. Efficacy estimand: estimates continued assigned treatment without prohibited rescue interventions.. Confidence interval not given in the cited summary. - **Weight change in selected extension participants**: −30.3% . Weight change in selected extension participants Original 12 mg, then maximum tolerated 9/12 mg; 104 weeks. Efficacy estimand: estimates continued assigned treatment without prohibited rescue interventions.. 532 selected completers across all arms; original placebo participants switched to retatrutide. - **WOMAC pain score change**: −4.3 points / 10. WOMAC pain score change Placebo: -2.24 points / 10. 12 mg; 80 weeks. Efficacy estimand: estimates continued assigned treatment without prohibited rescue interventions.. Knee basket, n=574; p<0.001 versus placebo. - **Apnea–hypopnea index change**: −36.1 events/h. Apnea–hypopnea index change Placebo: -11.1 events/h. 9 mg; 80 weeks. Efficacy estimand: estimates continued assigned treatment without prohibited rescue interventions.. Sleep-apnea basket, n=243; reported relative reduction 60.6%. - **Apnea–hypopnea index change**: −33.8 events/h. Apnea–hypopnea index change Placebo: -11.1 events/h. 12 mg; 80 weeks. Efficacy estimand: estimates continued assigned treatment without prohibited rescue interventions.. Confidence interval not given in the cited summary. - **Weight-related quality of life improved**: . Physical function and psychosocial domains improved; not a claim of treating depression. 4, 9, 12 mg; 80 weeks. Efficacy estimand: estimates continued assigned treatment without prohibited rescue interventions.. IWQOL-Lite-CT; conference report, p<0.001 versus placebo. - **Serious adverse events**: . Serious adverse events occurred in 7.7–10.5% across the retatrutide groups versus 5.5% with placebo. Serious events are not necessarily caused by treatment. Placebo: 5.5 . 4, 9, 12 mg; 80 weeks. Conference safety report, Overview of Adverse Events slide.. Range across separate dose groups, not a pooled rate or confidence interval; no causal conclusion. What this cannot tell us: Conference and sponsor reports, not a full peer-reviewed efficacy paper. Extension participants had completed and tolerated treatment. At 12 mg, adverse-event discontinuation was 11.3% versus 4.9% with placebo. Funding: Funded by Eli Lilly and Company; sponsor employees contributed to the research. [Primary source](https://www.lilly.com/hcp/congresses/ada-2026/dv039292-triumph-1) ### TRIUMPH-2: obesity with diabetes (2026) Sponsor announcement. Adults with overweight or obesity and type 2 diabetes; baseline HbA1c 7.7%.. 80 weeks. The trial met its weight endpoint in adults with diabetes, a population distinct from the obesity trials without diabetes. HbA1c also fell. These results were announced in July 2026. The largest glucose effect occurred at 9 mg and the largest weight effect at 12 mg. - **Body weight change**: −12.7% . Body weight change Placebo: -4 %. 4 mg; 80 weeks. Efficacy estimand: estimates continued assigned treatment without prohibited rescue interventions.. Confidence interval not given in the cited summary. - **Body weight change**: −19.1% . Body weight change Placebo: -4 %. 9 mg; 80 weeks. Efficacy estimand: estimates continued assigned treatment without prohibited rescue interventions.. Confidence interval not given in the cited summary. - **Body weight change**: −20.8% . Body weight change Placebo: -4 %. 12 mg; 80 weeks. Efficacy estimand: estimates continued assigned treatment without prohibited rescue interventions.. Confidence interval not given in the cited summary. - **HbA1c change**: −1.4 pp. HbA1c change Placebo: -0.2 pp. 4 mg; 80 weeks. Efficacy estimand: estimates continued assigned treatment without prohibited rescue interventions.. Confidence interval not given in the cited summary. - **HbA1c change**: −1.6 pp. HbA1c change Placebo: -0.2 pp. 9 mg; 80 weeks. Efficacy estimand: estimates continued assigned treatment without prohibited rescue interventions.. Confidence interval not given in the cited summary. - **HbA1c change**: −1.5 pp. HbA1c change Placebo: -0.2 pp. 12 mg; 80 weeks. Efficacy estimand: estimates continued assigned treatment without prohibited rescue interventions.. Confidence interval not given in the cited summary. What this cannot tell us: Only topline results were available in the verified source. Adverse-event discontinuation was 7.7% at 12 mg versus 4.9% with placebo. The sleep-apnea subset should not be assigned results from TRIUMPH-1. Funding: Funded by Eli Lilly and Company; sponsor employees contributed to the research. [Primary source](https://investor.lilly.com/news-releases/news-release-details/lillys-triple-agonist-retatrutide-successful-two-additional) ### TRIUMPH-3: established cardiovascular disease (2026) Sponsor announcement. Adults with BMI ≥35 and established cardiovascular disease, with or without type 2 diabetes.. 80 weeks. Weight and cardiovascular risk factors improved. Cardiovascular events were less frequent than expected, leaving wide confidence intervals. The prespecified MACE-5 estimate favored retatrutide numerically, while MACE-3 did not. Neither established cardiovascular benefit. These outcomes should remain visible beside the favorable changes in weight, lipids and blood pressure. - **Body weight change**: −21.6% . Body weight change Placebo: -3.2 %. 9 mg; 80 weeks. Efficacy estimand: estimates continued assigned treatment without prohibited rescue interventions.. Confidence interval not given in the cited summary. - **Body weight change**: −22.6% . Body weight change Placebo: -3.2 %. 12 mg; 80 weeks. Efficacy estimand: estimates continued assigned treatment without prohibited rescue interventions.. Confidence interval not given in the cited summary. - **MACE-5 hazard ratio**: 0.82 HR. 44 events with retatrutide, 52 with placebo. All-cause death, myocardial infarction, stroke, heart-failure event or coronary revascularization. Placebo: 1 HR. Pooled 9 and 12 mg; 80 weeks. Prespecified in-study time-to-first-event analysis, regardless of adherence.. 95% CI 0.55–1.22; includes 1. - **MACE-3 hazard ratio**: 1.12 HR. 27 events with retatrutide, 23 with placebo. Cardiovascular death, myocardial infarction or stroke. Placebo: 1 HR. Pooled 9 and 12 mg; 80 weeks. Prespecified in-study time-to-first-event analysis, regardless of adherence.. 95% CI 0.64–1.96; includes 1. - **Triglyceride change**: −37% . Triglyceride change 12 mg; 80 weeks. Efficacy estimand: estimates continued assigned treatment without prohibited rescue interventions.. Confidence interval not given in the cited summary. - **Non-HDL cholesterol change**: −16.5% . Non-HDL cholesterol change 12 mg; 80 weeks. Efficacy estimand: estimates continued assigned treatment without prohibited rescue interventions.. Confidence interval not given in the cited summary. - **Systolic blood pressure change**: −9.3 mmHg. Systolic blood pressure change 12 mg; 80 weeks. Efficacy estimand: estimates continued assigned treatment without prohibited rescue interventions.. Confidence interval not given in the cited summary. - **Waist circumference change**: −19 cm. Waist circumference change 12 mg; 80 weeks. Efficacy estimand: estimates continued assigned treatment without prohibited rescue interventions.. Confidence interval not given in the cited summary. - **High-sensitivity C-reactive protein change**: −51.2% . High-sensitivity C-reactive protein change 12 mg; 80 weeks. Efficacy estimand: estimates continued assigned treatment without prohibited rescue interventions.. Confidence interval not given in the cited summary. What this cannot tell us: Sponsor topline report with limited events, not proof of prevention or harm. At 12 mg, adverse-event discontinuation was 13.5% versus 4.8% with placebo. Post-hoc on-treatment analyses answer a different question. Funding: Funded by Eli Lilly and Company; sponsor employees contributed to the research. [Primary source](https://investor.lilly.com/news-releases/news-release-details/lillys-triple-agonist-retatrutide-successful-two-additional) ### TRANSCEND-T2D-1: published phase 3 diabetes results (2026) Peer-reviewed. Adults with type 2 diabetes inadequately controlled by diet and exercise alone; mean disease duration 2.5 years.. 40 weeks. Retatrutide monotherapy reduced HbA1c and body weight. The peer-reviewed report provides the treatment-regimen estimates shown here. Larger figures in sponsor headlines use an efficacy estimand instead. Most participants completed treatment. Normal-range HbA1c during medication use should not be described as drug-free diabetes remission. - **HbA1c change**: −1.69 pp. HbA1c change Placebo: -0.81 pp. 4 mg; 40 weeks. Treatment-regimen estimand: includes the effect of treatment discontinuation and relevant rescue treatment.. 95% CI for placebo-adjusted difference: −1.18 to −0.59 percentage points; p<0.0001. - **HbA1c change**: −1.86 pp. HbA1c change Placebo: -0.81 pp. 9 mg; 40 weeks. Treatment-regimen estimand: includes the effect of treatment discontinuation and relevant rescue treatment.. 95% CI for placebo-adjusted difference: −1.32 to −0.76 percentage points; p<0.0001. - **HbA1c change**: −1.94 pp. HbA1c change Placebo: -0.81 pp. 12 mg; 40 weeks. Treatment-regimen estimand: includes the effect of treatment discontinuation and relevant rescue treatment.. 95% CI for placebo-adjusted difference: −1.39 to −0.85 percentage points; p<0.0001. - **Body weight change**: −11.5% . Body weight change Placebo: -2.6 %. 4 mg; 40 weeks. Treatment-regimen estimand: includes the effect of treatment discontinuation and relevant rescue treatment.. Confidence interval not given in the cited summary. - **Body weight change**: −13.9% . Body weight change Placebo: -2.6 %. 9 mg; 40 weeks. Treatment-regimen estimand: includes the effect of treatment discontinuation and relevant rescue treatment.. Confidence interval not given in the cited summary. - **Body weight change**: −15.3% . Body weight change Placebo: -2.6 %. 12 mg; 40 weeks. Treatment-regimen estimand: includes the effect of treatment discontinuation and relevant rescue treatment.. Confidence interval not given in the cited summary. - **Body weight change assuming continued treatment**: −16.8% . Sponsor efficacy estimate; keep separate from the −15.3% treatment-regimen result. 12 mg; 40 weeks. Efficacy estimand: estimates continued assigned treatment without prohibited rescue interventions.. Confidence interval not given in the cited summary. What this cannot tell us: Early diabetes treated without other glucose-lowering drugs limits generalization to advanced disease or insulin combinations. Forty weeks cannot establish long-term complication prevention. Gastrointestinal adverse events were the main tolerability problem. Funding: Funded by Eli Lilly and Company; sponsor employees contributed to the research. [Primary source](https://pubmed.ncbi.nlm.nih.gov/42250575/) ### Body composition: fat and lean tissue (2025) Peer-reviewed. DXA substudy of the phase 2 diabetes trial; 189 enrolled.. 36 weeks. DXA measurements showed preferential reduction in fat mass with higher retatrutide doses. Lean mass also decreased. The largest mean fat-mass reduction occurred in the pooled 8 mg groups. These findings describe the composition of weight loss; they do not support claims that retatrutide builds muscle or prevents all lean-tissue loss. - **Total fat mass change**: −26.1% . Total fat mass change Placebo: -4.5 %. 8 mg pooled; 36 weeks. Model-estimated change from baseline; efficacy analysis.. SE 2.5%; placebo-adjusted difference −21.6 percentage points, 95% CI −27.1 to −16.1. - **Total fat mass change**: −23.2% . Total fat mass change Placebo: -4.5 %. 12 mg; 36 weeks. Model-estimated change from baseline; efficacy analysis.. SE 3.0%; placebo-adjusted difference −18.7 percentage points, 95% CI −25.1 to −12.3. - **Lean mass also decreased**: . Fat loss exceeded lean-mass loss. Preservation of every kilogram of muscle was not demonstrated. Study doses; 36 weeks. Model-estimated change from baseline; efficacy analysis.. Confidence interval not given in the cited summary. What this cannot tell us: Only 155 had baseline DXA and 103 completed treatment with paired scans. DXA lean mass includes water and tissues beyond muscle. This substudy cannot establish effects on strength, frailty or fractures. Funding: Funded by Eli Lilly and Company; sponsor employees contributed to the research. [Primary source](https://pubmed.ncbi.nlm.nih.gov/40609566/) ### Appetite and eating behavior in diabetes (2025) Peer-reviewed. Exploratory questionnaire analyses in 275 participants from the phase 2 diabetes trial.. 36 weeks. Higher doses reduced perceived hunger and the tendency to overeat. Changes in these eating-behavior scores tracked weight change. Results depended on the questionnaire and time point. Satiety and fullness ratings did not consistently separate retatrutide from placebo, a useful distinction from broad claims that it eliminates hunger or food-related thoughts. - **Placebo-adjusted hunger VAS change**: −15.7 VAS points. Placebo-adjusted hunger VAS change Placebo: 0 VAS points. 12 mg; 24 weeks. Least-squares mean difference versus placebo, exploratory.. p<0.05; no multiplicity adjustment. - **Perceived hunger and overeating tendency decreased**: . Eating Inventory scores improved versus placebo. 8 and 12 mg; 24 and 36 weeks. Model-estimated change from baseline; efficacy analysis.. Confidence interval not given in the cited summary. - **Placebo-adjusted dietary-restraint score change**: 3.7 EI points. Placebo-adjusted dietary-restraint score change Placebo: 0 EI points. 12 mg; 36 weeks. Exploratory least-squares mean difference.. p<0.05; other doses did not show this difference. What this cannot tell us: Exploratory analyses without multiplicity correction; self-report, no dietary records of calorie intake. Associations with weight do not isolate a causal receptor mechanism or establish treatment of binge-eating disorder. Funding: Funded by Eli Lilly and Company; sponsor employees contributed to the research. [Primary source](https://pmc.ncbi.nlm.nih.gov/articles/PMC12587234/) ### Eating behavior in obesity (2025) Peer-reviewed. Secondary Eating Inventory analyses from the phase 2 obesity trial.. 48 weeks. Participants reported lower hunger and less cue-driven overeating, with larger changes at doses of at least 4 mg. This publication adds a patient-reported perspective to the weight results from the same trial. It should be read as a secondary analysis, not counted as another independent demonstration of weight loss. - **Hunger and disinhibition scores decreased**: . The largest changes occurred at doses of at least 4 mg. ≥4 mg; 24 and 48 weeks. Secondary patient-reported analysis.. Confidence interval not given in the cited summary. What this cannot tell us: Questionnaire-based secondary analysis, without adjustment for multiple comparisons. Scores do not measure actual calorie intake and cannot establish that psychological or eating disorders have been treated. Funding: Funded by Eli Lilly and Company; sponsor employees contributed to the research. [Primary source](https://dom-pubs.onlinelibrary.wiley.com/doi/10.1111/dom.16662) ### What trial participants noticed (2025) Peer-reviewed. 40 blinded exit interviews after the phase 2 obesity trial; 36 retatrutide, 4 placebo.. 48 weeks. Interviewees described smaller meals, less frequent hunger, more control over eating and easier movement. Some reported more energy, confidence and participation in leisure or exercise. These accounts show which changes mattered in daily life. They are qualitative experiences from a selected sample, not estimates of how often future patients will benefit. - **Interviewees reporting easier movement**: 27 people / 36. 27 of 36 retatrutide interviewees. 1, 4, 8, 12 mg; Exit interview. Qualitative interview counts; no statistical placebo comparison.. Confidence interval not given in the cited summary. - **Interviewees reporting more energy**: 24 people / 36. 24 of 36; qualitative report, not an objective energy-expenditure measurement. 1, 4, 8, 12 mg; Exit interview. Qualitative interview counts; no statistical placebo comparison.. Confidence interval not given in the cited summary. - **Interviewees reporting early eating changes**: 31 people / 36. 31 of 36 described changes within the first eight weeks. 1, 4, 8, 12 mg; Exit interview. Qualitative interview counts; no statistical placebo comparison.. Confidence interval not given in the cited summary. What this cannot tell us: Only four placebo interviewees; no reliable treatment-effect comparison. Recall and selection bias are possible. Some reported weakness, social limitations, nausea or disappointment. Feeling happier does not establish an antidepressant effect. Funding: Funded by Eli Lilly and Company; sponsor employees contributed to the research. [Primary source](https://pmc.ncbi.nlm.nih.gov/articles/PMC12596213/) ### Atherogenic particles and inflammation (2026) Peer-reviewed. Two phase 2 cohorts: obesity with and without type 2 diabetes. 36 weeks with diabetes; 48 without. A later analysis of the original phase 2 trials examined blood particles and inflammation. Several markers associated with cardiovascular risk moved in a favourable direction, including ApoB and small LDL particles. The clearest inflammatory changes occurred in participants without diabetes. This publication extends the laboratory description of the earlier trials. It does not add a new independent clinical population or establish prevention of cardiovascular events. - **ApoB**: −24.2% . ApoB up to 12 mg; 48 weeks. Maximum placebo-adjusted change, obesity cohort. No clinical outcome demonstrated. - **Non-HDL cholesterol**: −26.9% . Non-HDL cholesterol ; 48 weeks. Maximum placebo-adjusted change. No clinical outcome demonstrated. - **Small LDL particles**: −32.3% . Small LDL particles ; 48 weeks. Maximum placebo-adjusted change. No clinical outcome demonstrated. - **High-sensitivity C-reactive protein**: −54.8% . High-sensitivity C-reactive protein ; 48 weeks. Maximum placebo-adjusted change, without T2D. Significant in the obesity cohort; not significant in the diabetes cohort. - **Interleukin-6**: −29.6% . Interleukin-6 ; 48 weeks. Maximum placebo-adjusted change, without T2D. No clinical outcome demonstrated. - **Triglycerides, ESC 2024 analysis**: −40.6% . Triglycerides, ESC 2024 analysis ; 48 weeks. Up to; earlier conference abstract of same obesity cohort. ESC abstract https://doi.org/10.1093/eurheartj/ehae666.1501; do not count as a new trial. - **Lipoprotein insulin-resistance score**: −32.5% . Lipoprotein insulin-resistance score 12 mg; 48 weeks. NMR-derived score, ESC 2024 abstract; same cohort. A biomarker score, not a 32.5% reduction in diabetes incidence. - **Fewer triglyceride-rich particles, including large particles**: . Fewer triglyceride-rich particles, including large particles. ; 36/48 weeks. Exploratory laboratory analysis; 2026 publication, abstract Results; total and large TRLP. No demonstrated reduction in clinical cardiovascular events. - **Lower triglyceride-rich lipoprotein cholesterol**: . Lower triglyceride-rich lipoprotein cholesterol. ; 36/48 weeks. Exploratory laboratory analysis; 2026 publication, abstract Results; TRL cholesterol. No demonstrated reduction in clinical cardiovascular events. - **Fewer total LDL particles**: . Fewer total LDL particles. ; 36/48 weeks. Exploratory laboratory analysis; 2026 publication, abstract Results; total LDL particles. No demonstrated reduction in clinical cardiovascular events. - **Lower ApoC-III**: . Lower ApoC-III. ; 48 weeks. Exploratory laboratory analysis; ESC 2024 abstract, Results; DOI 10.1093/eurheartj/ehae666.1501. No demonstrated reduction in clinical cardiovascular events. What this cannot tell us: Post hoc analyses of existing cohorts. Laboratory changes cannot be read as equivalent percentage reductions in heart attacks. The strongest reported effects can come from different doses. The two cohorts overlap with the weight, liver and kidney reports. Funding: Eli Lilly and Company [Primary source](https://pubmed.ncbi.nlm.nih.gov/42608321/) ### How fuel handling changes (2026) Peer-reviewed. Samples from the original phase 2 cohorts; 282 obesity participants in the omics analysis. 24 and 48 weeks. Detailed blood profiling found changes consistent with greater use of fatty acids and improved insulin-resistance markers. Ketone and carnitine measurements offer clues about glucagon signalling in people taking retatrutide. These are biochemical observations from existing phase 2 participants, not another efficacy trial. The authors describe plausible mechanisms, while recognising that weight loss and the shared incretin actions may also account for parts of the pattern. - **3-hydroxybutyrate**: 198% . 3-hydroxybutyrate Placebo: 19.1 %. 12 mg; 24 weeks. Change from baseline; Figure 2A. FDR-adjusted P < 0.001 versus placebo; mechanism marker. - **3-hydroxybutyrate**: 148.7% . 3-hydroxybutyrate Placebo: 7.8 %. 12 mg; 48 weeks. Change from baseline; Figure 2A. No clinical outcome demonstrated. - **Acetylcarnitine/free carnitine ratio**: 95.3% . Acetylcarnitine/free carnitine ratio Placebo: 6.4 %. 12 mg; 24 weeks. Change from baseline; Figure 2B. No clinical outcome demonstrated. What this cannot tell us: Exploratory molecular measurements. Higher ketones do not prove better cognition or heart function. No receptor-isolating comparison establishes how much of each change comes from glucagon rather than weight loss or the other receptor actions. Funding: Eli Lilly and Company [Primary source](https://academic.oup.com/jcem/advance-article/doi/10.1210/clinem/dgag201/8678492) ### Albuminuria and estimated filtration (2025) Peer-reviewed. 281 participants with T2D and 338 with obesity without T2D; mostly normal albuminuria. 36 or 48 weeks. Researchers re-examined the two phase 2 trials for kidney signals. Albumin leakage into urine declined at higher doses. Estimated filtration increased in the obesity cohort, with similar directions across creatinine and cystatin C calculations. The diabetes cohort did not show the same filtration change. These findings justify dedicated kidney studies, especially because few participants had established kidney disease and their starting albuminuria was generally low. - **UACR in type 2 diabetes**: −37% . UACR in type 2 diabetes 12 mg; 36 weeks. Placebo-adjusted change. 95% CI −57.3 to −7.0 - **UACR without diabetes**: −28% . UACR without diabetes 8 mg; 48 weeks. Placebo-adjusted change. 95% CI −46.0 to −4.1 - **UACR without diabetes**: −31.5% . UACR without diabetes 12 mg; 48 weeks. Placebo-adjusted change. 95% CI −49.3 to −7.4 - **Creatinine-based eGFR without diabetes**: +8.5 mL/min/1.73 m². Creatinine-based eGFR without diabetes 12 mg; 48 weeks. Difference versus placebo. 95% CI 4.9 to 12.1; eGFR unchanged in T2D What this cannot tell us: Post hoc, short duration, few participants with CKD and no hard kidney outcomes. UACR began at low absolute levels. Much of the eGFR increase reversed during the four-week washout; cystatin C and combined estimates no longer differed from placebo. Creatinine estimates remained higher at 8 and 12 mg. These changes do not establish kidney regeneration or durable protection. Funding: Eli Lilly and Company [Primary source](https://pubmed.ncbi.nlm.nih.gov/40630318/) ### Measuring kidney function directly (2026) Ongoing trial. Adults with overweight/obesity and CKD, eGFR 25–75; with or without T2D. 24-week primary measurement. TRANSCEND-CKD addresses an uncertainty left by the exploratory kidney analysis. It measures filtration with iohexol clearance and uses MRI to examine kidney structure and blood flow. Participants have chronic kidney disease, making this population more relevant to renal questions than the original obesity cohort. The retrieved paper reports study design and baseline characteristics. Its participant counts and starting kidney values are not treatment results. - **Measured GFR by iohexol clearance**: . Measured GFR by iohexol clearance maximum tolerated, up to 12 mg; 24 weeks. Planned primary endpoint. No treatment result in the design publication. What this cannot tell us: The cited source is a design and baseline report. No efficacy estimate is extracted from it. Measured filtration and imaging may clarify mechanism, but a short mechanistic trial cannot independently settle long-term kidney failure prevention. Funding: Eli Lilly and Company [Primary source](https://pubmed.ncbi.nlm.nih.gov/41160422/) ### Memory findings in diabetic rats (2026) Animal / laboratory. Male rats with streptozotocin-induced insulin-deficient diabetes and control groups. Experimental animal treatment. A study first circulated as a preprint has now appeared in a peer-reviewed journal. Treated diabetic rats retained some learning and memory performance, with lower hippocampal TNF-alpha and partial preservation of tissue structure. Improvement depended on the task and did not normalise every memory measure. Healthy treated animals did not outperform healthy controls. This is a reason to investigate brain effects, not evidence of cognitive enhancement in people. - **Partial preservation of learning and memory**: . Partial preservation of learning and memory ; . Morris Water Maze and Passive Avoidance. Task-dependent; no complete normalisation. - **Lower hippocampal TNF-alpha**: . Lower hippocampal TNF-alpha ; . Animal tissue assay. IL-1β trend was not significant. What this cannot tell us: Male animals with an insulin-deficient experimental model. Direct central exposure was not established, and relevance to insulin-resistant human T2D remains uncertain. The study demonstrates neither Alzheimer prevention nor improved cognition in healthy humans. Funding: Funding not verified in abstract; authors declare no competing interests [Primary source](https://pubmed.ncbi.nlm.nih.gov/42385950/) ### Preventing major liver complications (2026) Ongoing trial. Adults with high-risk MASLD identified with non-invasive tests. Approximately 224 weeks; optional 2-year extension. SYNERGY-Outcomes asks whether treatment prevents serious liver complications in people selected for high-risk metabolic liver disease. The master protocol assigns participants to retatrutide, tirzepatide or placebo. It does not test taking both drugs together. Non-invasive tests identify eligible participants, so this is broader than a biopsy-confirmed MASH trial. The registry estimates primary completion in August 2030, with an optional extension extending the overall study to 2032. - **Major adverse liver outcomes**: . Major adverse liver outcomes ; approximately 224 weeks. Planned clinical outcomes. No results posted in the verified record. What this cannot tell us: An ongoing outcomes trial supplies a research question, not a treatment effect. Planned enrollment and completion dates can change. Do not confuse this protocol with the smaller tirzepatide SYNERGY-NASH biopsy trial or describe the active drugs as a combination. Funding: Eli Lilly and Company [Primary source](https://clinicaltrials.gov/study/NCT07165028?aggFilters=status%3Arec) ### The cardiovascular and kidney outcomes test (2026) Ongoing trial. Age ≥45, BMI ≥27, established ASCVD and/or chronic kidney disease. About 5 years; estimated completion February 2029. TRIUMPH-Outcomes is designed to determine whether retatrutide reduces serious cardiovascular complications or worsening kidney function. It enrols a higher-risk population with established cardiovascular disease or chronic kidney disease. This is the study that can connect favourable laboratory findings with clinical events over years. Its planned size is much larger than the exploratory phase 2 analyses. At the verification date, the cited source contains trial information rather than results. - **Serious cardiovascular complications**: . Serious cardiovascular complications ; . Planned clinical outcomes. Outcome benefit remains to be established. - **Worsening kidney function**: . Worsening kidney function ; . Planned clinical outcomes. Outcome benefit remains to be established. What this cannot tell us: No outcome effect is available from this source. The sponsor page and regional pages differ on recruitment wording, so enrollment status is omitted. The date shown here is the verification date, not a new result announcement. Funding: Eli Lilly and Company [Primary source](https://trials.lilly.com/en-US/trial/479798) ### The direct comparison with tirzepatide (2026) Ongoing trial. Adults with obesity. About 89 weeks of participation. TRIUMPH-5 compares retatrutide directly with tirzepatide in adults with obesity. All participants receive an active drug, with no placebo arm. This design can answer comparative efficacy and tolerability questions that separate obesity trials cannot resolve. The sponsor reports completed enrollment and a target of 800 participants. The public trial page describes the study but does not provide a comparative treatment result at this verification date. - **Comparative efficacy and safety**: . Comparative efficacy and safety ; . Randomised active-drug comparison. No result available in the verified sponsor source. What this cannot tell us: No head-to-head result is extracted. The sponsor page gives December 2026 as the end date, while its ADA development presentation places the comparison in 2027. Completion, analysis and publication are different milestones. Funding: Eli Lilly and Company [Primary source](https://trials.lilly.com/en-US/trial/549215) ### Maintenance, pain and dosing questions (2026) Ongoing trial. Several distinct development-program populations. Program milestones extending to 2030. The broader program investigates practical questions after initial weight loss, including maintenance, dose selection and escalation schedules. It also examines chronic low back pain and diabetes treatment with different background therapies. These research areas matter because the largest weight reduction does not answer every question about long-term use. The sponsor presentation is a map of planned work. Its dates describe expectations and must not be read as established benefits. - **Weight-loss maintenance**: . Weight-loss maintenance ; program expectation: 2028. Research direction. Planned work; not a demonstrated maintenance result. - **Chronic low back pain**: . Chronic low back pain ; program expectation: 2027. Research direction. Separate from knee osteoarthritis evidence. - **Dose-escalation options**: . Dose-escalation options ; program expectation: 2028. Research direction. No personal dosing recommendation follows. What this cannot tell us: Program overview rather than a single trial or a result. No participant total is assigned, to avoid adding overlapping populations. The source was presented at ADA 2026; the date shown is this atlas verification date. Funding: Eli Lilly and Company [Primary source](https://investor.lilly.com/static-files/a2b20b4f-7944-4de2-9d10-00f384bbede2) ### Liver metabolism in mice and hamsters (2026) Animal / laboratory. Diet-induced obese MASH mice and hamsters, 8 per experimental group. 5 weeks. This animal study examined metabolic changes in two diet-induced models. Retatrutide reduced liver lipid measures and insulin-resistance markers, with changes in food intake and body composition. The models also exposed limits to the response. In mice, improved liver fat did not mean improved fibrosis or inflammation, and energy expenditure did not change significantly. The findings help refine experiments while keeping human liver claims tied to human evidence. - **Hepatic triglyceride content in hamsters**: −50% . Hepatic triglyceride content in hamsters ; 5 weeks. Animal comparison. P < 0.01; histopathology score not reduced. - **No improvement in mouse liver fibrosis**: . No improvement in mouse liver fibrosis ; 5 weeks. Figure 2 histology. A negative result alongside reduced steatosis. What this cannot tell us: Short animal experiments cannot establish human efficacy. Weight loss included lean tissue changes. Liver triglyceride reduction and histological improvement were not interchangeable. Group sizes were small, and the experiments do not support treating human MASH outside clinical evidence. Funding: Funding not confirmed in abstract; author affiliations include Physiogenex and Janvier Labs [Primary source](https://pubmed.ncbi.nlm.nih.gov/41741376/) ### Cancer progression in obese mice (2025) Animal / laboratory. Mouse pancreatic and lung tumour models. Model-specific experimental schedules. Researchers tested how retatrutide-associated weight loss affected implanted tumours in obese mice. Pancreatic tumours appeared later and grew less, with related findings in a lung tumour model. The work also examined immune changes. It belongs among early research questions because an implanted mouse tumour differs substantially from cancer developing in a person. The results do not establish an oncology use for retatrutide. - **Reduced tumour growth in mice**: . Reduced tumour growth in mice ; . Animal evidence. Human benefit unproven What this cannot tell us: These are laboratory and animal experiments, not cancer treatment trials in patients. Neither the model response nor the mechanistic findings establish cancer prevention, safe combination therapy or a survival benefit in people. Funding: Funding not extracted in this review [Primary source](https://www.nature.com/articles/s44324-025-00054-5) ### Chemotherapy response in obese breast-cancer models (2025) Animal / laboratory. Cells and female mice with triple-negative breast tumours. Model-specific experimental schedules. This study explored how fat cells influence breast-cancer treatment resistance. Retatrutide reduced tumour size and improved chemotherapy response in experimental models, alongside changes in a pathway regulating the YAP protein. Human tumour datasets helped frame the biological question, but patients were not treated with retatrutide in this work. The distinction matters when reading a paper that discusses possible future therapies for obesity-associated cancer. - **Improved chemotherapy response in mouse models**: . Improved chemotherapy response in mouse models ; . Animal evidence. Human benefit unproven What this cannot tell us: These are laboratory and animal experiments, not cancer treatment trials in patients. Neither the model response nor the mechanistic findings establish cancer prevention, safe combination therapy or a survival benefit in people. Funding: Emory School of Medicine, NIH and Winship support [Primary source](https://advanced.onlinelibrary.wiley.com/doi/10.1002/advs.202407494) ## How to read this atlas We searched primary literature and trial records under retatrutide and LY3437943, then followed conference reports to later publications. The review includes beneficial, neutral and unfavorable findings relevant to interpreting the molecule. It is an editorial evidence review, not a registered systematic review or meta-analysis. Publication status, study population and endpoint type are separate facts. A peer-reviewed animal experiment remains animal evidence. An impressive biomarker result remains a biomarker result. An ongoing trial contributes a research question, not a benefit. The library links to the source for each research record. Trial reports and substudies may share participants; their sample sizes must not be added. Numbers retain their reported analysis and comparator. Missing confidence intervals are identified rather than manufactured. Most clinical studies in this atlas were funded by Eli Lilly, the developer. Funding does not invalidate a result, but it matters when interpreting sponsor announcements and exploratory analyses. Dates refer to available evidence, not guaranteed regulatory or commercial milestones. ## A few useful translations - **HbA1c**: A measure of average blood glucose over recent months. Changes are usually reported in percentage points. - **WOMAC**: A questionnaire measuring osteoarthritis pain, stiffness and physical function. Check the scale used in each report. - **AHI**: Apnea–hypopnea index: breathing pauses or reductions per hour of sleep. - **MRI-PDFF**: An MRI estimate of the fraction of liver tissue signal attributable to fat. - **UACR**: Urine albumin-to-creatinine ratio, a measure of albumin leakage into urine. - **eGFR**: Estimated glomerular filtration rate. An estimate of kidney filtration, not a direct measurement of repaired tissue. - **MASLD / MASH**: Metabolic dysfunction-associated steatotic liver disease; MASH also involves inflammation and liver-cell injury. - **ApoB / MACE**: ApoB reflects the number of atherogenic particles. MACE is a composite of cardiovascular events whose exact definition varies by trial.