Retatrutide

At week 40, HbA1c had fallen by 1.69%, 1.86% and 1.94% with retatrutide 4, 9 and 12 mg, compared with 0.81% with placebo. Average body weight fell by 11.5%, 13.9% and 15.3%, versus 2.6% with placebo. Digestive side effects were the most common, and two people in the 4 mg group died.

• A 40-week Phase 3 trial, so results cannot be extended to longer treatment or to how long the effects last.
• Retatrutide was compared with placebo only, so the trial does not show how it compares with other diabetes medicines.
• Participants had type 2 diabetes managed with diet and exercise alone (HbA1c 7.0% to 9.5%) and had not recently used diabetes medicines, so results may not apply to people on other treatments or with other HbA1c levels.
• This summary uses the published abstract and the registry record; the full text was not reviewed, so subgroup results, additional endpoints and full safety tables are not covered.
• The abstract gives HbA1c differences from placebo with confidence intervals but gives weight changes only as group averages, without differences from placebo or p-values.
• The registry records the three retatrutide groups as Dose Level 1, 2 and 3 without milligram doses, and the sources reviewed do not say which dose level corresponds to which of the 4, 9 and 12 mg doses.
• The study was sponsored by Eli Lilly and Company, the developer of retatrutide.

Across all randomised adults, body weight fell by 17.6%, 23.7% and 25.0% with retatrutide 4, 9 and 12 mg versus 3.9% with placebo. In the knee-osteoarthritis group, pain scores fell more with retatrutide (9 and 12 mg) than with placebo, and in the sleep-apnea group breathing events per hour fell more with retatrutide than with placebo. The abstract says digestive events were the most common side effects.

• Only the PubMed abstract of the journal report was reviewed, not the full paper, so details of safety, discontinuations and baseline characteristics are not covered.
• Weight, knee-osteoarthritis pain and sleep-apnea results are three separate analyses with different participant groups and estimands; they should not be combined.
• The knee-pain and sleep-apnea results come from subgroups (574 and 243 people) and use a hybrid estimand that treats stopping treatment as failure, which differs from the intention-to-treat weight result.
• The 12 mg hybrid sleep-apnea value is printed without a sign in the abstract and could not be confirmed against the full text.
• The registry (2,335) and the paper (2,339) give slightly different participant counts, and the sources do not explain why.
• The trial excluded people with diabetes, so the results do not describe people with type 2 diabetes.
• The study was funded by Eli Lilly.

• Number of participants randomised: ClinicalTrials.gov reports 2,335 (actual enrollment); NEJM abstract reports 2,339 underwent randomization. The paper reports 4 more randomised participants than the registry count. The reason for the difference is not established from the sources reviewed.
• Sign of the 12 mg sleep-apnea AHI change (hybrid estimand): NEJM abstract, hybrid-estimand values reports -22.9, -34.3 and 31.7 events per hour versus -9.9 with placebo; NEJM abstract, ITT values and stated differences reports ITT 12 mg value -32.1; hybrid difference from placebo -21.9. The abstract prints the 12 mg hybrid value as 31.7 with no minus sign, while the stated difference from placebo (-21.9) and the ITT value (-32.1) indicate a decrease. The reason for the difference is not established from the sources reviewed.

At 24 weeks, HbA1c had fallen 0.43 percentage points on retatrutide 0.5 mg, 1.39 and 1.30 on the two 4 mg arms, 1.99 and 1.88 on the two 8 mg arms and 2.02 on 12 mg, versus 0.01 on placebo and 1.41 on dulaglutide (journal values). Every retatrutide arm except 0.5 mg beat placebo (P<0.0001). Only the 8 mg slow-escalation and 12 mg arms beat dulaglutide (P=0.0019 and 0.0002). At 36 weeks, body weight fell by 3.19% on 0.5 mg and 7.92% to 16.94% on the higher doses, versus 3.00% on placebo and 2.02% on dulaglutide. Doses of 4 mg and above lowered weight more than placebo and dulaglutide.

• A Phase 2 dose-finding study of 281 people over 36 weeks; the arms were small (23 to 47 people) and it was not designed to show long-term benefit.
• Eight groups with different dosing schedules were compared, so results for one schedule do not apply to another.
• Only the PubMed abstract of the journal report was reviewed, not the full paper.
• The registry and the journal give slightly different HbA1c figures for the placebo and 12 mg groups, and the sources do not explain why.
• All participants were adults with type 2 diabetes treated at U.S. centres, mostly White, so results may not apply to other groups.
• Only the dulaglutide comparison speaks to how retatrutide compares with another medicine; the study did not compare it with other weight-loss or diabetes treatments.

• Change in HbA1c at 24 weeks in the placebo group and the 12 mg group: Lancet abstract reports Placebo -0.01; 12 mg -2.02 percentage points; ClinicalTrials.gov posted results reports Placebo -0.05; 12 mg -2.01 percentage points. Differences of 0.04 and 0.01 percentage points; all other arms match. The reason for the difference is not established from the sources reviewed.

At 24 weeks (primary endpoint) average body weight changed by -7.2% (1 mg), -12.0% and -13.8% (4 mg groups), -16.6% and -18.3% (8 mg groups) and -17.4% (12 mg), compared with -1.6% on placebo. At 48 weeks the changes were -8.7% (1 mg), -16.3% and -17.8% (4 mg), -21.7% and -23.9% (8 mg) and -24.2% (12 mg), compared with -2.1% on placebo. At 48 weeks, 100% of people on 8 mg and 12 mg had lost at least 5% of their weight, and the share losing at least 15% ranged from 16% (1 mg) to 83% (12 mg) versus 2% on placebo (model-based estimates). Within each maintenance dose, the groups that started lower (2 mg) had less nausea than those that started higher (for example 6 of 35 versus 21 of 35 people on 8 mg).

• A Phase 2 dose-finding trial of 48 weeks in adults without diabetes, all enrolled in the United States; it does not show long-term weight loss or effects on health outcomes.
• The dose groups were small and unequal (33 to 70 people), and the registry reports the groups that started at different doses separately.
• The registry's weight figures are model estimates that exclude data collected after people stopped the study drug, and the shares of people reaching 5%, 10% or 15% weight loss are estimates from a statistical model.
• More people left the placebo group early (20 of 70 did not complete), mostly by withdrawing (13), than the retatrutide groups (5 to 9 per group, out of 33 to 69).
• The trial was funded by the drug's developer, and only the journal abstract was reviewed, not the full paper.