TRIUMPH-1 in NEJM, TRIUMPH-2 in The Lancet, same day. The grade stays B. Here’s what moved.

On September 29, Lilly published two of its retatrutide Phase 3 trials in full on the same day: TRIUMPH-1 in the New England Journal of Medicine and TRIUMPH-2 in The Lancet. With TRANSCEND-T2D-1, published in The Lancet in June, that makes three of the five reported Phase 3 results peer-reviewed. Two remain company topline releases.

The grade on Peptide Briefs stays B. This post is about why, and about what the papers say that the May and July press releases didn’t.

Two numbers, both correct

Lilly’s May headline for TRIUMPH-1 was 28.3% weight loss at 12 mg. The NEJM paper’s primary result is 25.0%. Both appear in Figure 1 of the paper, and the difference is not a correction; it’s two questions. The efficacy estimand asks what happens if every participant stays on drug for 80 weeks. The treatment-regimen estimand, which NEJM reports as primary, counts the people who stopped. Across all three doses the peer-reviewed numbers are 17.6%, 23.7% and 25.0% vs 3.9% on placebo, against topline figures of 19.0%, 25.9% and 28.3%.

When a vendor or a headline quotes the higher figure, it isn’t wrong. It’s the optimistic one. The same applies to TRIUMPH-2’s 20.8%; the Lancet intention-to-treat number is [insert from paper].

What TRIUMPH-2 adds

TRIUMPH-2 is the type 2 diabetes cohort: 1,152 adults, 80 weeks, 4, 9 or 12 mg vs placebo. People with type 2 diabetes reliably lose less on incretin drugs than people without it, and 20.8% at 12 mg is the largest weight loss reported in that population. A1C fell 1.4 to 1.6 points from a 7.7% baseline vs 0.2 on placebo, and up to 40% of participants reached an A1C below 5.7%, the non-diabetic range. Triglycerides fell 39.5%, systolic blood pressure 10.8 mmHg, and hsCRP 58.3% at the top dose.

What the safety tables add

Three things the press releases didn’t lead with.

Dysesthesia. A skin burning or tingling sensation occurred in 12.3% of participants on 9 mg and 12.5% on 12 mg in TRIUMPH-1, against 0.9% on placebo. It isn’t a known GLP-1 or GIP effect; the working assumption is the glucagon component. Lilly reports most cases were mild and resolved on treatment.

Dropouts by dose don’t line up across trials. In TRIUMPH-1, discontinuation for adverse events rose with dose: 4.1%, 6.5%, 11.2% vs 4.6% placebo. In TRIUMPH-2 the middle dose had the most: 3.8%, 11.6%, 7.7% vs 4.9%. Two trials, two patterns; the honest reading is that we don’t yet know the dose-tolerability curve.

Orthostatic hypotension. 8.8% at 12 mg vs 0.9% placebo in TRIUMPH-1. Small in absolute terms, ten-fold relative.

Heart rate, the question most readers have asked, isn’t in the main NEJM tables. It’s in the supplementary appendix, and the page will be updated when we’ve read it rather than summarized from a summary.

Why B, not A

A on our scale means the full Phase 3 program is peer-reviewed and a regulator has reviewed the safety file. Three of five is not the program. Lilly says the BLA goes to FDA in Q1 2027; nothing has been reviewed yet. TRIUMPH-5, the head-to-head against tirzepatide, and the 10,000-participant outcomes trial are still running.

B is not a hedge. It’s the largest weight-loss effect ever recorded in a drug trial, replicated three times in peer-reviewed form, with a safety profile that’s heavier than tirzepatide’s at the top doses and one new adverse event that needs explaining. That’s what B means.