Research reading group · call recap
Phase two is not phase three
17 May 202616 attendedHeld by Sunil
The retatrutide phase 2 paper — Jastreboff AM, et al. New England Journal of Medicine 2023;389(6):514 to 526. Sixteen people, and the reason we came back to this one is that it is quoted at members constantly, usually in a form that skips the words phase two entirely. The room did the arithmetic on what a phase 2 result can and cannot support and then spent twenty minutes on why this particular paper travels so well.
What the paper is and is not
A phase 2 dose-finding trial. Its job is to establish whether a compound does something and at what doses it might be worth testing properly. The population is smaller, the follow-up shorter, and the endpoints are chosen to inform a later trial rather than to answer a clinical question.
What it is not is evidence about long-term safety, about outcomes that matter to a person over years, or about how the compound performs against the best available alternative. Those are phase 3 questions and, at the time of the read, they were not answered.
Why the claim outran the paper
Because the effect sizes were striking and because a striking number in a small early trial is the most shareable object in medicine. Every member in the room had seen it quoted without qualification. Three had been told by somebody outside the community that it makes everything currently available obsolete.
Nils gave the plain version: early trials systematically produce larger effects than the trials that follow them, for reasons that have nothing to do with anybody lying. Nadja, who trusts nothing once, added the other half — that the paper itself is careful and the carefulness is in the parts nobody quotes.
What the room concluded
That the compound is genuinely interesting and that nobody in the room knows what it will look like in a completed programme, and that both of those statements are allowed to sit next to each other. Also, from me, the point I make in every circle where this comes up: material sold for research use only is not approved for human use, and a promising phase 2 paper does not change that in the smallest degree.