PEP-RTA-3 has an unusual problem for a research compound: too much publicity and not enough context. The headline figure from the obesity trial is quoted everywhere, usually without the dose, the duration, the comparator or the population it came from, and almost never with the observation that it is a human clinical result.
This post is the counterweight. Five primary papers, what each one was actually built to measure, the numbers as published, and where the boundary sits between that literature and anything you would do with a vial.
The five papers
| Paper | What it was | Primary question |
|---|---|---|
| Coskun 2022, Cell Metabolism | Discovery and pharmacology | What does this molecule do at the three receptors, and in obese mice? |
| Urva 2022, The Lancet | Phase 1b, multiple ascending dose | Is it tolerated in people with type 2 diabetes, and at what exposure? |
| Jastreboff 2023, NEJM | Phase 2, obesity | How large is the effect, and over what period |
| Rosenstock 2023, The Lancet | Phase 2, type 2 diabetes | What does it do to HbA1c at 24 weeks? |
| Sanyal 2024, Nature Medicine | Phase 2a, liver fat | What does it do to hepatic fat at 24 weeks? |
Note what is not in that column. There is no phase 3 outcome trial in the list, no head-to-head against PEP-TZP-2 at matched exposure, and no in-vitro characterisation of the sort a bench researcher would actually reuse beyond the discovery paper.
Coskun 2022: where the compound came from
The discovery paper is the one that matters most for laboratory work and gets quoted least.
It describes LY3437943 as a triple agonist at the glucagon receptor, the GIP receptor and the GLP-1 receptor, and reports the in-vitro balance as balanced glucagon and GLP-1 receptor activity with more GIP receptor activity. In obese mice, it decreased body weight and improved glycaemic control, with the weight effect attributed to both increased energy expenditure and reduced calorie intake.
That last clause is the interesting one, because it is the claim the third receptor was added to support. Our post on why PEP-RTA-3 adds a glucagon receptor works through the design logic.
The full text is paywalled, so the potency figures at each receptor are not quoted here. If you need them for a methods section, the paper is the source and a summary is not.
Urva 2022: the first human exposure
A phase 1b trial in people with type 2 diabetes. Phase 1b studies exist to establish tolerability and pharmacokinetics, not efficacy, and the useful output is the structure the later trials inherited — including the escalation schedule that turns up again in every phase 2 arm.
Jastreboff 2023: the number everybody quotes
338 adults with a BMI of 30 or higher, or 27 to 30 with a weight-related condition, randomised across several arms against placebo. The design detail that matters when reading any summary of it: percentage weight change at 24 weeks was the primary endpoint, and the 48-week change was secondary.
That distinction is the whole reason this paper is worth reading rather than quoting. The figure that circulates online is the 48-week secondary endpoint from the highest arm — a larger number than the primary endpoint produced, reported at a later timepoint than the trial was powered for. Neither fact is usually attached to it. The results are in the paper, linked below, and reading them there is the point of this post.
Two findings from the same abstract travel far less well than the headline, and both are about tolerability rather than effect: gastrointestinal adverse events were the most common and were related to the amount given, and there were increases in heart rate that peaked at 24 weeks and declined thereafter. A summary that reports the weight figure and omits those is not reporting the trial.
Rosenstock 2023: glycaemic control, with an active comparator
281 participants with type 2 diabetes, on diet and exercise or on stable metformin. This trial is more informative than the obesity one for a reason that has nothing to do with the size of its result: it had an active comparator, dulaglutide, not just placebo.
That is what makes it readable. A placebo-controlled result tells you a compound does something. An active-controlled result tells you where it sits relative to an existing single-receptor agonist — which is the only comparison that says anything about whether the extra receptors earned their place. The lower arms landed close to the comparator and the separation appears higher up, which is the shape worth understanding; the figures are in the paper.
Worth noting for the design bet described in the glucagon-arm post: a glucagon receptor agonist alone would be expected to raise HbA1c. It did not.
Sanyal 2024: liver fat
A phase 2a trial with relative change in liver fat at 24 weeks as the primary endpoint, measured against placebo. Every active arm was reported as statistically superior to placebo at P < 0.001, and the paper reports how many participants in the higher arms reached liver fat below 5%.
This is the trial most directly downstream of the glucagon arm, since hepatic fat oxidation is the mechanism that arm is there to engage.
Why none of this is your result
Everything above shares four properties, and each one severs it from a laboratory experiment:
- It is human clinical data. Generated in people, under protocol, with informed consent, monitoring and stopping rules. Nothing in this catalogue is for human use, and no result above is a claim about the material we sell.
- It used pharmaceutical-grade material of known provenance. Manufactured, released and characterised under a regulatory framework that a research-grade vial does not sit inside — whatever its certificate says.
- Every figure is a least-squares mean from a modelled analysis of a defined population, with a dose-escalation schedule and a 24- or 48-week horizon. It is not a property of the molecule you can carry into a cell assay.
- The endpoints have no in-vitro equivalent. Body weight, HbA1c and hepatic fat fraction are not things a plate measures.
The longer version of this argument is in why clinical trial data does not transfer, and it applies to PEP-RTA-3 harder than to almost anything else in the catalogue precisely because the clinical dataset is so visible. What "research use only" actually means covers the framing from the other direction.
What the literature is good for: knowing which receptors are engaged and in what balance, knowing that the triple design does something the dual design does not, and knowing which questions are already answered so your experiment can ask one that is not.
Frequently asked questions
Is there a phase 3 trial?
Not in the published record as of this writing. Everything above is phase 1b or phase 2. Treating a phase 2 result as a settled characterisation of a compound is the most common way the literature gets over-read.
Why do different sources quote different weight-loss figures?
Because they are quoting different endpoints, and usually not saying which. The primary endpoint was measured at 24 weeks; a larger figure exists at 48 weeks as a secondary endpoint; different arms give different numbers again. Several of the figures in circulation are all from the same paper and all correct, and none of them is "the" figure. That is the reason to read the paper rather than a summary of it — this one included.
Can I cite these papers in my methods section?
Cite them for what they establish — the receptor profile, the design rationale, the provenance of the compound. Do not cite a clinical endpoint as support for an in-vitro hypothesis; the reviewer will notice. Our post on reporting a peptide in a methods section covers what belongs there.
Does the certificate on a vial relate to these trials in any way?
No, and it is worth being blunt about it. A certificate reports purity by HPLC-UV and identity by LC-MS for one lot of material. It says the vial contains what the label says, to a stated purity. It says nothing about how that material would behave in any biological system, and the trials say nothing about the vial. Our certificate page has the current lots, and how to read an HPLC chromatogram covers what the purity figure is actually measuring.
Which compounds in the catalogue have a comparable literature?
PEP-TZP-2 and PEP-SMG-1, both with larger and longer datasets, and both with the same transfer problem. The receptor and target table lists the primary pharmacology reference for each compound we stock.
References
- Coskun T, Urva S, Roell WC, et al. Cell Metabolism 2022;34(9):1234-1247.e9. doi.org/10.1016/j.cmet.2022.07.013
- Urva S, Coskun T, Lou MT, et al. LY3437943, a novel triple GIP, GLP-1, and glucagon receptor agonist in people with type 2 diabetes: a phase 1b, multicentre, double-blind, placebo-controlled, randomised, multiple-ascending dose trial. The Lancet 2022;400(10366):1869-1881. doi.org/10.1016/S0140-6736(22)02033-5
- Jastreboff AM, Kaplan LM, Frías JP, et al. New England Journal of Medicine 2023;389(6):514-526. ClinicalTrials.gov NCT04881760. doi.org/10.1056/NEJMoa2301972
- Rosenstock J, Frias J, Jastreboff AM, et al. The Lancet 2023;402(10401):529-544. ClinicalTrials.gov NCT04867785. doi.org/10.1016/S0140-6736(23)01053-X
- Sanyal AJ, Kaplan LM, Frias JP, et al. Nature Medicine 2024;30(7):2037-2048. doi.org/10.1038/s41591-024-03018-2
Every product mentioned is sold for laboratory research use only and is not for human or animal use. Nothing on this page describes or recommends use of the material sold here in humans or animals.
