I have identified 5 main claims from the AOH1996 paper (https://www.cell.com/cell-chemical-biology/fulltext/S2451-9456(23)00221-0?rss=yes)
1) Biochemical mechanism of action
2) Achieves inhibition of growth in many cancer types in-vitro at concentrations around 300nM or less
3) No toxicity to cultured human non-cancerous calles up to 10 μM concentrations
4) Favorable pharmacokinetics (availability after oral doses) in animals
5) Reduced xenograft tumor growth in mice
The question resolves YES if A) at least three of those claims are tested in published research from teams independent of the original AOH1996 paper authors AND B) At most one is found mostly incorrect. If there's no suspicion of problems, I'll take pre-registered studies with external oversight (e.g. a clinical trial) as "independent" even if the original authors are involved.
The question resolves NO if at least two of those claims are contradicted by later research.
If no attempts to replicate/confirm/disprove at least three of the claims are made in 8 years, this market resolves NA.
This is biomed, so I won't be particularly pedantic about what exactly means "confirmed" or "correct" - e.g. if another team reports inhibition of cancer growth in vitro at 800nM, I'd say it is correct as it is the same order of magnitude. No inhibition up to 1.5μM would probably be a problem, if it is for a set of cancer lines substantially overlapping those reported in the original paper.
When applicable, I'll also take clinical results as possible evidence (e.g. if dose-finding studies in humans indicate good pharmacokinetics, I'll take that as confirming the corresponding claim; OTOH lack of efficacy in human trials does not necessarily imply lack of in-vitro efficacy).
If there are conflicting reports from different labs, I'll use my best judgement to reconcile them and may wait for additional results.
As the resolution is partly subjective I will not bet on this market.
Related questions
@PanosFilianos I admit I don't have a good feel for how common this is. I was personally asked to analyze data for an in-vitro study on an already approved drug (https://doi.org/10.1016/j.jcf.2021.07.006), so I presume it is done at least sometimes. More in-vitro work is probably unlikely if the drug loudly fails in the early phases of the clinical trials. There already is a market resolved purely based on success in trials, if you prefer that: https://manifold.markets/RatUziCat/is-the-pcnatargeting-cancer-drug-pa)
@PanosFilianos There's also the interesting option of "the drug works in the clinic, but via a different mechanism than assumed during drug design", which I've heard happens ocassionally.
@AlexAmadori If nothing but a succesful Phase 2/3 happened, the resolution would depend on details - if the studies end up measuring availability of the drug in the tumors themselves as well as tumor growth, it might be enough to convince me that there's inhibition (claim 2) and good pharmacokinetics (claim 4). If concentrations in non-tumor tissue is measured and there are no side effects it might be evidence for no toxicity (claim 3). If however the published data would look only on overall survival (and no other in-vitro/animal studies), it would lead to NA resolution.
Also the drug may be shown to be clinically useful and pass Phase 2 or even Phase 3, but it may turn out that the mechanism is different than expected and there is substantial toxicity even at low concentrations. In this case the market would resolve NO despite the favorable clinical profile.