A friend of mine runs a company with an unusually good clinical result and an unusually bad financing problem, and the two are causally linked. The drug is not a new molecule but a reformulation of two chemotherapy agents that have been treating bowel cancer for decades, combined into a single administration, and the early human data is better than the standard of care it replaces. It is close to unfundable for exactly that reason: the capital allocated to a cancer drug tracks how well it can be defended, not how well it works.
The problem is delivery, not chemistry
Bowel cancer is the third most common cancer in the world, with roughly 1.93 million new cases a year, and the backbone of treatment has been unchanged for a long time. Patients receive 5-fluorouracil, usually as an intravenous infusion, alongside leucovorin, which is given either as separate pills or as its own separate infusion. Leucovorin kills nothing on its own and instead stabilises the interaction between 5-FU and its target enzyme, which is what makes the chemotherapy work harder, so the two have to arrive in the right relationship to each other. Achieving that today means interwoven dosing schedules, two delivery routes and a lot of chair time.
The company's insight was that the combination itself was never optimised. Five separate Phase 2 trials had already shown that co-administering the two agents roughly doubled efficacy against previous treatment, but every one of those studies delivered them independently, in separate arms, through separate lines. Nobody had put them into one formulation and asked what the pharmacokinetics would do. That is a formulation problem rather than a discovery problem, which is why it was solvable by two university chemists working on an existing research budget rather than by a discovery programme.
The clinical signal is strong and cheap
The 2019 Phase 1 study ran 40 patients across both bolus and infusional schedules and reported a 69 percent disease control rate, meaning the cancer at minimum stopped growing in roughly two thirds of patients. More useful for a clinician is the tolerability: the reformulation safely delivered 5-FU at doses up to 40 percent higher than standard regimens. In chemotherapy the dose ceiling is usually toxicity, so raising the ceiling without raising harm is the whole game. On the strength of the reformulation argument the FDA has since agreed the programme can skip Phase 2 entirely, which removes years and tens of millions from the path.
The origin story is worth noting because it explains the cost base. A cancer clinician approached a university chemist in 2004 asking whether these two drugs could be given better. The work ran on an existing research budget and government commercialisation grants, through pre-clinical animal testing where a regulated housing unit for study rats runs upwards of fifteen thousand dollars. When the university tried to commercialise it and got nowhere after two years, its 25 percent interest was bought out for around a hundred and fifty thousand dollars. The science that produced a doubling of efficacy cost less than a graduate programme.
A patent you can read is a patent you can copy
Here the economics turn against the asset: a reformulation is patentable, and this one now has protection running to 2040, but the patent publishes the formula, and that formula is a combination of two generic drugs whose behaviour is thoroughly documented. Anyone competent can read the filing and reproduce it. The enforcement position is therefore weak in the specific way large pharmaceutical companies have learned to avoid, and the internal comparison they reach for is Viagra, a drug whose formulation was widely copied and whose patent proved expensive and unsatisfying to defend.
That is the crux. A novel molecular entity carries an unknown safety profile, a decade of trials and a genuine chance of failing outright, and it is worth funding because if it works nobody else can make it. A reformulation of two known agents carries a well-understood safety profile, a shortened regulatory path and a strong early signal, and it is hard to fund because if it works everybody else can make it. The system is not selecting for the drug most likely to help patients but for the drug most likely to sustain a monopoly.
The trial is the cost, and the trial is the wall
The capital required scales with the size of the human study rather than with the difficulty of the science, and Phase 3 is where that becomes prohibitive. The company is raising roughly five million dollars at listing to reach the door of Phase 3. Walking through it costs eighty million or more, and completing it runs into the hundreds of millions. Against a hundred and fifty thousand dollar buyout of a quarter of the asset a decade ago, the ratio is the entire argument: discovery was almost free, and proof is what nobody can afford.
Geography is a discount
Being Australian makes all of this harder in ways that have nothing to do with the molecule. Domestic research funding has tightened to the point where roughly 8 percent of grant applications succeed, against about 20 percent a decade ago, so the pipeline that should carry an asset from bench to clinic is largely closed. The local biotech investor base is small and trend-driven, and the handful of companies already in Phase 3 on the ASX absorb most of the available liquidity, which leaves earlier assets competing for what is left. American investors, meanwhile, discount Australian biotech almost reflexively. Add a biotech bear market that has persisted since the post-Covid boom unwound and you have an asset with good data in the wrong postcode.
The early capital came from the only places that would supply it: friends, family, doctors and academics, assembled through a government tax rebate scheme designed for exactly this gap. That is a reasonable way to fund a Phase 1 and no way at all to fund a Phase 3, and the gap between those two facts is where most reformulation science in this country ends.
What this says about drug development
Two problems sit underneath this, and neither is about oncology: the first is that which drugs reach patients is decided by institutions optimising for defensible returns rather than by clinical need, so an improvement that cannot be monopolised struggles to attract the capital required to prove it, no matter how strong the signal. The second is that the grant system meant to catch exactly these cases has narrowed to an 8 percent hit rate, which means the filter is now mostly noise and good projects are missed by default rather than by judgment.
I do not think this resolves through better intentions so much as through cheaper proof. Anything that lowers the cost of running a credible Phase 3, whether that is trial design, real-world evidence, or regulatory willingness to accept reformulation data as the FDA has done here, changes which assets are fundable at the margin. Until then the pattern holds: the molecule that helps the most patients and the molecule that gets funded are selected by different criteria, and only one of them is measured in outcomes.