When you read that a compound “fights cancer,” the question that matters most isn’t what did it do — it’s where did it do it? In a laboratory dish? In a mouse? Or in people with cancer?
That distinction — the evidence tier — separates a biologically interesting finding from a promising candidate from an intervention actually shown to help patients. It’s also the distinction that headlines, product labels, and social-media summaries routinely blur. A compound doesn’t become a cancer treatment simply because it showed anticancer activity somewhere in the research process.
The evidence ladder
Broadly, evidence about human benefit climbs from laboratory studies, through animal studies, to clinical research in people. Each rung moves closer to the reality of treating a patient. This doesn’t mean every animal study beats every cell study — it means each step tests the idea in a more realistic setting. We describe the tiers in more detail in the evidence hierarchy.
Cell studies (in vitro)
In in vitro research, the compound is applied directly to cancer cells in a controlled dish. This is where mechanisms are found — whether a compound blocks an enzyme, disrupts a growth signal, or triggers apoptosis. But a dish is not a body. The compound doesn’t have to survive stomach acid, cross the gut wall, escape the liver, or reach a tumor. So a substance can kill cancer cells at a concentration that could never be safely reached — or reached at all — in a person. Cell studies show an effect is biologically possible, not that it will happen after you swallow a capsule.
Animal studies (in vivo)
Now the compound is tested in a whole living organism, usually a mouse — so it must be absorbed, circulate, be metabolized, and actually reach the tumor. Positive in vivo evidence is a genuine step up: the effect isn’t confined to an artificial dish. But mice aren’t people, and most treatments that work in mice never pan out in humans — the reasons have their own companion article.
Human studies (clinical)
The real test: does it help actual patients, safely? This is the only tier that can honestly claim benefit — and for most nutraceuticals it’s the thinnest, or missing entirely. Even here, “studied in humans” isn’t “proven”: an early trial may only test safety and dose, and a small uncontrolled study can’t separate a real effect from patient selection or chance.
So what does “preclinical” mean?
Simply: everything before human trials — the cell and animal work. It’s where the scientific case is built, not where the clinical conclusion is reached. The U.S. Food and Drug Administration describes preclinical research as including both in-vitro and in-vivo studies, used to establish dosing and toxicity before a candidate is ever tested in people. A mountain of preclinical evidence is a reason to keep studying a compound — not a reason to take it as treatment.
How to read a preclinical claim
When something is called “anticancer,” ask what the evidence actually shows:
- Isolated cells — and was the tested concentration realistically achievable?
- Was the animal dosed orally or by injection, at a dose a human could tolerate?
- Did treatment slow tumor growth, or actually improve survival?
- Was toxicity assessed? Has it been reproduced? Is there any human data?
A positive animal study can fairly support: “the compound showed antitumor activity in a mouse model.” It cannot, by itself, support: “the compound treats cancer in humans.”
Where Mechanica Natura draws the line
This is why, across the site, we write “studied,” “associated with,” or “demonstrated in preclinical models” rather than presenting lab findings as established clinical effects. It’s also why some compounds receive only partial strategic roles: the mechanism may be credible and the animal findings genuinely encouraging, while the human evidence stays thin, indirect, or absent.
That isn’t caution for its own sake. It’s an honest description of exactly which rung of the ladder a compound has reached. Preclinical evidence is a reason to keep asking the question — not yet the answer.
← All articles