Research Headline: Mouse Study vs Human Trial

The gap the headline skips

"Scientists find a cancer breakthrough in mice" is a genre of headline. The finding underneath is often real. The problem is what sits between it and a person.

A laboratory result answers a question about biology: does this molecule do what we think it does, in this system, under these conditions? A clinical trial answers a different question: does giving this to people make them live longer or better than the alternative, and at what cost?

Nothing in the first answer guarantees the second. Mice used in cancer research are usually bred or engineered to be uniform. Their tumors are often implanted rather than arisen. Their immune systems, metabolism, and lifespan differ from ours. And the dose that works in a 25-gram animal may be intolerable in a person.

What the human path actually involves

NCI describes the sequence in four phases, and the numbers show why it takes years.

A phase 1 trial asks whether a new treatment is safe, what its side effects are, whether people can tolerate it, and the highest dose they can tolerate. It runs in a small group, roughly 15 to 30 people.

A phase 2 trial takes more people, roughly 50 to 100, and asks whether the treatment seems to work against the cancer — by shrinking tumors or slowing their growth. Safety monitoring continues.

A phase 3 trial compares the new treatment against current standard therapy, on both effectiveness and side effects. Participants are randomly assigned. NCI is explicit about why: randomization ensures that any differences found are real, and not the result of differences between the people in each group. These trials run from 100 to several thousand people.

Results from phases 1 to 3 feed regulatory decisions. A phase 4 trial then follows long-term safety and effectiveness in large, diverse populations after approval.

Our page on clinical trial phases covers each stage in more detail.

Why randomization is the hinge

NCI's explanation of bias is the clearest reason a mouse result cannot substitute for a trial.

If doctors could choose which patients went into which group, some might assign healthier patients to the treatment group and sicker patients to the control group without meaning to. The treatment would then look better than it is. Random assignment removes that possibility.

A laboratory experiment has controls too, but they control for laboratory variables. They cannot control for the thousand differences between one person and another that determine whether a drug helps. Our page on randomization in clinical trials explains the mechanics.

The step after the promising early result

Even in humans, an early promising result can fail to hold. That is not a hypothetical.

FDA's accelerated approval pathway lets a drug reach patients on an early signal — often tumor shrinkage — on condition that a confirmatory trial then shows real benefit. FDA maintains a public register of the ones that did not.

Atezolizumab is on it twice for bladder cancer. An indication approved on May 18, 2016 was withdrawn on April 13, 2021. A second, approved on April 17, 2017, was withdrawn on December 2, 2022. The same register lists the drug's triple-negative breast cancer indication, approved March 8, 2019 and withdrawn October 6, 2021.

These were not mouse studies. They were human data, in real patients, that then failed to confirm. That is the distance between a signal and a proof — even after the mouse stage is long past.

How to read a preclinical headline

Five questions do the work.

Was this done in people, in animals, or in cells in a dish? Has a clinical trial started, and if so, what phase? What outcome was measured — a molecular marker, tumor size, or how long people lived? What side effects are known so far? And is there anything a patient should do differently today?

For most preclinical headlines the honest answer to the last question is no, and that is not a failure of the science. It is where the science currently is.

When to get checked

Research literacy does not replace symptom awareness. NCI's guidance is that symptoms which do not get better after a few weeks deserve an appointment, and that cancer often causes no pain, so pain is not the threshold to wait for. NCI's list includes:

• A lump or firm area in the breast, neck, armpit, abdomen, or groin
• Blood in the urine or the stool
• A lasting change in bowel or bladder habits
• A cough or hoarseness that will not clear
• Trouble swallowing, or indigestion that will not settle
• Weight loss or gain with no explanation
• Fever or night sweats with no known cause
• Severe fatigue that persists
• A sore that does not heal, or a mole that changes
• Yellowing of the skin or the whites of the eyes

If you are in cancer treatment, a fever of 38 degrees Celsius, or 100.4 Fahrenheit, is an emergency rather than something to watch.

What this story cannot tell

• It cannot tell you whether a specific preclinical finding will ever become a treatment. Most do not.
• It cannot replace a treatment plan from a care team, or a trial publication, or a drug label.
• It cannot tell anyone to start, stop, or change treatment.
• It does not mean animal research is useless. Mechanisms found in the laboratory are how most human trials get their starting point.
• It does not mean a treatment that failed in a trial was never worth testing. Finding out is the purpose.

Sources

• NCI, How Do Clinical Trials Work? — https://www.cancer.gov/research/participate/clinical-trials/how-trials-work
• NCI, Cancer Clinical Trials Information — https://www.cancer.gov/research/participate/clinical-trials
• NCI, Symptoms of Cancer — https://www.cancer.gov/about-cancer/diagnosis-staging/symptoms
• FDA, Withdrawn: Cancer Accelerated Approvals — https://www.fda.gov/drugs/resources-information-approved-drugs/withdrawn-cancer-accelerated-approvals
• FDA, Clinical Trial Endpoints for the Approval of Cancer Drugs and Biologics: Guidance for Industry (December 2018) — https://www.fda.gov/media/71195/download

How this article was prepared

An AI-assisted editorial system helped prepare this page. No named medical reviewer has reviewed it unless one is listed.

The National Cancer Information Foundation publishes Cancer Explained. This page is for learning. It is not medical advice and does not suggest a test or treatment.

See an error, old source, or unclear wording? Tell us.