The Cancer Drug Mechanism That Fooled Its Own Designers

An experimental cancer drug already being tested in patients entered clinical trials on a mistaken understanding of its cancer drug mechanism. The drug, zavondemstat, was engineered to block a family of proteins called KDM4 — but researchers led by Professor Lenka Munoz at the University of Sydney found it largely works through a completely different route: shutting down an enzyme named DHODH. The findings were published in Nature Chemical Biology in October 2026, and they raise uncomfortable questions about how well science understands the medicines it puts into human trials, according to MedicalXpress.

Zavondemstat was developed for advanced and metastatic cancers, including colorectal, pancreatic and prostate tumors. Its close chemical cousin, the compound QC6352, had become a standard laboratory tool that researchers around the world used to study KDM4 biology — and the drug rode into human testing on the assumption that blocking KDM4 was the key to its effect. When the assumed cancer drug mechanism is wrong, the consequences ripple outward: trials enroll patients selected for the wrong biological signal, researchers interpret results through the wrong lens, and years of funding chase a target that was never the real one.

How a Brain-Cancer Side Quest Exposed the Real Cancer Drug Mechanism

The mix-up came to light during an attempt to repurpose the drug against glioblastoma, the most common aggressive brain cancer, which carries a median survival of only 14.6 months even with aggressive treatment. The Sydney team, working with collaborators at Goethe University, the University of Oxford and the Institute of Cancer Research in London, tested the compounds on patient-derived glioblastoma stem cells — the treatment-resistant cells that drive the tumor's return. The lab compound proved remarkably potent, cutting cell viability by more than half at a 200-nanomolar dose, and mice carrying human glioblastoma tumors lived 34.5 days on treatment versus 28.5 days without it.

Then came the inconsistency that unraveled everything. A rival KDM4 blocker with almost identical potency against the KDM4A protein had no effect on the tumor cells at all, even at far higher doses. If the protein family were the true target, the two compounds should have behaved the same way. Instead, the results pointed to a different cancer drug mechanism entirely. A battery of genetic and molecular experiments traced the anti-cancer effect to DHODH, the enzyme that manufactures pyrimidines — the molecular bricks cells need to copy DNA before dividing. Munoz described DHODH as "a machine producing bricks needed to build new DNA," a vivid shorthand for why starving the supply chain stops fast-dividing cancer cells cold.

Pinning down the correct cancer drug mechanism reframes what the compounds actually do. KDM4 proteins are epigenetic regulators that adjust how DNA is packaged, acting like dimmer switches on gene activity. The enzyme sits far downstream in the cell's supply chain, churning out raw materials for DNA synthesis. Hitting the supply chain explains why the drugs worked on tumors no one expected KDM4 blockers to touch — and why the lab's standard research tool had been quietly misleading the field about the protein family's role for years.

Why the True Cancer Drug Mechanism Changes What Comes Next

The most immediate consequence of the corrected cancer drug mechanism is a new therapeutic lead. DHODH inhibition is now a plausible avenue against glioblastoma, a disease with brutally few options, and several drugs that target the enzyme are already being investigated for other cancers, raising the possibility they could eventually be tested in brain-cancer patients, as reported by News-Medical. DHODH is already a validated drug target outside oncology — blockers of the enzyme are used to treat rheumatoid arthritis and multiple sclerosis — which could shorten the path to repurposing, as reported by Genetic Engineering and Biotechnology News.

The team also built new compounds that inhibit KDM4 without touching DHODH, finally giving researchers clean tools to study what the protein family actually does. That matters because years of published studies used the old compound as a probe of KDM4 biology, and some of those conclusions may now need revisiting — a reckoning the authors say the field should face head-on rather than quietly filing away.

A Familiar Plot Twist in Drug Development

Here is the counterpoint the headlines could easily miss: this is not a story about a drug that failed. Zavondemstat was shrinking tumors and stalling cancer cells the whole time — only the label on the mechanism was wrong. History is full of medicines that worked first and were understood later, and Munoz noted there are well-known cases of cancer drugs advancing through large trials before researchers realized the mechanism was not what they had assumed. The discovery arguably improves the drug's prospects: future trials can now select patients based on DHODH dependence rather than KDM4 status, the exact kind of patient-selection fix she argues the field needs.

There is also a lesson for anyone watching clinical trials from the outside. When a cancer drug mechanism is mislabeled, the error is invisible to patients and investors alike — everything looks normal until someone runs the control experiment that should have been run years earlier. For a generation that follows science news in real time, the takeaway is less cynical than it sounds: the correction came from the scientific process working as designed, with one team's careful follow-up catching what the rest of the field had assumed. Rigor before trials protects patients, steers scarce funding toward genuinely promising treatments, and, as this case shows, can open doors to therapies nobody was originally looking for.

Related reading: Bowel Cancer Is Rising in Young Adults. Scientists Want Answers — and explore more explainers in Deep Dives.