WEDNESDAY, Feb. 4 (HealthDay News) -- In a rude reminder that medicine is not yet an exact science, a Dutch study has found that adding a fourth anti-cancer drug to a three-medication treatment actually makes things worse for people with advanced colorectal cancer.

"The lesson is that there may be negative interactions between those inhibitors that may be detrimental to those patients, even when animal studies show benefits," said Dr. Cornelis J.A. Punt, a professor of medical oncology at Radboud University Nijmegen Medical Center in the Netherlands and lead author of a report in the Feb. 5 issue of the New England Journal of Medicine.

That lesson applies to proposed new treatments of all forms of cancer, Punt said. "A lot of new drugs are being developed that work through all sorts of new pathways," he said. "You should carefully design your studies of them before you use them in general practice."

The study of 755 people whose colorectal cancers had spread to other parts of the body was done because not only animal studies but two smaller human trials had found benefits from adding the antibody cetuximab (Erbitux) to a standard three-drug regimen of bevacizumab (Avastin), oxaliplatin (Eloxatin) and capecitabine (Xeloda), Punt said.

Each drug works in a different way. Capecitabine and oxaliplatin kill cancer cells directly, whereas bevacizumab inhibits vascular endothelial growth factor, a natural molecule that promotes cell division. Cetuximab inhibits the activity of another molecule, epidermal growth factor.

Yet the average survival time for people in the trial who got the four-drug combination was 9.4 months, compared with 10.7 months for those given the three-medication regimen. Also, adverse drug reactions were more frequent in those given the four drugs.

The reason for the negative results is unclear, Punt said. "It could be aggression between the antibodies of which the nature is unknown," he said. The fact that fewer side effects were seen in the animal studies could offer a clue, Punt said: "If you see less toxicity in the experimental animals, somehow the biological effect is less."

The finding holds a lesson applicable to all new treatments for cancer and other diseases, such as AIDS and tuberculosis, according to Dr. Robert J. Mayer, a professor of medicine at Harvard Medical School and vice chairman for academic affairs at the Dana-Farber Cancer Institute, who wrote an editorial accompanying the study.

"The message is that the way you learn whether adding more drugs to the treatment you have is beneficial is by conducting a proper clinical study," Mayer said.