A new cause of drug resistance in patients taking the lung cancer drug gefitinib (Iressa) has been discovered along with a strategy for reversing the resistance.

Researchers have found extra copies of a gene called MET in the tumors of some lung cancer patients who stopped responding to gefitinib. Experiments with resistant cells showed that this genetic change - an amplification - can cause the drug to stop working.

Dr. Pasi Jänne of the Dana-Farber Cancer Institute and his colleagues also found that the resistance could be reversed, at least in a cell model. They did this by treating resistant cells with a combination of gefitinib and a compound that inhibits MET.

It is not yet clear whether such a combination therapy will work in patients, but an answer should come in the next few years. Many companies are developing agents that target MET or its partners, and the first such drugs are in early-stage clinical trials.

The MET gene was discovered at NCI in 1984 in the laboratory of Dr. George Vande Woude. It has since been implicated in two dozen cancers.

At least for now, the new results are relevant only to a small subset of lung cancer patients who have mutations in the gene EGFR. (About 10 percent of lung tumors have EGFR mutations.)

Since the findings were announced last month (and reported online in Science on April 26), lung cancer patients have been contacting physicians about the research.

"Patients with cancers involving EGFR mutations should be aggressive in searching out clinical trials with novel approaches to treatment if the initial therapies stop working," says Dr. Jeffrey Engelman of Massachusetts General Hospital, the study's first author.

Most lung cancer patients with EGFR mutations respond well to gefitinib and a related drug, erlotinib (Tarceva), but resistance usually occurs within 2 years. In half the cases, this happens because of new mutations in EGFR or in the genes it controls.

The MET gene, however, is not controlled by EGFR, nor does it normally interact with gefitinib. So before this study, researchers did not suspect that MET played a role in gefitinib resistance.

It now appears that MET and EGFR may be interchangeable members of the same family of kinase genes. Both produce protein receptors that sit on the cell surface and relay messages to the interior, including signals about growth and proliferation.

"The MET receptor plugs into the same signaling network that the lung cancer cells were using," says Dr. Jänne. In effect, MET restores the flow of cancer-promoting signals into the cell.

The question now is how often and under what circumstances the MET amplification occurs, says Dr. John Minna of the University of Texas Southwestern Medical Center, who developed the lung cancer cells used in the study.

He agrees that patients whose tumors progress while they are taking gefitinib or erlotinib should be evaluated for participation in one of the clinical trials testing the new inhibitors.

"We need to bring some MET-targeted drugs into the clinic and see whether or not they are going to work on these tumors with MET amplifications," says Dr. Minna.

Without a cell model of resistance, this research might not have been possible, because tissue samples from patients who develop resistance are extremely rare. After a drug fails, patients rarely have biopsies to determine what caused the failure, yet researchers need these tissue samples to understand how resistance occurs.

"Fortunately, we were able to figure out what was going on in the laboratory and then test the hypothesis on the precious few clinical samples we had," says Dr. Engelman. He obtained 18 samples from patients around the world.

In the future, patients may elect to have biopsies after a drug fails if the resulting information is likely to improve treatment, Dr. Engelman notes. He believes cancer physicians are moving toward an approach to drug resistance pioneered for treating HIV and tuberculosis.

"That means you attack the cancer based on how it would likely become resistant and by doing so you delay resistance and achieve longer term remissions," he says.

By Edward R. Winstead