Test may aid lung cancer survival

Duke discovery will be put to trial

In most cases, lung cancer is a death sentence. But a genetic discovery at Duke University may save lives.

Researchers there invented the first genetic test to identify patients with fast-growing lung tumors. Starting this fall, U.S. and Canadian doctors will use the molecular fingerprint to select patients who need extra treatment.

If added treatments work, fewer cancers may recur. More people may live.

"This is breaking into a new frontier. Instead of giving everyone the same treatment, we want to individualize," said Dr. David Harpole, a Duke surgeon who will run the 1,000-patient clinical trial.

Lung cancer, the deadliest form of cancer in the United States, killed more than 5,000 North Carolinians last year. Most of the 174,000 people diagnosed nationally each year develop non-small cell cancer, usually from smoking cigarettes. Eighty-five percent die.

The Duke test, described today in the New England Journal of Medicine, will be used first on patients diagnosed with non-small cell lung cancer, before it spreads. Those patients now have a 50-50 chance of surviving.

The test was invented at Duke's Institute for Genome Sciences & Policy, where Dr. Anil Potti hunts for patterns in the behavior of genes involved with lung cancer. Duke intends to patent it.

Potti used technology that let him see which genes are turned on high, low or off in tissue from scores of lung cancer tumors. Then he used statistical strategies and computer power to find patterns of genetic activity in the tumors.

The right test

Potti found one notable pattern among 133 genes, some involved in important functions such as cell growth and death. When he applied the gene pattern to lung cancers, it successfully picked out aggressive tumors in up to 79 percent of cases.

"I have been waiting for the right test and finally I think we have it," said Potti, an oncologist and genetics researcher.

Now, doctors often use less reliable factors to predict the ferocity of a patient's cancer, including the size of a tumor, whether the cancer has spread, and a person's smoking history, Potti said.

In the case of patients with early stage non-small cell lung cancer, doctors usually recommend surgery but not toxic chemotherapy. No studies have proved that chemotherapy extends enough patients' lives, said Harpole, the surgeon.

But in the trial of the new Duke test, patients with Potti's genetic fingerprint will receive chemotherapy as additional treatment.

In time, Potti expects genetic descriptions of different lung cancer types will advance treatment in other ways, too. Genetic insight should lead to more specialized drugs targeting the molecular biology of different cancer strains. That has already occurred in breast cancer treatment.

For Beth Williams of North Raleigh, a lung cancer survivor, any progress toward saving lives is worthwhile.

A retired computer programmer, Williams was diagnosed with early stage lung cancer in 2000. She would have taken a test to better gauge her type of cancer in a minute, she said, if one had been available.

"I don't know anyone who would say this is not a step toward beating this disease," she said.

Inadequate research

Like many people concerned about the toll of lung cancer, Williams, 67, says the disease doesn't get enough research attention. The nonprofit LUNGevity Foundation, which raises money for lung cancer research, calculated that federal spending on breast cancer research in 2005 equaled $13,704 per death from that disease, compared with $1,627 per lung cancer death.

"Lung cancer is a neglected issue," Williams said. "But it's the No. 1 cancer killer of men and women."

Potti, the genetics researcher, said people have very pessimistic views on lung cancer. That doesn't help.

"There is a lot of nihilistic behavior related to lung cancer. People say these people did it to themselves," Potti said. "If this test and this approach does any good, maybe it can give some people hope."