Genetic test aids lung cancer fight

Scientists have developed a powerful new genetic test to determine which lung cancer patients should get chemotherapy, saying the advance could save thousands of lives every year and heralds the growing importance of genetic profiling in cancer care.

Patients whose lung cancer is caught at the earliest stage typically do not get chemotherapy, because the benefits outweigh the toxic side effects of the drugs only for a subset of patients likely to suffer a recurrence. The new test allows doctors for the first time to identify which patients are at greatest risk for a recurrence -- an almost certain death sentence -- and then give them chemotherapy.

The results, described in today's issue of the New England Journal of Medicine, are so promising that researchers are launching a broad, government-funded clinical trial of the test within the next six months, including some 1,200 patients at about 60 sites in the United States and Canada.

The new test is welcome news for the treatment of lung cancer, which kills more Americans than any other cancer. Researchers also said it is a sign that a much-anticipated transformation in cancer treatment is gathering momentum. Using new genetic tools, doctors are learning how to measure the characteristics of a particular patient's tumor and use this information to mount a more effective counterattack. This ``personalized medicine" is already saving lives of patients with breast cancer and leukemia, and the list of diseases is now expected to grow. Similar tests are being developed for colon, prostate , and other cancers.

``I think this is a very, very exciting tool," said Dr. Tom Lynch , director of the Center for Thoracic Cancers at Massachusetts General Hospital. ``The future of cancer therapy is going to be based on exactly what this paper is about."

One of the great frustrations of cancer medicine is that tumors that look the same do not always act the same, leaving doctors unsure how to treat patients. Now, scientists are learning to distinguish cancers by the genetic machinery at work in the tumors, paving the way for more effective therapies. Some tumors may have distinctive mutations in their DNA. For example, some breast cancer tumors have a mutation that leads them to produce an excess of a protein called ``HER2," and women with these aggressive tumors are being successfully treated with a drug called Herceptin.

Tumors may also have telltale patterns of genes that are active. At any given moment, some genes in a tumor are turned on, making proteins, and others are turned off. Identifying the ``on" genes is known as measuring ``gene expression," and this is the method used in the lung cancer study, which was led by scientists at Duke University.

The scientists at Duke identified patterns of active genes typical of more lethal tumors. The work could provide pharmaceutical companies clues that will help develop drugs and help scientists understand the disease's inner workings, according to Tyler Jacks , a lung cancer researcher who is also director of the MIT Center for Cancer Research.

The test could also have broad clinical applications, according to Dr. Anil Potti , who led the study and is an assistant professor at Duke. Initially, it will be used to identify patients with stage 1 non-small cell lung cancer, a common form of the disease, who would benefit from chemotherapy. The team also hopes to identify patients with later stages of the disease who are unlikely to benefit from chemotherapy, saving them the rigors of the drugs.

And, Potti said, the team has early data that suggest the same genetic tools can identify which chemotherapy drug will work best on which patient's tumors. This would represent another important advance and could also have applications in the treatment of breast cancer, where doctors use some of the same chemotherapy drugs, he said.

Between 40,000 and 45,000 patients are diagnosed with stage 1 non-small cell lung cancer every year in North America, Potti said. Typically, a surgeon will remove the tumor, but the patient is not given chemotherapy. Between 40 percent and 50 percent of patients will have a recurrence, and virtually all of these patients die. Preventing even some of these deaths would mean thousands of lives saved every year, Potti said.

The team analyzed samples of tumors from patients who had a recurrence and those who were cancer-free for five years, according to the journal article. The team discovered they could distinguish between the two groups based on patterns of activity in groups of genes. These patterns serve as a kind of fingerprint.

The team then used their test in three groups of patients and found that they were able to predict whose cancers would recur between 72 percent and 81 percent of the time.

There is currently a test, offered by California-based Genomic Health , that uses gene expression to predict the risk of recurrence in breast cancer. In June, the company said that early data showed that the same thing could be done with colon cancer.

As the research moves forward, it could also improve the process of developing cancer drugs, according to Dr. Todd Golub, director of the Cancer Project at the Broad Institute of Harvard and MIT. Genetics can be used to identify subpopulations of patients most likely to benefit from a particular drug ahead of time, meaning that clinical trials could be smaller, less expensive, and more likely to succeed, Golub said.

Gareth Cook can be reached at cook@globe.com.