Feb. 12 (Bloomberg) -- More than half of lung and colon cancer patients may benefit from high-speed tests that detect DNA flaws doctors can target with existing medicines, a study found.
Researchers used a gene test made by closely held Foundation Medicine Inc. to sequence 145 cancer-associated genes in 40 colon tumor samples and 24 lung tumors. They found that 53 percent of colon tumors and 71 percent of lung tumors had mutations that may be attacked with cancer medicines on the market or in human trials, according to the study published today in Nature Medicine. In some cases, the results revealed what drugs wouldn’t work against the tumors.
The study from researchers at Foundation Medicine and the Dana-Farber Cancer Institute in Boston, shows the value of using DNA sequencing machines to optimize treatment by matching drugs against specific gene abnormalities inside a patient’s tumor, said Pasi Janne, a study co-author.
“It is moving closer and closer to real personalized medicine,” Janne, a lung cancer specialist at Dana-Farber, said in an interview. “It is fantastic as we can tailor our therapy to the particular genetics of a patient’s cancer.”
The DNA sequencing field has drawn increased interest from pharmaceutical makers focused on developing gene-targeted therapies. Roche Holding AG, the world’s biggest maker of cancer medicines, last month began a $5.7 billion hostile takeover offer for Illumina Inc., the maker of gene sequencing machines that Foundation Medicine uses in its tests.
The researchers also spotted a previously unknown genetic flaw in 2 percent of 561 lung tumors tested. The flaw activates a growth-boosting protein targeted by Pfizer Inc.’s kidney- cancer drug Sutent, hinting that the treatment from the New York-based drugmaker may also work in these lung patients, said Janne. He wants to begin a trial of Sutent in lung-cancer patients with the gene change by year end, he said.
Researchers in Japan also reported finding the same new genetic change in a fraction of lung tumors, according to two other studies published today in Nature Medicine. Until the three new studies, the genetic change had never been seen in any cancer, said Janne.
The change fuses two unrelated genes together to form KIF5B-RET, turning on a growth-driving protein called RET that is usually not active in lung cells. When Janne and his collaborators treated cells with the aberrant gene using Pfizer’s Sutent or AstraZeneca Plc’s thyroid-cancer drug Caprelsa, the cells died. Both drugs block RET.
Finding Gene Abnormalities
Maureen Cronin, a study co-author and molecular pharmacologist at Cambridge, Massachusetts-based Foundation Medicine, said her company was finding new gene abnormalities at a much higher rate than they expected as it performs DNA scans on tumors.
“We expected to find new things, but not at the frequency we are finding them,” she said in a telephone interview. The results “are very surprising.”
The study also suggests cancer researchers may need to rethink the way they classify and treat the disease, Cronin said. The particular genetic abnormality inside tumor DNA may matter as much as what organ the tumor came from, she said.
Pfizer is aware of the new lung cancer gene finding and “believes the data are interesting,” said Jenifer Antonacci, a company spokeswoman, in an e-mail.
Laura Woodin, a spokeswoman for London-based AstraZeneca, said the company “is constantly alert to new developments and research in the science of oncology and we review relevant, peer reviewed studies for what they might mean for patients and drug development.”
Foundation Medicine performs a $5,800 test that takes tumor samples and sequences DNA from 200 genes relevant to cancer. It is funded with $33.5 million in venture capital from Third Rock Ventures, Kleiner Perkins Caufield & Byers and Google Ventures, according to its website.
--Editors: Angela Zimm, Andrew Pollack
To contact the reporter on this story: Robert Langreth in New York at firstname.lastname@example.org;
To contact the editor responsible for this story: Reg Gale at email@example.com