NIH Launches Cancer Genome Project

Still other cancers result when rogue molecules attach themselves to genes whose job is to control cell division. Such "epigenetic" changes are invisible on standard tests that look for mutated genes because the genes themselves are healthy but are being manhandled by the other molecules.

The Cancer Genome Atlas faces formidable challenges, not least of which will be collecting hundreds or thousands of tumor specimens -- each of which must be properly preserved, freely donated for research by patients who have been informed of the project's goals and accompanied by detailed clinical information such as the tumor's response to various drugs.

Then will come the tedious work of analyzing all the relevant genetic traits of those cells and correlating those details to the cancer's clinical characteristics. Collins likened that undertaking to "thousands of human genome projects," a reference to the $2 billion effort, completed in 2003, to learn the order of all 3 billion letters of genetic code that go into making and maintaining a human being.

Collins emphasized that gene research costs have dropped considerably in recent years and expressed hope that an initial "back-of-the-envelope" estimate of $1.5 billion for the cancer atlas will prove high.

As the atlas gets filled, its contents are to be placed on a computer database freely accessible to researchers, doctors, patients and the public.

Not everyone supports the effort. In letters to scientific journals and in other venues, some scientists have complained that now is not the time for such an effort, given the already pinched federal research budget. Others have suggested that the approach is bound to fail on scientific grounds.

Garth Anderson, a cancer geneticist at Roswell Park Cancer Institute in Buffalo, said that because cancers are by definition genetically unstable, even a single tumor has cells with different mutations, many of which played no role in its emergence. And while a few new molecularly targeted drugs have worked against cancers where other drugs have failed, he said, even those drugs typically buy patients just a few months before the tumor finds a way around the medicine.

"If it's going to cost me $100,000 to add 45 days of life, I'd rather drop dead and give my kids the money," said Anderson, who added that he puts more faith in better tumor imaging and old-fashioned surgery.

But Bruce Stillman, president of the Cold Spring Harbor Laboratory in New York and one of several experts who advised NCI on the proposal, disagreed.

"It would be ridiculous to say at this time, 'Okay, we finished the genome. Let's not do anything with it.' To say we shouldn't be doing this is to say we're not interested in solving cancer."