6 new genetic clues to breast cancer found

In a long-delayed harvest from the human genome project, researchers say they have found six new sites of variation in the genome that increase the risk of breast cancer.

Together with genes known earlier, the discovery means that a sizable fraction of the overall genetic risk of breast cancer might now have been accounted for, researchers say, and much of the rest could be captured within a few years.


The findings do not point to any new treatment and are too little understood to serve as the basis for a diagnostic test. But scientists say they are a critical step toward understanding the biology of breast cancer, from which new treatments should emerge.

Understanding the genetic basis of common diseases was the promised payoff of the $3 billion human genome project, which was completed in 2003 but has only slowly shown many tangible results.


But the logjam seems to be breaking. Last month, seven new DNA variations associated with diabetes were discovered, and advances with other diseases are expected to be announced soon.

The new findings have been made possible by new instruments, known as chips, that permit testing of up to 500,000 points of variation on the human genome simultaneously for possible association with disease. To attain statistical strength, large numbers of patients must be recruited, which has prompted otherwise competitive groups to work together.

Using the new chips, scientists can compare breast cancer patients with healthy people as they look for variant sites on the DNA of the human genome that seem associated with the disease. Because the chips sample each patient's entire genome, the new approach is known as a whole-genome association study.

With that approach, says a consortium of breast cancer scientists, led by Douglas F. Easton of the Cancer Research-UK Genetic Epidemiology Unit in Cambridge, England, it has found five new sites on the genome where a common variation confers a risk of breast cancer. Its findings were reported yesterday in the journal Nature.

Two of the same sites of variation have been found by a second team, led by David Hunter of the Harvard School of Public Health. Decode Genetics, a gene-finding company based in Iceland, has also identified two new sites of variation, one of which is the same as one found in Easton's study. The reports from Hunter and Decode Genetics were published online in the journal Nature Genetics.

Scientists involved in the three studies expressed confidence that most of the genetic risk of breast cancer will be detected through the whole-genome association approach. "Once the dust has settled, yes, it's possible we may have captured most of the genetic variation," Easton said.

Genes earlier known to be associated with breast cancer, such as BRCA1 and BRCA2, carry a high risk but are rare in the general population. The new DNA variations are common but confer a lesser risk.

A team lead by Simon N. Stacey of Decode Genetics has reported that 25 percent of women of European descent carry one copy of a DNA variant on the second of the 23 human chromosomes, conferring a 44 percent greater risk of breast cancer than for women without it. It also reported that 7 percent of women have inherited two copies, with a 64 percent greater risk.


The sites of variation discovered by Decode and the other two groups do not lie in genes and for the most part have no known biological function, though there is conjecture that some control the activity of nearby genes. It is on the basis of statistics alone that the sites of variation on the DNA are thought to be associated with disease.

Their discoverers say that the sites will, when understood, reveal new biology that underlies the progression toward breast cancer. But not everyone is convinced.

Mary-Claire King, a biologist at the University of Washington in Seattle who pioneered the search for the BRCA1 cancer gene, criticized the statistical basis of the studies and suggested that they should not have been published until the biological significance of the suspect sites had been established.

Her principal criticism of the three studies' statistics is that when 500,000 sites of variation are tested all at once for association with disease, many can come out positive by chance, not because of a link.