Breast cancer: hope for a genetic test.

The elation sparked by last fall's discovery of a gene linked to breast cancer gave way to pessimism when scientists reported finding 22 distinct mutations in the gene. This made routine screening for defects in the gene seem far beyond reach (SN: 12/3/94, p.372).

Now, scientists say they're up to 38 mutations in the gene, called BRCA1. But they also sound a note of hope. Most mutations appear to lead to a shortened version of the protein for which BRCA1 holds the code. If that trend holds for many patients, it could lead to a simple diagnostic test for the defect, says geneticist Mark H. Skolnick of Myriad Genetics and the University of Utah School of Medicine, both in Salt Lake City.

"We're hoping we can do something other than just detect specific mutations. It'll be much less expensive, more reliable, more automatable," says Skolnick. He and 41 colleagues report their findings in the Feb. 15 Journal of the American Medical Association.

Breast cancer kills some 46,000 American women each year. The cause of most breast cancers remains unexplained, but 5 to 10 percent stem from a defect in BRCA1 or another gene. Women with a mutant BRCA1 have about an 85 percent chance of developing breast cancer.

Skolnick's team pooled data on BRCA1 mutations from nine laboratories in North America and the United Kingdom. The data, some newly reported, included complete scans of BRCA1 in 372 women with breast or ovarian cancer who had a familial risk of these cancers. Three labs also looked for two mutations in the DNA of 714 other patients with these cancers, most of them with no known family history of risk.

Analysis showed 38 variations among 80 women with a mutation in BRCA1. Significantly, of the 63 patients whose defective gene was scanned completely, 86 percent appeared to carry errors coding for a shortened protein. Its role in the body remains unclear.

"It means that many people with many, many different mutations should have the same effect from that gene," Skolnick says. This might enable scientists to identify these people by screening for the snippet of BRCA1 that yields a short protein instead of poring over the entire gene, he says.

But his group still has a lot of work ahead. First, it must find out whether the truncated protein code is common to most mutations, including those not yet discovered. It also needs to learn how great a cancer risk the shortened protein poses. Even then, the test could hit technical snags, warns Francis S. Collins of the National Center for Human Genome Research in Bethesda, Md.

"It's definitely a big maybe," says one of the study's authors, Lawrence C. Brody, also of the center.