Genes cause more than one-quarter of three major types of cancer - more than previously thought, but not enough to sustain hopes that the cracking of the human genetic code could eliminate the disease, researchers say.

Scandinavian researchers concluded that genes account for 42% of the risk for prostate cancer, 35% for colorectal cancer, and 27% for breast cancer.

The rest of the cases are caused by what people do, such as smoking and diet, or what happens to them, such as on-the-job hazards, random genetic mutations or viral infections, the researchers said.

The finding, published in Thursday's New England Journal of Medicine, exceeds earlier estimates that genes account for 10% to 20% of all cancer, said Paul Lichtenstein, an epidemiologist who led the study at the Karolinska Institute in Stockholm.

Still, the conclusion runs contrary to the widespread belief that scientists ''will find solutions or cures to all diseases in the genes,'' Lichtenstein said. ''That won't be the case.''

The study is the largest of its kind yet, said Heather Spencer Feigelson, senior epidemiologist for the American Cancer Society.

Lichtenstein and colleagues studied 44,788 sets of twins in Sweden, Denmark and Finland. They looked at both identical twins, who have the same genes, and fraternal twins, who are no more closely related than any other brothers and sisters, to analyze the importance of genetics in each type of cancer.

Overall, genes probably account for 30% of various cancers, Lichtenstein estimated in an interview.

He emphasized that the figures do not mean that someone whose identical twin has prostate cancer has a 42% risk of developing it, too. The identical twin of someone with breast, colorectal or prostate cancer had an 11% to 18% risk of developing the same cancer before age 75, the researchers said.

The risk was only 3% to 9% for fraternal twins.

Those figures should go a long way toward dismissing the widespread belief ''that if your sister has cancer, you're doomed,'' said Dr. Robert N. Hoover, a director at the National Cancer Institute.

''Someone with the identical set of genes has a much greater chance of not developing cancer than of developing it,'' he said. ''I think that's a useful piece of information.''

In the Scandinavian study, only breast, prostate and colorectal cancer were common enough for full analysis. Lichtenstein said he expects enough information over the next 10 to 20 years to analyze 10 more types of cancer.
　　By The Associated Press

(US TODAY HEALTH)