The NEJM study attempted to tease apart the effects of genes and the environment on the risk of cancer. The researchers used the gold standard for such nature-nurture studies: twins. Combing the medical histories of 89,576 identical and fraternal twins from Sweden, Denmark and Finland, they got reliable enough statistics to analyze 11 types of cancers. They then compared the “concordance” of identical twins (in how many pairs both got or escaped cancer) with the concordance in fraternal twins. If concordance is higher among identical twins with their identical genes, then genetic effects are probably more important. Cranking through the math, Paul Lichtenstein of the Karolinska Institute in Stockholm and colleagues conclude that genes account for 42 percent of the risk of prostate cancer. In colorectal cancer, 35 percent of the risk is genetic; in breast cancer, 27 percent. The heritability of leukemia, and of cancers of the stomach, lung, pancreas, ovary and bladder is even lower. “For nearly all sites, the twin of a person with cancer has only a moderate absolute risk of having cancer at the same site,” say the researchers.
The study was far from perfect, however. For one thing, the statistics are so weak that the estimates of cancer risk from genes have tremendous wiggle room built in: the genetic contribution to colorectal cancer actually ranges from 10 to 48 percent; to breast, from 4 to 41 percent; to prostate, from 29 to 50 percent. That’s “a fairly large range,” Hoover noted in a NEJM editorial. Also, the mathematical model the scientists used assumes that genes and environment do not interact. But they do. The so-called genetic component to cancer is undoubtedly modified by the environment and so may be overestimated: if you carry a gene that gives you stomach cancer only if you eat a lot of red meat, then the genetic contribution to stomach cancer falls to zero if you avoid red meat.
For a sign of how hard it will be to deflate the my-genes-did-it hypothesis, you have only to look at the conclusion many researchers leaped to on the basis of the study. From the American Cancer Society to the Human Genome Project, they are pointing out that the cancer genes we know about, like BRCA1 and BRCA2 for breast cancer, are responsible for only a fraction of the now best estimate of the genetic contribution to cancer. That means, they say, that we must find more cancer genes! But the environment–what we eat, drink, breathe and smoke; how we live and what chemicals we are exposed to–accounts for roughly twice the risk of cancer that genes do. It’s time to grow out of our love affair with all things genetic and figure out what is really killing the 1,500 Americans who will die of cancer every day this year.
