How many scientists does it take to find a gene? If it’s a gene as important as one that raises your risk of lung cancer, at least 142—the number of researchers who conducted three separate studies being published today which all identify the same stretch of chromosome 15 as containing genes linked to lung cancer.
How many scientists does it take to figure out what the genes do and how important they are? More than 142, apparently, for the three groups have reached dramatically different conclusions about exactly how the genes function to make you more likely to develop lung cancer. Two conclude that the genes directly raise your risk of developing lung cancer whether or not you smoke, while one concludes that the genes make you more likely to smoke and only through that effect on behavior raise your risk of lung cancer.
The discrepancy gets to the heart of a longstanding debate about the value of genetic discoveries. Whenever scientists discover a gene “for” some disease or normal trait, it leads to the question of whether the gene causes that disease or trait directly, or is instead a bystander. (The classic here is the hypothetical chopsticks gene. Some genes are more common in Chinese people than in westerners, and Chinese use chopsticks more than Europeans do, but it would be a laughable mistake to conclude that the genes cause people to use chopsticks: the genes are merely markers for certain ethnic groups which, because of culture alone and not genetics, use chopsticks.) Now that debate has erupted in the clash of lung-cancer papers.
All three of the papers agree that the genes that increase the risk of lung cancer lie on chromosome 15, one of 23 (each part of a pair) in every human cell. The technique that each team used, called a genome-wide association study, cannot pinpoint which gene or genes on the chromosome is the culprit; all it can do is say the suspect gene lies in a certain region of the chromosome. Earlier studies have shown that among the five genes in the suspect region of chromosome 15 are three for parts of what is called the nicotinic acetylcholine receptor. All three groups agree that these nicotine-receptor genes are almost certainly the genes responsible for the elevated cancer risk. That is striking: it suggests that nicotine, long known to be what makes cigarettes addictive, might also cause lung cancer. And finally, all three groups agree that the dangerous gene variant is quite common: about half of people of European descent carry at least one copy, having inherited it from mom but not dad (or vice versa). About 10 percent to 15 percent carry two copies, inheriting it from both parents.
It is no surprise that there is a genetic component to lung cancer and smoking. Smoking causes an estimated 5 million deaths a year, but while many of those deaths are from lung cancer, not all of them are: only about 15 percent of smokers develop lung cancer, of which there are 1 million new cases worldwide every year. (In the United States, there are about 160,000 deaths from lung cancer every year.) The fact that relatively few smokers develop lung cancer strongly suggests that there is something inherently different between smokers who do and those who do not.
Two of the new papers conclude that the now-suspect nicotinic acetylcholine receptor gene is a large part of that difference. As its name suggests, this receptor acts as a docking port for nicotine molecules. In a study led by Paul Brennan of the International Agency for Research on Cancer and published today in Nature, the scientists conclude that the gene raises the risk of lung cancer directly: people who have this gene (to be precise, everyone has the nicotinic acetylcholine receptor gene; what matters is which variant, or chemical spelling, you inherited from mom and dad, since only some spellings raise the risk of lung cancer) have a greater risk of the disease even if they do not smoke.
So, too, concludes the study by Christopher Amos of the M.D. Anderson Cancer Center, published online in Nature Genetics. If you have ever smoked and have one or two copies of the gene variants, your risk of developing lung cancer is about 30 percent to 80 percent greater than if you do not have these variants, the researchers found. Putting actual numbers on that, the risk rises from about 15 percent to about 27 percent at the high end.
The extra risk of lung cancer conferred by these gene variants is unrelated to whether or how much you smoke, both the Amos and Brennan teams conclude. “What we do not understand is why some long-term smokers develop lung cancer and others don’t,” Amos said. “There are so many different cancer-causing compounds in tobacco smoke that it’s hard to separate them and we don’t fully understand the mechanisms that cause lung cancer.”
The third group reaches a strikingly different conclusion: that the genes affect how much you smoke, and only through that mechanism do they raise your risk of lung cancer—more cigarettes, more cancer. Each copy of the gene variant increases the risk of developing lung cancer 30 percent, they estimate. “This is a genetic predisposition to seek out a certain environment, to smoke,” said Kari Stefansson, CEO of the genomics company deCode who led the latter study, during a recent visit to Newsweek. “We know that complex diseases have both genetic and environmental components, and the challenge has always been to figure out their relative contributions. But if this gene causes you to seek out a ‘smoking environment,’ then it’s as if both gene and environment contribute 100 percent to the development of lung cancer.”
Scientists not connected to the new studies are scratching their heads over how the teams could come down on opposite sides of the question of whether the gene variants raise the risk of lung cancer directly (and therefore make you more likely to develop it whether or not you smoke) or indirectly (making it more likely that you will smoke). In the latter case, if you have these “lung cancer” genes and do not smoke, you’re fine—no extra risk of lung cancer. In the former, even non-smokers have cause for concern if they carry these gene variants.
“The extent of the disagreement in the conclusions of these studies is remarkable,” write Stephen Chanock of the National Cancer Institute and David Hunter of the Harvard School of Public Health in a commentary on the trio of studies. “They signal the need for greater methodological rigour in attempts to account for both the genetic and the environmental causes that we think underlie most diseases.”
Until that disagreement is resolved, don't rush to get tested for the nicotine-receptor genes. For one thing, even if you turn out to have the "good" gene variant and therefore to be resistant to developing lung cancer from smoking, you can't smoke with impunity: you’re unlikely to also be protected against emphysema, heart disease and other often-fatal consequences of lighting up. “I try to make a living by marketing genetic tests,” said Stefansson in a telephone briefing, “but I am not a proponent of this one.”
At least not yet. But if Stefansson’s interpretation turns out to be right, and the gene variants increase your risk of smoking but not of lung cancer otherwise, it will be one more piece of evidence that genes are not destiny: all you have to do to keep your risk of lung cancer near zero is to resist the urge to smoke. But you knew that already.