For the John Templeton Foundation, I wrote about the intriguing role of random mutations in cancer development.
Few people make it through life without being directly or indirectly touched by cancer. The disease is the second-leading cause of death globally, causing about one in six deaths. Over the past century researchers have identified many behavioral, environmental, and genetic risk factors that increase the likelihood an individual will get certain kinds of cancer, but an element of randomness remains when it comes to individual outcomes: some lifelong smokers never get lung cancer, and some people can come down with cancers even in the absence of known risks. For many researchers, areas of unpredictability in cancer have driven the search for unknown environmental or genetic factors. But a few years ago, Cristian Tomasetti, a professor of oncology at Johns Hopkins, began to wonder whether an underappreciated cause of cancer happens even without the influence of carcinogens or cancer-causing genes: the infinitesimal and inevitable random mutations that occur during normal cell division.
In 2015, Tomasetti and his Johns Hopkins colleague Bert Vogelstein published a seminal paper in Science that argued that a major driver of variation in cancer risk among different tissues could be explained by the number of cell divisions within that tissue over an individual’s lifetime — the higher the number of cell divisions, the greater the number of occasionally harmful random errors that occur and accumulate. Studying dozens of cancers, they showed a correlation across five orders of magnitude of cancer prevalence showing that the higher the lifetime number of cell divisions in a tissue the higher its cancer rate. At one end of the plot were rare cancers of slow-growing bone tissue; at the other were common cancers of the colon, which replaces the cells of its entire lining roughly every four days. Cancers occurring that way were, Tomasetti and Vogelstein said, significantly attributable to “bad luck” in the form of naturally occurring mutations, rather than to behavior or inheritance