Two years ago, a mathematician and cancer geneticist from Johns Hopkins kicked up a storm of controversy when they published a paper in the journal Science suggesting that most cancers are caused by random, unavoidable mutations—basically nothing more than bad luck. They were widely criticized for limiting their analysis to the U.S. and leaving out two of the most common and deadly tumor types, breast and prostate cancer.
Well now those scientists have expanded their study to include those cancers, and to encompass patients worldwide, and they’ve come to the same stark conclusion. About 66% of cancer-causing mutations are random “copying errors” that occur when cells go through their natural dividing process. Therefore they can’t be avoided, report Cristian Tomasetti, assistant professor of biostatistics at the Johns Hopkins Kimmel Cancer Center and the Johns Hopkins Bloomberg School of Public Health and Bert Vogelstein, co-director of the Ludwig Center at the Johns Hopkins Kimmel Cancer Center. Their latest research was published today in Science.
According to Tomasetti’s and Vogelstein’s calculations, just 29% of cancers can be blamed on environmental factors such as smoking and sun exposure. The remaining 5% of tumors are caused by inherited genetic mutations. They dub the cancer culprits in their study “E” for environmental, “H” for hereditary and “R” for replication errors.
The Hopkins researchers hope their study will bring fresh attention to the need to develop better strategies for early detection—a field of research that Vogelstein says is sorely underfunded. “With so many of the mutations caused by R, the internal enemies, one of the best strategies is to nip their attacks in the bud,” Vogelstein said during a press conference announcing the results. “We detect those attacks when they’re still early, before the cancers have widely spread, at a time when they can still be cured simply by surgery or perhaps surgery and adjuvant therapy, which always works better on smaller tumors.”
Before considering the prospects for early detection, it’s worth taking a look at some of the statistics Tomasetti and Vogelstein came up with when they expanded the scope of their analysis. For this study, they collected data from 423 cancer registries in 68 countries, amounting to 4.8 billion people and 32 cancer types. Among 17 of those cancers, there was a strong correlation between cancer incidence and the number of times normal cells divide—validating, they say, their original hypothesis that the more cells divide, the more likely they are to make mistakes in DNA that can cause cancer.
In the case of prostate, bone and brain cancer, for example, they found that 95% of cases are caused by mutations emanating from random copying errors. Most childhood cancers are also caused by R, Tomasetti and Vogelstein said. “None of these cases are considered preventable,” Tomasetti said.
This image depicts the mutations attributable to environmental factors (right), DNA replication... [+]
“We hope that this research offers comfort to the literally millions of patients who have developed cancer but who have led near-perfect lifestyles, who haven’t smoked, who avoided the sun without sunscreen, who eat perfectly healthy diets, who exercised regularly,” Vogelstein said, adding that he hopes it will also be reassuring to parents of children with cancer. “They need to understand that these cancers would have occurred no matter what they did.”
Vogelstein believes the new findings show that preventing cancer altogether may not be possible for many tumor types—but preventing cancer deaths should be. Well-established technologies such as mammography are still widely recommended, of course, but Vogelstein is calling for fresh ideas in early detection that could help many more patients. He cited ongoing research involving detecting cancer-causing microRNAs or proteins, as well as new imaging technologies that promise to not only detect tiny tumors, but also determine where in the body they might be hiding. “The problem we see is there’s so little work done on any of those modalities for early detection,” he said. “We’re still stuck with the same [technologies] that were used 50 years ago. We hope this will change the tide.”
There is one well-funded company that hopes to change the outlook for early detection: Grail, a San Francisco company that expects to raise $1 billion in venture capital this year, after getting $100 million in startup capital from the likes of Bill Gates, Jeff Bezos’ Bezos Expeditions and gene-sequencing pioneer
There will be plenty of challenges for companies developing new early-detection technologies, Vogelstein said during the press conference, not the least of which will be convincing the FDA and other regulatory bodies that they work with a high degree of accuracy.
But the first step toward changing the tide of cancer research, so that more resources are focused on early detection, is accepting the harsh reality brought to light by this latest research: For many patients, cancer really is just bad luck. During the press conference, Tomasetti and Vogelstein were asked if random cancer-causing mutations are merely the price everyone pays for evolution. “Yes,” answered Tomasetti. Echoed Vogelstein, “We couldn’t have said it better ourselves.”
