While there is a component of bad luck in most car accidents — “What if I had only left home 30 seconds later?” — attributing accidents to bad luck is no way to make progress in injury prevention. Such is the case, too, with cancer prevention — despite a recent article published in the prestigious journal Science that led to a spate of headlines, such as “Most cancer types — just bad luck.”
The vast majority of research does not show that cancer is the result of bad luck. Rather it shows that most cancers are theoretically preventable. Yet when the Science article in question: “Variation in cancer risk among tissues can be explained by the number of stem cell divisions,” was promoted with a press release entitled “Bad Luck of Random Mutations Plays Predominant Role in Cancer, Study Shows,” journalists can hardly be blamed for the resulting headlines.
We use the word “cause” with circumspection in the fields of medicine and public health. Even when a pedestrian dies after being hit by a car, there may be other factors — weather, road conditions, alcohol use by the pedestrian or driver — that contributed to his or her death.
Indeed, the authors of the stem cell study, trying to clarify the meaning of their first press release, followed up with an addendum in which they attempted to explain the initial statement with an automotive analogy. In thinking about the risk of a car accident, they compared the vehicle condition to inherited genetic factors, the road conditions to environmental factors, and the length of the trip to the number of stem cell divisions and random mutations a tissue undergoes. With this analogy, they attributed two-thirds of the risk of an accident to the length of the trip.
Unfortunately, however, the car analogy fails in this context. Most car accidents happen within 25 miles of home — a single longer trip doesn’t necessarily entail higher risk than a number of short trips. And no one would argue that the chief cause of car accidents is short trips.
The fundamental issue with communicating about causation of cancer is that we can generalize for cancer risk in populations, but that any specific individual’s cancer may or may not be due to any one or more specific causes.
For example, we know that lung cancer became more than 10 times more common as tobacco use increased, that countries with low smoking rates have low cancer rates, and that smokers have more than a 10 times higher risk of lung cancer — so we say smoking causes lung cancer. Nonetheless, we also know that most smokers do not die of lung cancer, and even a heavy smoker may have had their lung cancer caused by radon or one of the less common causes of lung cancer. What is unambiguous is that nearly all lung cancer would be prevented if nobody smoked.
Similarly, almost all cervical cancer in the world would be prevented if all girls received the human papillomavirus vaccine at young age. Most liver cancer would be prevented if everyone received the hepatitis B virus vaccine early in life, and those with hepatitis C virus received the new drugs that eliminate the infection. Most women infected with HPV do not develop cervical cancer, and most of those infected with HBV and HBC do not develop liver cancer, so one could technically say there was some element of chance for any individual patient.
Still, the large increases in cancer risk that occur in populations exposed to a risk factor are what determine the amount of cancers that are preventable. The frame of reference that is important in this debate is the population perspective — not the individual case. And the route to prevention is by focusing on populations, not individuals.
The scientists who came up with the “bad luck” press release subsequently stated that many had “found relief” in their statements due to the history of stigmatization and “burden of guilt” among patients who “frequently blame themselves” for their cancer. Stigmatization and guilt are not the aim of those who try and prevent — or treat — cancers. Rather, getting accurate information to populations is the first step in cancer prevention, followed by trying to ensure that we live in a society that promotes good health and healthy choices.
To return to the motor-vehicle accident analogy, changing the length of trips is not of much help in preventing accidents, but improving environmental conditions such as road quality and lighting, as well as reducing individual behaviors such as drunk or distracted driving, are all important to injury prevention. Similarly, the real question is whether we are going to take cancer prevention seriously by ensuring that populations in the United States and around the world have accurate information, are vaccinated against or treated for cancer-causing infections, and are less likely to engage in cancer-risk behaviors.
David J. Hunter is the Vincent L. Gregory Professor of cancer prevention and dean for academic affairs at the Harvard T.H. Chan School of Public Health.