¿ìè¶ÌÊÓƵ

Competing narratives of chance and lifestyle often frame the debate about cancer’s underlying causes.

This is based on wildly varying estimates of what proportions of the genetic mutations that characterise the disease are due to internal and external factors. These are frequently misinterpreted as evidence that cancer is either largely down to bad luck or almost entirely preventable.

At one end of the spectrum, this can lead patients on a futile path of reflection, trying to pinpoint a specific event or behaviour to blame for their disease. At the other, blaming bad luck risks diluting public health messages that modifiable lifestyle factors are important. The influence of diet, smoking, sun exposure and alcohol consumption are all well-established.

The debate hinges on the fact that cancer-causing mutations in our DNA can occur three ways: they can be inherited, produced by random errors when our cells replicate, or by environmental exposures. There is intense interest in understanding the relative contribution of all three.

Which is why . It attempts to resolve the conflicting evidence by combining genome-wide sequencing and population-wide disease data from 69 countries.

This is a follow up to , which caused a stir. That study suggested that the number of stem cell divisions, and hence the number of DNA replication errors, in different types of tissue was the main factor driving variations in how often cancer occurred in those tissues. Unfortunately it was ripe for misinterpretation, the most common being that .

Soon after, there was a backlash. In 2016, a New York team claimed that the stem cell model used could not distinguish between internal and external drivers of mutation and in driving the characteristic DNA changes in tumours.

Today’s follow up by the 2015 team addresses that, by showing that the correlation between stem cell divisions and cancer risk in different tissues holds across countries with very different environmental conditions. The authors also use a mathematical model that does not rely on stem cell divisions to calculate the relative contribution of hereditary, environmental or replicative factors in each cancer type.

For example, the authors cite the fact that nearly 90 per cent of the most common form of lung cancer is preventable, even if they find random stem cell copying errors account for 35 per cent of DNA changes in this tissue.

In this way, the new study resolves an apparent conflict between biological mechanisms and population-level estimates of how much cancer can be prevented by environmental differences. For example, Cancer Research UK puts the figure for preventable cases at 42 per cent, which the study authors say fits their work.

Where does all this leave the competing messages about cancer’s cause? Unsurprisingly, the answer lies somewhere in the middle.

The debate reinforces a cautionary tale in trying to translate population-wide data to causes in individual cases. It is almost never possible to say with any certainty what caused an individual tumour.

It also crystallises what for many is a difficult message to hear. We can still reduce our cancer risk through modifying lifestyle, but chance still plays a significant role.