The Blame Game

“Why do people get it?” — perhaps the most frequently asked question about cancer. But nowadays most of us can come up with a quick answer: “It’s mutations” — that is, damage to DNA, our genetic material. Such changes are most commonly to the sequence of DNA (alterations to individual bases or loss or gain of bits). That’s a molecular biologist’s answer. What we really want to know is “Why?” How do these changes come about and, of course, what can we do about them?

The things we do …

We’ve known part of the answer for a long time — it’s what we do to ourselves, stupid! The best known example is smoking, shown to cause lung cancer in the 1930s. We know now that chemicals in cigarette smoke damage DNA and they’re so good at it that over 90% of lung cancer is down to smoking. So pernicious is the habit that it killed about 100 million people in the twentieth century and, unless something pretty drastic happens, the number of deaths this century will be one billion.

It’s true, we have made some progress. Most countries now regulate smoking. Bhutan (bless ’em) was the first nation to outlaw smoking in all public places. In the UK tobacco advertising is banned, as is smoking in all work places. But none of this happened until we’d got well into the twenty-first century! Oh, and if you’re wondering how the country that likes to style itself the world leader is getting on, Congress has so far managed to avoid passing any nationwide smoking ban and left it to individual states — with the result that across the USA laws range from total bans to no regulation of smoking at all! All told, the saga has been one of the more staggering examples of political impotence.

I’m sure you can think of other daft things we do to propel us to our cancer graves but we need to move on by noting that, aside from what we do to ourselves, the world we live in contributes something of a helping hand. Thus some useful foods nevertheless contain harmful substances and even the ground we stand on gives off low levels of radiation. And there’s not a lot we can do about such things.

… And are done to us

Then there’s heredity — the state of our DNA when we get it. It’s been clear for some time that mutations passed to us at birth kick off about 10% of all cancers (see for example, A Taxing Inheritance).

Way back in 1866 Paul Broca suggested it might be possible to inherit breast cancer. He’d looked at his wife’s family tree and noted that ten out of twenty-four women, spread over four generations, had died from that disease and that there had been cases of other types of cancer in the family as well. This large proportion was not, he believed, mere chance. Now we know that a changed (mutated) form of a gene (a unit of heredity), passed from generation to generation, was almost certainly responsible for the suffering of this family.

So broadly speaking there are two long-recognized categories that cause cancer —‘environmental’ and ‘hereditary’ — and, although we cheat by lumping things that we can control (e.g., smoking, eating too much red meat and sunbathing) into the ‘environmental’ camp (there should be a separate group: ‘stupidity’), many factors really are beyond our control.

As ever, it’s worse than that

Lurking in the wings for many years now has been a potential third cause that arises from a slightly tricky concept — namely the fact that our DNA, the genetic rock upon which all life is built, isn’t rock-like at all. In fact the chemistry of DNA makes it inherently unstable. Thinking about it from the viewpoint of evolution, of course it’s unstable: it has to be to permit change as new genes, and hence new proteins, are made and unmade — allowing life forms to advance. Think of it like close relationships: we’re fond of calling such things ‘permanent’, ‘unchanging’, ‘solid as a rock’ even. But they’re not: they change all the time, adapting to our shortcomings and to how individuals develop and mature.

With that in mind maybe it’s less surprising to find that DNA reacts with a wide range of chemicals, some that we consume but others arising from the natural reactions of the body — products of metabolism in fact. And then, speaking of shortcomings, there’s the truism that ‘nobody’s perfect’ and the realization that this applies to the mechanics of DNA replication as well as everything else. In others words, every time we make a new cell its DNA differs from the original. Cells have remarkably smart methods for correcting most mistakes made during replication but, inevitably, some get through and become fixed in the new genome.

Although ‘replicative mutations’ have been known for a while, nobody had come up with a way of measuring how much they contribute to cancers. Step forward Bert Vogelstein and Cristian Tomasetti at Johns Hopkins University with the idea of looking at ‘stem cells’ — cells that can divide to make more of themselves or to turn themselves into specialized cell types. They reasoned, bearing in mind that with every division there’s a risk of a cancer-causing mutation in a daughter cell, that if you knew the number of stem cells in an organ and you could estimate the total number of divisions over a lifetime, that might relate to cancer risk.

Indeed it did. In spades, because it turned out to account for two-thirds of all cancers. In other words, the majority of cancers arise because of cumulative mutations caused by internal agents.

Quick test to see if that fits with something we know: cancers of the intestine. Cancers of the duodenum (the first section of the small intestine) are rare compared with those of the colon (the large intestine, into which the duodenum empties). For 2017 in the USA the estimates are 1390 and 50,260 deaths, respectively – that’s about 0.2% of all cancer deaths versus 8% for colon cancers. Sure enough, Tomasetti and Vogelstein estimated the cell division rate to be about 100 times greater in the colon over a lifetime.

Mice, somewhat curiously, are the other way round: they pick up more cell divisions in their small intestine — and more cancers — than in their colon.

This correlation of rising risk with increasing number of divisions held over 31 different types of cancer — as it did when extended from USA to world-wide data, thereby eliminating a bias from environmental factors.

The upshot of this is that to ‘environmental’ and ‘hereditary’ factors we need to add a third category of ‘replication errors.’

Real-life examples of the impact of replicative mutations on lung cancer and prostate cancer.

About 90% of lung cancers are preventable: heredity plays no significant role and the estimate is that 35% of all driver (i.e. cancer-promoting) mutations are due to replication errors. For prostate cancers there is no evidence that environmental factors are significant and hereditary factors account for 5 to 9% of cases. The remaining 95% of driver gene mutations are estimated to be replication errors. None of these cancers are preventable. Clouds represent contributions from environmental factors. Gray dots: Environmental mutations, Yellow dots: Replicative mutations, Blue dots + H: Hereditary mutations (from Tomasetti et al. 2017).

 Causes of driver mutations in 18 types of female cancer (UK).

The colour codes are the same for hereditary (left), replicative (centre), and environmental (right) factors and are from white (0%) to brightest red (100%).

The left-hand schematic indicates that inherited mutations are not statistically significant in these cancers (note that Paul Broca’s findings related to fewer than 10% of breast cancers – the proportion we now know to be caused by abnormal genes passed from parent to child).

B, brain; Bl, bladder; Br, breast; C, cervical; CR, colorectal; E, esophagus; HN, head and neck; K, kidney; Li, liver; Lk, leukemia; Lu, lung; M, melanoma; NHL, non-Hodgkin lymphoma; O, ovarian; P, pancreas; S, stomach; Th, thyroid; U, uterus (from Tomasetti et al. 2017).

Controversial or what …?

It’s fair to say that the estimate of two-thirds of all cancers being down to internal faults was a surprise to many.

It has to be said that there’s a continuing debate about the precise numbers — not least because figures for cell divisions in some tissues aren’t available and also because of somewhat vaguer problems, e.g., to what extent to external assaults contribute to replication errors.

Nevertheless, it now seems clear that what Tomasetti and Vogelstein call “bad luck” can be blamed for a significant number of cancers. That’s good because knowing that it’s not your ‘fault’ may help some patients but we need to be wary of promoting that message too strongly in the media.

The fact is that, whatever the proportion might be that we can put down to “bad luck”, there are still a great many cancers that can be prevented.

What’s to be done?

Now we know what to blame we can return to the question of what can be done. It won’t take long because at the moment the answer is ‘not much’. The accumulation of mistakes from replication errors is random, so we cannot predict who will find themselves with a critical (i.e. cancer-producing) set. But that scarcely matters as we have no way of preventing them happening. So all we can do at the moment is deal with what presents itself with the treatments currently available, comforting ourselves that in the long-term things like gene–editing might enable us to rectify critical replication mutations.

So, like a lot of fascinating advances in the cancer field, the take-home message here is “that’s all very interesting but in the meantime we need to keep focusing on the possible: the fact that if we stopped smoking, got people to eat sensibly and gave everyone decent sanitation we could cut cancers by half”. Give or take a few percent!


Tomasetti, C. et al., (2017). Stem cell divisions, somatic mutations, cancer etiology, and cancer prevention. Science 355, 1330–1334.

Tomasetti, C. and Vogelstein, B. (2015). Variation in cancer risk among tissues can be explained by the number of stem cell divisions. Science 347, 78-81.