Skip to main content

Together we are beating cancer

Donate now
  • Science & Technology
  • Health & Medicine

Reports that cancer is ‘mainly bad luck’ make a complicated story a bit too simple

by Michael Walsh | Analysis

24 March 2017

0 comments 0 comments

A little over 2 years ago a paper was published that sparked headlines claiming cancer was ‘mainly down to bad luck’.

As we and others explained at the time, this is a complex topic to put numbers on. And the message that accompanied the media coverage – that there’s nothing we can do to reduce our risk of cancer – didn’t tally with evidence showing more than 4 in 10 cancers could be prevented, largely through lifestyle changes.

Now, the same research team behind the previous study has taken their analysis further, publishing their findings in the high-profile journal Science. And as the first paper caused quite a stir, it’s no surprise that the results have been picked up in the media again.

The researchers, based at the Johns Hopkins Kimmel Cancer Center in Baltimore in the US, have addressed some of the original concerns around their approach, boosting sample numbers and adding data from more countries and cancer types.

And it seems to back up their previous conclusions, adding further evidence to their mathematical theory that cancer risk is most strongly associated with how quickly specialised cells, called stem cells, replicate.

As a result, they predict that mutations in 2 in 3 cancers are ‘due to bad luck’, which they define as the random genetic mistakes that occur naturally in cells as they divide, and may lead to cancer.

But could this message do more to confuse than inform?

Bad luck 2.0

The researchers looked at data from 32 types of cancer in patients from 69 countries. And they tried to link cancer-causing genetic faults (mutations) to different causes: inheritance, lifestyle/environment, and (the one they focus on) random errors that occur when cells divide.

“As a mathematical exercise to study mutations in cancer, the new paper is well done and interesting,” says Professor Richard Gilbertson, director of our centre in Cambridge. “But we have to remember that this study uses existing data to generate a correlation and hypothesis – it’s not a definitive answer to a very complex problem.”

This means they had to look at information that’s already out there to reach their conclusions. In other words, maths can only tell us so much about the real-life balance and biology of what causes cancer.

But one of the biggest problems, Gilbertson says, is that trying to get a clear answer of cause and effect from the analysis might oversimplify things.

“They’ve taken an approach focused too heavily on mutations, and while mutations are very important for cancer, the disease can’t be boiled down solely to specific mutations,” he says.

For cancer to develop, there are many other real-world factors at play, and it’s critical to understand these processes together and how they affect how cancer progresses to effectively prevent cases.

Mutations that can lead to cancer can be caused during a person’s lifetime by environmental and lifestyle factors, such as cigarette smoke. They can also be inherited, or they can happen by chance.

These chance mutations occur because errors can be made when a cell copies and divides its DNA, because the replication process isn’t perfect.

The researchers say that this third factor is a major player in cancer cases, and in their model it accounts for around 2 in 3 mutations in cancer. This, they say, is the ‘bad luck’ factor, something they want to be acknowledged more as a cause of cancer.

As they find such a large proportion of mutations being down to chance error, the researchers argue this makes a case for more to be done to use this information to try and diagnose cancers earlier.

It’s not exactly clear what this might mean in practice, but we’d agree that understanding more about the genetic faults fuelling tumours could lead to ways to detect them earlier. And spotting a cancer at an earlier stage means that treatment is more likely to be effective.

But this is just one part of the puzzle, and some would argue that it’s the balance of all these factors that needs to be taken into account.

Much like the work of Vogelstein, Gilbertson’s own research has focused on stem cells, which replenish tissues with new cells as others become old or damaged. These cells are thought to have an important role in how cancers start.

Cancer development does likely require mutated stem cells that are dividing,” says Gilbertson. “However, we know from our previous work that mutations can accumulate in stem cells that have the potential to divide but have gone to ‘sleep’.”

This means that there may be cells with stem cell potential capable of sitting in the body with mutations causing absolutely no problems if they’re not dividing.

In this situation, something extra is needed for a cancer to form. For example, tissue damage can make the cells divide and then this can lead to cancer.

This layering of events that push cells towards becoming cancerous suggests there’s an important interplay here that this latest study may overlook by focusing solely on chance mutations.

Mutations are one thing, but cancer is another

The chance of mutations happening is dramatically increased by certain environmental factors, thereby increasing the risk of cancer. Such factors also often cause a lot of cell death, meaning cells in the damaged tissue have to multiply to replace the damaged or dead cells, which can further increase the risk of cancer.

The latest study also analysed cancers on an individual basis, whereas it’s known that many environmental factors increase the risk of multiple types of cancer.

The researchers’ calculations didn’t seem to factor this in, which could lead to an underestimation of the contribution of environmental factors – something they point out themselves in their analysis.

And while there’s nothing we can do about random mistakes happening during cell division – this is an inherent part of being alive – we can do something about environmental factors in many cases.

Even if, as this study suggests, most individual cancer mutations are due to random chance, the researchers admit that the cancers they cause may still be preventable

Professor Mel Greaves from The Institute of Cancer Research, London

Others have agreed that the complicated nature of cancer makes it tricky to reduce multiple cases to a single cause. “Any model attempting to fully describe cancer risk needs to take into account a variety of extrinsic factors that influence cancer development and progression without inducing gene mutations, such as inflammation, immune responses and the tumour microenvironment,” says Professor Lawrence Young, director of the University of Warwick’s Cancer Research Centre.

Cancer is complicated and there’s no way of knowing what caused an individual person’s cancer to develop. The researchers have said that they hope that turning the spotlight on to chance genetic faults might offer comfort to “patients who didn’t smoke and who avoided the sun and ate healthily, done everything possible to avoid cancer but still got it.”

“It’s particularly important for parents of children with cancer, who may look on the web and see that cancer is caused by environment and heredity,” said Professor Bert Vogelstein, one of the study’s authors. “That says to parents, either I transmitted a bad gene to you or I exposed you to some environmental agent which caused your cancer. That causes an amazing amount of guilt.”

We absolutely agree that nobody should feel blamed for their cancer. And there are no lifestyle factors linked to cancer in children, so that should be reassuring for parents too.

But boiling this down to cancer being ‘mostly bad luck’ runs the risk of people feeling there’s nothing they can do to stack the odds in their favour.

There’s a huge body of evidence showing that more than 4 in 10 cancers could be prevented by a healthy lifestyle and if we were better able to limit exposure to environmental factors.

This latest update acknowledges the role that environmental factors and the ways we live our lives play in our risk of cancer, which is really important.

“Even if, as this study suggests, most individual cancer mutations are due to random chance, the researchers admit that the cancers they cause may still be preventable,” says Professor Mel Greaves from The Institute of Cancer Research, London. “We have good evidence to show that cancer is caused by a complex mix of environmental exposures, inherited risk, and random chance.”

In search of the complete picture

Gilbertson suggests an analogy to summarise the challenge in simplifying this complex situation: “Imagine the disordered process of mutations and dividing cancerous stem cells is like a vase falling off a shelf.

“There’s no denying that the vase smashed and was destroyed because of the fall – the cancer process was driven by the presence of mutations in dividing stem cells. But if you wind the process back you might see someone accidently pushing the vase off the shelf. This push is like the other factors contributing to cancer – like tissue damage and repair.

“So how would you want to prevent this in future? You might try to catch the vase – ‘correct’ or reverse the effects of mutations that have already happened, but it would also be very important to prevent the push.”

This means that while the latest results offer a solid prediction on one of cancer’s origins, the approach underplays other vital parts of the picture.

With more research from each of these groups of scientists that complete picture will continue to emerge.

And it’s this that will benefit those looking to reduce their risk of cancer, and attempts to diagnose cases earlier, the most.



Tomasetti, C., Li, L., & Vogelstein, B. (2017). Stem cell divisions, somatic mutations, cancer etiology, and cancer prevention. Science, 355 (6331), 1330-1334 DOI: 10.1126/science.aaf9011


This post was updated 10:00 29/03/2017 to clarify the distinction between mutations in cancer and cancer cases, following comments made here