Genes don’t explain why cancer spreads. Bring the body back into the picture

By Jerome Burne

Why are some cases of cancer cleared from the body after surgery while others lurk, waiting to metastasise and emerge elsewhere months or even years later? There is no good answer at the moment because the conditions that allow these invaders to flourish don’t fit with our gene-focused theory of how and why cancer develops in the first place.

To answer the question oncologists need to look up from their microscopes and gene sequencing machines, says a top cancer specialist, and investigate the systems that naturally protect against cancer developing in the patients’ bodies. Cancer cells are not oncogene packages containing all that’s needed to spread but invaders whose success or failure is heavily influenced by the body’s own defences.

The reason Japanese knotweed, a modest Jekyll-like plant at home, transforms into a rampaging Hyde in English gardens won’t be found in its genes; they don’t have a mutation that gives their roots super-powers. The explanation is that English gardens lack some of the features that keep knotweed well-behaved a home. The same goes for rabbits in Australia or beavers in Tierra del Fuego.

The radical (in cancer circles )notion that an invader – such as a metastasising cancer cell – depends more on the conditions surrounding it for success than any mutations in its DNA, is elegantly and compellingly set out in a long article just published in the New Yorker. That makes it worth reading in itself but the fact that it is written by celebrated oncologist and author Siddhartha Mukherjee, who wrote a history of cancer called The Emperor of All Maladies, published to rave reviews in 2011, means it is likely to attract serious attention.

Cancer research fails to see wood for trees

To pinch a term from political philosophy, it might herald the opening of a medical Overton Window (a useful phrase I’ve just discovered), that describes a time when ideas that weren’t part of normal discussion suddenly become acceptable. Before Corbyn’s election no sensible politicians talked about renationalisation or cutting income inequality; now it is the stuff of the opinion pages.

Similarly, questioning the idea that gene mutation is not the driving force of cancer is rarely discussed in polite oncological society; not with the 300 million pound 1000 genomes project promising to provide the tools to beat cancer. Of course mutations are involved but there is more to beating cancer than digging ever deeper into the genome. What’s needed is greater focus on the wood and less on the trees.

Mukherjee isn’t the first to point this out. Nearly two years ago a book called ‘Tripping over the Truth’ told the story of an idea that originated in the 1920s – the Warburg hypothesis. This proposed that what distinguished cancer cells from healthy ones was the way they generated energy. They had switched from using oxygen in the mitochondria (miniature power plants found in every cell) to relying on greatly increased supplies of glucose. For more details and how this provides a rationale for the ultra low-carb ketone diet, see here

Another pioneer of the ‘wood not trees’ approach is Professor Mina Bissell a leading authority on breast cancer at the University of California. The analogy she uses is good and bad neighbourhoods. (The technical term for her model is the “extracellular matrix” (ECM)).  Cells are far less likely to turn cancerous when they are in a healthy ECM. Then they nestle tightly against each other like terraced houses and any hint of leakage from a cell or abnormal behaviour is quickly dealt with by the neighbours. For more details see here.

Chemo poor at preventing cancer spreading

Mukherjee arrives at the idea that oncologists need to take a wider angled view by considering the mysteries thrown up by the way cancer metastasises. Predicting whether a patient’s cancer is likely to spread  is currently very hit and miss, as is advice on how to  reduce the risk because oncologists have no good model of how it works.

Preventing or reducing it would cut the death rate and the number of patients suffering pointless side effect  Removing a tumour is the easy part – what patients really want to know is: will it come back? The phrase clinicians use after successful surgery – ‘no evidence of disease’ – is obviously encouraging but it doesn’t tell you if that’s because methods for detecting it aren’t good enough or because it has actually gone.

This uncertainty means patients are routinely offered treatments such as chemotherapy or hormone blockers to cut the risk of a return. A small fraction benefit but there is no way to tell who they are in advance. The rest just suffer the various unpleasant and damaging side effects without the benefit.

Mukherjee tells how the puzzling nature of metastasis was first spotted in 1889 when a pioneering English doctor, Stephen Paget, investigated the common assumption that cancers behave like a stain; spreading out randomly from a central point. He looked at case files of 735 women and found that cancer spread was far from random. Certain organs were far more likely to be become new sites than others – the liver being the most popular by a long way with lung and spleen trailing behind. Bones were also common sites but only certain ones; almost never hands and feet.

Ask not what cancer is doing to you but what you can do to cancer.

But why? Were these areas more hospitable to cancer? The question was left on the table until researchers the late 1980′s when researchers began looking for mutations linked to the likelihood of a cancerous cell setting off on the metastasising journey and establishing itself in a new niche. As Mukherjee explains, if genes are the driver of cancer they must drive metastasis too.

But no individual genes or pattern of genes showed up as being vital. Then in 2001 an even greater problem for the genes-only theory emerged. It looked as if the ‘wood’ or the type of neighbourhood was very much involved in preventing or allowing a tumour to spread.

Animal work published in 2001 had found that tumours, like some creature on a coral reef, were constantly pumping clouds of cancer cells into the blood – around 3 million cells for every gram of tumour every 24 hours. If only genes controlled successful metastasis we should all be riddled with cancer in every organ. Something in the body was controlling it.

‘Ask not what cancer is doing to you,’ writes Mukherjee, ‘but what you are doing to cancer.’ It’s hardly a big surprise that the immune system emerges as a major player. He tells the story of a patient whose melanoma had been successfully treated with surgery and was declared cancer free. Some years later, this patient donated a kidney and patient receiving it was put on immune suppressing drugs to prevent rejection. A few weeks later the recipient’s new kidney was covered with hundreds of tiny black specks of melanoma.

Not everyone with flu virus gets flu

The donor’s immune system had rendered the cancer cells inactive; blocking the immune response in the recipient allowed them to come to life. Subsequent animal research has found that dormant cancer cells could also spring into action following the inflammation that comes with an infection such as pneumonia. Further research has shown that it was the NK – natural killer – cells that were keeping the melanoma in limbo.

The process at work here is similar to one of the mysteries of infectious disease – we know a certain bacteria or virus causes it but not everyone with the microbe in their body develops the infection. It’s not the bugs that are different but the conditions in the body in which they find themselves

Mukherjee’s solution to making cancer treatment more effective by getting better at preventing or controlling metastasis is to work out ‘how to alter a susceptible patient’s immunological and histological profile to resemble that of a resistant one.’  

It’s this shift to treating cancer by enhancing the body’s own invasion-beating abilities that seems to me to open a very significant Overton Window on cancer. Many many patients are all too familiar with oncologists’ indifference or outright hostility to questions about the benefits of changing their diet or boosting their immune system.

Oncology needs to get more complicated

Mukherjee is very much aware he is inviting oncologists to venture into unfamiliar and far more complex territory than they usually inhabit. Unlike identifying a gene or two and taking them out with a drug, ‘invasion ecology’ involves investigating lots of possible factors that could have allowed a species that was a well-behaved Jekyll at home to become a Hyde in a new environment.

‘Cancer biologists,’ Mukerjee comments ‘are trained to focus their attention exclusively on our ‘pathogen’: the cancer cell. Ecologists, meanwhile talk about webs of nutrition, predation, climate, topography, all subject to complex feedback loops.’ Clearly the ecology model provides a much more realistic picture of the multitude of factors that determine whether one of those cells sprayed out in their millions from an established tumour is going establish a foothold and begin its programme of rapid growth.

So what might an open-minded, ecology orientated and pioneering oncologist actually do differently to frustrate a cancer invasion? At this point Mukherjee’s brave and well-informed plea for a wider view of cancer runs our of steam.  up against a word that has not yet been glimpsed through an Overton Window – holistic. He clearly recognises that adopting the ecological model means taking into account the complex interactions between cells and systems in the body – what has been termed a holistic approach – but declares firmly that ‘holistic’  the word has become ‘a patchouli-scented catchall for untested folk remedies: raspberry-leaf tea and juice cleanses.’

A couple of years ago when I was writing about the rediscovery of the Warburg hypothesis and the work of Professor Bissell I asked a friend of mine – a senior professor specialising in prostate cancer – what he thought of these ideas. He replied with a single word: balderdash.

Wider view could boost cancer options

Much as I admire Mukherjee I have to describe his characterisation of holistic in the same way. What his response tells us is that he has not had the time or inclination to look at what else is available outside the drug model. However once you become aware of the effect that diet can have on your microbiome (big research programme on links between cancer and the microbiome underway at Imperial College in London) on your metabolism (ways to reduce high glucose and insulin levels that raise cancer risk) or the plant chemicals that affect genes involved in DNA repair) you are a very long way from raspberry leaf tea.

Professor Bissell has been very involved in what can be done to maintain your ECM as good supportive neighbourhoods. ‘If we took seriously at all the things that we know can help cancer prevention,’ she says ‘such as tackling obesity, exercising, eating fewer carbohydrates especially sugar, we would have many fewer fully fledged tumours and a reduction in mortality.’

She also favours some less familiar options such as  ‘combining treatments aimed at boosting our immune and energy generating systems with less toxic drugs and using fasting (calorie restriction) as an effective way to reduce the inflammation that’s linked with cancer.’

The ‘wood’ is full of many very promising ways to help control cancer that could support the severely restricted treatment options available now. My hope is that Mukherjee’s article will help to open another Overton Window leading to funding for a research programme that isn’t focused exclusively on drugs. 

Jerome Burne

Jerome Burne

Jerome Burne is the editor of HealthInsightUK. He is an award-winning journalist who has been specialising in medicine and health for the last 10 years and now works mainly for the Daily Mail. His most recent book “The Hybrid Diet” was written with nutritionist Patrick Holford, published 2018. Award: 2015: Finalist for 'Blogger of the Year' Medical Journalists' Association.


  • I had similar thoughts after reading it: a brilliantly considered piece that dismissed holistic and complementary approaches too readily. Mina Bissell’s work is a prime example of how useful if can be to think differently. We desperately need to start joining some of these dots. Thanks for another inspiring article.

  • I really enjoyed reading this article, because I have become ever more aware of how damaging blinkered thinking in the sciences has become.

    I remember, more or less as a child, puzzling over the fact that some physical injuries can become cancerous. That never made sense to me because I had read that cancer is caused by mutation. Likewise, I’d read suggestions that growth hormones might cause cancer because they promoted growth, and again this didn’t seem to fit very well with the concept of cancer being all about DNA damage.

    What I am getting at, was that even back then (late 60′s), there were hints, accessible to anyone who gave the matter any thought, that mutations were certainly not the full story. It is frustrating to think how slow it is for medical science to change.

    You wrote: “This uncertainty means patients are routinely offered treatments such as chemotherapy or hormone blockers to cut the risk of a return. A small fraction benefit but there is no way to tell who they are in advance. The rest just suffer the various unpleasant and damaging side effects without the benefit.”

    Having become aware of the huge NNT’s for statin treatments, I wonder if you can quantify that statement in terms of NNT values? As we age, the value of receiving drastic chemotherapy must diminish as the remaining life expectancy gets less.

    • David, I’m also interested in the benefits and adverse effects of chemotherapy. In 90% of patients chemotherapy increased five-year survival by 2.5%. I doubt many people are told that.

      The following extract is from an excellent BMJ article by Professor Peter H. Wise entitled ‘Cancer drugs, survival, and ethics’ followed by part of an editorial entitled ‘Too Much Chemotherapy by Dr Fiona Godlee.

      Peter H Wise, former consultant physician and senior lecturer

      “A meta-analysis published in 2004 explored the contribution of cytotoxic chemotherapy to five year survival in 250 000 adults with solid cancers from Australian and US randomised trials. An important effect was shown on five year survival only in testicular cancer (40%), Hodgkin’s disease (37%), cancer of the cervix (12%), lymphoma (10.5%), and ovarian cancer (8.8%). Together, these represented less than 10% of all cases.

      In the remaining 90% of patients—including those with the commonest tumours of the lung, prostate, colorectum, and breast—drug therapy increased five year survival by less than 2.5%—an overall survival benefit of around three months.”

      Editorial by Fiona Godlee

      People with cancer are living longer now than 40 years ago. This is clearly good news. But how much of this improvement can we attribute to drug treatment? Not much, concludes Peter Wise this week in an article I humbly suggest all oncologists should read (doi:10.1136/bmj.i5792). The nearly 20% improvement in five year survival over the past four decades is probably mainly due to improved early diagnosis and treatment rather than developments in cytotoxic chemotherapy, he says. And patients are being badly misled by over-enthusiastic accounts of what chemotherapy can achieve. Many expect a cure. In reality they will gain on average only a few months of extra life.”

      ‘Cancer drugs, survival, and ethics’

      ‘Too Much Chemotherapy’

      • Editorial

        Great post Stephen. many thanks for mentioning it. Wise’s conclusion should be engraved over the entrance of all oncology clinics. The very least patients should have is the information needed to make a decision.

        Am working on feature around an impressive book due out later this month called Keto for Cancer by an American woman (teacher at the time) whose four year old was diagnosed with a brain tumour in 2002. After over two years of first chemotherapy and then a series of operations, the tumour was still active and growing. She then came across the ketogenic diet on the internet and the work of Dr Seyfried – (very little known at the time). In face of professional scepticism she tried it and after 3 months the tumour had slowed significantly. Her son Raffi eventually died aged 13 – not common for brain cancer. One of the themes of the book (along with impressive and detailed biochemical and practical information) is how the oncology establishment is still responding to queries about the diet by claiming no evidence – i.e. no large scale RCT – reasonable response would seem to be “Why not? The other objection is that it is challenging to follow. These are people who deliver chemotherapy for a year.

        • Thank you. I’m afraid I can’t help thinking oncologists are partly blind to new developments because it helps to maintain their monopoly of treatment. They have no incentive to do anything that threatens their highly paid positions.

          Surely people with cancer have an absolute right to be given the truth about chemotherapy? The lack of true consent in the NHS is scandalous. It’s as if the law on consent is just a theory to be quietly ignored. Patients are merely components on the medical conveyer belt. At times the NHS looks more like a job creation scheme than healthcare.

        • I found the responses from doctors and surgeons to ‘Too Much Chemotherapy’ particularly telling. They’re well worth a look for anyone interested in the subject. Here’s a response from a surgeon:

          I find myself constantly at odds with optimistic oncologists with the “never say die” spirit, despite evidence of tumour progression and patients’ problems with side effects. Too much chemotherapy is of course expensive but the main thrust is the patients’ well being or lack thereof. To give an example of optimism: I have done bypass operations for patients with obstructed, non-resectable colon cancer and widespread mets. The next step of course is a visit to the oncologist. I prepare the family not to expect too much. To my dismay, when I see the patients on their following visit with me they are bubbling with confidence after the oncologist has assured them of a 10 year survival. To cut the story short, they are often dead within a year. We need to know our limitations and when to stop playing God.

          Arthur E. van Langenberg, colorectal surgeon, Canossa Hospital, Hong Kong

    • Editorial

      Apologies for late response to this post which raises the very interesting issue of cancer drugs and NNT (Numbers Needed to Treat – how many have to get a drug for one patient to benefit?). First point is that there is a very important but little known difference between the way cancer drugs and all other ones are calculated. Dr Malcolm Kendrick explains this in a piece about his excellent book ‘Doctoring the Data’ that he wrote for HIUK which describes how drug companies and medics can spin the results of research to make it look more beneficial than it is.

      “There is a fascinating little known difference in the way the benefits of statins and cancer drugs are calculated. No one ever tells you how much longer you might live from taking a statin but the benefits of all cancer drugs are calculated in terms of ‘increased life expectancy’.

      The result is that cancer drugs don’t seem very effective — an extra three to four months is not uncommon. But that is actually far more impressive than the benefit from a statin.

      This is because those few months make up the average increased survival time for everyone on the drug. Some will do much better, some much worse but almost everyone on a cancer drug will live a bit longer.

      But that is not true for a statin. After years of treatment only one extra person out of 200 will live a bit longer — probably a few months — the rest will die at the same time as they would without it.

      Cancer drugs are seen as being very expensive for every extra month of life they give, but looked at this way they might actually be better value than the far cheaper statins which are many times less effective.”

      Even so I did find that David Bailey’s suggestion had been made in a paper published last year entitled:
      “Number needed to treat for pricing costly anticancer drugs: the example of regorafenib in metastatic colorectal cancer” The paper suggested that the NNT would be a useful way of deciding if new drugs were worth licencing.

      In the particular cased of regorafenib the paper found that “on average 10 patients (95% CI 6–44) are needed to be treated with regorafenib to prevent one additional event (death) compared with placebo.’ Compared with a statin that looks very impressive but as Kendrick explains statins have been calculated in a different way.

      What this exercise showed was that the improvement in life expectancy was at the lower end of the poor results we have come to expect. “The median overall survival (OS) was 6.4 months in the regorafenib group versus 5.0 months in the placebo group. The study met the primary end point, with an improvement of 1.4 months in median OS favoring regorafenib.” If getting that marginal benefit is counted as a success it seems unlikely that basing licencing decisions on NNT would make a significant difference to which got licenced and which didn’t.

  • The “surgical approach” to cancer is pragmatic, often effective but without understanding of cancer. Radiotherapy and chemotherapy are the same. We cannot expect those who treat cancer to understand the causes. What we know is that about 50% of world human cancers are due to viruses – hepatitis B virus, hepatitis C virus, Epstein Barr virus, human papilloma virus, human T-cell leukemia-lymphoma virus. Mouse mammary tumor virus must be considered in human breast cancer, which can be muticentric. Microbiology remains in its infancy with exciting new developments and discoveries. Metastasis due to viraemia rather than spread of cells? Cancer is more common in those immunosuppressed, including by chemotherapy.

    The other important point is that many cancers are based on inflammation – “Tumours are wounds that do not heal”. Cancers can occur as developments of inflammation in the breast, the skin, the oesophagus, the lung, the colon, the liver. Again a suggestion of a microbiological cause and the need for maximising immunocompetence.

    Mutation is not the answer. It is likely to be the reprogramming of cells by viruses incorporated into the genome.

  • Siddhartha Mukherjee is the guest on Radio 4′s ‘Desert Island Discs’ next Sunday morning (8th October at 11.15 a.m.).

  • Were you aware of the Rethinking Cancer conference that took place in France, at the Institut Gustave Roussy, on September 21st? Several cancer researchers, including Thomas Seyfried, were there and it looks as though a research project is going to spring from it. I was there with my husband who gave a short talk about his own experience with oesophageal cancer and his total recovery thanks to what he learned from the work of Thomas Seyfried. The conference was organised, not by researchers, but by an airline pilot who is managing his own cancer using lifestyle methods.

Leave a Reply

WP-Backgrounds by InoPlugs Web Design and Juwelier Schönmann