A New Theory of Cancer
On a disease’s evolutionary history and the implications for treatment
When President Richard Nixon declared war on cancer in 1971, he set a goal for conquering the disease by 1976. The US National Cancer Institute (NCI) was empowered and expanded by the stroke of the president’s pen, and some serious public money injected into a mammoth research effort. In the intervening 47 years, in excess of US$100 billion of taxpayer money has been spent by the NCI alone on the search for an elusive “cure for cancer”. Its 2018 fiscal year budget is just shy of US$6 billion. Cancer research also attracts billions from drug companies and through charitable donations. Proportionately similar amounts have been spent in other developed nations, Australia included.
So what have we got for all the money spent? Not what was promised, clearly. Survival rates for many major cancers are little changed in several decades. Treatment regimes – a combination of surgery, radiation and toxic chemicals – remain much as they were in Nixon’s day. Life extension is often measured in weeks or months rather than years, and interventions are frequently a rearguard action against the inevitable. Optimistically touted “breakthrough drugs” are typically efficacious on only a few per cent of patients and often have dreadful side effects. Meanwhile, cancer incidence has grown remorselessly as overall life expectancy has risen, making cancer now the world’s number two killer, with about 14 million new cases per year. It’s fair to say that cancer touches every family on the planet.
Screening programs complicate the picture. They are very effective for skin and colon cancers because early surgical intervention can be 100 per cent successful. But breast and prostate cancer screenings are extremely controversial because even though doing nothing is often the best option – many early-stage cancers never progress to be life-threatening – patients are understandably reluctant to merely watch and wait when diagnosed with a potentially killer disease. Furthermore, as screening technology has improved, tiny tumours are being spotted earlier, which has the effect of skewing the statistics. The medical profession has defined a “cancer survivor” as someone remaining alive five years after diagnosis. Find tumours earlier and there will appear to be more survivors even if there was zero progress in treatment outcomes. So claims that we are slowly winning the war against cancer have to be nuanced. For the record, current overall five-year survival rates in Australia are still less than 70 per cent.
It is not all bad news. Some cancers have indeed been largely conquered by drugs, childhood leukaemia being a famous example. Lung cancer, the scourge of the Nixon era, has steadily declined as smoking has become unfashionable. And the careful use of drug combinations has progressively extended survival times for certain cancers. But others are on the rise, and the overall picture is far from rosy, as is obvious from the continual pleas for additional funding.
Unfortunately there is a widespread belief that if enough money is thrown at the problem a solution will be found. All science requires funding, of course, but the truth is we have to think our way to a solution, not spend our way to one. And we won’t outsmart cancer until we understand what it is. I don’t mean understand it as a disease, but understand it as a biological phenomenon. There are more than a million published research papers on cancer and yet the simple questions, “What is cancer?” and “Why does it exist?”, have no clear answers in the scientific community. Maybe progress against cancer is so slow because we have been thinking about the problem the wrong way?
I first became involved in this subject in 2008 when I received a phone call from Anna Barker, at that time deputy director of the NCI. She perceived the need for some radical new thinking to replace the sameold-same-old mindset of many cancer researchers, and wondered whether bringing in scientists from other disciplines, notably physics with its stunning track record of success, might provide a welcome shakeup of the field. I countered that I knew nothing about cancer, and she replied, “That’s perfect!” Thus began a US$35 million per annum NCI-funded program in physical science and oncology. In all, 12 new centres were created across the US, each pairing a physical scientist with an oncologist.
I was chosen to run the centre at Arizona State University (ASU), and I took seriously Dr Barker’s entreaty to rethink cancer from the bottom up. As a physicist and cosmologist I am used to sidestepping technicalities and searching for deep unifying principles. A culture clash was immediately apparent. Oncologists work at the sharp end of the subject, and every patient is different. It’s natural that physicians should regard cancer as a bafflingly complex disease. But most cancers follow a fairly predictable pattern: a tumour grows in a specific organ, and after a while some cells leave the tumour and spread around the body, invading remote organs and creating secondary tumours, usually with fatal consequences. As cancer progresses, it displays several distinctive hallmarks, including uncontrolled proliferation, increased motility, evasion of the immune system and organisation of its own blood supply. I wanted to know what underlies this bizarre phenomenon and why it happens.
What first struck me was the fact that cancer, far from being restricted to humans, is found right across the tree of life, in most multicelled organisms. Scientists have found cancer or cancer-like phenomena in all mammals, as well as in fish, birds, worms, insects and corals. There are even cancers among plants and fungi; at ASU we have just created a “cancer garden” displaying striking examples of cancer in cacti. Recently, cases of cancer have been found in the humble hydra, a tiny organism with only seven cell types.
In biology, the most widespread properties of organisms are usually the most ancient, and can often be traced to a common ancestor in the far past. Photosynthesis, for example, is used by all plants and many bacteria, and dates back more than three billion years. Could we use the tree of life, I wondered, to trace the origin of cancer?
For the greater part of our planet’s history, life was restricted to single-celled organisms: bacteria and archaea. The earliest uncontentious traces of life come from the Pilbara region of Western Australia, and date back about 3.5 billion years. Single cells have but one imperative: replicate, replicate, replicate! Bacteria that go on dividing and multiplying are in a sense immortal. It goes without saying, however, that cancer is a disease of bodies; it makes little sense to say a microbe has cancer. But bodies in the usual meaning of the word, in which there are different tissue types with specialised functions, had to await the onset of multicellularity, which first emerged about 1.5 billion years ago. By 600 million years ago, many of the basic body plans we see today had evolved.
A multicelled organism executes the life project very differently from a microbe. Immortality is outsourced to specialised germ cells (for example, eggs and sperm), which carry the genetic legacy down the generations, while the rest – the so-called somatic cells – accept eventual death. In an echo of their free-living past, somatic cells may undergo quite a few rounds of division – skin cells, for example, can divide up to 40 times – but eventually all normal somatic cells commit suicide, a process known as apoptosis. In the absence of the cells’ renewal (by so-called stem cells), the organism must finally die.
To police this ancient contract between individual somatic cells and the organism as a whole, layer upon layer of regulatory control has evolved. As always with a cooperative venture, there is a danger of cheating. Just as humans are tempted to accept the benefits of society but dodge their taxes, so somatic cells harbour inner instincts to freeload off the benefits the organism provides but evade the apoptosis police. The result is cancer – the uncontrolled proliferation of a population of somatic cells, triggered by a disruption of functions that evolved to regulate multicellular organisation.
Nothing I have stated above is controversial, but few oncologists choose to think about cancer in this way – that is, as a biological phenomenon with deep evolutionary roots stretching back to the Proterozoic era of life on Earth. But to make serious progress we need this broader context.
Cancer as a throwback
The standard explanation for cancer, known as the somatic mutation theory, is that random mutations – genetic defects in DNA – accumulate in cells over time, perhaps as the result of chemical damage or radiation, until a point is reached where a mutated cell embarks on a rampage, multiplying uncontrollably, forming a “neoplasm”, or population of new cells, and behaving in many respects like a parasitic independent organism. Though entrenched, the somatic mutation theory has poor predictive power. Moreover, it struggles to explain how random mutations confer so many fitness-improving gains of function in a single neoplasm in the space of weeks or months. It also seems paradoxical that increasingly damaged and defective genomes engage in such systematic and broadly predictable behaviour, acquiring the various hallmarks I mentioned. On top of this, it is clear that mutations cannot be the whole story; a tumour’s micro-environment is also critical in determining the tumour’s behaviour. Even highly mutated cancer cells can be “tamed” by surrounding healthy tissues.
Over the past few years, my colleagues and I have developed a somewhat different explanation of cancer that seeks its origins in the far past. We were struck by the fact that cancer never invents anything new. Instead, it merely appropriates already existing functions of the host organism, many of them very basic and ancient. Limitless proliferation, for example, has been a fundamental feature of unicellular life for aeons. After all, life is in the business of replication, and cells have had billions of years to learn how to do it well and keep going in the face of all manner of threats and insults. Metastasis – the process whereby a neoplasm spreads around the body – mimics what happens during early-stage embryo development, when cells surge in organised patterns to designated locations. And the propensity of circulating cancer cells to invade other organs closely parallels what the immune system does to heal wounds. These facts, combined with the predictable and efficient way that cancer progresses through its various stages of malignancy, convinced us that cancer is not a case of damaged cells randomly running amok but an ancient, systematic and brutally efficient pre-programmed strategy that is deployed when cells are challenged or threatened in some way. Crucially, we believe that the various distinctive hallmarks of cancer do not independently evolve as the neoplasm goes along, but are deliberately switched on and deployed in an organised strategy.
In summary, our view of cancer is that it is not a product of damage but a systematic response to a damaging environment – a primitive cellular defence mechanism. Cancer is a cell’s way of coping with a bad place. To be sure, it may be triggered by mutations, but its root cause is the self-activation of a very old and deeply embedded toolkit of emergency survival procedures. A helpful analogy is a computer that suffers an insult,
Few oncologists choose to think about cancer in this way – that is, as a biological phenomenon with deep evolutionary roots.
such as corrupted software, and starts up in safe mode. This is a default program enabling the computer to run on its core functionality even when defective. In the same way, we think, cancer is a default state in which a cell under threat runs on its ancient core functionality, thereby preserving its vital functions, of which proliferation is the most ancient, most vital and best preserved.
Although elements of the cancer default program are extremely ancient, dating back to the origin of life itself, some of the more sophisticated features revisit later stages in evolution, especially in the period between one billion and 600 million years ago, when metazoans (complex multicelled organisms) first emerged. Thus cancer is a type of atavism – a throwback to an ancestral form or “phenotype”. That essential idea was proposed as long ago as 1914 by the German biologist Theodor Boveri, but was sidelined until recently.
It is because cancer is so deeply integrated into the fundamental logic of multicellular life that combating it proves such a formidable challenge. People talk about our “inner fish”; well, rewinding the evolutionary tape still further back in time connects us to our inner cancer. Sadly, it seems that cancer is an accident waiting to happen.
Testing the new theory
I was lucky that when I embarked on the NCI project several other people had been thinking along the same lines. These included a longstanding collaborator of mine in the fields of cosmology and astrobiology, Charles Lineweaver of the Australian National University. Importantly, however, a number of clinical oncologists had either formed similar conclusions or shared our general unease about the inadequacy of the somatic mutation theory. Among these were Mark Vincent, medical oncologist at the London Regional Cancer Centre in Ontario, Canada, and David Goode of the Computational Biology Program at the Peter MacCallum Cancer Centre in Melbourne. In addition, one of the more far-sighted and renowned American oncologists, the late Don Coffey of the Johns Hopkins University School of Medicine in Baltimore, was a strong supporter of our endeavours. He even invited me to do “grand rounds” at Hopkins to present the basic ideas.
In science, for a new theory to be taken seriously it has to not only explain the known facts but also offer testable predictions. Fortunately our theory was developed just as gene-sequencing technology was advancing rapidly and very large genetic databases were being compiled. The key novelty is our emphasis on the ages of genes, which is not normally a consideration in cancer research. It is possible to estimate how old a gene is by comparing how gene sequences diverge across many species. This procedure is known as phylostratigraphy, and it enables scientists to reconstruct the tree of life, working backwards from common features today to deduce the convergence point in the past. Using this technique, we can trace the evolutionary origin of genes
that are implicated in cancer, a subset of which are called oncogenes. If the throwback theory is along the right lines, one expects that cancer genes would cluster in age around the onset of multicellularity.
A study by Tomislav Domazet-Lošo and Diethard Tautz in Germany used four different cancer gene datasets and confirmed the presence of a marked peak in age at around the time that metazoa evolved. A recent analysis of seven tumour types by Anna Trigos, Richard Pearson and Anthony T. Papenfuss in David Goode’s laboratory in Melbourne sorted genes into 16 groups by age and then compared how strongly the genes are expressed in cancer versus normal tissue for each group. The results were striking. Cancer over-expresses genes belonging to the two older groups and under-expresses younger genes, much as we predicted.
The Melbourne study also went further, suggesting that the regression to a more primitive cellular state was not an across-the-board affair, but a highly regulated and finessed process. Genes rarely act in isolation; rather, they form cooperative networks. Goode and Trigos found that the gene networks dating from the era of unicellularity are systematically decoupled from the more recently evolved multicellular gene networks, revealing a novel pattern of gene expression specifically tied to gene ages. They reported a strong association between the evolution of multicellularity and patterns of gene expression in cancer. Furthermore, they found that as cancer progresses to a more aggressive, dangerous stage the older genes are expressed at higher levels, confirming our view that cancer reverses the evolutionary arrow at high speed as it develops in the host organism.
Our own work at ASU, much of which was carried out by a geneticist, Kimberly Bussey, in collaboration with a physicist, Luis Cisneros, focused on mutation rates. The atavistic theory predicts that older genes should be less mutated in cancer (after all, they are responsible for running the “safe mode” program), while younger genes should be mutated more. My colleagues considered a total of 19,756 human genes and used an inventory of cancer genes, compiled by the UK’s Sanger Institute, called COSMIC. This data was combined with a database of genetic sequences from about 18,000 species across all taxonomic groups, which allowed an estimate of the evolutionary ages of the genes in the human genome. We found that genes younger than about 500 million years were indeed more likely to be mutated in cancer, while genes older than a billion years tended to suffer fewer mutations than average, as expected.
The most telling result came from addressing a rather different question: what are cancer genes good for? A gene classification tool called DAVID organises genes around their function. When my colleagues fed the COSMIC data into DAVID, what leapt out was that recessive genes older than 950 million years were strongly enriched for two core functions: cell cycle control, and DNA damage repair involving double-strand breaks (the worst kind of damage DNA can suffer). Looking at the evolutionary history of those genes involved revealed a startling and unexpected result. The cancer genes implicated in DNA repair turned out to match up with genes in bacteria employed for a critical survival function.
When bacteria are stressed, for example from starvation, they deliberately ramp up their mutation rates with a view to evolving out of trouble. The mechanism they employ was elucidated by Susan Rosenberg, who holds the Ben F. Love Chair in Cancer Research at the Baylor College of Medicine in Houston, Texas. Rosenberg found that stressed bacteria create a distinctive pattern of self-inflicted mutational damage around DNA double-strand breaks, extending either side of the repaired break for thousands of DNA bases, by strategically deploying a sloppy repair mechanism. Bussey and Cisneros at ASU found identical patterns of damage in cancer DNA, created by the same mechanism controlled by essentially the same genes. This discovery is important because the clustering of mutations in this manner is known to be associated with poor patient prognosis.
The most telling result came from addressing a rather different question: what are cancer genes good for?
Elevated mutation rates are one of the best-known hallmarks of cancer and the main reason why chemotherapy falters when neoplasms evolve drug-resistant variants. We collaborated with a research team led by Robert Austin at Princeton University to investigate the details of drug resistance, and specifically to address the question of whether resistance is acquired by random mutations plus Darwinian selection, as the somatic mutation theory predicts, or from a more deliberate and organised response. The Princeton group subjected human cancer cells to a therapeutic toxin (doxorubicin) and found a highly uneven pattern of mutations – hot spots of elevated mutation and cold spots that seemed to be protected from damage. And true to the atavism theory, they found that the genes in the cold-spot regions were significantly older than average.
The Princeton results explain why natural selection hasn’t eliminated the scourge of cancer. If tumours really are a reversion to an ancestral form, then we might expect that the ancient pathways and mechanisms that drive cancer would be among the most deeply protected and conserved, as they fulfil the most basic functions of life. They can’t be got rid of without disaster befalling the cells concerned. It makes sense that organisms should work hard to protect key parts of their genomes, such as those ancient genes responsible for running the
core functions of the cell, and devote fewer resources to the “bells and whistles” associated with more recently evolved and less critical traits.
Further support for the atavism theory may come from a worldwide effort to enlist zoos in a comprehensive study of cancer across species. The initiative is being managed by the Arizona Cancer Evolution Center, run by my colleague Carlo Maley. As part of the program, we have begun collaborating with Taronga Zoo in Sydney. I want to know whether cancer in marsupials differs from that in placental mammals, given their widely different developmental strategies. It has been known for decades that certain oncogenes play a crucial role in early-stage embryo development. Normally, these developmental genes are silenced in the adult form, but if something reawakens them cancer may result; a tumour is, in effect, a botched embryo developing inappropriately in adult tissue.
The disruption of gene regulatory networks that heralds cancer involves dramatic changes in the patterns of information flow between genes and between cells, just as safe mode on a computer represents a major reorganisation of the machine’s software. Our research group at ASU is trying to find “information signatures” of these gene network changes. We think it will prove possible to identify distinct “informational hallmarks” of cancer to go alongside the physical hallmarks I mentioned, providing a software indicator of cancer initiation that may precede the clinically noticeable changes in cell and tissue morphology, thus providing an early warning of trouble ahead.
Implications for therapy
The atavistic theory of cancer has important implications not just for diagnosis but also for therapy. Most approaches target the strengths of cancer. For example, many drugs try to block the propensity of cancer cells to replicate rapidly. However, as I have stressed, cells have had four billion years to combat threats to their proliferative ability and they usually find workarounds that render the drugs ineffective, such as pumping the toxins out or defensively switching on mutator genes to evolve resistant strains. The standard chemotherapy regime of applying the maximum tolerable dose to hit cancer hard therefore seems intrinsically flawed, because it risks provoking an aggressive atavistic response. We suggest instead an approach based on the minimum efficacious dose, with a view to containing and controlling cancer rather than trying to exterminate it. Two clinical trials along these lines are currently being conducted, one by the Arizona Cancer Evolution Center in collaboration with the Mayo Clinic, and the other at the Moffitt Cancer Center in Tampa, Florida.
Better still would be an alternative to toxic chemotherapy altogether. One of the more important “bells and whistles” that multicellular life has evolved over the past few hundred million years is the adaptive immune system, instrumental in fighting infections but also in surveilling for cancer. The atavism theory predicts that as
cancer progresses – and hence regresses in evolutionary terms – it should subvert this system, and indeed it does. A great deal of attention has been given recently to immunotherapy as a powerful new way to combat cancer. The basic idea is to boost the body’s immune system so as to strengthen the immune response. It is too soon to know whether immunotherapy will prove to be the decisive breakthrough claimed or yet another example of cancer outwitting whatever the physician throws at it. But early results are promising.
There is a curious backstory here. Over a hundred years ago, the American physician William Coley was intrigued by the fact that some cancer sufferers undergo spontaneous remission following an infection. Coley even experimented with deliberately infecting patients with streptococcus, a bold – some might say reckless – practice that soon went the way of leeches. But Coley may have been onto something. The conventional explanation is that the infections boosted the immune system, which then destroyed the cancer as incidental collateral damage. We contend, however, that at least part of the reason for Coley’s results is that cancer tumours are more vulnerable to infections than the rest of the body because they have decoupled from the adaptive immune system. In other words, by regressing to an immunocompromised state, tumours leave themselves unprotected against infections. The selective use of viruses and bacteria against some late-stage cancers therefore seems a rational approach.
Another therapy idea to come from the atavism theory also harks back a century, to the work of Otto Warburg, a Nobel Prize–winning physician. Normal human cells use oxygen to generate energy, but cancer often switches to fermentation, a low-oxygen, high-glucose process. It is less energy-efficient, but good for making biomass. Warburg discovered that cancer will switch to fermentation even in the presence of normal oxygen levels. It is tempting to speculate that, in reverting to an ancestral form, cancer is reprising a lifestyle adapted to the state of planet Earth at the time when multicellular organisms first evolved. And geologists have determined that before a billion years ago there was indeed far less free oxygen in the atmosphere. Some researchers have used this insight to advocate the application of hyperbaric oxygen therapy combined with a low-glucose diet to stymie the Warburg effect and slow the cancer.
I believe the search for a general-purpose “cure” for cancer is an expensive diversion. Being so deeply entrenched in the nature of multicellular life itself, cancer is best managed and controlled (not exterminated) by challenging the cancer with physical conditions inimical to its ancient atavistic lifestyle. It does, however, require a change in the culture of cancer care, away from Nixon’s all-out war and towards accepting cancer as a chronic disease. We don’t have to destroy cancer; we just have to prevent cancer destroying us. Only by fully understanding the place of cancer in the overall context of evolutionary history will a serious impact be made on human life expectancy. M