Cancer is evolution taking place inside multi-cellular organisms. Cancer cells are perversely adaptive, in the evolutionary sense of the word, because they proliferate faster than neighboring cells. Since evolution has no foresight, it doesn’t matter that the end result of cancer evolution is the death of the cancer along with its host.
Amazingly, only a tiny group of researchers are taking this approach to cancer. One of them is Athena Aktipis, an Assistant Professor at Arizona State University. Her training as a theoretical evolutionary biologist gives her a special advantage in thinking about cancer in the same way as cooperation and conflict among multi-cellular organisms. In addition to her dynamic career as a cancer researcher, she co-directs the Human Generosity Project.
DSW: Welcome, Athena.
AA: Thank you.
DSW: I want to talk with you about cancer from an evolutionary perspective. You’re a pioneer in this area and part of a relatively small group. That sounds strange to most people because they know that cancer research is a highly sophisticated area. Tons of money are thrown at it and the research is very high tech. So my first question for you is, how is it possible for cancer research to be highly sophisticated in some ways and yet in other ways to be “pre Darwinian” – not from an evolutionary perspective? Can you explain that to me and our audience?
AA: One thing that’s really interesting about the field of evolution and cancer is that it’s actually been around for a long time. It’s been accepted that evolution is the main theory of cancer for decades now, but what happened is that soon after the evolutionary framework for cancer was put into print and really started being used in the 70s, the molecular revolution in biology came around and that moved the field into studying mechanisms and focusing on particular pathways within the cells. That’s starting to change because there’s been a genetics and genomics revolution and now big data. There’s increasing awareness that we’re only going to understand all the data that’s available by taking an evolutionary perspective and using evolutionary tools and methods on the data that we already have.
DSW: This mirrors the history of molecular biology, which started out being purely mechanistic and is now becoming fully rounded in an evolutionary sense. Does that sound about right?
AA: Yes, that sounds right.
DSW: But you say the 1970s and I read from your work that the first paper was written in 1976. Is that right?
AA: Yes, that’s when the first paper was written but, apparently, it was largely summarizing what people believed at the time. So in the 70s it was the way people were thinking about and approaching cancer.
DSW: Even though it was only put in print in the 70s?
DSW: Okay. I use Niko Tinbergen’s four questions a lot, concerning function, history, mechanism, and development. My way of phrasing this is that cancer research being highly molecular is basically focusing on the mechanism question and not enough on the other three questions. Is that more or less accurate?
AA: I think that molecular approaches to cancer biology are still most dominant, although they interact of course with looking at the genetics, and now there’s increasing interest in the microenvironment, so that starts to put a little more of the history and the context in which selection is taking place into the picture.
DSW: Yes. This is where an ecological background in addition to an evolutionary background comes in handy. I know because I’ve been reading your work that you cite wonderful examples; for example, cancer cells that differ in their distance from a blood vessel could experience different selection pressures, just as animals that live at different distances from a river. Is that correct?
AA: Yes, that’s exactly right, and it might be very important for understanding what happens not just for the initiation of cancer but during treatment, because cancer cells close to a blood vessel will have higher levels of whatever chemotoxins are being used in therapy. Let’s say that a tumor has evolved and grown so much that the interior is totally absent of nutrients, like you see in some cities, where the slums are totally absent of resources and filled with garbage. There is intense selection for cells to survive better in those conditions, which we call hypoxic conditions. These cells become pre-adapted to live in regions that are far away from blood vessels, so when you give chemotherapy, they are able to hide out, a phenomenon called “refugia” in cancer treatment. There are some really important aspects of the environment, ecology, and diversity of ecological niches that get created in the course of cancer progression and changed during treatment that haven’t been fully considered from an evolutionary perspective. There’s a lot of opportunities.
DSW: So this puts cancer therapy in the same class as antibiotic resistance and also pest resistance in agriculture. What does that mean in terms of effective therapies that would not have occurred to somebody without making that connection?
AA: Yes there are certainly a lot of parallels there. One important difference is that with cancer, any resistance to treatment that evolves within a patient dies with that individual whether they die of the cancer or something else, whereas infectious diseases that have evolved antibiotic resistance can be passed on. That’s actually a hopeful fact–we might be able to do better with treating cancer in a way that minimizes the evolution of resistance and we’re not dealing with the problem of the transmission of resistant strains across individuals.
DSW: Right! Another important aspect of studying cancer from an evolutionary perspective is the realization that multicellular organisms have been evolving defenses against cancer for about a billion years, so our bodies are well protected. Can you flesh that out? What are some of the body’s mechanisms that protect us against cancer?
AA: This is one of my favorite topics. When it comes to thinking about what we can learn from taking an evolutionary approach to cancer, cancer can teach us about the evolution of life. Cancer is a process of evolution taking place among the cells of the body, so cancer defenses are largely mechanisms that evolved by evolution at the level of the organism to suppress evolution within the organism. So I think that’s a cool aspect of…
DSW: It’s multilevel selection that it’s a cool aspect of!
AA: Exactly! More generally, one can think of multicellularity as a supremely sophisticated cooperative system or set of systems – we have different types of multicellularity — but what they all do well is to coordinate their parts in ways that are cooperative and allow the higher-level units to be functional and effective. Doing that requires regulating the behavior of the cells. In particular, we’ve identified foundations of multicellularity that we see in many different forms of multicellular life, which all seem to break down when we have cancer or cancer-like phenomena. These include control of cell proliferation and death. They also include goods and services, the more economic side, which involves allocating resources effectively among cells, maintaining a shared multicellular environment and dividing labor. These foundations allow multicellular entities to function effectively and accomplish adaptive goals on the level of the collective–and they all break down in one way or another with cancer.
DSW: One thing you write about is cell cycle checkpoint genes, which makes me want a little tutorial on stem cells. I understand these as cancer prevention mechanisms that limit the number of generations for any given cell line. Can you elaborate?
AA: The literature on so-called cancer stem cells is controversial. The basic idea is that certain cells can maintain the tumor and recreate a tumor de novo if they are put into a mouse with a compromised immune system. What makes it a stem cell is that it can recreate the tumor. Then there are other cells that are part of the tumor mass but can’t start a new tumor. Those are the basics of stem cells.
DSW: So that’s a cancer stem cell, but what’s a normal stem cell? It’s like our skin cells, right? They’re derived from stem cells and they only have a certain number of cell divisions… and here I’m getting rusty.
AA: Right. A stem cell can divide many more times than “differentiating” cells, which are also sometimes called transient amplifying cells. Those are the cells that have some fixed number of divisions and then undergo controlled cell death.
DSW: So a given skin cell on you right now is derived from a stem cell and has a limited number of generations. Is that right?
AA: That’s right. A stem cell will produce a number of differentiating cells from it, and so any given cell on our body is likely to have come from a stem cell not so many generations ago. That’s right.
DSW: I was just reading a paper in Science magazine, which analyzed skin from eyelids that had been surgically removed for a condition called drooping eyelids. These were middle-age and elderly individuals without cancer. The scientists did a
detailed biopsy of tiny sections of the eyelids, showing an amazing diversity of cell clones with lots of mutations associated with cancer. This genetic diversity is the normal state of affairs for aging skin tissue. But I couldn’t quite understand how that relates to stem cells. There were all these different clones and they were all expanding a little bit, but I didn’t know if they were going to divide indefinitely or if this stem cell thing was going to kick in. Can you help me out a little?
AA : Sure. First is this observation that we are not genetically homogenous, as we like to think. We have lots and lots of pre-neoplastic lesions or areas where there are genetic mutations that with additional mutations would lead to cancer. Those are all over our bodies. So the question is, how are they kept in check long enough for us to reproduce and to survive as long as we do? Stem cells provide part of the answer. Each stem cell creates a proliferative unit, resulting in an internally regulated chunk of tissue. This is probably a cancer suppression mechanism in itself. If you get a mutant stem cell that produces factors that make the proliferative units divide more often than they otherwise would, that might possibly lead to an expansion on the skin. These are all questions that people are starting to look at. How is cancer suppression built into the metapopulation structure of the cells that make up epithelial tissue? I think that’s one of the important questions that we’ll be able to address with multilevel selection theory intersecting with cancer biology more and more.
DSW: Another important point that you raise concerns cross-species comparisons. I understand that the vast bulk of cancer research is conducted on humans and mice, just two species. Is that right?
AA: Yes, that’s right.
DSW: Talk to me about the importance of cross-species comparison and some of the new comparisons being made, including dogs as a new model species for cancer research, which is fascinating.
AA: I think there are two main benefits that we can gain from taking a cross-species perspective. First, what are the general principals across life that lead to cancer susceptibility and that enable better protection from cancer? By looking at species that are particularly susceptible or resistant to cancer– their genomes, their cancer suppression genes, their physiology, how their cells are interacting with each other, what factors they’re producing–all these things can point us in the direction of how to prevent and treat cancer. Dogs, as you mentioned, are being studied more and more extensively as are pets and veterinary animals more generally. There is the “One Health Initiative” linked to Barbara Natterson-Horowitz and Kathryn Bower’s book Zoobiquity, increasing awareness among veterinarians and medical doctors that we can look at health from a more generally biological perspective rather than just a species-centric perspective. And there are lots of plans in progress now to examine certain kinds of treatment options in dogs and also to look at complex cancers that we might not be able to look at it in humans, like brain cancers, by using dogs as a model. So it’s a really exciting area and Matthew Breen and Joshua Schiffman are great people to talk to about that more. Both are at the cutting edge.
DSW: I’ve read one of their papers in the material you gave me.
AA: Josh Schiffman is involved in public outreach too, so he’s working with Barnum and Bailey circus and with Utah’s Hogle Zoo on getting elephant blood to examine cancer suppression mechanisms. It turns out elephants get much less cancer than humans, and some of that probably has to do with the extra copies they have of a tumor suppression gene called TP53. Carlo Maley and Josh Schiffman have been working together on this project for years and now the work is finally being published in JAMA, showing that elephants have multiple copies of TP53 and that they seem to have much more sensitive responses to radiation as a result, which probably protects them from cancer.
DSW: So this brings us to Petos’s paradox, which is that you’d expect big and long-lived species to have high cancer rates, but they don’t, so they must be exceptionally good at suppressing cancer. This creates a special interest in studying long-lived species such as elephants and whales. I think that’s so cool. Yet, within a species you do get that trend, that larger and older individuals are more prone to cancer. Taller people are more prone to cancer than shorter people–something I didn’t know but makes sense from an evolutionary perspective. These are fascinating things that jump out at you, and you never would have thought about it otherwise. This leads to my next question because the book I am currently writing is based on the idea, to quote Einstein, that the theory decides what we can observe. A theory organizes perception, so when you approach a topic such as cancer from an evolutionary perspective, some things that were previously invisible become clear and other things that appeared reasonable can be seen to be fallacies leading to counterproductive results. Can you give me an example of a cancer therapy that made sense from another perspective but can be seen to be ineffective or counterproductive from an evolutionary perspective?
AA: Well, probably the best example would be the bias toward as aggressive treatment as possible. For a long time it was almost a moral imperative to use the highest dose and most aggressive treatment possible, but now that’s being reexamined and many top cancer treatment hospitals in the country and the world are backing off from that. Not all of them are backing off because of an explicitly evolutionary framework, but some of the work that has been done over the last decade has helped to show why an overly aggressive approach can be problematic. As for drug-resistant infectious agents and resistance to pesticides in agriculture, high doses in cancer treatment imposes the highest selective-pressure on a population of cancer cells. If you have a small population that’s not very diverse, then using a high dose can make sense. But if you have a large and diverse population of cancer cells then the higher the dose, then the greater the selection pressure for resistant cells. That’s one of the really important insights that comes from taking an evolutionary approach to treatment. We shouldn’t just think about killing cells; we should be more strategic about what we want to select for and against.
DSW: Right. Another one is the idea of classifying cancer according to tissue types, like breast cancer or lung cancer, whereas evolutionary theory tells you that for any given tissue you’ll get a diversity of cancers because every one is a separate evolutionary experiment.
AA: Yeah, the question of how to characterize cancer or name it is an open question even within various tissue types. Now it’s understood that there are many types of cancer. Every cancer is different in every individual and indeed every cell in every cancer is pretty much different, so it’s much more heterogeneous than previous paradigms took into consideration. But I think there’s some potential for classifying cancer based on the evolutionary dynamics that are going on. Some of the work by myself and Carlo Maley, Daryll Shibatta, and Matt Breen is to create a classification system that is based on evolutionary and ecological parameters of cancer.
DSW Great! Of course, from a mechanistic standpoint there’s a limited family of mechanisms so in some generic sense there are some cells that will mutate more than others, since they’re involved in things like cell division. So you can make some general statements even though the particular mutations might be different, etc..
AA: Yes, that’s right, so part of our goal in that paper is looking at cancer across the tree of life and taking the cheating and cooperation perspective that is general across species. I think that evolutionary theory, ecological theory, cooperation theory can all come together to give us a way of quantifying cancer that will potentially be helpful for understanding the general principles underlying it so we can better treat it.
DSW: So we have some examples of cancer therapies that seem to make sense but not so much from evolutionary perspective. What are some examples of therapies that make sense from an evolutionary perspective and were missed by other perspectives?
AA: There’s an approach that’s been pioneered by Robert Gatenby at Cancer Center called adaptive therapy, which is based on conditionally applying therapy when a tumor is growing but leaving it alone otherwise. It’s not really understood why that works. It might be because it’s not imposing large selection pressures that favor the resistant cells and when the tumor isn’t being treated the sensitive cells are able to do better. It might also be that this conditional treatment is affecting the life strategies of the tumor cells. So treating the tumor only when it’s growing and leaving it alone when it’s not growing might be selecting for slow life-history strategy.
DSW: And your own work examines movement as metastasis. Metastasis is the evolution of movement, which will be favored by some ecological circumstance and not others.
AA: Yeah, that’s right. You have a situation when cancer cells are over-consuming resources, which creates selection pressures for individuals to leave and move around. Yes, my work has shown that this could be the initiator of the evolution of cell motility.
DSW: That’s amazing. Now why has the study of leukemia and esophageal cancer led the way? You say that one of the barriers to thinking about cancer from an evolutionary perspective is methodological because it requires such things as studying clonal diversity, which is not easy in some cancers and more tractable in others. Is that right?
AA: One of the reasons that leukemia helped to illuminate the evolutionary dynamics underlying cancer is because it’s so easy to get longitudinal samples. All you need to do is take blood, you don’t need to have invasive biopsy, and you can also take blood easily after treatment. You can see not just how the cancer may be changing over time but also how cancer is changing with therapy. And esophageal cancer is another fascinating case because it’s been discovered that a premalignant condition leads to esophageal cancer but not that often–I think it’s 10%. It used to be the case that they’d remove the esophagus, but there were pretty high mortality rates from the surgery. So they stopped routinely removing it and began biopsying it every few years. This created a huge database that could be mined for looking at evolutionary dynamics associated with progression to cancer or not.
DSW: I read there’s a connection between cancer and microbiomes. Just as gall-forming insects will change plant development to create homes for themselves, something equivalent takes place in our bodies, and esophageal cancer can be involved with that.
AA: Yes. There is increasing evidence that microbes are associated with gastrointestinal cancers of many types. I think this is one of the new exciting areas where an evolutionary perspective can ask deeper questions than other perspectives. If microbes are playing a cardinal role in certain types of cancer, why is that? Is there a way that these microbes are increasing their fitness by changing tissue architecture? Why haven’t our bodies evolved a better way to defend against them? It starts raising questions that take us down a different route than just looking for a molecular mechanism.
DSW: I want to end up by talking about how fast the evolutionary perspective is catching on. In one of your recent articles, you do an analysis of the literature on therapeutic resistance and relapse, which seems to indicate that the evolutionary perspective is not catching on. Something like 1% of articles on this manifestly evolutionary topic mention any evolution-associated words and there’s no upward trend. Can you talk a little bit about that article?
AA: That article is from an analysis we did 5 years ago and it’s possible that things have changed since then. I haven’t gone back to look. It’s possible there are many unrealized opportunities in this field–whether it’s taking an evolutionary perspective or looking across species. We have new tools and methods that evolutionarily affords us to look at cancer.
DSW: And yet it’s not catching on fast. Randy Nesse has been fighting the good fight for a long time and Evolutionary Medicine is still not part of core medical training.
AA: And yet that’s changing. The new version of MCAT has evolutionary questions. Certain medical schools are teaching cancer and antibiotic resistance using an explicitly evolutionary approach. It could be a matter of bringing in some basic principles of evolution that are more explicit, such as life history theory.
DSW: One thing you say in that article is that not only are most medical articles not from an evolutionary perspective, but they’re not from any theoretical perspective at all.
AA: We did an in-depth analysis of the 20 or 30 most recent articles that were published on the topic of therapeutic resistance. We looked for whether they took an evolutionary or some other perspective–we basically coded them for whatever explanation they were giving for the results that they found. We found that about 10% used an explicitly evolutionary explanation somewhere in the paper. But most of the papers either gave an explanation such as differential patient sensitivity that essentially just restated the results, or they didn’t try to explain the results at all. I and my coauthors on the paper were dismayed — how is it possible that most papers don’t try to put their results into any framework? Most of the papers are not comparing their results to any theory at all. Then we thought about it more and realized it was actually a good thing. It’s not as if there are competing paradigms for explaining why resistance occurs. It’s just that the tools and the methods and the theory that evolution provides have not permeated enough for the researchers to be aware that they could be doing more.
DSW: There must have been something guiding the expectations of these authors. What’s guiding their expectations?
AA: It’s possible that they’re taking some implicit evolutionary approach. Whether that guided them I can’t say for sure, but with a more explicit approach researchers would realize the greater leverage they would get using the tools and methods that are available.
DSW: Well, this has been great. Such wonderful work! It’s been a pleasure to see it developing thanks to a small number of people with ecological and evolutionary training.
AA: We have a great group. It’s exciting for all of us. There are so many opportunities and ways that we can do something useful with our evolutionary and ecological training.
DSW: The idea that something as scientific and biological as medicine might not be evolutionary has to be explained to a lot of people. It’s wonderful to get this snapshot of cancer research from an evolutionary perspective and congratulations for all the great successes you’ve been having.
AA: Thank you so much. It’s been a pleasure to talk to you!
Special issue of the Royal Society Philosophical Transactions B on “Cancer across life: Peto’s paradox and the promise of comparative oncology”.
Aktipis, C. A., Boddy, A. M., Gatenby, R. a, Brown, J. S., & Maley, C. C. (2013). Life history trade-offs in cancer evolution. Nature Reviews. Cancer, 13(12), 883–92. http://doi.org/10.1038/nrc3606
Aktipis, C. A., Kwan, V. S. Y., Johnson, K. a, Neuberg, S. L., & Maley, C. C. (2011). Overlooking evolution: a systematic analysis of cancer relapse and therapeutic resistance research. PloS One, 6(11), e26100. http://doi.org/10.1371/journal.pone.0026100
Aktipis, C. A., & Nesse, R. M. (2013). Evolutionary foundations for cancer biology. Evolutionary Applications, 6(1), 144–159. http://doi.org/10.1111/eva.12034
Jansen, G., Gatenby, R., & Aktipis, C. A. (2015). Opinion: Control vs. eradication: Applying infectious disease treatment strategies to cancer: Fig. 1. Proceedings of the National Academy of Sciences, 112(4), 937–938. http://doi.org/10.1073/pnas.1420297111