The Big Bang Theory Of Cancer
Despite tremendous scientific advances, there’s still so much scientists don’t understand about cancer. One of the biggest remaining questions is how do tumors form in the first place?
Researchers are getting closer to an answer. For years, the prevailing theory of tumor growth was that cancer cells gradually acquire a series of mutations that enable them to outcompete healthy cells and run amok.
But improved genetic sequencing of cancers is revealing a more complicated picture. New technology has enabled a new theory of tumor development, called the big bang theory. It turns out that some types of cancer contain a whole hodge-podge of mutations right from the very beginning, even before the tumors are detectable on a scan. Researchers initially observed this pattern in colon cancer, and then replicated the findings in pancreatic, liver, and stomach cancers, too.
Guest host Roxanne Khamsi talks to Christina Curtis, associate professor of medicine and genetics at Stanford University’s School of Medicine about her research into tumor development, and how to improve cancer diagnosis and treatment.
Christina Curtis is an associate professor of Medicine and Genetics at Stanford University of Medicine in Stanford, California.
ROXANNE KHAMSI: This is Science Friday. I’m Roxanne Khamsi in this week for Ira Flatow. Cancer touches many of our lives. But there’s still so much we don’t understand about it. One big question, how do tumors form in the first place? Researchers are getting closer to an answer. For years the prevailing theory of tumor growth was that cancer cells kind of gradually acquire a series of mutations that enable them to somehow outcompete our healthy cells and run amok. But improved genetic sequencing of cancers is revealing a much more complicated picture.
From this technology comes a new theory of tumor development called the Big Bang Theory. It turns out that some types of cancers contain a whole hodgepodge of mutations right from the very beginning, even before the tumors are detectable on a scan. To help us better grasp what this means for our understanding of cancer and how to treat it is Christina Curtis. She’s an associate professor of Medicine and Genetics at Stanford University’s School of Medicine. Welcome to Science Friday. Thanks for being here.
CHRISTINA CURTIS: It’s my pleasure. Thanks for having me.
ROXANNE KHAMSI: So what’s the prevailing model of how scientists understand tumor growth in the spread of cancer? How have we understood it in the past?
CHRISTINA CURTIS: So really the prevailing view is that cancer growth invariably follows the principles of evolution that Charles Darwin outlined for us in his theory of natural selection. And that theory was that organisms with favorable traits are more likely to reproduce, pass their traits on to the next generation.
And the way that was translated in the cancer field is that cancer cells that arise from our normal cells and accrue mutations can occasionally acquire a mutation that confers a really strong advantage for that cell to grow, and such that every time one of these beneficial mutations comes along, they will take over the population in effectively a linear or a stepwise fashion, gradually increasing the sort of aggressiveness of a tumor, its fitness. And so this notion was that it’s very much stepwise, sequential, over time.
ROXANNE KHAMSI: So how is the Big Bang model different from that?
CHRISTINA CURTIS: Yes, so we were able to observe just this extensive diversity of mutations in cancers that suggested there wasn’t one dominant event that took place that overtook or outcompeted all the other cells. And this led us to go back and really formulate an alternate model. And in this model we use the term “Big Bang” because we postulate that early on the nascent tumor acquires a full house of mutations that initiate that growth.
And once it has those, it can expand rapidly, creating a wealth of additional diversity, much of which are mutations that are passengers. They’re non-consequential. They don’t actually fuel the growth. They’re just arising through the course of cell division. And so this really posits that there’s a tipping point at which this cancer can expand rapidly, that it’s not gradual or sequential, that much of the action happens really early on.
ROXANNE KHAMSI: So does this play in at all to why you called it the Big Bang Theory? Can you explain some of the parallels it has with this Big Bang Theory of the creation of the entire universe?
CHRISTINA CURTIS: So we use that model because what we observed really implies that we could go back and trace the origins of a cancer from what we detect in the clinic. So really we don’t know how the tumor became what it was. But the fact that we observe these very specific mutational patterns when we sequenced tumor genomes suggested that this was really a reflection of how the tumor began.
And so this parallels the notion that the cosmic microwave background is really a signature of the origins of our universe. And in both cases, these observations about the end product allowed us to reveal the origins of the universe and the origins of a tumor.
ROXANNE KHAMSI: A big part of your research is figuring out what happens before the cancer is even big enough to detect. But there’s a catch-22 there. How did you use the samples from fully formed tumors to then go back retroactively and kind of chart their growth and find that Big Bang of tumor mutations?
CHRISTINA CURTIS: So we took, in this case, patients that had colon cancer. And we used sequencing technologies that allow us to peer in and read out the genome sequence of our tumors. And in this manner we really were able to do this at a very high resolution using some advances in technology and assemble the patterns of diversity that we had observed.
And what it really took then actually was computational approaches and some theory from other fields, from fields of population genetics, where we studied the relationship between species to use that to trace back their ancestry. And you can kind of think of it that much as we can understand patterns of growth over time in a tree, if we take a cross-sectional view, the concentric rings of the circle tell us about how that grew, we’re able to look back in time and use the patterns of mutation to tell us about the origins of the tumor.
ROXANNE KHAMSI: And how many mutations are we talking about in the colon cancers that you looked at? Was it like 20 mutations after the Big Bang of mutations? Or 100? Or 2000?
CHRISTINA CURTIS: Many mutations, many mutations. But I want to stress that a lot of those mutations that we observe– in fact mutations are happening all the time in our cells. And so these mutations are accruing. Our cells have many ways to cope with them. Most of them are inconsequential. They don’t influence how the tumor will grow.
And so there’s many, many mutations. There’s an extensive diversity to the point that in fact, every cell within a tumor may be genetically distinct. And that poses real challenges for how we might treat patients and poses risks to the development of resistance to our very best therapies.
ROXANNE KHAMSI: And I know you looked at colon cancer, but are there other types of cancers that follow this Big Bang model, like for example in the liver or stomach? And do we know why some types of cancers follow the Big Bang model?
CHRISTINA CURTIS: Yeah, so there’s been some surveys to really try to understand how different tumors evolve. And I would say this has sparked, really reinvigorated the field and how we think about this. And you’re correct. There’s a whole host of tumors of the gastrointestinal tract.
You can think of liver, not only colon, the stomach as well as even pancreas, that tend to follow this model where that growth is rapid expansion, is rapid and the diversity is vast. Now why that’s the case, particularly in gastrointestinal cancers, is still unknown. But we would speculate that it may have to do with the organ structure and the architecture within that is innate to our different organs.
ROXANNE KHAMSI: That’s so interesting to me that you’re finding that this is the case for some cancers, but that not all cancers have this Big Bang phenomenon operating in them. It also makes me wonder, do we know what causes a cancerous cell mutation in the first place?
CHRISTINA CURTIS: So mutations accrue at random. That happens throughout the lifespan. And of course we have many trillions of cells in our bodies. So we are constantly coping with these insults. Now some of these mutations affect proteins that carry out critical functions in our cells. And sometimes those mutations can cause a growth advantage for a cell. And that is really what we think of as the key drivers of tumor growth.
Now we believe, and much work has suggested, that there’s probably a limited set of events that actually initiate a cancer, that form a cancer. They differ by our tissue types, our organ types. And yet there’s not so many of them. But a cell must have accrued a particular constellation of these events.
It’s very seldom that a single event is going to cause a cancer. So they clearly have a genetic origin. And there seems to be some preference in different tissue types for particular mutations, meaning that they promote growth in one environment but not in another.
ROXANNE KHAMSI: I’m wondering as you’re talking, once a tumor is fully formed, does it still continue to evolve over time?
CHRISTINA CURTIS: Absolutely. So that is one of the greatest challenges that we really face is that tumors are constantly evolving. They are not static. This is a highly dynamic process. And what that means is that given this diversity, given the vast array of mutations that a tumor can have, that we need to anticipate whether or not any of those mutations can confer or allow for resistance to therapy.
Because we’re really dealing with a very large population. So there’s a huge amount of variation to be acted upon by evolution. And that is a key challenge for the field.
ROXANNE KHAMSI: Knowing about this Big Bang Theory, how does that help scientists point doctors towards better cancer treatments?
CHRISTINA CURTIS: So while those mutations that we’ve talked about may not fuel the growth of the tumor, they may enable resistance to a particular therapy. And there’s such diversity that there’s really this huge reservoir of ways in which a tumor can evade therapy.
And so that’s a problem. We need to really anticipate resistance. And we need to harness the evolution therapeutically, meaning can we exploit this variation to actually impede cancer growth? And that’s really a very active area of ongoing research.
ROXANNE KHAMSI: So how would you do that? So you’re saying you could leverage the fact that there’s this Big Bang of mutations to somehow undermine the way the cancer can grow?
CHRISTINA CURTIS: A particular mutation may confer resistance to one drug but possibly sensitivity to another, or that there’s a tradeoff in the fitness of that cell that could be exploited. And so we need to clearly do much more research to know the many ways we can exploit this. Because each patient’s tumor is distinct. But arguably our best tools to tackle this ongoing evolution is the evolution itself and to design evolutionary-grounded therapies.
ROXANNE KHAMSI: So it sounds like there’s a really complex conundrum there for doctors to pick the right treatment, but that somehow looking at all the mutations might guide us towards these things where we can be more nuanced about how to treat the tumor. One thing I did wonder is about metastasis, so the idea of a tumor that’s kind of able to spread in the body. Does the Big Bang Theory inform our understanding of how cancers spread in the body?
CHRISTINA CURTIS: We’ve been able to do this on sort of an initial basis. And what we’re finding is really quite profound. And for example, when we compare these patterns of diversity between the originating tumor and its metastasis, which is really the cause of patient death, what we tend to find is that there’s really strong evidence that these tumors can leave home early. And by leave home, I mean leave the primary organ site and disseminate and colonize another organ.
And so that comes back to really the early origins of tumor formation and the notion that, in fact, what we had posited alongside this theory was that some tumors may be born to be bad, that their aggressive potential is specified really early, that it’s dictated by these early mutations.
And that has been borne out in a number of studies, both from our work and from others. And of course that gives us quite a bit of pause because, again, it really highlights the fact that this process is not necessarily gradual as it has long been assumed. And it places considerable emphasis on earlier detection.
ROXANNE KHAMSI: I’m Roxanne Khamsi. And this is Science Friday from WNYC Studios.
So it sounds like metastasis is something that could happen really early in a cancer development, which does point towards the value of early detection. But as you kind of hinted to earlier, there’s different kinds of cancers that may or may not follow this Big Bang model. Some of the mutated cells, as you mentioned, might be born to be bad. And you’ve got some current research on breast cancer that has some promising new insights into why some types of breast cancers return after let’s say, 5 or 10 or even 20 years of remission. Can you tell us a little bit about what you found there?
CHRISTINA CURTIS: So really using very different approaches, what we observed is that there are specific genomic differences amongst breast cancer patients, some of which had not been previously appreciated. And so really we can classify breast cancer into some 11 subgroups. So that’s a lot. But these groups really have distinct mutational landscapes, distinct features.
And remarkably some of these genomic alterations allow us to predict which patients are likely to relapse from their disease, and in some cases up to two decades after that initial diagnosis. So really again, as we saw in these earlier studies in colon cancer and has been extended to other tumor types, specific genomic alterations can provide a wealth of information about what that tumor’s trajectory may be and allow us to forecast its next steps.
ROXANNE KHAMSI: So you’re doing a lot of looking back at tumor histories with genetic sequences and also kind of using that insight into the mutations to look forward and maybe give us a better sense of what’s to come.
CHRISTINA CURTIS: That’s right. We think that if we can understand the origins of a cancer, that it really provides fundamental clues as to how to better detect those cancers, to intercept, and then to treat them in a far more personalized manner. And so it really is quite nuanced given the heterogeneity.
But equally there are some core principles here that I think really give us some hope that we might be able to understand what those alterations, those mutations are, and to go after them. So there is some silver lining there in terms of how we can use this information to improve patient outcomes.
ROXANNE KHAMSI: Well, that’s a great positive note. And I think that looking forward in that way sounds fantastic. I’d like to thank my guest Christina Curtis, associate professor of Medicine and Genetics at Stanford University’s School of Medicine. Thanks so much, Christina, for joining today.
CHRISTINA CURTIS: Thank you.
John Dankosky works with the radio team to create our weekly show, and is helping to build our State of Science Reporting Network. He’s also been a long-time guest host on Science Friday. He and his wife have four cats, thousands of bees, and a yoga studio in the sleepy Northwest hills of Connecticut.