Can we cure cancer?

In my previous post I paraphrased my understanding of why the cancer workshop was called as the premise: “cancer is an evolutionary process; the cure for cancer is within reach, and is mostly an engineering problem now that we have the right model; and what can we do collectively to work towards and achieve the goal of a cure”. Here’s the current scorecard in my mind:

“The cure for cancer is within reach” –>

  1. The nature of “cancer” is more complex than I originally imagined.
  2. Cancer is both real and fictional at the same time, just like every other model.*
  3. To make any progress on the fundamental understanding of cancer, we need to admit that whatever definition we use of the concept of “cancer” — and there are many proposed and competing ones — proscribes/limits the set of all possible models that you have to work with.
  4. Ultimately in order to make progress on what you thought was the goal (e.g. “cure cancer”), you have to be willing to change the definition (of “cancer” itself).
  5. But changing the definition, then changes the goal because the “underlying reality” you thought you were studying is ultimately just the model you created in the first place.
  6. For instance, if “cancer” turns out to be a normal consequence of life, then perhaps what we are really after is extending life by looking at cancer as — and treating it as if it is — a chronic condition which flares up from time to time.
  7. Ultimately the new models you use to understand the nature of cancer and what it would practically mean to “cure” it, are just new models and will inevitably be show to be wrong at some point.
  8. The new models we create will undoubtedly have explanatory power (by which I mean they are as true as anything is) in other domains — for instance we, we may find “cultural cancers” just as we find “computer viruses”.

“Cancer is an evolutionary process” –>

  1. Yes, but it is also a process that involves other dynamics like metastasis, self-organized criticality, ecology dynamics, etc.; “evolution”** alone cannot sufficiently describe cancer to give us the understanding we seek.
  2. And these are all just models, imperfect to the bone. Any model that adds predictive power to the current best hypothesis is worth exploring.

“Curing cancer is mostly an engineering problem now that we have the right model” –>

  1. We don’t have the “right” model and never will; we only have an ever more refined model as we learn.
  2. Do we have a refined enough model to cure cancer? No, because our conceptions of “cancer” and “cure” have shifted and will continue to shift as we learn more.
  3. Can we re-frame the question of curing cancer and be more precise about what we mean? Yes, for instance we can re-frame our notion of “cure” to be “extending life by detecting and halting tissue-level metastatic process in tumors indefinitely”.
  4. But something similar to cancer can crop up at a different level of organization, so be careful about using “indefinitely” and confining yourself to the tissue level.
  5. Can we say we don’t care if we die because we didn’t understand cancer enough to stop those same cancer-like behaviors at other levels? Yes, but we still end up dead.
  6. Can we say that we don’t care if we die “eventually” as long as we’ve staved off death from well-understood processes? Yes, as long as we are happy with the improvement in life-span and don’t mind dying of other causes.
  7. In other words, if we say we want to increase life-expectancy by a specific length of time (say 15 years) by controlling the class of of maladies commonly referred to as cancer, then I believe it’s an engineering problem (but a very large one).

“What can we do collectively to work towards and achieve the goal of a cure” –>
With a re-framing of cancer and a re-framing of cure, we can do the following in parallel:

  1. Work on engineering related to achieving the re-framed goal.
  2. Continue the “meta-science” of cancer, by which I mean engage in a process of continually challenging our assumptions (no matter how basic) about all levels of organization, including levels above the human body (such as mind, society, computation) and levels below the level of DNA (such as molecules, atoms, subatomic particles, etc).
  3. We need to develop new intuitions about how to think about complex adaptive systems (which is to say, everything) because our intuitions are what drive and constrain the models we build and use to understand the world.


* Which is to say, everything we can ever know as humans about reality is based on models, which are by definition convenient fictions. A cheeky way I heard of saying this was “all models are wrong, but some are more wrong than others”.

** Most people think of evolution as Darwin suggested, roughly, selection on populations of creatures with heritable traits. But you get more predictive power when you add an evolving environment (as in “co-evolution”), punctuation (as in “punctuated equilibrium”), meta-stability (as in a more dynamic notion of equilibrium), phase shifts, strange attractors, etc, etc, etc.

  • Bruce Bloom

    I wonder if “Cancer IS an Evolutionary Process?” I haven’t found good evidence that cancer itself is evolutionary at the whole human being level based on the “survival of the fittest” definition, although it may be evolutionary at the level of individual cells within a human being.

    Cancer seems to be caused by genes that are damaged or otherwise changed by outside factors (radiation ingested or inhaled harmful molecules, virus infections, etc.) or internal factors, like damage during cell reproduction. These damaged genes cause changes in the cells-usually the cells are destroyed because they no longer function properly. Some damaged genes do the opposite, endowing the cells with immortality, rampant reproduction and ability to travel.

    Left unchecked, cancer usually causes death of the whole human being, which then causes the death of cancer cells (although there are many reports of spontaneous remissions without treatment.) Cancer therefore seems to arise in spite of evolution at the overall human level, which should select out humans with genes that are more likely to suffer damage from these kinds of ubiquitous threats.

    Cancer generally, although not always, strikes beyond the age of reproduction, would negate the selection of human survival based on passing on the genetic material from humans who have cancer to the next generation. That passing of genes happens before the cancer kills people. Does that mean cancer is not evolutionary because it does not have the same kind of effect on survival of the fittest? I guess it depends on the level of evaluation (cellular, individual human, overall population.)

    After reading the same book as you, and attending the same conference, I am wondering whether “we often fail to cure cancer because some cancer cells within a patient EVOLVE ways to avoid the therapeutic effect of existing treatments.” By “evolve ways to avoid” I am not suggesting that cancer cells have a conscious desire or ability to change to beat the therapy. I am saying that through the many cycles of reproduction of these already unstable cancer cells some invariably continue to change in ways that by chance allow them to avoid the therapy. This seems clearer to me as an explanation of why therapies don’t work in the majority of patients with most types of cancer, but I have not yet seen the kind of scientific proof that would support it. It makes logical sense as a hypothesis.

  • Bruce, I think I agree with everything you clarify and posit above. I’ll make some further observations along the same lines.

    It is critical to be clear when speaking of “evolutionary processes” the organizational level at which replication/mutation/selection is happening. Most (if not all) of the time we are talking about cancer, we mean evolution of a population of cells within a particular individual leading to tumor formation and progression. But, just as cell populations adapt under selective pressure, so do populations of humans. And it logically follows that any population of “things” that can replicate with mutation under selective pressure, also evolve in the Darwinian sense. One of the Immunologists at the conference explained to me how the human immune system works, and there is clearly evolution going on within the macrophage population within each of us. It just so happens that in this case cell evolution is working to protect the body from harm, whereas in cancer, it works against the human body’s survival.

    But cells and humans (or animals) are not the only populations of “things” which could be evolving. Dawkins famously argued that our collective “genes” are such a population. It is a confounding factor that there is such a tight linkage between gene pool (genotype) and human pool (phenotype) because it causes us to argue over red herrings like primacy: we assume that only one population is *really* evolving (genes or humans) and the other must be somehow just “epiphenomena”. In reality, both are co-evolving, just like cultural memes have been co-evolving with humans for some tens of thousands of years now (see Durham’s “Coevolution”).

    If cancer is an evolutionary process, it behooves us to ask what other populations evolve within the body (and even outside the body), since we cannot be naive enough to think that the process of cancer must be unique to populations of cells. For instance, Dr. Henry Heng of Wayne State University has done some seminal work which highlights the distinction between chromosomal and genomic evolution (if I understand his work correctly). While it was previously thought that these populations evolve on a one-for-one basis and are essentially the same exact process, Heng is showing that not to be the case.

    And while it’s generally thought that what kills us is tumor progression (i.e. evolution at the cellular level), what other evolutionary processes are going on in our bodies that materially affect our early demise? If we find a way to halt or slow down cellular evolution and tumor progression, does that ensure we live longer, or will some other evolutionary disease become unmasked that’s just as bad (or worse)? If we can understand the dynamics under which non-cellular populations evolve, could that help us in containing cellular evolution and tumor progression? Are there diseases that we heretofore do not understand to be evolutionary in nature that really are — cancer at a level other than cellular? And if so, by making progress on curing cancer, can we also make progress on these other diseases?