Cancer Isn’t a Logic Problem
cancer.nautil.us
cancer.nautil.us
> Andy Grove has had prostate cancer, and now suffers from Parkinson’s, so it’s no wonder that he’s taken aback at how poorly we understand each of those diseases – not to mention all the rest of them. But his experience in the technology world has warped his worldview. We are not suffering from a lack of urgency over here – talk to anyone who’s working for a small company shoveling its cash into the furnace quarter by quarter, or for a large one watching its most lucrative patents inexorably melt away. And we don’t suffer from a lack of hard-charging modern management techniques, that’s for sure.
What we suffer from is working on some of the hardest scientific problems in the history of the species.
Not dying of the first cancer is better than dying of the first.
Just because there are N cancers in the series doesn't mean the problem is impossible, it just means we haven't ever gotten past the first levels to figure out how it works.
The secret to winning pacman isn't 256 separate strategies, it's a small set of strategies that fit together once you know how the game changes level to level.
The author is just fixated on the fact they finished the first level AND THE WAS ANOTHER!
This is the most important part. Life is cancerous. When people claim to cure cancer it's sort of an oxymoron. There are indeed no such things as cancer genes but more like a breakdown in the cooperative biochemical networks that make multi-cellular life possible. It's actually a great wonder how all those cells in your body manage to cooperate at all without stepping on each others' toes all the time.
In a way the delicate balance is a constant fight against entropy and entropy always wins. Enough breakdowns and misunderstandings in the biochemical networks and the multi-cellular ecology and you have cancer.
This description fits scurvy fairly precisely. But no one would say it's nonsense to talk about "curing" scurvy. Give someone some vitamin C, and their scurvy will be cured.
Would you also say it's an oxymoron to talk about "fixing" software bugs because the terminology assumed they were designed in?
I characterize scurvy as a sub-optimal stable operating state where optimality can be restored. Cancer does not have similar characterization. It's more like an unstable runaway loop.
But I'm neither a biologist nor a doctor. My understanding is basically a bunch of analogies between game theory and general systems in the context of biology and maybe that's not the proper set of analogies for the problem at hand. In that context cancer is the breakdown of feedback loops that keep the system stable.
This makes no sense. It's not a stable operating state. Vitamin C is necessary to hold your cells together. Scurvy is the failure path of not having vitamin C; it is a transition from "healthy" to "dead".
> cancer is the breakdown of feedback loops that keep the system stable
OK, but this is exactly what scurvy is. They're just different feedback loops. Infection by foreign bodies is also "the breakdown of feedback loops that keep the system stable". Your objection applies to everything, which makes it completely meaningless when you apply it to cancer specifically.
It's a little frustrating honestly. Don't people see that most things seem impossible until they aren't? Just two years ago a lot of people were quite confident that computers would never surpass humans at Go. https://www.reddit.com/r/baduk/comments/2wgukb/why_do_people... This is one of dozens of such links you could find from that period. Much moreso ten years ago. I get that cancer is a horse of a different color, but come on. If you are confident that ANYTHING will be impossible in 200 years, besides things which are physically so . . . well, I just cannot agree.
Cell division is an unstable positive loop that requires dampening with negative feedback loops at higher organization levels. When those negative feedback loops break down you get an unstable positive loop in the form of cancer.
So you see how having cell division in a multi-cellular context can not be "cured" of cancer? You'd need to stop cell division and somehow figure out a way around entropy. Cells divide to renew/reset and shed accumulated environmental damage. So no matter what you do as long as you have a system organized around unstable processes like this you will always have a control problem when the negative control mechanisms stop working.
What about animal species that get cancer much more rarely, or never? Are naked mole rats an oxymoron?
The universe is logical. Each cell in the human body is at its core just a very complex state machine.
What we call cancer is just a particular state for a cell, and its descendents.
Find a way to change the cell state, we end cancer. And we know we can change a cell state-we can transform skin cells into stem cells, so what would make it impossible to turn cancerous cells into non-cancerous cells?
We just haven't found out the proper way to do so. It takes data, and simulation, and lots of time. At some point we'll figure it out.
I think the only misguided attempts are chemotherapy-based - damaging cells indiscriminately, just because cancer cells multiply faster and thus accumulate dna damage faster and thus die faster than normal cells. This means that unless you kill all the cancerous cells your cancer will come back. The problem is chemotherapy drugs will also be damaging your body in other ways, depressing your immune system, and thus making it easier for any remaining cancerous cell to start it all over again.
I believe it's much easier to simply figure out the right cocktail that would change cancerous cells into normal cells again, and then find the delivery mechanism for that.
I can see a future where we can transform cancerous cells into cancer-killing cells (let's say T/B Cells) that would just cascade into the complete eradication of cancerous cells within an organism.
This guy's argument boils down to lack of imagination.
I think for people with cancer right now, waiting for magic would be a lot more misguided than chemotherapy.
"Cancer cells are not simply a disorder or breakdown in a mechanism, but an organism going on a full-tilt offensive, using multiple, often shifting strategies to produce and use molecular fuel, win resources, and evade the immune system. If so, then the rules of the game may change—these insights suggest that the war on cancer may be endless."
I don't have any expertise here, so I'm not saying he's right or wrong. But it doesn't seem to be a lack of imagination.
any person's cancer is millions/billions of semi-independent, extremely complicated, poorly understood state machines, with a variety of mutations, constantly interacting and evolving against any treatments applied and immune reactions.
imho, cancer is a fundamental disease/property of multicellular life, and trivial reductionist perspectives completely out of touch.
This is of course ridiculously ambitious compared to the current state of the art, but I think it's a reasonable start to countering the claim that a general cure for cancer is impossible in principle.
That said, I don't want to say that cancer is fundamentally beyond us. I think cheap sequencing and emerging highly targeted DNA editing systems, along with better understanding of the immune response to cancer, will get us most of the way there in a practical sense.
What you suggest is more or less what I mention; you could, say, try to add back some genetic machinery for the cancer cells to recognize that they're defective, and then they kill themselves. Or help the higher level repair mechanism (the immune system) target cancer cells.
To my understanding, trying to fix anything at the level between whole cell and DNA mutation is much harder.
I like analogizing between multicellular organisms and nation-states, in that both are complex systems built upon the coordination of many individual systems (cells/people). In the case of cancer, what we have is essentially rebellion, a failure of control: individual cells decide to stop cooperating and instead begin competing with the body that houses them. So I agree with you... rebellion is always a threat to any system of control.
Unlike actual rebellion, though, there's no way the cancer can "win" in the sense of taking over control of the body. If only cancer could realize the futility of its crusade...
It's like arguing that entropy always increases and therefore fighting disease is fundamentally hopeless, or requires non-logical approaches. Yes, of course "ill health" can't be cured in general, but we can drive disease rates ever lower.
Sure, but no one disputes this. Silicon valley investing in cancer research does not imply anyone takes such a silly position.
So the universe is just math, basically? Or maybe a computer program?
How so?
>at its core just a very complex state machine
Why do you say that?
Agreed that we don't yet know how to regulate the apoptosis, or understand how metastasis works. DNA Methylation, small RNA regulation, and purpose of non-coding RNA isn't very well understood yet, too.
> I believe it's much easier to simply figure out the right cocktail that would change cancerous cells into normal cells again, and then find the delivery mechanism for that.
I reckon much of this has to do with small RNA and miRNA based regulation of transcription. Essentially the specific portions of the RNA/DNA are "silenced" by several regulatory factors. If a growth regulator gene, say CDKNA is accidentally silenced then the cell would undergo an uncontrolled growth, _with_ mutations. This in turn will grow "diseased" cells, eventually causing cancer. Trouble is, we just do not understand this properly, yet.
I have little doubt cancer will be someday as well. It might not -- probably won't -- happen in my lifetime. But if humanity hasn't collapsed, I find it very unlikely that we're still dealing with cancer in a thousand years.
E.g. we think of stars as big balls of fusing matter. But that's an idealisation. Is a star really less complex than a human body? It all depends on which questions we want to answer. So there's an asymmetry in this argument.
Take a thing that we're trying to understand down to the finest detail, and compare it to a caricature of some other thing that we're not trying to understand at such a fine resolution.
If you simply gather a very large number of hydrogen atoms into a region of space at a certain density, you will create star. From this angle, it isn't terribly complex.
The instructions for creating a cell from scratch are... immensely more complicated.
The star's behavior, .e.g. the movement and changes of the convection zones, fusion dynamics, how the magnetic fields change over time, coronal mass ejections, sunspots, etc. may be extremely complex. I think few people who know much about stars would disagree on that.
But I would wager there are many more orders of magnitude of complexity going on in a human cell, from an information theory perspective. Even just describing individual proteins themselves and how they fold is phenomenally difficult.
Of course, I may be wrong about the cell's behavior having higher complexity. The important take away is that it is possible to make such comparisons in a meaningful way.
So, we're back to comparing thing A that we need to understand at the finest scale, to thing B where we ignore its individual complexity in favour of the stereotypical version. It's easy to estimate the gravitational attraction or mass of a cell. It's this fine-scale manipulation that causes these severe requirements for deep understanding.
EDIT> :) The instructions for creating life are even simpler. 1) Have a Big Bang. 2) Wait.
Even if the information theory metric of complexity is somewhat flawed, given our limitations, I think it is a useful tool. I'm not sure what would be a better way of comparing complexity.
At the very least, I think I can make statements such as: "a closed container of hydrogen gas at room temperature and pressure is a much simpler system than a Swiss watch", which we could roughly quantify in a somewhat robust way. The hydrogen part is easy, as the atoms are indistinguishable, and their movements can be closely approximated with simple formulas.
P.S. I just remembered a good (and short) minutephysics video on entropy & complexity, perhaps it is worth linking: https://youtu.be/MTFY0H4EZx4
Also: this is a really great, really stellar thread.
More concretely, the combinatorial complexity of the state-machine approach makes it infeasible.
Even if determinism holds (and, spoiler alert from about 1927, it doesn't), you can't avoid chaos. Lorenz found that out and defined Chaos: where the present determines the future, but the approximate present does not approximately determine the future.
I assure you, biological systems exhibit all the hallmarks of (possibly) deterministic mathematical chaos.
Its understandably very hard problem to nudge network states back to the normal or to kill itself once it's gone haywire and causes other networks go haywire too. The "error state" cells are changing and those who survive the last nudge continue to reproduce.
If I understand Kozubek's argument correctly, cancer is ecologic, not systemic because of this evolution of cancer cell populations. If the cancer is ecologic problem, solving cancer means that it's not enough to figure out cell's state and fix it. You must figure out how the cancer cell ecology works and find a way to drive it into extinction without survivals. Just setting more mousetraps is not working solution if the mouse population is evolving hour by hour and branching.
I think a common trope is to point out some system that humans are currently unable to model in a robust or effective manner, and then throw up one's hands and declare "science and logic can't be used to solve this problem!" The current lack of an effective model for some system in no way implies that one will not eventually be found.
To be fair, this article isn't quite doing that, and makes some very important points. I only bring this up as it is a misleading line of argument to watch out for that I run into from time to time, and a few lines in the article were slightly reminiscent of such reasoning.
Ecology is a scientific endeavor, and perhaps it will prove to be a more powerful paradigm for looking at a cancer. I hope we can make the most of any tools at our disposal to significantly reduce the harm caused by cancer.
That's not what the author was saying.
It's useless to judge articles based on their title or headline because they are either selected by editor or there are multiple titles and A/B testing is used to find one that gets most clicks.
It seems that most comments are just comments based on the headline.
I will say, however, that only responding to my first sentence is sort of like only discussing the headline :)
we need to consider that the system of cancer goes far beyond the biochemistry of cancer cells. Cancer relates to our fundamental constitution as multicellular organisms, our limited lifespan, the epidemiology of the aging population, socioeconomics and the future of society. Those who believe that the problem of cancer can be solved by killing or reprogramming cancer cells need to take a step back from the molecular technicalities and take a look at the bigger picture.
Let's take a step back ... my understanding of what the author's saying is that "solving" cancer is no easier in some sense than "solving" biology.
Yes, we have to think of cancer as some kind of adversarial game -- it's not enough to solve a snap-shot of that game. But ... that doesn't mean that we can't have active defences that work for the vast majority of people. It's just that maybe we need to know how to build people from molecules, first.
I think we'll solve biology in the sense of being able to engineer life from first-principles. I make no predictions on time-frame.
There’s 3.2Bn genetic base pairs.
There’s just over 7Bn people.
Maybe like 7M-20M people suffer from a disease or even well below or above that.
Finding someone at a given time in the exact modality that would be testable by a given trial is really hard.
Genes are sets of base pairs.
A lot of drug discovery is a situation where you have to test compounds empirically against a notion of a theory of a viable drug target.
Some drugs may be efficacious, but they may kill you.
Some genes may work in coalition with other genes, not at all, or against some genes.
There's also the fundamental notion about the druggability of a disease. This article does a great job in outlining it: http://omicsomics.blogspot.com/2017/03/targets-drugability-r...
There are no monstrously hard computational power problems in drug development that are waiting for mathematical brains to pursue. There are hard Empirical problems. Empirical problems are problems which require actual experiments to show resultant data.
Not sure if I would put it that way. For ex, in principle, you could use molecular dynamics and search to find protein sequences that form other proteins/antibodies that can hit targets you want or operate in a more complex fashion. So hard computational power can most definitely help out (and depending on level of advance, help out a great deal). Ofc that doesn't mean you don't need experiments at all.
Another example of why this sort of thinking is missing the bigger picture - many people today die not because we don't know how to treat something, but instead because they can't afford the treatment.
A) cancer cells constitute a separate species, or at least evolutionary "actor".
B) the tech naivity is not that things can't be deeply understood, but that it's a moving target, and less than complete erradication of the parasite just increases evolutionary pressure leading to decreased understanding of the foe.
Some, like Aristotle, did try to do that. To be fair, it was a mix of observation and first principles, but it was lacking adequate experimentation, as humans later found out.
Fair enough. I'm probably giving too much of my own spin to the word "logic".
I'm thinking of logic as being somewhat grounded in empirical observations. I agree that armchair logic completely removed from reality is of questionable value.
At this stage of my understanding, I would argue that our notion of logic and math largely stems from us creating models about the world that lead to abstract deductions, which often are eventually used in new models about the world... and so on. This cycle probably got started as early humans acquired better mental models for categorizing objects of different types and their apparent boundaries, and then began to "count" like entities.
If math wasn't useful for describing the world, I doubt we would have done much with it.
But there are hundreds of types of cancer, and taking the genetics/drug discovery approach for each of them is its own distinct and massively expensive undertaking. It'll never get done: the research community and its funding sources are not large enough to make significant progress on the whole of the problem in the next few decades. This isn't hacking cancer, this is hacking the cancer funding institutions in order to obtain funding to get on with the fundamental life science research goal of mapping all of cellular metabolism. Cancer is the excuse, but not the goal.
This is a cultural problem.
Hacking cancer would be to sit down and say, ok, the economics of this are not working, we need to find common points to target, a way to produce a universal cancer therapy at no greater expense than one of these single cancer therapies.
Some immunotherapies, like CAR-T approaches, are a small step in this direction, something that can be applied to multiple cancers with a lower cost of customization per cancer type.
The best target, however, is to block telomere lengthening; disabling both telomerase and alternative lengthening of telomere (ALT) mechanisms. This is a small area of biochemistry in comparison to the scope of most single cancer explorations. All cancers depend on telomerase or ALT or both. All of them. No exceptions. No cancer can evolve its way around a suppression of these mechanisms. It will simply die, losing its ability to replicate uncontrollably. Some noted research groups are making early inroads into disabling telomerase in a targeted way. ALT remains to be dealt with, but is easier to work with than telomerase interdiction by virtue of the fact that ALT doesn't happen in normal cells.
A telomere lengthening interdiction treatment would be applicable to all cancers, and it doesn't appear to be more expensive to develop than any other approach to a specific cancer. That is what hacking cancer research looks like; identifying a way to completely change the economics of the situation, and redirect the primary effort back to producing an effective cures as soon as possible.
Lastly, I have to fundamentally disagree with the final lines of this article; cancer can absolutely be dealt with robustly by killing cancer cells. You just need to be selective enough, early enough, and comprehensive enough, and raising the bar on these items is the whole point of the field.