in large part because "cancer" is not really just one disease, but a general term for a variety of related diseases with many different kinds of outcomes and treatment trade-offs.
If someone can find a very general commonality to most cancers, with a treatment for that commonality that is safe and effective (and not too expensive) for human beings around the world, that person will of course win a Nobel Prize in medicine. Here on Hacker News, over the 1704 days that I have been here, I have seen many kind submissions of breathless claims of breakthroughs in cancer research. We all desire that cancer breakthroughs happen. Follow-up, so far, has not suggested that any such cancer breakthrough actually has happened. I wish the researchers well, and I hope to hear later about placebo-controlled multicenter clinical trials that show that this approach is safe and effective for human medicine.
Hum... the article does not talk about anything that could be placebo-controlled multicenter clinically trialled.
It's just about some dudes deriving conclusions from the relationship between embryo development and (what they call) cancer, with a couple of predictions thrown in to make their theory falsifiable. As a press release is not that bad.
"deriving conclusions from the relationship between embryo development and (what they call) cancer"
Except that they repeatedly say "The new theory predicts that as cancer progresses through more and more malignant stages, it will express genes that are more deeply conserved among multicellular organisms, and so are in some sense more ancient". So, just to be clear, they're not deriving conclusions from a relationship between embryonic development and (what they call) cancer - they're deriving a relationship from their conclusions.
Physicists poking their noses into fields they know absolutely nothing about and making grand pronouncements isn't new.
Why this is a 2013 press release I don't know - they basically published on this in 2011 (http://iopscience.iop.org/1478-3975/8/1/015001) though they seem to be no further along now than they were then - which is nowhere, since in 2011 they were basically still saying "We think these are atavistic genes, here is the sort of research agenda that might reveal this..." and then they didn't undertake that research, as far as I can find. They're still just speculating on it.
Okay, you're right about deriving a relationship from their conclusions. At least mostly. There's certainly an interesting set of possibilities to be pursued where the idea is concerned with evolutionary processes and thinking outside the box of seeing cancer as purely a random mutation, and the problematic questions Davies and Lineweaver pose from that standing perspective.
Yes, they are speculating on a potential new direction of study based on being outside observers who specialize in physical systems and recognizing orchestrated, systemic actions.
However, let's dial down the disdain slightly, can we? These aren't physicists arbitrarily and whimsically "poking their noses into fields they know absolutely nothing about and making grand pronouncements". These are physicists who have been asked to poke their noses into something they know absolutely nothing about and help existing specialists uncover potentially new avenues of research. Davies makes this very clear in his opening statements of the article. He admits he knew nothing about cancer, and started asking questions that were not typically the questions asked by standing specialists--who show a pattern of questioning and investigating cancer from that perspective of a random genetic mutation that occurs in humans, and not investigating the possibilities that might be found in questioning cancer as a feature of multi-cellular life as a whole.
Davies and Lineweaver are not saying this is the fundamental theory of what cancer is. They are suggesting it as a possibility of future research and investigation and, given their work in their own fields, I strongly suspect they'll be willing to update their notions based on evidence.
Who knows where we can get if more respected scientists joined cross-disciplinary programs that were aimed at collaborating on investigating and understanding various human concerns. It's not like the physicists who've joined these programs (of which Davies and Lineweaver are but two) are going to set back cancer research and treatment.
"The new theory predicts that as cancer progresses through more and more malignant stages, it will express genes that are more deeply conserved among multicellular organisms, and so are in some sense more ancient."
The good news is that we already have that data! From 2005.
And the story, is, of course that it's waaay more complicated than that.
For that matter, we know a lot of cancers (e.g. CML) come about in no small part due to fusion proteins - which would be if anything, the creation of a novel protein, not a gene that is 'deeply conserved'.
"Cures cancer" might as well be a red flag when reading press releases.
I've recently won a hackathon (http://hack4ac.com) with an idea to cut the middlemen out of the science news cycle with only a single link remaining: from the scientist to the general public. It was built to prevent this sort of misinformation, intentional or not. I believe it usually is unintentional and is just a side product of all of the players in the science news cycle playing their part in the game of 'telephone', and information just gets mangled. The project (ScienceGist) puts scientists into the driver's seat of their own communication with the public.
"Davies and Lineweaver are currently testing this prediction by comparing gene expression data from cancer biopsies with phylogenetic trees going back 1.6 billion years, with the help of Luis Cisneros, a postdoctoral researcher with ASU's Beyond Center."
Press releases for untested scientific conjectures (further nerd-rage inducing by calling it a "theory" in official communication) raises a red flag for me: it suggests that the investigators aren't fully skeptical of whether their conjecture actually represents the real world, and increases the probability that they'll introduce bias in their investigation.
"Could it be, we wonder, that cancer’s predilection for a hypoxic environment reflects the prevailing conditions on Earth at the time when multicellularity first evolved, before the second great oxygenation event?"
"Wondering," alas, is not science! We need a model, a hypothesis, and a well-controlled experiment to actually discern truth.
Religious science types always strike me as odd. There is nothing bad about thinking about things in a new way, and nothing bad about writing a little paper about it - maybe they want to generate excitement, maybe they want to make sure everyone knows they came up with this first, who knows.
But getting an idea out there is in any case a net positive.
Scientific breakthroughs don't start with double blind studies; they start with an idea. To say it with Einstein: Imagination is more important than knowledge.
I only read the link above, but to me (MSc in biology) it totally sounds like a legitimate theory, I can easily see how it makes testable predictions. It may turn out to be wrong, but IMHO it's worth some testing.
"Could it be, we wonder, that cancer’s predilection for a hypoxic environment reflects the prevailing conditions on Earth at the time when multicellularity first evolved, before the second great oxygenation event?"
Could it be that once cells are approximately 2-3 cell-diameters away from a blood vessel they die of hypoxia, and thus cannot break through a basement membrane and achieve metastasis without evolving a tolerance for hypoxia? This is essentially the cancer-specific version of the anthropic principle.
I love it when physicists start spinning yarns about other people's fields.
They are hardly "spinning yarns". Did you miss the part where they were asked to be part of a cross-disciplinary program, along with other physicists, to bring fresh ideas to the field of cancer research? They are doing what they were asked to do, and wrote an article on it. It's hardly a setup for clinical trials in which human lives are at stake, so I think we can, while holding out for evidence that increases the probabilities of certainty in this theory, allow them to do what they've been asked to do.
no, the predilection for hypoxia is due to the warburg effect (which used to be thought of as a cause for cancer but now is accepted as the result of carcinogens, but a 'hallmark' of cancer). Fermentation of glucose in the cytosol is more efficient than using the oxygen-requiring mitochondria - However, it can only occur in less aerobic conditions. So the rapid division of cancer cells is assisted by this shift in metabolism, but that requires anaerobiosis.
The conjecture is reasonable. My point is that it's inappropriate to trump up an untested conjecture, as it creates a cost to the investigator for not being right. And you don't want to be incentivized in this way during your investigation: you're supposed to be fully skeptical.
>According to the article they are simply comparing gene expression from cancer cells to ancient lifeforms, which would directly support the theory.
No it wouldn't. Cancer expresses fundamental genes critical to all multicellular organisms. That is already known so the fact that cancer expresses 'older' genes doesn't prove anything.
Consider this nugget of wisdom that is 100% pseudoscience: A century ago the German biologist Ernst Haekel pointed out that the stages of embryo development recapitulate the evolutionary history of the animal. Human embryos, for instance, develop, then lose, gills, webbed feet and rudimentary tails, reflecting their ancient aquatic life styles.
That is wrong, it has been know to be wrong for a hundred years at least and basing a theory on it is ridiculous.
> You mean the current scientific consensus doesn't believe that is correct any longer.
No, I think that what the grandparent meant is that, for all the problems "ontogeny recapitulates phylogeny" has, it is quite falsifiable in its claims, and they have been falsified. Therefore, its false.
> Scientific consensus believed bleeding patients was a proper treatment for two thousand years.
"Proper" isn't part of a scientific conclusion, and for most of the time bloodletting was an accepted treatment, the scientific method didn't exist (it certainly hasn't for two thousand years), so it couldn't be the subject of a scientific consensus.
Scientific consensus does change, though, but the important thing is why. A theory which loses status as a consensus not because it has been falsified but because a more broadly applicable or parsimonious (two sides of the same coin, really) theory with equivalent predictions (at least, of those that have been tested) insofar as their domains overlap replaces it might still be correct, even though it is no longer the consensus theory. One that is rejected because its testable predictions have been shown to be false is not in the same position.
> From a common sense point-of-view it still seems like a legitimate theory to me.
A theory that comports with "common sense" (i.e., someone's intuition) but whose claims don't stand up to empirical scrutiny does not remain valid merely because it is (for someone) intuitive.
Physicists have made great contributions to molecular biology, but this isn't one of those cases. This one is so wrong at so many levels it's hard to to justice to it. Much of it just misinterprets things we already know.
In the article it says:
Davies and Lineweaver claim that cancer is actually an organized and systematic response to some sort of stress or physical challenge. It might be triggered by a random accident, they say, but thereafter it more or less predictably unfolds.
There's nothing new here. Cancer is basically deregulated cellular replication and proliferation. These are core processes in multicellular organisms, so of course they unfold predictably. The trigger is often a mutation ("random accident") in a key regulatory gene.
The article says the new theory "challenges the orthodox view that cancer develops anew in each host". But we know that the mechanism of proliferation does not develop anew in each cancer; only the trigger is different in different cancers.
The physicists say “We envisage cancer as the execution of an ancient program pre-loaded into the genomes of all cells,” and that "it will express genes that are more deeply conserved among multicellular organisms, and so are in some sense more ancient."
Since cellular replication programs are deeply rooted and ancient because they are fundamental processes, this shouldn't be surprising.
"Sure enough, cancer reverts to an ancient form of metabolism called fermentation, which can supply energy with little need for oxygen, although it requires lots of sugar." Guess they are unaware that your muscle cells do they same when they are starved for oxygen! Anaerobic "fermentation" is a standard part of our metabolic system under certain conditions.
The guy wrote an article on his 'idea' back in 2011 in a newspaper:http://www.guardian.co.uk/commentisfree/2011/apr/25/cancer-e... where he explicitly uses Haeckel as evidence for his theory: A century ago the German biologist Ernst Haekel pointed out that the stages of embryo development recapitulate the evolutionary history of the animal. Human embryos, for instance, develop, then lose, gills, webbed feet and rudimentary tails, reflecting their ancient aquatic life styles.
Of course this is total pseudoscience and I'm floored someone is receiving money for coming up with this. His descriptions of cancer (after two years of apparent study) simply do not make sense in any context.
Cancer's adaptation to anaerobic respiration has been know for a long time and isn't some magical 'atavism' or 'safe mode' but a necessity for a tumor.
I don't see you pointing out anything that is wrong?!
I also disagree that all that stuff should be obvious to everybody. Certainly wasnt to me, although it makes sense, which is why it's interesting. Rather than saying "hell, anybody can get cancer and we have no idea why" these guys seem to think about its role in evolution. They are thinking about it not as a random mutation but as an intentional genetic program which just makes much more sense.
Like in any criminal investigation, answering the cui bono question is imperative. I agree with the people in the paper actually: A deeper understanding is necessary. We dont know whether that will lead to a cure but as a software engineer its my experience that you can't fix a bug you don't understand. (And it amazes me so many of my fellow engineer try....)
I also disagree that all that stuff should be obvious to everybody.
It doesn't. But not everybody publishes something in journals. When you do, you have a responsibility to know your subject.
They are thinking about it not as a random mutation but as an intentional genetic program which just makes much more sense.
This actually makes little sense to me. The replication machinery and programming is deeply established, not a random mutation. The mutation is in the genes that regulate the system. The inappropriate execution of the replication is the key and that's well established to arise from mutational changes in regulatory elements.
As a biologist and chemist studying cancer, this sounds an awful like the "ontology recapitulates phylogeny" theory which, while a cute just-so story, turns out to be a particularly misleading way of looking at developmental biology.
and now I'm going to be pithy: Looks like some physicists think oncology recapitulates phylogeny.
As a biologist and chemist studying cancer, this sounds an awful like the "ontology recapitulates phylogeny" theory which, while a cute just-so story, turns out to be a particularly misleading way of looking at developmental biology.
It is misleading, but with a small kernel of truth that appears in the article.
That kernel is that there is little evolutionary pressure to cause changes in early development, and huge potential consequences to adjustment there, so early development is remarkably stable across many species. And this conservation goes right down to the genetic level. Thus it is wrong that we trace out our evolutionary heritage, however it is true that there are similarities between early development of distant species, and that commonality is due to conserved genes that evolved a very long time ago.
(There is, of course, another commonality. And that is the observation due to Gould in Ontogeny and Phylogeny that a lot of the raw material for evolution comes from altering the timing and rates of existing development processes, rather than from introducing "new" features.)
That cancer is predominantly not "just" a mutagenic defect but that it more specifically recruits developmental controls and make them go go awry is relatively new in biology, but not revolutionary.
Can you quantify "relatively new"? It was clear to me 10 years ago, and I thought it was common knowledge, but then everyone I talked to (especially people from medicine and biology) called it "an novel, interesting theory" when I talked about it (and had to explain why that would make sense).
If there's a name or reference for this theory, I would appreciate a link.
[ My argument that gets the "novel and interesting" theory remark is that in many cases cancer becomes "chemo resistant" after much shorter than any multicellular organism would -- thus, it can't just be a random mutagenic defect, and must involve non trivial developmental/evolutionary machinery. I don't remember where I picked it up, and it might have had a different form, but it was over 10 years ago ]
yeah, that's probably about right. I had first started hearing about it 3 years into grad school (i.e. 6 years ago), and I wasn't really a biologist with an interest in oncology until recently.
That said, the idea of targeting characteristic differences that show up when those controls are turned off was somewhat novel to me. But I'm not in the field, and that might be old hat as well.
I am surprised that the majority of comments ridicule or scorn this idea. I found it to be at least an explorable link, even if it would yield no practical cure. It was nice to know which genes are involved, which still makes no claim on the trigger of cancer. It merely states that forms of stress could activate this, and this still would be in line with current theory of one-off mutations (that is, the tangible result after the event).
I stopped reading PZ Myers quite a while ago - he's almost entirely negative, and doesn't seem to be able to do much more than flame other people.
eg. his little diatribe on Haekel. Ok, Haekel was wrong, but the initial processes behind embryo development must still be pretty old, right? Nope, apparently, even mentioning Haekel == wrong, and your whole argument is wrong and you're stupid and should be embarrassed, etc, etc.
I love the fact that physicists are looking at this problem. I really love the way they think--that there are underlying elegant solutions to inherently messy problems. I think it speaks to how much you think the universe makes sense underneath everything, and they see it on a daily basis. Hence, the success and reknown of an Einstein for example: someone who takes the currently-best-known-answers and makes them simpler, more elegant.
I think this sort of thinking needs to be applied in many more fields than it is, and it's very encouraging to see some people from outside the traditional medicine groups even looking at this supremely important problem. We don't need more drug studies, we need more root cause analysis.
Chiming in? These physicists were asked to join 12 cross-disciplinary centers with other cancer specialists, setup by the US National Cancer Institute, and Davies was asked to lead one such center. He's not chiming in on anything; he's doing the job he was asked to do.
Whether this theory ever gets beyond discussions with field experts is another story. But as it stands, he's not just lobbing fly balls to the outfield for no reason.
"This could provide clues to future treatments. For example, when life took the momentous step from single cells to multicellular assemblages, Earth had low levels of oxygen. Sure enough, cancer reverts to an ancient form of metabolism called fermentation, which can supply energy with little need for oxygen, although it requires lots of sugar.
Davies and Lineweaver predict that if cancer cells are saturated with oxygen but deprived of sugar, they will become more stressed than healthy cells, slowing them down or even killing them. ASU’s Center for the Convergence of Physical Science and Cancer Biology, of which Davies is principal investigator, is planning a workshop in November to examine the clinical evidence for this."
This stood out at me as particularly interesting - especially as someone that tries to limit sugar and carbohydrate intake. I've heard other theories about sugar helping cancer spread - does anyone have more information or research on this?
No, it is not true that the scientific consensus on cancer is that it is caused exclusively by random genetic mutations. Different cancers have different causes. There are, for instance, a small number of viruses that together cause a sizable fraction of cancers. Viruses make fairly specific genetic changes.
Sounds like they are coming up with testable predictions, so rock on! That's how it works in physics. Come up with a model, check the predictions, if the model is nuts but predicts well, you have no choice but to accept it until something better comes along. See General Relativity and it's bizarre time space bending and stretching. And Quantum Mechanics. All the biologists going nuts in this thread sound just like people who couldn't accept things can be in multiple states at once.
Myers visits the Davies/Lineweaver speculations three times, in April 2011 [0], November 2012[1], and December that year. PZ is quite scornful in all three posts, but after all, he is a biologist specializing in developmental biology -- so he can be assumed to know the literature, and the details of cell biology, which he claims they very clearly do not.
This doesn't really sound like a completely new theory; Haeckel proposed in the 1800s that "ontogeny recapitulates phylogeny", and it is widely understood that cancer is like a de-differentiated state.
From the source, Davies' argument (for those who don't want to register for Physics World):
Cancers prevalence among multi-cellular organisms indicates it is "deeply embedded in the logic of life". It is found among mammals, birds, fish, reptiles, and other organisms. The human genome is pre-loaded with a "cancer sub-routine" that is triggered by various factors. Once triggered, most cancers follow a similar pattern: reproduce uncontrollably in a specific organ, then spread throughout the host, invading and colonizing other organs and parts of the bodies with the help of tumors.
> The overall impression is of a carefully orchestrated and pre-programmed strategy--its aim to multiply cancer cells and colonize new sites--which is unleashed when neoplastic cells somehow evade the normal regulatory mechanisms of the organism and embark on their own agenda.
Seeing a "highly organized and efficient behaviour in biology" leads Davies and Lineweaver to suspect evolutionary mechanisms at play. Davies finds that orthodox explanations describing cancer as random genetic mutations starting from scratch leave too many unanswered questions--such as why they are fitter than the existing healthy cells in a body, why such random mutation provides "a whole series of mutually supportive survival traits ... in a period of just years of months", why dormant cancer returns in the same organ after a primary tumor is removed, or "why cancer cells deliberately transplanted into certain tissues, or cancer nuclei into healthy cells, often results in normal behaviour", while the reverse process results in cancerous cells.
Thus, from a physics perspective, "there are clues pointing to cancer as a phenomenon influenced by forces and fields--not one that is purely ruled by genetic instruction." Sadly, the article never dives into more on this physics perspective and what kinds of forces and fields may be at play, instead turning to evolutionary biological arguments from embryonic development applied to cancer development.
> Cells are usually regulated by mechanisms that instruct them when to multiply and when to die. What we believe is that when these mechanisms malfunction, the cells revert to the default option, a genetic subroutine programmed into their ancestors long ago, of behaving in a selfish way.
> Lineweaver and I suggest that genes that are active in early-stage embryogenesis and silenced thereafter – which, by our hypothesis, are generally the ancient and highly conserved genes – may be inappropriately reactivated in the adult form as a result of some sort of insult or damage. This trigger serves to kick-start the cascade of maladaptation events we identify as cancer. So the "cancer subroutine" is really just a re-run of an embryonic developmental program. We envisage a collection of ancient conserved genes driving the cancer phenotype, in which the metastatic mobility of cancer cells and the invasion and colonization of other organs merely reflects the dynamically changing nature of embryonic cells and their ability to transform into different types of tissues.
> The big picture is that we attribute cancer's survival traits to deep evolution on a billion-year scale, rather than orthodox explanations that point to evolution from scratch with each case of the disease. In our theory, the latter remains true, but is a small perturbation.
The ASU article covers much of the rest of the original article by Davies. ASU just left out much of the lead-up to how they arrived at their conclusions.
Physics World also has a lot more coverage in the current issue, entitled "Physics of Cancer". The entire issue can be downloaded as PDF here: http://physicsworld.com/cws/download/jul2013
Well, as a molecular biologist I can't say I ever saw cancer in any other way. What's so new about this?
We all know that embryonic, proliferation and stem cell associated pathways like the Wnt pathway and the PI3K pathway commonly are the ones that get out of hand in cancers. We also know that proliferation in single cells is the highest goal and that the uncontrolled division of a single cell (as it probably did when still living alone) leads to problems in higher organisms. If the fail-safes that developed to allow for multicellular organisms fail to prevent the egocentric behaviour of individual cells and a cell reverts back to a "divide as fast as you can"-state, they become what we call cancer cells.
Odd. I always assumed that cancer was a natural part of a living organism, and that what got disrupted was the "cancer suppression mechanism" (not a technical term). The fact that cancer evolves anew in every patient seems very... improbable.
Yeah. Similarly, I pretty much landed (around a decade ago) on seeing cancer as little more than a small subset of death for which we'd identified features and given it a name--i.e., before humans identified cancer, people just died for then-unknown reasons (which were likely, at given points in history, called 'natural causes'). Once we identified cancer, people 'got cancer' and, subsequently, a varying percentage of such people died, with that percentage declining as understanding and treatment improved. I remain fascinated by cancer research and the hope for a cure because I remain fascinated by seeing humans practicing science discover more ways to identify death and prevent it.
I get nervous when physicists wander too far from Physics. They're a smart lot, but look what happened when they went into Finance. :-) It takes a while to apply the math skills properly elsewhere.
"Arizona State University Regents' Professor in ASU's College of Liberal Arts and Sciences. “It is rather like Windows defaulting to ‘safe mode’ after suffering an insult of some sort.”
<snicker>
It's probably as attractive and simplistic as the old saw "ontogeny recapitulates phylogeny". Kinda true at some levels, but not really. Sometimes functionally useful. Certainly not new.
Wanna contribute to ASU? Have I got a real smart guy for you..... Or is that the London Bridge in Arizona?
I use it to store any file that I save because it's the first spot that pops up (and the easiest one to find no matter where you are) in Windows Explorer or when you right click and say "save as".
Then every few months I shove them all into an "old desktop" folder.
http://www.phdcomics.com/comics.php?f=1174
I wish these researchers well. Cancer has been tough to treat,
http://www.sciencebasedmedicine.org/why-havent-we-cured-canc...
in large part because "cancer" is not really just one disease, but a general term for a variety of related diseases with many different kinds of outcomes and treatment trade-offs.
If someone can find a very general commonality to most cancers, with a treatment for that commonality that is safe and effective (and not too expensive) for human beings around the world, that person will of course win a Nobel Prize in medicine. Here on Hacker News, over the 1704 days that I have been here, I have seen many kind submissions of breathless claims of breakthroughs in cancer research. We all desire that cancer breakthroughs happen. Follow-up, so far, has not suggested that any such cancer breakthrough actually has happened. I wish the researchers well, and I hope to hear later about placebo-controlled multicenter clinical trials that show that this approach is safe and effective for human medicine.