The complexity of cancer: A science-based view

Last week I participated in a panel discussion at NECSS with John Snyder, Kimball Atwood, and Steve Novella, who reported on the conference last Monday. What I mentioned to some of the attendees is that I had managed to combine NECSS with a yearly ritual that I seldom miss, namely the yearly meeting of the American Association for Cancer Research (AACR) meeting. There are two huge cancer meetings every year, AACR and the annual meeting of the American Society for Clinical Oncology (ASCO). AACR is the meeting dedicated to basic and translational research; ASCO, as the word “clinical” in its name implies, is devoted mainly to clinical research. Personally, being a translational researcher myself and a surgeon, I tend to prefer the AACR meeting over ASCO, not because ASCO isn’t valuable, but mainly because ASCO tends to be devoted mostly to medical oncology and chemotherapy, which are not what I do as a surgeon. Each meeting draws between 10,000 to 15,000 or even more clinicians and researchers dedicated to the eradication of cancer.

Having taken the Acela train from the NECSS meeting in New York straight to Washington, DC for the AACR meeting, I couldn’t help but think a bit about the juxtaposition of our discussion of the infiltration of quackademic medicine into medical academia with the hard core science being discussed at AACR. One session in particular at AACR highlighted what is one of the most significant differences between science-based medicine and the various forms of “alternative” medicine that we discuss here on SBM on such a regular basis. That difference, quite simply put, is the difference between the simple and the complex. “Alternative” medicine supporters often scoff at practitioners of science-based oncology, asking why we don’t have a “cure for cancer” yet—as if cancer were a single disease!—or why we haven’t made much more progress since President Richard Nixon declared “war on cancer” back in 1971. One part of the answer is that cancer is incredibly complicated. Not only is it not a single disease, but each variety of cancer is in and of itself incredibly complicated as well. To steal from Douglas Adams, cancer is complicated. You just won’t believe how vastly, hugely, mind-bogglingly complicated it is. I mean, you may think algebra is complicated, but that’s just peanuts to cancer.

On Tuesday morning, I attended a session that hammered home that cancer is complex. The session was called, appropriately enough, The Complexity of Cancer. It was chaired by Dr. Joan S. Brugge, professor of Cell Biology at Harvard Medical School and featured as speakers cancer stem cell expert Dr. Sean J. Morrison of the University of Michigan, as well as two faculty from UCSF, Dr. Lisa M. Coussens and Dr. Allan Balmain. Dr. Brugge spoke about mechanisms that control tumor cell anchorage, as well as the interface between genetics and metabolism in cancer; Dr. Morrison discussed cancer stem cells and how some tumors appear to follow the stem cell model while others didn’t; Dr. Coussens discussed how chronic inflammation can lead to cancer; and Dr. Balmain discussed genetic network analysis as a means of determining the susceptibility to cancer of various cells.

Right from the beginning, Dr. Brugge invoked Nixon’s war on cancer with a particularly appropriate observation, namely that the war has been far more difficult than anyone could possibly have ever envisioned in 1971. Back in 1971, in the wake of the discovery of the first oncogene, src, most scientists studied almost exclusively cancer cells, not appreciating the role of the surrounding matrix of normal cells and connective tissue in both preventing and modulating tumors. It’s true that, even back in 1971, scientists understood that the immune system has an important role in controlling cancer, but we lacked the tools to study this system in great detail. Since 1971, the list of discoveries about cancer has been long. Some examples include the discovery of many more oncogenes; tumor suppressor genes; the role of tumor angiogenesis in cancer; cancer stem cells; the rediscovery of the Warburg effect and metabolic derangements in cancer cells; and an enormous number of discoveries in tumor immunology. Each discovery helped us understand better how normal cells become tumors and how tumors grow, invade, and metastasize. But each discovery also led to additional complexities and more questions.

One characteristic that virtually defines a malignant cell in solid organs (as opposed to blood-derived tumors) is its ability to survive when not attached to other cells in its normal surrounding matrix of collagen and other connective tissues. This characteristic of tumor cells has long been recognized, having been first described back in the 1960s. Normal cells, when not attached to the proteins to which they normally cling, rapidly undergo programmed cell death (apoptosis). Apoptosis due to becoming detached is known as anoikis. Dr. Brugge started out discussing her interest in anoikis and understanding how breast epithelial cells survive detachment. Her talk ended up encompassing intracellular signaling pathways, metabolic derangements, and genetics. One observation that is likely underappreciated is that cells that undergo detachment develop metabolic deficiencies that lead to decreased ATP deproduction, decreased cellular energy (real energy, not the fake “energy” — or qi — that alt-med proponents often invoke), and ultimately programmed cell death. One of the more provocative observations is that antioxidants can actually help save these cells by neutralizing reactive oxygen species (ROS) and thereby rescuing fatty acid oxidation. For purposes of this discussion, the details aren’t important (although they are very important for cancer biology). What is important is that antioxidants are not a universal good when it comes to cancer; in the case of the models of breast cancer discussed by Dr. Brugge, antioxidants actually promote the survival of transformed cells because part of the mechanism by which these cells undergo programmed death is through the production of ROS. Does this result mean that antioxidants don’t prevent cancer? Of course not. It does however, when taken in context with other studies, suggest a great deal of complexity, where in some cases antioxidants prevent cancer and others may promote cancer.

Contrast this to the frequent alt-med claim that antioxidants prevent cancer and are virtually always good.

Dr. Morrison’s talk touched upon one of the most contentious issues in cancer today, namely the cancer stem cell hypothesis. This hypothesis goes something like this. There exist within cancer a population of cells that behave like stem cells. They are self-regenerating and each is capable of dividing indefinitely and recapitulating a tumor, while the vast majority of tumor cells have only limited replicative potential. The population of cells that can actually produce new tumors may be very small, much less than 1% of the cells in any given tumor. This concept has been most validated in leukemias, although there is good evidence that breast and a variety of other cancers may follow a stem cell-like model.

Cancer stem cell

Cancer stem cell

Under the stem cell model of cancer, these stem cells are highly resistant to chemotherapy, which wipes out all the non-stem tumor cells but leaves a few tumor stem cells, which can rapidly grow and then recreate the tumor, even from a single cell. In essence, the stem cell model postulates a hierarchy among tumor cells, as contrasted to the previous model, which was a more stochastic model in which any tumor cell could produce a tumor. To boil the concept down, in the stochastic model, any given cell in a tumor could be viewed as having, for example, a 1% chance of being able to form a new tumor if transplanted, while in the stem cell model only 1% of the cells of a given tumor can form new tumors, but they do so with very high efficiency. Moreover, these cancer stem cells have various protein and genetic markers that distinguish them from other cells in the same cancer.

The concept of the cancer stem cell is rather controversial. Personally, although the evidence has persuaded me that there is such a thing as what is commonly called a “cancer stem cell,” from my perspective I view the term as a poor descriptor of this cell mainly for semantic reasons. The term “stem cell” implies unlimited ability to produce different tissue types, which is not what this model is about at all. It’s long been known that tumors are made up of different populations of cells with different characteristics, and it’s not such a stretch to accept that many tumors might have a subpopulation of cells that are most responsible for tumor growth, with most of the other cells remaining quiescent or only slowly dividing. What Dr. Morrison argued is that some cancers follow a more “stem cell” model, while others follow a more stochastic model. He used the example of melanoma to illustrate this point: over 30% of the cells in a given melanoma studied can produce new tumors. While this observation might be consistent with the existence of a melanoma stem cell that makes up 30% of a typical melanoma, Dr. Morrison was unable to find any markers to distinguish the cells that could form tumors from those that could not, and he checked over 50 markers. While this does not entirely rule out a stem cell model (it’s possible that he hasn’t yet found the right marker), it is more consistent with a stochastic model, in which each cell in the melanoma has a 30% chance of being able to form a tumor when transplanted.

Why is this important? It’s important because it has great relevance to treatment. If a tumor is driven by stem cells, then to eradicate the tumor it is necessary to eliminate the stem cells. If it is driven by a more stochastic mechanism, a non-stem cell mechanism, then a “kill ’em all” approach is more likely to succeed. Of course, it wouldn’t surprise me if it turns out that most tumors actually fall somewhere on a continuum between being stem cell-driven and being stochastic. Cancer is just that complex. The term “either-or” rarely, if ever, applies to it.

Dr. Coussens’ talk is fascinating for what it revealed about the immune system and cancer. How many times have you heard “alternative medicine” believers and promoters brag that this nostrum or that potion “boosts the immune system”? As we’ve said before here, it’s a meaningless claim, because sometimes boosting the immune system is bad, as in autoimmune diseases. In cancer, it’s long been known that inflammation, particularly chronic inflammation, can lead to cancer. One of the most classic examples of this phenomenon is how gastroesophageal reflux disease (GERD) can lead to inflammation in the lower esophagus, which can lead to a change in the cells there known as Barrett’s esophagus, which can ultimately lead to esophageal cancer. Inflammation is a function of the immune system; consequently, when you take anti-inflammatories, you are suppressing part of the immune system on purpose in order to decrease inflammation. In any case, Dr. Coussens discussed how activation of certain parts of the immune system can suppress cancer development, while activation of other parts can promote tumor progression. This slide, taken on my iPhone, demonstrates the concept:


Dr. Balmain echoed this message but came at it from a different angle, namely from the complexity of changes in gene expression in cancer, and how a highly complex interaction between inflammation, stromal cells, the immune response, metabolism, and changes in gene expression in a tissue, specifically skin, can influence susceptibility to cancer. One of the big disappointments in cancer research is that relatively few cancers have easily identifiable genes driving them, even though many tumors have a strong heritable component. The reason may well be due to the inheritance of multiple susceptibility genes of low penetrance, meaning that they don’t individually have a strong effect on the characteristics of a cell. Cancer actually involves changes in the expression levels of hundreds, if not thousands, of different genes. In fact, the way we now look at cancer is through network analysis of the levels of thousands of genes in the cell. We’ve gone from looking at single genes to looking at thousands upon thousands of genes. As Dr. Balmain concluded, cancer susceptibility and progression depend upon the emergent properties of many genes, each of which individually has a small effect, and these genetic variants affect the tumor cell, the microenvironment surrounding the tumor cell, or both. Moreover, depending upon the tumor type and situation, inflammatory networks can play opposite roles, either promoting or inhibiting tumor susceptibility and progression.

Is that complicated enough for you yet?

Then let’s move on beyond this talk. On Friday, a bunch of us on our floor on the cancer institute got together to discuss interesting stuff we saw and learned at AACR this year. One topic that came up is the Cancer Genome Atlas, or TCGA (you gene geeks out there may find the initials amusing, but they explain why the word “the” was included). The idea behind the project is to sequence the genomes of many, many cancers. You might wonder why it’s necessary to sequence so many cancer genomes, and it’s not an unreasonable question. The reason is that so many cancers are driven by different mutations that it’s unlikely that any two tumors have the same set of mutations driving them. Consequently, TCGA seeks to sequence at least 500 cancers for each cancer type studied. It started with a pilot project and has since been expanded to 20 different tumors. By sequencing lots and lots of tumors, or so the idea goes, we can identify commonly occurring mutations and sets of mutated genes, perhaps even across cancers, that can be targeted for therapy. At the very least, it is thought that we will be able to develop a greater understanding of the complexity of cancer.

I must admit that when I first heard of TCGA, I was skeptical. To me, it struck me as perhaps the largest fishing expedition in the history of cancer research. Moreover, even this massive undertaking is only part of the picture. As I alluded to earlier, the metabolism of cancer cells is often hugely abnormal, and a “chicken or the egg” argument continues to some extent even now about whether it is the metabolic abnormalities that drive mutations or the mutations that produce metabolic abnormalities. More likely, it’s a little of both, the exact proportion of which depending upon the tumor cell. None of this even considers influences outside of the genome (epigenetic influences) or differences in how proteins are made. Part of our discussion also pointed out that so many mutations have been associated with cancer and that they are often so different in different tumors, even from the same tissue, that trying to figure out which mutations found in TCGA are even relevant to cancer and which ones are actually driving the development, progression, and spread of cancer will be a daunting task, every bit as challenging as the Manhattan Project or sending a man to the moon in less than a decade. In fact, when you consider how vastly, hugely, mind-bogglingly complicated cancer is, it’s amazing that we do as well as we do now and that we’ve made as much progress as we have, arguments over whether we are too conservative or whether pursuing riskier research strategies will bear fruit faster notwithstanding.

Compare this to the view of many practitioners of unscientific medicine. My favorite example of a vastly, hugely, mind-bogglingly simple pseudo-explanation for cancer is that of the late Hulda Clark, who claimed to be able to cure all cancers (not just all cancers, but all disease) but who died of multiple myeloma herself. Her idea was that all cancer is caused by a liver fluke, which she would claim to be able to kill (and thus cure the cancer) with device she called her “Zapper,” a cheap little electrical gadget that looked as though it were assembled from spare parts at Radio Shack.

Another quack, Nicholas Gonzalez, claims that all cancer is due to a deficiency in pancreatic enzymes, for which he prescribes pancreatic enzyme replacement, up to 150 supplement pills a day, a “nutritional” regimen consisting of various vegetable and fruit juices , and a “detoxification” regimen including coffee enemas. He made a name for himself with a cherry-picked case series of his own patients that appeared to have survived pancreatic cancer far longer than is generally anticipated based on historical controls. This lead to a highly unethical clinical trial that ultimately showed that Gonzalez’s patients did considerably worse than conventional therapy, as poor as conventional science-based therapy does against pancreatic cancer.

These are not the only ones, of course. Still another quack, Robert O. Young, ascribes all cancer to “acidity” in the blood, and his treatment is always diet and bicarbonate to try to “alkalinize” the blood:

Robert O. Young

Young even goes so far as to describe cancer as a “poisonous acidic liquid,” states that “there is no such thing as a cancer cell” and that cancer cells are cells that have been “spoiled by acid.” To him, the tumor is the “body’s protective mechanism to encapsulate spoiled or poisoned cells from excess acid that has not been properly eliminated through urination, perspiration, defecation or respiration.” Young’s ideas have sucked in unwitting cancer patients, including one named Kim Tinkham, who even appeared on Oprah’s show a couple of years ago. (In addition, Young also doesn’t believe that sepsis is caused by bacterial infection.) On a related note, another quack named Tullio Simoncini espouses a variant of Robert Young’s ideas in that he believes that all cancer is a fungus. The similarity is that he prescribes “alkalinization” for the fungus, some of which can involve injecting sodium bicarbonate directly into tumors.

If there’s one difference between science-based medicine and quackery when it comes to cancer, it’s that science-based medicine appreciates the sheer complexity of tumors, while quacks often go for risibly simplistic pseudo-explanations of cancer. The complexity of cancer as a set of related diseases is incredible. Indeed, one has to respect it and even stand in awe at its ability to grow, evolve, and ultimately develop resistance to almost any treatment we can come up with. That’s not to say that the situation is hopeless, but it is an explanation as to why, nearly 40 years after Nixon’s war on cancer commenced, our progress against this foe has been incremental. Despite this record, I remain nonetheless optimistic and expect this situation to change within my lifetime. The reason is that we are finally developing the tools, both scientific and technological, along with the computational power to analyze the data, that hold out hope of an understanding of different cancers deep enough to make real progress in reducing the incidence, morbidity, and mortality from cancer. This isn’t any comfort to patients suffering from cancer now or to those who have (as I have) lost loved ones to cancer, but it does give me hope that, should I be one of the unlucky ones who develop cancer, my chances of survival will be better than at any time in history.

No quack can even come close to giving me that sort of hope.

Posted in: Cancer

Leave a Comment (44) ↓