Era of Hope 2011 Plenary Sessions

Session Title: Afternoon Session – Breast Cancer Prevention and Risk

Title of Presentation: Metabolism: Relevance to Breast Cancer

Malcolm C. Pike, PhD, Memorial Sloan-Kettering Cancer Center: We now have Michael who has come to talk to us, also an expert on IGF-1.

Michael Pollak, MD, McGill University: Okay; well I’d like to thank the organizers for the chance to speak to you today. And in my talk, we will hear I think some of the themes that you’ve heard from the prior speakers and this will give you perhaps a flavor for some of the areas of research that are causing a great deal of excitement in the—in the scientific community. And so we’re going to talk about metabolism and its relevance to breast cancer. This is a huge topic. I’m just being selective. I’m also mindful that some people here may be specialists and other people may be advocates with less of a—of a science laboratory experience, so I try to kind of be in the middle in terms of the amount of technical detail.

So metabolism and cancer, what is metabolism; how do we define it? What do we mean by that? Well it’s a very broad term. We mean chemical processes that are necessary for maintaining life. And so metabolism would include for example making cell membranes, you know lipid synthesis, protein synthesis, you know necessary for the structure and function of the cells, nucleic acid synthesis, DNA, you know the information management system of cells, and particularly important for the points I’m going to try to make is ATP generation which is involved in maintaining cellular energy balance. That’s also part of metabolism.

So metabolism is regulated at many levels. And now I’m going to start of focus specifically on energy metabolism. And at the cellular level, an individual cell, each cell has to import fuels from the blood like glucose and make ATP which is the kind of gasoline that the cells use for the cellular functions. And so that has to be regulated at the level of the single cell.

But if you step back we’re not single cells; we’re organisms and so another level of organization is you know how much food is consumed, how much gas is taken in and that’s involved in appetite control, and still at the organism level, how is the fuel that you eat distributed, stored, and used by different organs within the body. That’s another level of metabolic control and even you can extend it further at the society level, you know how much and what kinds of food are available. So we’re really talking about a kind of infrastructure where energy is consumed as food, the organism has to partition it, store it, and then dispense it for each individual cell to use and—and that’s a fundamental part of life.

So cancer cells have very high energy requirements because they’re doing so much—so much bad stuff. They’re growing quickly, they’re invading; it all takes a lot of energy. So cancer cells generate and spend ATP in ways that differ from normal cells. So that led to a direction in research where people are wondering if there could be novel cancer prevention or treatment strategies related to these differences, in other words, that cancer cells have requirements for energy that are different so that—that could be exploited therapeutically.

So I have to now give you some background on the complexities of whole organism energy metabolism that are taught to you know the medical students and it’s—it’s well known in the textbooks. But if you stop to think about it, you are not continuously eating. You—but you are continuously—well that may be a problem in our society—you’re not—you weren't evolved—you’re not designed to eat continuously; you’re designed to have meals of reasonable size and then to survive between meals even though your each cell needs to burn energy. So this is going to connect to some of the themes you’ve heard about. The solution through evolution is that you eat, you store some of the energy mainly in the liver or in fat, and then it—the energy is released gradually from these storage organs to maintain a constant fuel or nutrient level in the blood, for example, glucose and that then is available for each cell to use.

So how does this regulation take place? Well you know hormones like insulin are crucial. And now you’ve heard about insulin, and so insulin is a pretty fundamental part of life and it may be a pretty fundamental part of oncology too because hormones—. Like, here’s how it works; you eat, your blood sugar, all your—everything that you take in through your mouth gets absorbed, so your blood sugar level goes up inappropriately high in the blood. Insulin then is called into action. The insulin level arises indicating a time of plenty. That signal means that the liver and the storage organs uptake the extra glucose, store it—now that could make you fat if you’re eating too much but then it’s available in the storage tissues to be gradually released over time between meals.

So that’s—that’s kind of basic physiology. But now we have to kind of try to link and ask questions about these two levels. You see in the picture there whole organism energy balance; you see someone eating a little bit too much in a well-known chain restaurant and beneath them you see people like we read about in the Horn of Africa who are absolutely starving. And so these are extremes of whole organism energy balance differences. But how does that relate to the actual energy available to the individual cancer cells or normal cells when the body is—is subjected to such extremes.

Well, this is a very famous paper which I won't show the data for lack of time but for sure how much you eat and how obese you are has a lot to do with your cancer risk. Your risk of getting cancer in the first place and if you have cancer your risk of doing badly or well with it, and it varies according to different populations but in some populations about 50% of cancers can be attributed to—to obesity. So these are important questions. And here’s the common view that here’s how it works—that there’s variations in host energy balance that is how much you eat and that causes a variation in the amount of energy available to each individual cell and then the cells’ proliferative behavior or how fast they multiply or how fast they’re likely to become cancerous is changed according to the amount of energy available. That’s a simplistic view and the—and the important point to realize is it’s completely wrong. This is not true at all. And what really happens is this; when you’re subject to different extremes of how much you eat that actually changes the hormonal environment in your body, and the hormonal environment in your body is what alters your proliferation rates in your cells. So it is true as I showed you from that New England Journal paper that there is a huge effect of macronutrient intake and cancer risk, and this is shown in many animal models, too, the more you eat the worse it is, but it’s not because the extra energy you consume goes straight into cancer cells. It’s more because the dietary variations cause variations in these hormones that are very important in controlling proliferative behavior.

So this is a reminder, the brown stain in those breast cancer cells are insulin receptors. And so breast cancer has insulin receptors and breast cancer has IGF-1 receptors. So even though the endocrinologists are classically used to saying well the insulin and IGF-1 receptors are the business of fat and the liver and it’s all diabetes research, the cancer cells and even the normal breast cells have these receptors. So when the levels of these hormones change, it’s plausible that the cancer biology is influenced.

So let’s look at some implications here clinically. We have to remember that cancers are visualized in PET scans and here’s the cancer in this patient—because PET scans detect glucose uptake. It’s labeled glucose so the tumor lights up but it’s not just an anatomical thing like a CT scan; this is showing you that the tumor is consuming more glucose than all the normal tissues. So you may have heard of some nonreputable clinics that advise as a treatment strategy with a naïve interpretation of some of this physiology that a good way to treat cancer is to starve, and you can go into a starvation kind of spa, not eat, and then your cancer will be starved. But this is completely erroneous; it will not work theoretically or practically and the reason is that the cancer cells are better than the normal cells at extracting nutrients from your blood. So if you try to starve your tumor, you know the very last glucose molecule in your blood would be taken by a tumor cell and they just would win the competition. So starving is—is—is not an effective strategy and it gets back to the same point, caloric restriction does inhibit tumor growth by not by the simplistic method of starving the tumors of nutrients but rather through these hormonal changes.

So I say these hormonal changes and I’m using insulin as an example but there’s many hormones that vary with nutrition state—insulin, IFGs, glucagon, inflammatory cytokines. It’s just interesting to look at this time line to remind ourselves that when the insulin receptor was first sequenced and cloned in the title Axel Ullrich said wait a minute; this looks like an oncogene. So that it’s—it’s plausible that this line of research is—is relevant—that there is an interaction between nutrition, nutritionally related hormones and cancer.

So do cancer cells care if their host is hungry? In other words, this is kind of trying to integrate these concepts, and you know since for 100 years we know that if you take a mouse and give them a carcinogen and starve them the cancers in that mouse will grow less aggressively. We now understand it’s not because the cancers have no nutrients. It’s because of the hormonal environment that’s changed. We now have models where we can feed mice junk food, elevate the insulin levels and that makes the cancers grow more quickly. And the latest subtleties are shown in the last point there that of course as we’ve often heard, tumors are heterogeneous; they’re not all like each other. It’s very hard to generalize and say something that’s true for every single tumor, and different tumors vary from each other in the extent to which host diet is important in determining their behavior.

So let’s look at some examples of data. Here you see an experiment where we’ve—where we’ve just given the same kind of mouse, the same kind of tumor and you’ll see that there’s two different growth rates—the blue rate and the red rate, and that’s—those different growth rates are just influenced by blue is control and red is a junk food diet. And the junk food diet on the left panel there you see is associated with higher levels of insulin which is exactly classic physiology. But the blocks shown at the bottom left show that the insulin receptors of the tumor cells are more active when the person is having the junk food. So this is—is—is a mechanistic way by which an overfeeding can influence the diet.

And then in another model here, we have the same kind of results where we have breast cancers growing in mice. The red arrow indicates the controlled growth, the green arrow indicates a case where we’ve poisoned the pancreatic beta cells so they cannot make insulin, so this is an insulin deficient mouse and even though because it’s insulin deficient it has too high glucose; it’s like a Type 1 diabetic. Even so the tumors are inhibited because these tumors want the insulin that they lack; they don’t really care too much about the glucose because there is enough glucose for the extra good glucose importing cells to pump it in. And then a new drug candidate that inhibits the insulin receptor also acts in a way that inhibits the growth of this tumor model.

So to summarize this part of the talk, the idea is that macronutrient variation impacts carcinogenesis and tumor behavior by varying the hormonal and cytokine environments with insulin as an example and eating more than you need to eat is—is not the only cause of cancer, but it does increase cancer risk and it does worsen the prognosis of a sub-set of cancers.

I’d like to turn now to Metformin, which you’ve heard about from the prior speakers, and say a few words in the next 5 or 6 minutes. It’s a very simple compound, very fascinating, comes from the French lilac and kind of a natural product. It’s been used for diabetes treatment for a long time. There’s the structure. And there were some remarkable retrospective epidemiological studies that showed something very unusual. Usually when we study drugs that are on the market, we find that there’s nothing—there’s nothing going on or that there’s some unexpected toxicity. But in this case, retrospective epidemiological studies showed an unexpected lack of cancer amongst the diabetics taking Metformin for their diabetes. And the effect size was huge. I mean there was—you know more than half the expected cancers were missing. And so this kind of study led to a lot of interest.

And because it’s not like many other cancer prevention strategies that are disease-specific, you know one drug prevents breast cancer and another drug prevents prostate cancer, so you'd have to have—you know take a blue pill, a yellow pill, and a green pill if you wanted to prevent several cancers, all cancers you know tend to be less amongst the Metformin users. So is this a valuable clue or a misleading artifact? What is going on with those studies? We have to clarify it. And here are the caveats; you’ve already heard some of them from other speakers. All the people in these studies were Type 2 diabetics. That’s why they’re on Metformin, so it’s not clear if this is relevant to non-diabetics. And all of these studies were retrospective so formally they should be regarded as hypothesis generating. And it’s encouraging that many people have been able to reproduce these results all over the world but we have to bear in mind that not every study has—has shown the same results. So is it plausible; could there be a mechanistic basis. And I’m just going to show you some results that stimulated by these population studies, scientists actually went to the lab with some skepticism to show that actually this was some statistical error. There couldn’t really be a laboratory counterpart. But surprisingly there was; in other words, the laboratory research was stimulated by the epidemiologic findings and here’s some examples.

This is a paper in press where we showed—I won't bore you with the technicalities but we showed that when we put the mice on a kind of junk food diet, the kind that makes some kinds of tumors grow more quickly if we give the mice the Metformin who are on the junk food diet their PET scans of their tumors are normalized. So PET scan looks like slow-growing tumor with the controlled diet, aggressive tumor if you’re eating the junk food, but junk food with Metformin it looks more relaxed. And even if we take away the whole organism, the whole animal and just look at cells in a dish when we put Metformin on the dish, we can make the breast cancer cells grow more slowly and they grow more slowly—I mean insulin can speed the cells growing—growing more quickly but when we add Metformin the cells are growth inhibited.

And in more sophisticated research that I won't go into but here’s one of the many references about this—we find that like so much cancer research, it ends up being complicated because the effects of Metformin are not uniform for all metabolic states nor for all kinds of tumors. So what does Metformin really do and what’s going on? Turns out that Metformin even though this is a very commonly used drug taken by tens of millions of people every morning for their diabetes, the real mechanism of action is quite obscure. And most doctors who prescribe it and most patients who take it don’t really know what it’s doing. But really it’s a mitochondrial poison. It’s a mild poison for the cellular organelles that are very involved in ATP generation. It prevents the cells from doing their oxidative phosphorylation to generate energy.

Now you heard from the last speaker that that’s the process—one of the things related to aging. And so when you have too much of this—when you’re burning too quickly, you age faster. Metformin slows down the rate of ATP generation.

Now a piece of background information, one slide when—when you have two—when a cell makes too little ATP, so it’s—it has—it feels it’s running out of energy, there’s a control system called the AMPK control system which says oh, I’m running out of energy. I better stop spending energy. I better proliferate more slowly, make less proteins, do everything I can to save the energy because I perceive I’m running low on energy. And when the liver in particular is exposed to Metformin, the liver has a little bit less ATP so each liver cell says I’ve got to save energy. I’m going—I’m going to do everything I can to save energy and for the liver that means exporting less glucose to the blood and then the blood glucose falls and because the blood glucose is lower the insulin gets lower. And that’s why Metformin works in diabetes. In Type 2 diabetes, which is associated with high insulin levels, the insulin levels are lowered by Metformin.

So I have a few cartoons here now that summarizes a lot of research. Metformin starts off; it poisons the mitochondria, there’s a little bit less ATP; those are the first steps and in the liver that leads to less glucose output, drops the insulin levels, so for the sub-set of tumors that cares about insulin level the Metformin could inhibit the growth by lowering the insulin level even though the Metformin doesn’t get near the tumor. It’s acting on the liver indirectly to control the tumor growth.

But supposing now that the Metformin actually can reach the tumor, not the liver, the first steps are the same; it’ll poison the mitochondria, there will be less ATP, AMP kinase thermostat kind of system will be activated, and so oh, this cell—I got to slow down. I’m running out of energy. But how could a tumor cell behave if it’s trying to respond to energy stress? The only way it can behave if it’s under the control of this AMP kinase system is to spend less energy. So spend—if you’re a tumor cell and you have to spend less energy, you can't be an aggressive tumor. You’re just going to sit around and grow less, make less protein and proliferate less. So that’s a model why Metformin could have a cytostatic effect on tumor cells.

But there’s another case, the last case; supposing that you’re a tumor cell whose thermostat is broken. The AMP kinase system doesn’t work. You take the Metformin, you run out of energy, but you cannot compensate for the running out of energy by spending less energy because you don’t realize that you’re running out of energy. And in that case you have high energy expenditure, but less energy available, and so you have an energy crisis and you die. So the last two slides here are just to say with all of this excitement, there’s been a lot of clinical interest to—to work with Metformin. Metformin is off-patent. It costs nothing; it’s effectively free, but we don’t have the discipline as if it was being coordinated by one office in a pharmaceutical company. You know and that’s both an interesting situation and a difficult situation. So there’s many clinical trials in progress or planned even though we still have important gaps in knowledge, and some of them are kind of amateur, some of them are more serious, because anyone can just prescribe it. Every drugstore stocks Metformin.

It may be useful for prevention but not for treatment but those pharmaco-epidemiological studies didn't suggest that cancer patients on Metformin were surviving longer. It just said that normal diabetics on Metformin got less cancer. So should we be using it in a prevention context or a treatment context? And maybe it will only work in certain kinds of patients. As you heard before, the people who have most to gain from the indirect mechanism of insulin lowering would be those who have high insulin levels. So maybe we shouldn’t be trying it for every woman. And pharmacokinetic data are incomplete. Maybe Phenformin would be better than Metformin; what’s the best dose to use? I’m trying to tell you that there’s still a lot of work to be done here, and we shouldn’t feel that we’re ready to put Metformin in the drinking water.

And—and a lot of people are saying well let’s just take any old existing cancer treatment and add Metformin to it. We don’t know if that—if there’s rational combinations or not. So I’ve introduced you here I hope to a field and—and these metabolic factors may explain certain aspects of the relationship between macronutrient intake, obesity, and cancer. This energy balance aspect may also be related to exercise and—and cancer risk, but we do need to—to build on these important clues and—and gain more information before we’re ready to really optimize the clinical applications. Thank you very much.