IRA FLATOW, HOST:
It's a phenomenon that has perplexed cancer researchers: A patient is taking an anti-cancer drug, it's tailored to the genetic makeup of his tumor, and right after the targeted therapy, the patient shows a dramatic recovery. Plus the side effects of conventional chemotherapy are minimum. Things look great. But then a few months later, the cancer is back again.
How does the cancer evade the drug treatment? Why do targeted therapies often produce short-lived results? Well, some scientists are looking for answers within the cancer cells. But one team of researchers decided to search for clues beyond the cancer cells, in the tumor's own neighborhood, and they say what they found could have a major impact on cancer research and personalized medicine.
Dr. Todd Golub was a senior author of a study published in Nature. He is chief scientific officer at the Broad Institute and investigator at the Dana-Farber Cancer Institute. Welcome back to SCIENCE FRIDAY.
DR. TODD GOLUB: Thank you for having me.
FLATOW: You're welcome. So the targeted therapies, they attack the cancer, and then what happens? They lose their power, or something's going on within the neighborhood?
GOLUB: Well, that seems to be the story for a number of different cancers that show the pattern of tumors initially showing a great response to these new targeted agents, and then the tumor comes back quickly. Or in many cases, the tumor doesn't shrink entirely, and this has really vexed the cancer research community in order to really see the full promise of personalized medicine.
FLATOW: When you say targeted, how do you target the tumor?
GOLUB: Well, the old-fashioned approaches to cancer treatment have been using chemotherapy that we're all familiar with. They're pretty non-specific. They kill many types of tumor cells, but they unfortunately kill many normal cells, as well, as a side effect of the chemotherapy. That's why your hair falls out, and you get lots of side effects.
The newer classes of anti-cancer drugs target specific proteins that are mutated in the cancer cell but not in the normal cell. So these show great promise for being much more specific only for the cancer cells, not having side effects. So there's a lot of excitement in our community now around using these new targeted agents that are specific for particular mutations in particular tumor types.
FLATOW: And this is what personalized medicine is all about?
GOLUB: Well that's the idea. The concept of personalized cancer medicine is that patients could have the DNA code within their tumor read out in a genetic test that would identify which mutations are causing their particular tumor and then use a drug that is specific for the types of mutations that are causing their tumor. That's where the personalized part comes in. So there's a lot of excitement about this concept.
FLATOW: But as we said before, something happens that stops the action, and the tumor comes back, and you think you have identified something in the neighborhood of the tumor cells?
GOLUB: Well, that's the story that we discovered in the case of melanoma, a type of skin cancer, which has been a huge treatment problem for many years. A lot of researchers had given up on melanoma entirely because it seemed that nothing would work. Even chemotherapy didn't work particularly well. The only patients that could be cured of melanoma were those where the tumor could be surgically removed entirely.
But then a number of years ago, scientists in England discovered a mutation in a gene called B-Raf that was present in about half about all patients with melanoma, the mutation only in their tumor cells, not in their normal cells. And then a company rapidly developed a drug that inhibited this mutant B-Raf protein. And then almost overnight, the melanoma research field changed because all of a sudden, patients whose tumors had this B-Raf mutation were showing dramatic responses to this B-Raf drug, this specific drug, this personalized medicine drug.
The problem was that in the patients that did respond, most of the patients, the tumors would only shrink by about half, sometimes more than that, but usually not completely, and in almost all cases, the cancer would come back within about six months on average.
And so we asked the question: Could it be that there is a mechanism for these tumors to develop resistance to this anti-B-Raf drug that's not coming from the tumor cells, but it's coming from the neighborhood, what we call the microenvironment of other cells that are living in and around the tumor itself?
FLATOW: And so you went in, and you found stuff that's going on in that microenvironment, and rather than you have us tell us right now, because we have to wait for a break, we're going to come back after the break and have you fill us in on the rest of this detective story on how you found this stuff and what's going on there.
Our number, 1-800-989-8255, if you want to talk with us, with Dr. Todd Golub, and we'll talk about other cancers as well and kinds of research that's going on in the field of cancer. You can also tweet us @scifri, @-S-C-I-F-R-I, go to our website at sciencefriday.com and leave a comment there during the show, 1-800-989-8255. We'll be back with Todd Golub after this break, so stay with us.
(SOUNDBITE OF MUSIC)
FLATOW: This is SCIENCE FRIDAY. I'm Ira Flatow. We're talking with Dr. Todd Golub, who is chief scientific officer at the Broad Institute and an investigator at the Dana-Farber Cancer Center - Cancer Institute. We're talking about why therapies that target cancer cells seem to work a little bit, and then - for a certain amount of time or attack 50 percent of the cells, and then the cancers come back.
I think we were talking about melanoma specifically, right, Dr. Golub?
GOLUB: That's right.
FLATOW: You went in there, into the neighborhood of where the cells were growing, and you found something else?
GOLUB: Well, we had this suspicion that maybe the microenvironment, the neighborhood of the tumors, not the tumor cells themselves, were contributing to this resistance. But we had no idea which cells in the neighborhood, in the microenvironment were doing this. And so we set up an experiment in the laboratory to test many different types of normal cells from the microenvironment, different types of fibroblasts, for example, types of blood vessel cells, fat cells, all these types of normal cells in the body that are in and around the tumor and to ask whether any of those cells, those normal cells, could make the melanoma cells resistant to drug therapy.
And what we found was that actually most of them don't make them resistant, but we found one particular cell type, a particular type of fibroblast, that was able to make the melanoma tumor cells completely resistant to the drug.
FLATOW: What's a fibroblast?
GOLUB: A fibroblast is a normal part of your body. It forms a lot of structural parts of your connective tissue, for example, and it's sort of thought of as a supporting cell in many different tissues of the body. What's interesting about this is that these fibroblasts were viewed as kind of an annoyance to cancer researchers for many years.
We often would go to great lengths to get rid of them and throw them out of our experiments because they were just getting in the way and because we were so sure that what we should be focusing on was the cancer cells themselves. And what this study taught us is that those fibroblasts, those normal cells, are actually contributing to the behavior of the cancer itself.
FLATOW: All right, give us a little more detail. What are they doing to get in the way, or how are they contributing?
GOLUB: Well, so our observation was that if we would mix the melanoma tumor cells with these special fibroblasts, all of a sudden the melanoma cells became resistant to the drug, but we then needed to figure out how were these fibroblasts doing that. Maybe they were producing some factor and releasing it into the area around the tumor cells that would make them resistant.
And so we tested as many different number of those factors as we could, several hundred of them, to ask were these fibroblasts releasing a factor that was able to make the cells resistant. And in fact again there was a single factor, a factor called hepatocyte growth factor, or HGF, that was all that was needed to make the tumor cells resistant. In fact, any fibroblast that produces this HGF protein was able to make the cancer drug-resistant.
FLATOW: Was this a known protein from before?
GOLUB: It was a known protein. It plays a role in a number of developmental processes. It's important in the liver, that's why it's called hepatocyte growth factor. But people hadn't been thinking about it as having anything to do with drug resistance or drug sensitivity for melanoma.
FLATOW: So this - so you have this HGF that's in the skin where melanoma is growing?
GOLUB: That's right.
FLATOW: And so something that normally would be working in the liver is now found in the skin?
GOLUB: Well, these HGF-producing fibroblasts are normally present in our skin, too. So it's not - we don't think it's so abnormal that they're there in the skin. Many times, genes and proteins get named for the first place that they're discovered, but in fact they're later discovered to have important normal roles throughout the body, and that's the case with HGF.
But it was not known that this could be problematic for cancer treatment to have these HGF-producing fibroblasts in your skin in and around your melanoma tumors.
FLATOW: Which raises two immediate questions for me when I hear about this is, one, could HGF be helping other cancers besides melanoma be resistant to treatment? Let me go with that one first.
GOLUB: Well, I think that's - we're less far along in studying that in detail, but we do have some early preliminary evidence that suggests that HGF may be playing this trick on other tumors, as well. Still more work to be done there, but there's some early evidence that there might be some subsets of particular types of colon cancer, or maybe certain types of brain tumors where this may be at play, as well. Still more research needs to be done there.
FLATOW: And of course the other obvious question is if you can get rid of the HGF, can you make the - can you get away - get rid of the resistance to the drugs?
GOLUB: Well, that's exactly the experiment that we did in the laboratory. We said, well, if HGF is causing this drug resistance, can we reverse the drug resistance by blocking the HGF protein? And that's exactly what happened. We treated - with a combination of the B-Raf drug and an inhibitor of the HGF pathway and found that now the tumor cells became sensitive again, the tumor cells died with this treatment, whereas without this HGF blocker, the cells were completely resistant.
FLATOW: So is this a potential effective treatment for melanoma, as you say, which is so hard to treat?
GOLUB: Well, we're very excited about this finding, again so far just in the laboratory, but because there are drugs already in clinical development that block components of this HGF pathway, either blocking HGF directly or the receptor for HGF, we think it should be possible to rapidly move this finding to a clinical trial to test this hypothesis in patients in a proper clinical trial.
I should also mention that, you know, one view of this study is that wow, this is really depressing: Tumors find all kinds of ways to develop drug resistance. I, on the other hand, think this is very exciting, and I'm optimistic about it because it shows that it's possible to get on top of understanding why it is that tumors become resistant in the first place. And if we can scientifically understand why resistance develops, then it's going to be much easier to combat it with appropriate combination treatments.
FLATOW: 1-800-989-8255 is our number. Of course tumors are tumors. Would it work with other cancers that are not in the form of a tumor?
GOLUB: Meaning like leukemia?
FLATOW: Yeah.
GOLUB: Well, I think we tend to think of leukemias as not being solid tumors, not necessarily requiring the same microenvironment support, but it turns out that even leukemias require signals to survive that come from the normal parts of our body, from the normal microenvironment, from the normal neighborhood. So I suspect that this concept will apply equally to all types of cancer.
FLATOW: And do you think that the tumor has tricked that neighborhood into releasing this protein, making it think it's for some other reason?
GOLUB: That's a really great question. It's possible that these HGF-producing fibroblasts just happen to be in the neighborhood, and they're there doing their thing. Or it could be that the tumor cells are somehow enticing these special fibroblasts into the neighborhood of the tumor because they like this HGF protein that we're making.
We don't yet have any experimental evidence that that's the case, but it wouldn't surprise me if the tumor cells were smart enough to do that.
FLATOW: Does - could this tumor microenvironment affect other things like the progression of cancer? You know, there are times when you have tumors that look benign and then suddenly turn into cancer.
GOLUB: There's been actually a considerable amount of work over the past decade or so on just that, on the tumor microenvironment providing supporting signals for the growth and progression and survival of tumor cells. It's really fascinating, if you try to take - just purify the tumor cells themselves from a tumor from a patient and put them in a laboratory, in most cases the tumor cells just die, because you've taken them away from the supporting microenvironmental cells, the normal cells in the body that are doing something - we still don't really fully understand what it is. They're doing something to help keep the tumor cells alive. So there's a very important signal being received there.
FLATOW: This would also explain the multifaceted aspect of cancer, would it not, how there are other - it takes two or more kinds of environmental or genetic combined with environmental factors for cancers to grow if, as you say, if it's something in the microenvironment plus the cancer itself?
GOLUB: I think that's a very reasonable hypothesis, that it takes multiple events, either genetic mutations or other abnormalities in the body, could be DNA, could be damage from environmental exposure for example, that are somehow collaborating to allow the tumor to grow. So I think that the fact that there are multiple steps needed to make a tumor doesn't necessarily mean that you need to inhibit all those multiple steps in order to kill it, because it's likely that all tumors have special Achilles' heels which if inhibited with the right drug could cause them to collapse.
FLATOW: Interesting. Let's go to the phones. Beth in Waynesburg, Pennsylvania. Hi, Beth.
BETH: Hi there.
FLATOW: Hi there.
BETH: Thanks for taking my call.
FLATOW: You're welcome.
BETH: I'm not exactly sure what HGF stands for, but it sounds a lot like human growth hormone or - and its varieties. And I'm wondering if this stuff is somehow related to human growth hormone, because I know that human growth hormone in excess - and sometimes when it's not even in excess but when it's being prescribed for low-growth hormone disorders - can increase the risks for certain kinds of tumors to grow. And I'm wondering what this might mean for people who have to take growth hormones.
FLATOW: OK. Beth, good question.
GOLUB: Good question. This is a different protein entirely. Growth hormone and hepatocyte growth factor are completely different. They have similar acronyms, but they're very different. They are related in the sense that they're both small proteins that are released from a cell and then can travel in the bloodstream and to different parts of the body and then activate a receptor on a different cell and then cause a series of events to occur. In this case, it's a different protein, but the concept is similar of a secreted protein that can have effects on other cells in the neighborhood.
FLATOW: Dan in Philadelphia, hi. Welcome to SCIENCE FRIDAY.
DAN: Hi, Ira. Dr. Golub, is the HGF activating the same intracellular pathways, just essentially doing an end-run around what is being blocked by the B-Raf drug, or is it doing something else to allow the tumor cells to keep growing? And if you - as George Sledge, the previous president of the American Society for Clinical Oncology once said not too long ago, one dumb cancer cell can beat 10 smart oncologists. Do you expect blocking HGF, if that's possible, to be temporary also?
GOLUB: Well, those are great questions. We do think that HGF is doing its normal thing of activating its normal receptor, which is called MET, M-E-T. It sits on the surface of the outside of the melanoma cells waiting for an HGF signal to come. And when the HGF stimulation comes, the MET receptor turns itself on and then activates a pathway within the melanoma cell, which results in an end-run, as you said, to get around the melanoma treatment. I think it is true that cancer cells are pretty smart, and they've figured out all these tricks to become drug resistant.
But I think we're starting to see that as we understand these different mechanisms of drug resistance, it's going to be able - it's going to be possible to get on top of them. So I think that the - in the end, the oncologists are going to win, even though the cancer cells are pretty smart.
I think, you know, a good analogy to this is HIV. If you treat HIV or AIDS with a single drug, the virus becomes resistant very quickly, and the patients show recurrence of the disease. If you treat with the right cocktail of drugs as is done now routinely, you can have very long-term efficacy of this cocktail therapy, and that's the model that we need to think about for cancer.
FLATOW: I'm Ira Flatow. This is SCIENCE FRIDAY from NPR. Talking with Dr. Todd Golub, chief scientific officer at the Broad Institute. We're going to have to take a break in a little bit. We've got a couple - time for a couple more questions. Our number, 1-800-989-8255. Here's a tweet from Russell(ph) who says: What is the most significant factor that differentiates different cancers? Genetic or origin? What differentiates them?
GOLUB: So we used to think of different types of cancers based on the part of the body they would be found in, that we would differentiate one tumor because it was a lung cancer and the other because it's a breast cancer. Now, we're starting to see that a better way to differentiate different tumors is based on the particular mutations that they have within the tumor cells, within the genome, within the DNA sequence of the different tumors. So in the case of melanoma, we know that melanomas that have a B-Raf mutation will respond to a B-Raf drug.
There are other tumors that also have B-Raf mutations, and those should be explored in clinical trials with B-Raf inhibitor drugs as well. So I think the new textbooks for oncology, for cancer treatment are likely going to be rewritten such that the tumors are differentiated based on the genetics of the tumor more than their origin.
FLATOW: So you think this is a whole new direction?
GOLUB: I think it's a whole new direction for the field, but we won't be able to declare victory until we can really understand drug resistance because these tumors are just too smart.
FLATOW: Yeah. And they figure out a way of...
GOLUB: Yeah.
FLATOW: ...beating - so you're optimistic then?
GOLUB: I'm very optimistic. You know, if the - the concept of trying to overcome drug resistance when you have no way of knowing what's going on, that's really depressing, because you don't know what to try. You don't know...
FLATOW: Yeah.
GOLUB: ...how to get over the problem. But if you can look into the cell and see what mechanisms are at play, then it's a different ballgame.
FLATOW: So where do you go from here with the next step in this kind of research?
GOLUB: Well, I think the next step is to try to fully understand what component the normal microenvironment is contributing to drug resistance and importantly to test these ideas in clinical trials. There are many examples of things that work well and seem promising in a laboratory and then don't work as well as we had hoped in the clinic. And so that's really the next important step for this project.
FLATOW: Well, good luck to you, Dr. Golub, and thank you for taking time to talk with us today.
GOLUB: Thank you for having me.
FLATOW: You're welcome. Dr. Todd Golub is chief scientific officer at the Broad Institute and an investigator at the Dana-Farber Cancer Institute. We're going to take a break and switch gears. When we come back, we're going to talk about high-speed railroads. California has said they're going to go ahead and think about linking up San Francisco and Los Angeles, go on and, you know, take steps to link up a high-speed rail. Is it economical? Would you like one? What about the rest of the country? We'll get into the science technology of high-speed railroads. 1-800-989-8255 is our number. You can also tweet us, @scifri, and on our website at sciencefriday.com. We'll be right back after this break, so do stay with us. I'm Ira Flatow. This is SCIENCE FRIDAY from NPR. Transcript provided by NPR, Copyright NPR.