303 | James P. Allison on Fighting Cancer with the Immune System

0:00:00.4 Sean Carroll: Hello everyone and welcome to the Mindscape Podcast. I'm your host Sean Carroll. Cancer is one of the most terrible things we have to deal with in human life. It's a potentially fatal disease. Of course, we're all gonna die someday. That's something that maybe we can make our peace with. But unlike many other diseases, cancer seems arbitrary in ways that are hard to pin down. It can happen to anyone. It can happen at any stage of your life. Young people can get it as well as old people. When you reach a certain stage of your life, like I have, not only you have to worry about checking for it yourself, but you know people who've had cancer and even who have died because of it. So it's very natural that as a species, we put a lot of effort into figuring out what is going on, how to stop this turns out to be really, really difficult as maybe you know. Today's conversation is going to be with one of the world's leaders in the field of fighting cancer.

0:00:58.5 SC: James Allison who won the Nobel Prize a few years ago in physiology and medicine for one of the ways, one of the various techniques that we can use to attack cancer once it starts. The idea being rather than going in and just zapping the cells with radiation or chemicals or whatever, that we can use immunotherapy. In other words, we learn how to cajole the body's own existing immune system to fight the cancer tumors. And if this can work, it works in some cases, doesn't work in others yet. This is what we're studying, and we'll talk about it in the episode, but it's not just a more kind of organic, natural way to fight the disease. But then the person who has had the therapy has that extra layer of protection against getting cancer going forward. You haven't just killed that tumor or most of that tumor, you've built up the body's defense systems.

0:01:56.4 SC: So, of course, that is for real world purposes, a fascinating and important development for science purposes. Of course, it's also fascinating because man, the body very, very complicated. [chuckle] very, very complex networks of reactions and cells and proteins and molecules and all that going on. We learn a lot about it because we are motivated for reasons of making people healthier. But what we learn is also equally fascinating. So we'll talk about how we got here where we're going and how I think the impression I get as an interested outsider is real progress is being made on one of the trickiest problems out there. So let's go.

[music]

0:02:56.1 SC: James Allison, welcome to the Mindscape Podcast.

0:03:00.4 James Allison: Thank you. Glad to be here.

0:03:01.9 SC: I guess this is a big topic. You do re research on cancer therapies and things like that. Let's start at the very lowest level here. We've all heard of cancer. It's bad. Something about cells dividing and going crazy. How do you think about what cancer is, broadly speaking?

0:03:20.4 JA: I think it's a plague of a sort that unfortunately involves your own body going awry in some way. And a lot of I don't know. There's a lot of, used to be a lot of stigma associated with it. Maybe there still is in some cases, but it's an unfortunate thing that happens to us. If we have a lot of cells that comprise our body if they aren't controlled [laughter] it's very, it's amazing to me that it works as well as it does that everything stays under control. But unfortunately it hits people. It hits a lot of people. Way too many. It's hit way too many in my family. And.

0:04:04.2 SC: I guess the question I have is is it the same phenomenon when we talk about different kinds of cancer? There's obviously a wide variety and know some therapies work better than others. Is it accurate to call it one thing?

0:04:18.0 JA: No, it's not. Well, it depends. The common thing is cells growing when they ought not to and where they ought not to. I mean, that's the common thing of all of them. But no, they're very different. They're different tissues. First of all, that's the way it used to be classified solid, and muscle or whatever. Solid tissues versus leukemias, in the blood. And then now it's lung versus colon or bladder versus prostate or whatever. But then with the advent of genomic sequencing and everything, the tendency was to attribute them to the causation. First crudely, obviously carcinogen induced, obviously virus induced or something, although viruses are a tiny fraction of it. It's not zero. But that's not the main thing.

0:05:16.3 JA: But the common thing is just mutation. And the functional classification, according to causation can be useful. And for a while it was thought was the real key. If know RAS is a molecule that really is involved in a pathway that takes signals outside a cell like wounding or something, and tells the cell, oh, you better divide because we need to cover up that scrape or that fix that cut or whatever. But if something happens and that pathway gets locked on, so the doorbell's ringing all the time, the cells just keep dividing. And the idea was, well, if we could inhibit that pathway, then we cured cancer. So this was all the excitement around nearly two thousands, 2010 to up to about 2015 was tyrosine most of these enzymes that do these or things or tyrosine kinases or forms thereof.

0:06:21.0 JA: So the is, you can just inhibit that enzyme. You can cure the cancer by attacking the cause. Turns out good idea. Way too simple because the pathways always have multiple steps and you can fix one of 'em, but then if there's another mutation downstream, you're back to where you started and your drug does nothing. And so that's what the problem is because tumors, as they progress, as the tumors become more and more strange as they start, they become unstable and the genome actually becomes very unstable. And you start getting a lot of more mutations. And as soon as that happens, you have more drivers that are called. And so you can take one out. It doesn't make any difference. You can cure 99%, kill 99% of the tumor cells doesn't make any difference because 1% or 0.1% will have another driver and they'll inevitably grow out and kill you.

0:07:17.1 SC: That's interesting. Actually, I don't think I knew that, that the rate of mutation in the tumor gets much larger. So as the tumor is growing you're not just fighting one kind of cell.

0:07:28.1 JA: Yeah. That's not totally universal, but generally it's true that the tumors become, as they progress, become less and less stable. And so the idea of having a magic bullet that can attack it at its source is really, was not unreasonable. And it was impossible to even think of doing anything about it until we had genomic sequencing, which we do now, but now we know it's probably not gonna be, it's certainly worth doing because it can prolong life. I've, as an immunologist tend to look at things differently because the tumor biologists who have the take the classical tumor biology view of a cancer would say, oh, well you got that mutation, let's attack it and we'll be done. So we should concentrate on what's going on there. On the other hand, so there's lots of mutations.

0:08:26.2 JA: The mutations really occur pretty much randomly. It's when they hit, so they've gotta hit generally a couple of genes before you've got problems. A couple of very types of genes that are very specific. But the other ones, the cancer biologists say, oh, well those are irrelevant. They're passengers. So they're not important. Because They're not drivers. The immune system on the other hand, trained as an immunologist. Your immune system is just, its purpose is to find things that shouldn't be there.

0:09:00.8 SC: Exactly.

0:09:01.2 JA: They don't Care what they do. It's, that shouldn't be on my cells. That that cell may have a virus in it. I'm gonna kill it. And so then you look at it that way. And those mutations, which to the cancer biologists are not important because they're not the drivers.

0:09:20.8 SC: Yeah. Okay.

0:09:23.9 JA: They're equally important to the immune system.

0:09:24.0 SC: The immunologist.

0:09:24.4 JA: The immune system doesn't know the difference. It just does. There's something different here. We better get rid of that guy.

0:09:30.6 SC: You already mentioned that it's a little bit surprising to you that the body lasts as long as it does and works as well as it does. I see that. And I also see someone else saying, it's kind of amazing to me that bodies haven't learned to fight cancer better since it's all over the place.

0:09:49.1 JA: Yeah. Well evolutionarily, it's all evolution. And unless it takes you out before you reproduce this evolution doesn't care. If you get cancer when you're 50, evolution doesn't care.

0:09:57.5 SC: Right tell me about it right.

0:10:00.0 JA: Doesn't matter right Species.

0:10:02.3 SC: But some animal species are pretty good at avoiding cancer. It does seem to be possible. Do we understand that?

0:10:08.9 JA: No, I wish we did. No, I think that people were, maybe they have a really good immune system. I doubt it. I doubt if it is even possible.

0:10:16.1 SC: Isn't there. Some this is very vague to me. Isn't there some paradox that larger animals, you might expect to get cancer more easily 'cause there's more cells, but in fact they get it more rarely.

0:10:25.6 JA: Yeah. That's true. I guess elephants don't, as far as I know, elephants don't get cancer very often. But then again, maybe they're not exposed to carcinogens. Like we're.

0:10:36.6 SC: That's true. But mice happily we're able to give cancer too, 'cause that's where we do a lot of our tests right.

0:10:41.4 JA: Yeah, not so good for the mice. But.

0:10:45.1 SC: So how much do we know about, I know this is not, you sort of just very nicely explained why this is not what you care most about, but how much do we know about how tumors start? And is it just a myriad of various different reasons or is there some central understanding?

0:11:02.8 JA: Well, there's some sense of understanding. I think what it comes down to, if you look at all the data that exists, is that it's basically mutations that cause it. And if you get out of the sun a lot and you expose your skin to ultraviolet radiation, the chances are you're gonna get mutations in your skin. That can cause melanoma because that's what the sunlight hits. And if you stay outta the sun, you're probably not gonna get melanoma. Although plenty of people, same in lung cancer, you smoke, you're highly, you're a lot more likely to get, but you're not that if you don't smoke it. I mean, you're not gonna get lung cancer because the cells divide. There are mistakes.

0:11:47.3 SC: And you're obviously gonna tell us about immunotherapies. But maybe put that in the context of other kinds of therapies. We've all heard of chemotherapy, radiation therapies, etcetera, least most common things you do.

0:11:58.9 JA: Let's go back, let's go back historically. Let's go back way to the start. There's evidence in that the Greeks, knew about tumors and they cut 'em off.

0:12:12.4 SC: Okay.

0:12:12.5 JA: And so the very first cancer therapy was surgery.

0:12:14.5 SC: [chuckle] got it, yes.

0:12:16.8 JA: And that still is perhaps the most effective if you can get it all. That's the problem though, because by the time you, certain kind of cancers in particular like melanoma, very often by the time you notice it and then you're kind of big way, it's already spread to other organs in your body, prostate, same way. So but if you can catch it early surgery is pretty effective. The next therapy that came along was radiation. The cures around many of the 19th century beginning of the 20th century. Madam Curie in particular developed radiotherapy. And, that was quite useful and curative in some cases. Of course, it also caused cancer.

0:13:08.3 JA: If people learned about immunogenic effects. But the idea was you blast the cell and given it enough mutations that it can't live. The problem, of course, is that it also kills the normal tissue. So you gotta be careful about that. And then there's, oddly enough, with the advent of mustard gases and things during World War I and chemical Warfares agents that were developed leading into World War II ultimately led to some pioneers in cancer therapy in the '50s, applying it to leukemias, childhood leukemias in particular, mustard, I mean, just it's toxic gases that are toxic chemicals that were used to. Mustard gas, for example. Those were the basis of the first chemotherapies. They kill dividing cells which.

0:13:57.0 SC: Why are they good at killing cancer cells just 'cause they can bypass?

0:14:06.7 JA: They cause a lot more mutations, as they screw up the DNA as it's dividing and make it sure that the cell can't successfully divide without making it.

0:14:13.0 SC: Okay.

0:14:14.7 JA: Unfortunately, you could also put it in a lot of mutations that'll make you more likely to get a cancer down the road too. But but they also, but the bottom line with both of those, both radiotherapy and chemotherapy, that's given, unless you kill every last tumor cell with those techniques, those approaches to it, the tumor's gonna win. 'cause It'll just come back. And so you have to blast them so hard, or with radiation or poison them so much that it makes you sick. Your hair falls out, the lining of your gut comes out. You don't make new blood cells, your immune system's blown away, and you're sick. And that's what I saw when I was growing up, and my mother, and two of her brothers, and ultimately my brother. And it's a devastating thing, not just the cancer, but the consequences of the therapies. And So.

0:15:14.8 SC: And did tumors spread just because the cells in the tumor sort of get carried around by the blood system or?

0:15:23.3 JA: It's a little more complicated than that. They're signals that tell 'em to stay where they are. Those could be lost, but they could also develop things on their surface that'll tell 'em where you really belong over here?

0:15:36.3 SC: Oh, okay.

0:15:37.7 JA: So it'll leave. That's the way they get there is through the blood, but it's more complicated. There has to be changes that allow 'em to get into the blood and other changes that allow 'em to get out of the blood into the tissue that they're going into. So that's a whole nother area, metastasis of study. That's underwriting model.

0:15:56.6 SC: And my understanding is that you started as a chemist an actual chemist more than a sort of biologist or medical person.

0:16:06.1 JA: Yeah. Well, a biochemist I got bored with chemistry and biochemistry quickly and actually then when I was but again, I had this family history of cancer, which made me, I was interested. Ultimately I was interested in biology, started in biochemistry because that's just, I don't know where I landed [chuckle] But then as I learned about the immune system, T cells had just been discovered shortly before I was an undergraduate. And I was really fascinated by the whole idea that immune system. And, but it was mostly antibodies and B cells at the time. But then when T cells came along I happened, I was lucky enough to have an immunology course as an undergraduate at University of Texas, Austin, and Bill Mandy, the professor, towards the end of the, he was a antibody guy. Towards the end of the semester, he talked about these new cells that have been discovered called T cells that percolate all through your body. It was known, they go all through your body. Not only just going around in the blood, in the lymph, but they actually go through your tissues.

0:17:15.1 SC: Oh, okay.

0:17:16.5 JA: Screen your cells and see what's going on. Every, the remotest corner of your body, is being surveyed, surveilled, I guess the proper verb, by the immune system to make sure nothing's going awry. And that's one of the reasons we're able to keep it all together without going completely bananas. One of the reasons early on, first of all, most of the mechanisms for replicating are pretty damned accurate, but if they're accurate, 99.9% of the time, that's not enough.

0:17:49.6 SC: Not enough. Yeah.

0:17:51.5 JA: You're gonna get problems.

0:17:54.2 SC: And T cells are cells there are a variety of white blood cells.

0:17:57.9 JA: Yeah. Yeah. A variety of 'em. Yeah. B lymphocytes are one kind, but they're also macrophages. They're largely, I guess I would say two families of lymphoid cells and myeloid cells. Myeloid cells are like macrophages and things like that, that engulf bacteria or infected cells are dying debris from dying cells and cleanup wounds and all that stuff, and help wounds repair. And they have innate signals, they're called that they can recognize a lot of viruses, a lot of bacterial, just because they have carbohydrates and things on their surface that is classes of molecules are different than those found in our cells, mammalian cells. So there's sort of this, we call 'em being innate, because everybody's got those and they can protect you against a lot of organisms. But that's not enough either. And so late and evolution with development of actually not shortly after vertebrates and things, but anyway, evolutionarily this other system came along, which we call the adaptive immune system, where you actually have receptors that are made by random recombination of DNA sequences. I don't wanna go the details, but basically, essentially random process. They give you a random set of different receptors.

0:19:27.0 JA: It's been calculated that you can make all things if you had the full ability to make every receptor that can be made with the structures that are in there. It's really a fascinating area of biology that. And it's on. But it's 10 to the 15th, maybe 10 to the 17th power. You only have about 10 to the 10th cells, 10 to the 12th cells. So it's 1000 times more cells than total cells, at least 1000 times more total cells than you have in your body. So each of us really realize only a fraction of the possible diversity of different T cell receptors that could be made in your body and to the species. So the population is protected, probably, but not the individual necessarily.

0:20:27.5 SC: Well, I was gonna ask, do the T cells in our body develop new receptors? Do they learn on the job, or does the body make new T cells?

0:20:36.4 JA: No, they're generated essentially randomly. The cells come out of the bone marrow. For T cells they come out of the bone marrow, they go to this organ called the thymus, which sits right above the heart. And there they start developing from this precursor stem cell into functional T cells by random rearrangements. And they're put together, and there's this fascinating testing system where there's a scaffold on cells that presents the antigens, which are little bits of protein. There are only 8 to 12 maybe amino acids long, that are presented in the surface of this thing. It looks like a hot dog bun, and it's got a peptide in the middle of it.

0:21:24.2 JA: There's a peptide for virtually every protein that's made in your body that'll bind there and will be put on the surface of your cells. So the immune system knows a sample of everything that's going on in the cell even if the thing normally wouldn't be on the cell surface. If it's a virus that's just infected a cell and the cell starts making bits of the virus, I mean making the parts of the virus that it needs to reproduce and go out and infect other cells. As they're making it, those proteins will be cut into little pieces and pieces of 'em will be put on the surface. And the right T cell comes by, it says, "Whoops."That shouldn't be there. It'll kill it. It's essentially random.

0:22:08.7 SC: So the cell is sort of doing its own annual checkup at all times.

0:22:14.0 JA: Exactly. But you got to get rid. A lot of those are gonna be harmful. They may be harmful. And so they got the tumor. I mean, the thymus not only educates the cells as to which ones can be useful, but gets rid of the ones which will hurt you, hopefully. If they don't, you end up with diabetes or you end up with different autoimmune syndromes and stuff.

0:22:38.4 SC: Sorry, these are T cells that can be dangerous if they're not in the right kind?

0:22:42.6 JA: Yeah. If they're autoreactive, they react with cell proteins. Or sometimes if a virus comes in and tricks the immune system into thinking something's foreign. When it's closely enough to a self thing, occasionally you'll get a spillover, can be damaged like for measles and stuff, down the road. But anyway, basically, they're pretty damn good though.

0:23:06.4 SC: Yeah. This is where my simple physics brain rebels at the complexity of all the networks inside the human body. It's kind of an amazing edifice.

0:23:16.1 JA: Yeah. How many stars are there?

0:23:18.5 SC: Well, there's a lot of stars, but stars are pretty similar to each other. There's 10 to the 22 stars out there. But they're not that different. There's no lock and key in there. I guess that's my question. That your talk about receptors reminds me of people who were trying to study smell and how we're sensitive to different kinds of molecules. Is it a similar kind of thing going on?

0:23:41.6 JA: Yeah, in a way. Except that there, what you have is one kind of receptor for each kind of bit of a smell. And it's the sum of all those that tells you what the overall smell is. But each one only detects one thing and the brain integrates it all. Whereas here you've got all these different ones that are flowing all around. All you need to do is trigger one.

0:24:06.3 SC: And so the job of a T cell is to understand what the normal healthy cells are like and target anything that is not that.

0:24:14.7 JA: Exactly. Not self recognition. That's what it's called. Philosophically, a very, how do you recognize not self?

0:24:23.7 SC: [laughter] But they do a pretty good job like you said and yet we still get cancer. So there's some reason why I guess in principle they can attack cancers, but they don't do as well as they could.

0:24:35.4 JA: Yeah, that's first of all because cancer cells are not necessarily all that different early on especially, although they get weirder and weirder with time and often with a lot more mutations. But they also have ways of protecting themselves 'cause cells don't like to be killed either. For example in tumor cells there's a process called apoptosis. And they're mechanisms that guard cells built into the cell or mechanisms for detecting mutations. If there's too many, the cell tries to commit suicide. It's told to kill yourself because you're gonna get may cause cancer.

0:25:19.4 JA: At least that's the thought. But there are these suppressor genes which do that. So really, in order to get cancer, you've got to not only get an activating gene which will tell the cell it ought to be a cancer, but you got to get rid of those suppressor genes which would shut that down. So genetically it's complicated too, 'cause you really have to have both in order to get it. So that's why some people with retinoblastoma gene, for example, if you have two copies of that, kids get tumors of the eyes when they're about 2 years old. It's a devastating disease. But in other kinds of cancer, you don't get them in your germline, but you can get 'em in your somatic cells. And if you lose the RB genes that makes you a lot in a cell. That makes that cell a lot more likely to get cancer.

0:26:12.7 SC: That helps. Because I did have the question, do tumors or do cancer cells defend themselves? They don't pass on their genes in some sense, but I guess the answer is but they're versions of or made of ordinary cells which do have defense mechanisms.

0:26:30.2 JA: Yeah, yeah. And they also one of the things that we found recently that's even more interesting to me is that the immune system, every now and then, these macrophages who play a role in cleaning up after wounds and wound healing and replacement, they'll protect the tumor too.

0:26:50.0 SC: Oh okay.

0:26:51.4 JA: They think there's a wound and so your own immune system can turn around and we're finding that that's one of the reasons that happens big time in pancreatic cancer and in glioblastoma, which are tumors that are very lethal. And it's not that we got the T cell issues solved with those, but what we know is there are myeloid cells there that are trying to stop the T cells from killing the tissue 'cause they're just doing what that's their normal functions, protect tissue and help it heal. They get in a weird situation where they are treating the tumor like a wound and protecting it from the immune system. So that's one of the things that we're working on now is trying to find out some way to override that or change the myeloid cells where they'll help.

0:27:47.2 JA: They could also help the T cells. Just depends on what they're doing. It's a whole new area of biology, that single cell RNA techniques and things are making us realize that when you talk about T cells, there used to be B cells and T cells. Now there are T cells specifically. First there were two kinds, and there were five kinds. Now there are about seven kinds. The truth is that those kinds are just constructs that we make up. In biology it's a continuity. It's just a continual spread. In myeloid cells it's really pronounced. You can see all these different cells. They differ by two or three genes that are being expressed, and you can just shift the population one way or another, and that's what we're trying to figure out, and then shift them from this protective overall effect to the helping, thing to help the immune system rather than hinder it.

0:28:41.5 SC: One impression I get from reading about these or talking to people is that everything is about switches turning things on and off. Like, once you realize that every cell has the same DNA in it, I guess it makes sense 'cause they do very different things. But it's all a matter of which parts of them are playing a role at this particular moment.

0:29:00.4 JA: Yeah. Which genes are turned on and how much and it's relative amounts of different genes in some cases. The complexity is amazing. And so our brains cut and box things, but nature doesn't.

0:29:14.8 SC: [chuckle] It sounds like with the receptor stuff on the wall of the T cell. That's right down to the level of atomic and molecular structure. Is it very different how we're studying that now than when we were in like the 1980s or whenever it was? Has the technology changed things?

0:29:36.4 JA: Well, yeah. In the '70s, nobody knew what the T cell receptor was. That was my first thing when I got interested in immunology was what is the antigen receptor? What is it that the T cell uses? And nobody knew what it was. And so I put on my biochemist hat and said, well, what should such a molecule look like? And worked it out and published that for the first time in '82, it's gonna look like this. And I had data and stuff. And then the genes were cloned a couple of years later and confirmed that that's what it was. An alpha chain, a beta chain, two chains of protein, both of which have constant regions and then variable regions. And outside it gives you the combining site. But how is it presented? And so there was this. I won't go into the details too much, but that's another complicated thing. How the molecule, that hot dog, it doesn't, every wiener won't fit into that bun. Everybody bun's a little bit different. And holds a different kind of wiener. And so if you've got the right bun for the peptide, you could present it. If you don't, you won't. So, that's another. If you can't present a peptide even though it's there and it's from a cancer, then the T cells aren't gonna see it.

0:31:00.5 SC: We don't mind a little bit of details here. You're allowed, like, if you really want to go into details, we'll go with you a little.

0:31:06.5 JA: Okay. I won't go anymore. But it's a complicated thing. But we didn't know this stuff until the '80s and the '90s. And now we're beginning to understand some detail. And now with a lot of the work in terms of really understanding the biology of it, some of this, the new work from the Nobel Prize this year, David Baker for example, and understanding how sequence informs shapes and how shapes influence sequence, all this, we're beginning to understand at the molecular level how these interactions occur and how to manipulate them a little bit better in ways we couldn't even think about 10 years ago, much less in the '80s.

0:31:53.0 SC: I do love the idea that even before you really had any direct evidence of what it looked like, you could sit back and think, well, what should it look like to do the job that it does? In some sense that's what people did for DNA too.

0:32:04.7 JA: Yeah. I'm one of these old school guys that thinks hypotheses are useful.

0:32:08.2 SC: Yeah [laughter]

0:32:09.3 JA: Now it's get a whole bunch of data and try to pull something out of it.

0:32:13.7 SC: Well, are we at the level now where we take pictures of not just T cells but the actual receptors that are on them?

0:32:21.8 JA: Yeah, pretty much. Pretty much. [0:32:25.8] ____ That's really helping with vaccine design and stuff now. So we're on the verge of being able to, in some cases, develop prophylactic vaccines I think. Certainly for HPV we're there. We can prevent HPV-induced cervical cancer and head and neck cancer by the HPV vaccine if people would just use it. More importantly, in the clinical realm are therapeutic vaccines now. And we understand, we're beginning to understand enough about the antigens that are induced by the mutations where we can begin to come up. This was something that was greatly informed about our recent epidemic with the vaccine strategies that were developed then where you've got, basically, a cassette approach now where we know how to really put a gene in the middle of this cassette. I think we're gonna come up. There's gonna be eventually a small set of adapters you could use. It'll have all the signals that you need to properly activate the innate immune system and all that. But you put your specific thing in the middle of it, and we can generate a vaccine in weeks.

0:34:02.6 SC: Pretty good. Is this part of the sort of CRISPR revolution of gene editing?

0:34:08.3 JA: That's part of it. That's part of it being able to get the genes out. It's more useful right now in testing as to whether you're right by actually doing fine, structured, seeing if you can initiate or terminate an immune response by manipulating things. But, yeah, it's that and just all the work that went in developing the coronavirus vaccines, the COVID vaccines. But that sort of framework, it's not exactly the same thing. It's going to be something very similar, which now you can theoretically take a melanoma patient, sequence the genome, and with computer algorithms predict which of the mutations would actually elicit in your own body, because different people's, it's called MHC type. The genes that you have that influence what you present. If you know that and you know the sequence of the different peptides, you can predict which ones will be presented, and then you can design the proper vaccine to snap it into that thing and start vaccinating people. So that's coming along. I think that preventive vaccines, pretty soon we're gonna be able to maybe protect people that have, what's called, what is it, Lynch syndrome.

0:35:27.2 JA: They get a lot of polyps. They eventually go on and develop colon cancer. It's definitely a predisposition you can predict. But there are mutations, common mutations associated with that that I think if you immunized early enough, you could protect those people from ever developing the disease by immunizing them later. Maybe the same will happen with BRCA. BRCA mutations are associated with, in women, with breast cancer and ovarian cancer, in men also with certain kinds of cancer. But it may be possible to prepare vaccines there. But most of the times the mutations are so individual, individually specific to you that there's no way you're gonna have a vaccine. You could prophylactically treat the population. For most cancers you're gonna have to come up with something that's really powerful and allows you to jumpstart a response as soon as you detect it. And I think we're getting close. We're not there yet, but.

0:36:27.8 SC: I was actually gonna ask about the role of simulations. In physics and astronomy, of course, that's what we do all the time. We simulate things, we test them against the data. I've always had the impression that in biology the state of the art was that biological things are too complicated to do that and too specific and too individual. So, we have to actually test the pharmaceutical in a living thing rather than just putting it on a computer.

0:36:51.7 JA: Yeah, we're still there, although it's better now. We could guess, we could make educated guesses now. We can't get exactly there.

0:37:04.3 SC: So, okay, so we have this basic security force in our body of the T cells roaming around looking for interlopers. You mentioned a little bit about this already. But why is it that they aren't better at attacking cancer? And how can we make them better? That seems to be the project, right?

0:37:22.8 JA: Yeah. Well, again, it's because one of them is big and largely they're individual. So your tumor has mutations that nobody else's has. So, you got to tailor it to the individual. And secondly there are ways that the tumor can lose. For example, the peptide has to be containing the mutation has to get on the cell surface by these MHC molecules. And if you lose expression of MHC molecules, then they're invisible to the immune system. The tumor will still be there, still making the mutation will still be being made, but the immune system can't see it 'cause it doesn't get to the surface of the cell. There are ways around that that we've got to have them. But that's what happens and has been shown in melanoma, for example. Patients just quit making the molecule that carries it.

0:38:19.4 JA: But there are ways, there are other ways around that that we're beginning to come up with. But that's one easy mechanism. Another one is that one of the main ways that tumors kill tumor cells is by making gamma interferon, which is a cytokine that, you've heard of, but one of its activities, besides helping protect you against viruses, it'll causes tumor cells to quit dividing and can kill many of 'em. But if you lose anything along the receptor downstream of the gamma interferon, the thing that detects gamma interferon, it tells the cell it's around. If you mutate anything in that pathway, then that doesn't work anymore either. And so the T cells can make all the gamma interferon. They won't, the tumor won't respond to it. So there's that too. So.

0:39:12.0 SC: So what is our goal as immunotherapists? Are we trying to teach the T cells to ignore some of these problems?

0:39:20.3 JA: Well no. Well, first it's just to get the T cells. That's the, and then secondly in importance is there's something about the myeloid cells, which we're not there yet. We're getting close on the T cells, the myeloid cells we're only beginning there. But ultimately it's how do we deal with these things that comes down. So it's not that you can get the T cells to do any better, but you could come up with ways of making them work at a distance. It's kind of hard to explain, but a different type of T cell, it'll make more soluble factors. Maybe see the antigen on a myeloid cell, even though the tumors can't present it. If these myeloid cells, which will be gobbling up or tumor cell dies, they'll pick up pieces of it, put it on their surface. If they have the antigen and a T cell sees it there and they can make enough of these gam interferons or other things, then they can kill the tumor cells at a distance. Even though they don't interact physically directly with the tumor cell anymore. That's, we could show that happens in animal models, at least. I'm pretty sure it happens in people as well. But.

0:40:29.9 SC: Okay. I read a little bit about the T cells before talking to you, but I didn't read anything about myeloid cells. Maybe you should tell me something about that is sounds like they're important.

0:40:39.1 JA: Yeah, that's, well, the reason. I used to avoid the things like a plague because they're so complicated, there's so many different types. And, but what we found started, what, I mean I'm speaking, they're not as just me, but the field began to realize is the tumors which don't respond well to T cell based therapies usually have really large populations of myeloid cells in them. So that raised the hack, raised the interest. And sure enough, you can show that the myeloid cells can inhibit T cells. And we've found two molecules on myeloid cells that do that. And we've shown if you take 'em away, then we can make the T cells more effective. So that leads rise to some more strategies where you target the myeloid cells. But you're gonna have to do both, particularly in things like pancreatic and glioblastoma.

0:41:32.3 SC: What are the myeloid cells? What's their role when everything is going well?

0:41:38.1 JA: Gobble up bacteria that come in, deal with if antibodies have, if you've made antibodies and they're binding the viruses or bacteria and clumping them up to gobble up and get rid of the bugs that way as I said, to wound heal, to help wounds heal to.

0:41:55.6 SC: Okay, good.

0:41:56.4 JA: Make growth factors and stuff, just sort of generally handymen, they're just sort of general handymen that do whatever fixer uppers or you need. But they can get in the way.

0:42:11.0 SC: And they have this spinoff effect that they can basically communicate the existence of a tumor to the T cells.

0:42:17.0 JA: If you treat 'em right, yeah. Or they can they can hide them from the tumor cell too. Or they could hide the tumor cell from the T cells too.

0:42:27.2 SC: Right. If things are going badly and isn't there also, this is my very vague understanding popping up. But you can get in trouble if you make too many T cells.

0:42:37.7 JA: Yeah. If you make too many T cells, you can definitely get sick. And especially if they react with normal cells or with something that makes some big soluble factors that affect, and that unfortunately is one of the main side effects of the therapies is cost reactivity with normal tissues and things. But there's no free lunch. You start messing with those things. It's not that it always happens. I know of at least one marathoner who got melanoma and went through the whole course of therapy and never missed a race.

0:43:16.2 SC: Oh wow. Okay.

0:43:18.8 JA: Other people are getting very sick and sometimes there can be things which can be fatal. Thank God they're rare. And the clinicians now know from experience just developed algorithms for recognizing when it's about to happen and heading it off, or at least reversing it early on while still got time.

0:43:39.4 SC: And by the therapy it sounds like what we're doing is sort of trying to regulate the amount and the sensitivity of these different cells in our immune system. Presumably by giving people drugs.

0:43:54.5 JA: Drugs in the form of, right now, popular things are antibodies. These are just proteins that we can make that are very specific. That's what we did with CTLA-4 in the '90s, there was this molecule, and it was funny, it looked like a molecule that was the gas pedal on T cells. But this molecule called CTLA-4. And we showed that it was actually not a gas pedal, it was a break. And that the people that called it a gas pedal, what they did was they had T cells that put 'em in culture, activate 'em, and then add an antibody and they'd get more. And they said, okay, well that's a gas pedal. That's what we did earlier in the '80s showed that CD20 eights and other molecule, it's like the gas pedal. So that antigen receptor you can think of is the ignition switch.

0:44:42.0 JA: That's all it is. It tells a T cell, but a T cell has to get a second signal at the same time, which is like the gas pedal. And that's a molecule called CD28 that we showed in the late '80s, was the gas pedal. And if you don't push on that, activating through the antigen receptor doesn't do any good at all. The T cells don't do anything. Tumors never have on their, solid tumors, I should say, don't have the structure that binds to that gas pedal. So it's complicated. So tumors are inherently solid. Tumors are invisible to the immune system because they don't have that second signal. The only way they get 'em is to the tumor. And we worked this out in the early '90s. The tumor gets picked up or whatever causes inflammation. The innate immune system comes in.

0:45:35.5 JA: The innate immune system has those things that they're called B cell for, they could be called Frank. It doesn't matter. It's just the name means nothing. But they bind into CD28. And that says go. So the only way that the immune system was solid tumors, and this is that we published this in the early '90s, was when you, that they grow until there's tumor cell death and the innate immune system comes in and primes the T cells and then you start generating T cells. But the thing is you probably got 10Th the 9th different. Nobody really knows for sure, but different T cells in your body with different receptors. And of course that means you've only got a few hundred maybe of any given clone. And that's not enough to protect you against anything you need hundreds of thousands.

0:46:35.3 JA: Yeah. And so you have to, they have to expand. And so that's what this, so the T cell receptor Act sees the ignition switch and then when you, but nothing happens really until you push on the CD28 molecule gas pedal, then they take off, you gotta generate a hundred thousand. You've got hundreds of thousands to millions of cells to swarm through the body and look for things. That's what's so cool about, because then you don't need to know where to cut or know where to shine the radiation or whatever. You let the T cells go find it, go find the micro metastases, the little clumps of tumor cell under your toenail or wherever [chuckle] that, I mean that's exaggeration obviously. But anyway, they go find the tumor cells wherever they are and take 'em out or at least keep 'em in control where they don't grow anymore. But that process has to happen fast. So the tumor wins. And so.

0:47:38.0 SC: Because tumors are gonna go very fast, generally.

0:47:44.6 JA: Right. And see, well, not that fast. It takes years. Generally it's at the end when they start damaging stuff.

0:47:47.4 SC: Oh Okay.

0:47:52.8 JA: With the problem. I mean, a lot of tumors can get quite large as long as they aren't in a place. You got a lot more degrees of freedom than a tumor that's in your abdomen than the tumors in your brain, for example. Just the sheer pressure in the brain will kill you.

0:48:05.2 SC: Right. Is it something where different kinds of cancer are still gonna need different kinds of T cells to come to life with the gas Pedal?

0:48:13.1 JA: No. It's not pretty different. They're all, they're basically the same kind of T cell. So it's commonality, but it's those myeloid cells again that get in there. Because they can interfere in different ways. So we gotta work on those. But anyway, there's this other molecule called CTLA-4, which we showed in the early '90s, again, was a break. And so at the end of that expansion phase, that CTLA-4 molecule starts coming up and stops the T cells from providing. They've gotta stop or they'll kill you.

0:48:48.1 JA: Because they gotta stop. I mean, it's not, they don't become a cancer, something like that. But they'll just grow normally and use up all your nutrients and nothing left for everybody else. So you gotta stop that process. And that's what this molecule called CTLA-4 does. And so the thing that we figured out in the early '90s was if you block that molecule, you can let the T cells keep going a little bit longer than they would normally long enough to out the tumor. And then you stop the therapy and everything comes back to normal. And so that works spectacularly well in some kinds of cancer. Just taking the breaks off for a while.

0:49:26.5 SC: Right. I know one of the most terrible things you can hear when you have cancer is that it has spread. It spread around the body. So it sounds like maybe this kind of therapy will be more amenable to even dealing with that.

0:49:42.1 JA: Yep. Melanoma kills you because it's up in the liver or the bone or the brain. Jimmy Carter had a melanoma in his brain. He got immunotherapy was cured to be a hundred and whatever hundred.

0:49:58.7 SC: Oh, okay.

0:50:00.8 JA: He was cured about seven years ago with immunotherapy of brain cancer.

0:50:04.8 SC: So I guess that that answers my next question. Like how much is this in the clinic now? Is there a pill that we can take? Is this growing? Is It being tested?

0:50:12.4 JA: Yeah. No, it's not. No, it's all over the world. There are millions of people literally that have been treated in fact in melanoma. Now immunotherapy is pretty much the standard of care. It's the first thing you'll get because it's so effective. The drug that we developed in late '90s ipilimumab that was approved by the FDA in 2011, that's given all over the world. As I said, millions of people have been treated and it cures overall by itself about 20% of people with metastatic melanoma. To put this in context, melanoma was one of the earliest targets for several reasons, but one of which is no drug had ever had any effect at all in melanoma ever. And so when you go with immunotherapy, and also there's some indications that maybe it was immunogenic. 'cause there has a lot of mutations. But any event with our drug, 20% of people are cured by one or two injections of an antibody. The therapy, by the way, is you sit in a chair for about an hour and they run a, there's a drip bottle with a drug in it and it's infused in your blood and you go home and that's it.

0:51:25.6 SC: And that's better than chemotherapy as we know it.

0:51:29.4 JA: Yeah. And with luck, most patients there's some scratchiness, there's some diarrhea, there's some stuff. But usually nothing really bad. Occasionally there can be bad stuff. Even, patients, some patients get type one diabetes, which is autoimmune conditions. It may be if they already had it borderline, I didn't know it. It just makes it worse. But there's a downside a lot of places, but most stuff it's not. Anyway, it was 20%. But then after we after we started, our stuff started coming out Tasuku Honjo in Japan plus Arlene Sharp and Gordon Freeman at the Dana-Farber discovered this other checkpoint. CTLA-4 was the first checkpoint defined as a cell intrinsic a molecule on the surface of a T cell itself, that helps give a negative signal.

0:52:25.9 SC: Okay.

0:52:27.1 JA: So the T-cell receptor obviously is a positive signal. It's the distinct switch, CD28 is the gas pellets, another positive signal. C24 says, stop all that stuff, it's time to quit. And so that's the one we chose to focus on was, you can either make the other ones better. We chose, let's just take the brakes off, let's disable the brakes. And it works spectacularly well, as I said in the phase one trial, which normally is just safety. You make sure you're not killing anybody. And then irate it up until they start getting really sick and then back off a little bit. And that's the way that cancer therapy used to be. So there you prove it's safety. You find the maximum tolerated dose, how far can you go before you make people sick? And then you give that dose and you give it until all the tumor's gone.

0:53:23.4 JA: And if the tumor grows at all, it's a failure. All of that's out the window with immunotherapies. First of all, there is no maximum tolerated dose. Usually people either, they may have adverse events, but you go up and up and you might get, it gets more frequent in the population, but people don't necessarily get sicker. It's not like there's a poisonous level of it. And with in melanoma in the first 14 patients, phase one, there were three whose tumors completely went away, which was just unheard of at the time, particularly in melanoma. 'cause In 2011, when the phase one, phase three trial, sorry, that I was associated with was unblinded and reported to the FDA at that time, if you were diagnosed with metastatic melanoma, the median survival, 50% of people would be dead in seven months.

0:54:22.3 JA: Fewer than 3% would be alive at five years after that. It was minuscule. It's not to say that everybody always died, but it was much less than 1% would survive. Now with just this one drug, 20% are alive at 10 years plus. With no other therapy, just one, one round of therapy, you're done. When Tasuku came along with this other molecule called PD-1, which works a slightly different way, but the overall pictures similar enough, it's another checkpoint that works differently if you put 'em together. There was just a trial that was just reported about a month ago. There was over a thousand people randomized, 10 different countries. I don't know how many different PIs, I mean the gold standard of clinical trials, 10 years follow up, 55% of the patients were still alive.

0:55:18.3 SC: Wow.

0:55:20.5 JA: So now we went from a cancer, which was almost uniformly fatal in less than a five years to we could cure more than 50% of the people with that disease.

0:55:34.7 SC: That is amazing. But, so my immediate reaction is that 20%, 50, 55% numbers are on the one hand, super impressive. On the other hand, why not a hundred percent? What do we gotta do? Like [laughter]

0:55:47.8 JA: Exactly. And that's exactly what we're working on now is how do we get that to a hundred. And unfortunately, I don't want to get that, but I'm not sure that the drug companies seem to be happy with 55% or so.

0:56:00.0 SC: Yeah. [laughter]

0:56:01.4 JA: Anyway it gets harder now. But that is the question. To me that's why we have this thing called, we have this institute that's been founded in Anderson and it's whole goal is how do we make that better? And the way you make that better again, is by bringing the myeloid cells in there. What are the ways, what are the things that are now we know, for example, there are a lot more of these things called checkpoints. Probably there's a small number. It's probably a dozen, maybe, I don't know for sure. But did influence the immune system in various ways that they regulate it. And some of 'em only pop up when you take one off. In one sense, it's whack-a-mole, you take one off and another one. 'cause the immune system tries to regulate itself.

0:56:51.1 JA: I the biology's just wonderful, just wonderfully complicated. And the way it all fits together to make sure that everything that can happen, there's a counterweight to it. Multiple ways of built in. But 'cause it doesn't wanna kill you. There are more ways of turning an immune response off than there are turning it on, simply because the consequences of having it work when it shouldn't are too devastating. Particularly if it kills her when you're young. So it's all tuned to do that. But anyway, so one of the cardinal rules and performing drug development is make sure something has a single agent activity before you combine it with something else. That paradigm is shot here because some of those molecules aren't even expressed until you give one of the other ones. And so we've gotta get off of that sort of thing and understand it mechanistically.

0:57:53.0 JA: And so that's, what we're doing is going into humans as soon as we can, as soon as we know it's safe. And giving combinations of getting biopsies and seeing what happened and what new molecules came up. 'cause Now we can measure the advances that have been made in biology in terms of being able to do single cell transcriptomics and knowing every gene that's expressed essentially. And every protein that's made allows us to really understand what's going on. So if we can get biopsies after treatment, we can begin to unravel all this stuff. And so that's what we're doing now, is we try to work stuff out in mice, get a good idea, it works, and then as soon as we can with the help of PAB Sharma who runs our clinical thing, going to patients, do a 12 patient trial safety and mechanism, not necessarily looking for an antitumor effect yet, but just, did we at least not hurt the patient? And did we learn something about the mechanism? Maybe Vista is the name of another one of these checkpoints. Maybe Vista plays a role, maybe the next time we need to add Vista to the cocktail. So we go back and we add Vista then. See if that works. So that's the idea is do this iterative thing. That's how we get it from 55% to a 100%.

0:59:15.9 SC: Right. And do I get, my impression is, or my guess would be that this kind of therapy might also have the benefit that can, thinking of vaccines as an analogy it sticks around in the body and might help prevent things coming next.

0:59:30.1 JA: Yeah. I think in a way, we do ourselves a bit of disservice by calling like the antibody that we made to C24. That's the drug. It's not the drug, it's the thing that, it's the pro drug. It's the T cell that's the drug.

0:59:45.0 SC: Right.

0:59:47.8 JA: And so that antibody's gone in a few weeks, the T-cells there for the rest of your life.

0:59:53.2 SC: Okay. So we're running to the end of the podcast, but, so I'll ask one slightly crazier question. I mean, all these ideas about networks and switches and non-linearities, I'm a complexity scientist among other things. It just makes me think of the study of complex systems. But is your work and the work of other people trying to do what you do? Or is it just so focused on cancer and immunology that you don't have time to?

1:00:18.9 JA: No, I think, no, that's an excellent question. But I think that if we can learn how to reverse this, we could treat autoimmunity. There's increasing evidence that a lot of neurodegenerative diseases involve dysfunctions in the immune system.

1:00:32.4 SC: Exactly.

1:00:33.6 JA: We can really understand what we're doing. We might be able to do something about Alzheimer's down the road or Parkinson's or diseases like that. We're thinking about that all the time, and hoping that, some of the data that we have will just people, it's the very simplest oversimplification. We can just do the opposite of what we're doing now. Maybe we can cure those diseases, and, so that, no our institute is, our goal is to use the advances in basic science to advance medical practice. Right now it's cancer. After we cure cancer, then we'll go after neurological diseases.

1:01:17.6 SC: [laughter] Never satisfied. Well, one of the goals of the podcast is to give young, curious people food for thought about areas that are exciting and changing very rapidly right now. You've certainly done that for us.

1:01:33.6 SC: I think, of course I'm biased, but I think that it's a wonderful time to be doing this work and it's become a systems biology problem more and more. And so single cell analytical techniques the ability to take a slide, now to diagnose a cancer, you get a piece of the cancer and you state in it with hematoxinomy and he said, you get this purplish thing that you could look at shapes. And the pathologist could say, that's a cancer. That's not, there's some immune cells over there. Now we can look at that. And by doing some different analytical techniques, we could tell you every gene that's expressed in that cell. And what we already know is that the same cell that's here next to the cancer cell is gonna be different than an otherwise pretty much identical cell over here that's not in contact with the cancer cell and the immune cell, the same thing. The immune cell that's next to the tumor cell is not gonna be like the immune cell that's in the blood. And those, so now we can start to unravel what are those differences and how can we send the cells down there? We know that there's certain inhibitory molecules that the myeloid cells, particularly those that are in these complexes near the tumor cells expressed that turn the T cells off, if we can figure out how to just turn those molecules off And the myeloid cells.

1:03:00.1 JA: Then, but we gotta know what they are first. I'd bring that up. 'cause I was just discussing with one of my postdocs, two molecules of that sort that we recently found that we're in the process of trying to do that with. But as the science is the engineers and the computational people and all that are coming up with all these magical new things I couldn't even dream of five years ago. That's how fast it moves. There's stuff we're doing now that wouldn't even have guessed that we'd be doing five years ago. That it's just such a fascinating time to be in biology. And I know my motto is a, I mean, I've always been fascinated in science and biology and everything is just as a, you might as well do something on the area of a, something helps people.

1:03:52.6 JA: Yeah. 'cause, I got into it initially because I just, my family situation. But on the other hand, it's fun and it's rewarding and it helps people too. And so I think that, it's a lot of it's drudgery too. But it's just ex puzzle solving. And with the tools that we're getting that are coming from bioengineers and computational people are just allowing us to ask questions that we couldn't even dreamed of.

1:04:27.5 SC: It's exciting times. Absolutely. Jim Allison, thanks so much for being on the Mindscape podcast.

1:04:31.7 JA: Thank you. By the way. I see you have a guitar. Do you play much or do you.

1:04:36.4 SC: That's a bass guitar that I'm very, very, very bad at. If we had more time or if you wanna take another five minutes, I was gonna say like, tell us the Willie Nelson story. Come on. It's so good.

1:04:48.4 JA: Oh, I'll tell you one funny one. I had a knee replacement last what was it, last year? Yeah. Anyway, and Willie, I played with him occasionally, but he they asked me to come to Austin. I'm in Houston but his 4th of July picnics in Austin. So they said, would you come down? He didn't, but his wife and his harmonica player, a friend of mine said, Jim, come down, join us. I said, I can't, I just had an knee surgery and I don't want to go down there. They said, well, we just happened that they were playing in this place called the Woodlands, which is just north of Houston, near this lake where I have a lake house that was recuperating from the surgery there. They said, well, hell, we're gonna be there on Saturday night, so why don't you come and play there?

1:05:35.0 JA: I said, well, I can't walk very well. They said, well get a wheelchair. You don't have to travel. So I said, okay, okay. So my son got a wheelchair. We got some friends. We went down, wheeled me in. We were still on the edge of the stage. And so I was waiting to play, but I still wasn't sure I could go out there. And one song came by and they said, come out. And I said, no, I can't stand yet. I had a, I was in a wheelchair. I had a cane. Anyway, Willie got into the gospel. He always closes his concerts with a gospel medley. I saw the light, will the circle be broken? I'll fly away. All the great old songs. Anyway, so that's so much fun to play. And he said, Mickey, waves me out there because they had a microphone for me.

1:06:23.9 JA: And I said, no. And Willie's wife got Bobby. He said, Allison, get your out. And so anyway, so I jumped up and took a couple of steps before I didn't even think, I just stood up and I had a cane in one hand, a harmonica in the other. And then I realized this isn't gonna do because I need both hands when I got out there. So I turned and I threw my crane behind me and then turned back around and took a couple of steps towards the mic. And this woman screamed, he's healed. Willie healed him. Praise the Lord.

1:06:55.9 SC: Evidence [laughter] It was not a double blind study, but some data there.

1:07:01.7 JA: Everybody starts clapping, Willie, starts singing I saw the light. It was just so perfect [laughter]

1:07:07.0 SC: Congratulations on being the recipient of a miracle. It sounds like you deserved it there, [laughter] All right. Once again, Jim Allison, thanks very much for being on the Mindscape podcast.

1:07:13.6 JA: Okay, Bye-Bye. Thanks.

[music]