Jim Allison believed in the power of T cells—when hardly anyone else did

Throughout Jim Allison’s early career, the field of oncology appeared to be unfazed by his T cell research.

Allison’s introduction to the now-famed immune cell came in an undergraduate immunology course, where even the professor was apprehensive about Allison’s fascination with the recently discovered T cells.

“I remember going to see the professor after class and asked, ‘Well how does that work? How do they recognize that something is wrong?’” Allison, a Nobel laureate whose research led to the discovery of the T-cell antigen receptor and the anti CTLA-4 antibody ipilimumab, and director of James P. Allison Institute, Regental Chair of Immunology, and vice president of immunobiology at MD Anderson Cancer Center, said to The Cancer Letter. Allison is also the Olga Keith Wiess Distinguished University Chair for Cancer Research and executive director of the Immunotherapy Platform.

“And he said, ‘Well, nobody knows.’ He didn’t even think there was such a thing as T cells, because he was an antibody guy, through and through. He just thought it was some kind of weird macrophage that picked up an antibody molecule.”

Allison was not convinced. He wanted to understand how these cells worked.

“At the time, the T cell receptor hadn’t been discovered. The thing that T cells use to recognize things had not been discovered. I mean, there was a lot in the literature, but it was all over the place and confusing, with a variety of different structures, and not particularly informative from what I could tell,” Allison said. “So, I set out to do that, and ended up working out the structure of the T cell antigen receptor.”

He continued to work with T cells, but his transition from fundamental immunologist to Nobel Prize-winning tumor immunologist came from the discovery of CTLA-4’s therapeutic potential.

Allison and his team had been doing research with CD28, a co-stimulatory molecule that he describes as the “gas pedal” of the immune response. He became interested in a highly homologous molecule, CTLA-4, which others in the field believed to be another “gas pedal” molecule.

Allison’s experiments showed, however, that CTLA-4 was actually a “brake” to stop immune responses. With this discovery, the pieces began to fall into place.

Data speaks to you. Design an experiment right, you’ll get a good answer, but you’re never going to get a perfect answer.

“By then we’d done a lot of experiments in the lab, showing solid tumors were invisible to the immune system,” Allison said. “It occurred to me that, maybe because of that, the tumor got a head start. The immune system was late to catch up.

“Unlike viruses and bacteria and things that the innate immune system picks up right away—they have those second signals that say, ‘Go. This is foreign. Do something about it.’ Tumors are a little less obvious than that.”

Allison and his team quickly began doing experiments on CTLA-4. All they had to do was inject antibodies to CTLA-4, and “the tumors just melted away.”

Despite the experiment’s success, Allison’s colleagues were not enthusiastic.

“We presented the data at the International Congress of Immunologists in San Francisco in 1995 and it was greeted by a very wide and mediocre yawn by most of the science community. ‘Oh yeah, somebody else curing tumors in mice.’”

It was a matter of bad timing.

“Unfortunately, the world of cancer therapy had sort of passed through the immunology phase that was popular in the 60s and 70s,” Allison said. “There had been so many failed trials, particularly with vaccines, that people began to think tumor immunology was strictly a property of mouse models and would never work in humans. The net result was me shouting into an empty room most of the time for years.”

Allison trusted his data, and it led to a novel treatment that is life-altering for many.

“Data speaks to you. Design an experiment right, you’ll get a good answer, but you’re never going to get a perfect answer.”

The important thing is to stay open-minded, Allison said.

“No scientists should ever say the words, ‘It never works or it never will.’ I mean, that’s what a scientist is about; right? Moving forward, making things, finding out something new—where you can make something happen that never had before,” Allison said.

Jim Allison spoke with Alexandria Carolan, associate editor of the Cancer History Project.

Transcript

James Allison: When I was little, I saw a lot of cancer in my family. My mother died of lymphoma when I was about 10. In the next few years, two of her brothers were diagnosed, one with melanoma, the other with lung cancer, and both passed away at that time.

Since then, there have been several other instances, including both my brothers and myself. A lot of cancer history.

I was always interested in biology, and my dad was a doctor—a country doctor in a little town in south Texas. So, I kind of gravitated towards medical issues. He wanted me to follow in his tracks.

But I kind of got sidetracked. I got really interested in doing research. When I was in high school, I was part of a biology program sponsored by the National Science Foundation, at the University of Texas in Austin.

At the age of 14, I attended college lectures in biology and worked in a lab, which pretty much did it for me. It was so much fun. I just decided—no pre-med. I’m going to go into science.

I was doing biochemistry because I really liked chemistry. One project was related to cancer therapy, even then. It was when asparaginase was first being used for the treatment of leukemia, so I embarked on a project to try to identify multiple sources that can be used as they became immunogenic, which was a real problem at the time.

During that time, I took a course on immunology, and T cells had just been discovered. We went through antibodies and that was cool—B cells secreting antibodies, and connecting with things with high degrees of specificity—that was a well-developed science, but T cells were not.

My professor just taught that T cells go through your blood, through the lymph nodes, and percolate through your tissues to look for viruses, infections, or maybe even cancer, and do something about it.

I remember going to see him after class and said, “Well how does that work? How do they recognize that something is wrong?” And he said, “Well, nobody knows.”

He didn’t even think there was such a thing as T cells, because he was an antibody guy, through and through. He just thought it was some kind of weird macrophage that picked up an antibody molecule.

He just couldn’t imagine that nature would have solved that problem of specificity twice.

So, I decided to focus on immunology, and I left to do a postdoc in a lab where I actually learned some immunology. After my postdoc, I took my first job, which was with MD Anderson. Not at the main Houston campus, but at a new lab near Austin, called Science Park.

There, I used immunological approaches to make monoclonal antibodies to detect surface changes that might have occurred.

There were some very well-characterized rat models with liver carcinogenesis, where you could go through a hyperplastic phase, and then a neoplastic, and then frank neoplasia and metastases. I succeeded in making monoclonal antibodies to find specific proteins to all the different stages, but to me, that wasn’t really immunology.

At the time, the T cell receptor hadn’t been definitively discovered. The thing that T cells use to recognize things had not been discovered. I mean, there was a lot in the literature, but it was all over the place and confusing with a variety of different structures, and not particularly informative from what I could tell. So, I set out to do that, and ended up working out the structure of the T cell antigen receptor protein.

I published this side project and it got a lot more attention than the other work. Then I received an offer to join the immunology program at The University of California at Berkeley as a full professor.

JA: Everything. T cells have the ability to go through your body and tell exactly what’s the problem—not getting hung up on stuff that’s not a problem—and then they change into something that actually can solve the problem.

Back then, we knew some things—that they made IL-2, and a few really crude things like that—but not the full gamut of what they can do.

So, just the possibility of all that, the mystery of how it’s controlled and how it’s regulated. These were as interesting a fundamental biological question as there was, you know?

When I went to Berkeley, there was a strong contention of faculty that thought only first principles were important—like, how does DNA reproduce itself? How do you make proteins?

Those were the things we ought to be asking about, not something so mundane as how the immune system works. The feeling was that it really wasn’t science, that it was more applied, somewhat suspect – particularly when applied to cancer research.

JA: Well, I think the fundamental work was well received. Others had shown that there was a costimulatory signal that was necessary which proved it wasn’t just a switch that you flipped.

We knew about a molecule called CD28, and there was this other interesting molecule that was highly analogous to CD28, called CTLA-4, and nobody knew what it did. We showed that we were behind on that one.

Other people thought that it was another sort of gas pedal-like molecule like CD28,but I saw some holes in the data and I wasn’t completely convinced.

When Max Krummel was a graduate student in the lab, we decided to do the extra experiments to explore the possibility that CTLA-4 was a gas pedal. But this work showed that it was actually a brake – stopping immune responses.

That immediately caused me to think, “Well maybe the reason that the immune system isn’t better at dealing with cancer is because the cancer does not have the wherewithal to give that co-stimulatory signal.”

By then we’d done a lot of experiments in the lab, showing solid tumors were invisible to the immune system. It occurred to me that, maybe because of that, the tumor got a head start. The immune system was late to catch up.

Unlike viruses and bacteria and things that the innate immune system picks up right away—they have those second signals that say, “Go. This is foreign. Do something about it.”

Tumors are a little less obvious than that.

So, we quickly moved on to doing experiments with anti-CTLA-4 antibodies. When we injected these into mice with tumors,, the tumors just melted away.

At the time, I didn’t really consider myself a tumor immunologist. I was more of a fundamental immunologist dabbling in tumor immunology. We played around a little bit with using ligands for CD28 and, of course, vaccines and things like that. They worked okay, but it wasn’t anything earth shattering.

But CTLA-4 really was. I mean, we could inject it to well-grown tumors, and they just melted. At that point, I guess, I became a full-fledged tumor immunologist.

JA: We presented the data at the International Congress of Immunologists in San Francisco in 1995 and it was greeted by a very wide and mediocre yawn by most of the science community. “Oh yeah, somebody else is curing tumors in mice.”

So, we worked for the next three years making sure it wasn’t a fluke, by doing it in different strains of mice and with different kinds of tumors.

We found that sometimes anti-CTLA-4 treatment worked by itself. A lot of tumors considered to be strongly immunogenic—ones with higher mutational loads, that were more consistent with human melanoma, lung cancer, bladder cancer—we could cure those tumors with anti-CTLA-4 treatment alone.

Poorly immunogenic tumors you couldn’t really immunize—even with kill tumor cells or lysates of tumor cells—you couldn’t even prophylactically immunize. For these, we tested combination therapies and could cure the cancer when combining anti-CTLA-4 with chemotherapy or radiation.

Unfortunately, the world of cancer therapy had sort of passed through the immunology phase in the 60s and 70s. There had been so many failed trials, particularly with vaccines, that people began to think that tumor immunology was strictly a property of mouse models and would never work in humans.

By then, the Genome Project had come along and driver mutations were discovered. So, the idea of making TKIs—that was felt to be the cure.

Everybody that I was around said, “Oh, well, that’s the cure to cancer. Forget immunology. It’s not going to work.”

So, for three-and-a-half years, we tried to find a biotech company. If I had it to do over again, I’d find somebody to fund my own company, but I really wasn’t interested in that then. I had confidence that somewhere out there, somebody would believe in the science.

The net result was me shouting into an empty room most of the time for years.

JA: Yeah. And then luck would have it, I complained to a friend of mine named Alan Korman, who had a friend named Nils Lonberg, the founder of GenPharm, which had been bought by Medarex.

He had mice that you could immunize. At the time, when you made an antibody, you’d make them in mice. You can’t inject mouse antibodies into humans, because it will make a mouse protein response and the drug won’t be useful for very long, so you have to have a human antibody. So, people were humanizing mouse antibodies, which is a tedious process that can take a long time.

Nils had mice that had their immunoglobulin genes replaced with human genes. So, right out of the mouse, you’ve got a fully human antibody.

JA: He didn’t really have any applications for the mice, so it was an ideal match. They made an antibody in no time and we were off to the races.

There was a lot of interest in it, first through the Ludwig Cancer Research Institute, from an academic pursuit.

Medarex sought out people like Lloyd Old and some of his affiliates at Sloan Kettering, like Pam Sharma.

The real telling thing, though, was a phase I trial in melanoma. At the time, the mean survival after a melanoma diagnosis was seven months and fewer than 3% of people lived for five years.

But, three patients in this phase I trial had complete responses.

JA: Bam.

And that was a safety trial. It wasn’t even intended to look for efficacy.

JA: I heard about it a roundabout way from Medarex.

But I said, “Wow. Okay. We’re on the right track here. I better take this seriously.”

What dawned on me, at the time—as a scientist and learning a little bit more about cancer—I realized I better learn something about clinical trials. In cancer, people were used to drugs that attack the cancer cells, but my drug had nothing to do with cancer at all.

It was directed against a molecule in the immune system, released T cells to go after the cancer cell. I was afraid that was such a different paradigm for therapy, that it would be really easy for it to get screwed up in a trial.

So, it happened that Harold Varmus took over as head of Memorial Sloan Kettering and wanted to rebuild their immunology program. I’d had a pretty good record at Berkeley of rebuilding their program, so he asked me to come to Sloan Kettering.

So, I thought, “Yeah. That’s where all the trials are—in New York. I better get there.”

It was a lucky thing. I was embedded right where a lot of the work was going on. By then, Bristol Myers Squibb was collaborating with Medarex and had decided to co-develop the antibody and I was brought on as a consultant.

JA: It was kind of overwhelming for a while. I got to Sloan Kettering and at least 90% of people there were into TKIs and genomically-targeted therapies, not immunology.

Still, some began to change their minds and realize, “Wow, there’s something here.”

So, that was encouraging.

JA: Some did. But melanoma had a reputation as being immunologically responsive anyway. Also, patients occasionally would have a spontaneous remission that would be accompanied by a big pigmentation that was immune-mediated.

They said, “Oh well, melanoma is an inherently immunological disease.”

So, a lot of people came around but still said, “You’ve got a melanoma drug, but that’s it. It won’t treat anything else.”

JA: So, even in 2011 when ipilimumab was approved, I got pats on the back and they said, “Well, you cured melanoma, but it’s not going to cure anything else.”

JA: I think lung and there were positive results in small trials with bladder.

Of course, the situation shifted because the response rate of this monotherapy with melanoma was only a little over 20% at this point. But, it was clear right away that those people had very, very durable responses—20 years, we now know. Some of the people from the phase I are still alive.

People began to ask, “Well, it’s only a fraction. Why isn’t it 100%?”

One reason is because there might be other checkpoints.

So, then Tasuku Honjo teamed up with Arlene Sharpe and Gordon Freeman, two immunologists at Dana Farber, and they showed that the molecule he had been studying, programmed death 1 (PD1), was actually another checkpoint with a different mechanism of action.

We’d shown CTLA-4 works during the initial priming of T cells, whereas PD1 works on late-stage effector cells. Anyway, it was shown to be effective and, since it worked on already committed T cells, it worked a lot faster.

By 2014, studies were showing that the combination of those two drugs was more powerful than either one alone. It is now the gold standard for a lot of cancers, including melanoma. We now have trial data where 60% of patients with melanoma are alive after five years. 60%. Not 3%, 60%.

JA: And they are still going to be alive in 10 years too, you know?

JA: I stayed at Sloan Kettering as it was beginning to explode. But then I decided to learn what Pam Sharma was doing—mechanistic-based trials. It became clear that it would be a lot better if we really understood how these drugs worked in people, instead of just assuming we knew something from mice.

There’s never been a better time to be in biomedical research and in cancer research. For the first time, we can use the words cure and cancer in the same sentence—for a fraction of patients, to be sure—but it’s beginning to be within reach.

So, I moved to MD Anderson. Together with Pam, we set up an immunotherapy platform to do just that.

We realized our immune system is very, very complicated. It has to be carefully balanced by a combination of positive and negative signals to keep it under control. Otherwise, it can really cause a lot of damage.

We now know of a dozen or more checkpoints. CTLA-4 and PD-1 are the two major ones, one working very early and one very late in the immune response. But there are a lot of others that play important roles in myeloid cells, helper T cells or regulatory T cells. Targeting these checkpoints may not elicit a response on their own, but it could give you an extra 10% boost when targeted in combination.

I think that’s how we’re going to get from 60% up to 80% or 90% in melanoma. We’re going to have to add more things in rational combinations. And we’re going to have to change the way we do the trials to see these improvements. I think that the field is maturing, now, where people are beginning to see that all checkpoints aren’t created equal.

There is a place, I think, for conventional therapies – chemo, radiation, surgery and targeted therapies. Not as curative agents, necessarily, but as part of a combination attack on the tumor.

These treatments can cause some antitumor effects to prime an immune response. By using them essentially as vaccines in conjunction with immunotherapy, it lets the immune system do the heavy lifting.

JA: The time that I felt I was on to something was very clear—when my postdoc did the experiment, came back, and told me that all the tumors went away.

“What the hell? I’ve never seen that,” I said, “Set it up again, double blind this time. Set it up. Don’t tell me what’s what.”

I went back and measured the tumors myself. I went through this educational process, where I saw that all the tumors grew at the same rate.

I had multiple cages of mice with tumors, and they all looked the same—until they didn’t.

In some of the mice, the tumors stopped growing, and then ulcerated and went away. Those mice, six months or so later, we challenged them again and they rejected the tumor. So, I said, “We’re onto something here.”

That’s when I knew. I never doubted after that. I thought I might go to my death being one of only a few people who believed in this work.

It was amazing to me at the time that scientists were so set in their ways. Immunotherapy had its day and it didn’t work, so everybody walked away from it. The TKIs were the way to go until they weren’t curative, and people began to walk away from them a little bit.

What bothered me, and still does, is when people should know better. When scientists would say, “Oh, that’s never worked and it never will.”

No scientists should ever say those words. I mean, that’s what a scientist is about; right? Moving forward, making things, finding out something new—where you can make something happen that never had before.

I mean, isn’t that what science is? It is to me.

If you ever say, “Oh that’s never worked and it never will,” then you ought to do something else.

JA: First of all, look at your data very closely. Learn it. Don’t just say, “Oh, I made a hypothesis. Here’s the data that supported or didn’t support my hypothesis.”

That’s a very short-sighted way of doing it. Look at your data closely, because it may tell you something that had nothing to do with your hypothesis at all.

It’s very seldom that you get an answer that’s a crisp yes or no—particularly an important one—where there isn’t already a lot of knowledge. You get something that’s close, but not quite there.

Nature doesn’t move things between boxes, as we’re learning now with single-cell RNA sequencing, for example. We can’t say, “All T cells fall into five boxes.” There’s more than that. The myeloid cells aren’t two or three boxes, they seem to be a continuum.

Biology changes as it needs to deal with problems, particularly the immune system. So, we need to be more flexible in our thinking. We should not have sharp edges and sharp boundaries, and not have a short-sighted look at our data.

I like to look at the data and first ask “Did it support my hypothesis or not?” And then look at it and say, “Well, maybe it’s not quite there, but what else is there? What else is it telling me?” Really study it for a while before moving on to the next experiment.

That’s the main thing. Data speaks to you. Design an experiment right, you’ll get a good answer, but you’re never going to get a perfect answer.

Nature is more complex than we can imagine. Every time a new technique comes along, it will enable us to look at more of the complexity.

That doesn’t mean that everything is hopeless. It means there’s more to learn. More shots on goal, if you’re trying to figure out some way to treat cancer. And cancer has its own tricks.

JA: Nope. I didn’t know what the field would be now, to tell you the truth. It’s been one revelation after another. Over the years, I continue to learn new things and be surprised at where it’s going.

I’m about to move into a new era. MD Anderson launched the Allison Institute in 2022, which is going to take this research to the next level in a data-based way: small trials where we look for mechanistic signals rather than clinical signals, and where we use the clinical techniques I’ve talked about, that Pam Sharma developed.

JA: There’s never been a better time to be in biomedical research and in cancer research.

For the first time, we can use the words cure and cancer in the same sentence—for a fraction of patients, to be sure—but it’s beginning to be within reach, I think.

I think that we got the basics down. There’s still a hell of a lot to learn and to do and it’s going to be a different field in five years than it is now, I’m sure.

But after a long time of nothing really working too well, we’re at the point of being able to have some successes. Now is the time to just build more on those.

That’s where the Allison Institute comes in. We believe that gaining a complete understanding of immunobiology and integrating immunotherapy effectively with conventional treatments will allow us to bring the benefits of immunotherapy to many more patients.

We’re not looking for incremental advances—we’re aiming for cures for as many patients as possible.