In 2018, James P. Allison, PhD, won the Nobel Prize in Physiology or Medicine for his lifetime of contributions to the treatment of advanced cancer patients. His discovery of the CTLA-4 immune checkpoint led to the development of the anti-CTLA-4 drug ipilimumab (Yervoy®), the first FDA-approved checkpoint blockade therapy. Today, the drug combination pairing ipilimumab with another checkpoint blockade drug, the anti-PD-1 agent nivolumab (Opdivo®), has produced longer survival for advanced melanoma patients and many other cancer patients than any other treatment in history.
Dr. Allison is chair of the Department of Immunology, the Vivian L. Smith Distinguished Chair in Immunology, director of the Parker Institute for Cancer Research and the executive director of the Immunotherapy Platform at The University of Texas MD Anderson Cancer Center in Houston. He is widely regarded as the father of checkpoint blockade immunotherapy.
At this year’s ASCO annual meeting, Dr. Allison offered a presentation on why some patients don’t respond to checkpoint blockade immunotherapy and what further investigation needs to be done for these patients. Mark Teich interviewed Dr. Allison about the most important trends in research for the future treatment of advanced melanoma.
Mark Teich: I want to congratulate you, of course, on the Nobel. What a wonderful achievement, reflecting a lifetime of work. But your time is valuable, so let’s cut right to it.
Even with the immense improvements in survival produced by the checkpoint blockade therapies in recent years, only about half of treated patients with distant metastatic melanoma are alive at the four-year mark. Yet in your presentation, you expressed the belief that these therapies are going to remain the most important focus of research, taking us closer to 100 percent survival. Why have these therapies had greater success than any others, and why do you believe they will continue to be key?
James P. Allison, PhD: The immune checkpoints were a complete paradigm shift. They gave us insights into entirely new therapeutic mechanisms. It’s these different mechanisms of action that have made the difference and will continue to make the difference. In the past, cancer therapy was always built around personalized driver mutations, creating drugs such as chemotherapies or the targeted BRAF-MEK therapies to inhibit the drivers. The problem with these treatments is that, given the heterogeneity of tumors, by the time you detect one driver and target it, other drivers almost always arise. So, there’s almost always disease relapse. For example, the limitation with the BRAF-MEK-targeted therapies is that by the time you reach stage IV, even stage III, the cancer cells are so genetically unstable that they already have varied driver mutations; or, the tumor cells have sort of figured out how to not be dependent on mutant BRAF anymore, so you have to start looking for another drug to block another driver mutation. That’s the limitation of the therapies targeted against driver mutations, because in virtually all cases, the cancers eventually come back.
Immunotherapy is different. The immune system doesn’t care if you handle the drivers or not; you’re in a sense ignoring the cancer cell. With checkpoint blockade, you aren’t targeting the drivers but the immune system itself. You’re removing the brakes on the immune system — not harnessing it so much as unleashing it. You’re releasing hundreds of millions of T cells. And the wonderful thing is that once you’ve turned on the T cells, you’ve got them for life. They keep being re-created.
MT: You also expressed the somewhat surprising belief that the anti-CTLA-4 checkpoint and the anti-CTLA-4 therapy ipilimumab, rather than anti-PD-1 therapy or other new therapies, is going to prove to be the most important bulwark of advanced melanoma treatment in the future. How so?
JA: Well, I have to admit that I’m not entirely unbiased. Of course, since the PD-1 checkpoint was discovered and the anti-PD-1 drugs and anti-PD-L1 drugs were developed, they are much safer drugs than ipilimumab; the adverse events are similar, but they’re less frequent with the anti-PD-1 therapies. And other types of drugs have been developed that are even safer, but they work very differently and they remain to be proven effective.
Despite all the new drugs being tested and new checkpoints explored, it’s the combination of ipilimumab and nivolumab that has had the best possible results to date. Nivolumab and the other approved anti-PD-1 drug pembrolizumab (Keytruda®) have had very good results as monotherapies, and if you place their characteristics side by side against ipilimumab, it obviously favors the anti-PD-1 therapies. But it’s combination ipilimumab-nivolumab that has the greatest durability and appears to be saving the most lives. It has had the greatest success of any two-way combination therapy. Furthermore, algorithms have been developed that are becoming increasingly successful in dealing with the adverse events, as long as you pay close attention as treatment goes on; so more and more, the combination is prevailing.
Another important thing about ipilimumab is that it’s given only four times in the course of therapy, and that’s the end of it. The 10-year follow-ups on ipilimumab as monotherapy are just appearing now, and they show that after just those four treatments, 20 percent of patients are alive 10 years after you stop therapy, essentially cured. The anti-PD-1 drugs, on the other hand, are given every two weeks for a long time, until progression or until the side effects are too bad. That appears to be important, because the T cells are getting exhausted and you have to present the antigen again.
When you put anti-CTLA-4 and anti-PD-1 together, the number of T cells actually decreases from what you had in monotherapy, but the nonexhausted cells go way up. These nonexhausted cells get activated, and you’ve lost the exhausted cells. We’re not clear why, but you’re reversing the exhausted state. You’re radically changing the mechanism of action, and exhausted cells are no longer important. It’s an epigenetic change in the cells. You’re teasing out the nonexhausted T cells, and they may never get exhausted. The combination is key; by targeting multiple checkpoints, you can get durable remission.
To me, the 10-year results on ipilimumab are perhaps the most exciting news at the ASCO annual meeting this year, because they probably mean that all the approved checkpoint drugs will plateau, going 10 years and beyond, and patients who respond for that length of time are essentially cured. The 20 percent of patients responding at the four-year mark after treatment with ipilimumab are still responding 10 years and longer, and while the anti-PD-1 drugs haven’t been around long enough for 10-year results, the same will probably be true with the anti-PD-1 monotherapies as well as combination ipilimumab-nivolumab, meaning that 50 to 60 percent of patients on the combination will be cured. The combination has an additive response rate. So far, the responses with the combo therapy are holding at about four and a half years, and the chances are they’ll be holding at 10 and beyond in most patients. I think we can consider many of these patients cured.
MT: How can we get durable responses in more than 50 to 60 percent of patients?
JA: That’s the goal, to get the percentage as high as we can. With the basic standard of care now leading to about a 50 to 60 percent durable response, and with new combinations and other checkpoints being explored, we want to get as close to 100 percent as we can. We’ll be combining the new agents with those we already have. That’s the focus now, adding third or fourth drugs to the combinations we already have.
MT: Which of these new agents holds the most promise?
JA: There are a lot of candidates. But the goals have shifted: We’re no longer looking for a single agent home run as we always did in the past; we’re beginning to realize the importance of just adding 5 to 10 percent here and 5 to 10 percent there to the response rate, without toxicities. A key factor we learned along the way is this: When patients didn’t respond to ipilimumab, one of the reasons was that there are multiple checkpoints. So, for example, many who didn’t respond to ipilimumab then responded to the anti-PD-1 therapies, which blocked a different checkpoint. There are also a lot of other reasons patients may not respond, but this is a very fundamental reason. The more checkpoints you uncover and block, the more T cells you have to fight the tumor. And one simple way you keep increasing durability of response is by putting these checkpoint blockers together. I think checkpoint combinations are going to become the standard of care for almost everything. Melanoma treatment is leading the way, but it’s going to be the same sort of situation with other cancers, for example, kidney cancer, where the response rates are still under 50 percent.
MT: When you say that checkpoint combos are going to be the standard of care, are you talking specifically about ipilimumab combined with anti-PD-1 and other checkpoint blockers, or entirely different checkpoint combos?
JA: I think that combos of anti-CTLA-4 with anti-PD-1 drugs, with new agents added to them, are going to be the key. People disagree on the number of checkpoints out there — it depends on how you define checkpoints — but given the data to date with melanoma and other cancers, those are the two most important ones. There are others that may play another role, depending on the type of cancer, but those are the two most important in melanoma. For example, one relatively new molecule called VISTA, which is PD-1-like, is a newly discovered checkpoint present in a lot of patients with gastric cancer, but not so much in melanoma. So I think what we’re trying to do, rather than just find new blockbuster checkpoints, is find ones in different tissues and just add antibodies to block those. You might have a cocktail of an anti-PD-1 drug and an anti-VISTA drug, or another drug blocking a different checkpoint, to get better and better response rates.
There are also many other agents out there and other checkpoints, but they’re still in testing and not widely adopted yet. I also think that combining the checkpoint blockade therapies with BRAF-MEK therapies may end up a pretty amazing combination. It’s just possible that these drugs can inhibit melanoma cells during their proliferation phase, but we’ll have to design a strategy, figure out some different way of combining them or sequencing them than we have now, since now they interfere with one another in some ways. We need to make sure these additive therapies enhance rather than impede the immune response. Also, there’s some liver toxicity now.
It’s similar with other techniques being combined with immunotherapy, including radiation and chemotherapy, which are already proving very promising. What’s happened is that people who do radiation therapy and/or some form of chemotherapy are now realizing that you don’t have to kill every last tumor cell in the patient, you just have to kill enough to cause inflammation, which kicks on the immune system, and the immune system may take out the rest of the cancer. You just need to find the right dosage and means of administration of whatever it is so that it keeps toxicity down and enhances rather than impedes the immune response. With radiation, for example, you may kill a bunch of tumor cells, and then you come back a week later to kill some more, and immune cells are already priming T cells where the tumor was, or let’s say where it still is. Its size has been reduced, and then you do another bolt of radiation, and you kill more tumor cells, but you’re also killing the immune cells there. So, it’s just a matter of figuring out how to do it in a way that it works against just the tumor and not against the immune system. There are pretty interesting results coming out of radiation therapy, and they’re taking a careful look at what it does to the immune system.
MT: Will we ever assemble combinations that will achieve a 100 percent response?
JA: Well, with melanoma we’re already getting near 60 percent, which isn’t 100, but it’s a lot better than 5 percent or whatever it was when we started. I would like to think of 100 percent as the goal, but there can be a limit to how many drugs you combine. You can’t necessarily put 13 different things together. So I think there will be some limitations.
One thing that bothers me is that there are so many possibilities with so many combinations, so many diverse trials being run, and they’re repetitive, and many aren’t adding anything new. Even worse, there are some pivotal trials with, say, 100 or more patients, which is a lot, and some aren’t being controlled well enough to leave a meaningful trail. In my opinion, if you do a trial with any kind of cancer right now, with any kind of treatment, and you don’t do biopsies, it’s just a waste. Whether something meaningful happens clinically or not, you can learn from it — you can learn from what doesn’t work as well as what does work. You need to be looking for any changes in the tumors. We run a big lab at MD Anderson called Immunotherapy Platform, and that’s one of our goals. On every trial we’re involved in, we’re looking at every clinical signal, every potential biomarker of success or failure, every change that occurs with therapy that might indicate if we’re going in the right direction. And it’s been useful so far.
MT: How long do you think we are from going the rest of the way, close to 100 percent responders?
JA: You just keep adding things to the mix, and you keep trying. I’d like to think of it as a 10-year plan.