The Next Wave in Checkpoint Inhibition

Transcript:Mark A. Socinski, MD: That actually brings up the next point that I want to make, and I’ll start with Howard here. We’ve had this tsunami of CTLA4 and PD-L1 inhibitors. Then we start to hear about things like LAG-3 and OX40. What’s the next wave going to bring us?

Howard L. Kaufman, MD: Well, I think these checkpoints really block the inhibition of the T cell and keep them alive, and the next rounds are going to be focused on the agonist antibodies that really stimulate the T cell. So, things like OX40, 4-1BB really work to give the T cells a signal to stay on and in many cases, to go on and enter into a more memory-type phenotype. I think these are going to be quite powerful potentially, especially when you can combine something that turns the T cell on with something that prevents it from coming off, I’d be very careful about toxicities that we may not be able to even anticipate, given what we know now. But, I think that’s going to be the future. And these studies are beginning to be reported already this year.

Mark A. Socinski, MD: John, I want to ask you because I think one of the other unmet needs that we’ve talked about is that right half of the curve where there are more people alive. We know, at least currently, that the immunotherapeutic drugs are not great answers for every patient. And we’ve talked a lot about how we can develop strategies to increase the percentage of patients. But, the other approach is to identify patients that are getting the real benefit, and it gets back to biomarkers. We talked about PD-L1 status, and that’s currently a complementary diagnostic. But, we hear about things like mutational load, immunoscore, and all these other things. What’s your perspective on where we may be 2, 3 years from now?

John V. Heymach, MD, PhD: I think, in some ways, we’ve been spoiled by the success of some genomic markers that are absolutely clear-cut—for example, starting with the BCR-ABL, where it’s the target and it’s the biomarker for the drug, moving on to KIT in GI stromal tumors, and then EGFR and ALK in lung. If you use that as a standard for a biomarker, where if you’ve got that target, you’ve got a 70% or 80% chance of responding and if you don’t have that target, you have almost a 0% chance of responding, I think we’re going to be disappointed in immunotherapy biomarkers for a long time, because it isn’t a single driver oncogene. It’s a more complex environment. So, I think biomarkers are going to come along more slowly, and they’re going to be more multidimensional. And by multidimensional, we already believe there are several different influences both in the microenvironment and in the tumor cell. To give you some examples, work from Naiyer Rizvi and others has shown that the number of mutations in lung cancer correlates with better response. This was shown in melanoma and other diseases.

But, we’ve actually seen independently—and this is something just reported from Don Gibbons and the rest of our group—that epithelial mesenchymal transition (EMT) status, where a tumor undergoes a change where it adopts a more aggressive phenotype, actually turns up its checkpoint factors and it suppresses the immune system. And mutation burden and EMT status are completely independent of one another. But, they’re both associated with response to checkpoint factors, or at least levels of checkpoint factors. And I mentioned before that there are some tumors that have a high mutation burden, but are poorly responsive to immunotherapy.

So, really, we have to be agnostic and to not think we know more than we do about this, and think that it’s settled. What I’d say so far is just starting with the simplest PD-L1, it’s a pretty good biomarker. It’s widely criticized. There are a lot of different ways to measure it. But, no matter how you measure it in everybody’s studies, more of it is associated with more benefit.

Mark A. Socinski, MD: It makes a difference.

John V. Heymach, MD, PhD: But, we still can’t distinguish just what you said: who is going to be the long-term survivor and who isn’t. This is going to be a long-term problem, and there’s not going to be a quick solution with one marker. But, I think there are already three or four different directions that are going to be promising to pursue.

Transcript Edited for Clarity