Everett Vokes, MD: Roy, we increasingly have patients who progress on a PD-1 or a PD-L1 inhibitor. What do you do then? I guess Lung-MAP covers that a little bit, but can you expand on that a bit?

Roy Herbst, MD, PhD: Yes and no. I was thinking about this with some people this morning. You look at the phenomenal progress we’ve made in lung cancer. Five years ago, to think that we had immunotherapy? And now we’re using it in pretty much all patients, and they’re all benefitting (to some extent). But what we have to remember is that it’s really only about 10% of the patients who have those amazing responses, and many have activity and then they progress. And the problem is, what do we do in that setting? That really is a challenge, and I think we all face that issue.

Certainly, you have to look at how someone has progressed. There are different groups. If someone receives an immune-oncology (IO) therapy and they haven’t benefitted at all—in fact, their tumor’s progressed—I think that’s one group versus a group of patients who might have stable disease or minor response, or an even response, and then they progress. So, I think you have to look at it in 2 different groups.

For the patients who get a PD-L1 inhibitor and they progress rapidly, the standard there would be to give chemotherapy. Interestingly, we’ve seen (and there were reports at last year’s ASCO Annual Meeting) higher than expected response rates for certain chemotherapies in that setting—mechanism not completely known, but certainly chemotherapy would be worth a try. We also talked about how chemotherapy plus IO combinations might be reasonable, and I think that’s important.

Then, of course, we may consider some sort of immune combination. What we like to do at Yale is we like to rebiopsy those patients, and we’re in the process of trying to adapt trials based on what we see on that biopsy. But, in common practice, I think chemotherapy or some sort of early phase trial is reasonable.

Then, of course, there are the patients who have had some response and then they progress. And for that group, that’s where I think Lung-MAP will focus most on. And there you have to ask, why is someone not responding? Is it because PD-L1 doesn’t drive their tumor? That’s probably about 25% of the cases, and that’s where there might be another checkpoint in play—let’s say LAG3, TIM-3. Maybe they’re part of the 40% group where the tumor is noninflamed? We described that in a paper, not long ago, as the “immune desert.”

Or they might be what we call immune-excluded, where the T cells are there, but they don’t get through. In those cases, you have to think about what you can do to sort of prime that situation. Many interesting things are going on right now—from PARP inhibitors in combination to pegylated interleukins to the VEGF inhibitors. We have interesting phase I data with ramucirumab, the VEGF-R2 agent that many of us have worked with in combination with pembrolizumab. That, hopefully, will go forward to some more confirmatory trials. So, it really is going to take a lot of effort to figure this out. I think it’s great that we’ve seen so much benefit from immune therapy, but this chapter is just the first chapter of many to come. It is going to take, in my opinion, a little bit more science to do that. We’re going to have to work in both directions (some reverse translation), treating patients with these combinations and working back, but also trying to generate things forward for the laboratory and whatever animal models exist in this tumor type.

I have a very active phase I group at my center, and I know all of you do. We’re testing, probably, about 20 or 25 different combinations now, which means we really don’t know what we’re doing. So, it’s really time to sort of figure it out, scientifically.

Everett Vokes, MD: Yes. It reminds me a little bit about what used to be described as alphabet soup with chemotherapy agents, where they did have different mechanisms of action and people actually were trying to be scientific about it. But in the end, it was just empirical combinations. I am not sure that we are that much better at this point, although I think reverse translation is what will likely be most informative since the trials are already ongoing.

Roy Herbst, MD, PhD: What is important though, Everett, is that today, when we do clinical trials, we ensure that the clinical trials have the sufficient translational components, and we include this type of research into the clinical trials. I think that is strongly needed, today. We have, traditionally, only done clinical trials and have just focused on the clinical parameters while we, these days, need to really implement translational research into our clinical trials.

Transcript Edited for Clarity