Transcript:Benjamin P. Levy, MD: Govindan, I’m going to ask you to put down your squamous crystal ball, and put up your immunotherapy crystal ball and tell us where you think we’re heading. I see it as a ‘choose your own adventure’ here. There are a lot of different possibilities. One is single agent in a PD-L1—restricted patient population. There are data with combination chemotherapy. There are combination strategies with other checkpoint inhibitors. There are certainly novel strategies in a second-line setting. Can you talk a little bit about where you think we’re heading? Maybe mention some of the data we have so far with chemotherapy, and what’s on the horizon?

Ramaswamy Govindan, MD: So, first, I think we are in this for the long haul. This is not going to be a drive-by-night approach. This is going to be an entirely new era that’s going to be with us for a long time. I see there are two radically different approaches, and they may come together, they may complement each other. The first one is, how can we optimally combine the inhibitor checkpoints, as well as the agonistic elements? For the immune system to act, there are built-in checkpoints to slow down the immune system activity. That’s the CTLA4/PD-L1/PD-1 axis. But, also, there are other things that are involved, like IDO inhibitors. These things are moving along very nicely in combination. So, we’re going to see combination studies in this setting. Of this, the immediate one is the combination of the ipilimumab/nivolumab in the frontline setting of non-small cell lung cancer. The CheckMate-012 study showed, in the frontline setting, that the combination actually produces impressive response rates. And also, the 1-year survival data are upwards of 70%, and the median progression-free survival upwards of 8 months. I think we can build upon that and it’s reasonably well tolerated, it’s not that hard to manage. The results may not be as impressive as what we see in melanoma, but certainly really gives hope that the blocking CTLA4 and the PD-1/PD-L1 axis is an important one.

In a very small subset in the CheckMate-012 study, if I’m not mistaken of 12 or 13 patients with a high level of PD-L1, about 50%, the response rate was a fantastic 70% to 80%, and more than 90% were alive a year later or so. Again, these are small numbers, but it brings up an issue that is PD-L1 expression may be a better marker for a combined inhibition rather than for a single nivolumab. So, all these things will be fleshed out in the coming years. That’s, right now, how we see it, and we’re going to see a lot more in the coming years. There are second-line studies in a number of inhibitors, number of combinations, IDO inhibitors that are going on. And there are now master protocols. The BMS (Bristol-Myers Squibb) has this fraction study where you have multiple different arms being replaced for the post-nivolumab patients. So, that’s really going on.

And, yesterday in the development of therapeutic section, I was really impressed with the early data from OX40 agonists along with atezolizumab. The important thing is you would worry OX40, by the way, is upregulated after the T cells are engaged. And they also suppress the T-regulatory cells, so it’s a double whammy in a good solid way. You can actually make the T cells activated, and then you eliminate the repressive T-cell population. You would think that the combination may produce toxicities, but it’s extremely well tolerated with virtually no grade-4 toxicities or recall. There are only two responses, but it’s a very early study, and the fact that we can combine them gives us a hope we can select them. That’s interesting, and the 4-1BB and all those things are going to be targeted. So, that’s one set of approach.

The other set of approach, where it’s moving forward very quickly in liquid malignancies, is those CAR therapies, these adoptive therapies. And, in fact, there was even a randomized study from Steve Rosenberg’s group in melanoma about these adoptive therapies going on. And then, along those lines, apart from CAR therapies, now we are able to sequence the tumors, find the neoantigens and pick those neoantigens, create this, take the patient’s T cells, and expand them in relation to the artificial peptides and infuse them. In fact, we have already treated our first several patients. You can induce antibodies for more than 50% neoepitopes, and then you have to take it forward with checkpoint inhibitors and combine them. So, it’s really moving along in these two kinds of directions. It’s really a fascinating approach.

Benjamin P. Levy, MD: I would piggyback. Also, I’ve been impressed with the data that you mentioned. I’ve also been impressed, and I didn’t think I would, with the unselected adding of chemotherapy with PD-L1 drug upfront, atezolizumab with carboplatin/pemetrexed, at least in data presented at World Lung, eliciting response rates in the 75% range, although it was a small number of patients. But, I think the data are really compelling and certainly there are competing strategies here. As you mentioned with OX40, there are CAR-T cell approaches, very novel approaches that can be done here. And I think we’re just at the beginning. Any thoughts from the global perspective, European perspective? Where do you think we’re heading with these drugs?

Marina Garassino, MD: We are waiting, I think, like you, for the results of the phase III trial in first-line. I think that the majority of these drugs will come in the first-line setting. I don’t know if it will be combined with chemotherapy or be alone or combined with CTLA4, but it’s possible that a fraction of patients will be treated with these kind of drugs.

Benjamin P. Levy, MD: Any other thoughts from the panelists where we’re heading? I think we’ve covered a lot here. I’m certainly impressed with the combination data I saw yesterday with CTLA4 in a PD-L1—positive patient population, certainly the combination strategies with chemotherapy. I think it’s going to take a lot of work and it’s going to make our jobs a lot more interesting, but a lot harder in a lot of ways to sort these data out.

Mark G. Kris, MD: I think one fundamental difference now is that we’re doing a lot more biology-based research, and that’s what got us where we are today. We could have empirically, I guess, found who gefitinib worked in, but knowing those mutations there and knowing what we do now about the immune system, giving chemotherapy with a PD-L1 drug, is that the right thing to do? Well, not really. What seems to be the right way to do it is to figure out how chemotherapy would modulate T-cell populations to best take advantage of the immune attack that you initiate by giving an anti—PD-L1 agent. So, that’s where I think we’ve got to go, and I think we just have to remember the ‘stay with the lady we brought to the dance’ here. It’s biology that got us here, and we really need to pay attention to that and just stop throwing things together because they were produced by the same company, for example. That’s not necessarily the way to do it, but pay attention to the biology.

Benjamin P. Levy, MD: Yes. Looking at these drugs in the first-line setting, and certainly there are data in the second-line setting in combination with CTLA4, a tremelimumab/durvalumab combination looks good in the second-line setting. But, I agree, I think our treatment decisions have to be predicated on biology, and we’re moving that way. I think it makes our jobs a lot harder, a lot more interesting, but a lot more meaningful for us and our patients.

Transcript Edited for Clarity