The Emergence of PD-L1 Inhibitors

Transcript:Mark A. Socinski, MD: So, Jared, talk to us about atezolizumab as well as durvalumab, another PD-L1 inhibitor, in non—small cell lung cancer.

Jared M. Weiss, MD: These are PD-L1 inhibitors with very interestingly named trials. On the topic of atezolizumab, we’d start with the tree-hugger trials. We have the BIRCH trial. BIRCH allowed any line of therapy. This was first-, second-, or third-line lung cancer. What was common to these patients were high levels of expression of PD-L1. Our response rates were in the 24% to 27% range, and I can’t quote you most of the PFS or OS because they’re very happily immature at this point; most of them are not reached. Then you get into POPLAR. This was a randomized phase II trial with previously treated patients. PD-L1-positive and PD-L1—negative patients were allowed on this trial, but the results were reported separately by PD-L1 status. So, in the PD-L1-positive patients, we had a nearly tripling of response rate from 13% to 38% and we had a doubling of our PFS from 3.9 to 7.8 months. Even in the PD-L1-negative patients though, you can’t disparage that data too much. Response rate was exactly the same, 15% in each group. PFS was 2.7 in 3 months, so basically just about the same. And even survival, it showed 9.7 and 12.6 months favoring the immunotherapy arm. So, they are two very different trials in terms of design and which patients are on them, but both are inspirational and following the same themes that we discussed in the context of the PD-1 inhibitors.

Mark A. Socinski, MD: And then we’re all waiting for the OAK trial, which is the large phase III trial. POPLAR was phase II, but OAK is phase III. I think there are upwards of 1200 patients on that trial. Again, the comparison is immunotherapy versus our old friend docetaxel. I would predict that that’s going to look very similar to what we’ve seen with the CheckMate and the KEYNOTE trials; and durvalumab, another PD-L1 agent.

Jared M. Weiss, MD: Yes. So, in head and neck, we have a series of trials named after birds of prey. The HAWK trial is for cisplatin refractory PD-L1-positive head and neck cancer. Defining lines of therapy is a little harder in head and neck cancer than it is in lung cancer because the default patient is treated for cure. And the default patient who is now incurable has received multiple prior lines of therapy. The cleanest way to define these populations, perhaps, is just as cisplatin refractory. That was a cisplatin refractory PD-L1-positive study of durvalumab monotherapy that’s ongoing.

We have our CONDOR trial which is for PD-L1-negative patients, also platinum refractory. Here we have a 2:1:1 randomization though between durvalumab/tremelimumab—so PD-L1 and CTLA-4 inhibition versus durvalumab alone—versus standard of care chemotherapy. And then we have the EAGLE trial which is our second-line trial, unselected by PD-L1 and a 1:1:1 randomization. And finally, most people haven’t heard of this bird, but the KESTREL, also a bird of prey, is the first-line trial, which is a 2:1:1 randomization. The two-arm is durvalumab with tremelimumab. We have a monotherapy PD-L1 arm and then we have, again—on the theme of toxic chemotherapy regimens—a somewhat toxic regimen in the trial called EXTREME, platinum/5-FU and cetuximab.

John V. Heymach, MD, PhD: Jared was discussing the BIRCH trial, and one of the things as a lung cancer doctor I was very excited about, is that we’re used to response rates always dropping as you got in first-line, then second-line, then third. And if you think about chemotherapy, your response rates drop 20% to 30% down to less than 10% in second-line, and by the time you’re in third-line, you’re down to 1%. When you looked at the second and third-line responses to immunotherapy…

Jared M. Weiss, MD: Twenty percent range.

John V. Heymach, MD, PhD: Right, yes. So, in other words, second-, third-line patients are responding to immunotherapy just about as well as first-line patients.

Mark A. Socinski, MD: One of the themes that we’ve seen, and John addressed this earlier, is that there seems to be similar activity between PD-1/PD-L1 agents across all different tumor types. And the observation of whether you get it first-, second-, or third-line seems to be pretty consistent. One of the agents that we’ve talked about but I didn’t get a read on the data in bladder cancer is durvalumab. Do you have any data there that would be notable?

Dean F. Bajorin, MD: We do, and actually the FDA got the data and made it a breakthrough designation in February. That’s being presented at the meeting. Durvalumab is clearly active in patients who are PD-L1-positive. The response rate was just over 50% in previously treated patients, that’s a pretty high number. It’s a small number of patients, actually it was a 61-patient trial. We have the data for 42. We’ll get more data during this meeting, but it certainly proves the point that it’s active. I was listening to John talk about MYSTIC, and the next study is exactly the same thing. It’s durvalumab versus durvalumab plus tremelimumab versus standard of care chemotherapy for 1:1 randomization for first-line therapy. And the chemotherapy is basically a platinum-based chemotherapy, which is standard of care. We should get a readout. I can tell you, the accrual to these trials are getting very rapid. It may not be where melanoma or renal was a while ago, but I think we’re getting there.

John V. Heymach, MD, PhD: Yes. We’ve had waiting lists for these studies.

Mark A. Socinski, MD: Yes.

Transcript Edited for Clarity