H. Jack West, MD: We got the overall survival results from the PACIFIC trial earlier this year at the World Conference on Lung Cancer in Toronto, and that also included subset analysis work that suggested that there were a couple of groups that may not benefit as much. But this was a post hoc exploratory analysis and included in that were the patients with PD-L1 [programmed death-ligand 1] that was negative and patients who were EGFR-mutation positive. It hasn’t been our clear standard to do this testing in this stage III setting, in part because there hasn’t been a clear reason to use this information. So, first of all, do you do this testing routinely? And, if so, how do you use this information? And do you exclude patients with PD-L1—negative disease or EGFR mutation-positive from recommending durvalumab?

Charu Aggarwal, MD, MPH: So I guess two questions. Do I get routine testing? PD-L1 is part of our pathway where we get reflex testing, and, at the time that I see a new patient, I usually have their PD-L1 results back already, at least for my locally advanced and my metastatic non—small cell lung cancer patients. So that’s a given. I usually know their PD-L1 status. Do I discuss this at their first visit? No. I think there is just so much to discuss at the first visit. I’m talking to them about chemotherapy; I’m talking to them about concurrent chemoradiation; I’m talking to them about the timeline; I’m talking to them about immunotherapy, that it’s going to be a year. So I’m not really discussing the intricacies of PD-L1 subset analysis at that visit. However, I do tend to bring it up at the time of the durvalumab initiation. Once I have established a relationship, they’re more comfortable with what they’re going through, we discuss their scans.

And I don’t necessarily hold back DURVA [durvalumab] just yet for my PD-L1—negative patients because, to your point, it was a subset analysis. The whole trial was overall positive. I feel that maybe those patients will benefit, maybe they won’t. I just don’t know if I can draw a conclusion out of those small numbers enough to make a binary distinction for a patient that’s sitting in front of me who may benefit. And we all know that PD-L1 is not the perfect biomarker. If I’m seeing a patient who’s a smoker, who has a squamous histology, who otherwise I think would be a great immunotherapy candidate, I’m inclined to offer it to them.

H. Jack West, MD: Yeah. I have seen a little bit of data to suggest that this is a subset analysis. Not everyone had PD-L1 data. There were some imbalances in the groups, enough to give us pause about being too confident about this. Hoss, what are your thoughts on testing, and do you use this to exclude patients or do you recommend it routinely?

Hossein Borghaei, DO: We do testing reflexively also. We get our molecular testing and PD-L1 for locally advanced and metastatic patients sort of reflexively now. So the information is out there, but because as you said, the intent-to-treat patient population benefitted from durvalumab, that’s how I look at the data. So I offer it to anybody regardless of PD-L1 or molecular subset right now. Again, in the absence of clear data in that particular group of patients, I tend not to rely on the analysis of a subset of patients in these phase III studies. I think they’re hypothesis-generating. I think they’re interesting to look at it, and let’s also be honest that the subset analysis was done at the request of health authorities in other countries that want to cut down on costs perhaps.

There are all the other areas. Not that it’s not an interesting analysis, it’s just that when the intent-to-treat patient population was completely positive for survival with a simple treatment in an area where we have had no improvement in over 2 decades, it’s kind of hard not to offer that to a patient. As you said, this is curative intent type of a treatment, so I don’t want to go by some subset analysis with small patients and all this stuff that Charu just said, so I offer it.

H. Jack West, MD: I agree, in general. I also do feel a post hoc subset analysis is not a reason to exclude patients from a trial that was broadly positive. But I do have a discussion with my patients, and particularly I think it’s maybe more of a dilemma for the EGFR mutation-positive for a couple of reasons. 1) We have a pretty good degree of evidence now in the stage IV setting, at least with monotherapy with these checkpoint inhibitors, that they are not particularly effective. And then secondly, there’s a concern, I think, a real concern that giving immunotherapy prior to an EGFR TKI [tyrosine kinsase inhibitor] may predispose to pneumonitis risk. And these are patients who have the potential to benefit profoundly from EGFR TKIs. So I think that the risk there may be a little greater, not to say that I would not offer it, but I would have a very careful discussion, and I think a reasonable informed EGFR mutation-positive patient may decide that the risk/benefit doesn’t fall in their favor for DURVA.

Hossein Borghaei, DO: I don’t know. I look at it a little bit differently. We haven’t cured anybody with osimertinib or any of the other TKIs. That’s the honest truth. We have cured people with chemo-RT [chemoradiation therapy], and now we see that we have really good survival curves with the addition of DURVA plus chemo-RT. And I see your point, but the metastatic setting is different than locally advanced setting.

H. Jack West, MD: I agree.

Hossein Borghaei, DO: Single-agent I-O [immuno-oncology] in somebody who hasn’t had chemo-RT to prime their tumor is different. So I think there is sufficient reason to think that immunotherapy in the setting of chemo-RT for even an EGFR-positive patient might be beneficial. Again, in the absence of a clear study that tells us not do to so because it could be harmful, I think the patients actually deserve to get the drug. And I agree with you. A good, informed consent is really important here. Just like you said, you discuss it with the patient. You might not benefit because you have negative PD-L1 or because you have an EGFR mutation. But the data are a little bit unclear right now, and hopefully you can help the patient decide how they want to proceed with it. But I look at it a little bit differently, that maybe the story with the post-chemo-RT tumor is a little bit different than just a standard EGFR.

Charu Aggarwal, MD, MPH: Although I feel that EGFR-mutant lung cancer may be biologically different and may not benefit in the same way that a PD-L1—negative patient with a smoking history may benefit from immunotherapy consolidation. So I think the risk/benefit ratio maybe is a little more skewed for EGFR-mutant patients compared to the PD-L1–negative subset. But overall, these are subset analyses and I think we have to discuss this with the patients. I’m less inclined to offer durvalumab to an EGFR-mutant patient.

Transcript Edited for Clarity