Naiyer Rizvi, MD: Jacob, if you have somebody who was on osimertinib, he progresses, and he doesn't have anything come up on biopsy. And you put him or her on chemotherapy for 9 months and then they progress. Do you ever come back to osimertinib?

Jacob Sands, MD: Well, that's not something I have yet done, although it’s something that I hear others discuss that I think is reasonable to consider. It depends. If you're doing retesting, and you're finding that again there's a large population, and you're seeing a significant increase in cell-free DNA, sensitizing mutations present, I think it's reasonable to consider. In many of these patients, if you do get some control, that can help. But I wouldn't necessarily expect to get long-term durable control again, and in some cases I think it starts feeling a little bit like that game Whac-A-Mole. You keep hitting, and more pop up. So I think it's reasonable to consider. That's not something I've broadly done.

Naiyer Rizvi, MD: Leora?

Leora Horn, MD, MSc: I have done it in some patients. It more seems like my patients who are on erlotinib a long time ago progressed. I put them on chemotherapy. Maybe they were T790M negative, and maybe they've been on chemotherapy for a couple of years, and they progress again. And so using it at that time. I think if a patient progresses after 3 to 6 months on chemotherapy, I’m not going to go back to a TKI [tyrosine kinase inhibitor]. It really depends on that disease-free interval, how long that has been in helping me select what therapy I'm going to give those patients next.

Naiyer Rizvi, MD: One of the common situations that I get questions about is a patient comes in newly diagnosed, EGFR-mutant lung cancer, but their PD-L1 [programmed death-ligand 1 expression] is 100%. Josh, what do you do? I know what you'll do, but tell me why.

Joshua Bauml, MD: You give them osimertinib, or you give them an EGFR TKI-based therapy. And the reason for that is that based upon the paper done by Aaron Lisberg, MD, at the University of California Los Angeles as well as Edward Garon, MD, where they answered this question., It was a critical question, because this is something that comes up all the time. Patients with EGFR-mutated non—small cell lung cancer do have PD-L1 overexpression. And so they said, “OK, if you have PD-L1 overexpression and EGFR mutation, we're going to give you pembrolizumab.” They treated 12 people, and no one responded. Actually, 1 of them responded, and then they looked back and said that patient actually was EGFR wild type. So out of 11 patients no one responded, and importantly, we had 2 toxicity-related deaths after stopping the pembrolizumab and beginning a TKI. So the toxicity synergy that can occur with these drugs occurs well after stopping the immunotherapy. It's really important to wait for that molecular data and to use the targeted therapy when you can.

Jacob Sands, MD: Something Josh said earlier is so relevant to this as well, where you said it related to the chemotherapy adverse effects, and patients not wanting that—it ends up educating patients about that. And I think starting therapy is a big education point where you have people that come in, and they can wait. The majority of people can wait. They don't need to immediately go on therapy. But there's a lot of anxiety to start treating the cancer right away. And the big thing that I end up saying to patients that I think resonates is, “We want you on the best therapy, not the first one.” It's a matter of educating them that it's OK to wait, and that you actually prefer to wait to get them on the right therapy to be able to put them on targeted therapy if that's an option for them.

Joshua Bauml, MD: Yes, that's so true. What I'll say to patients is, “Look, you know that the treatments have adverse effects, and they have serious adverse effects. So we better make sure we give you the right thing.” That usually helps to explain to patients why they need to wait because they are aware and often afraid of the adverse effects. So that can help to explain things.

Transcript Edited for Clarity