Naiyer Rizvi, MD: What is your approach for up front, Jacob? I know that the Dana-Farber Cancer Institute has been doing research in plasma testing methods for really a long time. You and Geoffrey Oxnard, MD, of the group have been leaders in the space. What do you do with your newly diagnosed patients?

Jacob Sands, MD: We agree. We have a low threshold for sending plasma for cell-free DNA testing. Also, when we see people from the community, sometimes knowing whether there is tissue available, just as Josh outlined, can be a challenge. If we have tissue, and the tissue result is back, then we don't send plasma necessarily, but if we don't have anything back, if we end up testing both, I’m comfortable with that as well. Often, to Josh's point, the plasma testing may come back fairly quickly. For many people it does come back in a week or a little bit more than a week. And if we have that answer before we get tissue, then we don't necessarily have to test the tissue.

Naiyer Rizvi, MD: Now, I don't particularly do it, but I see people who measure serial plasma DNA to track plasma response. Leora, is there a value to that? Do you think it's useful to do serial plasma testing on patients?

Leora Horn, MD, MSc: No, I think it's something that's interesting, and there are studies that are looking at that. But the thing with the serial plasma, if you track it and the frequency of the mutation goes down, and then it comes up but the CT [computed tomography] scan looks fine, I'm not necessarily going to switch therapy based on that until there are good data to suggest that switching therapy based on plasma rather than a scan is the way to go. I think it's an interesting clinical question, but I don't think it's ready for prime time and helping us make decisions on what to do with our patients in the clinic right now.

Joshua Bauml, MD: I was confronted with this very situation a couple of years ago where I had a patient where molecularly I saw T790M mutation in a patient who was on a first-generation TKI [tyrosine kinase inhibitor], but radiographically they were fine. So that sort of molecular transformation where I know I have another treatment for it, it created a very difficult cognitive dissonance, where I know what's going to happen and I don't know whether intervening early makes a difference, to your point, Leora.

Leora Horn, MD, MSc: Yes. And the reality is that there are a finite number of TKIs that these patients have available to them right now. If you switch quickly, they're going to develop resistance to that TKI as well.

Jacob Sands, MD: I was following 2 patients as part of a research study. This is a few years back as well, where T790M showed up without any radiographic changes, and in both cases radiographic progression was about 3 months after the initial presence of detecting T790M. I think the one area where this is particularly compelling—there was a nice publication out of Leora's group in small cell lung cancer, we’re seeing cell-free DNA levels going up in small cell. And in small cell, this ends up being a population where progression happens very rapidly when we do see it. Maybe in that population we will be able to detect this, although this is entirely investigational at this time. This is not something that is yet ready or widely used in clinical practice, but I would like to see that studied a bit more to provide that data that then could guide practice.

Naiyer Rizvi, MD: I assume everyone uses osimertinib first line for a sensitizing EGFR mutation, so I don't think we need to spend much time talking about that. What about the recent approval of erlotinib with bevacizumab as a first-line treatment option? I'm not going to ask Jacob about that because he doesn't like bev [bevacizumab] anyway. So, Josh?

Joshua Bauml, MD: Sure. I think that was a study that was designed before osimertinib was the standard of care, which is, I think, across America. But it was designed, and they said, “Well, based upon this, if we reach our end point, will it get approval?” And they said yes—so I think it's very reasonable that it was approved. Would I use it? No. I don't really have much of a reason to use it. Osimertinib is much easier to administer. It has a better toxicity profile than erlotinib, and so therefore I would expect it to have a better toxicity profile than erlotinib plus something else.

But that being said, if I have a patient for whom I cannot administer osimertinib for whatever reason, I think we then have multiple options available to us. I think erlotinib/bevacizumab is a reasonable option to consider. I think also the data that have come out of India and Japan regarding the combination of chemotherapy with gefitinib is also something that I would consider. Given the long-term outcomes there, that would probably be my second choice, with the RELAY study regimen of erlotinib plus bevacizumab being my third choice just based upon the long-term outcomes that we see there. But at the end of the day, am I going to use this regimen? Probably not because I'm sitting from a position of privilege where I can give osimertinib to whomever I wish. But if I could not, that would be a situation where I would talk about that.

Naiyer Rizvi, MD: Leora, regarding the FLAURA trial, the survival data showed that there wasn't really any difference between third- and second-generation TKIs in the Asian population. Is that a setting where you may use erlotinib and bevacizumab?

Leora Horn, MD, MSc: No. I think it's hard to justify bringing someone in for infusion when they can just come in every 3 months for a CT scan on an oral drug. And there are more toxicities with the bevacizumab: the hypertension, the proteinuria, the fatigue. So it's hard, to make any distinctions—we all look at the studies, and then if it's a positive study, we try to find subsets where maybe we should exclude it. If it's a negative study, we try to find subsets that maybe we should give the drug to. It's hard to make anything of a small subset analysis. I think the only thing that would have changed is if the overall survival was negative in the whole patient population. I might have given pause to my first-line choice of therapy, but at this time osimertinib is the drug that I’m using in everybody who's eligible.

Transcript Edited for Clarity