Benjamin P. Levy, MD: Let’s talk about testing. We all agree that fusion is the way to go. The fusion is predictive; amplification point mutations are not. What tests are able to pick these up? Is it Foundation [Medicine]? Is it Archer? You and I were having a conversation before on the panel about these different panels. What’s the recommendation here? Is it to make sure that whatever company you’re using is picking this up? Is it just to make sure you’re doing next-generation sequencing, and that should capture it? What’s the statement that we have here?

John L. Marshall, MD: It’s not just the CHEM-20. You don’t know what you’re getting in your CHEM-20. We think there are 3 TRK fusions: 1, 2, and 3. They’re different levels of expression, and they may be different in terms of their biology, but we need to, in theory, catch them all. And so most of the groups are working as quickly as they can to dial up their assay so that it captures all these things.

You can get it from any of the major players that are out there, or any of the major clinical partners that we all use. I’m not so sure that’s true of the internal groups. The Mass Generals [Massachusetts General Hospital] and the Memorials [Memorial Sloan Kettering Cancer Center]—they do their internal stuff. I’m not sure what those panels have. I’m sure they know, but I don’t know.

Benjamin P. Levy, MD: I can tell you, we’re not doing in-house at Johns Hopkins Sidney Kimmel Cancer Center right now.

John L. Marshall, MD: So you’re not doing this?

Marcia S. Brose, MD, PhD: We are doing it at Penn [the University of Pennsylvania].

Philip Agop Philip, MD, PhD, FRCP: When I first heard about the drug—and obviously we have to do it—I checked with the… We are a large cancer center [the Barbara Ann Karmanos Cancer Institute]. We have a partner we work with in terms of the NGS [next-generation sequencing]. We asked them, and we got the answer that we will get what we want and all that. But I have a lot of patients who have been with the same partner for the last 4 or 5 years, but that wasn’t done then. So what do I do?

John L. Marshall, MD: I think that’s the question. Do you go back?

Philip Agop Philip, MD, PhD, FRCP: Do I go back and do the same sample? For some of those patients, obviously most of them aren’t alive, but some of them are still on treatment or are changing treatments. Do I go back and say, “Hey, do you have the sample sent to you before? A new biopsy?” What would you do?

John L. Marshall, MD: I would love to hear what the lung people have to say too, because we have lots of tissue usually. As far as rebiopsying somebody, what is it? Is it 0.2%? Is there a chance that this would be positive?

Benjamin P. Levy, MD: I think it’s a great question. We struggle with this all the time when tissue is insufficient, and this brings into the fold the whole idea of plasma genotyping and liquid biopsies. We’ve got a lot of data in lung recently coming out showing that it can detect alterations that weren’t found in tissue if you do it concurrently with tissue interrogation.

John L. Marshall, MD: You guys are liquid people. Does that work? Are you all getting all 3 fusions?

Benjamin P. Levy, MD: You get just 1. You get NTRK1 off Guardant360 right now.

John L. Marshall, MD: OK.

Edward S. Kim, MD: We encourage people to rebiopsy at progression. I think that’s a good thing to do. When you have basket trials out there like NCI-MATCH [The National Cancer Institute's Molecular Analysis for Therapy Choice trial] or ASCO’s [American Society of Clinical Oncology’s] paper study, it allows you to use archived tissue. But I like getting the fresh tissue. A few years ago, Mass General reported the transformation to small cell that occurs 15% to 18% of the time. And that’s an EGFR-mutant patient. Again, you’re looking for T790M and, all of a sudden, wow, it’s small cell. It’s a totally different therapy you’re going to use here. In the lung field, I think we are trying to retrain. So I don’t think, John, to answer your question, we would specifically go after a 0.2% marker. But there are other things associated that allow us to justify getting it.

Mark Agulnik, MD: And I think it’s easier to retest. In the sarcoma community, it’s common that we retest and rebiopsy. I think the reason we do it is we also think that the moleculars will change based on prior treatment. I do think it would be a challenge to retest someone who hasn’t progressed on a treatment and is doing well. Really, there has to be something that would lead you to think there is good reason. Treatment failure, for us, is usually the marker that sort of gets us to rebiopsy and resend.

Marcia S. Brose, MD, PhD: And I’d just add that I think that Ed also brought in a good point, which is [that] we’re not dealing with TRK fusions in a vacuum. There are so many others. You know, we have the pleomorphic, or the heterogeneity in the tumors, and we have all these other points. So the point is, if we’re really doing a good job picking up other mutations, we can do them often together.

And that leads me to talk about 1 of the other tests that we should talk about: There is a development of an immunohistochemistry test. The problem I have with that is I don’t think working in sort of like a vacuum, just looking for TRK fusions, is really going to be a viable approach. The reason for that is the community doctor doesn’t even know what to do. They know what to do in so many ways, but they have to become experienced and become experts in so many things now. I think that it’s much easier to say, “Let’s send the test off, but let’s just make sure that test gets everything.”

Benjamin P. Levy, MD: Yeah.

Marcia S. Brose, MD, PhD: And do 1 test and get everything as opposed to these scenarios of, “Oh, I’m going to remember, on my 0.1%, that I better send off an immunohistochemistry for this. And then, if it’s positive…” I just don’t think that’s really…maybe if you had something for which you knew was like a 20% or 30%, then you want to do it as screening, but I just don’t think that’s going to happen.

John L. Marshall, MD: Ben, Ed brought up this point earlier about the culture around rebiopsying and how that was not the right thing to do. We’ve also been through a decade for which NGS has been kind of available, or broad sequencing has been available but not paid for.

So there’s been this cultural and financial barrier to this. But we’ve even seen that kind of go away. Thanks to these kinds of medicines that rely on these rare subtypes, it is now part of our culture—just as we get CT [computed tomography] scans on everybody. Even though the yield is low, I think we’re going to see a culture in which we’re doing this in everybody. So I’m hopeful that barrier falling then makes this sort of retesting concept more and more common.

Benjamin P. Levy, MD: I agree. I can remember NGS coming into the fold in clinical practice probably around 5, 7 years ago, and a lot of talk about cost and it being prohibitive. Certainly, I think that there’s been a greater depth and breadth of what we can find, and it’s gotten cheaper. Hopefully, again, that will be supportive for a broad use of next-generation sequencing.

Edward S. Kim, MD: The issue is not going to be the CHEM-20 or the blood test. It’s going to be a CBC [complete blood count] and the NGS.

John L. Marshall, MD: You’re right. That’s going to be part of the base staging that we’re going to see for almost every cancer we manage.

Philip Agop Philip, MD, PhD, FRCP: So what about the provocative question: NGS reflex test in stage IV disease? What do you think of that? In my practice, I have to admit that I do NGS in stage IV disease. I do that. I don’t do it in stage III. I don’t do it in stage II or stage I. But I do on every patient. I do NGS on every patient when I first see them.

John L. Marshall, MD: But I could push back. In pancreatic cancer, for example, we need to know BRCA on everybody. All of a sudden, we’ve got some need to do some additional testing.

Philip Agop Philip, MD, PhD, FRCP: The other thing is that NTRK mutations are driver mutations. Driver genes compete with one another. If 1 is present, the other 1 is not present. If you take, for example, colon cancer, maybe in half the patients who don’t have a RAS mutation, maybe that’s more relevant. But again, I do NGS on every stage IV patient. The question is, is this something that we have to talk about as a more practical thing?

Transcript Edited for Clarity