Tony S.K. Mok, MD: I hear diversity in practice, which is understandable because there’s no standard practice. But let’s just focus on a few things. We have the transformation. We have the MET amplification and have C797S [mutation]. Let’s talk about what to do with each 1 of them. Let’s start with this small cell transformation. Do you just give chemotherapy and that’s the only thing you do, or if you give chemotherapy, do you continue with the TKI [tyrosine kinase inhibitor] when you give the chemotherapy when you have small cell transformation? Yi-Long?

Yi-Long Wu, MD, PhD: Both. Because of all the histology translocation, we stop the TKI and then switch to the chemotherapy only. Because the [INAUDIBLE] clinical trial [was impressive], in the first generation, continued TKI past the chemotherapy is not superior to the chemotherapy only. If the patient is resistant to the TKI, why are we continuing the TKI in this version? So I only switch the chemotherapy.

Tony S.K. Mok, MD: Lu Shun?

Shun Lu, MD, PhD: Yes, it’s the same. I only switch to the chemotherapy.

Tony S.K. Mok, MD: OK.Now, the MET is a bit more interesting. Nowadays, there is the TITAN study that’s suggesting that chemotherapy together with a MET inhibitor, and also Yi-Long's study using gefitinib with capmatinib that has shown some evidence of response. The 2 questions: No. 1, how do you define MET positive? And no. 2, are you comfortable to treat the patient with the TKI plus the MET inhibitor without the phase 2 or phase 3 data? Myung-Ju, what do you think?

Myung-Ju Ahn, MD: The MET is difficult because we don’t have any good biomarker. So far, I think the FISH [fluorescence in situ hybridization] is the most reliable biomarker, and then the C-amplification by NGS [next-generation sequencing]. We don’t have any good cut offs, so it’s very challenging. Then, immunohistochemistry is another method, but there is just some correlation between the FISH and the 3 positive immunohistochemistry, but there are a lot of challenges. Anyway, if the patient [had] a c-MET maybe off-label I can use the TKI plus crizotinib. So far, that’s the available drug. Other than that, they do the clinical trial.

Tony S.K. Mok, MD: Japan, I presume you are very restricted on what you can do, right?

Tetsuya Mitsudomi, MD, PhD: Right. Yes, I’m very interested in MET amplification as a cause of the resistance, but as you said, we cannot use the TKI in addition to the EGFR TKI outside of a clinical trial basis. Also, I agree with Myung-Ju, the definition of the MET amplification is a bit tricky. If you set the threshold higher, then the response rate may be better, but what is the most appropriate point is very difficult to determine.

Tony S.K. Mok, MD: Go ahead, Yi-Long. Go ahead. Sorry.

Yi-Long Wu, MD, PhD: I have been focused on the c-MET since 2010, and since that time we have done so many clinical trials focused on the first generation past the c-MET inhibitor to overcome the c-MET amplification. So far, the so-called golden biomarker criteria is the FISH. If the FISH amplified, I think more than 60% of patients comes to this combination. But for the NGS, I think this is very difficult. We don’t know what is the exact type of barrier. Recently, we had a consensus among the Chinese [INAUDIBLE] and NGS, the molecule and the [INAUDIBLE]. We find if the gene copy number is more than 5 and 6, then in this situation the patient responds to this combination.

Tony S.K. Mok, MD: Thank you. Now Lu Shun, I want you to tell me the ideal study design, phase 3 design, to prove the role of a MET inhibitor in an osimertinib failure knowing your involvement in this study.

Shun Lu, MD, PhD: The key issue is what are the cutoffs that matters for amplification? This is the very tough answer. Previously we believed [it was] MET 5 or more than 5 or 6. Actually, the progression-free survival was quite similar with the platinum-based chemotherapy, like 5 months or 6 months. But [this is] only for the very high amplifications like maybe 9 months to 10 months, but the patient number is very small. In addition to this, go back to the previous [INAUDIBLE] although this is negative study, but they do the subgroup analysis. They’re published in JCO [Journal of Clinical Oncology] you know. They’re using the immunohistochemistry If 90% cancer cells are 3+ immunohistochemistry positive [they] still get quite a longer progression-free survival. What are diseases at the design of the study, we selected a cutoff number like 10, plus 90% of cancer cell immunohistochemistry is 3+ positive patient, randomized to the osimertinib plus savolitinib vs. platinum-based, based on the chemotherapy. In China, we already launch the studies called a [INAUDIBLE] study, called a search study, but they also allowed patients previously using the first or second generation TKI, had failure and no T790M mutation, but a MET amplification. In this study we used the cutoff of 5. A study is already ongoing. I have already 2 patients that are enrolled this study.

Myung-Ju Ahn, MD: I think that’s a very interesting study. In China the recruitment might not be the problem, but the selection of patient with a high expression of the c-MET more than 90% is going to be very challenging for us. China is no problem.

Tony S.K. Mok, MD: I think Lu Shun’s smart point is that how do you make the data to be strong enough for a drop registration? I’m quite sure once it’s registered that people lower than the cutoff may still be considered to use that. So I think that’s the tricky point about this type of study design. You can set a very high bar, but I’m quite sure that upon generalization people will look into the situations very intuitively.

Transcript edited for clarity.