Biomarker Testing for Gastric and GEJ Cancers 

Sam Klempner, MD: This is a nice segue into general biomarker testing of advanced patients. I’ll lead off with high-level description, and then we can get into some of the role for next-gen [generation] sequencing. It’s pretty clear although the labels may argue otherwise that certainly, MSI [microsatellite instability] needs to be done for every patient. We would both agree that this should done including in nonmetastatic patients. It’s a different biologic group. That is a subset that we know has exceptional outcomes with IO [immunotherapy]. HER2, certainly a well-validated target, needs to be tested for everyone. We can talk about whether we’re doing that by IHC [immunohistochemistry] with FISH [fluorescence in situ hybridization] or whether we’re adding ctDNA [circulating tumor DNA] on top of that. Then, I would argue that PD-L1 [programmed death-ligand 1], I would agree with you, it was still a relevant biomarker, even though both the label for the [KEYNOTE]-811 regimen and [CheckMate] 649 don’t really require it. At a minimum that should be the assessment for every advanced patient that walks in the door. Beyond that, what are you doing for these patients?

Daniel Catenacci, MD: At a minimum, the answer should be MSI-high testing by standard classical IHC and or by a next-generation sequencing panel, which inherently will have that HER2, of course, which again could be high IHC and then FISH of its equivocal or by next-generation sequencing, which is on the panel. PD-L1, as you heard, my practice incorporates that and uses it for a decision. I’ll point out, we have to at least, be cognizant that one of the arguments as to why PD-L1 should not be used, and we should just treat everyone, is that there is some operator dependence on this in terms of the pathologist scoring and the antibody and assays used. But, that said, the argument should be get it done at a place where there’s the experience and they’re using the right assay and then it should be fine just like it was for HER2 where we want to make sure that there’s quality control on anything that we’re doing with respect to this. That’s an important thing to mention. Then outside of that, our practices, and a lot of academic centers, I have incorporated next-generation sequencing panels of the tissue and or the blood with ctDNA analysis and the advantages there include, that we can get MSI high in HER2, of course, but we can sort of get a broader view of what’s going on in that patient’s tumor. There are a number of other biomarkers that we may get to later on here, that are on the horizon that are promising, that may become standards in the near future. Including say, for example, FGFR2 [fibroblast growth factor receptor 2] amplification or fusion. Other things that, I know you and I both have a passion for like EGFR amplification and MET amplification, et cetera, clot infusions that would be detected on genomic sequencing, particularly RNA sequencing. We look at that, and then ctDNA is just an extension of the tumor-based but receives a broader view of all of that spatial disease that’s going on in that patient when they have multiple sites. All sites are dumping into the blood and we’re getting a full view of the whole disease as opposed to one biopsy that may or may not be representing what’s actually going on in the patient. We’ve studied that prospectively in our PANGEA study where a number of patients were changed in terms of their targeted therapy assignment based on the metastatic site compared to the primary site that was captured by the blood-based testing. That’s more on the cutting edge and things that we should be doing in the future looking for, but certainly, NGS [next-generation sequencing] testing tissue testing is becoming more commonplace as the years go by and it’s becoming part of routine care for the level one of the things that we talked about, but also for that next level of things that are on the verge offer options for studies, et cetera, if not in first-line, but for later lines. That’s my take on biomarker testing but how about you? Any differences there?

Sam Klempner, MD: Yes. In both of us practicing in academic centers a lot of our job is both to understand the standards, but also to try to advance them. To understand the patients, we need that level of detail, and we do the same. We do next-generation sequencing tissue-based from everybody. To the viewers, I think it should be noted that that is actually noted in the NCCN [National Comprehensive Cancer Network] guidelines as something that can be considered, and we do blood as well. We do it both for heterogeneity assessment and then later on we can talk about using it as the delta in ctDNA as an early predictive biomarker in addition to maybe being prognostic. It affects the management of maybe up to half of our patients in academic settings with the number of trials we may have, the number of biomarkers. And then I’ll just make a point that you’ve made several times, which is we really don’t know if FOLFOX [folinic acid, fluorouracil, and oxaliplatin] plus nivo [nivolumab (Opdivo)] performs well in an EGFR-amplified PD-L1 positive versus an EGFR wild-type PD-L1 positive patient. We’d like to have that granular level of detail for all of our studies and learn the most we can from all of our data but ideally to shepherd patients to therapies that may improve upon the standard. One could envision FOLFOX plus IO plus an EGFR drug as something that may be better for an EGFR-amplified patient. That’s on us to try to prove that. We do biomarker testing extensively for everyone primarily to make sure that we’re opening every possible door.

TRANSCRIPT EDITED FOR CLARITY