News
Video
Assessing HER2 Expression in Metastatic Colorectal Cancer
Author(s):
Switching their focus to HER2+ metastatic colorectal cancer, panelists discuss best practices in testing HER2 to aid in the timely selection of best available therapies.
Transcript:
John L. Marshall, MD: Back to talk about an old target that’s new to colorectal cancer. We knew it was out there all along, but we didn’t really know how much to do with it. It’s a rare expressing target around HER2. Dr Cohen, we’re going to pick you to set the stage for this marker and why it’s made a splash.
Stacey Cohen, MD: I think we are very excited about HER2 because coming into it late, we already have a lot of targeted therapies. So of course we’re always wanting to look for things that can help treat patients. We expect to see HER2 alterations in about 3% to 5% of patients. But we’re not looking just for mutations here, we’re looking for overexpression, for amplification. And we know that the degree of that overexpression or amplification really matters. We want 3+ [score] by IHC [immunohistochemistry], some studies are allowing 2+ and ISH [in situ hybridization] positive, but they’re still not getting the same response rates that we see for 3+. So we really want the highest expressing patients, and these are the ones who seem to benefit from HER2 inhibitors.
John L. Marshall, MD: Is it just breast cancer rules?
Stacey Cohen, MD: It’s interesting because we didn’t know. In my institution, we had long discussions with our pathologists because they said, how can you order this test that we don’t know how to interpret? Because it’s not a direct response, it’s a categorized response. As it turns out, when they’ve looked at the data from MOUNTAINEER and other studies, you can use the breast rules, you can use the gastric cancer rules, you just want the high [expressing patients]. Those are the ones who seem to benefit from these treatments.
John L. Marshall, MD: If I’m partnering with a gene testing company that’s doing ErbB amplification, is that good enough?
Stacey Cohen, MD: There are some data to suggest that if you’re going to look at tissue amplification or blood amplification on an NGS [next-generation sequencing] platform, the tissue seems to have a higher correlation with what we expect from IHC or FISH [fluorescence in situ hybridization]. You can use blood, but it’s not quite as representative as what we would expect from our classic tests.
John L. Marshall, MD: Let’s say I have a patient I’ve been treating for a couple of years now, and I got my outside company test that didn’t show any change, and I didn’t do an IHC or anything for this. Is this a patient for whom I should go back to a block and order IHC on even though maybe the ErbB was not amplified?
Stacey Cohen, MD: I’m ordering it for patients who are KRAS, NRAS, and BRAF wild-type. If they have a pathway activating mutation already, we’re not really seeing overexpression of HER2. So if you had a classically wild-type patient, we’d need to go back and find those patients, and they may be the ones who weren’t responding to the EGFR inhibitors anyway.
John L. Marshall, MD: I am so glad you brought that up because I’ve been in the mutually exclusive club around these. But then I saw the data yesterday [showing that] it was a fairly high percentage of patients who had a RAS mutation.
Aparna R. Parikh, MD: It was. What Dr Marshall is referring to is the data with the ADCs [antibody-drug conjugates], so trastuzumab deruxtecan, T-DXd. And in the data we saw yesterday, which were looking at dosing too, there was a number of patients who had RAS mutations who benefited. It makes sense for the mechanism of action for an ADC because you’re looking at HER2 as the target to give your cytotoxic payload and not looking at HER2 targeting approaches.
But it was a lot higher number than I anticipated. So I think generally, reflexively, and I think we talked about this in the first segment, just doing HER2, IHC, and FISH on all patients. If you get it from the outside [sources], even if it’s not on the commercial testing, we’ll still go back and do that because the copy numbers can be a bit tricky on some of the tissue-based testing. There are clear copy number cutoffs, 7 or 8, where it seems to definitely be beneficial. But what’s a copy number of 2 or 3? So I still use the IHC and FISH too.
John L. Marshall, MD: This is what I did in this patient. I didn’t find it, but I felt obligated to look because it’s often these left-sided tumors, and they’re younger, and you don’t want to miss any chance at any therapy going forward. For the first time yesterday [I was] thinking about our 2 agents that we’ll talk about in this space. One was enriched for just the RAS wild-type group, and you had to be RAS wild type to go in. The other basically says no, if you have a RAS mutation, we have an angle for you.
Testing, Tani, again, this is something we [have] to make sure you don’t miss these patients, even though it’s fairly rare.
Tanios Bekaii-Saab, MD, FACP: Yes. Testing is at the first line, before deciding on first line and making the case for adding HER2 as part of the testing, for 3 main reasons. One is, we need planning, and when we think about colon cancer, we’re thinking about a marathon now, so I have to have all my plans for my patients, from point A to point Z. But the other is, HER2 amplifications can predict for EGFR inhibitor resistance. A number of studies have shown that, we’ve presented some data at ESMO [European Society for Medical Oncology annual meeting].
John L. Marshall, MD: Let me say that very bluntly back at you. Will you give an EGFR drug to a patient if you don’t know their HER2 status?
Tanios Bekaii-Saab, MD, FACP: No, not at this point.
John L. Marshall, MD: This is where I’ve gotten.
Tanios Bekaii-Saab, MD, FACP: Now, can I give it in later lines? Perhaps when I run out of other options. But in the [first or second line], the answer is no, although our European colleagues may not agree as much. Everything relates back to biology. If biologically it makes sense, and it does, HER2 amplifications, which is part of the same family of genes, HER1, HER2, and HER3, then it [also] makes sense clinically. [It has] been shown clinically. That’s one of the main reasons.
The third reason is [that] we have MOUNTAINEER-03. We have a clinical trial that certainly is going to be challenging for us to accrue. We need to capture every single patient because if we want to keep moving these biologic agents and make a bigger difference in our patients in the first line, then we have to be testing consistently. The way I describe it to my fellows is [to] think about if you have a 1% expression of a certain target, if you test consistently, you have to test 100 patients to [get] 1. Imagine if you do it randomly through the door, the 100 becomes 10,000. You won’t have an opportunity to find that target at the same level. So we have to test consistently.
John Strickler [MD] is presenting data on behalf of all investigators at ASCO [American Society of Clinical Oncology annual meeting] this year [where] they looked at the concordance rate between the different platforms where there were centrally assessed platforms. Most of the patients had the 2 tissue platforms, next-generation sequencing through liquid, tissue, and then IHC and FISH. There’s a very high level of concordance between the different platforms, to the point that was brought.
The tissue platforms have the highest level of concordance, about 93%, so FISH and next-generation sequencing, using 6 as the cutoff. The liquid and the tissue [were] about 80%, 81%. So it’s not too bad, but not as good. The good news is that any positive [result] on any of those tests is a good predictor of a really good response, and that’s important to keep in mind. I think that certainly helps us quite a bit in thinking about our testing platforms and the flexibility we have around our test platforms.
Transcript edited for clarity.
%20(2)%201-Recovered-Recovered-Recovered-Recovered-Recovered.jpg?fit=crop&auto=format)
%20(2)%201-Recovered-Recovered-Recovered-Recovered-Recovered.jpg?fit=crop&auto=format)
%20(2)%201-Recovered-Recovered-Recovered-Recovered-Recovered.jpg?fit=crop&auto=format)
