HER2 Alterations in NSCLC: Mutations, Testing, and Treatment Selection

Drs. Joshua Sabari, Alfredo Addeo, Julia Rotow, and Helena Yu examine the different types of HER2 alterations in non-small cell lung cancer (NSCLC), including mutations, overexpression, and gene amplification. They discuss the prevalence of HER2 mutations, the importance of comprehensive next-generation sequencing (NGS), and how testing practices differ across clinical settings. The panel also explores the distinction between HER2 tyrosine kinase domain and non-kinase domain mutations, emphasizing the need to determine whether an alteration is oncogenic before selecting targeted therapy. They highlight the importance of reviewing molecular reports and using resources such as OncoKB to identify actionable HER2 mutations that may otherwise be overlooked.

Transcript:

Dr. Sabari:  So excellent transition here to HER2. If you think about the HER2 space, there are many alterations very similar to what we just discussed in the MET space. There’s mutation, most commonly in lung cancer, overexpression, and amplification. Helena, how do you think about these three separate groups, and can you help define what it looks like in lung cancer versus breast cancer versus maybe GI or gastric cancer?

Dr. Yu: Yeah, I think someone brought up the Venn diagram, which I think is pretty helpful. When we think about HER2 mutations, those are most often in the kinase domain, in that exon 20 space. That is about 2% of non-small cell lung cancers that have that.

When we talk about IHC or expression, very high-level expression, that 3+ in the majority of cells, is also quite rare, probably 3 to 5%. When you’re looking at lower thresholds, that can be up to 30 or 40%.

And then gene amplification, so that’s found on NGS and/or FISH, looking at copy number gain. I think the focus in lung cancer has really been the gene mutations because we have those indications and approvals.

Dr. Sabari: Yeah, so Alfredo, what are the common HER2 gene mutations that you see? I know you practice in Europe, but probably a similar population to what we see in the US. What are the common HER2 mutations? How often do you see them in your practice? How are you identifying them?

Dr. Addeo: Thank you very much for the question. I think it very much depends on where you work, in which country, whether you have access to full NGS analysis, and whether you have a custom panel or use a commercial one.

So there’s a lot of variation, which is one of the problems. We keep talking about NGS in general, but we should perhaps explain what we should be looking for and how difficult and different it could be to detect different mutations.

So in my center, at least in Switzerland, we are somewhat lucky because we can get full-panel NGS. We also have FoundationOne if needed. Therefore, I think it’s a rare mutation. It gets even rarer if you don’t test for it, and it’s about 2% at the moment.

The indication for the drugs we have is purely for mutation. There was discussion about overexpression for HER2, but at the moment, the only population we can treat is patients who have HER2 mutations, and only in the second line.

Things are going to change quite quickly, hopefully, given the data that we are going to discuss perhaps in a second. But at the moment, it’s the second line and through NGS. We don’t routinely do immunohistochemistry for lung cancer as we used to do for breast cancer.

Dr. Sabari: Yeah, we’ll definitely dig into the data. But before that, Julia, thinking about the different HER2 mutations, right? The TKD versus non-TKD. Do these matter to you? How do you differentiate these two? And imagine putting yourself in the shoes of community oncologists, where maybe all these mutations may be new, sometimes listed as ERBB2, sometimes as HER2, sometimes as exon 20.

It’s quite confusing. So how do you make sense of these different mutations, and does it matter for how you select therapies?

Dr. Rotow: And even worse, if you tested your patient a few years ago, they might not even be listed as an actionable mutation if there were no drugs approved yet. It might be hidden in the tier three mutation list on your genomic report. So definitely worth looking back if you have patients coming off first-line therapy to make sure that the field has evolved and that we’re actually catching all of these patients. I agree it’s about 2 to 4% of lung cancer where we’ll find these alterations, and I think of them as being in concentric circles, right?

So you’ve got HER2 mutations, most of which are pathogenic, but a few are not. They’re red herrings. My first question is always, “Is this mutation oncogenic? Is it pathogenic?” And then within that, of course, HER2 is interesting because you get mutations not just in the kinase domain, where the majority are, and that’s common for many of our actionable oncogenes. You can also see activating oncogenic HER2 mutations outside that kinase domain, outside the business end of the molecule, in the transmembrane domain, the juxtamembrane domain, and the extracellular domain.

And the question is, how can you identify what these are? A few are very common, like the exon 20 insertions, analogous to the EGFR exon 20 insertions. Some of these, like YVMA, are common enough that you might recognize them on sight if you see them a lot. But the uncommon ones or the less frequent ones are tricky. So I really don’t hesitate to look these things up online when I see them in my reports.

There’s a lot of great online resources. I often use OncoKB, which is a nice online resource that will give you a listing of the domain and whether it’s predicted or known to be pathogenic or not. And I’ll look these up routinely when I see patients to be sure I know what I’m treating.