Adam Brufsky, MD, PhD: The next question really gets to late-stage disease, and one thing we didn't talk about here in late-stage disease are the genomic assays or the next-generation sequencing assays. Do you use those, and when do you use those?

Komal Jhaveri, MD, FACP: That's such a fantastic question. Certainly, now that we have approved therapies—such as a specific PIK3 inhibitor like alpelisib for our HR [hormone receptor]-positive patients in combination with fulvestrant based on the SOLAR-1 trial [NCT02437318]—we now want to think about this a lot more. We were doing that in our practices, but I think now we really have a reason. We also have a reason to do germline testing for BRCA, as we spoke about, because we have approval for PARP inhibitors. For my HR-positive patients, I think at the very minimum, and the most important time point for considering this would be after they progress on a CDK4/6 inhibitor, which is really what the majority of our patients do get, based on the overall survival and the progression-free survival benefits that we've seen with a CDK4/6 inhibitor.

I have utilized these tests before first-line [therapy] not to decide whether I should give a patient CDK4/6 or not, but because we have trials available at our institution including trials that look at a triple combination of, say, a PIK3 inhibitor with CDK4/6 and fulvestrant in the first-line setting. Short of that, I think the most important use would be in the post–CDK4/6-progression patient, where we want to decide what kind of treatment they should be given. Now, PIK3CA mutations are clonal mutations, truncal mutations; if you have this information even before they start their first-line therapy, it will come in handy when the patient progresses so that you don't necessarily have to wait for this 2 or 3 weeks or whatever turnaround time you might have for the kind of assay you are using—maybe plasma, maybe tumor. But I think [concerning] the utility, as you point out, when is post-CDK4/6 for HR-positive patients?

Adam Brufsky, MD, PhD: Right, and the next question is related to that. You mentioned plasma. Are you preferably using plasma for this, like a Guardant assay, or are you using tumor? Now, Memorial [Sloan Kettering] is a little bit different because your team does it on everybody that walks in the door, but I mean would you prefer plasma over a biopsy of tissue and using one of the other next-generation sequencing tests?

Komal Jhaveri, MD, FACP: I think that's a great question. I think if you think about what is the gold standard, tumor, I think, is the gold standard. Having said that, PIK3CA mutations, as we discussed, can be truncal and clonal, and so once we have that available, we know that that is what the tumor might have. ESR1 mutations, on the other hand, are subclonal mutations, so when you think about the prevalence of ESR1 mutations, it's about 20 percent in the tumor tissue but nearly up to 40 percent in the plasma. Because of the subclonal nature, you're not going to be able to necessarily find it from that 1 biopsy site because they are subclonal. I think it's a good idea to think about plasma as a way to decide what your endocrine backbone would be because you would think about utilizing a SERD should a patient have ESR1 mutations after the first-line aromatase inhibitor therapy, which is where we've seen these mutations occur.

and for other mutations that might help you assign, say, potentially for a clinical trial that might be applicable for that patient. And on the turnaround time, you mentioned Guardant, which is an approved test—a CLIA [Clinical Laboratory Improvement Amendments]-certified assay that's available. The turnaround time for that is pretty good. It's about 7 business days, so that does come in handy in clinic if one had to use that. Having said that, there is really a lot of importance to having the tumor biopsy given the heterogeneity that we see post-CDK4/6 as well and that could be valuable, but I think the patient could start some therapy if the plasma test was available and the results were available.

Adam Brufsky, MD, PhD: Got it, OK. That sounds really good. I think the one thing that's really interesting that may be a marker has to do with ER amplification. There's some emerging data that suggests ER amplification is related to response to estrogen, and that's something that we're starting to use. The other one that I think we're using that we haven’t mentioned are HER2 mutations. Do you find them usually in about probably the 5 percent, 6 percent range, and are you using a HER2 TKI [tyrosine kinase inhibitor] when that happens?

Komal Jhaveri, MD, FACP: Such a great question, again. I think when you talk about the prevalence of HER2 mutations, it's about 2 to 3 percent in breast cancer; in metastatic diseases, it's up to 8 percent, but the prevalence is really high in our lobular breast cancer patients. It can be as high as 15 percent as shown from our [research] at Memorial Sloan Kettering Cancer Center as well.

Adam Brufsky, MD, PhD: OK, fair enough.

Komal Jhaveri, MD, FACP: It's a decent number, and certainly we have trials such as the SUMMIT trial [NCT03433274] that evaluated neratinib and its role. We have now antibody drug conjugates that are looking into this as well. Certainly [they are] very important especially for our lobular patients, [and] of course [they] can be seen in ductal as well and that information can help us decide what treatment to offer.

Transcript edited for clarity.