Hope S. Rugo, MD: That leads us into the idea of NGS [next-generation sequencing] and when we should be doing NGS and in what patient. Tiffany, do you do NGS testing on all of your patients with triple-negative breast cancer [TNBC] in the metastatic setting?

Tiffany A. Traina, MD: I do in the early line setting, so first-line metastatic triple-negative breast cancer. Actually, at our institution, any subtype first-line metastatic breast cancer we do next-generation sequencing. In triple-negative breast cancer particularly, I'm sorry to say that we don't often find actionable alterations. I think that with the potential to uncover germline BRCA mutations in certain settings, if you're utilizing a test that would be reporting both somatic and germline, that can be important for finding a role for PARP inhibitors. There are trials accruing looking at somatic alterations that are non-BRCA that could potentially be predictive of benefit to a PARP inhibitor.

I think we're looking for potential mutations to understand the driving biology of that subtype of TNBC. Although rare, NTRK mutations are possible, so this could also be a role for next-generation sequencing. Notch alterations, where there are trials looking at gamma secretase inhibitors, are out there and accruing. And certainly to look at Akt PI3 kinase and other pathways that could be important in the heterogeneity of triple-negative breast cancer to choose particular trials at this point.

Hope S. Rugo, MD: Yes, I think that's important. We should be doing genetic testing for everybody who has certainly triple-negative disease, and I think there's a big call to do it for HER2-negative metastatic breast cancer in general. That's a really important thing for everybody to keep in mind. You hate to see a patient who hasn't been tested who should be. We tend not to because almost everybody has been tested now. When we're thinking about next-generation sequencing, personalized medicine, and the impact on care, there are so many issues to talk about here.

For a drug approved with a companion diagnostic, should it cost more? Does that warrant a higher price, and how do the payers make sure that you are using the appropriate companion diagnostic? Are they the adjudicators of who gets a drug or who doesn't based on which companion diagnostic? Can Joe Smith make a similar companion diagnostic and still have it work because it's half the price? How does that work, John?

John Fox, MD, MHA: It's interesting. If you go way back to the first molecular diagnostic other than HER testing, it was an ALK mutation in patients who were potentially eligible for crizotinib in non–small cell lung cancer, and Abbott came out with a test that cost $1500. The lab said, “Well, we can do that same mutation analysis for half that amount or even less.” I think there is this mythology, at least among payers, that all molecular diagnostics are created equal. I think we've shown in PD-L1 testing that's not true, and especially in this setting in the IMpassion130 trial where the SP142 assay produced different results than other laboratory tests, we have to re-examine that.

I would say today that most payers when they're looking for evidence to support coverage with a specific molecular diagnostic, they're agnostic to which lab is being used. I think it's really incumbent upon the NCCN [National Comprehensive Cancer Network] and investigators like you to help us understand which molecular diagnostic should be used and dispel this myth that they're all created equal. But I would say today that we're simply looking for the molecular diagnostic. I want to go back to something that Claudine talked about, the question you posed about patients who had a germline mutation in the BRCA1 and BRCA2 genes and who were also PD-L1 positive. I actually agree that today that's not in the NCCN guidelines, and the best thing for those patients would be to be in a clinical trial setting.

If you look at NTRK mutations for which we now have 2 drugs, the only way to really find out whether somebody is eligible for that is to do next-generation sequencing. With that exception, I don't know that there are any other drugs that you would want to look for unless you're interested in entering a patient into a clinical trial. But that's the common frustration that payers have, that while you may want to get next-generation sequencing done on a patient, what useable information is available outside of potential eligibility for a clinical trial?

It’s still an outstanding question. Payers can't control the drugs that are being used, especially in the Medicare population, but we can control when molecular testing is done, so there is an imperative to demonstrate how you're going to use that information, the clinical utility of that information, and how it's going to change the potential treatment for the patient.

Transcript Edited for Clarity