Debu Tripathy, MD: Let’s shift gears into the HER2 [human epidermal growth factor receptor 2]—positive subtype. This is clearly an area where a lot of advances, both in terms of practice changing as well as understanding the biology, have happened. But in terms of clinical practice, believe it or not, after using HER2-targeted therapies for so long, there’s still controversy about the testing. We still have patients, hopefully fewer now with the amended guidelines that are in this borderline. Then there is also the issue of tumor heterogeneity that has been reported on. Adam, very briefly, can you discuss the critical issues regarding HER2 testing?

Adam M. Brufsky, MD, PhD: It’s just very sobering. It’s been 20 years, or 21 years if I’m not mistaken, since we all went to Orlando, Florida, and saw the data presented. I think we were all at that meeting in Orlando when they presented the data. It’s been 20 years, or 21 years, since they did that, and we’re still arguing about what the proper test for response is.

It’s gone back and forth. But I think the ASCO [American Society of Clinical Oncology] Guidelines are probably reasonable now, as the ones that we’re settling on. Just to recap them, if you have a copy number for HER2 by FISH [fluorescence in situ hybridization] of 6 or greater, you’re considered positive; 4 or under, you’re considered negative; in between, you’re obviously considered equivocal. But now the equivocals are being adjudicated, obviously. They’re being adjudicated with copy numbers—I mean, with HER2/CEP17 ratio. If you have a ratio of less than 2 in an intermediate range, you’re considered to be negative.

That’s a very reasonable way of doing it. In our institution—I think most institutions do this—if you’re 3-plus, we don’t do FISH. If you are 2+-plus, we then reflex you to FISH and then go through this whole copy number plus ratio. Again, we are trying to really eliminate the equivocal. I think that’s been the problem for the last 10 years or so, and part of it may be that you have to do the heterogeneity. If you don’t happen to count the right spot—our pathologists know how to count the spot that has the most enriched area when they look at a specimen. Most pathologists go that way as well. I’m curious to know what other people do at their institutions.

Debu Tripathy, MD: I would just add that there’s also the equivocal based on the HER2 copy number. That is now adjudicated by IHC [immunohistochemistry]. So you have to use both. A pathologist clearly has to be involved. About heterogeneity, I think it’s important for it to be documented. We don’t quite know how to use that data.

Adam M. Brufsky, MD, PhD: But how do you document it?

Debu Tripathy, MD: But there are ASCO/CAP [College of American Pathologists] guidelines to actually define heterogeneity. Whether those are the best ones or not, I think it is important to document that. We do probably have crude tools at this point. They are not as elegant as the ones that have been used in the lab. But we need to do that.

Ian E. Krop, MD, PhD: I completely agree with you that these new guidelines are really helpful at giving clear-cut guidance on what is positive and what is negative; and how to manage all the equivocal patients who, you know, came out based on the 2013 guidelines. We really didn’t know what to do with them, and people were sending other centromeric probes. I think these new guidelines tell you that you should not be doing that. This gives you clear-cut guidance on what to do. You do need your pathologist to be well versed in this, and it does include both IHC and FISH for the ambiguous ones. But at the end, you have a yes-no answer.

Adam M. Brufsky, MD, PhD: Biology is not a yes-no.

Debu Tripathy, MD: Right.

Adam M. Brufsky, MD, PhD: We’re not in a yes-no. It’s not a yes-no thing.

Ian E. Krop, MD, PhD: I think it is yes-no for trastuzumab, at least so far. But I as these new drugs come out, as we’ll talk about later, it may not be so clear-cut. Then there’s the heterogenic question, which we’re still not sure what to do with. We did have a study that showed very clearly that in patients treated neoadjuvantly with T-DM1 [trastuzumab emtansine] and pertuzumab, heterogeneous cancers do not do as well as those that are homogeneous. We are now going back and are trying to define with more granularity exactly how to define heterogeneity in a way that will be clinically useful. But right now, it’s not ready for prime time.

Adam M. Brufsky, MD, PhD: I want to ask this question: You would not treat someone with a ratio of 1:9, but you’d treat someone with a ratio of 2:1?

Ian E. Krop, MD, PhD: Right.

Debu Tripathy, MD: Like anything else, I guess you need cut points. But what has happened very recently now is that another assay, which is the human bioassay, is now actionable. For years, we’ve been talking about responses in the neoadjuvant setting—what to do with the residual disease in both triple-negative and HER2-positive cancers. Now we actually have actionable therapies.

Hope, do you want to review what’s new? Also, what are some new insights from the neoadjuvant trials that actually guide therapy? I’m pointing to options that are actually now impacting outcome.

Hope S. Rugo, MD: It’s been interesting, and I actually think response in HER2-positive disease in the neoadjuvant setting has driven a lot of research in the neoadjuvant setting for other subtypes. Early on, we saw that adding trastuzumab to chemotherapy in a really small University of Texas MD Anderson Cancer Center trial resulted in high pCR rates [pathologic complete response]; and those pCR rates translated into better outcome. And then there have been lots and lots of trials, mostly not powered to look at event-free survival, that have shown that different agents added in either improved pCR or didn’t. But it’s been a real way to sort of test new drugs in HER2-positive breast cancer. Of course, this led to the initial approval, accelerated approval, of pertuzumab based on a really small neoadjuvant trial looking at pCR and NeoSphere, for which you could give neoadjuvant pertuzumab. We have learned that if you get a pCR, regardless of the way you get there, your outcome is very good.

We’ve also learned that even though you get a pCR, you still have a small risk of getting CNS [central nervous system] metastases, which is unfortunate. This is an unmet need that maybe some of the new data will address in the future. I think we’ve also learned something about the heterogeneity in terms of ER [estrogen receptor] and PR [progesterone receptor]. Patients who have HER2-positive, ER-positive disease often have a lower ratio, a lower copy number, and less response to trastuzumab-based therapy. But in the adjuvant trials, they still did much better if they got trastuzumab.

So pCR, again, even with HER2 for ER-positive disease, is not the end all. But then we learned this next step. We wanted to be able to change therapy based on response for early-stage disease and change outcome forever. We go back to the Aberdeen trial, where we just sort of changed around the chemotherapy for everybody. You know, 1 size fits all. We didn’t know enough. Obviously, just changing chemotherapy has never worked so far.

But in the recent data from the KATHERINE trial, which is so interesting, it’s exactly what you would expect—70% of people who had ER-positive disease failed to achieve a pCR. And then some people had little bits of residual disease. And then they randomized them to continue the trastuzumab or get T-DM1. As we all know, the outcome was so much better in the T-DM1—based treatment group, which is very interesting. A little less than 20% got pertuzumab, but we don’t really expect, based on the adjuvant data we’ll talk about in a moment, for that to really change what we would do.

We also learned from the KRISTINE trial that if you gave T-DM1 with pertuzumab, which probably doesn’t add much to T-DM1 as far as we know, versus a standard TCHP [docetaxel, carboplatin, trastuzumab, pertuzumab] regimen, your pCRs were higher with the standard chemotherapy. But if you got a pCR with T-DM1, you did really well. So there’s a differential now in what you get preoperatively, being able to help with your prognosis and determining the next steps of therapy. Of course, this is now going to be tested in a cooperative group trial as well as others, where there will be some variation in treatment based on response and a way do deescalate as well as escalate.

Transcript Edited for Clarity