Adam M. Brufsky, MD, PhD: Joyce’s perspective is the glass half full, which is great. There’s always a glass half empty, but why isn’t it better? Let’s just step back for a minute. One thing that you talked about a little bit, Debu, but I want to push a little bit further, is in the various lines these drugs appear to have a survival benefit—the combinations appear to have a survival benefit over individual endocrine therapy alone.

I guess the issue is that a survival benefit has never been demonstrated to chemotherapy in the setting as far as we know. Yet we’re having the PFSs [progression-free survivals] that are the same in the PEARL study. With small numbers I know, but will there be a survival benefit?

Hope S. Rugo, MD: It’s a very different population of patients. What I really have taken from this is that we should be giving the endocrine therapy with CDK4/6 [cyclin-dependent kinases 4 and 6] inhibitor therapy early. Maybe that’s what our targeted therapy problems are with some of the endocrine agents, even the BOLERO-2 study, where in the past people had received prior chemotherapy, so you’re generating a multiple resistance mechanism. For the first-line, second-line, MONALEESA-7, MONALEESA, and MONARCH trials in the second line, we’re seeing these really striking survival benefits. None of those patients received chemotherapy in the metastatic setting.

Debu Tripathy, MD: I think that it makes sense to use endocrine therapy earlier for quality of life reasons. But I must say that with the patients’ heterogeneity, the fact that these patients are chosen for clinical trials, it’s difficult to say that sequence actually matters. We’ve hypothesized a lot through some models what it does, but I think we do have flexibility as to what order we might use….

Adam M. Brufsky, MD, PhD: At the 2019 San Antonio Breast Cancer Symposium they presented data on real-world analysis, and I think this is where we’re going to answer some of these questions, if we believe in that. The Flatiron database is now 2.2 million records. Say what you want about how accurate we think it is, but at the 2019 meeting, a real-world analysis was presented in the poster session on palbociclib and letrozole, and I think within the first-line setting. What they did is they looked at about 600 of each in the Flatiron database with about 3 or 4 years of follow-up. And they matched them because what you have to do is you is propensity score matching to be sure that both cohorts were roughly the same as they could be. In terms of PFS, they replicated PALOMA-2, replicated the first-line trial, and interestingly enough, actually had an OS [overall survival] benefit.

The interesting thing about this real-world data is will that actually be borne out when PALOMA-2, when the trials come out? If it does, then this may be an approach that we can use to answer some of these questions, if we believe in that. A lot of these questions will not be able to be answered by a randomized clinical trial.

Ian E. Krop, MD, PhD: But this trial isn’t addressing the survival benefit of a CDK inhibitor. It’s asking a sequence question.

Adam M. Brufsky, MD, PhD: Right.

Ian E. Krop, MD, PhD: I think the important point here is that the recommendations have always been to use hormonal therapy first, except in the cases of true visceral crisis. But if you look at what doctors are actually doing, at least in the United States, the most recent data were showing something like 60% of patients were still receiving chemotherapy as their first line of metastatic treatment for ER [estrogen receptor]-positive disease. I think the reassuring thing here is now, this is just 1 more piece of data showing that hormonal therapy in first-line treatment works well. There’s really no reason to use chemotherapy as your first line.

Adam M. Brufsky, MD, PhD: That’s a very good way to take that.

Debu Tripathy, MD: Let me move on now to factors that may tell us a priori if patients may exhibit resistance to some of these newer targeted therapies, like CDK4/6 in particular, and how we might be managing them more from a diagnostic point of view. Joyce, any insights that you would say from, we have a couple of trials that are looking at molecular testing both blood-based and tissue-based, and informing decisions of what we’ve learned from these studies?

Joyce A. O’Shaughnessy, MD: Yes, I think there was a very nice oral presentation from Nicholas C. Turner, MD, PhD, and his group on the long-awaited plasmaMATCH trial, where about 1000 women with metastatic breast cancer underwent ctDNA analysis, circulating tumor DNA analysis, by a couple of different methodologies. One is the commercially available Guardant technology. There is also the droplet digital PCR [polymerase chain reaction]. The first good news was there was good concordance between them. Secondly, looking at the ctDNA results, vis a vis the metastatic tissues, also with NGS [next-generation sequencing] was good concordance as well. That was good because I think there have been a lot of questions there about the reproducibility of these findings. But that was quite encouraging.

Secondly, the blood found what we expected we would find on NGS, is the way I took it away. These patients had had some pretreatment, about 5% had HER2 mutations or so, and 20% to 30% had PIK3CA mutations. There were AKT1 mutations. There were ESR1 mutations in a good number, etcetera. So it recapitulated the NGS, which is also very helpful.

Then patients were treated, insofar as possible, with agents such as neratinib for HER2-positive expression, alpelisib, everolimus. Patients benefitted and at about the level we would expect. Also for the AKT inhibitor with capivasertib, those patients benefitted nicely, which was great.

This is very encouraging. First of all, I thought it was a good proof of concept, a large proof of concept, which honestly was needed because we’ve been getting this in practice. We need a group of proof, number 1.

Number 2, when it comes to the issue you raised about resistance to CDK4/6 inhibitors, the list is growing. This is because there are going to be people who are primary resistant—only about 20%—but there’s going to be some, and we’re going to need to know that up front. We’re going to need these rapid turnaround technologies that are reproducible and reliable for us, and then, of course, we’ll be able to do it serially as well.

It was very encouraging, and so it was good. We can see that would be hopeful both in regard to resistance mechanisms as well as prospectively what we would treat the patient with next, and who is progressing on a CDK, which is a huge question that we have.

Adam M. Brufsky, MD, PhD: I’m not sure plasmaMATCH answered a priori, what primary resistance looks like. I think the trial answered what afterward it looks like. It’s a continuation what they did with PALOMA-3 a couple years ago. It confirmed a lot of what they had. But I’m not sure we know. I know there was stuff at San Antonio that maybe some interferon response signature or something like that. But do we really think we have anything by that now, who won’t get it up front?

Debu Tripathy, MD: The 1 marker that I think has the most evidence but probably still needs to be validated prospectively is cyclin E overexpression….

Adam M. Brufsky, MD, PhD: Right, I think that’s reasonable.

Debu Tripathy, MD: RB [retinoblastoma] tumor suppressor loss is pretty uncommon, and it does evolve over time but not commonly. So I think you’re right. The majority of the mechanisms I think still remain undefined.

Hope S. Rugo, MD: Well, also how to target those mechanisms, I think we’ll talk about it a little bit more. I think it’s really interesting to see these data presented and what can happen, but it’s very heterogeneous. It’s something like looking at resistance in triple-negative disease, incredibly heterogeneous. It’s hard to know. Whether in the future we’ll be able to use the same technology to figure out what’s gone wrong and how to target that maybe even in the earlier stage setting is interesting. This rise of ctDNA testing is so incredibly important for our patients because a lot of people in ER-positive disease, it’s really tough to get tumor. You’re biopsying bone. People are decalcifying it and missing all the receptors.

Joyce A. O’Shaughnessy, MD: There’s heterogeneity within the organ too. We had that issue too, allele frequencies can be very low, and what is the clinical relevance of that? We have a lot to learn here.

Hope S. Rugo, MD: But being able to do ctDNA is good.

Adam M. Brufsky, MD, PhD: Apparently, I didn’t realize, do you know how much fold they go, what sequencing depth they use actually in this, in Guardant? It’s like over 100,000, over 200,000.

Debu Tripathy, MD: Yes, it’s very deep.

Adam M. Brufsky, MD, PhD: It’s incredibly deep—they say they can pick up 1 molecule, apparently.

Joyce A. O’Shaughnessy, MD: Yes, but does that help us?

Adam M. Brufsky, MD, PhD: That’s a good question. In fact, if anything, it makes it worse.

Joyce A. O’Shaughnessy, MD: Yes, exactly.

Adam M. Brufsky, MD, PhD: Right, exactly.

Transcript Edited for Clarity