Breast Cancer at ASCO 2019: Immunotherapy, CDK4/6 Inhibitors, and Incorporating Recent Approvals Into Clinical Practice

Kathy D. Miller, MD;  Harold J. Burstein, MD, PhD


June 07, 2019

This transcript has been edited for clarity.

Kathy D. Miller, MD: Hi. It's Dr Kathy Miller, coming to you from the 2019 American Society of Clinical Oncology (ASCO) annual meeting. I'm joined by my good friend, Dr Harold Burstein, to talk about updates in breast cancer. He is a medical oncologist in the breast cancer program at Dana-Farber and is now professor of medicine at the Harvard Medical School.

Harold J. Burstein, MD, PhD: I want you to have some respect for me now.

Miller: I always have. Thank you so much for coming in.


Burstein: As always, a lot is going on in breast cancer here, and I think there are several important stories that people need to know about as we move forward into several new areas of treatment. Let's start with immunotherapy (IO) because everybody's doing IO these days. At this meeting, in contrast to last year's, we have a couple of negative reports which show some of the challenges in moving forward with IO and breast cancer. A randomized study looking at chemotherapy with eribulin, given with or without pembrolizumab, in women with hormone receptor–positive breast cancer[1] really showed no signal of activity from the pembrolizumab. Then a press release came out about a week ago about a study looking at single-agent pembrolizumab versus chemotherapy for women with triple-negative breast cancer . Again, there was no survival signal for the IO approach. It's interesting to learn from the previous experience with the IMpassion130[2] data, which looked at atezolizumab and nab-paclitaxel in triple-negative breast cancer, and try to understand how we can build on that. In that study they seemed to have a biomarker with the PD-L1 immune cell expression in the tumor. These latter two studies did not use that.

Miller: I want to ask you about the biomarker. Progression-free survival (PFS) in IMpassion130 for the entire group was 2 months' improvement with atezolizumab, and in the PD-L1-positive group it was 2.5 months, which does not sound like great selection.

Burstein: It's not great, but if you look at the survival data, the benefits are more robust. And if you look at our friends in lung cancer and some of the other malignancies where they are using IO regularly, I think it's fair to say that they are still teasing out where the biomarker threshold should be. I would not maintain that PD-L1 is the be-all and end-all; I think it's just where we start. Probably enriching patients who have some biomarker-driven [disease] is going to be the way to go forward with the IO.

Miller: We did have an update for the IMpassion130 study here, looking at more overall survival (OS) data.[3] The first report had very immature OS data. It's still immature OS data but we're getting a bit more. Hazard ratios are starting to go up a bit. Do we know when we might actually see final OS data?

Burstein: I don't know the answer to that. I will confess that we are using this in our practice, so I will tell you personally as I test more patients that I don't see the 40% of cases with the PD-L1 positivity. I don't know if you have been looking at that in your series.

Miller: We have only tested a few. The testing itself is just a complicated mess of chaos to me, because what antibody you use to test and what cells your pathologist looks for is different depending on what IO agent and what tumor type you are talking about. While they have three of those assays at Indiana University, they don't have the one used in the atezolizumab trial so we have to send our samples out. We've only tested a handful so I don't have a good sense yet if it looks like we're near 40%.

Burstein: I think we have sort of put our foot in the water for the immuno-oncology piece in breast cancer, but we actually have a lot to figure before this becomes more widely used.

CDK4/6 Inhibitors

Burstein: I think a second big story has been updates on the role of cyclin-dependent kinase 4 and 6 (CDK4/6) inhibitors. The updated data released today on the MONALEESA-7 trial[4] showed what we already knew, which is that there is an improvement in PFS but now actually a reasonably robust improvement in OS. You perhaps more than anyone in the breast oncology community have wrestled with the question of PFS versus OS as endpoints, but I think it actually is very gratifying to see that there is a survival benefit for these drugs. It validates the approach that I know your center and ours has already taken, which is that these are very commonly used products in the management of estrogen receptor (ER)–positive metastatic disease.

Miller: It's starting to become a consistent message across a couple of CDK inhibitor trials now with a longer follow-up. These are patients with ER-positive disease, so survival events thankfully are often years later. It does require patience to get to those results but a consistent story is building up.

Burstein: There was a nice update of an old SWOG trial in the New England Journal of Medicine, where adding fulvestrant first-line to anastrozole actually improved survival.[5] I think it's unclear how that relates to a more contemporary practice setting with the use of the CDK4/6 inhibitors, but it underscores the point that good drugs are going to have to have a long-term survival impact and it's important to keep looking for that. The other study I liked in the CDK4/6 space was an international study where they compared palbociclib with endocrine therapy versus capecitabine chemotherapy to see which approach would be better.[6]

Miller: I thought that one was really important because there are still many people in the community who are too quick to jump to chemo, especially for somebody who has recurred with a shorter period of time on adjuvant hormone therapy or if they have visceral disease.

Burstein: As a sidebar, Hope Rugo and I updated the ASCO guideline on management of ER-positive metastatic disease,[7] and when we pulled the data for that in contemporary practice, something like 40%-50% of patients still get chemotherapy as initial treatment. In this study—which was really quite clever, I thought—the CDK4/6-based regimen with palbociclib outperformed capecitabine chemotherapy in first-line, again using PFS estimates. But exactly to your point, this underscores that endocrine therapy with targeted treatment should be the first approach for women with ER-positive metastatic disease in almost all situations. I'm sure you and I occasionally see the true visceral crisis, but those are rare.

Miller: They are particularly rare as an initial presentation of metastatic disease with ER-positive disease. Probably the time when I would not use a CDK4/6 inhibitor is at the other end of the spectrum—the lady who is just now recurring and not symptomatic at age 85 when she had a first diagnosis of breast cancer at age 60. I'm not sure I need to jump to that as initial therapy for her.

Burstein: That may be. The US Food and Drug Administration (FDA) has looked in terms of clinical benefit, age, all the biologic subsets, volume of disease measures, and disease interval, and none of those actually predict benefit or not from the CDK4/6 inhibitor. Of course, there are patients who just don't need it at the moment, and you make a good point.

PI3K Inhibitor Therapy

Miller: We have a recently approved targeted therapy in ER-positive metastatic disease. I'm interested in your experience with it and particularly interested in your thoughts about a quandary we're going to face now when somebody with a PIK3CA-mutated ER-positive tumor pops up. Is it a CDK4/6 inhibitor? Is it a PI3K kinase inhibitor? What are they best served by getting?

Burstein: With 20 years of research in PI3K inhibitors, the problems were probably drug targetability—mostly because they were relatively dirty drugs. Alpelisib is a relatively alpha-selective PI3K inhibitor but also has a better toxicity profile. The SOLAR-1 data from Novartis showed that the drug improved PFS in second-line management of ER-positive metastatic disease. The FDA approved it based on a cell-free companion diagnostic, which is exciting because it kind of opens the door for that.

Miller: This is particularly helpful because many of these patients will have bone-only disease, where getting a biopsy is more challenging and tissue processing is more difficult. A liquid companion diagnostic makes this easier to employ.

Burstein: Agreed, and the mavens tell me that apparently there are 11 PI3K mutations that might predict sensitivity to alpelisib. Fortunately, I think more than 80% are from one or two mutations so you don't have to look that hard—you won't miss that many cases. I think this is going to begin to open the door to cell-free plasma testing for wider use in breast cancer genomics. Again, we're catching up to our lung cancer friends. At the moment, I think this is probably a second-line drug after the CDK4/6 inhibitors. I say that because we now do have the survival benefit for the CDK4/6 inhibitors and also because in subset analysis they actually show that alpelisib still has robust activity in patients previously treated with a CDK4/6. So, to my mind, that is an argument for treating those patients after the CDK4/6 inhibitor therapy. How will you approach this?

Miller: Since this is a new approval, I have not faced this situation yet but we've certainly thought about it. I think we also would tend to use a CDK4/6 inhibitor first followed by the PI3K inhibitor and be more swayed by the toxicity profile.

Burstein: That is important because oncologists are going to have to deal with the toxicity profile. About 60% or more of patients are going to end up with hyperglycemia and many of them will need to start a metformin-type treatment. Inevitably these laboratory results will be called in at 4 o'clock on a Friday afternoon, so we're going to have to brush up on how to effectively use these drugs to control the symptoms. It is expected that that goes away once the patient stops the alpelisib, so it's not a lifelong commitment to this. I liken it to more of an antinausea treatment for chemotherapy; it's something you have to do to support people through the treatment.

Miller: It's going to be challenging because we are going to have to brush up on management of hyperglycemia and management of diabetes. And it also means that we're going to have to remember that we put the patient on metformin, so when we stop their PI3K inhibitor we don't end up having to brush up on management of hypoglycemia.

Burstein: Good point. The other thing people often get with this is a rash, and it's not the typical EGFR inhibitor acneiform rash. It's more of an urticarial-type rash, and I believe that the recommendation is for one of the over-the-counter antihistamines like Zyrtec [cetirizine]. These are things that medical oncologists in the breast cancer space will have to work on and sort of figure out how to tweak a little bit.

HER2-Positive Breast Cancer

Miller: I think, for the first time ever, when asking you about things you are excited about at the meeting, you did not mention HER2 a single time.

Burstein: Right. This is progress—we've got something else to occupy our thoughts. Of course, there are data for HER2-positive disease. Two interesting neoadjuvant studies looked at the use of TDM-1, the antibody drug conjugate ado-trastuzumab emtansine, as a neoadjuvant approach compared against chemotherapy plus trastuzumab plus pertuzumab-based approaches. One study was an update of the KRISTINE trial,[8] a Genentech-sponsored study, and the other was a presentation from Jonas Bergh and colleagues in Sweden.[9] The interesting thing here is that in some of the reports, chemotherapy plus trastuzumab plus pertuzumab actually had a higher pathologic complete response (pCR) than did the TDM-1-based regimen. But what they showed in the long-term follow-up with KRISTINE is that if you did achieve a pCR with either approach, your long-term results were the same. This fits with this theme in breast cancer of trying to get away from the chemotherapy. A lot of people are calling this de-escalation, but it's really adding different drugs and not really getting rid of drugs. But the emerging idea here is: Let's use enough therapy that we give the patient a good chance to have a pCR. For those who do, they might be done with their treatment; and for those who don't, then we can add more sort of out back. This is consistent with the message that emerged from the KATHERINE trial,[10] where patients who had residual HER2-positive disease after trastuzumab-based chemotherapy benefited from the addition of TDM-1 afterwards.

Miller: It is a great opportunity to think about what we can do [to minimize] toxicity in our patients. And perhaps not everybody needs everything to get that good outcome.

Burstein: The numbers are pretty similar, so it's not like there was 50% pCR with one approach and 10% with another. In the KRISTINE trial it was like 55% versus 45%, and in the Jonas Bergh study it was like 45% for each group. It's a very exciting approach. What we truly still don't know is whether in the long, long term there is some benefit for chemotherapy. That has yet to be proven.

Miller: That is always going to be challenging because we give people neoadjuvant therapy for the same reason you give them adjuvant. It's not really about the breast; the goal is preventing distant recurrence. Those neoadjuvant studies will continue to follow patients, but because the primary endpoint was what happened at surgery, they are always way too small to do anything but tease us.

Burstein: Even in the KRISTINE trial the data they are talking about is a 3-year follow-up which, again, isn't huge. There was a related study[11] that my colleague at Dana-Farber, Otto Metzger, ran, where they were specifically looking at outcomes in so-called heterogeneous HER2-positive cases—those cases where there is a little bit of fluorescent in situ hybridization amplification but the tumor is mixed. For those cases it's a mixed picture. A lot of the idea of the de-escalation will hinge on the quality of the pathologic review, which is not a trivial thing to figure out.

Miller: It's always a pleasure to talk to you. Thanks for coming in and updating us about this year's meeting.

Burstein: Happy to be here. Enjoy the meeting.

Miller: Thanks to you, our listeners. This is Dr Kathy Miller from ASCO 2019.

Follow Medscape on Facebook, Twitter, Instagram, and YouTube


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.