Pembrolizumab Shows Durable Responses in TNBC

Kathy D. Miller, MD; Hope S. Rugo, MD


December 21, 2015

This feature requires the newest version of Flash. You can download it here.

Kathy D. Miller, MD: Hi. I'm Kathy Miller, Professor of Medicine at Indiana University School Medicine in Indianapolis, Indiana. Welcome to Medscape Oncology Insights and the 2015 San Antonio Breast Cancer Symposium.

For the first time, this year's leadoff oral session focused on immunology and immunotherapy, clearly indicating the emerging importance of this area in breast cancer, and really quite a surprise to many of us. Here to talk with me about this new wave of therapies is Dr Hope Rugo, Professor of Medicine at the University of California, San Francisco, Helen Diller Family Comprehensive Cancer Center in San Francisco. Welcome, Hope.

Hope S. Rugo, MD: Thanks.

Rewriting the Book on Breast Cancer

Dr Miller: Years ago, textbooks and journals said that breast cancer is not immunogenic. It's not like renal cell cancer and melanoma. For me, the first evidence that maybe that wasn't true came a couple of years ago, when we started seeing reports about tumor-infiltrating lymphocytes (TILs) and their impact on prognosis.[1,2,3] Have we learned more about that pathology finding and what it means for us?

Dr Rugo: It's been really interesting, and very interesting for me, having worked in the field of cytokines when cytokines were coming up; we were going to use cytokines to try to cure everything. We've understood that there are immune infiltrates in aggressive cancers for a long time, including in breast cancer. We see inflammatory breast cancer, and we know that there's an immune infiltrate there.

But what's been found—and this came from neoadjuvant therapy as well as a better understanding of the host response to cancer—is that at the end of neoadjuvant therapy, you see some cancer cells and an immune infiltrate, in many situations. That led us to see what cells comprised that immune infiltrate and how they affected prognosis.

What we've discovered in some really beautiful work encompassing multiple studies and all the tissue samples available in that area—really a huge amount of work—is that more infiltrating lymphocytes are present in more aggressive cancers, as you might expect. The slower-growing cancers are indeed less immunogenic, and the presence of these immune cells or lymphocytic infiltrates correlates with a generally better outcome. This signifies, as you said, that the host's response to this cancer makes a really big deal. Those patients with TILs tend to do better.

So, of course, that's led to the idea that if you could stimulate the immune response and direct it towards the cancer, maybe you could change outcome. Potentially, this would be a way for us to identify tumors to which the host isn't responding well vs those to which the host is responding well.

But this is complicated. A lot of this is your T-cell response, which is a little bit different from the whole immune response. And macrophages play a really big role. Trying to understand exactly how to manipulate this is a big task for the future.

Deactivating the Tumor's "Cloaking Device"

Dr Miller: So that is our task. To me, the TIL brought a big "aha" moment: If the host's immune response is important, then it should be possible to stimulate or harness that response for therapy. That sounds simple. But there are a lot of steps between here and there. Early looks at the first immunotherapies, such as ipilimumab in breast cancer, really were not successful. We've seen more hope with the programmed cell death 1 (PD-1) agents and the PD ligand 1 (PD-L1) agents. We finally have some early clinical data. Has that been more promising?

Dr Rugo: It's actually fascinating. We and particularly our patients think about immunotherapy as being vaccines that stimulate the host response. But we haven't been very good at doing that. Also, we don't stimulate a T-cell response much with vaccines, and it's been hard to find antigens to stimulate. This idea of taking a whole different approach—rather than use the vaccine to increase your immunity, which is a great idea, right? It worked for measles.

Dr Miller: But we use it very differently in cancer. The terminology has never made sense to me as a nonimmunologist. For infectious disease, a vaccine is something you get before you get the disease to prevent it, but a cancer vaccine has been studied in advanced cancer. That's not how we use vaccines in any other setting.

Dr Rugo: People are looking at it certainly for early-stage disease to augment the immune system in that setting. But it still doesn't get over the problem. Our patients, particularly where I live, want to be taking agents that stimulate their immune system against the cancer or to protect them against chemotherapy.

Dr Miller: I get that question a lot; what can I do to stimulate my immune system?

Dr Rugo: I say: The problem is that your immune system isn't recognizing the cancer. It's agnostic to the cancer. The only reason the cancer can grow—and we've understood that for a long time—is that it's somehow escaped the immune system. The cancer has become like self, so the immune system can't see the cancer cells as foreign and get rid of them, which we must be doing all the time as our cells are becoming abnormal in the course of dividing and proliferating.

It's really such a cool idea that the tumor can harness the host's immune system to allow the tumor to escape recognition. It's like putting on the cloak of invisibility.

Dr Miller: It's the cloaking device.

Dr Rugo: Exactly, it's the cloaking device. The tumor causes itself to be invisible. So the task is to take away that process, which is an active process. The PD-1 protein, which is a receptor on the cell surface, interacts with ligands and allows the host to turn off immunity to that target.

Dr Miller: This is the PD-1 and PD-L1 interaction that functions in that cloaking way.

Dr Rugo: And PD-1 means programmed cell death protein. So it's really good. It makes sense.

Dr Miller: You want that PD-1 to be on.

Dr Rugo: Right. You want this programmed cell death to be turned on. And when it binds to its ligands, PD-L1 and PD-L2, it turns off that interaction. And it turns off that interaction in a complicated way, which may be a challenge for us in the future. But the idea was that if you block that so-called immune checkpoint inhibitor—the checkpoint being the cloaking device—you might be able to stimulate the host's immune system to attack the cancer. That's clearly worked for melanoma and non-small cell lung cancer in a rather striking way.

We knew from the days of interleukin (IL)-2 that those cancers were very immunoresponsive, whereas breast cancer wasn't, as you pointed out. If you look at the literature and especially more recent studies in which people looked at the expression of PD-1 and PD-L1 in breast cancer tissue, the expression of PD-1 and PD-L1 has been associated with a worse outcome, which makes sense. Also, it's been associated with more aggressive cancers, so cancers that are proliferating more rapidly—triple-negative breast cancer is a good example as well as luminal B type, and even HER2-negative. So hormone receptor-positive, more proliferative tumors have more of this pathway driving the lack of the host's ability to attack the cancer.

Durable if Weak Responses in Triple-Negative Breast Cancer

Dr Rugo: Of course, we have antibodies against PD-1, pembrolizumab, and PD-L1, atezolizumab. Those are two examples, although there are certainly many others that are being studied—and we have evidence of single-agent activity in our most highly proliferative cancer, where we don't have good treatment—namely, triple-negative breast cancer.[4] And that is the cancer that has the most expression of PD-L1.

In terms of the trials that have been looking at using these agents to treat the cancer, we've looked at expression of the ligand—PD-L1 being the most commonly expressed ligand—and then treated patients whose cancers express PD-L1 with either the PD-1 inhibitor or PD-L1 inhibitors that are out there being tested.

And it's interesting. These aren't huge single-agent response rates. But they are single-agent and sometimes very durable response rates to just an antibody. That's the first time we've seen anything like that in triple-negative breast cancer, so that's very exciting.

Dr Miller: The pembrolizumab data we saw even before this meeting.[4] Was it at ASCO that we saw the earliest results?

Dr Rugo: San Antonio 2014.[5]

Dr Miller: About a year ago, we saw those first results with real single-agent activity. What's happened with pembrolizumab in breast cancer in the last year?

Dr Rugo: That single-agent activity really held up in the first presentation of the data. There are some patients who were on the drug for 3, 4 years—just unbelievable in heavily pretreated triple-negative breast cancer. And it doesn't seem to be that there's another predictor. In the pembrolizumab trial, all you had to do was have PD-L1 expression of 1% or more. There are fewer than 30 patients in these trials because they were phase 1B expansions, but response rates are a little under 20%, and some are very durable.

Then we have atezolizumab, which is a PD-L1 antibody, and those data were presented at the American Association for Cancer Research meeting in the spring of this year.[6] What's very interesting is that it had about the same number of patients and showed about the same response rate and about the same durability. These are two different antibodies targeting different parts of this mechanism and showing very durable, low-level response rates.

Of course, then the question is: Okay, what can we combine this with to make it work better? That's the next step. You can't get enough of a response to make this the killer of all triple-negative breast cancers.

Dr Miller: But it does sound very immunotherapy-like. If you reflect back to the early days of high-dose IL-2 in renal cell cancer, the single-agent response rate was there, but it was in the neighborhood of 15%, so it's not huge.

Dr Rugo: It was even lower and very toxic.

Dr Miller: I think what got everyone excited was that, among the responses, some of them were very durable, and this feels very similar. The single-agent response rate is definitely there. It's not huge. But what is so striking is how durable some of them are.

Chemotherapy Can Stimulate an Immune Response

Dr Rugo: There are a few issues as we think forward about this. The determination of PD-L1 expression is different with every study because everybody's using their own antibody preparation and immunohistochemical test. There was a study presented here at San Antonio looking at avelumab,[7] which is another PD-L1 antibody, in a much larger number of patients. But they didn't require PD-L1 expression. Then they looked at how many patients had it. Most of the patients had low-level expression, and they only saw a response rate of 3%. It makes you think that: (1) the antibodies aren't equivalent; and (2) we really need to come up with a test that everybody agrees to and that we can do across the different trials so that we have some kind of similarity in reporting.

The combination studies are all underway and very exciting.[8,9,10,11,12] It's really interesting. What we've found is that some chemotherapy drugs more than others actually stimulate an immune response. We saw that with the TILs. You give some chemotherapy in the neoadjuvant setting, and you have more of an immune infiltrate afterwards. The idea is that if you could stimulate this immune response and then activate it by turning off the immune checkpoint, you could enhance host immunity and get some cytotoxic effect as well from the chemo drug.

There's been a lot of interest in combining pembrolizumab and atezolizumab with certain chemotherapy drugs[8,9] and also with other targeted biologic agents.[10,11,12] That's sort of the next wave of trials. Even these drugs called histone deacetylase agents might be used in cases where you amplify the T-cell compartment. I think that's the next wave that we're looking at in this area.

Dr Miller: In our last couple of minutes, I have to ask you about steroids because many of the chemotherapy agents that we use require or commonly use steroids as part of a premedication regimen to prevent significant nausea. This is predicated on a T-cell response, and my rudimentary immunology knowledge recalls that T cells and steroids are not a good match. Do we know if steroids are really a problem in hindering activity here?

Dr Rugo: It's a really good question. And it's just going to be answered by not knowing and getting rid of them. There are studies that are going on—for example, the phase 3 trial in the first-line metastatic setting with atezolizumab is using nab-paclitaxel.[12] And all of these trials are working very hard on keeping patients away from steroids.

Now, the toxicity of these agents is quite modest, and they're well tolerated, but it's these immune-related toxicities such as colitis and autoimmune colitis that concern us, and those are treated with steroids. It's not clear whether those patients lose their response. I think that we're just not going to know, but we're going to go the route of not using steroids. And that'll be interesting.

I do want to make one point, which is that in this sort of luminal A/luminal B area, we haven't done as much work. Here at San Antonio, we're presenting data on the effect of pembrolizumab in estrogen receptor-positive disease,[4] and there are data that suggest that there's more expression of PD-L1 in luminal B-like proliferative disease than luminal A low-proliferative disease. We did see a low-level, durable response in these estrogen receptor-positive patients, lower than triple-negative disease. But I think the next area of study will be differentiating out the more hormone-resistant group where immunotherapy may be quite effective; again, trying not to use steroids, for whatever it's worth.

The issue will come when we get into using carboplatin and cisplatin in these combinations because steroids are a cornerstone of antiemetics. I think that it'll be interesting to see if we can tease out how that affects the response.

Dr Miller: Thank you, Hope, for coming in and sharing your expertise with us. We will be fascinated to see this area develop. Thank you for joining us. This is Dr Kathy Miller at the San Antonio Breast Cancer Symposium 2015.


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.