Pembrolizumab Extends EFS in Early Triple-Negative Breast Cancer


In an interview with Targeted Oncology, Lee Schwartzberg, MD, further discusses the findings of KEYNOTE-522 and pembrolizumab in patients with triple negative breast cancer.

KEYNOTE-522 (NCT03036488), a phase 3 trial which studied pembrolizumab (Keytruda) in patients with early triple-negative breast cancer (TNBC), showed positive results in event-free survival (EFS).

The trial enrolled 1174 patients with previously untreated stage II or stage III TNBC. Once randomized 2:1, they received neoadjuvant chemotherapy using carboplatin, paclitaxel, doxorubicin, and cyclophosphamide in addition to either 200 mg of pembrolizumab or placebo every 3 weeks, according to Schwartzberg. Every 3 weeks for up to 9 cycles following surgery, patients were given adjuvant pembrolizumab or placebo.

Pathological complete response (pCR) at the time of definitive surgery and EFS were the primary end points of the study. Some secondary end points included pCR rate using alternative definitions, overall survival (OS) in all patients randomized, pCR rate according to all definitions.

The first interim follow-up showed success as a pCR of 64.8% with pembrolizumab versus 51.2% with chemotherapy alone was reported. Additionally, a median follow-up of 39 months showed pembrolizumab to produce an improved EFS. When compared with chemotherapy alone, patients in the PD-L1–positive group had a 33% reduced risk of EFS events with pembrolizumab versus the placebo group (HR, 0.67; 95% CI, 0.49-0.92). In the PD-L1–negative group, patients who received the pembrolizumab regimen had a reduced risk for EFS events by 52% in comparison  to the placebo-chemotherapy group (HR, 0.48; 95% CI, 0.28-0.85).

The 36-month EFS rates were 84.5% versus 76.8%, respectively. Due to these results, pembrolizumab is becoming the standard of care for both node-positive and node-negative patients with stage II or III TNBC.

In an interview with Targeted OncologyTM, Lee Schwartzberg, MD, chief medical director of West Cancer Center and chief of the division of hematology/oncology at the University of Tennessee Health Science Center, further discusses the findings of KEYNOTE-522 and pembrolizumab in patients with TNBC.

Targeted Oncology: Can you provide a brief overview of what the current landscape of immunotherapy and breast cancer looks like?

Breast cancer was a little bit late to the I-O party. There were a lot of studies launched early on in breast cancer and in the variety of the different subgroups. What we found early on was that first of all, lateline patients who received I-O therapy in the form of checkpoint inhibitors didn't have a very good response rate. That was likely due to the fact that they had exhaustion of T cells, and were not very immune cells infiltrated at that time, probably as a result, multiple lines of therapy and then the progression of the disease. However, when we looked at patients who had PD-L1 positivity, and particularly in the triple-negative breast cancer patients, we did see some evidence of response. That led to studies in the first line of I-O therapy with checkpoint inhibitors with chemotherapy in TNBC, and those studies were positive.

One of the studies, KEYNOTE-355, which was a first line trial of chemotherapy and uses a variety of different chemotherapies that could be used in that setting, with or without pembrolizumab. What it showed was for patients who had a PD-L1 positive TNBC, they had a better outcome, both with improved progression free survival and improved overall survival. That led to the approval of pembrolizumab with chemotherapy in that setting. About 40% of TNBC that are metastatic have Pdl1 positivity defined in that trial as a cognitive performance scale score greater than 10%. That has become the biomarker which gives the best differentiation in terms of response or not. We have a therapy for about 40% of the triple-negative population that includes I-O, and it's become very important.

The problem is that that's a fairly small population and although patients do much better than they would with that very difficult subgroup of breast cancer, triple-negative metastatic disease, they still continue to progress. We need new strategies as well. The most important new strategy that we have developed over time is using checkpoint inhibitors earlier in the therapy of patients and again, it started with TNBC which has the least options. Chemotherapy is the main option for that group of patients, and they have a poor prognosis in general. Many patients with TNBC receive neoadjuvant chemotherapy, and that's beneficial in a number of ways. It’s prognostic if you get a pathologic complete response rate to whatever the therapy is giving it before surgery, then that is a very good predictor of excellent outcome. Secondly, we can change therapies based on that, and we can use that pathologic complete response rate as a surrogate marker for long term outcome or the ability not to see the cancer come back event free survival.

A variety of chemotherapy programs have been used successfully for TNBC. The pathologic complete response rate hovers around 30 to 40%, which means that most of those patients do very well but that means that the majority of patients without additional therapy still have a relatively poor prognosis. I-O has been added to TNBC chemotherapy in the neoadjuvant and sometimes adjuvant setting and the results have generally been very positive.

The study that has led to approval in this case with pembrolizumab, again, is KEYNOTE-522. That was a large trial and patients received aggressive chemotherapy with 4 drugs, and that included carboplatin, paclitaxel, [doxorubicin] or another anthracycline, and cyclophosphamide, all given prior to surgery. And they also either received pembrolizumab every 3 weeks with that, or placebo.

The results when initially published showed an improvement in the pCR rate which was substantial. So that's the first surrogate marker that pembrolizumab was getting. And now we have 3-year EFS of the data, which is, if you will, a harder end point. And it shows almost an 8% improvement in the risk of developing recurrent disease if you get pembrolizumab, which is really very substantial. So that has become the standard of care.

Now in KEYNOTE-522, the pembrolizumab was given both in the neoadjuvant setting and in the adjuvant setting, meaning after the surgery, whether you had a pCR rate or not, patients continued pembrolizumab and so it's based on that entire year of therapy; although not all patients did that, the majority did. So that has become the standard of care now to use a KEYNOTE–522-type of program to treat patients with stage II or III TNBC, and that includes both node-negative and node-positive patients.

An interesting thing about using I-O therapy in the early stage is that these patients have more intact immune systems. Although all patients were enrolled regardless of their PD-L1 status in both the metastatic and the adjuvant setting, in the adjuvant setting or new adjuvant setting, it didn't make a difference if they were PD-L1 positive or not in terms of their outcome, although they were a little more likely to have pCR rates for the high PD-L1. It is a prognostic marker, not a predictive marker for pembrolizumab. Most patients who present with newly diagnosed breast cancer have a more intact immune system. It looks like treating patients earlier is more important.

Do you believe anything from the study was left unanswered or do you have any further questions?

There's lots of questions that are still left after the studies. We have a new standard of care for stage 2 and 3 TNBC in the early setting, and we have a new standard of care for the 40% of patients in a PD-L1 positive metastatic setting, but it leaves many questions. Some of those questions are: Do we need that aggressive chemotherapy with pembrolizumab? What's the most important component of that chemotherapy? What of those agents synergize with the pembrolizumab? Do we need to continue pembrolizumab after achieving a pCR? There was a 2% difference in outcome favoring continuing it but we don't really know what that really means. What about populations who don't get a pCR? Should they receive additional therapy beyond pembrolizumab in the adjuvant setting?

That was not done in KEYNOTE-522, but we have now approved therapies for BRCA mutated patients with elaborate, so should we combine it with pembrolizumab in this sequence? We often use capecitabine in the adjuvant setting after a standard chemotherapy patient does not achieve a PCR, so should capecitabine also be added to pembrolizumab in the adjuvant setting for patients who don't achieve a PCR? Then turning to the metastatic setting, how do we make patients more inflamed or more immune responsive?

I think in that area, the best approach will likely be combination I-O therapy to try to re-stimulate other aspects of the immune system beyond just the PD-1 axis, which is what the checkpoint inhibitors are doing now. There is exciting work that's been done with other agents like TIGIT or a host of other approaches to attack different components of the immune system and to increase the T-cell response in patients who don't have a PD-L1 positive tumor, or even those that do and to do better, so a lot of work to do there.

There is some cost to adding pembrolizumab in the adjuvant setting. Remember that many of these patients would be cured by standard therapy. We don't know which ones those are, but we also know there's a substantial relapse rate. We are trying to improve that relapse rate and reduce it, but we have to acknowledge that treating patients with standard therapy would be cured, and the toxicity is evenly distributed among all patients. We have to manage the risk benefit ratio.

The biggest issue with pembrolizumab in the neoadjuvant setting is obviously immune effects, and most commonly, effects on the thyroid. About 20% of patients get either hypothyroid or hyperthyroid, and we think that that's most likely going to be long standing. We have to lay the benefit of treating patients and trying to prevent the cancer coming back, at which point it's typically not curable versus long term effects on the thyroid gland, or in rare cases, on the adrenal gland or on the skin. The severe acute toxicities of I-O therapy haven't really been seen very much. They're in the single digits for patients who received in a new agent setting, which is very good.

What unmet needs still exist in this area that needs to be filled by these combinations?

Even with I-O therapy there will be a substantial number of patients who relapse with TNBC. Biotherapy is definitely a major advance in the neoadjuvant setting, but there's still room for further improvement, whether that is with some other combination therapy, like I mentioned, or whether it's with a combined I-O approach. It is also a very aggressive regimen, so understanding deescalation, in terms of the intensity of therapy and preserving the high pathologic response rates and event free survival will definitely be an area as we go forward because that's what we've done in other areas of breast cancer. Once we establish an excellent outcome, then we see what components of that therapy are absolutely necessary to avoid the toxicity, long term or even short term.

In metastatic breast cancer, we are still not curing patients in any of the three major subgroups. Where I-O therapy is going to fit in or not as we've seen with other cancers, is that there's a long tail. It is a small percentage of patients, but if you look at melanoma or even lung cancer, there are patients who are now out many years with therapy. Now in breast cancer, we've made good strides. Even in the metastatic setting, other than triple-negative, our patients are living in medium times in the multiple years, 5,6,7 years. We would like to convert those to cures if possible, and I-O does have the potential to help if we can figure out the best biomarkers to determine which patients are going to respond to I-O and also to make patients more sensitive to I-O.

Disseminating this information to the community is an important aspect. It's great to have clinical trials that show good outcomes. We have to make sure that those trials are successfully and rapidly transmitted to the community. We have organizations doing that one of the founding members of the CCCI-O Institute has been in existence now from the very beginning of the I-O eras, dating back about 6 or 7 years now. That has been a wonderful tool to disseminate information and to alleviate logistical and operational logjams and financial logjams with therapy. The science is great, but unless you get it to all the patients, not just including those that have access issues and equity issues, then we are not really serving our patients well.

What in this space are you most excited for in 2022?

I'm most excited about seeing whether we can expand the use of I-O therapy. That would include the other 2 major subgroups that we have, which are the ER-positive HER2-negative subgroup and the HER2-positive subgroup. We already have some preliminary data that you can add I-O therapy to anti HER2 therapy. Although we have many good anti energy therapy patients now, we'd like to do even better in the high risk patients. I'm also very excited to start to see combination I-O therapies in breast cancer, and I think that's where will make the biggest impact in patients with with metastatic ER-positive breast cancer and potentially the TNBC that are here and to date have not responded to our standard single agent therapy with chemo.

Related Videos
Aditya Bardia, MD, MPH, FASCO, and Laura Huppert, MD, experts on breast cancer
Aditya Bardia, MD, MPH, FASCO, and Laura Huppert, MD, experts on breast cancer
Aditya Bardia, MD, MPH, FASCO, and Laura Huppert, MD, experts on breast cancer
Aditya Bardia, MD, MPH, FASCO, and Laura Huppert, MD, experts on breast cancer
Aditya Bardia, MD, MPH, FASCO, and Laura Huppert, MD, experts on breast cancer
Aditya Bardia, MD, MPH, FASCO, and Laura Huppert, MD, experts on breast cancer
Aditya Bardia, MD, MPH, FASCO, and Laura Huppert, MD, experts on breast cancer
Related Content