Elizabeth Mittendorf, MD, PhD, discusses checkpoint inhibition studies in triple-negative breast cancer, potential ways to make other subtypes more immunogenic to improve response, the role of PD-1, and whatâ€™s next in the field of immunotherapy in breast cancer.
Elizabeth A. Mittendorf, MD, PhD
There is still a lot to be learned regarding the use of checkpoint inhibitors in breast cancer, but the field is advancing rapidly. Several agents have shown potential in recent trials.
In the phase Ib JAVELIN study, treatment with the PD-L1 inhibitor avelumab demonstrated promising results for patients with metastatic breast cancer. After a median follow-up of 10 months, the overall response rate (ORR) in all patients with PD-L1-expressing metastatic breast cancer was 33.3%. The highest response rate of 44.4% was seen in patients with PD-L1- positive triple-negative breast cancer (TNBC), although other subtypes also saw benefit.
The trial includes patients with tumors that are HER2- negative/HR-positive (42.9%), triple-negative (34.5%), and HER2-positive (15.5%). Around 7% of breast cancer cases in the trial were not identified as a particular subtype. In the full population of the study, the ORR included 1 complete response and 7 partial responses. An additional 23.2% of patients experienced stable disease with avelumab, for a disease control rate of 28%. Of those who responded, 5 had TNBC (8.6%) and 4 were PD-L1-positive. Responses were also observed in patients with HER2-negative/HR-positive breast cancer (2.8%) and in those with HER2-positive disease (3.8%).1 In addition, in the ongoing phase Ib KEYNOTE-028 basket trial—which is evaluating pembrolizumab (Keytruda) in patients with PD-L1-positive advanced solid tumors of all types—an ORR of 12% was seen in patients with ER+/HER2- advanced breast cancer.
In previously reported data from the phase Ib KEYNOTE-012 trial, pembrolizumab demonstrated an ORR of 18.5% in patients with PD-L1—positive TNBC. Elizabeth Mittendorf, MD, PhD is an associate professor in the Department of Surgical Oncology at the University of Texas MD Anderson Cancer Center.
OncLive: How has the use of checkpoint inhibition in breast cancer evolved in the past year?
In the following interview with OncLive, Mittendorf talked about checkpoint inhibition studies in TNBC, potential ways to make other subtypes more immunogenic to improve response, the role of PD-1, and what’s next in the field of immunotherapy in breast cancer.Mittendorf: We have additional data now for multiple agents in TNBC; this time last year we just had pembrolizumab.
There is a cooperative group study trial proposed to evaluate the role of checkpoint blockade in TNBC patients who receive neoadjuvant chemotherapy and have persistent disease. Our group at MD Anderson is just starting a trial looking at checkpoint blockade in combination with chemotherapy in the neoadjuvant setting for what we are defining as high-risk TNBC.
Very importantly, at SABCS (San Antonio Breast Cancer Symposium) in December 2015 there were data presented on the use of checkpoint inhibitors in subtypes of breast cancer other than TNBC. The JAVELIN trial included ER-positive, HER2-positive, and TNBC breast cancers. There are also several ongoing studies and concepts in development that are moving checkpoint inhibitors, which are showing themselves to be safe, into the neoadjuvant setting as well as the adjuvant setting.
How could that potentially be achieved?
What role does PD L1 status play in the use of checkpoint inhibitors in breast cancer?
What’s on the horizon for immunotherapy in breast cancer?
Much of the checkpoint inhibitor research in breast cancer has been focused on TNBC. Do you see potential for other subtypes as well? TNBC is an area with no targeted agents, so much of the focus has been based on need. However, if you think about checkpoint blockade, it is a drug that works by taking the breaks off of T cells, so in order for it to have an effect there needs to be T cells present. If you look at the published data, it has shown that the presence of tumor-infiltrating lymphocytes (TILs) is greatest in the TNBC subtype when compared to the other subtypes. Based on that, the thought is that TNBC will be more susceptible to checkpoint blockade. However, there is an opportunity to turn these other breast cancer subtypes, like HR-positive breast cancer, more immunogenic and then come in with checkpoint blockade after that.There are studies that have looked at strategies like cryoablation and radiation. There are data that suggest that chemotherapy in some cases can be immune stimulating. Vaccines are also a possibility. There are strategies investigating these novel intratumoral injection agents like talimogene laherparepvec (T-VEC). There are also strategies using other drugs that modify the innate immune system. There are a lot of possibilities.When this class of drugs initially started showing benefit in breast cancer, people suggested that the expression of PD-L1 was required—that it was a biomarker needed to predict response to treatment. I think most of the experts now feel that is not the case. The reason for that is that it is very dynamic how PD-1 expression can go up and down in a tumor depending on aspects of the microenvironment. The JAVELIN trial did have some interesting data that suggested that the expression of PD-L1 by the tumor was not predictive of response, but the expression of PD-L1 by other immune cells in the microenvironment did appear, at least in TNBC, to potentially be predictive. They refer to these as ‘immune hotspots.’ I do not think the trials should be designed to require PD-L1 positivity in order for patients to enroll, but that needs to be further evaluated in biospecimens collected from patients who are on the studies.I think we are going to move to an era where it is not limited to anti-CTLA-4 or anti-PD-1/anti- PD-L1 agents. There will be an opportunity to move agents that are currently being evaluated in other tumors to be evaluated in broad, phase I basket trials where many tumor types are included and to look at different agonist antibodies against co-stimulatory molecules. An example of that is OX40 or 4-1BB. I think we will also look at antagonist antibodies against inhibitory molecules other than just CTLA-4 or PD-1.
In breast cancer, the really interesting work is going to be focused on better characterizing the immune aspects of the tumor microenvironment with our current standard of care therapies, so that we can rationally add them in.
There will likely be a role for immunotherapies in most if not all of our breast cancer patients. But we need to—as a group—approach this in a very thoughtful manner. We have therapies that work in breast cancer; we cure a lot of patients. We need to figure out what those therapies are doing to the immune response so that we can better design trials that will incorporate these immune agents to be a benefit to our patients. We have a different challenge than colleagues in fields like melanoma had, where they really had few effective systemic therapies.