Translating Trial Data to Treatment in Metastatic TNBC

Yuan Yuan, MD, PhD

Immunotherapy, PARP inhibitors, and antibody-drug conjugates (ADCs) have shown efficacy in patients with metastatic triple-negative breast cancer (TNBC), but enhanced genomic understanding and proper patient selection are necessary to determine the optimal use of these agents, explained Yuan Yuan, MD, PhD.

“As a medical oncologist, I feel that for the first time ever, we have options for our patients, which is very exciting, said Yuan, “but as a research community, we need to do a better job of identifying patients properly so we can figure out [how to treat them] appropriately.”

In an interview during the 2019 OncLive^® State of the Science Summit™ on Breast Cancer, Yuan, an associate professor of breast oncology in the Department of Medical Oncology & Therapeutics at City of Hope, discussed how treating physicians are optimizing these agents for patients with metastatic TNBC.

OncLive: What are the biggest updates in metastatic TNBC?

Yuan: [My presentation] focused on 3 areas: immune checkpoint inhibitors, PARP inhibitors, and ADCs. Now we have the first FDA-approved indication in metastatic TNBC [with atezolizumab (Tecentriq) and nab-paclitaxel (Abraxane)]. The area of PARP inhibitors and potential combination approaches is still evolving, although we do have 2 approved PARP inhibitors for patients with metastatic germline BRCA-mutant TNBC. [ADCs] are very exciting; one may potentially gain FDA approval later this year.

Could you speak to the impact of the FDA approval of atezolizumab? Did its indication leave it open to use beyond the frontline setting?

The efficacy of atezolizumab is really limited to the patient population examined in the IMpassion130 trial. However, I understand that the FDA approval did not limit its use to the frontline setting due to the huge area of need. We have patients in the clinic who are PD-L1—positive and have a high tumor mutational burden but are beyond frontline therapy. To deprive them of the opportunity of getting this potentially effective therapy is detrimental. There is a huge need in the real-world setting.

Regarding the impact of the study, we now have, for the first time, this exciting combination. This is especially exciting for patients who had a 10% complete response rate. We have yet to see the matured overall survival data, but it is the beginning of an era with immunotherapy. Now we need to figure out how to turn the 59% of patients who didn’t have PD-L1 expression [into those who do] or, figure out how to turn a “cold” tumor into a “hot” tumor. We’re confident that we have a signal for immunotherapy. We just need to do a better job of selecting patients and [finding optimal combinations].

What biomarkers were examined as potential predictors of response in IMpassion130?

Investigators looked at several potential predictors such as CD8-positive cytotoxic T cells, PD-L1 expression in tumor cells, BRCA mutations, and tumor-infiltrating lymphocytes (TILs). Earlier data indicated that these patients tend to have a better survival and better responses to neoadjuvant therapy. These were looked at to see whether these predictors held in the context of this clinical trial.

I'm intrigued by the fact that we did not see a quantitative correlation between the percentage of PD-L1 in the immune cell versus tumor response. It only makes sense when you see higher responses in a higher percentage of patients. That is something I didn't understand fully. I understand that TILs and CD8-positive T cells are probably important, but they do not trump PD-L1 expression.

As far as PD-L1 expression is concerned, it is a moving target. There is evidence showing that it's a temporary venue on a biopsy. For PD-L1—negative patients, there has been evidence showing that these patients may respond to immune checkpoint inhibitors. Many research groups are looking into the biopsies of treatment-naïve patients. Like I said, it's a moving target. Once you expose a patient to an immune checkpoint inhibitor, can you turn a PD-L1–negative tumor into a PD-L1–positive tumor? The biomarker is interesting but doesn’t explain everything completely. Our understanding is probably just the tip of the iceberg. We need to better understand the interplay with other cells—natural killer cells, dendritic cells, macrophages, collagen, and even the stromal cells, which are usually immunosuppressive.

Should future investigations with immunotherapy be restricted to the frontline setting?

Probably not. I have a phase I trial for patients with heavily pretreated TNBC in which we did not use PD-L1 expression as a limitation factor. Patients were treated with entinostat, which is an HDAC inhibitor, in combination with ipilimumab (Yervoy) and nivolumab (Opdivo). We've seen responses in heavily pretreated patients in whom you may not expect activity. The host’s immune system and health [is likely related], but there's no simple answer.

Do PARP inhibitors have potential activity in patients without germline BRCA mutations?

Likely. A big question for us is BRCA-ness or those tumors that may not carry the germline mutation, although they may have methylation levels. They may have hypermethylation in which the tumor mimics a germline mutation; however, we have no reliable way to test them. Myriad has an assay called myChoice HRD. Beyond that, there is also a recently arising marker [that is a] simple immunohistochemistry [test]. If it works, it may [enable] a functional read out of BRCA-ness.

We’re in the process of proposing a study to expand the use of PARP inhibitors beyond patients with germline mutations because they account for 10% to 20% of all patients with TNBC, whereas BRCA-ness HRD tumors may account for approximately 40% to 45% of patients. Therefore, the question is, can we secure these biomarkers and then apply a de-escalated approach to therapy? These patients really don’t need a huge dose of chemotherapy. We're in a dilemma. For patients who have already achieved a pathologic complete response with chemotherapy, why would we want to subject them to the risk of lifetime immune toxicity? We need to do a better job of finding out who these patients are, but it's not easy.

Could you discuss the differences between the ADCs being investigated?

They’re quite different. For example, sacituzumab govitecan operates on an anti—TROP-2 receptor. Ladiratuzumab vedotin (SGN-LIV1A) operates by way of an anti-LIV receptor. Beyond that, we have an internal study with principal investigator Mihaela C. Cristea, MD, of City of Hope. She’s studying mirvetuximab soravtansine; on the surface of triple-negative cells, we think it’s naked. I explain to patients that there’s no estrogen receptor (ER), no progesterone receptor, and no HER2. However, there are other receptors that we have not discovered yet. There's a lot more work to be done.

It’s important to emphasize the importance of tissue and re-biopsy. We heard the case about tumor phenotype shift [during the panel discussion]. We often give second opinions on cases that are presumably ER positive, but [end up having] TNBC when we redo the biopsy. Patients also need to be educated on tissue re-biopsy and get a biomarker via a genomic analysis. We're at the beginning of the genomic era. I know cost is an issue, but if we can identify the right patient, that may ultimately drive the cost down.