PD-L1–Positive Recurrent or Metastatic Triple-Negative Breast Cancer

Although the phase 3 confirmatory trial IMpassion131 (ClinicalTrials.gov Identifier: NCT03125902), which evaluated atezolizumab in combination with paclitaxel compared with placebo plus paclitaxel, found improvement in median overall survival in patients with programmed death-ligand 1(PD-L1)–positive metastatic triple-negative breast cancer (TNBC), researchers were unable to declare statistical significance for progression-free survival, the trial’s primary end point, in the frontline setting.¹ Do you think these updated findings will influence the drug’s final approval with the US Food and Drug Administration (FDA) or enthusiasm for immunotherapy combinations in TNB

Data from several large phase 3 studies involving immunotherapy combinations in TNBC, including IMpassion130 (ClinicalTrials.gov Identifier: NCT02425891) and KEYNOTE-355 (ClinicalTrials.gov Identifier: NCT02819518) in advanced and metastatic settings, as well as IMpassion031 (ClinicalTrials.gov Identifier: NCT03197935) and KEYNOTE-522 (ClinicalTrials.gov Identifier: NCT03036488) in the neoadjuvant setting, appear promising for immune checkpoint blockade in TNBC. Although longer-term follow-up is needed to determine the impact of immune checkpoint blockade in the neoadjuvant setting for patients with localized disease, improvement in the rate of pathologic complete response (pCR) is a strong initial signal of efficacy.

Discordance between the results of IMpassion130,² which used nab-paclitaxel as the chemotherapy backbone, and IMpassion131, which used paclitaxel, is an interesting observation. These data suggest that paclitaxel might not be the ideal chemotherapy backbone to use in combination with immune checkpoint blockade in TNBC. Some have speculated that the use of steroids with paclitaxel might limit the efficacy of immunotherapy in this setting.

Therefore, although a deeper analysis of data from IMpassion131 is needed to better understand this apparent discordance, enthusiasm for immunotherapy combinations in TNBC remains high, based on signals of efficacy observed in the other large phase 3 studies we discussed earlier. Notably, in KEYNOTE-355, patients received a variety of chemotherapy backbones, including paclitaxel, nab-paclitaxel, and the combination of gemcitabine and carboplatin.³ Data on whether specific chemotherapy regimens used in combination with pembrolizumab in KEYNOTE-355 affected efficacy have not been released but will provide greater clarity on this issue.

Based on the phase 3 KEYNOTE-355 trial, the FDA will give priority review to a new supplemental biologics license application that seeks accelerated approval for pembrolizumab in combination with chemotherapy for PD-L1–positive patients with TNBC who have metastatic or locally recurrent unresectable disease.⁴ Was this expedited review warranted? What will subsequent confirmatory trials have to demonstrate?

Two — among several — criteria for granting priority review are (1) evidence of increased effectiveness in treatment and (2) evidence of safety and effectiveness in a new subpopulation. Data from KEYNOTE-355 support those 2 relevant criteria. I believe that the decision by the FDA decision to grant priority review was certainly warranted.

Assuming that the FDA grants accelerated approval, the 2 key data points we should look out for in subsequent confirmatory studies — whether based on longer follow-up from KEYNOTE-355 or additional trials — are overall survival data when they mature and any long-term safety data and quality-of-life metrics.

Please discuss response biomarkers in TNBC. Have there been recent developments in this area of study?

In the metastatic setting, the most recent example is the use of PD-L1 by immunohistochemistry as a predictive marker of response to immunotherapy and chemotherapy. Another example is the use of germline BRCA1 and BRCA2 mutations to select for patients with a tumor most likely to respond to poly(ADP-ribose) polymerase inhibitors.

In addition, several ongoing studies in the metastatic setting are evaluating the role of phosphatidylinositol 3‑kinase (PI3K)/protein kinase B (Akt) signaling pathway alterations in predicting response to PI3K/Akt inhibitors.

Another area of active research is identifying markers of increased activity in the androgen receptor pathway that can potentially serve as predictive biomarkers of androgen receptor antagonists.

However, as you can imagine, reality is a little more complicated because TNBC is a heterogeneous disease. This heterogeneity results in an admixture of multiple variables that can influence response to therapy and prognosis in different directions within a single patient. Therefore, our research group is leveraging machine learning to integrate multiple clinical, radiologic, pathologic, genomic, and immunologic characteristics to better predict response to therapy within the context of our neoadjuvant clinical trials in TNBC. Being able to predict response is the first step to individualizing therapy and achieving the dream of precision medicine.

Ongoing randomized phase 2 trials are examining the role of platinum in early-stage TNBC. Which of these is most promising?

The study that is probably the most relevant to clinical decision-making at this point is CALGB 40603 (ClinicalTrials.gov Identifier: NCT00861705).⁵ In that study, addition of carboplatin led to a statistically significant improvement in pCR rates — 54% with carboplatin vs 41% without. It is important to note that, although long-term follow-up of patients enrolled in this study did not reveal an improvement in event-free survival, the study was not statistically powered for that end point.

Notably, adding carboplatin comes at a cost of increased toxicity. Grade 3 or greater cytopenia was more common in CALGB 40603 in patients receiving carboplatin, who were also more likely to miss more than 2 doses of weekly paclitaxel. In our practice at MD Anderson, we generally recommend an individualized approach to the use of carboplatin in the neoadjuvant setting. Most physicians here use it in select patients, such as patients with a larger tumor, node-positive disease, and/or young age at diagnosis.

I want to emphasize that the decision to use, or not use, carboplatin in the early-stage setting should be individualized. I encourage patients to have a thorough discussion with their treating physician when making this decision.

Another study evaluating carboplatin that warrants attention is NeoSTOP (ClinicalTrials.gov Identifier: NCT02413320).⁶ In this randomized, phase 2 study, patients with stage I-III TNBC were randomly assigned to receive (1) carboplatin and docetaxel or (2) carboplatin and paclitaxel, followed by doxorubicin and cyclophosphamide. Essentially, this study was designed to determine whether we can reduce the use of anthracyclines (eg, doxorubicin) in TNBC without compromising clinical efficacy.

The pCR rate was similar in both arms of NeoSTOP, but the toxicity profile was more favorable in the arm that did not contain the anthracycline. However, this trial is different from CALGB 40603, in that patients in both arms received carboplatin. Depending on what long-term data show, this might give us an opportunity to reduce the use of anthracyclines in patients with stage I-III TNBC.

A vaccine for TNBC is being investigated in combination with T-cell agonist BMS 986178, followed by sequenced administration of the programmed cell death protein-1 immune checkpoint inhibitor nivolumab. What are your concerns, if any, with a patient with TNBC receiving so many different types of therapy in a short time?

I cannot comment specifically on an ongoing trial but, in general, the biggest challenge when combining multiple agents that enhance antitumor immunity is the risk of cumulative immune-related toxicities, some of which can be quite devastating. With cancer vaccines, the current paradigm is that the vaccine will have greater efficacy if target molecules are more highly expressed in cancer cells than in normal cells. Of course, this paradigm would also suggest that the ideal strategy is to develop a vaccine against molecules that are expressed exclusively in cancer cells. Scientifically, that is a challenging endeavor.