Early Results of Immunotherapy in Breast Cancer

Phoebe Starr

June 2016, Vol 7, No 5 - Breast Cancer


Immunotherapy is a hot topic in cancer right now, with approved checkpoint inhibitors for melanoma and non–small-cell lung cancer. Checkpoint inhibitors are also making inroads in other solid tumors. Separate preliminary studies of checkpoint inhibitors in the treatment of patients with breast cancer, the JAVELIN trial and KEYNOTE-028, had less impressive results; however, these are early studies, and investigators believe that subsets of patients may benefit from this approach. Both studies were presented at the 2015 San Antonio Breast Cancer Symposium (SABCS).

JAVELIN Trial

The JAVELIN trial evaluated the investigational anti–PD-1 antibody avelumab in patients with locally advanced or metastatic breast cancer. The overall response rate (ORR) was low (4.8%), but it appears that there may be a benefit in the subset of patients with triple-negative breast cancer. Of the 8 patients who responded in this trial, 5 had triple-negative breast cancer. Overall, tumor shrinkage of ≥30% was observed in 16 (9.5%) patients. “Despite the modest overall response rate in unselected metastatic breast cancer patients, there were signs of greater clinical activity…among patients with triple-negative breast cancer and patients expressing PD-L1 [PD-1 ligand 1]. Further analysis of PD-L1 expression and clinical activity of avelumab is ongoing,” according to Luc Y. Dirix, MD, PhD, Sint-Augustinus, University of Antwerp, Belgium, who presented the results. The 168 patients enrolled in the trial were unselected for PD-L1 expression. Overall, 58 patients had triple-­negative breast cancer, 72 had estrogen receptor (ER)-positive, HER2-negative or progesterone receptor–positive disease, and 26 had HER2-positive disease; another 12 patients had an unknown molecular subtype. Avelumab was given as 10 mg/kg intravenously every 2 weeks until disease progression. The response to avelumab appeared to be related to PD-L1 expression on immune cells within the tumor—33% for those with positive PD-L1 expression versus 2.4% for tumors that were PD-L1–negative. Of the 5 responders with triple-negative breast cancer, 4 had PD-L1–positive immune cells. The safety of avelumab was acceptable, and grade ≥3 treatment-related adverse events were reported in 13.7% of patients. Immune-related side effects (ie, hypothyroidism, thrombocytopenia, and/or autoimmune hepatitis) were reported in 17 patients, and 8 (4.8%) patients discontinued treatment with avelumab. Treatment-related deaths were reported in 2 (1.2%) patients. At the time of the data cutoff, 9 patients were still receiving avelumab.

KEYNOTE-028

Pembrolizumab was evaluated in the KEYNOTE-028 “basket” trial that included patients with several types of advanced, heavily pretreated solid tumors who were screened for PD-L1. Of the 261 patients in the breast cancer cohort, 48% were PD-L1–positive; among the 25 evaluable patients with ER-positive, HER2-negative, PD-L1–positive breast cancer, the ORR was 12% (3 partial responders), with another 4 (16%) patients having stable disease. A total of 5 (60%) patients had disease progression while taking pembrolizumab, and 3 (22%) were not evaluable at the time of SABCS. The clinical benefit rate (response rate plus stable disease) was 20%. The responses were durable, ranging from 8.7 weeks to more than 44 weeks. At the time of the data cutoff, all 3 partial responders remained on therapy with pembrolizumab. Lead investigator Hope S. Rugo, MD, Director, Breast Oncology Clinical Trials Program, University of California San Francisco, was upbeat about the study. “Based on these data, we believe that further investigation of immune therapies in HER2-positive, ER-negative breast cancer is warranted, particularly if combination therapies can expand the T-cell compartment,” she told attendees. Dr Rugo and other investigators at SABCS believe that the standardization of PD-L1 assays is very important. Right now, different assays with various cut points are used by various drug companies, confounding the interpretation of the results. Dr Rugo also pointed out that some patients who are PD-L1–negative do respond to checkpoint inhibitors, so it is still questionable whether PD-L1 will be a useful biomarker. Grade 3 or 4 adverse events were reported in 4 of 25 patients in the trial; 4 immune-related grade 3 events occurred, and there was 1 grade 4 event. No treatment-related deaths occurred at a median follow-up of 7.3 months.