Background: The role of immune checkpoint inhibitors (ICI) has not been elucidated in ER+/Her2- metastatic breast cancer (MBC). Previous studies have demonstrated a subset of these patients may derive benefit from immunotherapy in combination with endocrine therapy. We hypothesized that pembrolizumab in combination with fulvestrant would demonstrate disease control with acceptable toxicity profile.

Methods: This was a multicenter, single arm, open label Phase II Simons two-stage optimal design study in patients with ER+/HER2- MBC. The study was conducted through the BIG TEN Cancer Research Consortium. Patients received no more than 2 prior lines of endocrine or chemotherapy in the metastatic setting. Treatment consisted of pembrolizumab 200 mg every 3 weeks in combination with fulvestrant. The primary endpoint was to evaluate the clinical benefit rate (CBR: SD+PR+CR). The anti-tumor activity of pembrolizumab plus fulvestrant was measured by RECIST 1.1, irRECIST, and progression free survival (PFS). Scheduled staging scans occurred every 3 cycles (9 weeks). Secondary endpoints included safety and tolerability. PD-L1 status and next generation sequencing (NGS) by Tempus was performed on available samples to investigate biomarkers of response and the underlying mutational landscape. (NCT03393845).

Results: Forty-seven patients were enrolled, with median age of 61. Majority of patients (n=32, 70%) received 1 prior line of therapy in the metastatic setting (range 0-2). Forty patients (87%) had prior treatment with CDK4/6 inhibitors, with 1 patient who received it in the adjuvant setting on a clinical trial. Eleven patients (24%) received prior fulvestrant, and 1 received chemotherapy in the metastatic setting. Forty-six patients were treated with at least 1 cycle, 44 were efficacy-evaluable, and 43 remained on study at the time of first restaging imaging. One patient came off study due to clinical progression and was included in the final response analysis. Two patients without radiographic or clinical progression withdrew after adverse events unrelated to the study drugs and were not included in the final response analysis. The median PFS in evaluable patients was 3.2 months (range 0.2- 23.3). The CBR at 18 weeks was 36.4% (n=16/44) by RECIST 1.1. Of the 43 patients who underwent first restaging scan, 22 (50%) patient achieved disease control, including 17 patients with stable disease and 5 patients with partial response. Amongst patients who achieved disease control, the median duration of response was 6 months (range 2.3-21.3 months). Twenty-two patients had either radiographically confirmed progression on first restaging scan or clinical progression. The most common treatment related adverse effects (TRAEs) were AST elevation (n= 15, 31.9%), ALT elevation (n= 12, 25.5%), and fatigue (n= 14, 29.8%). Majority of TRAEs (93.5%) were G1-2. TRAEs of interest included hypothyroidism (n= 8, 17.0%), hot flashes (n= 3, 6.4%), arthralgia (n= 4, 8.5%), and pneumonitis (n= 1, 2.1%, G2). There were 7 (15%) patients who experienced G3 toxicities (arthralgia, anemia, hypocalcemia, decreased lymphocyte, rash, weight loss) and of these one patient (2.1%) discontinued the study drug due to elevated AST/ALT. There were no G4+ TRAEs. In patients with available tissue for NGS utilizing the Tempus platform (648 gene panel for next generation sequencing, whole exome, and RNA sequencing), genomic alterations and tumor mutational burden were evaluated. Response will be correlated with PD-L1 status by combined positive score (CPS), prior fulvestrant exposure, BMI, and biomarkers identified on NGS as exploratory analyses.

Conclusions: The combination of pembrolizumab and fulvestrant in ER+/HER2- MBC demonstrated a manageable toxicity profile with durable response in patients who achieved disease control. The combination may be considered for further exploration to better understand biomarkers of response to ICI in this population.

VIEW THE PUBLICATION