Safety and antitumor activity of pembrolizumab in advanced programmed death Ligand 1-positive endometrial cancer: Results from the KEYNOTE-028 study

Abstract

Endometrial cancer (EC) identified at an early stage is successfully treated in a majority of patients with surgery with or without radiotherapy or chemotherapy. For patients with advanced disease, however, the prognosis is poor; 5-year survival rates are less than 50% in patients with lymph node metastases and less than 20% in those with peritoneal or distant metastases. Novel, more effective therapies are needed for patients with advanced or recurrent EC. Immune approaches are promising. The interaction of programmed death ligand 1 (PD-L1) with programmed death 1 (PD-1)-an immune checkpoint receptor expressed on tumor-infiltrating T cells-can lead to their functional inactivation and prevent tumor immune evasion. Targeted immune therapy directed against PD-1 or PD-L1 can result in objective tumor responses in a wide spectrum of cancers and could play a major role in treatment of EC. A positive correlation between tumor PD-L1 expression and clinical activity against a number of tumors using PD-1 pathway blockade was the rationale for assessing the safety and efficacy of a PD-1-blocking antibody, pembrolizumab, in 20 advanced, PD-L1positive solid tumor cohorts in the KEYNOTE-028 (NCT02054806) trial. The investigators report results from the EC cohort in this article. Eligible patients had locally advanced or metastatic PD-L1-positive EC progressing after standard therapy were eligible. Pembrolizumab 10 mg/kg was administered intravenously every 2 weeks up to 24 months or until disease progression or intolerable toxicity. The primary study efficacy end point was objective response rate by RECIST (Response Evaluation Criteria In Solid Tumors, version 1.1). Secondary end points assessed included safety, duration of response (DOR), progression-free survival, and overall survival. The data cutoff for this analysis was February 17, 2016. Seventy-five patients with advanced EC were screened for PD-L1 expression; 36 of these (48.0%) had PD-L1-positive tumors, and 24 (32.0%) were enrolled. Among these 24 patients, 15 (62.5%) had received at least 2 previous lines of therapy for advanced disease; 3 of these patients (13.0%) achieved confirmed partial response (95% confidence interval, 2.8%-33.6%); median DOR was not reached. At time of data cutoff, 2 patients were still receiving treatment and an ongoing response. An additional 2 patients (13.0%) achieved stable disease; median duration was 24.6 weeks. A partial response was achieved in 1 patient who had a polymerase E mutation. Treatment-related adverse effects (AEs) occurred in 13 patients (54.2%); more than 10% of patients experienced fatigue (n = 5; 20.8%), pruritus (n = 4; 16.7%), pyrexia (n = 3; 12.5%), and decreased appetite (n = 3; 12.5%). Only 4 patients (16.7%) experienced grade 3 treatment-related AEs. There were no grade 4 AEs or immune-mediated AEs or treatment discontinuation because of an AE. These data show that pembrolizumab has a favorable safety profile and durable antitumor activity in patients with heavily pretreated advanced PD-L1-positive EC. The safety and efficacy of pembrolizumab and the utility of biomarkers including PD-L1 in a cohort of patients with EC are being investigated in the phase 2 multicohort KEYNOTE-158 trial.