Serum PD-1 is elevated after pembrolizumab treatment but has no predictive value

Abstract

Immune checkpoint blockade (ICB) employs antibody-targeting of specific inhibitory receptors and ligands. The major limitations of ICB, such as high cost, limited success rate and immune-related adverse events (irAEs), highlight the need for predictive biomarkers. We analyzed pre- and post-immunotherapy serum samples of 24 patients treated with pembrolizumab for changes in PD-1 and over 1,000 additional protein markers using a multiplex proximity extension assay (PEA) to identify potential predictive biomarkers of response and/or toxicity. Candidates were selected based on the criteria that at least two patients within any of three patient groups (responders without irAEs, responders with irAEs or non-responders with irAEs) had either a > 4-fold increase or 4-fold decrease in expression post-immunotherapy. Female and male control samples were used as technical duplicates. A patient group with no response and no irAEs was used to exclude candidates. Following treatment with pembrolizumab, there was a relative increase of PD-1 in the serum of all patients, compared to controls (average 4.4-fold). We identified 7 additional serum proteins that met our candidate selection criteria. These candidate markers did not have any significant association with response or toxicity to pembrolizumab. Overall, we show that serum PD-1 increases post-therapy with pembrolizumab treatment but has no predictive value for response or toxicity in this small set of patients.