Monitoring Patient Response to Pembrolizumab With Peripheral Blood Exhaustion Marker Profiles

Abstract

Exhausted T cells are effector T cells that are silenced due to continuous T cell receptor (TCR) stimulation from persistent antigens. Characteristics of exhaustion include the increased expression of multiple inhibitory receptors such as programme death-1[PD-1], lymphocyte activation gene 3 [LAG-3], T cell Ig and mucin domain [TIM-3], the loss of effector cytokine secretion and altered transcriptional profile. The PD-1/PD-L1 interaction induces functional exhaustion of tumor-reactive cytotoxic T cells and interferes with anti-tumor T cell immunity. T cell exhaustion has been observed in metastatic melanoma patients where the exhaustion of tumor specific T cells suggests that tumor clearance has been impeded and contributed to tumor immune escape. Checkpoint immunotherapies are antibodies designed to block the interaction between the inhibitory receptors expressed on T cells and their respective ligands. Therapies such as anti-PD-1 (Pembrolizumab and Nivolumab) block these inhibitory receptors and are associated with a significant improvement in overall survival and progression free survival. However, only 20–40% of metastatic melanoma patients experience long-term benefit. In a cohort of 16 metastatic melanoma patients receiving pembrolizumab, blood was serially collected before each infusion (mean 8.3; range 1–12 cycles). The presence of inhibitory markers LAG-3, TIM-3, and PD-1 on the surface of T cells was examined and assessed in relation to patient response to identify if inhibitory markers can be used to differentiate responders from non-responders for Pembrolizumab. We confirmed that across a range of cycles (range 1–26) of pembrolizumab, PD-1 expression was significantly higher on CD4+ T cells from non-responders compared to responders and TIM-3 expressed on the surface of CD8+ T cells was significantly higher in non-responders compared to responders. This longitudinal data confirms previous studies that assessed single timepoints. This study provides preliminary evidence that PD-1 and TIM-3 may be predictive of non-responders when assessed over multiple treatment cycles.