OS1.1 Plasma PD-L1 levels over time differ between glioblastoma and lower-grade glioma patients

Abstract

BACKGROUND: Immune modulating therapies are a promising treatment approach in patients with primary brain tumors, however their clinical efficacy is still under investigation as the systemic inflammatory responses are potentially challenged by the brain's tight immune regulatory mechanisms. Here, we aimed to investigate soluble PD-L1 (sPD-L1) in the plasma of recurrent glioblastoma (GBM) and recurrent WHO grade II-III glioma (LGG) patients over their course of disease. MATERIAL AND METHODS: 30 patients with recurrent GBM and 10 patients with recurrent LGG were treated with bevacizumab (400mg, q14 days) as a salvage therapy at the Medical University of Vienna. EDTA plasma samples were drawn prior to each treatment cycle. sPD-L1 was analyzed using a sandwich ELISA (Millipore ABF133 and clone 5H1). The lower limit of detection was 0.050 ng/ml as determined by serial dilutions of recombinant human PD-L1. RESULTS: sPD-L1 values could be retrieved in a median of 6 timepoints per patient (range: 2-24). No significant difference in sPD-L1 concentrations between GBM and LGG patients was present at baseline (p > 0.05). After 2 weeks of bevacizumab treatment, LGG patients presented with statistically significant lower sPD-L1 concentrations as compared to baseline (p = 0.036, Wilcoxon signed rank test), while no difference in sPD-L1 concentration was observed in GBM patients. Significantly higher sPD-L1 intrapatient variability was evident in LGG patients compared to GBM patients (p = 0.012, Mann-Whitney-U). In LGG patients, sPD-L1 correlated with leukocyte count (Spearman's r = 0.397, p = 0.001) as well as with C-reactive protein level (Spearman's r =-0.424, p = 0.001), while in GBM patients no correlation of sPD-L1 and systemic markers of inflammation was evident (-0.3 < Spearman's r < 0.3). CONCLUSION: sPD-L1 as a systemic marker of inflammation was lower in recurrent LGG patients compared to recurrent GBM patients after bevacizumab-based treatment, arguing that the gliomaimmune system interactions differ between LGG and GBM.