Pre-treatment immune biomarkers of pembrolizumab efficacy in patients with glioblastoma treated with standard of care.

2078 Background: Immune checkpoint inhibition with pembrolizumab has shown clinical benefits in multiple cancer types. However, success with checkpoint inhibition remains elusive in glioblastoma (GBM). Identifying patients a priori who would benefit from checkpoint inhibition with pembrolizumab in GBM may allow for improved survival. Currently there are no immune biomarkers that allow for selective treatment with pembrolizumab in GBM. Methods: The exploratory endpoint for our prospective single-arm study (NCT 03347617) evaluated longitudinal changes in immune cell sub-populations (B-cells, T-cells, NK cells, monocytes, and myeloid cells) in 50 newly diagnosed GBM patients who received standard of care chemoradiation with concurrent pembrolizumab after maximal safe resection. Flow cytometry was performed on blood collected at predetermined timepoints (pre-treatment, post-radiation, suspected progression, and confirmed progression) and immune cell sub-populations were evaluated as covariates in univariate and multivariate overall survival analysis while stratifying by IDH and MGMT methylation status and controlling for age and KPS using Cox proportional hazards regression in R. Results: Of the 56 patients enrolled in this trial, 50 were included in the analysis who had completed the pre-treatment blood collection. Median overall survival was 13.84 months. Nineteen patients (38%) were female and 31 (62%) were male. Median survival was 58 years (IQR: 46.25-69). Median KPS at study start was 90 (IQR: 80-90). MGMT methylation status included 23 (46%) patients not methylation and 25 (50%) hypermethylated patients, as well as 2 (4%) patients unidentified. IDH status included 39 (78%) wild-type patients, 10 (20%) mutant patients and 1 (2%) unidentified. At the pre-treatment timepoint, univariate analysis identified the following serum derived immune populations as being significantly associated with overall survival: cytotoxic T-cells (CD3+CD8+; HR: 0.98; p=0.094); CD4:CD8 ratio (HR:1.35; p=0.003); regulatory T-cells (4+127-BRIGHT25+; HR: 0.99; p=0.075); B-cells% (CD19+; HR:1.07; p=0.045). Multivariate analysis revealed significant association of CD4:CD8 ratio (HR:1.80; p=0.002) and regulatory T-cells (HR:0.98; p=0.015) with overall survival. Conclusions: GBM patients with increased serum CD4:CD8 ratios are associated with worsened overall survival prior to receiving standard of care chemoradiation and pembrolizumab. These results suggest that pembrolizumab should be withheld in patients with pre-treatment elevated levels of CD4:CD8 ratios. Alternatively, patients with higher serum regulatory T-cells have improved overall survival and may benefit from the addition of pembrolizumab. Validation of these results is needed in patients treated with standard of care versus standard of care plus pembrolizumab before implementing into clinical practice. Clinical trial information: NCT03347617 .