2990 - Circulating PD-L1 levels vary across brain tumor entities and are oppositely linked to survival in glioblastoma and lower grade glioma patients

Background

Immune therapies targeting the programmed cell death receptor (ligand) 1 (PD-1/PD-L1) axis have shown remarkable activity in a variety of solid tumors. While substantial responses were observed in asymptomatic patients with brain metastases (BM), their clinical activity in primary brain tumors remains limited. In a cohort of adult brain tumor patients, we aimed to analyze soluble PD-L1 (sPD-L1) levels in patient plasma as a systemic marker of tumor - immune system interactions.

Methods

We obtained EDTA plasma from 55 glioblastoma, 26 lower-grade (WHO grade II - III) glioma (LGG), 17 meningioma and 43 BM patients (29/43 lung cancer, 6/43 renal cell carcinoma, 4/43 breast cancer, 4/43 melanoma) as well as 25 controls. sPD-L1 concentrations were measured by a sandwich ELISA using an anti-PD-L1/CD274 antibody (Millipore ABF133) as capture antibody and a non-commercial anti-PD-L1 antibody (clone 5H1) as detection antibody. The lower detection limit as determined by serial dilutions of recombinant human PD-L1 was 0.050 ng/ml.

Results

sPD-L1 was identified in 41/81 (50.6%) glioma, 5/17 (29.4%) meningioma and 6/43 (14.0%) BM patient samples and in 9/25 (36.0%) controls (p = 0.001, Chi-square), with a median concentration of 0.415 ng/ml (range: 0.050 – 42.150 ng/ml). sPD-L1 concentrations were significantly lower in BM as compared to glioma patients (p < 0.001, Mann-Whitney-U) or controls (p = 0.028). There was no observable difference in sPD-L1 concentrations and detectability over WHO grades in glioma and meningioma patients. In BM patients, a trend towards different sPD-L1 levels between primary tumors was seen (p = 0.122, Kruskal-Wallis). sPD-L1 was not correlated to prognosis in meningioma and BM patients. In contrast, glioblastoma patients had longer overall survival (OS) when sPD-L1 was detected (p = 0.006, log-rank). Conversely, sPD-L1 detectability was linked to worse OS in LGG (p = 0.028).

Conclusions

Circulating PD-L1 levels differ between patients with distinct CNS malignancies, suggesting diverse tumor - immune system interactions between primary and secondary brain tumors. Moreover, our results indicate a differential prognostic impact of sPD-L1 in glioma which should be further studied.

Clinical trial identification

Editorial acknowledgement

Legal entity responsible for the study

The authors.

Funding

Medical University of Vienna.

Disclosure

A. Ilhan-Mutlu: Advisory / Consultancy: MSD; Advisory / Consultancy: Servier. M. Preusser: Research grant / Funding (self): Boehringer Ingelheim; Honoraria (self), Advisory / Consultancy, Research grant / Funding (self): GlaxoSmithKline; Honoraria (self), Advisory / Consultancy, Research grant / Funding (self): MSD; Honoraria (self), Advisory / Consultancy, Research grant / Funding (self): Roche; Honoraria (self), Advisory / Consultancy: Bristol-Myers Squibb; Honoraria (self), Advisory / Consultancy: Novartis; Honoraria (self), Advisory / Consultancy: Gerson Lehrman Group (GLG); Honoraria (self), Advisory / Consultancy: CMC Contrast; Honoraria (self), Advisory / Consultancy: Mundipharma; Honoraria (self), Advisory / Consultancy: AstraZeneca; Honoraria (self), Advisory / Consultancy: AbbVie; Honoraria (self), Advisory / Consultancy: Lilly; Honoraria (self), Advisory / Consultancy: Medahead; Honoraria (self), Advisory / Consultancy: Daiichi Sankyo. A.S. Berghoff: Honoraria (self), Advisory / Consultancy, Research grant / Funding (self): Daiichi Sankyo; Honoraria (self), Advisory / Consultancy, Research grant / Funding (self), Travel / Accommodation / Expenses: Roche; Honoraria (self), Advisory / Consultancy: Bristol-Myers Squibb; Honoraria (self), Advisory / Consultancy: Merck; Travel / Accommodation / Expenses: Amgen; Travel / Accommodation / Expenses: AbbVie. All other authors have declared no conflicts of interest.