Objective:

To assess if cell-free CSF RNA sequencing of patients with glioblastoma (GBM) can identify relevant cancer and GBM-specific biomarkers.

Background:

Glioblastoma (GBM), an aggressive primary brain tumor, is diagnosed via tissue biopsy. Cerebrospinal fluid (CSF) is not typically obtained from patients with GBM, due to the difficulty of access, and lack of validated CSF biomarkers. Although cell free (cf) DNA biomarkers are now commonplace in some malignancies (e.g., lung cancer, melanoma, IDH-mutant glioma), GBM mutations are variable and difficult to detect, limiting the use of cfDNA in this patient population. With newer trials utilizing ventricular access devices to monitor and deliver novel agents into the CSF, we have the opportunity to study CSF RNA biomarkers in GBM that can offer valuable insight into GBM treatment response and resistance.

Design/Methods:

Cell-free CSF (cfCSF) samples from ventricular reservoirs were collected from 6 patients with recurrent GBM, after standard therapy with radiation, chemotherapy, and resection prior to receiving B7H3 CAR-T cell therapy (NCT05474378). Normal pressure hydrocephalus (NPH) CSF samples obtained during shunt placement were used as non-cancer controls. Total RNA was extracted from 2 mL of cfCSF and sent for whole transcriptome sequencing. Differential expression was calculated between GBM and NPH using three varied statistical models (DESeq2, edgeR, and limma) and taking the commonly expressed RNAs across all methods.

Results:

Principal component analysis demonstrated clear separation of experimental and control groups. There were 165 differentially expressed RNAs between GBM and NPH cfCSF, including cancer-relevant RNAs such as the suppression of PTEN and TP53 and upregulation of TGF-Beta in GBM

Conclusions: