BIOM-15. CIRCULATING EXOSOMES CORRELATE WITH RADIOGRAPHIC PROGRESSION IN GLIOBLASTOMA PATIENTS AFTER ULTRASOUND-BASED OPENING OF THE BLOOD-BRAIN BARRIER

Abstract

Abstract Non-invasive methods to detect progression of glioblastoma (GBM) are needed to optimize patient management. Circulating blood-based biomarkers for GBM progression are limited by the blood-brain barrier, which restricts leakage of tumor-related elements into circulation. To address this, we investigated circulating GBM-associated exosomes using samples obtained as part of a clinical trial where the blood-brain barrier was temporarily opened with a skull-implantable ultrasound system (NCT04528680). Samples were collected after an intra-operative sonication procedure at multiple time points, as well as longitudinally before and after subsequent outpatient sonication cycles. Plasma exosomes were isolated using a newExoChip microfluidic device, and enrichment of tumor-derived exosomes was based on expression of Phosphatidylserine (which binds to Annexin V on the device). Tumoral origin of the exosomes was validated with GBM cell lines as positive controls, using Western blots and variant-specific ddPCR to confirm tumor contents. We observed a 2.1-fold increase in exosome concentration in plasma samples obtained intraoperatively between 2 and 45 minutes after sonication, relative to pre-operative baseline (n = 4 patients; 19 plasma samples; one-tailed t-test p = 0.02). In a longitudinal analysis of eight patients, volumetric MRI analysis of enhancement was compared with changes in exosome concentration after sonication. Among patients with short progression-free survival (PFS) (n = 4; PFS < 90 days), multi-subject analysis yielded a repeated-measures correlation of -0.59 (p = 0.06), compared to 0.01 among long PFS patients (p = 0.96). In both cohorts, use of raw pre-sonication exosome concentrations had less association with MRI enhancement (p = 0.67 and 0.98 respectively). Considering each sample independently, we found the average Pearson correlation among short PFS patients was 0.73 (avg. p = 0.39), while no correlation was seen among long PFS patients (avg. r value = 0.10; avg. p-value = 0.54). Circulating exosomes hold promise as a non-invasive method to track GBM progression upon opening of the blood-brain barrier. Further refinement and validation in larger cohorts is warranted.