OS01.6.A Through the Looking Glass: CSF Proteomics for Glioma Monitoring and Response Assessment

Abstract

Abstract Background The development of novel glioma therapies necessitates rapid, longitudinal, and individualized feedback from each patient's tumor. Due in part to the relative inaccessibility of dynamic gliomas, little information is available regarding mechanisms of in situ therapeutic response and resistance which may be needed to iteratively develop candidate therapies and combinations thereof. To address this challenge, we are performing glioma biomarker discovery by serial cerebrospinal fluid (CSF) sampling from Ommaya reservoirs to determine how the CSF proteome can reveal early longitudinal intelligence regarding glioma status, biology, and therapeutic response. Material and Methods Global proteomic analysis of CSF was performed on the Somalogic platform -- an aptamer-based technology for highly sensitive and specific analyses of over 7,000 proteins. Discovery analysis comprised of the top-500 ranked proteins in CSF from seven patients with high-grade gliomas (HGG) versus non-glioma controls. The top-500 HGG proteins were then preliminarily filtered to include only proteins that met two additional criteria: 1) decrease with lower tumor burden as defined by tumor resection, and 2) increase with greater tumor burden as defined by recurrence in individual paired patient samples. Results Proteomic enrichment analysis revealed a conserved HGG CSF proteomic signature defined by 79 proteins, including ones known to be over-expressed in solid tumor malignancies, such as retinoblastoma binding protein 4, heat shock protein 90, and sorcin. When queried in an independent validation cohort, the HGG proteomic signature was consistently enriched across 13 gliomas diverse in primary versus recurrent status, subtype, and grade, when compared to control CSF samples. In two patients for whom CSF was available prior to and immediately after resections, proteins in the HGG signature decreased as indicated by a lower tumor burden. Interestingly, the glioma CSF proteomic signature was highly enriched for blood-associated proteins (as defined by two paired sets of CSF samples with low vs. high hemoglobin), despite these patients' samples containing little-to-no hemoglobin proteins. Conclusion Global CSF proteomics acquired from longitudinal neurosurgical access to unique gliomas has the promise to yield biomarkers which may be used to gain insights regarding glioma biology, tumor burden, and evolution throughout a patient's disease and treatment course.