P11.18.b aiming towards differentiating glioblastoma progression from pseudo progression -patients cohort establishment and preliminary results

Abstract BACKGROUND Longitudinal follow-up of glioblastoma patients includes advanced imaging modalities to identify tumor progression, which is of limited certainty and occasionally necessitates additional tissue sampling that has potential morbidity. Lately, the possibility of tumor-associated protein markers detection in patients' blood samples - liquid biopsies - has shown useful for this purpose in other cancers and has also gained significant interest in glioblastoma. METHODS This study presents data collected from 36histopathology-confirmed IDH-WT glioblastoma patientstreated in Sheba Medical Center between the years 2009 to 2021. Following their diagnosis, these patients were treated with radiation according to the Stupp protocol and subjected to repeated resection due to suspected tumor progressionbased on imaging studies. The data retrieved from patients' medical files include demographics, clinical outcomes, imaging characteristics, tumor molecular, andhistopathological markers. Notably, patients' selection and data retrieval were accomplished using a novel institutional AI algorithm software followed by complementary manual data retrieval of selected patients' medical files. RESULTS Patients demographics analysis showed that patientmedian age-at-diagnosis was 55.4 years (range, 21-82), and the male-to-female ratio was 3.5:1. The median overall survival (OS) was 25.1 months (range, 5.2-80.3) and the median follow-up duration was 23.1 months (range, 4.8-80.2).27 out of 36 patients had pathologically confirmed tumor progression while the remaining patients had post-radiation changes along with viable tumor or merely radiation necrosis identification. Pre-surgical resection blood samples were available for 27 out of 37 patients and will be used for a comprehensive glioblastoma-related proteins search using mass spectroscopy. In cases of patients' blood samples available both from the initial diagnosis surgical resection procedure and tumor progression surgical resection, a comparative analysis of those samples will be performed. The preliminary proteomics analysis and clinical outcomes measurements and volumetric tumor data will be presented. CONCLUSION s: Here, we present a glioblastoma patient cohort that was meticulously selected with the aim to examine our hypothesis that measuring glioblastoma-related proteins in patients' blood can be achieved and serve as a tool to distinguish between disease phases and monitor disease progression. Our comprehensive preliminary clinical, as well as proteomics data analysis, will be presented and discussed.