BIOM-28. AN EXOSOME BASED LIQUID BIOPSY DIAGNOSTIC FOR GLIOMA

Abstract

Abstract BACKGROUND Liquid biopsy has gained popularity of late in the diagnosis and management of several cancers. This study aims to diagnose suspected lesions suspected of being gliomas on neuro-imaging by obtaining a peripheral venous blood sample- with a specificity and sensitivity that approaches a tissue biopsy. MATERIALS AND METHODS After clearance from Ethics Committees, specimens of the tumour tissue along with blood from a peripheral or vein was collected from all 80 adult patients with a suspected diagnosis of glioma. Clinical details along with the neurological status, the functional score, the extent of resection, pathological diagnosis and survival data were also collected. The tumour tissue transported in RNA-later and stored at -80C. Following exosome isolation, RNA-seq was performed on the exosomal sample obtained from the serum and the tumour sample. Exome analysis was performed on the tumour tissue (somatic mutations) and the WBC (for germline mutations) to a depth of 100. RESULTS The analysed data was quality checked for all glioma associated genes including but not limited to ATRX, H3F3A, IDH1, MGTM and TERT driver genes related to DNA repair, cytoskeleton and cellular proliferation-related genes EGFR, FLT1/(VEGFR), BRAF, GFAP, MKi67, NES, OLIG2, PIK3CA, SMAD3, S1001A and VIM and tumour suppressor genes SMARCB1/INI1 and TP53. Open-source cloud based genomic analysis platforms were used for analysis. Machine learning clustering algorithms identified serum exosome based clusters that corresponded to already identified glioma tissue based genomic subgroups, with specificity and sensitivity approaching 95%. CONCLUSION Liquid biopsies provide a reliable and effective way of diagnosing gliomas in patients who present with an intracranial lesion on radiology.