BIOM-64. TARGETING IDENTIFICATION OF HOST-BRAIN CANCER SYSTEMIC CYTOKINE SIGNALLING FOR EARLY DETECTION AND TUMOUR SUB-TYPING. (14th November 2022)

• Record Type: Journal Article
• Authors: Brennan, Paul; Baker, Matt
• Is Part Of: Neuro-oncology. Volume 24(2022)Supplement 7
• Page Start: vii19
• Page End: vii19
• Publication Date: 2022-11-14
• Subjects: Brain Neoplasms -- Periodicals; Brain -- Tumors -- Periodicals; Brain -- Cancer -- Periodicals; Nervous system -- Cancer -- Periodicals; 616.99481
• DOI: 10.1093/neuonc/noac209.074
• Languages: English
• ISSNs: 1522-8517

Abstract

Abstract: Background: The host-tumour interaction is fundamental to brain tumour biology, orchestration of which is mediated in part by cytokines. We reasoned that detection of a tumour-specific cytokine profile in patient serum could contribute to early diagnosis strategies, and permit prediction of tumour subtypes before histological diagnosis. METHOD: Patient serum blood samples (300ul) were collected prospectively from patients with radiological evidence of a new brain tumour diagnosis pre-treatment, and from symptom-matched patients without radiological evidence of a tumour (controls). Our in-house cytokine assay incorporates antibodies captured onto a nitrocellulose slide. 64 capture antibodies are printed per slide, with 4 replicates per antibody. Fluorescence is measured from a biotin secondary antibody, the signal compared to a standard curve, and the averaged relative intensity across 4 replicates determined. We used a hierarchical clustering strategy to examine the relative expression of all 64 biomarkers for each patient class, and for cancer versus non-cancer. We refined this to identify the most discriminatory biomarkers. We made a univariate examination of the association of each antibody with brain tumour type. RESULTS: Cytokine profiles were collected from 83 patients (42 glioblastoma, 8 meningioma, 12 metastatic cancer, 6 grade III gliomas, 3 grade II gliomas, 2 other brain tumours, 10 non-cancer patients). The top 4 biomarkers with greatest separation for brain cancer versus non-cancer were osteopontin, MMP3, MMP9 and TIMP-1. A principal component analysis demonstrated clear separation between clusters for these biomarkers. Univariate analysis evidenced discrimination of tumour subtypes, for example MMP9 and oligodendrogliomas. CONCLUSION: Cytokine signals can be identified in small volumes of patient serum and used to identify brain cancer 'risk.' This could support prioritization for brain imaging, and prediction of tumour subtype to guide pre-operative management decision-making. Combination with other liquid biopsy strategies such as serum spectroscopy may improve spectroscopy test performance and utility.