Abstract 3343: Liquid biopsy screening tool for the earlier detection of brain tumors

Abstract

Abstract Background: There is a need for a rapid, cost-effective and non-invasive screening tool for the earlier detection and diagnosis of brain tumors. Serum based ATR-FTIR combined with machine learning algorithms can reliably predict which patients with symptoms will actually have a tumor on brain imaging. This technique gives a broad overall serum composition of the patient. However, an approach combining spectroscopy and specific cytokines within the serum could potentially provide a more accurate detection and permit prediction of tumor subtypes before histological diagnosis. Method: Patient serum samples (300uL) were collected prospectively from patients with a new brain tumor diagnosis, and patients without a brain tumor as non-cancer controls. Each patient was analysed using both ATR-FTIR and an electrochemical approach targeting specific cytokine markers. MMP-3, MMP-9 and osteopontin were selected from an unsupervised screening of 64 cytokines. The patients were discriminated using machine learning techniques in three approaches; spectral data alone, biomarker data alone and the combination of both spectral and biomarker data. Receiver operator characteristic (ROC) curve analysis was completed to display the discriminatory ability of each approach. Results: The spectral and biomarker data was collected for 49 patient samples (25 non-cancer, 24 GBM). All three approaches gave excellent discriminatory ability between cancer and non-cancer control patients. The spectral profiles alone had an area under the curve (AUC) of 0.958, the biomarker data alone had an AUC of 0.937 and the combination of spectral and biomarker data for each patient had an AUC of 0.975. The clinical datatset is being expanded to include more patients and this paper will discuss further patient analysis as well. Conclusion: The combination of using patient data from both the spectroscopic and electrochemical approaches improved the discrimination between cancer and non-cancer patients. The addition of specific marker information could support prioritization for brain imaging, and prediction of tumor subtype to guide pre-operative management decision-making. Citation Format: Ashton G. Theakstone, Paul M. Brennan, Matthew J. Baker. Liquid biopsy screening tool for the earlier detection of brain tumors [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2023; Part 1 (Regular and Invited Abstracts); 2023 Apr 14-19; Orlando, FL. Philadelphia (PA): AACR; Cancer Res 2023;83(7_Suppl):Abstract nr 3343.