Abstract 1067: Label free Raman spectroscopy of extracellular vesicles for early stage liquid biopsy cancer detection

Abstract

Abstract Aim: To improve patient outcomes, there remains a critical need to develop faster, less invasive platforms capable of identifying biomarkers. Multi-omics approaches, while promising, are high in cost and complexity, low throughput, slow, and require large sample volumes, thus are impractical for many stages of clinical care. Raman spectroscopy (RS) addresses many of these needs: it requires little to no sample preparation, is non-destructive, does not need exogenous dyes or labelling agents, and can be performed directly in aqueous solutions. Surface enhanced Raman spectroscopy (SERS) offers further improvements in terms of sensitivity but increases complexity. In this study we carried out comprehensive RS and SERS measurements on a >100-person cohort of blood and saliva from head and neck cancer (HNC) patients and benign controls. Analyses were carried out on whole biofluids, in addition to extracellular vesicles (EVs) isolated from the biofluids to investigate the added value. EVs are emerging biomarkers of interest found in high numbers in circulating biofluids with great promise to increase the specificity of diagnostic platforms. Results: We analyzed a robust clinical dataset of HNC patient saliva and plasma to build a diagnostic model. We found distinct Raman and SERS signatures of metabolites in all cancer samples uniquely present compared to non-cancer controls. Raman spectral measurements fused together for EVs and whole biofluids across saliva and plasma delivered accuracies of >95% using a novel machine learning module based on convolutional neural networks. Our platform features an automated method to isolate EVs amenable to rapid clinical use. Conclusions: The results of this study indicate an exciting step in validating Raman as a robust diagnostic tool revealing that complementary chemical information spread out across biofluids and biomarkers of interest can strongly improve simple, quick Raman based diagnostics of whole biofluids and EVs. Citation Format: Randy P. Carney, Maria Navas-Moreno, Andrew Birkeland. Label free Raman spectroscopy of extracellular vesicles for early stage liquid biopsy cancer detection [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2024; Part 1 (Regular Abstracts); 2024 Apr 5-10; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2024;84(6_Suppl):Abstract nr 1067.