Abstract 6514: Liquid biopsy for lung cancer based on extracellular vesicles

Abstract Background Liquid biopsy approaches for lung cancer are becoming increasingly common for diagnosis, prognosis, and management of the disease. Of the Approximately 236K new lung cancer cases that will be diagnosed in 2023, over 70% of those will be regional or distant with diminished chances of curative resection. Therefore, highly sensitive and specific approaches for early detection are key for mortality reduction. Extracellular vesicles (EVs), a new class of blood-based cell-free biomarker, can be informative for the diagnosis of lung cancer at its earliest stages. Methods A case-control cohort of blood plasma samples from 146 pathologically confirmed lung cancer cases (stage I = 90, II = 42, III = 14; median age = 63 yrs.) and 491 controls (median age = 58 yrs) was used in this study. EVs were isolated using a proprietary technology and the EV protein cargo was analyzed via immunoassay. A machine learning engine was employed to determine the most informative biomarkers and algorithm for differentiation between cases and controls in a detection setting. Results Using a stratified cross-validation approach, we found a biomarker signature that yielded an AUC of 0.966 (95% CI: 0.949 - 0.979) with an overall sensitivity of 91.1% (CI: 85.4% - 94.7%) at a specificity of 91.0% (CI: 88.2% - 93.3%). By stage, the following sensitivities were obtained: stage I: 92.2%, stage II: 88.1%, stage III: 92.8%. The algorithm developed includes 13 EV-protein biomarkers using the adaptive boosted tree methodology. Conclusions Our pilot study demonstrates utility of a blood-based approach that can detect lung cancer at early stages, when treatment can be more effective. By combining multiple EV protein signatures, it is possible to achieve high sensitivity and specificity. Independently collected cohorts including confounding conditions, such as benign lung nodules and COPD, are being evaluated for validation purposes to improve performance in clinical settings. Citation Format: Juan P. Hinestrosa, Harmeet Dhani, Jean M. Lewis, Gregor K. Schroeder, Heath I. Balcer, Razelle Kurzrock, Paul Billings. Liquid biopsy for lung cancer based on extracellular vesicles. [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 6514.