Blood metabolites–based MCTarg for multi-cancer screening and clinical diagnosis.

• DOI: 10.1200/jco.2024.42.16_suppl.e15028
• Publication Date: 2024-06-04T20:15:36Z
• Authors: Yan Li; Xinping Xu; Chunyan Zeng; Bei Qing; Li Liu; Guodong Song; Shuqi Yu; Tianqi Shao; Qingyan Wei; He Wen; Junyuan Hu; Wei Zhang; Youxiang Chen; Zhenkun Xia
• Keywords: 03 medical and health sciences; 0302 clinical medicine

e15028 Background: Multi-cancer early detection (MCED) tests have emerged as promising tools for reducing cancer-related healthcare costs and mortality. As metabolome is closely linked to the phenotype of bio-individual, metabolomics analysis enables high throughput investigating pathophysiological conditions. Based on machine learning, we developed Multiple Cancer Target (MCTarg) multiple cancer screening diagnosis by a single blood draw. Methods: A total 1153 individuals (172 healthy participants, 623 patients with lung cancer, gastric colorectal 358 these organ-related benign disease patients) were enrolled from three independent clinical sites in China. Plasma samples collected measured mass spectrometry (MS)-based platforms (LC-MS polar lipid). MCTarg was learning modules scenarios (community non-healthy individuals, classifying malignant diseases, tracing tumor origin). Results: exhibited an Area Under Curve (AUC) 0.96 92.82% sensitivity at 80% specificity differentiate controls, AUC 0.86 75.41% 79.66% discriminating multi-cancer diseases. Notably, classification accuracy distinguishing between conditions organ reached 90.74%, 83.87% 100%. Furthermore, overall detecting origin 84.78% average 78.08% 90.7%, respectively. Conclusions: The outstanding results indicated that blood-metabolites-based poised be cost-effective, precise, versatile tool diagnosing origin. Its potential applications encompass health checkups supplementary diagnostic scenarios. More deadly types will extended performance validated prospective population-scale cohorts longitudinal follow-up.