PS4-05-24: Development of a Novel Platelet RNA Based Biomarkers for Breast Cancer Detection

Abstract Emerging evidence suggests that tumor-educated platelets (TEPs) undergo dynamic RNA splicing changes in response to cancer-derived signals, offering a minimally invasive avenue for cancer detection. In this study, we leveraged platelet RNA sequencing (RNA-seq) data to identify breast cancer-associated exon-exon splice junctions whose expression is significantly altered in cancer patients compared to healthy controls. Using a stepwise feature selection method guided by the Akaike Information Criterion (AIC), we identified 11 discriminatory splice junctions from platelet RNA profiles. These junctions were used to train a logistic regression classifier, which achieved strong predictive performance with an AUC of 0.873 on public test data and 0.949 on newly generated clinical samples, highlighting the model’s robustness and clinical applicability. Dimensionality reduction (t-SNE) and heatmap visualizations confirmed distinct separation between breast cancer and control samples based on these junction features. Gene Ontology enrichment analysis of the corresponding genes revealed significant enrichment in pathways such as endothelial cell development, establishment of endothelial barrier, and regulation of protein localization to the cell surface (FDR < 0.05), suggesting that platelet splicing alterations may reflect tumor-induced vascular and immune interactions. Our findings demonstrate that platelet-derived exon-exon junction expression profiles serve as promising biomarkers for breast cancer detection. Citation Format: J. Cheun, C. Lee, D. Shin, S. Park, T. Ahn, W. Han, H. Kim. Development of a Novel Platelet RNA Based Biomarkers for Breast Cancer Detection [abstract]. In: Proceedings of the San Antonio Breast Cancer Symposium 2025; 2025 Dec 9-12; San Antonio, TX. Philadelphia (PA): AACR; Clin Cancer Res 2026;32(4 Suppl):Abstract nr PS4-05-24.