Early Diagnosis and Recurrence Monitoring of Colorectal Cancer

Colorectal cancer (CRC) has an insidious onset and its survival rate is closely related to clinical stage. While the 5-year survival rate of stage I patients exceeds 90%, it drops to 14% in stage IV. In China, only 15.2% of CRC cases are diagnosed at stage I, far below the rate in developed countries like the United States (24.1%). Early detection and screening are key to reducing CRC mortality and improving patient outcomes. However, current screening methods-including colonoscopy, fecal immunochemical tests (FIT), and emerging stool or blood-based biomarkers-face limitations such as invasiveness, low sensitivity, high cost, or lack of large-scale clinical validation. Importantly, these methods fail to dynamically reflect tumor evolution or assess prognosis and treatment response. Liquid biopsy has recently emerged as a promising non-invasive strategy for early cancer detection. It enables detection of tumor-related components in body fluids such as circulating tumor DNA (ctDNA), circulating tumor cells (CTCs), tumor-educated platelets (TEPs), exosomes, and cancer-specific proteins. These approaches offer a comprehensive view of tumor heterogeneity, epigenetic modifications, and treatment response. Our multidisciplinary team, from Fudan University, Xiamen University, and Xuzhou Medical University, has previously developed a cfDNA-5hmC-based diagnostic model using 5hmC-Seal technology, achieving 88% sensitivity and 89% specificity in detecting CRC. To further improve tissue specificity and diagnostic accuracy, the investigators integrated multi-omics data and advanced AI techniques (including a Transformer-based deep learning model) to deconvolute tissue origin signals. In parallel, the investigators established a charge-selective CTC enrichment platform and discovered novel metabolic markers such as HPD that may indicate tumor recurrence and metastasis. The investigators team also developed a series of nano-biosensing platforms and tumor-targeting aptamer-based d