Ó³»­´«Ã½

Tumors likely develop over years/decades, implying that assessing long-term exposure to risk factors is crucial in cancer epidemiology. An increasing trend of long-term risk factor exposures starting from early life since the mid-twentieth century appears to have contributed to the epidemics of early-onset cancer (EOC) worldwide. A rising incidence of EOC has been reported in various body sites such as the bone marrow, bile duct, breast, colorectum, esophagus, gallbladder, head and neck, kidney, liver, pancreas, stomach, and uterine corpus. To address an intractable gap between long-term exposure assessments and tumoral molecular/microenvironmental profiling in EOC research, here we describe a framework using the prospective cohort incident-tumor biobank method (PCIBM), which was recently conceptualized. The PCIBM enables prospective molecular pathological epidemiology research that can link long-term exposures with tumor pathogenic signatures. We illustrate this framework using the study of early-onset colorectal cancer (EOCRC). First, one recognizes overlaps of the characteristics of EOCRC and later-onset counterparts. Second, EOCRC tumoral, multi-omic, or microenvironmental features are discovered and replicated. Third, using the PCIBM, long-term exposure variables are examined in relation to the incidence of all-age colorectal cancer subtypes possessing EOCRC tumoral features. Fourth, identified putative risk factors are tested for EOCRC incidence. This framework, which has provided etiological insights and advanced our understanding of EOCRC pathogenesis, is widely applicable to EOC in various organs. In addition, this research modality with artificial intelligence-driven computational tools should be used in lifecourse and other prospective cohort studies to improve our knowledge of EOC pathogenesis.