Characterizing the epidemiology and natural history of colorectal cancer using fecal immunochemical test data from screening programs: a modelling study

Abstract

<h3>Background</h3><p>The epidemiology and natural history of colorectal cancer (CRC) is poorly understood across many populations, hindering effective disease prevention. We hypothesize that quantitative fecal immunochemical test (FIT) data from CRC screening programs can be used to characterize the prevalence and disease progression of CRC.</p><h3>Methods</h3><p>We developed an inferential framework to characterize the epidemiology of CRC using quantitative FIT data from 248,692 first-time screenees of the Hong Kong CRC Screening Program during 2016- 2021 (aged 49-77; 56.4% female). Each screenee submitted two fecal samples and was referred for colonoscopy if one or both FIT values exceeded 100 ng/ml (13.1% FIT-positivity). A natural history model, comprising both the conventional adenoma and serrated pathways, was embedded to infer unobserved disease stages in FIT-negative screenees.</p><h3>Findings</h3><p>We estimated that without screening, 41.1% (95% credible interval = 39.4%-44.0%) of males and 31.0% (29.7%-32.4%) of females would have colorectal neoplastic diseases at age 50. Among them, approximately 7-8% would have advanced colorectal neoplasia (advanced adenoma, serrated lesions, or CRC). The prevalence of advanced neoplasia increases with age, approximately by 0.5% per year after age 50. The annual progression rate to CRC was around 4% for advanced adenoma and 1-2% for serrated lesions. Males with advanced neoplasia generally had higher FIT concentrations than females. In detecting advanced neoplasia, the negative predictive value of FIT was above 99% across all ages, while the positive predictive value was 15-20% at age 50, increasing by around 1% each subsequent year. FIT demonstrates high sensitivity in detecting CRC, ranging from 83% to 98% in both genders.</p><h3>Interpretation</h3><p>Quantitative FIT-based CRC screening data can be used to effectively characterize the epidemiology and natural history of colorectal cancer. Our novel framework provides a solid foundation for optimizing the effectiveness and cost-effectiveness of screening programs.</p><h3>Funding</h3><p>Health and Medical Research Fund (HMRF), Hong Kong Special Administrative Region, China.</p>