Combining genome-wide polygenic scores with registry data for colorectal cancer risk-based screening

Abstract Background Aggregating effects of common genetic variants into a polygenic risk score (PRS) has shown potential for risk-based colorectal cancer (CRC) screening. However, its utility must be evaluated in ancestrally diverse populations with diverse tumor characteristics and compared to the current screening standard, the fecal immunochemical test (FIT). Objective Evaluate the utility of PRS for risk-based CRC screening. Design The cohort included 112,204 individuals from the Copenhagen Hospital Biobank (8,995 with adenoma and 9,246 with CRC), all with linked genetic and health registry data. A subset (N=20,658) also had available FIT results. CRC PRSs were evaluated for their association with lifetime adenoma and CRC risk and their predictive value individually and combined with FIT. Results PRS stratified population-calibrated lifetime adenoma and CRC risk independently of ancestry and sex. Incidence curves showed that individuals with a high PRS reached the incidence levels of low-PRS individuals around 10 years earlier, between ages 45 to 60. PRS also stratified risk across tumor location and histology but showed no association among individuals with deficient mismatch repair tumors (N=623). Predictive modeling indicated that combing PRS with FIT did not meaningfully improve prediction of adenoma or CRC at first screening, negative colonoscopy outcomes among FIT-positive participants, or adenoma or CRC occurrence within two years after a negative FIT. Conclusion PRS effectively stratifies lifetime adenoma and CRC risk and shows promise for guiding risk-based screening initiation and intensity. When combined with FIT, it adds limited additional predictive value.