Performance Evaluation of Predictive Models for Detecting MMR Gene Mutations Associated with Lynch Syndrome in Cancer Patients in a Chinese Cohort in Taiwan

Novelty and Impact Our study in an ethnically Chinese cohort suspected of Lynch syndrome demonstrated high accuracy for four widely used MMR mutation risk predictive models. We compared their accuracy, suitability for specific subgroups, and accessibility. By establishing risk thresholds for each model, we refined pre-test genetic counseling, leading to more precise recommendations for genetic testing. The robust methodology with subgroup analyses strengthens the applicability and impact of these models in personalized medicine strategies.