Population-based molecular screening for Lynch syndrome: implications for personalized medicine.

PURPOSE Molecular screening techniques are available to identify hereditary Lynch syndrome in people with newly diagnosed colorectal cancer (CRC). We aimed to determine whether decisions of patients or clinicians reduced detection of Lynch syndrome. PATIENTS AND METHODS A prospective cohort of 245 consecutive individuals with mismatch repair-deficient CRC recruited from a population-based molecular screening program of all incident patient cases of CRC in a health care region of 1.2 million inhabitants. All incident CRCs were analyzed for mismatch repair protein loss, supported by BRAF mutation and microsatellite instability testing. Advice regarding referral for germline testing was provided to treating surgeons. RESULTS The mean age of patients was 72.5 ± standard deviation of 12 years; 64% were women; 65% had BRAF-mutant cancers. Consent for germline testing was received from 194 patients (79%): 120 with low and 74 with high likelihood of Lynch syndrome based on tumor molecular profile. Of patients who consented, 143 provided samples for germline analysis, with 12 of 143 showing a mutation (8.4%; 95% CI, 4.4% to 14.2%). Among the 102 patients who chose not to provide a sample or did not consent, an estimated 5.3 of 102 had germline mutations (5.2%; 95% CI, 2.0% to 17.5%). CONCLUSION A universal screening strategy for Lynch syndrome is potentially effective because the overall estimate of germline mutations was 17.3 of 245 patient cases (7.1%; 95% CI, 2.8% to 18.2%). However, the true value of screening is likely to be greatly limited by the decisions and circumstances of patients in taking up germline testing.