Diagnostic yield of genetic screening in a diverse, community-ascertained cohort

Abstract

AbstractBackgroundPopulation screening for genetic risk of adult-onset preventable conditions has been proposed as an attractive public health intervention. Screening unselected individuals can identify many individuals who will not be identified through current genetic testing guidelines.MethodsWe sought to evaluate enrollment in and diagnostic yield of population genetic screening in a resource-limited setting among a diverse population. We developed a low-cost, short-read next-generation sequencing panel of 25 genes that had 98.4% sensitivity and 99.98% specificity compared to diagnostic panels. We used email invitations to recruit a diverse cohort of patients in the University of Washington Medical Center system unselected for personal or family history of hereditary disease. Participants were sent a saliva collection kit in the mail with instructions on kit use and return. Results were returned using a secure online portal. Enrollment and diagnostic yield were assessed overall, and across race and ethnicity groups.ResultsOverall, 40,857 people were invited and 2,899 (7.1%) enrolled. Enrollment varied across race and ethnicity groups, with the lowest enrollment among African American individuals (3.3%) and the highest among Multiracial or Other Race individuals (13.0%). Of 2,864 enrollees who received screening results, 106 actionable variants were identified in 103 individuals (3.6%). Of those who screened positive, 30.1% already knew about their results from prior genetic testing. The diagnostic yield was 74 new, actionable genetic findings (2.6%). The addition of more recently identified cancer risk genes increased the diagnostic yield of screening.ConclusionsPopulation screening can identify additional individuals that could benefit from prevention, but challenges in recruitment and sample collection will reduce actual enrollment and yield. These should not be overlooked in intervention planning or in cost and benefit analysis.