Use of the Utah population database to evaluate statewide use of genetic testing for hereditary breast/ovarian cancer.

Abstract

e22529 Background: Genetic testing for hereditary breast/ovarian cancer (HBOC) continues to be underutilized, and options for population-based assessment of testing barriers and outcomes are lacking. This project uses linkages between statewide data sources available through the Utah Population Database (UPDB) to establish an infrastructure for studying the use of HBOC cancer genetic testing across a state. Methods: Clinical HBOC testing data from 1994-2018 was obtained for the University of Utah Huntsman Cancer Institute, Intermountain Healthcare, Utah Cancer Specialists, and the Salt Lake Veterans administration via electronic imports of tests attributed to these healthcare systems from three commercial laboratories. Genetic testing was linked to external data through the UPDB to determine demographic and urban/rural designation. Cancer diagnoses were obtained from the Utah Cancer Registry, and genealogies from the Utah Resource for Genetic and Epidemiology Research. These variables were matched to data available for the individual at the date of testing. For individuals with multiple genetic tests, the date for the first test was used. Results: Testing data was available for 12983 individuals who linked to additional records within the UPDB. Tested individuals were 86% White, 9% Hispanic, and 16% lived in rural/frontier areas. 75% of tests were performed between 2011-2018. 1575 (12%) had >1 pathogenic variant (PVs) identified in an HBOC gene, with the majority of PVs being in BRCA1/2 (89%), and TP53, CHEK2, and ATM each accounting for 2% of PVs. 7178 cancers were diagnosed in 5980 individuals (46%, avg. 1.2 cancer/person). Cancer cases were evaluated to determine if National Comprehensive Cancer Network (HBOC 2018) criteria were met. Cancer cases who have a relative with BC < 50 years of age or a relative with ovarian cancer (OC) were more likely to have a have BRCA1/2 PV than cases not meeting those criteria (17.5% vs 6.1% and 22.3% vs. 6.3% respectively). Cancer cases meeting criteria due to family history of pancreatic cancer also had a higher rate of PVs (13.1% vs. 8.4%) predominately due to additional PVs in BRCA2. Conclusions: This project begins to address the challenge of population assessment of HBOC genetic testing. We established a regulatory infrastructure to share testing data between multiple healthcare systems. In collaboration with commercial laboratories, genetic testing data was obtained in a consistent, discrete format even though it is stored differently within each health care system. The majority (54%) of HBOC testing in Utah is happening in people who did not have cancer at the time of their test, and focusing on assessing testing of cancer patients will not provide comprehensive information on testing done in the state. Among individuals with cancer, access to family history information is crucial for assessing the rate of PVs and utility of testing criteria.