Patterns of genomic testing for epithelial ovarian cancer across a large community-based health care network- a real world experience

Background: NCCN guidelines recommend germline and somatic tumor testing for all women with invasive epithelial ovarian cancer (EOC). Despite this recommendation, testing rates remain low and an optimal strategy to achieve the recommended testing has not been defined. Purpose: We elected to review the patterns of germline and somatic tumor testing for patients with EOC across a large health care network to identify barriers to testing. Methods: Clinical, pathologic, demographic and genomic testing (GT) information including involvement of a genetic counselor, specific test(s) ordered, test vendor, test turn-around time, and test results were obtained from the diverse dataset within the Providence St. Joseph Health (PSJH) Electronic Medical Records and the system-wide cancer registry data mart. PSJH is the third largest non-profit health care system in the US and treats roughly 43,000 cancer patients annually across a 7 state region. Patients with a diagnosis of EOC (ICD C56.x) who had at least a single in-person visit to a PSJH oncology department or a PSJH oncologist for EOC during the period between January 2015 and January 2020 were identified. GT data was manually abstracted, where structured data was un-available; data were analyzed in aggregate and to evaluate for trends over time in patterns of testing. Results: Within this EOC cohort (n=3,007), 1,1027 (34%) had GT results available in the EMR. Germline testing (GMT) was the initial testing approach in 728 (71%) of women tested, and 210 (29%) of the 728 women who had GMT first went on to have tumor tissue testing (TTT). Of the 300 patients who had TTT first, 79 (26%) went on to have GMT. A BRCA1/2 mutation was identified in 153 (14.9%) patients on GMT and/or TTT. Mutation results were discordant in 7 of 289 (2.4%) who had both GMT and TTT. GT rates increased over time but remained low (46% in 2019). Involvement of a genetic counselor (GC) increased uptake of GT, however, only 62% of patients completed recommended GC referral. GT was ordered from 17 different vendors (12 GMT; 11 TTT). Median time from initial diagnosis to GT order date decreased over time and was 8 weeks and 12 weeks in 2019 for GMT and TTT, respectively. The median time interval between GMT and TTT in patients who had both tests decreased from 130 weeks in 2015 to 6 weeks in 2019. Despite improvement in median time to testing, multiple outliers were observed. Conclusion: The uptake of GT for EOC patients has increased over time but remains low. There is substantial heterogeneity in testing approach including the timing, sequencing, and ordering of tests. Genetic counseling for patients with EOC increases uptake of testing. Significance: These findings highlight the challenges of developing a standardized testing approach across a diverse health care system. There is a need to develop a comprehensive network wide testing strategy that can be effective in multiple settings