Germline genetic testing among patients with pancreatic cancer (PC): A Pancreatic Cancer Action Network (PanCAN) patient survey.

10594 Background: Universal germline genetic testing for PC is endorsed by multiple professional organizations like the National Comprehensive Cancer Network (NCCN), yet implementation remains limited. We sought to investigate practices related to germline testing in PC and identify potential causes of under-testing through a large patient survey. Methods: We used the PanCAN patient registry to administer this HIPAA-compliant survey electronically to patients and caregivers. Logistic regression was used to determine the association between patient-reported demographic and clinical characteristics and the odds of undergoing germline genetic testing. A multivariate model was built to include all characteristics that were significant on univariate analysis. The association between receipt of genetic counseling and cascade germline genetic testing among first degree relatives (FDR) of patients with germline mutations was assessed via a chi-square test. Results: A total of 1,046 patients with PC were included, of which, 724 (69.2%) reported undergoing germline genetic testing. On multivariate analysis, race (p = 0.01), insurance type (p < 0.01), stage (p < 0.01), and treatment facility type (p < 0.01) were significantly associated with the odds of undergoing germline genetic testing after PC diagnosis. Black patients (compared to White) [OR = 0.42 (95% CI 0.23-0.78)], uninsured patients (compared to Medicare) [OR = 0.11 (95% CI 0.03-0.44)], patients insured through Veterans Health Administration (compared to Medicare) [OR = 0.26 (95% CI 0.08-0.92)] and patients receiving care at large [OR = 0.62 (95% CI 0.46-0.84)] and small community practices (compared to academic/teaching hospitals) [OR = 0.51 (95% CI 0.29-0.88)] were at significantly decreased odds of undergoing germline genetic testing. Patients with stage IV (compared to stage I) [OR = 2.23 (95% CI 1.46-3.40), stage II/III PC (compared to stage I) [OR = 1.58 (95% CI 1.11-2.23)] and those with private insurance (compared to Medicare) [OR = 1.51 (95% CI 1.11-2.05)] were at significantly increased odds of undergoing germline genetic testing. Of 724, 167 (23.2%) patients reported testing positive for germline mutations. Only 103 of 167 (61.7%) of these patients reported cascade testing among FDR. Patients undergoing genetic counselling had higher rates of cascade germline genetic testing among FDR (67.7% v 38.2%, p < 0.01). Conclusions: This is one of the largest patient-reported surveys from recent times suggesting sub-optimal implementation of guideline-based germline genetic testing in PC. This study highlights key inequities in testing based on race, insurance and practice setting. Our findings also underscore the critical role of genetic counselling in facilitating cascade testing among FDR of patients with PC and germline genetic mutations.