Statewide Retrospective Review of Familial Pancreatic Cancer in Delaware, and Frequency of Genetic Mutations in Pancreatic Cancer Kindreds

We conducted a retrospective review of pedigrees of families with a pancreatic adenocarcinoma cancer diagnosis held in the statewide Ruth Ann Minner High Risk Family Cancer Registry at the Helen F. Graham Cancer Center and Research Institute, Christiana Care Health System, Newark, DE, USA, from 2002 to 2013. The registry was queried based on how many first-, second-, or third-degree relatives of the proband were affected with PC, genetic testing status, and (if applicable) the results. These data were then categorized into families that meet familial PC (FPC) criteria, defined as two first-degree relatives with PC (FPC families), families that did not meet the FPC definition but had one first-degree relative affected with PC (first-degree families), and probands with PC (probands). Each family was counted only once in the analysis, even if multiple family members were tested.

Our analysis revealed that 175 of 597 families fitting any of the above criteria completed genetic testing. Of this cohort, 52 had pathogenic alterations with nine different genes implicated. Overall, 164 of the 175 families that fitted into any of the three categories previously identified had BRCA1 or BRCA2 testing, either by DNA sequencing or next-generation sequencing via a panel test that included BRCA1/2. BRCA1 pathogenic alterations were noted in 17/164 (10.4 %) and BRCA2 pathogenic alterations were noted in 23/164 (14.0 %). FPC families (n = 46) 42/46 of the FPC families underwent BRCA1/2 testing, and 11/42 (26 % [95 % CI 12.89–39.49]) had pathogenic alterations. Specifically, 4/42 = BRCA1 (9.5 %) and 7/42 = BRCA2 (16.7 %). Additionally, 16/46 of the FPC families underwent exclusively Lynch syndrome (LS) testing, and pathogenic mutations in a mismatch repair protein were identified in 2/16. Specifically, 1/16 = MLH1 (6.3 %) and 1/16 = MSH2 (3.6 %). Overall, a genetic mutation within any gene associated with an increased PC risk was found in 28 % of FPC families. First-degree families (n = 106) 99/106 of the families with one first-degree relative underwent BRCA1/2 testing, and 21/99 (21.2 % [95 % CI 13.16–29.27]) had pathogenic alterations. Specifically, 11/99 = BRCA1 (11.1 %) and 10/99 = BRCA2 (10.1 %). 32/99 first-degree families underwent exclusively LS testing, and pathogenic mutations were identified in 4/32. Specifically, 3/32 = MLH1 (9 %) and 1/32 = MSH6 (3 %). 25/99 of the families pursued panel testing, and pathogenic alterations in any gene were identified in 3/25. Specifically, the mutations were found in 1/25 = ATM (4 %), 1/25 = CHEK2 (4 %), and 1/25 = RAD51D (4 %). Affected probands (n = 23) Lastly, all 23 probands affected with PC pursued genetic testing. Of these, 11/23 were found to have pathogenic alterations. All 23 underwent BRCA1/2 testing, and pathogenic alterations were identified in 8/23 (35 % [95 % CI 15.32–54.25]), specifically 2/23 = BRCA1 (9 %), and 6/23 = BRCA2 (26 %). 10/23 patients underwent panel testing and pathogenic alterations were found in 3/10 (30 %) patients, of whom 1/10 = MSH6 (10 %), 1/10 = ATM (10 %), and 1/10 = TP53 (10 %).

This study demonstrates that a statewide high-risk family cancer registry is an important instrument in studying the risk of PC in families. Our analysis revealed 14 mutations associated with FPC, among which hereditary breast and ovarian cancer and LS were most prevalent. BRCA1 was found to have the same association with PC as BRCA2, which appears unique to our population. We plan to use our knowledge of these mutations in developing a PC screening program.