Tumor sequencing with germline genetic testing: Identification of patients with hereditary cancer and precision treatment eligibility.

person Scott T. Michalski Invitae, San Francisco, CA info_outline Scott T. Michalski, Daniel Esteban Pineda Alvarez, Meaghan Russell, Shan Yang, Guru Sonpavde, Edward D. Esplin

Authors person Scott T. Michalski Invitae, San Francisco, CA info_outline Scott T. Michalski, Daniel Esteban Pineda Alvarez, Meaghan Russell, Shan Yang, Guru Sonpavde, Edward D. Esplin Organizations Invitae, San Francisco, CA, Dana-Farber Cancer Institute, Boston, MA Abstract Disclosures Research Funding Other Background: Cancer is a fundamentally genetic disease, as such, somatic and germline mutation analysis is used in the comprehensive assessment of patients with cancer. Studies report that approximately 10% of patient’s tumors have clinically significant variants known to predispose to hereditary cancer, with medical implications for both patients and family members. We retrospectively reviewed a series of patients where providers suspected a somatic variant also existed in the germline and followed up with clinical germline genetic testing. We report the rate of concordance between germline and somatic results and their clinical impact. Methods: Our study used de-identified data from 1043 consecutive patients who underwent somatic genetic testing followed by germline testing with NGS-based hereditary cancer gene panels. Results: Somatic results most frequently prompting germline testing included variants in BRCA2 (290), BRCA1 (174), TP53 (158), ATM (70), MLH1 (65), APC (65), PMS2 (61), MSH6 (58), PTEN (54) and CDH1 (42). In 364/1043 cases (35%) the variant was detected as likely pathogenic or pathogenic (LP/P) in the germline. Genes confirmed as germline variants in 60-100% of cases included: FANCA, AXIN2, RAD50, MUTYH, BLM, PALB2, CHEK2, FANCD2, MITF, SDHB. Variants in: FH, BRCA2, RET, ATM, SDHA, BRIP1, MSH2, BRCA1, BAP1, EGFR and RAD51D confirmed in the germline in 25-60%. Variants were rarely germline for TP53 (3%), APC (3%), PTEN (2%) and none in CDKN2A, NF1 and STK11. In 24 (2%) cases a LP/P germline variant was detected but not reported in the tumor. Conclusions: Approximately ⅓ of patients suspected to have hereditary risk after tumor testing had LP/P germline variants. Notably, some genes had a high probability of variants occurring in the germline, while others were primarily seen in tumors. Interestingly, 6% of the germline variants were not included on the somatic report due to technical and gene content differences in either assays or due to differences of clinical classification between somatic and germline testing. Adding germline results to somatic testing may inform options for precision treatment, prevention, or early detection of, secondary malignancies and guide genetic counseling of family members.