Integrated germline and somatic cancer testing provides opportunity to identify cancer risk and resolve variant origins.

Abstract

10589 Background: Germline and somatic genetic testing are established tools for the management of cancer patients. Somatic testing is primarily used to inform therapy and germline testing is used to diagnose hereditary cancer predisposition syndromes. While somatic testing can detect germline variants, the interpretation and reporting algorithms are optimized to predict therapeutic efficacy. As a result, germline variants may be missed or only interpreted in context of their potential to act as a therapeutic target. We retrospectively reviewed a series of patients who received both germline and somatic testing to examine the opportunities for concurrent germline testing to improve somatic reporting. Methods: Our study reviewed data from 43 patients with solid cancer diagnoses who were otherwise unselected and underwent testing with a 435-gene somatic genetic test and an 83-gene germline test. The most frequent cancers were pancreatic (18), ovarian (8), and prostate (7). Results: Out of the 43 patients, 7 (16%) harbored a pathogenic or likely pathogenic germline variant (PGV) in a cancer susceptibility gene. PGVs were identified in MLH1, MSH6, CHEK2, PALB2, CDKN2A, NBN, and MUTYH. Notably, 3 of these genes ( CHEK2, PALB2, MUTYH) were not considered therapeutic targets, and therefore were only included as ancillary findings near the end of the preliminary somatic test reports (generated prior to integration of germline test results). In addition, 40 of 43 (93%) patients had at least one variant detected by somatic testing in at least one of the germline panel genes (mean number variant genes = 4.1, maximum = 10); all of these variants were within the reportable range of the germline assay, and therefore germline test results were able to resolve their germline versus somatic origins. The genes that most frequently had somatic variants identified were TP53 (79% of patients), CDKN2A (37%), SMAD4 (30%), and FLCN (21%). Conclusions: Due to the size of commonly ordered somatic gene panels, there is a high probability of detecting variants in hereditary cancer predisposition genes (> 90% of patients in this study) that can provide either therapy or cancer risk information or both. Given that a significant proportion (16% in this study) of cancer patients harbor PGVs (which can further inform treatment, disease surveillance, preventive measures, and risk assessment for family members), it is crucial to resolve the somatic versus germline origin of these variants. Since interpretation and reporting algorithms for somatic testing are optimized for therapy prediction, and variables such as specimen tumor purity, tumor ploidy, and variant allele fraction render estimates of variant origin unreliable for diagnostic purposes, it is important to take advantage of germline testing concurrently in patients receiving somatic testing to glean this critical information.