When DNA testing is clinically justified
Consequently, constitutional DNA testing has been restricted by clinical features of the cancer, or family history associated with high risk of a hereditary causal mutation. But, as larger numbers have been tested, it is possible to use clinical and research data \cite{Lek_2016} together to show where the panels and criteria might be usefully extended. For instance, in one study, 5589 women with breast cancer (but not ovarian cancer) meeting genetic testing criteria for high risk breast and ovarian cancer (GC-HBOC) were tested for a panel of 8 cancer-causing genes \cite{Hauke_2018}. In this group, 17% were expected to carry BRCA1 or BRCA2 mutations. 5 other mutated cancer genes were detected in 312 cases, one gene was not associated with breast cancer and two genes require further investigation in a larger cohort. In a second study using a 17-gene panel, pathogenic mutations were found in 12% of 10,901 women, specifically those diagnosed with the high-risk triple negative subtype of breast cancer previously associated with BRCA1 mutation \cite{Shimelis_2018}. Here they indeed found mutations in BRCA1 but also BRCA2. Further, 3.7% of triple-negative cancer was associated with 3 other high-risk genes (odds ratio >5) and 3 genes conferring a moderate risk (odds ratio >2). A third study investigated the performance of a gene panel in all types of female breast cancer \cite{Couch_2017}. Among 41,611 women with breast cancer tested with a 21-gene panel usually used only for hereditary cancer, 4.02% carried BRCA1 or BRCA2 mutations, and 6.18% of the women carried mutations in 5 genes conferring high or moderate risk of breast cancer. 8 genes associated with other cancers were not found associated with increased breast cancer risk.