Evaluation of Next-Generation Sequencing as a clinical and research modality in the diagnosis of hereditary breast cancer
Dougherty, Kristen Elizabeth
MetadataShow full item record
Next-Generation Sequencing has opened the doors to nearly limitless amounts of genomic data, but the clinical utility of this data is not yet clear. From examining at sequencing data of known familial cancer genes in hereditary cancer patients, the NCGENES study found a clear molecular diagnosis in about 5% of patients and an uncertain molecular result in about 15% of patients. The remaining 80% of hereditary cancer patients received a negative result for the screening of known cancer genes. These latter patients were followed up by whole exome sequencing analysis, and the data was used to perform a research sweep to potentially identify mutation(s) in gene(s) that have yet to be clearly associated with their phenotype. Hereditary breast cancer has a relatively well-established set of susceptibility genes, yet a large percentage of the molecular etiology is still unknown. There are many genes that are good candidates for breast cancer genes based on their protein's function, but they may not actually contribute to breast cancer susceptibility. The ClinGen consortium is aiming to establish the clinical validity of gene-disease associations so that clinicians and patients can better interpret and utilize sequencing results. Six breast cancer susceptibility genes were evaluated using the ClinGen clinical validity framework with the goal of both evaluating the genes already on hereditary breast cancer panels and evaluating genes not yet widely tested to determine if there is enough evidence to support their role in disease to warrant widespread testing. These genes have varying levels of evidence supporting their role in breast cancer susceptibility. The variants in each of the six genes were compared between a cancer patient cohort and a non-cancer patient cohort enrolled in the NCGENES whole exome sequencing study. One likely pathogenic variant and several variants of unknown significance were identified in various genes, and the burden of variants in cancer cases versus controls was evaluated, although the controls were not matched to the cancer cohort in any way. Research sweeps were performed for patients with VUSs to ensure that there were no other mutations in genes that would better fit the phenotype. This thesis presents a method for evaluating gene-disease associations and for utilizing whole exome sequencing data to pinpoint a molecular diagnosis in hereditary breast cancer patients. Overall, it was found that the ClinGen method of evaluating clinical validity of gene-disease associations could be helpful when determining if variants are pathogenic or benign. A new gene, RINT1, was found to have enough evidence to be moderately associated with hereditary breast cancer and it was subsequently added to the diagnostic list so that all cancer patients will now be screened for RINT1 variants. In addition, it was found that two of the genes currently on the diagnostic list, RAD51C and RAD51D, have "disputed" evidence with respect to breast cancer susceptibility. Interestingly, they have much more evidence for an association with ovarian cancer, so if variants are found in these genes, the patient's phenotype should be considered when evaluating them. It was also shown that PALB2, an established breast cancer susceptibility gene, indeed is definitively associated with breast cancer, and the NCGENES cancer patients have more truncating variants than the controls, further validating the clinical validity assertion. Finally, an ovarian cancer patient with two interesting variants, one in SLX4 and one in GEN1, were evaluated. Studies showed that knocking out both of these genes' pathways was highly destructive to the cell. A VUS was found in each of these genes, and it was hypothesized that perhaps these two variants together may be sufficient to contribute to this patient's cancer susceptibility.