Invitae announced new studies to be presented at the American Society of Breast Surgeons Annual Meeting (ASBrS) held in Orlando from April 10-14, 2024. The featured research will highlight how machine learning can reduce variants of uncertain significance (VUS) in patients who have received genetic testing for breast cancer, in addition to results from real world data showing that uncertain results do not lead to an overuse of mastectomies for breast cancer patients. New findings show machine learning models can help reduce the burden on clinicians and patients in the process of resolving uncertain genetic test results for hereditary breast cancer, particularly for underrepresented populations.

Due to limited evidence, classifying uncertain variants can be challenging. Individuals from certain racial, ethnic and ancestral (REA) populations are often underrepresented in large genomic databases and medical literature, leading to less definitive results, resulting in VUS which are at the center of healthcare disparities. To generate more definitive genetic testing results for these groups, the study sought to develop gene-specific machine learning models and evaluate their usefulness in genetic testing for patients with breast cancer.

After validating gene-specific machine learning algorithms, the impact on variant classification was analyzed in patients with breast cancer who underwent hereditary cancer genetic testing from January 2022 - May 2023. The study found that among breast cancer patients who had at least one variant with machine learning evidence applied, more than 15% resulted in a definitive (non-VUS) variant classification. However, when stratifying the analysis by REA populations, a higher percentage of Black (22%), Asian (28%), and Hispanic (19%) individuals with breast cancer had a definitive classification that was determined by machine learning evidence compared to White (12%) individuals, suggesting that machine learning evidence is particularly useful for patients from historically underrepresented populations with breast cancer who undergo hereditary cancer testing.

By reclassifying these VUS into definitive results, laboratories can help clinicians better manage the care of their patients and help reduce gaps in healthcare disparities across racial, ethnic, and socio-economic groups. Results offer reassurance that variants of uncertain significance in genetic testing results among breast cancer patients do not lead to overuse of breast surgeries, like mastectomies. Genetic testing for hereditary disease, which looks at genes that can directly inform surgical, medical and surveillance strategies for breast cancer risk-reduction and treatment, has become the standard of care for assessing hereditary cancer predisposition.

Although this testing can lead to high rates of uncertain results, it has been unclear if these results impact clinical management. Conflicting results have been reported regarding whether patients with VUS are more likely to undergo breast surgeries, compared to those with negative results. This study titled, "Real-World Breast Surgery Utilization among Breast Cancer Patients with Germline Variants of Uncertain Significance," examines breast surgery uptake in a large, recent sample of breast cancer patients undergoing genetic testing.

The study found, through a sample of more than 9,000 breast cancer patients, those with VUS results did not undergo surgical procedures more frequently than patients with negative results. These results offer reassurance that VUS results do not lead to overuse of mastectomies for breast cancer patients.