Using Functional Data for VUS Reclassification in Cancer Predisposition Genes
Laboratory Genetics and Genomics
-
Primary Categories:
- Laboratory Genetics
-
Secondary Categories:
- Laboratory Genetics
Introduction:
The classification of clinical gene variants relies on various evidence sources, such as population frequency, case-control studies, and familial segregation studies. However, for many variants, these types of data are either unavailable or provide limited information. This leads to the classification of many variants as variants of uncertain significance (VUS) which cannot be utilized for clinical decision-making. Variant-level functional data from Multiplexed Assays of Variant Effect (MAVEs) have the potential to resolve VUS classification by assessing all possible single nucleotide variants in a single experiment. These assays provide functional data which may be applied to variant interpretation using the BS3/PS3 ACMG/AMP criterion. In this study, we evaluated the ability of utilizing MAVE data for reclassification of VUS in four cancer predisposition genes within a cohort of individuals seen in adult genetics clinics at the University of Washington and Fred Hutch Cancer Center.
Methods:
Using the Brotman Baty Institute Clinical Variant Database (BBI-CVD), we identified 37 unique missense VUS in BRCA1 (16), BRCA2 (4), MSH2 (11), and TP53 (6). These VUS were noted in individuals seen for clinical testing due to personal and/or family history of cancer that warranted genetic testing. VUS reclassification was done according to recommendations put forth by the ClinGen Variant Curation Expert Panel for each gene. Multiple sources of validated functional data have been approved by the VCEPs for all four genes. We utilized the recommended MAVE datasets with the appropriate strength of evidence for each gene. Conflicting information across assays, or functional scores falling in an intermediate range, or results not predictive of protein function were not used for VUS reclassification in BRCA1, BRCA2, and TP53 as recommended by the VCEPs. For MSH2, no guidance is available on using conflicting assay results. Hence, we used data from a single MAVE (Jia et al., 2021) with the recommended evidence strength, as this MAVE has the strongest strength of evidence for both pathogenic and benign variants (PS3_strong, BS3_strong).
Results:
In our cohort, functional data was available for 29/37 variants (78%). In 7 cases the data showed conflicting effects across multiple assays and was not informative for variant reclassification. Of the 22 VUS with informative functional data, 50% were reclassified based on this data: 9 to likely-benign/benign (LB/B), 2 to likely-pathogenic (LP), with 11 remaining VUS. For the 11 variants that remained uncertain, functional data added valuable evidence such that only one additional supporting/moderate criterion would be required for reclassification of these variants to likely benign or likely pathogenic. In addition to the variants that were reclassified based on functional data, 10 variants were reclassified during the process of curation using other data types and updated VCEP guidelines. Therefore, in total, we reclassified 21/37 (57%) VUS in BRCA1 (13), BRCA2 (1), MSH2 (5), and TP53 (3).
Conclusion:
In this study we evaluated the value of MAVE data for reclassifying VUS in four cancer predisposition genes. Informative functional data was available for 60% of variants, and half of those were reclassified based on this data. Our findings demonstrate that functional data can be readily used for reclassifying VUSs in four cancer predisposition genes within a cohort of individuals seen in adult genetics clinics, enabling more precise and clinically actionable genetic interpretations. The study underscores the importance of MAVEs in interpreting VUSs and suggests the need for further research to expand this approach to other clinically relevant genes in larger cohorts. Future efforts should focus on refining guidelines for integrating conflicting MAVE results, and exploring point-based frameworks to further enhance VUS reclassification for cases with discordant information.
The classification of clinical gene variants relies on various evidence sources, such as population frequency, case-control studies, and familial segregation studies. However, for many variants, these types of data are either unavailable or provide limited information. This leads to the classification of many variants as variants of uncertain significance (VUS) which cannot be utilized for clinical decision-making. Variant-level functional data from Multiplexed Assays of Variant Effect (MAVEs) have the potential to resolve VUS classification by assessing all possible single nucleotide variants in a single experiment. These assays provide functional data which may be applied to variant interpretation using the BS3/PS3 ACMG/AMP criterion. In this study, we evaluated the ability of utilizing MAVE data for reclassification of VUS in four cancer predisposition genes within a cohort of individuals seen in adult genetics clinics at the University of Washington and Fred Hutch Cancer Center.
Methods:
Using the Brotman Baty Institute Clinical Variant Database (BBI-CVD), we identified 37 unique missense VUS in BRCA1 (16), BRCA2 (4), MSH2 (11), and TP53 (6). These VUS were noted in individuals seen for clinical testing due to personal and/or family history of cancer that warranted genetic testing. VUS reclassification was done according to recommendations put forth by the ClinGen Variant Curation Expert Panel for each gene. Multiple sources of validated functional data have been approved by the VCEPs for all four genes. We utilized the recommended MAVE datasets with the appropriate strength of evidence for each gene. Conflicting information across assays, or functional scores falling in an intermediate range, or results not predictive of protein function were not used for VUS reclassification in BRCA1, BRCA2, and TP53 as recommended by the VCEPs. For MSH2, no guidance is available on using conflicting assay results. Hence, we used data from a single MAVE (Jia et al., 2021) with the recommended evidence strength, as this MAVE has the strongest strength of evidence for both pathogenic and benign variants (PS3_strong, BS3_strong).
Results:
In our cohort, functional data was available for 29/37 variants (78%). In 7 cases the data showed conflicting effects across multiple assays and was not informative for variant reclassification. Of the 22 VUS with informative functional data, 50% were reclassified based on this data: 9 to likely-benign/benign (LB/B), 2 to likely-pathogenic (LP), with 11 remaining VUS. For the 11 variants that remained uncertain, functional data added valuable evidence such that only one additional supporting/moderate criterion would be required for reclassification of these variants to likely benign or likely pathogenic. In addition to the variants that were reclassified based on functional data, 10 variants were reclassified during the process of curation using other data types and updated VCEP guidelines. Therefore, in total, we reclassified 21/37 (57%) VUS in BRCA1 (13), BRCA2 (1), MSH2 (5), and TP53 (3).
Conclusion:
In this study we evaluated the value of MAVE data for reclassifying VUS in four cancer predisposition genes. Informative functional data was available for 60% of variants, and half of those were reclassified based on this data. Our findings demonstrate that functional data can be readily used for reclassifying VUSs in four cancer predisposition genes within a cohort of individuals seen in adult genetics clinics, enabling more precise and clinically actionable genetic interpretations. The study underscores the importance of MAVEs in interpreting VUSs and suggests the need for further research to expand this approach to other clinically relevant genes in larger cohorts. Future efforts should focus on refining guidelines for integrating conflicting MAVE results, and exploring point-based frameworks to further enhance VUS reclassification for cases with discordant information.