Development of an Open Database of Genes Included in Hereditary Cancer Genetic Testing Panels Available From Major Sources in the US

Hereditary cancer syndromes occur when germline variants increase an individual’s risk of developing cancer. Genetic testing can identify these variants, enabling clinicians to intervene through increased screening or prophylactic surgery. Advances in DNA sequencing have driven a shift to sequence-based testing of multiple genes and to subsequent classification of the identified variants. This approach has generated copious data, but a serious problem remains: New variants in cancer genes are discovered daily that cannot be classified (scientifically or clinically). This problem of variants of unknown significance (VUSs) is exacerbated in groups with predominantly non-European ancestry who have not been included in data sets used to interpret variants because of underrepresentation in genetic studies and diminished access to clinical genetic testing.¹ Fewer data lead to a higher proportion of VUSs.² Policies for data sharing and the related infrastructure are emerging to support variant classifications.³ To inform these efforts, we have compiled a list of genes included in hereditary cancer genetic testing panels. We also describe publicly available variant data for those genes. Our analysis identifies substantial variability across panels and gaps in available variant data that should be a focus for those engaged in building a robust system for interpreting inherited cancer risk.

This quality improvement study was reviewed and approved by the Arizona State University Institutional Review Board. Informed consent was not possible because this study uses publicly available gene variant data that are not attached to any participants.

We identified 17 major hereditary cancer testing companies and extracted data on their available hereditary cancer panels. We compiled a list of the genes included in at least 1 panel, and we extracted data on them from the ClinVar public database⁴ in October 2020. We also compiled information on genes included in clinical guidelines for hereditary cancer management and variants evaluated through the Clinical Genome Resource (ClinGen).⁵ We used Gephi software to display the overlap among companies and the genes offered in their hereditary cancer panels. In addition, we used Stata (StataCorp LLC) to perform descriptive statistical analyses. The complete data set is registered with the Open Science Framework.⁶

A total of 706 genes were included in at least 1 laboratory’s panel. Only 13 genes were included by all 17 companies. Only 110 genes appeared in at least 1 clinical guideline for hereditary cancer or had a ClinGen gene-disease relationship assessment. The Table provides summary statistics of variant-related characteristics available from ClinVar. The 699 genes with variants reported in ClinVar had a mean of 362 variants (median, 100 [range, 9-12 699]). The proportion of VUSs was a mean of 27.5% per gene, and 51.8% of variants per gene were provided by a single submitter. Genes included in clinical guidelines reported more variants, were included in more testing panels, had a higher proportion of variants identified during clinical testing, and had a higher proportion of VUSs. The Figure illustrates the considerable variability with which companies include genes in their panels, among a subset of laboratories submitting the most data to ClinVar.

Multigene, sequence-based genetic testing for an inherited risk of cancer has increased dramatically since 2012. Our analysis suggests that (1) there is a lack of standards for which genes are offered in hereditary cancer panels, with companies offering tests for genes with very little publicly available variant data to support clinical interpretation, and (2) clinical guidelines are also lacking. We also describe the importance of clinical testing as a source of variant data and thus the intertwined nature of equitable access to clinical testing and the return of informative results. One limitation of this study is its cross-sectional design, which restricted our analysis of very dynamic data to what were publicly available at one time point.

We offer this description of the US hereditary cancer panel landscape as an empirical basis for ongoing policy discussions about how to improve data sharing to reduce the proportion of VUSs. We also demonstrate how analytical software and visualization tools can support future empirical work. The data are openly available and we welcome others to use these data and this landscape.⁶

Accepted for Publication: November 19, 2021.

Published Online: February 3, 2022. doi:10.1001/jamaoncol.2021.7639

Open Access: This is an open access article distributed under the terms of the CC-BY License. © 2022 Geary J et al. JAMA Oncology.

Corresponding Author: Janis Geary, PhD, Barrett and O’Connor Washington Center, School for the Future of Innovation in Society, Consortium for Science, Policy, and Outcomes, Arizona State University, 1800 I (Eye) St NW, Washington, DC 20006 (jdgeary@asu.edu).

Author Contributions: Dr Geary had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.

Concept and design: Geary, Majumder, Cook-Deegan.

Acquisition, analysis, or interpretation of data: Geary, Guerrini, Cook-Deegan.

Drafting of the manuscript: Geary.

Critical revision of the manuscript for important intellectual content: All authors.

Statistical analysis: Geary.

Obtained funding: Cook-Deegan.

Administrative, technical, or material support: Geary, Majumder, Cook-Deegan.

Supervision: Cook-Deegan.

Conflict of Interest Disclosures: Dr Geary reported receiving grants from the National Cancer Institute and a fellowship from the Canadian Institutes of Health Research. Dr. Majumder reported receiving funding from the National Cancer Institute during the conduct of the study and other funding from the National Institutes of Health and National Human Genome Research Institute outside the submitted work. Prof Guerrini reported receiving grants from the National Cancer Institute during the conduct of the study. Dr Cook-Deegan reported serving as an expert witness in a genetic testing case (Williams v. Quest/Athena, 2016-2019) and received grant funding from the Wellcome Trust, Robert Wood Johnson Foundation, Chan Zuckerberg Initiative, and National Institutes of Health (National Human Genome Research Institute, National Cancer Institute, and National Institute of Mental Health) during the study period.

Funding/Support: This work was funded by grant R01CA237118 from the National Cancer Institute and grant 202012MFE-459170 from the Canadian Institutes of Health Research.

Role of the Funder/Sponsor: The funders had no role in any other feature of the work and publication, including the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication.

Additional Information: The complete data set described here is registered with the Open Science Framework (https://osf.io/qwyg2/) and includes the list of genes and variant-related characteristics along with instructions for using the network visualization tools.