Specification of frequency criteria for secondary findings genes to improve variant classification concordance

ACMG 2026

Mar 10, 2026
Presentation

Jennifer Johnston, Kristy Lee, Deborah Ritter, Steven Harrison, Leslie Biesecker

Introduction
The return of secondary findings (SF) is well-established in clinical testing and is frequently employed in clinical research. The American College of Medical Genetics and Genomics (ACMG) maintains a list of monogenic disease entities (MDEs, defined as a gene-phenotype pair) for SF variant return, which is annually reviewed and updated. In 2015 the ACMG and the Association for Molecular Pathology (AMP) published criteria for classifying sequence variant pathogenicity. The Clinical Genome Resource (ClinGen), in support of maintaining and streamlining the use of ACMG SF, is providing SF-specific guidance. As part of this effort, we set out to calculate maximum credible population disease allele frequency thresholds (DAFT), and specify values for BA1 (the stand-alone benign evidence code, associated with variants too common to be pathogenic) and BS1 (the strong-benign evidence code, associated with variants likely too common to be pathogenic) for SF MDEs that do not yet have specified classification criteria from a ClinGen Variant Curation Expert Panel (VCEP). This work covers 58 MDEs (involving 47 genes) suggested for SF return (v3.3).

 

Methods
We calculated DAFTs for SF MDEs using values for disease prevalence, gene contribution, and penetrance identified from the following resources: GeneReviews, the ClinGen Actionability Working Groups, and published literature. ACMG criterion BS1 was set at the calculated DAFT, with BA1 set at ten times the calculated DAFT. For genes associated with multiple SF MDEs without a clear genotype-phenotype correlation, DAFT values were combined. gnomAD Grpmax filtering allele frequencies (FAF) for variants classified as pathogenic/likely pathogenic in ClinVar and novel pLOF variants were compared to calculated BA1 and BS1 thresholds.

 

Results
DAFT/BA1/BS1 values were calculated for 58 SF MDEs (47 genes). For nine genes included in this study, multiple MDEs are recommended for SF return. For eight of these genes without a clear genotype/phenotype correlation, the DAFT values were combined across reportable MDEs. Variants in SCN5A are recommended for return for three distinct MDEs, loss of function variants are associated with Brugada syndrome, gain of function variants are associated with Long QT Syndrome (LQTS), and no clear genotype/phenotype correlation exists for variants associated with Dilated Cardiomyopathy (DCM). For SCN5A, the DAFTs for Brugada and LQTS were determined and the greater of the two was combined with the DAFT for DCM to create a final DAFT for SCN5A. For the 47 genes considered here, comparison of Grpmax FAF for ClinVar P/LP and novel pLoF variants to the calculated values for BA1/BS1 did not identify any P/LP or pLoF variant (in genes where loss of function is a mechanism for the SF MDE) with a Grpmax FAF greater than the calculated BA1 value for the relevant gene. With regards to BS1 thresholds, a total of nine P/LP or pLoF variants in eight genes had a Grpmax FAF greater than the calculated BS1 value. These nine variants represent 0.2% of a total of 4,305 P/LP or pLoF variants.

 

Conclusion
ClinGen VCEPs combine the expertise of geneticists, physician specialists, laboratory geneticists, genetic counselors, and basic scientists to consider the genetics, biology, and the molecular diagnostic approach of the MDE to derive variant classification recommendations, specifically, adaptations of ACMG/AMP 2015 recommendations. In the absence of VCEP specified criteria, few laboratories have the expertise or resources to set MDE-specific criteria for all SF genes. We provide calculated SF MDE BA1/BS1 values with supporting prevalence, penetrance, and gene contribution information as a resource for the community and plan to make them available on the ClinGen website. Importantly, VCEP specifications, when available, will supersede this work. While full VCEP specifications are ideal, a set of available and consistent thresholds from available published literature should improve consistency of SF MDE classifications.