Review Article

Beyond DNA: An Integrated and Functional Approach for Classifying Germline Variants in Breast Cancer Genes

Table 1

Classification scheme for high penetrance autosomal dominant breast cancer genes.

ClassClassificationCategoryCriteria

5PathogenicA
1 needed
(stand-alone)
(i) Alterations resulting in premature truncation (e.g., reading frame shift, nonsense)
(ii) Other ACMG-defined mutations (i.e., initiation codon or gross deletion)
(iii) Strong segregation with disease (LOD >3 = >10 meioses)

5PathogenicB
4 needed
(strong)
(i) Confirmed de novo alteration in the setting of a new disease (appropriate phenotype) in the family
(ii) Significant disease association in appropriately sized case-control study(ies)
(iii) Being detected in individuals satisfying established diagnostic criteria for classic disease without a clear mutation
(iv) Last nucleotide of exon
(v) Good segregation with disease (LOD 1.5–3 = 5–9 meioses)
(vi) Deficient protein function in appropriate functional assay(s)
(vii) Functionally validated splicing mutation
(viii) Well-characterized mutation at the same position
(ix) Other strong data supporting pathogenic classification (e.g., structural)

4Likely pathogenic1 needed(i) Alterations at the canonical donor/acceptor sites (± 1, 2) without another strong (B-level) evidence supporting pathogenicity

4Likely pathogenicC
4 needed
(supportive)
(i) Rarity in general population databases (dbSNP, ESP, 1000 Genomes, ExAC)
(ii) In silico models in agreement (deleterious) and/or completely conserved position in appropriate species
(iii) Moderate segregation with disease (at least 3 informative meioses) for rare diseases
(iv) Other data supporting pathogenic classification (e.g., structural)
3 of B
2 of B and at least 1 of C
1 of B and at least 3 of C

3VUS Insufficient or conflicting evidence
Gross duplications without strong evidence for pathogenic or benign

3Likely benignD
1 needed
(strong)
(i) Intronic alteration with no splicing impact by RT-PCR analysis or another splicing assay 
(ii) Other strong data supporting benign classification

3Likely benignE
2 needed
(supportive)
(i) Cooccurrences with mutations in the same gene (phase unknown)
(ii) Cooccurrences with mutations in other high penetrant genes that clearly explain a proband’s phenotype
(iii) Subpopulation frequency in support of benign classification
(iv) Intact protein function observed in appropriate functional assay(s)
(v) In silico models in agreement (benign)
(vi) Not segregating with disease in family study (genes with incomplete penetrance)
(vii) No disease association in small case-control study
(viii) Other data supporting benign classification

1BenignF
1 needed
(stand-alone)
(i) General population or subpopulation frequency is too high to be a pathogenic mutation based on disease/syndrome prevalence and penetrance
(ii) Not segregating with disease in family study (genes with complete penetrance)
(iii) Internal frequency is too high to be a pathogenic mutation based on disease/syndrome prevalence and penetrance
(iv) Being seen in trans with a mutation or in homozygous state in individuals without severe disease for that gene
(v) No disease association in appropriately sized case-control study(ies)
1 of D and at least 2 of E
2 or more of D
>3 of E w/o conflicting data
>4 of E w/conflicting data

The variant classification scheme is not intended for the interpretation of alterations complicated by epigenetic factors including genetic modifiers, multifactorial disease, or low-risk disease association alleles and may be limited in the interpretation of alterations confounded by incomplete penetrance, variable expressivity, phenocopies, and triallelic or oligogenic inheritance.