Oral Presentation Joint 2016 COSA and ANZBCTG Annual Scientific Meeting

Genetic testing for germline variants in susceptibility genes for breast and other cancers frequently identifies variants of uncertain clinical significance, including missense, small in-frame insertion/deletion, splice and regulatory region variants. Unclassified variants are a major clinical challenge as they complicate test reporting and genetic counselling, and prevent guided clinical management of patients and their relatives.

The ENIGMA (Evidence-based Network for the Interpretation of Germline Mutant Alleles) international consortium undertakes research to improve methods for classifying variants in breast cancer predisposition genes.  To promote standardised classification, ENIGMA has established detailed criteria to classify germline variants in BRCA1 and BRCA2 using a 5 tier system that reflects probability of pathogenicity. ENIGMA is a ClinGen-designated expert panel for BRCA1/2 variant classifications, and ENIGMA classification criteria are currently being applied to variants identified by research and clinical testing sites internationally.

 ENIGMA expert panel and research activities to date have:

• provided evidence for existing robust classifications of “not-obviously-truncating” variants to ClinVar
• accrued clinical information for quantitative multifactorial analysis of additional non-truncating sequence variants, with altered classification anticipated for >390 unique variants
• shown that the qualitative criterion of 1% allele frequency used to classify a variant as “not pathogenic” is conservative and reliable
• provided evidence that spliceogenicity and pathogenicity are not equivalent, and demonstrated that expert review of (likely) spliceogenic variants, including those at acceptor/donor “consensus” dinucleotides, aids variant interpretation
• provided statistical estimates from large-scale studies to utilize breast tumour pathology data for predicting variant pathogenicity
• shown that use of multiple splicing bioinformatic tools does not improve prediction of spliceogenicity of exonic variants, and developed schema to identify exonic variants that may disrupt mRNA splicing

These findings demonstrate the value of large-scale international collaborations with gene- and disease-specific expertise to improve variant classification methods and processes, and deliver clinically meaningful standardised disease gene variant classification.