Documents & Announcements

ClinGen Myeloid Malignancy Variant Curation Expert Panel recommendations for germline RUNX1 variants

Xi Luo, Simone Feurstein, Shruthi Mohan, Christopher C. Porter, Sarah A. Jackson, Sioban Keel, Michael Chicka, Anna L. Brown, Chimene Kesserwan, Anupriya Agarwal, Minjie Luo, Zejuan Li, Justyne E. Ross, Panagiotis Baliakas, Daniel Pineda-Alvarez, Courtney D. DiNardo, Alison A. Bertuch, Nikita Mehta, Tom Vulliamy, Ying Wang, Kim E. Nichols, Luca Malcovati, Michael F. Walsh, Lesley H. Rawlings, Shannon K. McWeeney, Jean Soulier, Anna Raimbault, Mark J. Routbort, Liying Zhang, Gabriella Ryan, Nancy A. Speck, Sharon E. Plon, David Wu, Lucy A. Godley.

Standardized variant curation is essential for clinical care recommendations for patients with inherited disorders. Clinical Genome Resource (ClinGen) variant curation expert panels are developing disease-associated gene specifications using the 2015 American College of Medical Genetics and Genomics (ACMG) and Association for Molecular Pathology (AMP) guidelines to reduce curation discrepancies. The ClinGen Myeloid Malignancy Variant Curation Expert Panel (MM-VCEP) was created collaboratively between the American Society of Hematology and ClinGen to perform gene- and disease-specific modifications for inherited myeloid malignancies. The MM-VCEP began optimizing ACMG/AMP rules for RUNX1 because many germline variants have been described in patients with familial platelet disorder with a predisposition to acute myeloid leukemia, characterized by thrombocytopenia, platelet functional/ultrastructural defects, and a predisposition to hematologic malignancies. The 28 ACMG/AMP codes were tailored for RUNX1 variants by modifying gene/disease specifications, incorporating strength adjustments of existing rules, or both. Key specifications included calculation of minor allele frequency thresholds, formulating a semi-quantitative approach to counting multiple independent variant occurrences, identifying functional domains and mutational hotspots, establishing functional assay thresholds, and characterizing phenotype-specific guidelines. Preliminary rules were tested by using a pilot set of 52 variants; among these, 50 were previously classified as benign/likely benign, pathogenic/likely pathogenic, variant of unknown significance (VUS), or conflicting interpretations (CONF) in ClinVar. The application of RUNX1-specific criteria resulted in a reduction in CONF and VUS variants by 33%, emphasizing the benefit of gene-specific criteria and sharing internal laboratory data.