Is 'likely pathogenic' really 90% likely? Reclassification data in ClinVar

Steven M Harrison; Heidi L Rehm

doi:10.1186/s13073-019-0688-9

Is 'likely pathogenic' really 90% likely? Reclassification data in ClinVar

Genome Med. 2019 Nov 21;11(1):72. doi: 10.1186/s13073-019-0688-9.

Authors

Steven M Harrison^{1

2}, Heidi L Rehm^{3

4

5}

Affiliations

¹ Medical & Population Genetics Program and Genomics Platform, Broad Institute of MIT and Harvard, Main Street, Cambridge, MA, 02142, USA. steven@broadinstitute.org.
² Department of Pathology, Harvard Medical School, Shattuck Street, Boston, MA, 02115, USA. steven@broadinstitute.org.
³ Medical & Population Genetics Program and Genomics Platform, Broad Institute of MIT and Harvard, Main Street, Cambridge, MA, 02142, USA.
⁴ Department of Pathology, Harvard Medical School, Shattuck Street, Boston, MA, 02115, USA.
⁵ Center for Genomic Medicine, Massachusetts General Hospital, Fruit Street, Boston, MA, 02114, USA.

Abstract

In 2015, professional guidelines defined the term 'likely pathogenic' to mean with a 90% chance of pathogenicity. To determine whether current practice reflects this definition, ClinVar classifications were tracked from 2016 to 2019. During that period, between 83.8 and 99.1% of likely pathogenic classifications were reclassified as pathogenic, depending on whether LP to VUS reclassifications are included and on how these classifications are categorized.

Publication types

Research Support, N.I.H., Extramural

MeSH terms

Genetic Association Studies / methods*
Genetic Predisposition to Disease*
Genetic Testing* / methods
Humans
Practice Guidelines as Topic

Abstract

Publication types

MeSH terms

Grants and funding