348 posters,  11 topics,  1808 authors,  584 institutions

ePostersLive® by SciGen® Technologies S.A. All rights reserved.

Clinical laboratories implement the ACMG/AMP guidelines to resolve differences in variant interpretations submitted to ClinVar

Primary tabs


Average: 5 (1 vote)


2034 reads

Clinical Laboratories Implement the ACMG/AMP Guidelines to Resolve Differences in Variant Interpretations Submitted to ClinVar

Steven M. Harrison1,2, Jill S. Dolinsky3, Lisa Vincent4, Amy Knight Johnson5, Danielle R. Azzariti1, Tina Pesaran3,Elizabeth C. Chao6, Soma Das5,  Sherri Bale4, and Heidi L. Rehm1,2   

 1Partners HealthCare Personalized Medicine and 2Brigham & Women’s Hospital and Harvard Medical School, Boston, MA; 3Ambry Genetics, Aliso Viejo, CA; 4GeneDx, Gaithersburg, MD; 5Human Genetics, University of Chicago, Chicago, IL; 6University of California, Irvine, Irvine, CA

Background and Methods

 •The ClinVar database allows labs to share variant interpretations that previously had been unpublished or unavailable to the larger community.

 •ACMG published guidelines for variant interpretation provide a framework to classify variants; however, given the complexity of variant interpretation, application of the guidelines still require subjective interpretation

 •Through a ClinGen initiative, 4 clinical labs, Ambry Genetics, GeneDx,  Laboratory for Molecular Medicine (LMM), and University of Chicago, worked to resolve variants with interpretation differences in ClinVar with the following process:

•Compare labs’ interpretation previously submitted to ClinVar to labs’ most recent interpretation
•Reassess variants using the ACMG/AMP guidelines
•Share evidence & internal data used, when applicable
•Identify persistent interpretation differences due to varying application of ACMG/AMP rules
•Update variant classifications in ClinVar as needed

ClinVar and Interpretation Data

•49,734 unique variants have been submitted to ClinVar by ≥1 of 4 participating labs (1/1/2016)
•6169 variants were submitted by ≥2 participating labs allowing interpretation comparisons
•724 (12%) had one or two-step differences between: Pathogenic/Likely pathogenic (P/LP), Uncertain significance (VUS), and Likely benign/Benign (LB/B) 

Resolution Results

P/LP vs VUS/LB/B Reassessments (104 variants)

•Majority (57%) resolved as VUS; 21% resolved as P or LP
•Consensus not reached for 23 P/LP vs VUS/LB/B differences (22%):
•For 8 variants, labs applied the same pathogenic criteria but differed on the application of benign criteria, specifically deciding how to account for observations of the variant in controls (BS2) or data suggesting no functional impact (BS3), despite other evidence for a pathogenic interpretation
•For 15 variants, labs differently applied pathogenic criteria included functional studies (PS3), reputable source (PP5), and hotspots/functional domains (PM1)

VUS vs LB/B Reassessments (128 variants)

•Majority (76%) resolved as B or LB; 16% resolved as VUS
•Consensus was not reached for 10 VUS vs LB/B differences (8%), mostly due to differences in how labs applied benign criteria for observation in controls (BS2) and functional studies suggesting no impact (BS3)


•Labs reached concordance on 86% of the 232 reassessed variants
•Figure 2 shows the breakdown of reasons for discordant interpretations (87 variants)
•Sharing internal evidence, such as segregations, co-occurrences, and de novo observations, facilitated resolution of 23 interpretation differences


•78% of clinically actionable differences & 92% VUS vs. LB/B differences were resolvable after consensus efforts were applied; 492 differences have not yet been assessed
•Interpretations in ClinVar do not always represent a lab’s current interpretation of a variant; more frequent submissions to ClinVar are needed
•Further specification regarding functional assays and weighting of conflicting pathogenic and benign criteria may further facilitate resolution of interpretation differences with ACMG/AMP guidelines
•Application of the ACMG/AMP criteria and sharing internal evidence and classification rationales increased the overall concordance rate between these four labs from 88% to 92%


Enter Poster ID (e.gGoNextPreviousCurrent