Clinical laboratories collaborate to resolve differences in variant interpretations submitted to ClinVar

Abstract

Purpose Data sharing through ClinVar offers a unique opportunity to identify interpretation differences between laboratories. As part of a ClinGen initiative, four clinical laboratories (Ambry, GeneDx, Partners Healthcare Laboratory for Molecular Medicine, and University of Chicago Genetic Services Laboratory) collaborated to identify the basis of interpretation differences and to investigate if data sharing and reassessment resolves interpretation differences by analyzing a subset of variants. Methods ClinVar variants with submissions from at least two of the four participating laboratories were compared. For a subset of identified differences, laboratories documented the basis for discordance, shared internal data, independently reassessed with the ACMG-AMP guidelines, and then compared interpretations. Results 6,169 variants in ClinVar were interpreted by at least two of the participating laboratories, of which 88.3% were initially concordant. Laboratories reassessed 242/724 initially discordant variants, of which 87.2% (211) were resolved by reassessment with current criteria and/or internal data sharing. 12.8% (31) of reassessed variants remain discordant due to differences in application of the ACMG-AMP guidelines. Conclusion Participating laboratories increased their overall concordance from 88.3% to 91.7%, indicating that sharing variant interpretations in ClinVar, allowing identification of differences and motivation to resolve those differences, is critical to move toward more consistent variant interpretations.

Variant Interpretation: Functional Assays to the Rescue

“…Answering this question is a central challenge because it is clear that clinical laboratories are currently having trouble with using functional data to agree on variant interpretations. 80 The large scale of MAVE data provides several key advantages that could help alleviate the problematic nature of using functional data for interpretation. First, all variants within a MAVE are tested simultaneously so that measurements for different variants are readily comparable to each other, whereas one-at-atime assays are performed by different personnel in different research labs at different times.…”

Adapting crowdsourced clinical cancer curation in CIViC to the ClinGen minimum variant level data community‐driven standards

“…However, NGS also yields increasing numbers of variants that predominantly are of unknown significance and compounds the challenge of variant interpretation (Good, Ainscough, McMichael, Su, & Griffith, 2014; Kamps et al., 2017). As clinical analysis of large volumes of patient variant data becomes increasingly difficult, inconsistencies increase both in variant interpretation and reporting between laboratories (Harrison et al., 2017). This issue is compounded by propagation of these inconsistencies to widely accessed knowledgebases (Hoskinson, Dubuc, & Mason‐Suares, 2017; Yorczyk, Robinson, & Ross, 2015).…”

Clinotator: analyzing ClinVar variation reports to prioritize reclassification efforts

“…In many of the instances of discrepancies in assertion, consensus has been achieved simply by sharing evidence previously unavailable to one party 7– 9 . Harrison et al found that 87.2% of discordant variants were resolved by reassessment and data sharing 8 . Several initiatives 5, 10– 14 have had success in encouraging the public sharing of datasets and new studies, but for the foreseeable future, private data will continue to be a challenge to achieving consensus.…”

“…Recently, in the field of cardiovascular disease, there have been several high-profile instances of cardiovascular variants deemed to be highly pathogenic, yet not segregating with disease 9, 16, 17 . This unfortunate outcome is inevitable owing to the aforementioned reasons and illustrates a key issue: the continual need to share and reconcile new information with old data and reclassify clinical assertions as appropriate on a regular basis 6, 8 .…”