Scaling resolution of variant classification differences in ClinVar between 41 clinical laboratories through an outlier approach

Abstract

ClinVar provides open access to variant classifications shared from many clinical laboratories. Although most classifications are consistent across laboratories, classification differences exist. To facilitate resolution of classification differences on a large scale, clinical laboratories were encouraged to reassess outlier classifications of variants with medically significant differences (MSDs). Outliers were identified by first comparing ClinVar submissions from 41 clinical laboratories to detect variants with MSDs between the laboratories (650 variants). Next, MSDs were filtered for variants with ≥3 classifications (244 variants), of which 87.6% (213 variants) had a majority consensus in ClinVar, thus allowing for identification of outlier classifications in need of reassessment. Laboratories with outlier classifications were sent a custom report and encouraged to reassess variants. Results were returned for 204 (96%) variants, of which 62.3% (127) were resolved. Of those 127, 64.6% (82) were resolved due to reassessment prompted by this study and 35.4% (45) resolved by a previously completed reassessment. This study demonstrates a scalable approach to classification resolution and capitalizes on the value of data sharing within ClinVar. These activities will help the community move toward more consistent variant classifications, which will improve the care of patients with, or at risk for, genetic disorders.

Keywords: clinvar; data sharing; variant interpretation.

Figure 1.. Comparison of classification concordance and discordance between all ClinVar submitters and the 41 clinical laboratories participating in this study.

Only variants with classifications from ≥2 submitters were included from either group. For variants with greater than two classification terms submitted, the two most discordant terms were used to assign the level of discordance. Data from April 2017.

Figure 2.. Flowchart and outcome of variant resolution efforts.

Classifications submitted to ClinVar from 41 clinical laboratories were first compared to identify medically significant differences (MSDs; 650 variants). Next, to identify MSDs that reach a majority consensus with an outlier classification, only MSDs with ≥3 submitters were analyzed (244 variants), of which 87.3% (213 variants) had a majority consensus in ClinVar. A further breakdown of the 213 variants into three majority consensus/outlier scenarios is provided. Clinical laboratories with outlier classifications were each sent a report and asked to reassess those variants. Clinical laboratories returned results from 204 variants, of which 62.3% (127 variants) were reclassified by the outlier laboratory resulting in classification resolution. MC, majority consensus.

Figure 3.. Distribution of interpretation differences and resolution outcome per disease area.

Distribution of outlier scenarios within each disease area before (“Initial”) and after (“Outcome”) reassessment, including proportion resolved. Initial and outcome variant counts differ due to incomplete reassessments by outlier submitters. MC, majority consensus.

Figure 4.. Outcome of variant reassessments.

Resolution status of 204 variants reassessed by clinical laboratories. For those variants where resolution was achieved, the pie chart insert depicts factors underlying the resolution.