Response to Cragun et al.

To the Editor: Thank you for your attention to our recent publication (1). As the primary focus of this manuscript was to improve understanding in how multiplex testing and Next-generation sequencing (NGS) technology are being integrated into clinical practice, details on our approach to specific variant classification are outside the scope of this report. However, the classification of DNA variants according to pathogenicity is clearly relevant to the quality of multiplex testing results, and to the diagnostic yields reported by Cragun et al. Numerous approaches to variant classification have been described and validated (2–5), yet specific variants continue to challenge even the most robust approaches, which can be limited by bias and limited or conflicting available data. Based on the evidence available to our laboratory, the MSH2 p.P349A is reported as a class 4 (likely pathogenic) variant because the combined evidence achieves a likelihood ratio>20:1. This included clinical phenotype and family segregation data, molecular tumor studies, and is quantified as recently described by InSiGHT (6). We appreciate that our evidence appears to differ significantly from that which authors describe below to support their benign assessment; this highlights a specific challenge to variant classification when conflicting data are fragmented and not made rapidly and publically available for comprehensive assessment. Authors’ comments additionally support the proposition that it is through the sharing of data and collaborative efforts that we will achieve the most accurate predictions of cancer risk for our patients (6–9).