Enter Note Done


Keywords: Cancer; Clinical testing; Diagnostics; Mutation detection; Variant calling11

S. Pan; A. Brown; B. Leclair; M. Elias; D. Chen; J. Kidd; K. Bowles; B. Coffee; B. Roa; D. Mancini-DiNardo

Affiliation: Myriad Genetic Laboratories, Inc., Salt Lake City, UT

Background: PMS2, a Lynch syndrome-associated DNA mismatch-repair gene, often is included in NGS hereditary pan-cancer panels. Molecular testing of PMS2 is complicated by the interference of highly homologous pseudogenes. The most homologous pseudogene, PMS2CL, is >98% identical to PMS2 exons 11-15. Therefore, additional analysis is required for variants identified in exons 11-15 to determine whether they are located in PMS2 or PMS2CL. Correct allocation of variants identified by NGS in this region is critical for proper clinical management.
Objective: The purpose of this analysis was to evaluate the frequency with which common PMS2CL variants occur in PMS2, with a focus on those that are considered pathogenic when they occur in PMS2.
Methods: Pathogenic/likely pathogenic sequence variants (PV/LPVs) detected from July 2016-April 2019 and large rearrangements (LRs) detected from May 2017-April 2019 in the PMS2 pseudogene region (exons 11-15) were evaluated in individuals tested using an NGS hereditary pan-cancer panel. These variants were initially identified by NGS, then confirmed by orthogonal assays as being present in either PMS2 or PMS2CL. PV/LPVs were selected for this analysis if they had >100 observations upon NGS testing but were confirmed to be in PMS2 in <1% of cases.
Results: Two sequence variants and three LRs were assessed (four variants in exons 13 and 14; one in exon 15). Collectively, these variants were detected in 12,217 individuals. In 99.91% (12,206/12,217) of cases, the variants were confirmed orthogonally to be present in PMS2CL. In 11 (0.09%) individuals, the variant was located in PMS2. The rarest PV in PMS2 was c.2186_2187del (p.Leu729Glnfs*6); only one (<0.01%) patient was confirmed to carry this PV in PMS2.
Conclusion: Comprehensive testing and a large testing population enabled identification of PV/LPVs that are predominantly present in PMS2CL but can occur in PMS2 with extremely low frequency. These data highlight the need to disambiguate PVs in PMS2 versus PMS2CL. It is tempting to assume that when certain sequencing variants and LRs are known to occur in PMS2CL >99% of the time, additional analysis is unnecessary. We have demonstrated that, though rare, these variants can occur in PMS2. Failure to confirm the PV/LPV location can produce a false negative result with significant implications for clinical management.