Targeted RNA-sequencing of cancer predisposing genes in cancer patients increased the yield of diagnostic

High-throughput sequencing is becoming more and more popular in diagnosis of hereditary cancers. Besides improving the rate of diagnostic, high-throughput sequencing results in identification of increasing number of variants of uncertain significance (VUS), which tends to become one of the major challenges for clinicians. Variants which are potentially altering RNA splicing are particularly interesting as their direct impact on protein is not clear. Currently the bioinformatics prediction tools are used to assess canonical splice-site variants in-silico and the results can be validated by RT-PCR and direct sequencing of cDNA. Despite their great potential in predicting the effect of canonical splice site variants, these prediction tools are not efficient on interpreting the effect of intronic variants which may result in aberrant transcripts. Here we investigate the utility of targeted RNA-seq as an additional method to genomic DNA-sequencing to improve the diagnostic rate. We performed targeted RNA-sequencing on RNA extracted from whole blood cells of cancer patients. A panel was designed to capture all known exons of 34 cancer predisposing genes. In parallel to targeted RNA-seq we carried out RT-PCR and Sanger sequencing of cDNA fragments. Taking this approach we were able to confirm the transcript-level impact of splice region variants in STK11 (c.597 G> A), TP53 (c.375+5 G> A), MSH2 (c.2211-10T> A, and 1979 A> G) and consequently improving the classification of these variants from variant of uncertain significance to likely pathogenic and pathogenic. On the other hand, targeted-RNA seq were not able to detect low expressed aberrant transcripts in genes such as BRCA1, BRCA2 and CHEK2 most likely due to low expression level of these genes in whole blood. Overall, our results emphasize the pros and cons of the targeted RNA-sequencing as an additional method to WGS and WES.