Detection of multiple types of cancer driver mutations using targeted RNA sequencing in NSCLC.

Abstract

e15043 Background: DNA NGS sequencing has been widely used as a companion diagnostic method for matching NSCLC patients with appropriate personalized target therapy. RNA-based NGS has been proved to be valuable in detecting fusion and exon skipping mutations and therefore recommended by NCCN guidelines for these mutation types. It would be helpful to have a diagnostic method that suits different types of clinically actionable mutations in a single tube. Methods: We developed an RNA-based hybridization panel targeting actionable driver oncogenes in solid tumors. Experimental and bioinformatics pipelines were optimized for detection of fusion, SNV and indel mutations, as well as gene expression quantification. Reference materials were used for analytical validation of the panel, and 1253 FFPE samples from NSCLC patients were analyzed by DNA panel sequencing and RNA panel sequencing in parallel to assess the performance of the RNA panel in detection of multiple types of clinically actionable mutations. Results: Analytical validation showed a detection specificity of 100.0% for SNVs and fusions. The targeted RNA panel achieved a 1.45-3.15 copies/ng limit of detection (LOD) for SNVs and a 0.21-6.48 copies/ng LOD for fusions. Gene expression analysis revealed a correlation greater than 0.9 between the RNA-seq and targeted RNA panel results. Among 1253 NSCLC FFPE samples, multiple mutation types were called at both DNA and RNA levels and compared between the two assays. Twenty-one fusion and six METex14 skipping events were only detected by the RNA panel. The positive percent agreement (PPA) and positive prediction value (PPV) were 98.08% and 98.62% for targetable SNVs and 98.15% and 99.38% for targetable Indels, respectively. We also explored the relationship of RNA expression with gene copy number and protein expression. The RPKM (reads per kilobase per million mapped reads) of EGFR transcripts exhibited a linear relationship with copy number, whereas MET gene expression seemed to be regulated in a more complexed manner. In IHC analysis, all MET IHC 3+ samples had significantly higher RPKM levels than samples with MET IHC level of 2+/1+. Conclusions: Parallel DNA and RNA systematic analysis demonstrated the accuracy and robustness of the RNA sequencing panel for detection of multiple types of clinically actionable mutations. The simplified experimental workflow, lowered cost and sample consumption, along with the superior performance in simultaneous detection of multiple types of mutations will make RNA panel sequencing an effective method in clinical testing for NSCLC.