Relationship of MET exon 14 splicing variants and MET exon 14 skipping expression in NSCLC FFPE samples.

e14603 Background: MET is the target of recurrent somatic alterations in approximately 7% of non-small cell lung carcinomas (NSCLC). A key class of therapeutically relevant MET mutations results in aberrant mRNA splicing, deletion of exon 14 from the mRNA transcript (METex14 skipping), and MET overexpression. METex14 skipping assays routinely use next-generation sequencing (NGS). However, because most NGS assays are DNA based, the relationship between METex14 DNA alterations and METex14 skipping mRNA expression is not well understood. Furthermore, because METex14 skipping may have low background levels, setting a threshold for calling METex14 skipping based on RNA levels alone is challenging. To investigate these relationships, we initiated a research study in 3 CLIA/CAP laboratories using an NGS panel (OCAv3, for research use only, not for use in diagnostic procedures) that includes 161 cancer genes and detects METex14 splicing variants and METex14 skipping mRNA isoforms. Methods: OCAv3 libraries were prepared following kit instructions with templating and sequencing performed on the Ion Chef and S5 XL systems. Reads were aligned to hg19 using Torrent Suite Software and variant calling was performed using Ion Reporter software. Two cell lines (HS746T and NCI-H596) known to contain METex14 splice site mutations were used as controls. Results: Of the first 28 NSCLC FFPE specimens selected having a range of METex14 RNA expression, we detected METex14 skipping alterations in 9 cases: 6 cases with variants at exon 14 canonical splice sites and 3 cases of novel deletions in the intron 13 poly-pyrimidine tract. METex14 skipping alterations were exclusive to other oncogenic driver variants but occasionally associated with MDM2 amplification. All cases with splice site variants showed elevated METex14 skipping expression. Cases without splicing variants had no or lower METex14 skipping expression and commonly harbor activating mutations in other oncogenes (6 KRAS, 5 EGFR, 1 ERBB2, 1 MYC and 1 CD74-NRG1 fusion). Conclusions: We demonstrate a framework for interpreting MET alterations that considers both MET splice site variants and METex14 skipping transcript expression in the context of additional oncogenic mutations. An update will be presented as additional cases accrue.