FFPE RNA-based ddPCR assay: An effective and efficient confirmation for <i>MET</i> exon 14 skipping in NSCLC.

Abstract

e20643 Background: MET exon14 skipping ( MET ex14sk) mutations occur in 3-5% of non-small cell lung cancer (NSCLC) and are responsive to MET tyrosine kinase inhibitors. While DNA profiling is commonly used to detect MET ex14sk, it has several limitations. DNA-based analysis heavily relies on in silico predictions, which may not reliably confirm transcriptional consequence. Furthermore, low tumour purity and formalin-induced DNA damage can compromise analysis, resulting in missed low frequency variants. Additionally, deletions outside the detection range of amplicon-based analysis may be overlooked, leading to false negatives. Although RNA sequencing (RNA-Seq) can detect MET ex14sk transcripts, formalin-fixed paraffin-embedded (FFPE) RNA is often degraded, making it challenging to produce high-quality NGS data that distinguishes true signals from background noise. To address these issues, we established an RNA-based digital droplet PCR (ddPCR) assay to confirm MET ex14sk in FFPE RNA samples. Methods: We analysed FFPE RNA available from 142 NSCLC patients, including 22 cases with potential MET ex14sk variants identified by DNA profiling. SpliceAI was used to predict spliceogenic effects, with Δ scores of &lt;0.20 and &gt;0.80 set as negative and positive cutoffs, respectively. All samples were tested using the RNA-based ddPCR assay. Results: Twenty potential MET ex14sk variants were identified in exon 14 and adjacent intronic regions in the 22 cases. Of these, 50% (10/20) were intronic, and 60% (12/20) were novel including 4 intronic changes. SpliceAI confidently predicted spliceogenic effects in 18 variants (15 positive and 3 negative, based on Δ scores &gt;0.80 and &lt;0.20), all of which were confirmed by ddPCR. For two variants with intermediate SpliceAI confidence (Δ scores: 0.57 and 0.58), ddPCR and RNA-Seq demonstrated skipping. Among the remaining 120 cases, 18 failed ddPCR testing due to poor RNA quality. In the 102 successfully analysed cases, ddPCR excluded MET ex14sk in 101 cases and detected one unexpected positive in a case without known driver mutations. Despite the absence of an underlying DNA variant, RNA-Seq confirmed MET ex14sk in this unexpected case, ruling out a false positive. Conclusions: RNA-based analysis is accurate and essential for clinical confirmation of MET ex14sk, particularly in cases with novel intronic variants or no detectable driver mutations. The ddPCR assay is a cost-effective method with a rapid turnaround time, offering a reliable solution for MET ex14sk confirmation in NSCLC patients.