Accurate detection of MET exon 14 skipping alterations in non–small cell lung cancer (NSCLC) may require complementary DNA- and RNA-based next-generation sequencing (NGS), according to a new real-world study. While participating laboratories achieved high testing accuracy using tissue samples, the researchers identified ongoing challenges in liquid biopsy testing and highlighted technical and bioinformatics factors that can lead to missed variants.
Published in The Journal of Molecular Diagnostics, the study examined 379 NSCLC samples collected between 2016 and 2024. Researchers identified 171 distinct MET exon 14 and adjacent splice-site variants, illustrating the considerable genetic diversity of these clinically actionable alterations. Of these, 114 variants were assessed using both DNA- and RNA-based NGS assays.
RNA analysis confirmed exon 14 skipping in 107 of the 114 variants. However, three clinically relevant alterations were initially missed by DNA-based analysis: two large deletions that were not detected by the bioinformatics pipeline and one synonymous splice-site variant that had been filtered from reporting despite causing exon 14 skipping. Reanalysis of the sequencing data identified the missed deletions, while RNA testing confirmed exon skipping in all three cases.
The study also demonstrated that not every variant located near MET exon 14 affects splicing. Seven suspected variants did not produce exon 14 skipping, whereas another variant located 10 bases upstream of exon 14 did. The findings illustrate why DNA findings alone may not always predict functional splicing effects.
RNA testing could not be performed for all samples because of limited tissue or inadequate RNA quality. Fifty-seven variants could not be evaluated, although 18 had previously been reported in the literature as exon 14 skipping variants.
The researchers also reported results from the first multinational external quality assessment schemes organized by Quality in Pathology GmbH. Among 50 laboratories participating in formalin-fixed, paraffin-embedded tissue testing, 49 (98 percent) met the required performance standard using a range of DNA-, RNA-, or combined NGS approaches.
Performance was lower for liquid biopsy testing. Pass rates improved from 38 percent in 2022 to 63 percent in 2024, but false-negative results remained common. Review of unsuccessful cases showed that inadequate sequencing coverage and bioinformatics filtering thresholds were the main reasons clinically relevant variants were missed. Those present at low allele fractions or involving intronic deletions were most likely to be missed.
Overall, the findings highlight that the wide range of MET exon 14 alterations presents ongoing diagnostic challenges. The researchers concluded that comprehensive assay validation – including bioinformatics pipelines – is essential, and that combining DNA- and RNA-based NGS, when sufficient material is available, can improve detection and help confirm whether identified variants produce exon 14 skipping.
Newsletters
Receive the latest pathologist news, personalities, education, and career development – weekly to your inbox.
