Validation of a DNA-Based Next-Generation Sequencing Test for Molecular Diagnostic Variant and Fusion Detection in...

This study evaluated two DNA-based next-generation sequencing approaches for detection of single-nucleotide variants (SNVs) and fusions in formalin-fixed, paraffin-embedded (FFPE) tissue specimens and liquid biopsies (AVENIO Targeted and Surveillance Panels). Reference standards (n = 7 with SNVs and structural variants) and real-world FFPE tissue specimens (n = 26 lung, colorectal, pancreas, ovary, breast, prostate, melanoma, and soft tissue cancer cases with n = 27 samples), liquid biopsies [n = 29 cases with n = 40 plasma/cell-free DNA (cfDNA) samples], and one pleural effusion (lung cancer) were analyzed by the AVENIO workflow for known SNVs (BRAF, BRCA1/2, CTNNB1, EGFR, KRAS, MET exon 14 skipping, NRAS, PIK3CA, and TP53), insertions and deletions (ERBB2 and KIT), and fusions (ALK and ROS1). Detection of SNVs, insertions and deletions, and fusions was reliable in 24 of 26 FFPE tissue specimen cases and at 1% allele frequency in 5 of 5 cfDNA reference standards and 37 of 40 plasma/cfDNA samples. Pitfalls were identified for the AVENIO workflow in calling and listing of clinically relevant variants, requiring additional manual inspection. Moreover, laboratory workflows are distinct for FFPE tissue specimens and liquid biopsies as well as time-consuming for sample quality control assays. In summary, the DNA-based next-generation sequencing approaches may be suitable for routine molecular pathology diagnostics on careful data interpretation and further optimization of the technical and laboratory workflows.