Assessing the diagnostic yield of targeted next-generation sequencing for melanoma and gastrointestinal tumors

Abstract A common rationale in molecular diagnostic laboratories is that implementation of next-generation sequencing (NGS) enables simultaneous multi-gene testing, allowing increased information benefit as compared to non-NGS assays. However, minimal published data exist to support this justification. In this study, we compared clinical diagnostic yield of TruSight Tumor 26 Sequencing Panel (TST26) in melanoma, colorectal (CRC) and gastro-intestinal stromal (GIST) tumors to non-NGS assays. 1041 formalin-fixed, paraffin embedded (FFPE) tumors, of melanoma, CRC and GIST were profiled. NGS results were compared to non-NGS single-gene or single-variant assays with respect to variant output and diagnostic yield. 79% melanoma and 94% CRC tumors were variant-positive by panel testing. TST26 panel improved BRAF variant detection in melanoma as compared to single-variant BRAF V600E/K routine tests by 24% and also detected variants in genes other than BRAF, NRAS and KIT which could impact patient management in 20% additional cases. NGS enhanced diagnostic yield in CRC by 36% when compared to routine single gene assays. In contrast, no added benefit of NGS based testing for GIST tumors was observed.TST26 panel either missed or inaccurately called complex insertion/deletion variants in KIT exon 11, which were accurately identified by non-NGS methods. Our findings demonstrate the differential impact of cancer site and variant type on diagnostic test information yield from NGS assays.