Comparing Next-Generation Sequencing Assays: Which Is Best for NTRK Fusion Detection?
Posted: Tuesday, January 4, 2022
A study published in the Journal of Molecular Diagnostics evaluated the specificity and clinical performance of four commonly available RNA-based next-generation sequencing assays: ArcherDX’s FusionPlex (AFL), Illumina’s TruSight Oncology 500 (TSO500), and Thermo Fisher’s Oncomine Precision and Oncomine Focus Assays (OPA and OFA). According to Brandon S. Sheffield, MD, of William Osler Health System, Ontario, Canada, and colleagues, all of the assays showed high specificity in detecting NTRK fusions and reported no false-positive results. However, each assay had its own strengths and limitations.
Study experiments were conducted using contrived samples [formalin-fixed, paraffin-embedded cell lines (n = 3) and SeraSeq formalin-fixed, paraffin-embedded reference material (three lots)], NTRK fusion–negative clinical samples (n = 30), and NTRK fusion–positive samples (n = 14). The samples were processed once using each of the four assays, and performance was subsequently compared.
The estimated limit of detection varied among assays. For the AFL assay, this was 30 to 620 fusion copies. TSO500 had a detection limit of 30 to 290 copies, and OFA/OPA had a limit of 1 to 28 copies. The quality control pass rate was also variable. TSO500 and OFA yielded success rates of 77% and 83%, respectively, but AFL samples had a lower rate of 27%. However, when substituting the assay’s original ReliaPrep extraction kit for Illumina’s AllPrep kit, this figure jumped to 43%. Despite this variability in detection limit, the NTRK fusion detection rate in quality control–validated samples was 100% across all four assays.
Although the tests were all highly specific, there were still discrepancies among them. Although the AFL assay required a high RNA input, it was able to detect novel gene fusions with moderate sensitivity. The OPA and OFA assays required small amounts of RNA, but they had less accuracy in detecting novel fusions that were not covered on the amplicon-based model. The TSO500 assay was well balanced in this regard.
Disclosure: For full disclosures of the study authors, visit jmdjournal.org.