(686.22) Optimization and Validation of Nanopore Based Sequencing Method for Molecular Testing of CNS Tumours

Poster Board Number: D22

Mashiat Mimosa (University of Toronto, University of Toronto), Wafa Al-ameri (St. Michaels Hospital), Michael Nakhla (University of Toronto), Jared Simpson (University of Toronto, University of Toronto), Karel Boissinot (St. Michaels Hospital), David Munoz (University of Toronto, University of Toronto), Sunit Das (University of Toronto), Ramzi Fattouh (University of Toronto, University of Toronto), Rola Saleeb (St. Michaels Hospital, St. Michaels Hospital, St. Michaels Hospital)

Background

The World Health Organization (WHO) introduced molecular identifiers for the diagnosis and prognosis of CNS tumors including the mutational status of isocitrate dehydrogenase or IDH genes in glial tumors. Currently used immunohistochemistry (IHC) is not capable of detecting the non-canonical mutations, and sequencing is often required as a follow-up. Current next-generation sequencing (NGS) technologies used in tumor molecular marker detection introduce key challenges including high capital cost, complex infrastructure requirements, and long turnaround times. These challenges considerably limit the ability to perform NGS testing in many pathology laboratories. In this study, we aimed to use third generation nanopore sequencing technology to resolve these limitations. The Oxford Nanopore MinION, a pocket-sized nanopore sequencing device, has minimal capital costs and infrastructural requirements, and shorter turnaround times. However, the nanopore technology has not been validated in clinical practice and has not been optimized on formalin-fixed paraffin-embedded (FFPE) tissue.

Methods

DNA extraction of selective tumor areas was performed from the corresponding FFPE tissue blocks from a cohort of gliomas with confirmed IDH1 and IDH2 gene statuses (n=65). A PCR amplicon-based approach was used to amplify hot spots of the IDH1 and IDH2 genes, starting with 30ng DNA material. The amplicon libraries were sequenced for 2 hours in multiplex on the MinION device and IDH SNPs were called with the Nanopolish software.ASIP Abstract Mashiat Mimosa

Results

26 IDH mutant samples were identified: 21 IDH1 R132H, 2 IDH1 R132G, 2 IDH2 R172G, and 1 IDH2 D177H. All cases showed concordant IDH mutational status when compared to the reference methods (IHC or NGS) and both analytical sensitivity and specificity were 100%. Precision analysis of variant allele frequency (VAF) showed the coefficient of variation was less than 5% (both inter and intra runs), and the limit of detection for VAF was 3%. The range of read depth obtained was 882X to 43,000x with an average of 20,000x. This assay revealed a $50-$100 material cost per sample, and the time taken from extracted nucleic acid to final result generation was 1-2 business days.

Conclusion

This project is the first to optimize and validate an approach to detect SNP mutations in FFPE samples using nanopore technology. It has demonstrated the feasibility and efficacy of the nanopore amplicon sequencing method in cancer FFPE tissue with excellent test performance characteristics, significantly shorter turnaround times at considerably lower costs and without any infrastructural needs. Thus, it can be used to circumvent challenges to current NGS testing platforms and can be the milestone that would make cancer NGS testing available for every laboratory.





Figure 1: Example of detection of IDH1 mutation (R132H c.395C>T) on two samples (top: mutant; bottom: wildtype); Table 1: Concordance analysis of 65 FFPE samples sequenced in multiplex