P1.32 Hand-held rapid whole genome nanopore sequencing to predict neisseria gonorrhoeae antibiotic susceptibility: steps towards clinic based tailored antimicrobial therapy. (8th July 2017)

• Record Type: Journal Article
• Authors: Phillips, Laura T; Witney, Adam; Izquierdo-Carrasco, Fernando; Mayes, Simon; Wright, Amber; Laing, Ken; Gould, Kate; Pond, Marcus; Hall, Catherine L; Harding-Esch, Emma M; Butcher, Phillip; Zhou, Liqing; Sadiq, Syed T
• Is Part Of: Sexually transmitted infections. Volume 93(2017)Supplement 2
• Page Start: A56
• Page End: A56
• Publication Date: 2017-07-08
• Subjects: Sexually transmitted diseases -- Periodicals; HIV infections -- Periodicals; 616.951005
• DOI: 10.1136/sextrans-2017-053264.140
• Languages: English
• ISSNs: 1368-4973

Abstract

Abstract : Introduction: Next generation sequencing can accurately predict antibiotic susceptibility in Neisseria gonorrhoeae (NG) allowing preservation of first-line treatments in the face of widespread antimicrobial resistance (AMR). The rapid nature of novel hand-held nanopore sequencing (NPS) gives promise for utility at the point of care. We evaluated time to result post DNA extraction, and accuracy of MinION (Oxford Nanopore Technologies) NPS to predict phenotypic antimicrobial susceptibility (PAMS) of NG to ciprofloxacin and azithromycin. Methods: One-directional (1-D) NPS using bar-coded DNA library preparations from 48 NG isolates, prospectively collected from a London clinic, were run on NPS flow cells (3 per R9.0 flow cell) and illumina MiSeq as a comparator. NPS raw sequences were transferred to the cloud for base-calling, alignment, and variant calling using standard tools. Results: Mean time for 1-D library preparation was roughly 1 hour; NPS and alignment took <40 min per sample with single nucleotide polymorphism (SNP) calling adding little extra time. NPS genome coverage was >30X per isolate. Of 48 samples, PAMS to ciprofloxacin, and azithromycin was 74% and 87% respectively. Accuracy of NPS-based genotypic susceptibility, defined as absence of any known AMR-associated SNP's, to predict PAMS for ciprofloxacin and azithromycin, was 34/34 (100%; 95% CI 89.8%–100%) and 35/40 (87.5%; 95% CI 73.9%–94.5%) respectively. Accuracies improved significantly for azithromycin when only high quality reads were included, and with Illumina sequencing. 30 of the 34 isolates susceptible to azithromycin were also susceptible to ciprofloxacin, and 3 of 6 isolates resistant to azithromycin were also resistant to Ciprofloxacin. Conclusion: NPS accurately predicted ciprofloxacin PAMS but was less accurate for azithromycin. With new iterations of the technology, imminent rapid barcoded library preparation (10 min) and rapid DNA extraction from clinical samples, NPS may allow accurate ceftriaxone-adjunctive treatment combinations, for a substantial proportion of patients.