Rapid identification of pathogens and antimicrobial resistance genes in positive hemoculture using nanopore sequencing

Abstract

Bloodstream infection (BSI) is a major burden of the healthcare system that has contributed to high mortality of patients. Rapid and reliable detection of pathogens affects antimicrobial management and patient outcomes. The current conventional techniques for diagnostic microbiology have been limited by uncultivatable pathogens, low sensitivity, and slowness. Our study aimed to identify pathogens and antimicrobial resistance (AMR) genes from positive blood cultures using Oxford Nanopore Technologies (ONT). The qualification and quantification of extracted DNA for sequencing were compared by using three different extraction kits (DNeasy PowerSoil Pro Kits, QIAamp DNA Mini Kit, and ZymoBIOMICS DNA Miniprep Kit). Moreover, the time to result and the comprehensiveness of bacterial identification were compared among conventional routine methods, BioFire FilmArray Blood Culture Identification (BCID), and ONT. By 30 positive blood cultures, we detected 31, 29, and 32 isolates using routine, BCID, and ONT methods, respectively. ONT could identify all isolates to the species level, whereas routine identified 83.87% (26 out of 31) at the species level and 89.65% (26 out of 29) of isolates could be identified by species name using BCID. There was a consensus among the three techniques that Escherichia coli was the highest prevalent pathogen, followed by Klebsiella pneumoniae. Moreover, we detected the human genome from sample number 30, alerting it as a false positive from the routine laboratory. Whole genome sequencing (WGS) of all isolates was performed using bioinformatics analytic software. Carbapenem resistance genes were detected that associated with the resistance phenotype. Time to get results since the point of hemoculture alarming positive by using routine, BCID, and ONT methods were 47.8 h, 1.07 h, and 3 h, respectively. We also determined the clones of our isolates by in silico multilocus sequence typing (MLST) and demonstrated the relatedness to other isolates recovered from Asian countries by generating a minimum spanning tree. Two K. pneumoniae demonstrated identical STs to isolates from other regions in Asia. In conclusion, the ONT with an appropriate extraction method provides rapid and comprehensive identification of BSI pathogens and their AMR genes. Therefore, this technique may help clinicians in an early de-escalation of broad-spectrum antimicrobials and could be applied to the field of infection control. Implication of the thesis: ONT, a long-read sequencing, could provide comprehensive analysis for rapid diagnosis that is promising for future applications in clinical microbiology and healthcare surveillance purposes.