A rapid two-step algorithm detects and identifies clinical macrolide and beta-lactam antibiotic resistance in clinical bacterial isolates

摘要

Purpose: Aiming to identify macrolide and beta-lactam resistance in clinical bacterial isolates rapidly and accurately, a two-step algorithm was developed based on detection of eight antibiotic resistance genes.Methods: Targeting at genes linked to bacterial macrolide (msrA, ermA, ermB, and ermC) and beta-lactam (bla(TEM), bla(SHV), bla(CTX-N-1), bin(CTX-M-9)) antibiotic resistances, this method includes a multiplex real-time PCR, a melting temperature profile analysis as well as a liquid bead microarray assay. Liquid bead microarray assay is applied only when indistinguishable T-m profile is observed.Results: The clinical validity of this method was assessed on clinical bacterial isolates. Among the total 580 isolates that were determined by our diagnostic method, 75% of them were identified by the multiplex real-time PCR with melting temperature analysis alone, while the remaining 25% required both multiplex real-time PCR with melting temperature analysis and liquid bead microarray assay for identification. Compared with the traditional phenotypic antibiotic susceptibility test, an overall agreement of 81.2% (kappa = 0.614, 95% Cl = 0550-0.679) was observed, with a sensitivity and specificity of 87.7% and 73% respectively. Besides, the average test turnaround time is 3.9 h, which is much shorter in comparison with more than 24 h for the traditional phenotypic tests.Conclusions: Having the advantages of the shorter operating time and comparable high sensitivity and specificity with the traditional phenotypic test, our two-step algorithm provides an efficient tool for rapid determination of macrolide and beta-lactam antibiotic resistances in clinical bacterial isolates