Comparative Evaluation of Assays for Broad Detection of Molecular Resistance Mechanisms in Enterobacterales Isolates.

Rapid detection of antimicrobial resistance in both surveillance and diagnostic settings is still a major challenge for the clinical lab, compounded by the rapid evolution of antibiotic resistance mechanisms. This study compares four methods for the broad detection of antibiotic resistance genes in Enterobacterales isolates: two multiplex PCR assays (the Streck ARM-D beta-lactamase kit and the OpGen Acuitas AMR Gene Panel u5.47 (research use only [RUO]) and one microarray assay (the Check-MDR CT103XL assay), with whole-genome sequencing as a reference standard. A total of 65 Gram-negative bacterial isolates, from 56 patients, classified by phenotypic antimicrobial susceptibility testing (AST) as showing resistance to beta-lactam antimicrobials (extended-spectrum beta-lactamase [ESBL] positive or resistance to third-generation cephalosporins or carbapenems) were included in the study. Overall concordance between the molecular assays and sequencing was high. While all three assays had similar performance, the OpGen Acuitas AMR assay had the highest overall percent concordance with sequencing results. The primary differences between the assays tested were the number and diversity of targets, ranging from 9 for Streck to 34 for OpGen. This study shows that commercially available PCR-based assays can provide accurate identification of antimicrobial resistance loci in clinically significant Gram-negative bacteria. Further studies are needed to determine the clinical diagnostic role and potential benefit of such methods.

Evaluation of rapid phenotypic identification and antimicrobial susceptibility testing in a pediatric oncology center.

Identification (ID) and antimicrobial susceptibility testing (AST) remain rate limiting steps in producing actionable data for clinical care of bloodstream infections. Rapid, automated phenotypic ID and AST by fluorescent in situ hybridization and automated microscopy were used to characterize blood stream infections in a predominantly pediatric oncology patient population. Results were compared to standard of care (SOC) phenotypic methods. The Accelerate Pheno System (AXDX) had a sensitivity of 91.2% and an accuracy of 100% to the genus level for identification, and an overall categorical agreement 91.2-91.8% for susceptibility, depending on the breakpoints used. The AXDX required a mean time of 1.4hours for identification and 6.6hours for susceptibility testing compared to SOC, requiring 32.5 and 46.7hours, respectively. Identification and susceptibility by rapid phenotypic methods shows a high degree of accuracy; the marked reduction in time to results may have significant implications for patient care.