Ability of two rapid phenotypic AST methods to detect carbapenem-resistance in Gram-negative bacteria directly from positive blood cultures

Background: Bacteremia and sepsis are life-threatening diseases associated with high mortality. Ineffectiveness of empiric treatments underscore the urgent need for accurate and timely detection of resistant organisms from blood cultures. The EUCAST Rapid antimicrobial susceptibility testing (RAST) provides phenotypic susceptibility data from positive blood cultures in 4-8 h. Recently, the QuickMIC system was launched, which provides MICs in 2-4 h, while reducing hands-on time. While rapid AST may improve patient management, clinical usefulness is determined by their sensitivity to detect carbapenem-resistant organisms (CRO). This study evaluates the ability of RAST and QuickMIC to detect carbapenem-resistance in Gram-negatives directly from positive blood cultures using Broth microdilution (BMD) as reference.

Methods: 101 carbapenem-resistant or carbapenemase-producing Gram-negative isolates, (E. coli n=24, K. pneumoniae n=23, A. baumannii-complex n=29, P. aeruginosa n=25), were spiked into blood culture bottles and underwent testing using QuickMIC and EUCAST RAST and BMD as reference method. Genomes of all isolates were sequenced and analysed.

Results: QuickMIC compared to BMD resulted in an overall categorical agreement (CA) of 87.7 % for meropenem and third-generation cephalosporins. RAST CA ranged between 84.2-86.3% with 4, 6, 8 and 20 h readout (Figure1). The CA for meropenem alone was 77.3% with QuickMIC, and ranged from 59.4-71.6% with RAST at 4, 6, 8 and 20 h. QuickMIC displays higher major discrepancies (MD) for ceftazidime-avibactam compared to RAST at any assay endpoint, whereas RAST hat significantly higher very major discrepancies (VMD) for meropenem. The VMD for meropenem in RAST and QuickMIC was mostly associated with blaOXA and blaNDM gene groups.

Conclusions: QuickMIC and RAST perform similarly, but QuickMIC has a lower rate of VMD compared to RAST and provides actionable results at 2-4 h while best RAST performance is at 8 h. Genotype influences the rapid AST result quality, especially in meropenem and other beta-lactam antibiotics, which is important to think of when implementing rapid AST methods in clinical practice.