Variable zinc concentrations among commercially available Mueller-Hinton Broth brands affects SIR interpretations during broth microdilution in vitro antimicrobial susceptibility testing with the novel metallo-beta-lactamase inhibitor APC148

The well-known discrepancy between the in vitro and in vivo efficacy of {beta}-lactam antibiotics in metallo-{beta}-lactamase (MBL) containing Gram-negative bacteria has during recent years been found to be at least partially explained by zinc levels at infection sites being much lower than those found in conventional cation-adjusted Mueller-Hinton Broth (caMHB) media used for in vitro susceptibility testing. Previous studies have also demonstrated that a high variability exists with respect to zinc content in caMHB from different manufacturers, potentially leading to differences in SIR interpretations for {beta}-lactam antibiotics with MBL-carrying isolates, depending on the brand of caMHB used for testing. APC148 is a zinc-chelating compound acting as an inhibitor of MBL enzymes and is currently undergoing phase one in clinical trials. In this study, ten clinical isolates of Klebsiella pneumoniae, Escherichia coli, Pseudomonas aeruginosa and Acinetobacter baumannii harbouring MBLs (NDM, n = 6; VIM, n = 4; IMP, n = 1) were tested in a broth microdilution assay with meropenem and APC148, employing caMHB of various brands. One K. pneumoniae strain carrying only a serine-{beta}-lactamase (KPC-2) was included as control. Antimicrobial susceptibility testing (AST) by broth microdilution was performed according to the European Committee on Antimicrobial Susceptibility Testing (EUCAST). MICs of meropenem alone and in combination with inhibitors were tested in four cation-adjusted Mueller Hinton II (caMHB).The four caMHBs used were analysed by ICP-MS/MS and found to have highly vaiable zinc content (in the range 0.4 to 1.9 {micro}g/mL, corresponding to 5.6 to 29.4 {micro}M). At 16{micro}g/ml APC148 lowered the MIC in nearly all strains and in caMHBs from all manufacturers. At lower concentrations of APC148 (4 or 8 {micro}g/mL), this MIC reduction could however only be retained when the zinc concentration in the broth was low, indicating that higher concentration of inhibitor is needed during in vitro Mic testing when using caMHB from certain manufacturers. The present work clearly shows that not taking the zinc concentration of the caMHB used into consideration when estimating MIC performance of compounds functioning through interactions with zinc may be a considerable source of error, and specifically when investigating potential inhibitors of metallo-{beta}-lactamase (MBL) enzymes. The present work supports the call for standardising zinc content in caMHB to be used for this purpose, to ensure that MIC results for drug combinations involving the use of zinc-chelating compounds are consistent and reproducible across laboratories.