Zinc Concentration Affects Metallo-Beta-Lactamase Susceptibility Testing of Enterobacterales

LETTER

I  am writing in response to conclusions the authors of the referenced publication (1)   have drawn with regard to the effect of variable zinc concentrations on antimicrobial susceptibility testing of metallo-β-lactamase-producing Enterobacterales.

The premise of the authors was based on initial broth microdilution testing of metallo-β-lactamase-producing strains of Enterobacterales with three different lots of Mueller-Hinton (MH) broth obtained from three different manufacturers. Those experiments appeared to show differing MICs for meropenem between lots and manufacturers. In addition, cations were depleted in those lots by addition of various concentrations of EDTA (from 3 to 100 mg/liter).

The authors showed that the zinc concentrations (measured by gas-plasma analysis) were between approximately 1.4 and 0.3 mg/liter for all lots from the three manufacturers. With cation depletion, only three Escherichia coli strains with Klebsiella pneumoniae carbapenemase (KPC) enzymes showed no reduction in meropenem MICs over the range of EDTA added.

They suggested that variable zinc ion concentrations even at or below approximately 1 mg/liter can have a marked effect on broth-microdilution susceptibility testing results (at least for meropenem).

As a previous chairholder of the CLSI Culture Media Subcommittee and as the deputy convenor of Working Group 4 (Microbiology and Molecular Diagnostics) of ISO TC212: Clinical Laboratory Testing and In Vitro Diagnostic Test Systems, I have reviewed the data provided in the manuscript and would comment on some of the conclusions based on their methodology for the study.

There are two major standards for broth microdilution susceptibility testing. These are CLSI document M7:A11 (11th edition) (2) and ISO 20776-1:2019: Susceptibility Testing of Infectious Agents and Evaluation of Performance of Antimicrobial Susceptibility Test Devices, Part 1 (3). Neither of these documents were referenced in the authors’ manuscript.

We agree that zinc ion concentration in Mueller-Hinton broth is not directly referenced in these standards (4). Measurement of zinc from different lots of Mueller-Hinton broth is also a requirement for manufacturers to achieve these low concentrations of zinc ions, particularly for quality control of carbapenem susceptibility testing. As the authors noted in Table 2 of their manuscript, zinc ions in all the MH broth samples tested were well below the requirements of the standard methods.

The authors used very large concentrations of EDTA for cation depletion, which would affect not only zinc but also other cations and susceptibility testing results (e.g., calcium and magnesium). Their experiments would suggest that this is not a zinc-specific issue, Further, they did not analyze for zinc or other cations in any of the EDTA-depleted samples. It is therefore difficult to attribute their susceptibility-testing results to variability only of small differences in zinc ions.

Mueller-Hinton medium is a complex culture and testing medium. It would be impossible to have a concentration of zinc ions that did not vary by more that 0.1 or 0.2 mg/liter even at low levels. The standards are correct. The broth works for quality control of carbapenems within 1 to 2 doubling dilutions with low zinc concentrations in broths from different manufacturers. Those broth media may well have differences in other cations that can contribute to variability as the authors have described. There are likely other factors that contribute to carbapenem (and metallo-β-lactamase) MIC variability, especially in strains with multiple resistance mechanisms. It is apparent that many factors need to be considered in these types of studies.