To the Editor—We read with interest the letter by Kazi et alReference Kazi, Shetty and Rodrigues ¹ in which the authors reflect on the question of probable higher carbapenem minimum inhibitory concentrations (MICs) when dual carbapenem resistance mechanisms coexist.

Greece was the first country where double carbapenemase-producing Klebsiella pneumoniae (CPKP) were detected in 2009.Reference Meletis, Tzampaz, Protonotariou and Sofianou ² Since then, they have reached approximately 5% of CPKPs isolated in Greek hospitals.Reference Giakkoupi, Papagiannitsis and Miriagou ³ These double CPKPs commonly bear KPC-2 plus VIM-1 carbapenemases. More recently, NDM-1-producing K. pneumoniae have also emerged in Greece.Reference Voulgari, Gartzonika and Vrioni ⁴

To further investigate the aforementioned question with a different carbapenemase combination than that used by Kazi et al, we performed the imipenem E-test (bioMérieux, Marcy l’Etoile, France) in 4 groups of nonduplicate CPKPs isolated from various clinical samples from patients hospitalized in AHEPA University Hospital of Thessaloniki. Group A consisted of 20 KPC-2-producers; group B consisted of 20 VIM-1-producers; group C consisted of 20 NDM-1-producers; and group D consisted of 20 KPC-2+VIM-1-producers. Bacterial identification and initial antimicrobial susceptibility testing of the isolates was performed using the VITEK2 automated system (bioMérieux, Marcy l’ Étoile, France), and the presence of carbapenemase-encoding genes was determined by phenotypic tests followed by polymerase chain reaction assays with specific primers to identify the presence of carbapenemases (ie, bla _VIM, bla _KPC, or bla _NDM).Reference Tsakris, Poulou and Pournaras ^{5 ,} Reference Nordmann, Poirel, Carrër, Toleman and Walsh ⁶

The imipenem MICs for the 3 groups are shown in Table 1. Moreover, we calculated the geometric mean of the MIC values for each study group as the nth root (n=the number of isolates) of the product of the MIC values using a geometric mean calculator. (A value of 64 was used conventionally for isolates with MIC>32 μg/mL.) Groups A and C had lower geometric means (49.496 and 47.130, respectively) compared to the conventional 64.000 of B and D.

TABLE 1

Relation of Minimum Inhibitory Concentrations of Imipenem to Carbapenemase Presence in K. pneumoniae Isolates

NOTE. MIC, minimum inhibitory concentration; KPC-2, Klebsiella pneumoniae carbapenemase-2; VIM-1, Verona integron-encoded metallo-β-lactamase-1; NDM-1, New Delhi metallo-β-lactamase-1.

Apparently, when studying large groups of carbapenemase-producing bacteria, more mechanisms may code for more resistance. This finding is in accord with the results of Kazi et al and with our previous study on carbapenem-resistant Pseudomonas aeruginosa, in which various mechanisms (eg, the VIM-2 carbapenemase, efflux pump over-expression, loss of the OprD porin, and the presence of inducible AmpC beta-lactamases) where taken into account.Reference Meletis, Vavatsi and Exindari ⁷

For single isolates, however, this cannot be considered the rule. Kazi et al found a Klebsiella spp. coproducing NDM-1 and VIM-1 with carbapenem MICs<2 μg/mL. Lower MICs in double-carbapenemase producers than those observed commonly in single-carbapenemase-producing isolates have also been observed elsewhere.Reference Meletis, Chatzidimitriou and Malisiovas ⁸ These findings can be attributed to the variable expression levels of the carbapenemase-encoding genes within bacterial cells and to the different hydrolytic potential of each enzymatic type. It is known, for example, that OXA-type carbapenemases have a lower potential of carbapenem hydrolysis than the metallo-beta-lactamases and Klebsiella pneumoniae carbapenemases.Reference Poirel, Potron and Nordmann ⁹

Interestingly, in our study, differences were not detected between groups B and D, probably because the presence of VIM-1 was enough to raise the imipenem MICs over 32 μg/mL for all 20 isolates of group B. Differences between these 2 groups may exist at higher MIC levels; however, these differences would be not relevant in clinical practice because the use of carbapenems, alone or in combination with other antibiotics, is not applicable with MIC values >32 μg/mL.

Overall, we agree with Kazi et al that the emergence of double-carbapenemase-producing bacteria is alarming and further confines the already limited treatment options.