In vitro evaluation of different antimicrobial combinations against carbapenemase-producing Klebsiella pneumoniae: the activity of the double-carbapenem regimen is related to meropenem MIC value

Objectives and methods: We evaluated the in vitro activity of different antimicrobial combinations with and without colistin against 39 carbapenem-resistant Klebsiella pneumoniae (CR-Kp) strains (colistin! meropenem/ doripenem, colistin ! tigecycline, colistin ! rifampicin, gentamicin! meropenem, gentamicin! tigecycline and the double-carbapenem regimen meropenem! ertapenem) using the chequerboard method. The triple combination colistin ! meropenem! tigecycline was also tested. In addition, killing studies were performed for meropenem! ertapenem. Results: Gentamicin-based combinations showed a high level of synergy. Meropenem ! ertapenem was synergic in 12/39 (30.7%) of the strains, whereas based on killing studies 1 % MIC meropenem ! 1 % MIC ertapenem and 2 % MIC meropenem ! 1 % MIC ertapenem combinations were bactericidal and synergic at 24 h [mean area under the bactericidal curve (AUBC) 54.9+26.1 and 44.2+15.3 compared with 1 % MIC meropenem (134.5+40.1) and 2 % MIC meropenem (126.4+5.4), respectively, P , 0.0001]. When the results were stratified according to meropenem MIC, we found that the degree of synergy significantly increased for isolates with lower meropenem (and not ertapenem) MICs, up to an MIC of 128 mg/L. Among colistin-containing combinations, synergy was observed in 18/39 (46.1%), 33/34 (97%), 24/39 (61.5%) and 17/39 (43.5%) of the strains for colistin ! meropenem, colistin ! rifampicin, colistin ! tigecycline and colistin ! doripenem, respectively, including colistinresistant strains. Colistin ! meropenem ! tigecycline at subinhibitory concentrations resulted in the absence of growth of 37/39 strains (94.8%). Conclusions: Our in vitro data suggest that colistin might be a valid therapeutic option against CR-Kp, even in the presence of colistin resistance, whereas the double-carbapenem regimen represents a viable option when colistin is not recommended, especially if the meropenem MIC is 128 mg/L. Since traditional antimicrobial susceptibility reports are not sufficiently informative for clinicians, synergy testing as well as actual meropenem MIC evaluation should always be performed in the case of CR-Kp infections.