Impact of ompk36 genotype and KPC subtype on the in vitro activity of ceftazidime/avibactam, imipenem/relebactam and meropenem/vaborbactam against KPC-producing K. pneumoniae clinical isolates

Abstract

Objectives: The availability of new β-lactam/β-lactamase inhibitors ceftazidime/avibactam, meropenem/vaborbactam and imipenem/relebactam have redefined contemporary treatment of Klebsiella pneumoniae carbapenemase-producing Klebsiella pneumoniae (KPC-Kp) infections. We aimed to characterize and contrast the in vitro activity of these agents against genetically diverse KPC-Kp clinical isolates.

Methods: We analysed genomes of 104 non-consecutive KPC-Kp isolates and compared the in vitro antibiotic activity by KPC subtype and ompK36 genotype. MICs were determined in triplicate by CLSI methods. Twenty representative isolates were selected for time-kill analyses against physiological steady-state and trough concentrations, as well as 4× MIC for each agent.

Results: Fifty-eight percent and 42% of isolates harboured KPC-2 and KPC-3, respectively. OmpK36 mutations were more common among KPC-2- compared with KPC-3-producing Kp (P < 0.0001); mutations were classified as IS5 insertion, glycine-aspartic acid insertion at position 134 (GD duplication) and other mutations. Compared to isolates with WT ompK36, ceftazidime/avibactam, imipenem/relebactam and meropenem/vaborbactam MICs were elevated for isolates with IS5 by 2-, 4- and 16-fold, respectively (P < 0.05 for each). Against isolates with GD duplication, imipenem/relebactam and meropenem/vaborbactam MICs were increased, but ceftazidime/avibactam MICs were not. In time-kill studies, ceftazidime/avibactam-mediated killing correlated with ceftazidime/avibactam MICs, and did not vary across ompK36 genotypes. Imipenem/relebactam was not bactericidal against any isolate at trough concentrations. At steady-state imipenem/relebactam concentrations, regrowth occurred more commonly for isolates with IS5 mutations. Log-kills were lower in the presence of meropenem/vaborbactam for isolates with GD duplication compared with IS5 mutations.

Conclusions: Our investigation identified key genotypes that attenuate, to varying degrees, the in vitro activity for each of the new β-lactam/β-lactamase inhibitors. Additional studies are needed to translate the importance of these observations into clinical practice.

Figure 1.

Distribution of ompK36 genotypes stratified by KPC subtype. Mutations described as ‘other’ are reported in Table S1.

Figure 2.

MIC distribution for each β-lactam and β-lactam/β-lactamase inhibitor agent tested against KPC-Kp. Median ceftazidime, imipenem and meropenem MICs were reduced by 256-, 32- and 512-fold with the addition of avibactam, relebactam and vaborbactam, respectively. CAZ, ceftazidime; CZA, ceftazidime/avibactam; IPM, imipenem; I-R, imipenem/relebactam; MEM, meropenem; MVB, meropenem/vaborbactam.

Figure 3.

Comparison of ceftazidime/avibactam, imipenem/relebactam and meropenem/vaborbactam MICs stratified by ompK36 genotype. Dotted lines represent CLSI susceptible and resistance breakpoints for each agent. Solid horizontal lines represent the median MIC for each group of isolates. * denotes P < 0.05. CZA, ceftazidime/avibactam; I-R, imipenem/relebactam; MVB, meropenem/vaborbactam.

Figure 4.

Comparison of ceftazidime/avibactam, imipenem/relebactam and meropenem/vaborbactam MICs stratified by KPC subtype and ompK36 genotype. Dotted lines represent CLSI susceptible and resistance breakpoints for each agent. Horizontal lines represents the median for each group of isolates. * denotes P < 0.05. CZA, ceftazidime/avibactam; I-R, imipenem/relebactam; MUT, mutant ompK36 genotype; MVB, meropenem/vaborbactam.

Figure 5.

24 h log–kills across KPC-Kp stratified by ceftazidime/avibactam, imipenem/relebactam and meropenem/vaborbactam MICs. 24 h log-kills were calculated as the difference in cfu/mL at 0 and 24 h and plotted according to the baseline MIC for each isolate. Dotted lines indicate the best fit linear regression line. CZA, ceftazidime/avibactam; I-R, imipenem/relebactam; MVB, meropenem/vaborbactam; 4×, 4× MIC; Tr, simulated trough concentration; Ss, simulated steady-state concentration.

Figure 6.

Mean log-kills across KPC-Kp exposed to various concentrations of ceftazidime/avibactam, imipenem/relebactam and meropenem/vaborbactam. Individual graphs demonstrate the mean log-kills at each time point over the 24 h incubation period. Error bars show the standard deviation. CZA, ceftazidime/avibactam; I-R, imipenem/relebactam; MVB, meropenem/vaborbactam; 4×, 4× MIC; Tr, simulated trough concentration; Ss,  simulated steady-state concentration.