Evaluation of five carbapenemase detection assays for Enterobacteriaceae harbouring blaKPC variants associated with ceftazidime/avibactam resistance

Sir, Ceftazidime/avibactam is an antimicrobial combination of a third-generation cephalosporin associated with a novel non-β-lactam β-lactamase inhibitor. Ceftazidime/avibactam exerts potent in vitro activity against class A, class C and some class D β-lactamases, while it is ineffective against class B carbapenemases (i.e. MBLs). In this context, rapid detection and characterization of carbapenemase type could guide clinicians regarding optimal antimicrobial regimen, prevent and limit diffusion of MDR pathogens and have an epidemiological purpose. In recent years, several phenotypic and molecular tests have been proposed to identify carbapenemase type in a shorter time than conventional diagnostic approaches.1 Although carbapenemase detection assays exhibit high specificity and sensitivity to detect KPC producers,1–3 recent studies reported failure to detect uncommon blaKPC variants associated with ceftazidime/avibactam resistance.4,5 In the present study, we aimed to evaluate the incidence rate of ceftazidime/avibactam resistance among carbapenemase-producing Enterobacteriaceae (CPE) strains in three Italian hospitals located in the province of Bologna and the performance of five methods for carbapenemase detection in ceftazidime/avibactam-resistant KPC-producing clinical isolates. Between 1 February 2019 and 1 August 2019, we collected all potential CPE from clinical samples and rectal swabs for active surveillance screening.6 Antimicrobial susceptibility testing was performed using the MicroScan WalkAway System (Beckman, USA) and MIC values of ceftazidime/avibactam were confirmed by MIC test strip (Liofilchem, Italy) and results were interpreted following EUCAST clinical breakpoints (Version 9.0). WGS of KPC-producing Klebsiella pneumoniae (KPC-Kp) strains was performed to identify the molecular mechanism of ceftazidime/avibactam resistance. Briefly, genomic DNA was sequenced using the MiSeq platform (Illumina, USA) and antimicrobial resistance genes and STs were investigated as previously described.7 Allele frequency was assessed by comparing the ratio between number of reads with mutant alleles to total reads mapped against reference blaKPC-3.7 In order to evaluate the sensitivity of the carbapenemase detection assays, we included six ceftazidime/avibactam-resistant KPC-Kp clinical strains harbouring different blaKPC variants described in previous studies.7–9 Carbapenemase production was investigated by MALDI-TOF MS specific-peak (11.109 m/z) detection analysis,10 the molecular assay Xpert Carba-R (Cepheid, USA), the immunochromatographic assay NG-Test CARBA 5 (NG Biotech, France), the Neo-Rapid Carb Screen test (Rosco Diagnostica, Denmark) and the Disc Diffusion Synergy test (Rosco Diagnostica, Denmark). All strains producing MBLs were excluded from the study. During the study period, we collected a total of 788 CPE strains producing class A or class D carbapenemases. In detail, 601 strains (582 KPC and 19 OXA-48 producers) were isolated from rectal swabs and 187 (181 KPC and 6 OXA-48 producers) from clinical samples. Confirmatory analysis demonstrated that five CPE strains were resistant to ceftazidime/avibactam, thus resulting in a prevalence of 0.63% among the total of KPC and OXA-48 producers. Of note, all ceftazidime/avibactam-resistant strains were KPC producers and most of them were isolated from patients treated with ceftazidime/avibactam-based therapies (BOT-CA-R, BOT-EMO, 27318, KpBO28, KpBO29, KpBO30 and KpBO32). A summary of the phenotypic and genotypic characteristics of the ceftazidime/avibactam-resistant KPC-Kp strains included in this study are shown in Table S1 (available as Supplementary data at JAC Online). Interestingly, specific mutations within blaKPC were found in 9 out of 11 (81.8%) ceftazidime/avibactam-resistant KPC-Kp strains, thus demonstrating that mutated KPC is the main mechanism responsible for ceftazidime/avibactam resistance in our setting. Evaluation of carbapenemase detection assays showed that all ceftazidime/avibactam-resistant strains were correctly identified as KPC producers by MALDI-TOF MS specific-peak detection analysis and the molecular assay Xpert Carba-R, while negative results were observed for six out of nine (66.6%) mutated KPC-Kp strains by the immunochromatographic assay NG-Test CARBA 5, the Neo-Rapid Carb Screen test and the Disc Diffusion Synergy test (Table 1). Interestingly, all these strains possessed a single amino acid mutation within the Ω-loop of KPC. In particular, five CPE strains harboured a blaKPC-31 variant (D179Y substitution) and one harboured a blaKPC-39 variant (A172T). In contrast, positive results were observed for CPE strains harbouring blaKPC-3, blaKPC-33 and blaKPC-36 variants, as well as one blaKPC-31 variant. Phenotypic and genotypic characteristics of KPC-Kp strains resistant to ceftazidime/avibactam and carbapenemase detection results of different assays ETP, ertapenem; IPM, imipenem; MEM, meropenem; CAZ/AVI, ceftazidime/avibactam; POS, positive; NEG, negative. Positive for class A carbapenemase. Percentage of mutated allele. Phenotypic and genotypic characteristics of KPC-Kp strains resistant to ceftazidime/avibactam and carbapenemase detection results of different assays ETP, ertapenem; IPM, imipenem; MEM, meropenem; CAZ/AVI, ceftazidime/avibactam; POS, positive; NEG, negative. Positive for class A carbapenemase. Percentage of mutated allele. Characterization of ceftazidime/avibactam-resistant strains demonstrated that all KPC-Kp strains with mutated KPC with allele frequency >90% exhibited significant reduction of MICs of carbapenems, especially MICs of imipenem and meropenem, and gave negative results by phenotypic assays (i.e. NG-Test CARBA 5, Neo-Rapid Carb Screen and Disc Diffusion Synergy). At the same time, KPC-Kp isolates harbouring mixed subpopulations composed of WT blaKPC-3 and mutated blaKPC-3 with an allele frequency of ∼50% showed high carbapenem MICs and gave positive results by phenotypic tests (Table 1). Of note, all ceftazidime/avibactam-resistant KPC-Kp strains grew on selective medium CHROMagar-KPC. Our study demonstrated that false-negative detection of KPC production by the immunochromatographic assay NG-Test CARBA 5, the Neo-Rapid Carb Screen test and the Disc Diffusion Synergy test occurred in KPC-Kp strains with subpopulations harbouring specific mutations within the Ω-loop region of the blaKPC gene and conferring resistance to ceftazidime/avibactam. These mutations are commonly associated with reduced carbapenem MICs, mainly due to low carbapenemase activity, thus explaining the false-negative results by the Disc Diffusion Synergy test and the Neo-Rapid Carb Screen test. In addition, we hypothesized that the false-negative results by the immunochromatographic assay NG-Test CARBA 5 could be related to low binding affinity to mutated KPC enzyme. In conclusion, we report that false-negative results can occur with carbapenemase detection assays when testing ceftazidime/avibactam-resistant KPC producers and that these assays must be considered carefully in CPE detection workflow to limit the diffusion of ceftazidime/avibactam-resistant strains and because of the potential impact of negative results in the choice of ceftazidime/avibactam-based therapy. We would like to thank the bacteriologists involved in isolate collection. This work was supported by the Italian Ministry of Health (Ricerca Finalizzata to P.G., Giovani Ricercatori, GR-2018-12367572). None to declare.