Abstract
Carbapenem-resistant Klebsiella pneumoniae (CRKP) has become a serious challenge for clinical treatment and public health. We found that both KPC-2-producing K. pneumoniae (KPC-KP) and NDM-1-producing K. pneumoniae (NDM-KP) are epidemic in a teaching hospital in Northeast China. The main aim of the present study was to compare antimicrobial susceptibility differences between KPC-KP and NDM-KP and elucidate complex resistant genotypes of the KPC-KP and NDM-KP by PCR and sequencing. Among 82 CRKP isolated between January 2015 and December 2016, 59 isolates were KPC-KP and 23 isolates were NDM-KP. All 59 KPC-KP had no susceptibility to gentamicin, tobramycin, levofloxacin, and ciprofloxacin, had very low susceptibility to amikacin (3.39%) and fosfomycin (8.47%), whereas the susceptibility of NDM-KP to the above antibiotics was 21.74%, 13.04%, 17.39%, 17.39%, 69.57%, and 73.91%, respectively. Although the susceptibility of NDM-KP to tigecycline (95.65%) and polymyxin B (73.91%) was higher than that of KPC-KP (84.75% and 69.49%, respectively), the difference was not statistically significant. The MIC90 of KPC-KP and NDM-KP to aztreonam-avibactam were 4 and 2 μg/mL, respectively. All 82 CRKP carried 2 or 3 Extended Spectrum Beta-Lactamase (ESBL) genes, and 79/82 CRKP carried the AmpC gene blaFOX. The aminoglycoside resistance gene rmtB was detected in 96.61% of KPC-KP and in 21.74% of NDM-KP. It seems that KPC-KP was more resistant to antibiotics than NDM-KP in this study, so that available therapeutic regimens against KPC-KP are very limited. Aztreonam-avibactam may be a promising and valuable option against both KPC-KP and NDM-KP.
Introduction
K
Tigecycline, polymyxins (polymyxin B, colistin), and fosfomycin have been reported as alternative choices for the treatment of carbapenem-resistant enterobacteriaceae (CRE).6–9 In addition, avibactam (also named NXL-104), a new β-lactamase inhibitor, combined with ceftazidime or aztreonam was active against carbapenemase-producing isolates. 10
The reports of KPC-producing K. pneumoniae (KPC-KP) or NDM-producing K. pneumoniae (NDM-KP) have markedly increased in China in recent years.11,12 Due to the vast territory in China, the epidemic characteristics of carbapenem-resistant K. pneumoniae (CRKP) have significant regional differences. We found that both KPC-KP and NDM-KP are epidemic in our hospital located in Dalian, a coastal tourist city with high population mobility. The aims of this study were to clarify the epidemic characteristics and resistance mechanism of CRKP in this area and further compare the differences in antimicrobial susceptibility between the KPC-KP and NDM-KP isolates, elucidating the underlying resistant genotypes. A better understanding of the antimicrobial susceptibility patterns of KPC-KP and NDM-KP will be helpful for choosing effective therapeutic regimens.
Materials and Methods
Bacterial strains and antimicrobial susceptibility testing
Bacterial isolates were collected between 2015 and 2016 from a 3,700-bed teaching hospital located in Dalian in Northeast China. Isolates identification and antimicrobial susceptibility testing (AST) were performed by an automated MicroScan WalkAway 96 plus system. E. coli ATCC 25922 and P. aeruginosa ATCC27853 were used as quality controls. AST results were interpreted according to the Clinical Laboratory Standards Institute (CLSI) guidelines (2014).
Determination of susceptibility of tigecycline, polymyxin B, aztreonam, aztreonam-avibactam, and fosfomycin
The MICs of aztreonam (Solarbio® Life Science, China), tigecycline (Pfizer) or polymyxin B (Solarbio Life Science, China) were determined by the broth microdilution method according to CLSI guidelines. The susceptibility results for tigecycline and polymyxin B were interpreted according to EUCAST (European Committee on Antimicrobial Susceptibility Testing) breakpoints (available at www.eucast.org). The breakpoints of tigecycline were as follows: susceptible, MIC ≤1; intermediate, MIC = 2; and resistant, MIC ≥4. For polymyxin B (referring to colistin), isolates with MICs of ≤2 μg/mL were categorized as susceptible, and those with MICs of >2 μg/mL were categorized as resistant. The susceptible result for aztreonam was interpreted according to CLSI guidelines.
The in vitro activities of aztreonam over a range of dilutions from 0.25 to 1,024 μg/mL combined with avibactam (Meilunbio®, China) at a fixed concentration of 4 μg/mL were determined by the broth microdilution method.
Susceptibility testing of fosfomycin (Oxoid) was determined by the Kirby-Bauer disc paper method, and isolates with an inhibition zone diameter ≥16 mm were categorized as susceptible according to CLSI guidelines.
Detection of drug-resistant genes and sequencing
Carbapenem-resistant genes were detected by PCR, including Ambler class A (blaNMC, blaSME, blaIMI, blaKPC, and blaGES), class B (blaNDM-1, blaIMP-1, blaIMP-2, blaVIM-1, blaVIM-2, blaSPM-1, blaGIM-1, and blaSIM-1), and class D (blaOXA-48). We also screened for common Extended Spectrum Beta-Lactamase (ESBL) genes (blaCTX-M, blaTEM, and blaSHV), AmpC genes (blaMOX, blaFOX, blaDHA, blaCIT, and blaEBC), plasmid-mediated quinolone resistance genes (qnrA, qnrB, qnrS, and aac(6′)-Ib-cr), and aminoglycoside resistance genes (16S rRNA methyltransferase genes: armA, rmtB, and rmtC) in the CRKP isolates. The primers used in this study were described previously (Supplementary Table S1).5,13–17 The positive PCR products were sequenced by Sangon Biotech (Shanghai) Co., Ltd.
Statistical analysis
Antimicrobial susceptibility differences between KPC-KP and NDM-KP isolates were determined by χ 2 test. p < 0.05 was considered statistically significant.
Strain genotyping by pulsed-field gel electrophoresis and multilocus sequence typing
The genetic relatedness of KPC-KP and NDM-KP isolates with a representative resistance phenotype was assessed by pulsed-field gel electrophoresis (PFGE) analysis with XbaI-digested genomic DNA as described previously. 18 Multilocus sequence typing (MLST) was performed with seven standard housekeeping loci (gapA, infB, mdh, pgi, phoE, rpoB, and tonB) according to a previous description. 19 Sequence types were analyzed by using the Institute Pasteur database.*
Results
Clinical features of carbapenem-resistant K. pneumoniae
A total of 82 CRKP isolates were isolated between January 2015 and December 2016 in a teaching hospital in Northeast China. The CRKP isolates were isolated from different sites of infection, including blood (25.61%), urine (29.27%), tracheal aspirate (21.95%), bile (2.44%), pus (10.98%), pleural effusion (6.10%), and catheter tip (3.66%). The CRKP isolates were mainly from ICU (intensive care unit, 46.34%) and EICU (emergency intensive care unit, 42.68%).
We detected carbapenem-resistant genes, as mentioned in the Materials and Methods section, among 82 CRKP isolates. We found 59 isolates to be KPC-2 carbapenemase producers and 23 isolates to be NDM-1 producers. Other carbapenem-resistant genes such as blaIMP, blaVIM, and blaOXA-48 were not found. The top KPC-KP infection sites were urine (32.20%), blood (22.03%), and tracheal aspirate (18.64%), while the NDM-KP isolates were mainly isolated from blood (34.78%), tracheal aspirate (30.43%), and urine (21.74%).
Temporal distribution of CRKP
A total of 56 CRKP isolates were identified in 2015, including 44 KPC-KP and 12 NDM-KP. In 2016, we identified 26 CRKP isolates, which included 15 KPC-KP and 11 NDM-KP. KPC-KP was the prevalent carbapenem-resistant isolate in the hospital from 2015 to 2016. As shown in Fig. 1A, we found that KPC-KP caused hospital outbreaks in January, April, June, and July in 2015, and in October in 2016. However, NDM-KP showed a sporadic characteristic and had a small-scale epidemic from June to July in 2015 and in March in 2016.

Temporal distribution and genotyping of the KPC-KP and NDM-KP isolates.
Antimicrobial susceptibility patterns
All 82 CRKP isolates were resistant to cephalosporins (100%), aztreonam (100%), piperacillin/tazobactam (100%), ertapenem (100%), meropenem (100%), and imipenem (100%). These isolates had low susceptibility to ciprofloxacin (4.88%), levofloxacin (4.88%), gentamicin (6.10%), tobramycin (3.66%), amikacin (21.95%), tetracycline (28.05%), fosfomycin (26.83%), and trimethoprim/sulfamethoxazole (36.59%). However, relatively high susceptibility to tigecycline (87.80%) and polymyxin B (70.73%) was observed (Table 1).
Antimicrobial Susceptibility Differences Between KPC-KP and NDM-KP
AMK, amikacin; CIP, ciprofloxacin; CRKP, carbapenem-resistant Klebsiella pneumonia; FOS, fosfomycin; GEN, gentamicin;
Through further analysis, we found significant differences between the susceptibilities of the KPC-KP and NDM-KP isolates to antibiotics. All 59 KPC-KP isolates had no susceptibility to gentamicin, tobramycin, levofloxacin, and ciprofloxacin, whereas the susceptibility of NDM-KP to the above antibiotics was 21.74%, 13.04%, 17.39%, and 17.39%, respectively. The KPC-KP isolates had lower susceptibility to amikacin (3.39%) and fosfomycin (8.47%) and had slightly higher susceptibility to trimethoprim/sulfamethoxazole (49.15%) and tetracycline (35.59%). Conversely, the NDM-KP isolates had a higher susceptibility to amikacin (69.57%) and fosfomycin (73.91%) and a lower susceptibility to trimethoprim/sulfamethoxazole (4.35%) and tetracycline (8.70%). All of the above differences were statistically significant (p < 0.05) (Table 1). The MIC50 and MIC90 of the CRKP isolates to tigecycline were 1 and 2 μg/mL, respectively. The MIC50 and MIC90 of the KPC-KP isolates to tigecycline were 1 and 2 μg/mL, respectively, while the MIC50 and MIC90 of the NDM-KP isolates to tigecycline were 1 and 1 μg/mL, respectively. The MIC50 and MIC90 of the CRKP isolates to polymyxin B were 2 and 4 μg/mL, respectively. The MIC50 and MIC90 of the KPC-KP isolates to polymyxin B were 2 and 8 μg/mL, respectively, while the MIC50 and MIC90 of the NDM-KP isolates to polymyxin B were 2 and 4 μg/mL, respectively. Although the susceptibility of the NDM-KP isolates to tigecycline and polymyxin B was higher than that of the KPC-KP isolates, the difference was not statistically significant (p > 0.05) (Table 1).
The MIC90 of the CRKP to aztreonam was 512 μg/mL. Our results show that 97.56% (80/82) of all the CRKP were inhibited by aztreonam-avibactam at a concentration of 8 μg/mL. The MIC50 and MIC90 of the KPC-KP isolates to aztreonam-avibactam were 2 and 4 μg/mL, respectively, and the MIC50 and MIC90 of the NDM-KP isolates to aztreonam-avibactam were 0.5 and 2 μg/mL, respectively. Avibactam dramatically lowered the MIC90 of aztreonam for the KPC-KP and NDM-KP isolates by 128- and 256-fold, respectively (Table 2).
Minimum Inhibitory Concentration Distributions for Aztreonam-Avibactam Against Carbapenemase-Producing Isolates
Fold reduction in comparison to the aztreonam MIC90 tested alone.
Drug resistance genes of the CRKP isolates
To explore the complex resistant genotypes, the ESBLs, AmpC, quinolones, and aminoglycoside resistance genes were detected among CRKP isolates. We found that all 82 CRKP isolates carried at least 2 ESBL genes (blaTEM and blaSHV). There were 33 CRKP isolates harboring 3 ESBL genes (blaTEM, blaSHV, and blaCTX-M). The gene blaFOX, a kind of AmpC, was detected in 79 of 82 CRKP isolates (Table 3). The sequencing results showed that TEM-1 and SHV-12 were the main types identified in all the CRKP isolates. The blaCTX-M gene was detected in 27.12% (16/59) of the KPC-KP isolates and 73.91% (17/23) of the NDM-KP isolates. Among 16 blaCTX-M-positive KPC-KP, 14 isolates (87.50%) harbored CTX-M-14, and 2 isolates (12.50%) harbored CTX-M-15. Among the 17 blaCTX-M-positive NDM-KP, 10 isolates (58.82%) harbored CTX-M-3, 4 isolates (23.53%) harbored CTX-M-14, and 3 isolates (17.65%) harbored CTX-M-15.
Detection of ESBLs, AmpC, Quinolinones, and Aminoglycosides Resistance Genes
The qnrB, qnrS, and aac(6′)-Ib-cr genes were detected in 1.69% (1/59), 54.24% (32/59), and 32.20% (19/59) of the KPC-KP isolates, respectively, and in 21.74% (5/23), 21.74% (5/23), and 34.78% (8/23) of the NDM-KP isolates, respectively. The aminoglycoside resistance gene rmtB was detected in 96.61% (57/59) of the KPC-KP and in 21.74% (5/23) of the NDM-KP. The armA gene was detected only in 4 (17.39%) of the 23 NDM-KP (Table 3). PCR results of resistance genes were shown in Supplementary Figure S1.
Genotyping of the KPC-KP and NDM-KP isolates
To identify the genetic relatedness of the KPC-KP and NDM-KP isolates, PFGE and MLST were performed among the strains with a representative resistance phenotype (Fig. 1B). MLST analysis showed that eight of the nine KPC-KP isolates belonged to ST11, and only CRKP55 belonged to ST1236. The NDM-KP isolates belonged to different ST types, and three new ST types were found, named STnew1, STnew2, and STnew3. PFGE analysis indicated that most of the KPC-KP isolates were highly homologous, while the NDM-KP isolates showed relatively low homology.
Discussion
Although KPC-KP and NDM-KP have been reported worldwide, both KPC-KP and NDM-KP were prevalent in this hospital. KPC-KP isolates were the main epidemic strain with high homology and caused several outbreaks from January 2015 to December 2016, whereas the NDM-KP isolates with relatively low homology showed a sporadic characteristic in this period. Notably, comparisons of antimicrobial susceptibility differences between the KPC-KP and NDM-KP isolates in the same hospital are rarely reported. In the present study, quinolones (LEV and CIP), aminoglycosides (AMK, GEN, and TOB), and fosfomycin were almost ineffective against the KPC-KP isolates from our hospital. High resistance of the KPC-KP isolates to quinolones appeared to be associated with a high prevalence of the plasmid-mediated quinolone resistance determinants qnrS (54.24%) and aac(6′)-Ib-cr (32.20%). The high prevalence of the 16S rRNA methyltransferase gene rmtB (96.61%) in the KPC-KP isolates might be responsible for high aminoglycoside resistance. The high susceptibility of the NDM-KP isolates to AMK (69.57%) and fosfomycin (73.91%) indicated that AMK and fosfomycin can be considered as options to treat NDM-KP-causing infection most of the time. It has been reported that fosfomycin could be used in combination with other antimicrobials for the treatment of infections caused by CRE, particularly those involving urinary tract infections. 20 A previous study suggested that fosfomycin could be combined with carbapenems or aminoglycosides, and these two kinds of combinations can be synergistic against CRKP 9 or KPC-KP. 21 Moreover, an in vivo study showed that fosfomycin might have a protective effect on aminoglycoside-induced nephrotoxics. 22 However, our study indicated that fosfomycin was almost not effective against the KPC-KP isolates in this hospital. A better understanding of the antimicrobial susceptibility patterns of KPC-KP and NDM-KP would be helpful for clinical treatment, especially for choosing effective antibiotics. Although rapid carbapenemase gene detection, such as multiplex real-time PCR, 23 was not performed in the clinical laboratory in our hospital, it is necessary to continue to assess the susceptibility patterns of KPC-KP and NDM-KP among newly isolated CRKP. If a susceptibility difference still exists between KPC-KP and NDM-KP isolates, rapid carbapenemase gene detection should be performed, which would be beneficial for severely infected patients.
In this study, both KPC-KP and NDM-KP maintained relatively high susceptibility to tigecycline and polymyxin B. Tigecycline and polymyxins are two of the few antimicrobials that still retain activity against CRKP and make a key component of anti-CRKP regimens. Based on currently available data, many studies recommend using combination therapy instead of monotherapy in CRKP-infected patients. Qureshi et al. reported that combination therapy, in particular with either polymyxins (polymyxin B or colistin) or tigecycline and a carbapenem, seems to improve the survival of patients with bacteremia caused by KPC-KP. 24
Avibactam, a non-β-lactam β-lactamase inhibitor, is capable of inhibiting Ambler class A, class C, and some class D β-lactamases, including KPC and OXA-48 type carbapenemase.25,26 Avibactam in combination with ceftazidime, which was approved by the U.S. Food and Drug Administration (FDA) in 2015, represents an important advancement to treat infections caused by CRE.27,28 However, the ceftazidime-avibactam combination does not inhibit MBL-producing isolates. A unique feature of aztreonam is its resistance to hydrolysis by MBLs. It has been reported that aztreonam-avibactam was active against CRE, including isolates with Ambler class A, class B, and class D carbapenemases. Furthermore, the antimicrobial activity of aztreonam-avibactam is currently under clinical development.29,30 However, the activity of aztreonam-avibactam against CRKP is very rarely reported in China. The only study is that by Wang et al., which reported that in vitro activity of aztreonam-avibactam only included 15 meropenem-nonsusceptible K. pneumoniae. 31 In this work, we assessed the in vitro activity of aztreonam-avibactam against both KPC-KP and NDM-KP isolates. The MIC50 and MIC90 of the NDM-KP isolates to aztreonam-avibactam were lower than those of the KPC-KP isolates. The above result indicated that NDM-KP seemed to be more susceptible to aztreonam-avibactam compared with KPC-KP. Although the resistant genotypes of CRKP were very complex in this study, all 82 CRKP isolates harbored at least 2 or 3 ESBL genes, and 79 of the 82 CRKP isolates harbored the AmpC gene blaFOX, aztreonam combined with avibactam still showed a very potent inhibitory effect on both KPC-KP and NDM-KP isolates.
Conclusions
Currently, available therapeutic regimens to treat infections caused by carbapenem-resistant K. pneumoniae are very limited. Clinicians can only rely on the antimicrobial drugs available today until new antimicrobial agents come on the market. Two or even three antibiotic combinations among tigecycline, colistin, aminoglycosides, and fosfomycin seem to confer decreased mortality. Notably, due to almost full resistance to aminoglycosides and fosfomycin of the KPC-KP isolates in this study, there seem to be fewer choices against KPC-KP-causing infection. The results on the in vitro activity of aztreonam-avibactam presented here and other currently available data indicate that aztreonam-avibactam may be an expectant and valuable option for treating infections caused by KPC-KP or NDM-KP.
Footnotes
References
Supplementary Material
Please find the following supplemental material available below.
For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.
For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.
