Could Frequent Carbapenem Use Be a Risk Factor for Colistin Resistance?

Abstract

Aims:

The antibiotic colistin, which had been previously abandoned, is being brought back as a last line of defense against bacterial infection. However, colistin resistance was reported shortly after its reintroduction. This study evaluated the risk factors for colonization/infections due to colistin-resistant Acinetobacter baumannii (ColR-Ab) and Klebsiella pneumoniae (ColR-Kp) strains and characterized the molecular epidemiology of these two strains.

Results:

Age, previous hospitalization duration, and previous use of carbapenem and colistin were risk factors for ColR-Kp, whereas previous use of carbapenem and colistin was a risk factor for ColR-Ab. According to pulsed-field gel electrophoresis analysis, most ColR-Kp strains could be grouped into two major pulsotypes. This appears to be an indicator of cross contamination of ColR-Kp strain, since different isolates appeared to be belonging to the same clones. The existence of colistin-susceptible (ColS) and colistin-resistant (ColR) strains in the same pulsotypes might also be an indicator of the recent emergence of resistance mechanisms.

Conclusions:

The results highlight the emergence of ColR pathogens in Turkey, which is considered to be developing country, and that carbapenem use coupled with insufficient infection control measures might increase the risk of ColR outbreaks.

Introduction

After their discovery in 1947, polymyxins were delivered parenterally to treat infections caused by Gram-negative bacteria. Dueto its high nephrotoxicity, use of colistin eventually fell out of favor except for a few rare instances. The increasing prevalence of infections caused by multidrug-resistant Gram-negative bacteria raises interest in returning colistin to the market as a last line of defense.¹ Colistin is now used as a drug of last resort for cases that cannot be cured by standard antibiotic therapy. However, after colistin was reintroduced, colistin resistance has been seen in clinics.^2–4

The emergence of colistin-resistant (ColR) strains will further limit antibiotic treatment options to fight bacterial infections. Among ColR bacteria, Acinetobacter baumannii (ColR-Ab) and ColR Klebsiella pneumoniae (ColR-Kp) are of crucial importance as they are major hospital-associated pathogens.^5,6 In susceptible individuals, these strains can cause a range of infections, such as pneumonia, bacteremia, meningitis, and blood stream infections, all of which have limited treatment options. Being major nosocomial pathogens, infections caused by ColR-Ab and ColR-Kp can lead to high morbidity and mortality rates.^6–8 As the dissemination of these strains raises a global risk for pan-drug-resistant strains, surveillance studies of ColR strains and examination of the molecular characteristics of these bacteria, and the epidemiology of patients infected with them, are critical. Despite the importance of ColR, few such studies, especially in developing countries, have been conducted. Here, we aimed to (i) evaluate the epidemiological characteristics of at-risk patients and risk factors for colonization or infections due to ColR-Ab and ColR-Kp; and (ii) determine the molecular epidemiology of these strains.

Material and Methods

Setting

Medical Faculty of Erciyes University Hospital is a 1,300 bed tertiary-referral teaching hospital in the urban setting of Kayseri (Turkey) that has a population of 1,000,000. Approximately 910,000 patients (adult and pediatric) are admitted to the hospital annually. Of these, 166,000 are hospitalized for an average stay of 2.8 days. As part of infection control measures, each patient is evaluated for risk factors for colonization with resistant pathogens upon admission. Surveillance culture specimens are obtained from those patients found to be at risk. Routine surveillance for carbapenem-resistant K. pneumoniae is also performed for intensive care unit (ICU) patients.

Identification and susceptibility testing of clinical isolates

Two groups of clinical isolates from a case and control group were included. The case group consisted of adult and pediatric patients colonized or infected with ColR-Ab (n = 60) or ColR-Kp (n = 51) between 2011 and 2016. A control group of 120 patients infected with colistin-susceptible (ColS) A. baumannii (n = 60) or ColS-Kp (n = 60) who were hospitalized during the same period was also selected. A total of 231 A. baumannii (n = 120) and K. pneumoniae (n = 111) strains isolated from different body fluids from all 231 patients were included in this part of the study. Identification and antimicrobial susceptibility testing of each isolatewere conducted following conventional methods using a Vitek-2 automated system (BioMérieux, Durham NC). A colistin minimum inhibitory concentration (MIC) was determined using the Vitek-2 and was confirmed by E-test. Data interpretation was done in accordance with Clinical and Laboratory Standards Institute (CLSI) break points.⁹ One isolate per patient per episode was included in the study.

Demographic data for case and control group patients

To conduct a risk analysis, all case group (n = 111) and control group (n = 120) patients were compared interms of epidemiological variables. Epidemiological data for all study subjects were retrospectively collected from hospital e-records gathered by the hospital's infection control committee and from patient files. Variables such as age, hospitalization duration before isolation of ColR and ColS strains, whether the patient was transferred from another unit or hospital, underlying medical conditions, antimicrobial agents given before isolation of pathogens and ICU admission were collected. Risk analyses were conducted to determine the potential risk factors for infection or colonization with ColR strains.

Microbiological and molecular assays for the collected strains

A total of 43 ColR (37ColR-Kp, 6 ColR-Ab) and 3 ColS-Kp strains collected between March 2014 and 2016 were assesed in microbiological and molecular studies. In addition to the above-mentioned tests, MIC values for colistin for the 43 ColR strains were confirmed by the broth microdilution method. The results were interpreted according to CLSI susceptibility break points.⁹ Antimicrobial susceptibility testing was performed using the disk diffusion method according to CLSI guidelines.⁹ The following antimicrobial disks were used: piperacillin-tazobactam, ceftazidime, cefepime, aztreonam, imipenem, meropenem, amikacin, gentamicin, ciprofloxacin, levofloxacin, tetracycline, tigecycline, and trimethoprim-sulfamethoxazole. The presence of mcr-I and mcr-II genes was assessed by PCR.^4,10 Genetic relatedness of the isolates was determined by pulsed-field gel electrophoresis (PFGE) using a CHEF DR III system (Bio-Rad, Hercules, CA). SpeI and ApaI restriction enzymes were used for pulsotyping of K. pneumoniae and A. baumannii, respectively. Three ColS-Kp isolated from the same patients before the isolation of ColR pathogens were also included in the PFGE study. The PFGE data were interpreted according to the criteria proposed by Tenover et al.¹¹ Genetic relatedness among the isolates was assessed by the unweighted-pair group method using Bionumerics version 6.01 (Applied Maths). The hierarchial clustering cutoff was set to 80% similarity to identify the pulsotypes.

Statistical analysis

Univariate statistical analysis considering patient demographic data was performed on the case and control groups using Pearson's Chi-square test. Age, previous hospitalization duration (in days), transfer from another unit (Boolean variable), transfer from another institute (Boolean variable), underlying diseases (categorical variables), previous use of antibiotics (categorical variables), and clinical specimens (categorical variables), were considered as variables. Significant features were selected using a p < 0.05 significance level. The selected features were then input into a logistic regression model for multivariate analysis. Using step-wise backward elimination, alogistic regression was performed iteratively, again considering a p < 0.05 significance level to eliminate confounding factors or correlated features. The final risk assessment was done by reporting the remaining statistically significant variables. SPSS version 22 was used for all analyses.

Results

A total of 111 patients were infected with or colonized by ColR-Ab (60) or ColR-Kp (51) between 2011 and 2016. Demographic data for these patients and 120 control patients are given in Table 1. According to the results, 50 patients received colistin treatment before the isolation of ColR strains. The most frequent reasons described in the patient files for colistin use were infections caused by carbapenem-resistant A. baumannii (n = 30 patients) and empirical therapy (n = 12 patients) (Table 2). In the case group, 83 out of 111 patients were treated in the ICU. During the study period, 1,813 culture positive episodes were found to be caused by A. baumannii and 3,419 by K. pneumoniae. The rate of ColR-Ab infection or colonization was 3.3%, whereas that for ColR-Kp was 1.5%.

Table 1.

Demographic Data for Case and Control Patients Hospitalized Between 2011 and 2016

Categorical variables	Case group,^a n = 111 (%)	Control group,^b n = 120 (%)
Median age (min–max)	55 (0–86)	47.5 (0–95)
Previous hospitalization days (min–max)	62 (1–144)	53.5 (1–153)
Transfer from another unit	33 (29.7)	37 (30.8)
Transfer from another institution	1 (0.9)	7 (5.8)
Underlying diseases
Diabetes mellitus	16 (14.4)	14 (11.6)
Chronic renal failure	22 (19.8)	14 (11.6)
Chronic heart failure	5 (4.5)	4 (3.3)
COPD	13 (11.7)	11 (9.2)
Previous use of antibiotics
Carbapenems	61 (54.5)	33 (27.5)
Beta-lactams	63 (56.8)	69 (57.5)
Glycopeptides	43 (38.8)	26 (21.6)
Colistin	50 (45.0)	20 (16.7)
Tigecycline	10 (9)	3 (2.5)

Patients with colistin-resistant Acinetobacter baumannii or Klebsiella pneumoniae.

Patients with colistin-susceptible A. baumannii or K. pneumoniae.

COPD, chronic obstructive pulmonary disease.

Table 2.

Reasons for Previous Colistin Use Before Isolation of Colistin-Resistant Strains

Reason for colistin use	ColR A. baumannii, n = 31^a	ColR K. pneumoniae, n = 19^a
Empiric therapy	7	5
Infection caused by carbapenem-resistant
ColS A. Baumannii	22	8
ColS K. pneumonia	0	6
ColS Escherichia coli	1	0
GNB	1	0

The number of patients who recieved colistin treatment before the indicated ColR pathogen isolation.

ColR, colistin-resistant strain; ColS, colistin-susceptible strain; GNB, Gram-negative bacteria.

According to the chi-square test, previous hospitalization, previous colistin use, and previous carbapenem use were found to be significant (p < 0.05) risk factors for ColR-Ab (Table 3). Age, previous hospitalization, previous colistin use, previous carbapenem use, and previous glycopeptide use were among the risk factors for ColR-Kp (Table 4). Multivariate analysis confirmed that age (OR 1.036, p < 0.00), previous colistin (OR 3.114, p = 0.037), and previous carbapenem (OR 2.896, p = 0.040) use were risk factors for ColR K. pneumoniae, whereas previous carbapenem use (OR 2.403, p = 0.045) and previous colistin use were risk factors for A. baumannii (OR 3.367, p = 0.004).

Table 3.

Risk Factor Evaluation for Isolation of Colistin-Resistant A. baumannii from Hospitalized Patients

Categorical variables	ColR, n = 60 (%)	ColS, n = 60 (%)	p^a	Multivariate analysis OR (95% CI), p
Median age (min–max)	53.5 (0–86)	56.5 (0–95)	0.429
Previous hospitalizationdays (min–max)	73 (1–144)	47 (3–88)	0.000
Transfer from another unit	18 (30%)	25	0.253
Transfer from another institution	0	4	0.119
Underlying diseases
Diabetes mellitus	8 (13.3)	11 (18.3)	0.618
Chronic renal failure	13 (21.7)	11 (18.3)	0.820
Chronic heart failure	1 (1.7)	3 (5)	0.619
COPD	8 (21.7)	9 (15)	1.000
Previous antibiotic therapy
Carbapenems	39 (65)	20 (33.3)	0.001	2.403 (1.101–5.312), 0.045
Beta-lactams	32 (53.3)	36 (60)	0.581
Glycopeptides	23 (38.3)	16 (26.7)	0.242
Colistin	31 (51.7)	12 (20)	0.001	3.367 (2.008–6.439), 0.004
Metronidazole	3 (5)	0	0.244
Macrolides	1 (1.7)	3 (5)	0.619
Tigecycline	6 (10)	2 (3.3)	0.163
Linezolide	3 (5)	1 (1.7)	0.364
Clinical specimens
Blood	10 (16.7)	20 (33.3)	0.057
Urine	8 (21.7)	1 (1.7)	0.032
Endotracheal aspirate	10 (16.7)	15 (25)	0.369
Wound	4 (6.7)	5 (8.3)	1.000
Drain	1 (1.7)	2 (3.3)	1.000
Cerebrospinal fluid	1 (1.7)	4 (6.7)	0.364
Bronchoalveolar lavage	11 (18.3)	10 (16.7)	1.000
Others	15 (25)	8 (21.7)	0.496

Bold entries indicate statistically significant (p < 0.05) variables.

Univariate analysis.

Table 4.

Risk Factor Evaluation for Isolation of Colistin-Resistant K. penumoniae from Hospitalized Patients

Categorical variables	ColR, n = 51 (%)	ColS, n = 60 (%)	p^a	Multivariate analysis OR (95% CI), p
Median age (min–max)	59 (19–84)	36 (0–93)	0.020	1.036 (0.427–2.311), <0.00
Previous hospitalization days (min–max)	51 (1–116)	60 (1–153)	0.033
Transfer from another unit	15 (28.8)	12 (20)	0.274
Transfer from another institution	1 (1.9)	3 (5)	0.623
Underlying diseases
Diabetes mellitus	8 (16)	3 (5)	0.252
Chronic renal failure	9 (15.7)	3 (5)	0.620
Chronic heart failure	4 (7.8)	1 (1.7)	0.178
COPD	5 (9.8)	2 (3.3)	0.158
Previous antibiotic therapy
Carbapenems	22 (43.1)	13 (21.7)	0.024	2.896 (1.512–5.817), 0.040
Beta-lactams	31 (60.8)	33 (55)	0.568
Glycopeptides	20 (39.2)	10 (16.7)	0.010
Colistin	19 (37.3)	8 (13.3)	0.004	3.114 (2.005–6.489), 0.037
Tigecycline	4 (7.8)	1 (1.7)	0.178
Clinical specimens
Blood	14 (27.5)	19 (31.7)	0.681
Urine	8 (15.7)	13 (21.7)	0.474
Endotracheal aspirate	5 (9.8)	9 (15)	0.568
Wound	8 (15.7)	8 (13.3)	0.790
Drain	4 (7.8)	2 (3.3)	0.411
Abcess	5 (9.8)	1 (1.7)	0.092
Others	7 (13.7)	8 (13.3)	1.000

Bold entries indicate statistically significant (p < 0.05) variables.

Univariate analysis.

Microbiological analysis of collected strains

All 43 ColR isolates (ColR-Ab, n = 6; ColR-Kp, n = 37) were confirmed as ColR with colistin MICs ranging from 4 to 256 mg/L in broth microdilution. According to the antimicrobial resistance profile, 90.7% (39/43) of the isolates were extensively drug resistant (XDR) showing a lack of susceptibility to ≥1 agent in all, but ≤2 categories.¹² All ColR-Ab isolates were susceptible to tigecycline, whereas 81% of ColR-Kb isolates were susceptible to tigecycline. After tigecycline, gentamicin and amikacin were the most effective drugs for ColR strains, whereas 2/6 ColR-Ab and 18/36 ColR-Kp were resistant to one of the aminoglycosides tested. Nearly all (36/37) ColR K. pneumoniae strains were extended spectrum beta-lactamase producers.

Molecular studies

In addition to all 37 ColR-Kp isolates, PFGE was also performed on 3 ColS-Kp strains, isolated from two different patients before ColR-Kp isolation. Using a cutoff of 80% similarity, the 37 ColR-Kp and 3 ColS-Kp strains were grouped into five common pulsotypes (CTs) and four single pulsotypes (STs). Among the CTs, two presented as major clones with 14 (CT3) and 16 (CT4) strains (Fig. 1). The demographic data associated with these major clones are given in Table 5. The pair of ColS isolates obtained from the same patients (Kp42 and Kp43, Fig. 1) were categorized as CT4. This pair showed 100% PGFE similarity (indistinguishable restriction profiles-0 band difference) with one ColR-Kp (Kp3) strain isolated from the same patient within 3 months. The other ColS-Kp (Kp44) was reported as a single pulsotype. Among ColR strains (Kp36, Kp38, and Kp39) isolated in 2016, 3/6 did not belong to any of the defined major clones, and instead displayed single pulsotypes (ST2, ST1, and ST3). The maximum similarity level between these three strains and those remaining in the set was around 75%.

FIG. 1.

PFGE profiles of 37 ColR-Kp and 3 ColS-Kp isolates. ColR, colistin-resistant strain; ColS, colistin-susceptible strain; PFGE, pulsed-field gel electrophoresis.

Table 5.

Demographic Data for Patients with Common Pulsotype 3 and Common Pulsotype 4

Categorical variables	CT3, n = 14 (%)	CT4, n = 14^a (%)
Median age (min–max)	49 (25–84)	59 (20–78)
Previous hospitalization day (min–max)	17 (2–48)	18 (5–72)
Transfer from another unit	3 (21.4)	4 (28.6)
Transfer from another institution	1 (7.1)	0 (0.0)
Underlying diseases
Diabetes mellitus	0 (0.0)	1 (7.1)
Chronic renal failure	2 (14.2)	0 (0.0)
Chronic heart failure	0 (0.0)	0 (0.0)
COPD	1 (7.1)	0 (0.0)
Previous antibiotic therapy
Carbapenems	8 (57.1)	5 (35.7)
Beta-lactams	7 (56.8)	6 (42.9)
Glycopeptides	5 (35.7)	3 (21.4)
Colistin	6 (42.9)	5 (35.7)
Tigecycline	2 (14.2)	0 (0.0)

Two patients included in CT4 were excluded in this table as they were infected with ColS-Kp.

CT3, common pulsotype 3; CT4, common pulsotype 4.

Among six ColR-Ab strains, four pulsotypes, consisting of one common and three single pulsotypes, were detected among the PFGE profiles. No ColR-Kp or ColR-Ab strains were found to be PCR-positive for mcr-1 or mcr-II genes.

Discussion

Colistin is the last drug of choice for the treatment of multidrug-resistant (MDR) Gram-negative pathogens and can be administered as a stand-alone medication or in combination with other antimicrobial agents.¹³ However, due to inappropriate use of colistin, resistant strains have emerged worldwide, which increases the epidemiological significance of these pathogens.^14–16 Over the past 10 years in Turkey, colistin has been increasingly used to treat hospitalized patients and colistin-resistant pathogens have also emerged.^17,18 This 5-year retrospective survey revealed that 111 patients admitted to our university hospital were infected with or colonized by ColR-Ab or ColR-Kp. This number is alarmingly high for a single center. However, the inconsistent pattern in the number of patients with ColR-Ab and ColR-Kp between 2011 and 2016 was insufficient to allow determination of whether there was an increasing or decreasing trendin the number of detected ColR strains.

Among the 111 patients, 45% (50/111) received colistin before isolation of ColR strains to treat carbapenem-resistant pathogen infections or for empirical purposes, which is lower than the 82.9% and 63% colistin use rates reported by Matthaiou et al.¹⁵ and Kontopidou et al.,¹⁶ respectively. Colistin has been used in Turkey since 2014 as an empirical therapy for patients with MDR A. baumannii, which is endemic in our ICUs.¹⁹ In this study, 74.8% of the case patients were treated in the ICU and most of their ColR-pathogens were XDR. Emergence of such resistant strains can be attributed to both evolution of susceptible strains due to colistin stress and to cross-transmission. The issue of colistin resistance among pathogens in ICU patients is important because these bacteria are often resistant to many other antimicrobials, including colistin.^14,20 This resistence renders the choice of a rational antimicrobial regimen extremely problematic.

In this study, PFGE results revealed two major clones for ColR-Kp strains, which imply the presence of cross-transmission of endemic strains between patients over the last 2 years. Each clone contains strains isolated at scattered time points within the sampling period, suggesting that they might be resident inhabitants of the ICU biogeography. The results further indicate that current infection control measures are vulnerable and might result in ColR strains outbreaks. Moreover, antibiotic resistance tests showed that 90% of strains were XDR. This means along with other antimicrobials, colistin, the last weapon remaining in the antibiotics arsenal, would also be ineffective. Insufficient infection control measures could also lead to a conducive climate for the growth of pan-drug-resistant strains, which would bring us one step closer to a postantibiotic era crisis. Based on our analysis, two ColS-Kp strains were found to belong to the same clone as a ColR-Kp strain isolated from the same patient 3 months later in 2014. This finding could suggest that the existence of colistin-pressure-dictated mutations is associated with the corresponding resistant phenotype. On the other hand, horizontal transfer of resistance genes other than mcr-1 or mcr-2 might be responsible for the observed colistin resistance. In both cases, a recent gain of resistance event likely occurred in the region of concern.

A number of studies claim that colistin resistance is caused by mutations of genes in clusters that respond to colistin stress and encode proteins that are involved in lipopolysaccharide biosynthesis pathways.^21,22 While these mutations raise significant concerns, recent findings that colistin resistance genes (e.g., mcr-1 and mcr-2) can be carried on plasmids presents a far greater danger.^4,10 In this study, none of the ColR strains tested were positive for mcr-1and mcr-2 genes. Thus, the mechanism of colistin resistance in our set of strains might indeed arise because of mutations in lipopolysaccharide biosynthesis pathways (or other vertical mechanisms) rather than by gene transfer through a resistome-mobilome ecosystem. However, the fact that current knowledge regarding horizontally transferable colistin resistance mechanisms is limited to only a few recently discovered genes should be taken into consideration. Thus, further genome-based studies are needed to identify unknown colistin resistance mechanisms so that proper infection control measures for colistin resistance can be identified before a potential pan-resistance crisis.

Based on the risk factor analysis in this study, risk factors such as previous hospitalization duration, and colistin and carbapenem administration were associated with the presence of ColR-Kp and ColR-Ab strains in patients. However, previous hospitalization duration could be a confounding factor for carbapenem and colistin use. Similarly, the use of combined drugs is associated with some classes of antimicrobials risk factors, even though such drug combinations may not play a primary role in colistin resistance emergence. The presence of glycopeptide resistance of K. pneumoniae is one outcome of combination antibiotic treatment, but here the incidence of glycopeptide use in ColR was found to be insignificant in multivariate statistical analyses. Based on Mann–Whitney U test and multivariate analysis, patient age is indeed a risk factor for ColR-Kp emergence, suggesting that the underlying mechanism for this association might be more demographic than biological. Moreover, ColR-Kp was not observed in pediatric patients, indicating that younger patients may have a statistically lower risk of ColR-Kp.

Multivariate risk analyses results showed that previous colistin and carbapenem administration is independently and strongly associated with isolation of ColR strains. Although many earlier studies identified previous colistin administration as a risk factor,^15,23 our data suggest that previous carbapenem use is another major risk factor for the emergence of ColR strains. Meanwhile, in a recent study, exposure to the antiseptic agent chlorhexidin was reported to be associated with emerging ColR in K. pneumoniae.²⁴

In our clinics, carbapenems are increasingly used as the preferred antimicrobials. The relationship between carbapenem use and the emergence of colistin resistance is unknown, but the association in terms of risk factors could stem from several mechanisms, including correlated application patterns of colistin and carbapenem in treatment strategies and pleiotropic modifications that also trigger colistin resistance and are governed by carbapenem pressure or horizontal acquisition of these two resistances simulataneously (e.g., existence of mobile genetic elements correspondingto MDR gene islands). Nevertheless, the claim that frequent carbapenem use positively affects the emergence of colistin resistance requires further investigation, particularly studies that involve comparative genomics.

This study has some limitations. We could not include all 111 strains for the microbiological and the molecular study since strain collection began only in 2014. There were only a small number of ColR-Ab isolates included in the microbiological and molecular analyses such that conclusions about whether the emergence of ColR-Ab strains was sporadic or epidemic could not be determined. Furthermore, as noted above, additional studies designed to investigate the potential relationship between carbapenem use and colistin resistance are needed to support conclusions about this association.

In conclusion, this study emphasizes the emergence of XDR-ColR pathogens in Turkey as a developing country and suggests risk factors associated with their emergence. The results highlight that previous use of carbapenem increases the risk of emergence of these pathogens. Moreover, insufficient infection control measures might lead to outbreaks of XDR strains that are resistant to the antibiotic of last resort, colistin.

Footnotes

Acknowledments

We thank Mrs. Burcu Arslan and the Erciyes University Hospital Central Bacteriology Laboratory staff for their technical help. Part of the study was presented at the 26th ECCMID Congress in 2016.

Disclosure Statement

No competing financial interests exist.

References

Giamarellou

2010. Multidrug-resistant gram-negative bacteria: how to treat and for how long. Int. J. Antimicrob. Agent., 36:50–54.

Cai

, Chai

, Wang

, Liang

, and Bai

. 2012. Colistin resistance of Acinetobacter baumannii: clinical reports, mechanisms and antimicrobial strategies. J. Antimicrob. Chemother., 67:1607–1615.

Y.M.

, Kim

A.J.

, and Lee

J.Y.

. 2014. Colistin resistance in Klebsiella pneumoniae. Int. J. Antimicrob. Agents., 44:8–15.

Liu

Y.Y.

, Wang

, Walsh

T.R.

, Yi

L.X.

, Zhang

, Spencer

, Doi

, Tian

, Dong

, Huang

, Yu

L.F.

, Gu

, Ren

, Chen

, Lv

, He

, Zhou

, Liang

, Liu

J.H.

, and Shen

. 2015. Emergence of plasmid-mediated colistin resistance mechanism MCR1 in animals and human beings in China: a microbiological and molecular biological study. Lancet Infect. Dis., 3099:1–8.

Boucher

H.W.

, Talbot

G.H.

, Bradley

J.S.

, Edwards

J.E.

, Gilbert

, Rice

L.B.

, Scheld

, Spellberg

, and Bartlett

. 2009. Bad bugs, no drugs: no ESKAPE! An update from the Infectious Diseases Society of America. Clin. Infect. Dis., 48:1–12.

Weiner

, Webb

, Limbago

, Dudeck

, Patel

, Kallen

, Edwards

, and Sievert

. 2016. Antimicrobial- resistant pathogens associated with healthcare-associated infections: summary of data reported to the national healthcare safety network at the centers for disease control and prevention, 2011–2014. Infect. Control Hosp. Epidemiol., 37:1288–1301.

Sunenshine

R.H.

, Wright

M.-O.

, Maragakis

L.L.

, Harris

A.D.

, Song

, Hebden

, Cosgrove

S.E.

, Anderson

, Carnell

, Jernigan

D.B.

, Kleinbaum

D.G.

, Perl

T.M.

, Standiford

H.C.

, and Srinivasan

. 2007. Multıdrug-resistant Acinetobacter infection mortality rate and length of hospitalization. Emer. Infect. Diseas., 13:97–103.

Capone

, Giannella

, Fortini

, Giordano

, Meledandri

, Ballardini

, Venditti

, Bordi

, Capozzi

, Balice

M.P.

, Tarasi

, Parisi

, Lappa

, Carattoli

, and Petrosillo

; SEERBIO-GRAB Network. 2012. High rate of colistin resistance among patients with carbapenem-resistant Klebsiella pneumoniae infection accounts for an excess of mortality. CMI., 19:E23–E30.

Clinical and Laboratory Standards Institute. 2012. Performance standards for antimicrobial susceptibility testing; Twenty-First Information Supplement, CLSI document M100–S121. Wayne, PA.

10.

Xavier

B.B.

, Lammens

, Ruhal

, Kumar-Singh

, Butaye

, Goossens

, and Malhotra-Kumar

. 2016. Identification of a novel plasmid-mediated colistin-resistance gene, mcr-2, in Escherichia coli, Belgium, June 2016. Euro. Surveill, 21:pii: 30280.

11.

Tenover

F.C.

, Arbeit

R.D.

, Goering

R.V.

, Mickelsen

P.A.

, Murray

B.E.

, Persing

D.H.

, and Swaminathan

. 1995. Interpreting chromosomal DNA restriction patterns produced by pulsed-field gel-electrophoresis-criteria for bacterial strain typing. J. Clin. Microbiol., 33:2233–2239.

12.

Magiorakos

A.P.

, Srinivasan

, Carey

R.B.

, Carmeli

, Falagas

M.E.

, Giske

C.G.

, Harbarth

, Hindler

J.F.

, Kahlmeter

, Olsson-Liljeguist

, Paterson

D.L.

, Rice

L.B.

, Stelling

, Struelend

M.J.

, Vatopoulos

, Weber

J.T.

, and Monnet

D.L.

. 2012. Multidrug-resistant, extensively drug-resistant and pandrug-resistant bacteria: an international expert proposal for interim standard definitions for acquired resistance. Clin. Microbiol. Infect., 18:268–281.

13.

Pogue

J.M.

, Mann

, Barber

K.E.

, and Kaye

K.S.

. 2013. Carbapenem-resistant Acinetobacter baumannii: epidemiology, surveillance and management. Expert Rev. Anti Infect. Ther., 11:383–393.

14.

Qureshi

Z.A.

, Hittle

L.E.

, O'Hara

J.A.

, Rivera

J.I.

, Sved

, Shields

R.K.

, Pasculle

A.W.

, Ernst

R.K.

, and Doi

. 2015. Colistin-resistant Acinetobacter baumannii: beyond carbapenem resistance. Clin. Infect. Dis., 60:1295–1303.

15.

Matthaiou

D.K.

, Michalopoulos

, Rafailidis

P.I.

, Karageorgopoulus

D.E.

, Papaioannou

, Ntani

, Samonis

, and Falagas

M.E.

. 2008. Risk factors associated with the isolation of colistin-resistant gram-negativebacteria: a matched case-control study. Crit. Care Med., 36:807–811.

16.

Kontopidou

, Plachouras

, Papadomichelakis

, Koukos

, Galani

, Poulakou

, Dimopoulos

, Antoniadou

, Armaganidis

, and Giamarellou

. 2011. Colonization and infection by colistin-resistant Gram-negative bacteria in a cohort of critically ill patients. Clin. Microbiol. Infect., 17:E9–E11.

17.

Keske

, Yılmaz

G.R.

, Güven

, Güner

, and Taşyaran

M.A.

. 2015. A health care-associated pneumonia case due to colistin resistant Acinetobacter baumannii. J. Microbiol. Infect. Dis., 4:111–113.

18.

Güven

, Yilmaz

, Güner

H.R.

, Kaya

, Eser

, and Taşyaran

M.A.

. 2014. Increasing resistance of nosocomial Acinetobacter baumannii: are we going to be defeated. Turk. J. Med. Sci., 44:73–78.

19.

Alp

, Kiran

, Altun

, Kalin

, Cosgun

, Sungur

, Akin

, Percin

, and Doganay

. 2011. Changing pattern of antibiotic susceptibility in intensive care units: ten years experience of a university hospital. Anaerobe., 17:422–425.

20.

Oikonomou

, Sarrou

, Papagiannitsis

C.C.

, Georgiadou

, Mantzarlis

, Zakynthinos

, Dalekos

G.N.

, and Petinaki

. 2015. Rapid dissemination of colistin and carbapenem resistant Acinetobacter baumannii in Central Greece: mechanismsof resistance, molecular identification and epidemiological data. BMC Infect. Dis., 15:559.

21.

Adams

M.D.

, Nickel

G.C.

, Bajaksouzian

, Lavender

, Murthy

A.R.

, Jacobs

M.R.

, and Bonomo

R.A.

. 2009. Resistance to colistin in Acinetobacter baumannii associated with mutations in the PmrAB two-component system. Antimicrob. Agents Chemother., 53:3628–3634.

22.

Moffatt

J.H.

, Harper

, Harrison

, Hale

J.D.

, Vinogradoy

, Seemann

, Henry

, Crane

, St Michael

, Cox

A.D.

, Adler

, Nation

R.L.

, Li

, and Boyce

J.D.

. 2010. Colistin resistance in Acinetobacter baumannii is mediated by complete loss of lipopolysaccharide production. Antimicrob. Agents Chemother., 54:4971–4977.

23.

Giacobbe

D.R.

, del Bono

, Trecarichi

E.M.

, De Rosa

F.G.

, Giannella

, Bassetti

, Bartoloni

, Losito

A.R.

, Corcione

, Bartoletti

, Mantengoli

, Saddioti

, Pagani

, Tedeschi

, Spanu

, Rossolinj

G.M.

, Marchese

, Ambretti

, Cauda

, Viale

, Viscoli

, and ISGRI-SITA. 2015. Risk factors for bloodstream infections due to colistin-resistant KPC-producing Klebsiella pneumoniae: results from a multicenter case-control-control study. Clin. Microbiol. Infect. 21:1106.e1–1108.e8.

24.

Wand

M.E.

, Bock

L.J.

, Bonney

L.C.

, and Sutton

J.M.

. 2016. Mechanisms of increased resistance to chlorhexidine and cross-resistance to colistin following exposure of Klebsiella pneumoniae clinical isolates to chlorhexidine. Antimicrob. Agents Chemother. 61:pii: e01162.