Comparison of Antimicrobial Resistance Profiles among Extended-Spectrum-β-Lactamase-Producing and Acquired AmpC β-Lactamase-Producing Escherichia coli Isolates from Canadian Intensive Care Units

Abstract

Resistance profiles were compared among 18 extended-spectrum-β-lactamase-producing (ESBL) and 27 acquired AmpC β-lactamase-producing Escherichia coli isolates collected from Canadian intensive care units from 2005 to 2006. ESBL-producing E. coli isolates were more likely to be gentamicin resistant (P < 0.03), fluoroquinolone resistant (P < 0.0001), and multidrug resistant (P < 0.0001) than AmpC-producing E. coli isolates.

Escherichia coli is the most commonly encountered pathogen within the Enterobacteriaceae family, causing significant infections worldwide. The spread of resistance to extended-spectrum cephalosporins in E. coli has been increasing in both hospital and community settings. Resistance to cephalosporins often is mediated by the production of extended-spectrum β-lactamases (ESBL). However, it also may be attributed to AmpC cephalosporinases, which can result either from the overexpression of the chromosomally encoded AmpC enzyme in E. coli or by the acquisition of a plasmid-mediated AmpC enzyme (13).

Generally, isolates with AmpC differ from ESBL producers phenotypically by yielding a negative ESBL disk test result, and also they are resistant to cefoxitin and usually are susceptible to cefepime/cefpirome. Resistance to other classes of antimicrobials among ESBL-producing E. coli isolates and, to a lesser extent, among AmpC-producing E. coli isolates has been escalating and limits therapeutic options (4, 11).

The purpose of this study was to compare the prevalence as well as the resistance patterns of ESBL-producing E. coli (ESBL-EC) isolates and acquired AmpC β-lactamase-producing E. coli (AmpC-EC) isolates from Canadian intensive care units (ICUs).

Study isolates were obtained as part of the Canadian ICU (CAN-ICU) surveillance study. CAN-ICU is a laboratory-based surveillance study coordinated by the Health Sciences Centre in Winnipeg, Canada. From 1 September 2005 to 1 August 2006, inclusive, 19 tertiary-care hospitals from across Canada with active ICUs voluntarily participated as study sites. Each center collected a maximum of 300 consecutive pathogens isolated from blood, urine, wound tissue, and respiratory specimens of ICU patients. As only pathogens causing infectious diseases were of interest, surveillance cultures and eye, ear, nose, and genital swabs were excluded. Duplicate isolates from the same site also were excluded. No anaerobes or fungi were assessed in the CAN-ICU study. All isolates were identified at participating sites by routine procedures performed at each laboratory. Isolates were shipped to the reference laboratory at the Winnipeg Health Sciences Centre on Amies charcoal swabs, subcultured onto blood agar, and stocked in skim milk at −80°C until MIC testing was carried out. The most common infectious sites and pathogens causing infection in Canadian ICUs have been reported previously (3) (www.can-r.ca). In this study, only the analysis of all unique genotypically confirmed ESBL- and AmpC β-lactamase-producing E. coli isolates obtained from CAN-ICU will be reported.

After two subcultures from frozen stock, the in vitro activities of ampicillin, piperacillin-tazobactam, ceftriaxone, cefepime, cefoxitin, gentamicin, ciprofloxacin, levofloxacin, meropenem, ertapenem, doxycycline, tigecycline, and trimethoprim-sulfamethoxazole were determined in triplicate by broth microdilution in accordance with CLSI guidelines (2).

Any E. coli isolate for which the ceftriaxone MIC was ≥1 μg/ml was identified as a putative ESBL or AmpC β-lactamase producer. All 50 E. coli isolates identified as putative ESBL or AmpC producers were tested genotypically for ESBL and AmpC genes by PCR. Universal primer sets were used to identify bla_SHV, bla_TEM, bla_CTX-M, and bla_{OXA-1,-2,-10-like} ESBL genes by PCR as previously described (1, 6, 12).

Genes encoding EBC (ACT-1/MIR-1-related genes), DHA (DHA-related genes), FOX (FOX-related genes), and CIT (CMY-2-related genes) phylogenetic groups of AmpC-acquired enzymes were detected using a previously described multiplex PCR (8). The PCR products were purified and sequenced. A BLAST search of the DNA sequence was conducted to determine the specific gene.

All ESBL-producing and AmpC-producing E. coli isolates were subtyped by pulsed-field gel electrophoresis by the standardized protocol for E. coli as previously described (6).

χ² analysis or t tests were used to evaluate statistical significance, as appropriate, using Graphpad Quickcalcs.

A total of 4,092 clinical isolates were obtained from the CAN-ICU study. Of the 4,092 isolates, 493 (12.0%) were E. coli isolates. The ceftriaxone MIC for 50 of the 493 (10.1%) E. coli isolates was ≥1 μg/ml, and these isolates were identified as putative ESBLs. Eighteen (3.7%) and 27 (5.5%) of 50 E. coli isolates were identified genotypically as unique ESBL and acquired AmpC producers, respectively. The mechanism producing elevated ceftriaxone MICs for the remaining five E. coli isolates was not identified. No isolates carried both ESBL and acquired AmpC genes, as determined by PCR. All 18 (100%) ESBL-EC isolates were of the CTX-M genotype, with 13 (72.2%) being bla_CTX-M-15, 2 (11.1%) being bla_CTX-M-2, and 1 (5.6%) each being bla_CTX-M-9, bla_CTX-M-14, or bla_CTX-M-1. Of the 27 AmpC-EC isolates, 26 (96.3%) were bla_CMY-2 and 1 (3.7%) was bla_ACT-1. A comparison of the patient demographics for the 18 ESBL- and 27 AmpC-producing E. coli isolates is summarized in Table 1. AmpC-producing E. coli isolates were significantly more likely to be isolated from eastern Canada (P < 0.005) and from urine (P < 0.018) than ESBL producers, which were more likely to be isolated from central Canada (P < 0.002) and from a variety of specimen sources (Table 1).

TABLE 1.

Comparison of patient demographics, region of isolation, and specimen source of ESBL-EC and AmpC-EC isolates from Canadian ICUs

Isolate type (n)	Patient age		Gender (%)		% Isolates from regiona:			% Isolates from specimen type:
	Mean age	Range	Female	Male	W (n = 7)	C (n = 7)	E (n = 5)	Urine	Blood	Wound	RTb
ESBL-EC (18)	60	25-83	55.6	44.4	11.1	55.6	33.3	50.0	27.8	11.1	11.1
AmpC-EC (27)	66	5-96	77.8	22.2	11.1	11.1	77.8	85.2	7.4	3.7	3.7
P value	NSc		NS	NS	NS	<0.002	<0.005	<0.018	NS	NS	NS

^aThe west region (W) encompasses British Columbia, Alberta, Saskatchewan, and Manitoba; the central region (C) encompasses Ontario and Quebec; the east region (E) encompasses New Brunswick and Nova Scotia. The numbers in parentheses indicate the number of sites from which isolates were obtained.

^bRT, respiratory tract.

^cNS, not significant (P > 0.05).

A comparison of antimicrobial susceptibilities of the AmpC- and ESBL-producing E. coli isolates is shown in Table 2. Susceptibility testing demonstrated that resistance to fluoroquinolones (ciprofloxacin and levofloxacin), trimethoprim-sulfamethoxazole, doxycycline, or gentamicin occurred in 83.3 and 11.1% (P < 0.0001), 61.1 and 26.6% (not significant; P < 0.064), 33.3 and 18.5% (not significant; P < 0.304), and 27.8 and 3.7% (P < 0.03) of ESBL-EC and AmpC-EC isolates, respectively (Table 2). The AmpC-EC isolates were more susceptible to all antimicrobials tested, with the exception of cefoxitin (by definition). All ESBL- and AmpC-producing isolates were susceptible to meropenem, ertapenem, and tigecycline (Table 2). Eighteen (100%) ESBL-EC isolates and nine (33.3%) AmpC-EC isolates were multidrug resistant (P < 0.0001) (i.e., resistant to three or more different antimicrobial classes).

TABLE 2.

Comparison of antimicrobial susceptibilities of ESBL-EC and AmpC-EC isolates from Canadian ICUs^a

Antibiotic	ESBL-EC isolates (n = 18)						AmpC-EC isolates (n = 27)
	MIC50	MIC90	Range	% Sensitive	% Intermediate	% Resistant	MIC50	MIC90	Range	% Sensitive	% Intermediate	% Resistant
Ampicillin	>256	>256	>256			100	>256	>256	>256			100
Ceftriaxone	128	>256	8->256	5.6	11.1	83.3	16	64	2-128	44.4	44.4	11.1
Cefepime	8	32	≤1-256	50.0	38.9	11.1	≤1	≤1	≤1-2	100
TZP	2	8	≤1-64	94.4		5.6	2	8	≤1-16	100
Cefoxitin	8	32	4-32	61.1	22.2	16.7	256	>256	32->256			100
Ciprofloxacin	>16	>16	≤0.06->16	16.7		83.3	≤0.06	>16	≤0.06->16	85.2	3.7	11.1
Levofloxacin	16	32	≤0.12->32	16.7		83.3	≤0.06	16	≤0.06->32	88.9		11.1
SXT	>32	>32	0.12->32	38.9		61.1	2	32	≤0.12-32	70.4		26.6
Doxycycline	4	64	1-128	61.1	5.6	33.3	2	64	0.5->256	66.7	14.8	18.5
Gentamicin	1	16	≤0.25-16	66.7	5.6	27.8	≤0.25	8	≤0.25-64	88.9	7.4	3.7
Meropenem	≤0.12	≤0.12	≤0.12	100			≤0.12	≤0.12	≤0.12	100
Ertapenem	0.12	0.25	0.03-0.25	100			0.12	0.5	0.03-2	100
Tigecycline	0.5	1	0.25-1	100			0.25	0.5	0.12-2	100

^aMIC₅₀, MIC at which 50% of the isolates tested are inhibited; MIC₉₀, MIC at which 90% of the isolates tested are inhibited; TZP, piperacillin-tazobactam; SXT, trimethoprim-sulfamethoxazole.

A dendrogram depicting the genetic relationships among ESBL-EC and AmpC-EC isolates from Canadian ICUs is shown in Fig. 1. Both genetically related (>80% homology) and unrelated clusters of E. coli isolates were observed. On the basis of macrorestriction patterns, intrahospital and interhospital spread may have occurred in eastern Canada and in each case that involved two AmpC-EC isolates.

FIG. 1.

Dendrogram depicting the genetic relationships among ESBL- and AmpC-producing E. coli isolates from Canadian ICUs. The darkly shaded region highlights the intrahospital spread of two AmpC-EC isolates, whereas the lightly shaded region highlights the interhospital spread of two AmpC-EC isolates.

In 2005 and 2006, AmpC-EC isolates (5.5%) were found to be more prevalent than ESBL-EC isolates (3.7%) in Canadian ICUs. The increasing awareness and improved recognition of ESBL producers have led to improved infection control measures to minimize the spread of these emerging pathogens. Therefore, it is not surprising that the prevalence of ESBL-EC isolates is lower than that of AmpC producers, which generally are unidentified and not subjected to such infection control measures (4). However, the prevalence of ESBL-EC isolates in this study may be underestimated due to the use of a single extended-spectrum cephalosporin as an indicator for screening. This study, in agreement with other studies, found that CTX-M and CIT-type were the predominant enzymes being produced, with CTX-M-15 and CMY-2 being the predominant genotypes among ESBL-EC and AmpC-EC isolates, respectively (4, 5, 7, 9-11, 13). Since 1999 to 2000, CTX-M has replaced SHV as the most common ESBL in Canada; however, CIT-type remains the predominant acquired AmpC enzyme and has been increasing in prevalence (6, 7). No E. coli isolate carried both ESBL and acquired AmpC genes, an occurrence that has yet to be described in Canada. However, there were two ESBL-producing isolates that were resistant to cefoxitin. These isolates could have chromosomal promoter ampC mutations or porin mutations, or they may have acquired an unknown bla_AmpC gene. In contrast to a previous study of cefoxitin-resistant E. coli isolates from Canadian hospitals, cefoxitin resistance in AmpC-EC isolates was attributed to the acquisition of plasmid-mediated AmpC genes rather than the overexpression of the chromosomally encoded AmpC gene (7). AmpC-EC isolates were significantly less likely to be multidrug resistant and more likely to be susceptible to fluoroquinolones and gentamicin than ESBL-EC isolates. Although AmpC-EC infections are, at the moment, more susceptible to many more antimicrobials than ESBL-EC infections, they are an increasing public health concern (10, 13). Like the ESBL genes, the AmpC genes have localized to multidrug-resistant plasmids and have further limited our therapeutic options (5). With no guidelines for proper identification and infection control measures, AmpC-producing E. coli infections may become a greater concern than ESBL-producing E. coli infections, as they are increasing in prevalence worldwide. In addition, multidrug-resistant plasmids bearing both ESBL and AmpC genes recently have been reported and will most likely spread among bacteria and become the new emerging threat (11, 13).