Phenotypic methods for screening carbapenem-resistant Enterobacteriaceae and assessment of their antimicrobial susceptibility profile

Introduction: In this study, we used phenotypic methods to screen carbapenem-resistant Enterobacteriaceae (CREs) and evaluated their antimicrobial sensitivity profile. Methods: One hundred and seventy-eight CREs were isolated at a university hospital in south Brazil in a one-year period. Samples were assessed using disk diffusion tests with inhibitors of β-lactamases such as phenylboronic acid (AFB), cloxacillin (CLOXA), and ethylenediaminetetraacetic acid (EDTA). Strains with differences in zone diameters ≥ 5mm for disks supplemented or not were considered producers of carbapenemases. Results: Klebsiella pneumoniae was the most prevalent CRE, which appeared in 80.3% cases (n = 143). Among clinical materials, the rectal swab was responsible for 43.4% of the isolations (n = 62), followed by urine (18.9%; n = 27). Among the CREs identified in this study, the growth of 56.7% (n = 101) isolates, which were putative producers of Klebsiella pneumoniae carbapenemase (KPC), were inhibited by AFB, whereas 7.3% (n = 13) isolates were inhibited by both AFB and CLOXA and were considered as putative producers of plasmid-mediated AmpC; approximately 3.4% (n = 6) were inhibited by EDTA, which possibly produced metallo-βlactamase. Lastly, 32.6% (n = 58) cases showed negative results for AFB, CLOXA, and EDTA sensitivity, and represented another class of β-lactamases and/or mechanism of resistance. Conclusions: Phenotypic screening of CREs is important for clinical laboratories that monitor outbreaks of resistant microbes. Phenotypic tests that use carbapenemase inhibitors and enhancers such as AFB, CLOXA, and EDTA are necessary since they are good screening methods for the detection of carbapenemases.


INTRODUCTION
Emergence of antibiotic-resistant bacteria is a major global public health concern, and the Gram-negative bacilli of the Enterobacteriaceae family are well-known for exhibiting drug-resistance.Drug-resistant microorganisms cause recurrent infections in patients in hospital units, limiting treatment alternatives and increasing morbidity and mortality rates 1,2 .Since the isolation of Enterobacteriaceae that produce an extended spectrum of β-lactamases (ESBL) capable of hydrolysing almost all cephalosporins, use of carbapenems (imipenem, meropenem, ertapenem, and doripenem) in treating Enterobacteriaceae infections has become mandatory 1 .These antimicrobials are crucial for preventing and treating infections in high-risk patients such as those undergoing transplantation surgery or any other surgical procedure or admitted in intensive care units (ICU) 1 .
A wide variety of carbapenemase-producing Enterobacteriaceae have been reported worldwide 3,4 .Carbapenem resistance is mediated the by transfer of mobile genetic elements such as plasmids and transposons, which are easily transferred to other bacterial genera and species, i.e., Enterobacter cloacae, Citrobacter freundii, Salmonella spp., Escherichia coli, among others.
The mechanisms via which Enterobacteriaceae resists different classes of antimicrobials vary; for example, the mechanisms may be associated with the decrease or loss of porin in bacterial outer membranes (OMPs) and efflux pumps, mutations in the active site of antimicrobials that decreases their affinity for microbes, and the presence of β-lactamase-encoding genes 3,5-7 .Among the carbapenemases produced from plasmids, the Ambler class A (Klebsiella pneumoniae carbapenemase -KPC and Guiana-Extended-Spectrum-GES) has been identified in clinical isolates 8 .The other types of carbapenemases include Ambler class B or metallo-β-lactamases (MBL) (Verona imipenemase-VIM, Imipinemase-IPM, and New Delhi metalloβ-lactamase-NDM) and oxacillinases or Ambler class D (Oxa-carbapenemases-OXA-48) 1,3,[9][10][11] .
From the epidemiological point of view, bacteria that produce KPC carbapenemase are the most worrisome owing to their rapid worldwide dissemination 12 .These bacteria are considered important agents of nosocomial infections because they produce carbapenemase, which is an enzyme that hydrolyzes the β-lactam ring of not only carbapenem antibiotics, but also those of cephalosporins, penicillin, and monobactams 13 .
Infections caused by carbapenem-resistant Enterobacteriaceae (CREs) increase the morbi-mortality rates of patients, especially those admitted in hospitals or with weakened immune systems, and make therapeutic alternatives scarce 14,15 .
In this study, we used phenotypic methods to screen carbapenem-resistant Enterobacteriaceae (CREs) isolated at a university hospital in South Brazil in a one-year period (July 2014 to July 2015), and evaluated their antimicrobial sensitivity profiles.

Study site
The study was performed at the Laboratório de Bacteriologia do Departamento de Análises Clínicas e Toxicológicas (LaBac) at Centro de Ciências da Saúde of Universidade Federal de Santa Maria (UFSM), Santa Maria, Rio Grande do Sul.Samples were provided by the Laboratório de Análises Clínicas of the Hospital Universitário de Santa Maria (HUSM), Santa Maria, Rio Grande do Sul.

Samples
One hundred seventy-eight samples of CREs were isolated between July, 2014 and July, 2015 from several biological materials, including epidemiologic vigilance research comprising patients admitted to a university hospital (HUSM) in the southern region of Brazil.Samples were subsequently sent to LaBac and subsequently stored in 15% glycerol at -80°C for further phenotypic tests.

Bacterial identification test
All cultures were collected and processed per the standard operating procedure (SOP) at the Laboratório de Análises Clínicas of the hospital.Identification tests of the isolated bacteria were performed using the automated system, Vitek ® 2 (BioMérieux, France).

Sensitivity profile
Sensitivity profiles of the isolates were assessed through the automated methodology Advanced Expert System (BioMérieux, France), following recommendations of the Clinical and Laboratory Standards Institute 16 .Sensitivity cards were used with the following antimicrobials: ertapenem, meropenem, imipenem, amikacin, gentamicin, norfloxacin, nitrofurantoin, sulfamethoxazole/ trimethoprim, ciprofloxacin, tigecycline, and colistin.

Phenotypic tests with phenylboronic acid, cloxacillin, and ethylenediaminetetraacetic acid
Samples stored in 15% glycerol at -80°C were reactivated in plates containing trypticase soy agar (TSA/Oxoid LTD, England), and incubated at 35 ± 2°C for 18 to 24h.A bacterial suspension was subsequently prepared in 0.9% sterile saline solution, with turbidity similar to the 0.5 McFarland standard, and humidified with a swab sowed in Mueller-Hinton agar (MHA/HiMedia Laboratories, India) in 15 × 150mm plates.Next, ertapenem, meropenem, and imipenem disks (Diagnósticos Microbiológicos Especializados, Brazil) were placed on a Petri dish, supplemented with a 10μL solution of AFB (40mg/mL, Sigma-Aldrich), CLOXA (75mg/mL, Sigma-Aldrich), or ethylenediaminetetraacetic acid (EDTA) (0.1mol/L, Proquimios Comércio e Indústria Ltda, Brazil), with a drying time of 20 minutes such that they could be applied on the bacterial suspension in MHA at a distance of 3cm from one another.Non-supplemented ertapenem, meropenem, and imipenem disks served for comparison with supplemented disks.In a plate there were placed non-supplemented ertapenem, meropenem, and imipenem disks; ertapenem, meropenem and imipenem disks supplemented with AFB; meropenem and imipenem disks supplemented with CLOXA; and meropenem and imipenem disks supplemented with EDTA.The plates were then incubated at 35 ± 2°C for 18 to 24h 13 .
Subsequently, the difference of the inhibition zone diameter was compared between non-supplemented disks and those supplemented with AFB, CLO, or EDTA.Isolates with an inhibition zone difference ≥ 5mm for ertapenem, meropenem, and imipenem disks supplemented with AFB were considered possible KPC producers.Isolates with a difference ≥ 5mm for antimicrobial disks supplemented with AFB and CLOXA were considered possible producers of plasmid-mediated AmpC.Isolates with zone difference < 5mm for antimicrobial disks supplemented AFB, CLOXA and EDTA were considered possible producers of another β-lactamase (ex.OXA-48) or porin loss, and the ones that showed a zone difference ≥ 5mm only for disks supplemented with EDTA were considered likely producers of MBL 13 .

Ethical considerations
This study was approved by the Ethical Research Committee of the Federal University of Santa Maria under n° 0285.0.243.000-09.

DISCUSSION
The prevalence of CREs has increased worldwide, which represents an alarming threat to public health 15 .In this study, we showed that a large incidence of K. pneumoniae was detected in the analyzed samples, and the rectal swab, a surveillance culture, was the clinical material with the maximum number of isolates (43.4%).The most frequent carbapenemase detected in rectal swab isolates was KPC (64.5%).Similar results were reported by Pinto et al 15 , who assessed 701 CREs isolated from hospitals in Porto Alegre, in which 47% cases were represented by K. pneumonia, and 66% of these were KPC producers.In addition, 51.7% samples with CREs were from rectal swabs, which corroborated the results of our study 15   which 41.1% (n = 30) were KPC-producing K. pneumoniae (via the AFB test), which is similar to the results reported in this study where 56.7% clinical isolates were found to be KPCpositive using the same test.
Among the isolated CREs, 58 (32.6%) were carbapenemresistant but were not positive in any phenotypic tests, indicating the presence of another type of β-lactamases as a resistance mechanism (e.g.OXA-48 or porin loss).This was the secondmost frequent resistance mechanism identified in our study.Since the global frequency of occurrence of this class of carbapenem-resistant bacteria is still low (which corroborates the results of Pinto et al. 15 ), a detailed investigation into alternative mechanisms of resistance is required to control the dissemination of such strains in future.The majority of the isolates showed decreased sensitivity to carbapenems, which are the most commonly used therapeutic choices against these infections 1 .The isolates identified in this study showed increased resistance to carbapenems, quinolones, and glycylcyclines, which is similar to that shown by Hayder et al. 17 , where isolates producing KPC showed 100% resistance to carbapenems, cephalosporins, quinolones, and penicillin.Our results also in agreement with those reported by Singh et al. 12 , where the greatest resistance was observed for third generation cephalosporins (100%) and penicillin (93.3%).In addition, Singh et al. 12 have verified increased sensitivity to tigecycline (86.7%) and polymyxin (93.3%), which is different from the results of our study as we observed greater sensitivity to aminoglycosides and colistin.
However, 32 strains isolated in this study were resistant to colistin, which is an antimicrobial used in the treatment of infections caused by multidrug-resistant bacteria.Liu 18 , observed that the resistance to colistin is mediated by plasmids in K. pneumoniae and E. coli in China, and colistin-resistant bacteria were found in animals and isolated from humans.It is important to highlight that polymyxin B or colistin are used for the treatment of infections caused by CREs and are associated with one or more antimicrobials such as aminoglycosides (gentamicin or amikacin), carbapenems, and tigecycline 19,20 .This increases the concern associated with the indiscriminate use of these drugs in treating nosocomial infections and in veterinary medicine 18 .
The phenotypic detection of CREs is of great importance for clinical laboratories and for monitoring the emergence of resistant bacterial strains.The fast dissemination of genes and mechanisms of resistance to antimicrobials limits therapeutic options and increases the morbi-mortality of patients 15 .Therefore, phenotypic tests that use inhibitors and enhancers of carbapenemases such as AFB, CLOXA, and EDTA are necessary since they provide a good screening method for detection of carbapenemases.In addition, these methods are easy to adapt in the laboratory routine.However, the results obtained from these phenotypic tests should be confirmed by further molecular tests, if required, for identifying resistant strains.

Limitations of the study
This research presented limitations in terms of non-detection of other resistance mechanisms (OXA, MBL), which could interfere with sensitivity and specificity.In addition, the molecular tests were not performed to confirm the presence of the enzymes.

TABLE 1
Distribution of 178 CREs* isolated at the Hospital Universitário de Santa Maria (HUSM) from July 2014 to July 2015.
. Singh et al. (2015) 12 evaluated 73 samples from various clinical specimens like urine, pus, swabs, body fluids, among others, in India, of

TABLE 2
Distribution of CREs in clinical specimens and the resistance mechanism obtained in phenotypic tests.

TABLE 3
Resistance profile and sensitivity of Enterobacteriaceae isolated in the study.