Metallo-β-lactamase-producing Pseudomonas aeruginosa in the Netherlands: the nationwide emergence of a single sequence type

Authors


Corresponding author: A. K. Van der Bij, Department of Microbiology, Reinier de Graaf Hospital, PO Box 5011, 2600 GA Delft, the Netherlands
E-mail: akke.van.der.bij@rivm.nl

Abstract

Clin Microbiol Infect 2012; 18: E369–E372

Abstract

Recently, the first outbreak of clonally related VIM-2 metallo-β-lactamase (MBL)-producing Pseudomonas aeruginosa in a Dutch tertiary-care centre was described. Subsequently, a nationwide surveillance study was performed in 2010–2011, which identified the presence of VIM-2 MBL-producing P. aeruginosa in 11 different hospitals. Genotyping by multiple-locus variable-number tandem-repeat analysis (MLVA) showed that the majority of the 82 MBL-producing isolates found belonged to a single MLVA type (n = 70, 85%), identified as ST111 by multilocus sequence typing (MLST). As MBL-producing isolates cause serious infections that are difficult to treat, the presence of clonally related isolates in various hospitals throughout the Netherlands is of nationwide concern.

Introduction

Resistance to the carbapenems in Pseudomonas aeruginosa is most often due to the loss of the outer membrane protein OprD, the over-expression of efflux systems or a combination of both [1]. However, since the 1990s there has been a substantial increase in the reporting of metallo-β-lactamases (MBLs) among carbapenem-resistant P. aeruginosa isolates globally [2]. MBLs are capable of hydrolyzing most of the β-lactam antibiotics and MBL-encoding genes are usually located on integrons that frequently carry additional genes encoding for resistance to non-β-lactam antibiotics, resulting in multidrug resistance (MDR). Currently, the most prevalent and widespread MBL genes in P. aeruginosa are the VIM and IMP types; in particular, VIM-2 has become the dominant MBL type among P. aeruginosa worldwide [2].

The Netherlands is considered to be a low-prevalence country for antimicrobial resistance [3], and the reporting of MBLs among P. aeruginosa has been rare. However, a recent study at a tertiary-care centre identified a large number of VIM-2 MBL-producing P. aeruginosa isolates [4]. Genotyping by multiple-locus variable-number tandem-repeat analysis (MLVA) demonstrated one main cluster, suggesting the presence of a common source and/or nosocomial transmission. In the present study, we collected P. aeruginosa isolates non-susceptible to imipenem or meropenem according to the EUCAST 2010 recommendations (http://www.eucast.org) from 21 laboratories (c. 30% of the Dutch laboratories) serving over 30 hospitals, including three university teaching hospitals and various large community teaching hospitals, to assess the occurrence of MBL-producing P. aeruginosa in the Netherlands.

In total, 206 non-repeat isolates that were either retrospectively stored or prospectively collected from clinical and screening specimens between November 2009 and May 2011 were included. Isolates from cystic fibrosis patients were excluded. Identification to the species level was performed by oxidase activity (DrySlideTM Oxidase; Becton Dickinson, Franklin Lakes, NY, USA) and resistance to C390 (DIATABS™; Rosco Diagnostica A/S, Taastrup, Denmark). Of isolates susceptible to C390, further identification was performed by VITEK 2 (VITEK AMS; bioMérieux VITEK Systems Inc., Hazelwood, MO, USA). Antibiotic susceptibility was determined by disk diffusion (Oxoid Ltd, Basingstoke, UK) using EUCAST 2010 breakpoints (http://www.eucast.org).

Isolates that were positive in a phenotypic screening test for MBL production (i.e. imipenem and doripenem combo disk test with EDTA) [5], were tested by PCR for the presence of blaVIM and blaIMP genes [6]. Sequencing of the genes was performed using the class 1 integron primers 5CS (5′-GGC ATC CAA GCA GCA AG-3′) and 3CS (5′-AAG CAG ACT TGA CCT GA-3′) in combination with IMP and VIM primers, as previously described [7]. Isolates that were negative in the phenotypic screening test and isolates that were phenotypically positive but negative for blaVIM or blaIMP were further evaluated for carbapenemase activity by a modification of the bioassay of Masuda et al. [8] using a solution containing 50 mg/L imipenem on Mueller–Hinton agar plates, inoculated with Escherichia coli ATCC 25922 as described before [9]. Genotyping of all MBL-producing isolates was performed by MLVA as described by van Mansfeld et al. [10] using the MLVA9-Utrecht protocol. Additionally, MLST was performed using seven conserved housekeeping genes for P. aeruginosa (acsA, aroE, guaA, mutL, nuoD, ppsA and trpE) (http://pubmlst.org/paeruginosa/). All tests, except MLST, were performed in a centralized laboratory at the Erasmus University Medical Centre.

In total, 77 isolates tested positive in the phenotypic screening test, of which 76 carried blaVIM. The isolate that was negative for both blaVIM and blaIMP was considered falsely screen positive because carbapenemase activity was excluded by the modified bioassay. Of the 129 isolates that were negative in the phenotypic screening test, six were positive for carbapenemase activity in the bioassay. All six isolates carried blaVIM, resulting in 82 (40%) MBL-producers. Most of the isolates were isolated from the respiratory tract (n = 70, 34%); 32 (16%) were isolated from urine, 11 (5%) from blood, 22 (11%) from soft tissue or bone, 13 (6%) from intra-abdominal specimens and 45 (22%) from various other specimens. For 13 isolates (6%) the specimen site was unknown. MBL-producing P. aeruginosa were more often isolated from patients admitted to an ICU (51% vs. 36%, p = 0.05) and were more resistant than non-MBL-producing isolates (p < 0.01, Fig. 1). MBL-producing isolates were detected in eight hospitals, two burn wound centres, one long-term healthcare facility and in two community patients attending a general practitioner. The majority of the care facilities with MBL-producing isolates were located in the western part of the Netherlands, and five of these care facilities had more than one patient with an MBL-producing isolate within the same ward.

Figure 1.

 Resistance % for 206 carbapenem non-susceptible Pseudomonas aeruginosa isolates isolated from various care facilities in the Netherlands, from November 2009 to May 2011. IMP, imipenem; MEM, meropenem; PIP, piperacillin; CAZ, ceftazidime; TOB, tobramycin; AMK, amikacin; CIP, ciprofloxacin; AZT, aztreonam; COL, colistin.

Among the MBL-producing isolates, MLVA identified one large cluster consisting of 36 patients (MLVA 364) and a few closely related clusters consisting of 13 patients (MLVA 255), seven patients (MLVA 382) and four patients (MLVA 404), and nine closely or possibly related small clusters consisting of one or two patients each (Fig. 2). This large cluster of closely and possibly related isolates was present in all of the care facilities with MBL-producing isolates. MLST identified this large cluster as sequence type (ST) 111. MLVA also demonstrated a non-related cluster consisting of 11 patients (MLVA 68 and MLVA 409), identified by MLST as ST446. This smaller cluster was only present in one university hospital. There was also one patient with an MBL-producing isolate that was not related to any of the other isolates (MLVA 384). This patient attended a general practitioner in the north-eastern part of the Netherlands. Sequencing of 25 randomly selected isolates among all major MLVA types revealed the following gene cassettes in all 25 isolates: aacA29a, blaVIM-2 and aacA29b in the same order as previously described [11].

Figure 2.

 MLVA genotyping of MBL-producing P. aeruginosa isolates of 82 patients from various care facilities in the Netherlands, from November 2009 to May 2011, shows one large cluster of 70 patients, a smaller cluster of 11 patients (no. 68 and 409), and a single strain unrelated to any of the three clusters (no. 384).

The Netherlands has so far succeeded in keeping the incidence of major resistant nosocomial pathogens at a low level, [3] due to a combination of low antibiotic pressure in human healthcare [12] and the implementation of national prevention guidelines to minimize the nosocomial spread of resistant bacteria. The low incidence of methicillin-resistant Staphylococcus aureus (MRSA) is an example of the success of Dutch measures to prevent and control infection and transmission [13]. The presence of a multi-resistant VIM-2-MBL-producing P. aeruginosa clone in several hospitals in the Netherlands is therefore remarkable.

Pseudomonas aeruginosa has a non-clonal structure with a huge diversity of sequence types (STs) and nosocomial P. aeruginosa infections or colonization have generally been characterized by polyclonality [14]. In this study we only identified two sequence types among the MBL-producing P. aeruginosa, of which ST111 was the major type found. ST111 belongs to the international clonal complex CC111 (serotype O12, corresponding to CC/BURST Group 4) and has mostly been described in Europe as the major European multi-resistant P12 clone that has been associated with the emergence, spread and persistence of MDR strains in hospitals, mainly ICUs [14–17]. ST446 seems less widespread than ST111, but has also been described among MDR isolates in France and Spain [15,17].

The source and transmission modes of this particular P. aeruginosa strain in the Netherlands are currently unknown. The fact that most of the patients with an MBL-producing isolate were admitted to an ICU, a setting that is characterized by an increased risk of cross-transmission and high antimicrobial pressure [18], might have favoured clonal spread. Additionally, all Dutch hospitals are connected by referred patients who have the potential to facilitate dissemination of MDR organisms between hospitals [19]; subsequently, patients might serve as important reservoirs and transmission sources, stressing the importance of hand hygiene compliance, which was only 20% in a recent observational study in 24 Dutch hospitals [20], and patient precautions.

Since MBL-producing isolates can cause serious infections that are difficult to treat, their presence in various hospitals in the Netherlands is of nationwide concern. In the absence of new agents for the treatment of infections caused by these bacteria, the spread of this particular clone may lead to treatment failures with increased morbidity and mortality. Therefore, MBL-producing P. aeruginosa require rapid identification and the timely implementation of infection control measures in combination with systematic surveillance to monitor its potential clonal spread.

Acknowledgements

We thank Nicole Lemmens for her technical support of this study, Rosa van Mansfeld and Rob Willems for their help with the implementation of the MLVA, and Luc Sabbe, Joke Visser, Bernard Moffie, Hans Koeleman, Annemarie van der Aa, Mariëtte Muijsken, Carla Westra-Meijer, Ben Ridwan, Marjon van Merriënboer, Tjaco Ossewaarde, Jeroen Keijman, Bent Postma, Thuy-Nga Le and Ellen Oord for isolate and data collection. We acknowledge the infection control practitioners at the Erasmus University Medical Centre for their work on Pseudomonas aeruginosa. Results of this study were presented at the Interscience Conference on Antimicrobial Agents and Chemotherapy, Chicago, 20 September 2011.

Conflicts of Interest

Nothing to declare. No sources of funding were obtained for this study.

Transparency Declaration

JDD Pitout has previously received funding from Merck and Wyeth. All other authors none to declare.

Appendix

Appendix: Members of the MBL-PA Surveillance Study Group

Academic Medical Center, Department of Medical Microbiology, Amsterdam (M. Kolader).

Diagnostic Centre SSDZ, Department of Medical Microbiology, Delft (A. K. Van der Bij, R. W. Vreede).

Erasmus University Medical Centre, Department of Medical Microbiology and Infectious Diseases, Rotterdam (A. K. Van der Bij, W. H. F. Goessens, J. A. Severin, M. Van Westreenen, D. Van der Zwan).

Gelderse Vallei Hospital, Department of Medical Microbiology, Ede (R. W. Bosboom).

Groene Hart Hospital, Department of Medical Microbiology, Gouda (R. W. Vreede).

Isala Clinics, Laboratory of Medical Microbiology and Infectious Diseases, Zwolle (M. J. H. M. Wolfhagen).

St Jansdal Hospital, Department of Medical Microbiology, Harderwijk (P. C. R. Godschalk).

Leiden University Medical Center, Department of Medical Microbiology, Leiden (K. E. Veldkamp).

Laboratory for Infectious Diseases, Groningen (A. Ott).

Maasstad Hospital, Maasstad Laboratory, Rotterdam (A. Van der Zee).

Meander Medical Center, Department of Medical Microbiology, Amersfoort (P. C. R. Godschalk).

Orbis Medical Center, Department of Medical Microbiology and Infection Control, Sittard-Geleen (E. R. Heddema).

Regional Laboratory of Public Health, Haarlem (B. M. W. Diederen).

Rijnstate Hospital, Department of Medical Microbiology and Immunology, Arnhem (R. W. Bosboom).

Slingeland Hospital, Department of Medical Microbiology, Doetinchem (R. W. Bosboom).

University of Calgary, Department of Pathology and Laboratory Medicine, Microbiology and Infectious Diseases, Calgary, Alberta, Canada (G. Peirano, J. D. D. Pitout).

VieCuri Medical Center, Department of Medical Microbiology, Venlo (J. De Vries).

Ancillary