HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

This Article Abstract Full Text (PDF) Alert me when this article is cited Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via CrossRef Citing Articles via Google Scholar Google Scholar Articles by Kuriyama, T. Articles by Hosein, I. K. Search for Related Content PubMed PubMed Citation Articles by Kuriyama, T. Articles by Hosein, I. K. Agricola Articles by Kuriyama, T. Articles by Hosein, I. K.

Molecular characterization of clinical and environmental isolates of vancomycin-resistant Enterococcus faecium and Enterococcus faecalis from a teaching hospital in Wales

¹Department of Oral Surgery, Medicine and Pathology, Dental School, University of Wales College of Medicine, Heath Park, Cardiff CF14 4XY, UK ²Department of Infection Prevention and Control, Cardiff and Vale NHS Trust, Heath Park, Cardiff CF14 4XN, UK

Correspondence David W. Williams williamsdd{at}cardiff.ac.uk

Received November 15, 2002
Accepted June 2, 2003

The present study describes the first molecular characterization of environmental and clinical isolates of vancomycin-resistant enterococci (VRE) in Wales. Over a 3-month period (May–July 2000), 134 isolates of VRE (89 Enterococcus faecium and 45 Enterococcus faecalis) were isolated from the patient environment of the University Hospital of Wales (UHW) in Cardiff, Wales, UK. In addition, over the same time-period, 24 clinical isolates of VRE (20 isolates of E. faecium and four isolates of E. faecalis) were obtained from 14 patients. All study isolates were subjected to PFGE typing and their van genotypes were determined by using multiplex PCR. The vanA PCR product (231 bp) was evident in 146 (92 %) of 158 VRE isolates; the remaining 12 isolates (8 %) were positive for the vanB gene. All isolates of E. faecalis were found to be vanA-positive. In total, 16 PFGE banding profiles (pulsotypes) were observed for environmental isolates of E. faecium, whilst eight pulsotypes were found for isolates of E. faecalis. Some of these pulsotypes were isolated from multiple sites, whereas others were more restricted in their distribution. Eleven pulsotypes were evident for clinical isolates and eight of these (representing 11 isolates) were also encountered in environmental isolates. Eleven clinical isolates of E. faecium (55 %) shared an identical pulsotype that was not detected in environmental isolates. These results demonstrate a heterogeneous environmental population of VRE and an association of certain strains with clinical isolates. Predominance of a single pulsotype (not detected in the environment) amongst clinical isolates suggests non-environmental transmission between patients.

	INTRODUCTION

TOP INTRODUCTION METHODS RESULTS DISCUSSION ACKNOWLEDGEMENTS REFERENCES

 
Enterococci are important nosocomial pathogens that are often recovered from patients with urinary tract infections, wounds, bacteraemia, endocarditis or meningitis (Murray et al., 1999). Enterococci exhibit intrinsic resistance to several antibiotics and have the ability to acquire antibiotic resistance rapidly (Hayden, 2000). Whilst a treatment regimen with penicillins, alone or in combination with aminoglycosides, is recommended to combat enterococcal infections initially, vancomycin is the antibiotic of choice for infections caused by strains that are resistant to these antibiotics (Murray, 2000). Detection of vancomycin-resistant enterococci (VRE) in the mid-1980s was therefore a major therapeutic concern. VRE were first recovered in France and England (Cetinkaya et al., 2000; Hayden, 2000) but strains have subsequently been reported on a worldwide scale (Goossens, 1998; Cetinkaya et al., 2000; Hayden, 2000). Acquired resistance to vancomycin by enterococci greatly reduces the number of treatment options for disease management (Hayden, 2000; Mundy et al., 2000) and the problem is further compounded by the fact that resistance genes can potentially be transferred to other pathogenic organisms, such as Staphylococcus aureus and Streptococcus species (Hayden, 2000; Murray, 2000). It is also accepted that prevention of infection and control strategies for VRE have been of limited benefit once a clinical problem has become established (Hayden, 2000; Mundy et al., 2000).

Seven genotypes (vanA, vanB, vanC1, vanC2/3, vanD, vanE and vanG) of vancomycin resistance have been reported for enterococci to date (McKessar et al., 2000). vanA and vanB are the principal resistance genotypes reported for Enterococcus faecium and Enterococcus faecalis, the species isolated most frequently from clinical sites, and are an acquired form of resistance (Murray et al., 1999). Strains with the vanA genotype are characterized by high-level vancomycin and teicoplanin resistance, whereas those with the vanB genotype exhibit moderate to high resistance to vancomycin only (Cetinkaya et al., 2000; Murray, 2000).

A patient can acquire VRE from a range of sources (Weber & Rutala, 1997; Falk et al., 2000; Hayden, 2000) and the hospital environment may be an important reservoir (Weber & Rutala, 1997; Hayden, 2000). Epidemiological analysis of VRE has been used to elucidate the movement of VRE between patients and their environment and the extent of clonal transmission within a hospital setting (Cetinkaya et al., 2000). There is still a need for data that clarify the role of the environment in transmission of VRE to patients within a hospital setting.

In 1995, six index patients were found to be culture-positive for VRE at the University Hospital of Wales (UHW), a 920-bed teaching hospital and tertiary care referral centre. Since 1995, there has been an annual increase in isolation of VRE in patients at UHW; 94 patients were found to be positive for VRE in 2001. Between 1995 and 2001, 393 VRE-positive patients have been reported.

PFGE has been used previously with success as a molecular typing technique for VRE (Gordillo et al., 1993; Gambarotto et al., 2000; Kawalec et al., 2000). The aim of the present study was to use PFGE to characterize isolates of E. faecium and E. faecalis from the environment (wards) and patients in UHW over a specific time-period in 2000. The van genotypes of the VRE isolates were also determined, by using multiplex PCR.

	METHODS

TOP INTRODUCTION METHODS RESULTS DISCUSSION ACKNOWLEDGEMENTS REFERENCES

 
Isolates of VRE.

VRE isolates were obtained between 1 May and 31 July 2000. Clinical isolates were recovered from blood, sputum, urine, bile and dialysis fluids from patients in UHW over this time-period. Environmental isolates were obtained from selected sites within the Intensive Care Unit, Surgical High-Dependency Unit, haematology ward and nephrology unit at UHW. A saline-moistened sterile swab was applied to a variety of surfaces to collect environmental isolates (Table 1). Each swab was rolled five times over the selected sampling site. Swabs were inoculated on to Slanetz and Bartley agar and Staphylococcus/Streptococcus agar (both from Oxoid). All plates were incubated aerobically at 35 °C for 48 h.

Identification of VRE.

Enterococci were identified in accordance with American Society for Microbiology (ASM) recommendations (Murray et al., 1999). Resistance of each isolate to vancomycin was determined by the disc-diffusion method recommended by the National Committee for Clinical Laboratory Standards (2000).

Molecular typing of VRE by using PFGE.

Test isolates were subcultured on blood agar for 4 h at 37 °C. DNA was prepared by using a previously described method (Bartie et al., 2000). Restriction enzyme SmaI (New England BioLabs) was used as described previously (Bartie et al., 2000). PFGE parameters comprised ramp pulse times of 1–35 s for 30 h at 6 V cm^-1. Comparison of banding patterns was done by visual analysis (Tenover et al., 1995) and through the use of GelCompar software (Applied Maths; Bartie et al., 2000).

Characterization of vancomycin-resistance genotype.

Identification of van genotypes (vanA or vanB) for each isolate of VRE was performed by using multiplex PCR as described by Bell et al. (1998).

	RESULTS

TOP INTRODUCTION METHODS RESULTS DISCUSSION ACKNOWLEDGEMENTS REFERENCES

 
Isolates of VRE

In total, 24 clinical isolates of VRE (20 isolates of E. faecium and four of E. faecalis), obtained from 14 patients, were studied (Table 1). All patients were from the Intensive Care Unit, haematology ward or nephrology unit of UHW. A total of 134 isolates of VRE (89 isolates of E. faecium and 45 isolates of E. faecalis) was recovered from targeted environmental sampling sites (Table 1).

PFGE

All study isolates were amenable to typing by using the PFGE methodology described (Fig. 1). Sixteen banding profiles (pulsotypes) were identifiable by visual analysis of environmental E. faecium isolates (n = 89; Table 1). In the case of the 45 isolates of E. faecalis, eight pulsotypes were evident (Table 1). Certain pulsotypes (e.g. E. faecium C5 and E. faecalis A4) were noted for isolates recovered from multiple environmental sites within the hospital, whereas others were more limited in their distribution (e.g. E. faecium C11 and E. faecalis A5) and were often restricted to individual wards.

View larger version (92K): [in this window] [in a new window]   Fig. 1. PFGE of SmaI-restricted genomic DNA of VRE. Lanes contain DNA from the following strains (origin, pulsotype): 1, Q41 (patient, C17); 2, Q48 (patient, A6); 3, e30 (environment, A6); 4, e14 (environment, C12); 5, G75 (patient, C12); 6, H3 (patient, C12); 7, e55 (environment, C4); 8, P35 (patient, C19); 9, e117 (environment, C4); M, DNA marker.

In total, 11 pulsotypes were encountered from the 24 clinical VRE isolates. Eight of these pulsotypes, representing 11 isolates (P3, K28, K95, I95, K30, G75, H3, K29, Q48, L94 and J24) were identical to some of the environmental pulsotypes (C2, C9, C13, C4, C12, A6, A3 and A4; Fig. 2). Of the remaining 13 clinical E. faecium isolates, 11 were found to share an identical pulsotype (C17) that was not evident for any of the environmental isolates. Two clinical isolates (E. faecium P35 and E. faecium N95) had unique pulsotypes.

View larger version (48K): [in this window] [in a new window]   Fig. 2. Correlation of pulsotypes from 24 clinical (bold) and 22 environmental VRE isolates. Arithmetic mean dendrograms were generated from similarity values determined by the Dice coefficient. ICU, Intensive Cure Unit. Environmental VRE pulsotypes represent all previous visually distinct pulsotypes (C1–C16, A1–A6).

van genotype analysis

Multiplex PCR detected either the vanA or vanB genotype for all VRE isolates studied. The vanA PCR product was represented by a 231 bp fragment, which was clearly distinct from the 330 bp vanB product. The vanA genotype predominated (146 of 158 isolates; 92 %) and was the only van genotype encountered for isolates of E. faecalis.

	DISCUSSION

TOP INTRODUCTION METHODS RESULTS DISCUSSION ACKNOWLEDGEMENTS REFERENCES

 
It has been suggested that hospitalized patients can become colonized by VRE from the local environment (Weber & Rutala, 1997; Hayden, 2000). However, there have not been any studies to assess the molecular relationship between environmental and clinical isolates from hospitalized patients in Wales. Clarification of the patterns of VRE dissemination in the hospital environment and determination of genetic similarity between isolates could provide information on possible routes of transmission.

Although amplified fragment-length polymorphism analysis and multilocus sequence typing schemes have recently been developed for typing enterococci (Willems et al., 2001; Homan et al., 2002), PFGE remains the most widely used tool for enterococcal typing (Tenover et al., 1995). Several methodologies for PFGE analysis of enterococci have been reported (Gordillo et al., 1993; Gambarotto et al., 2000; Kawalec et al., 2000), although experience gained during the present study revealed that suggested procedures were inconsistent with regard to lysis of some VRE strains (data not shown). Previous studies have recommended that test cultures should be incubated on blood agar at 37 °C for 16–24 h to generate plugs for PFGE. However, this incubation time was reduced to 4 h in the present study, which led to a noticeable improvement in DNA extraction and subsequent band intensity. Furthermore, initial evaluation of the band profiles obtained by using electrophoresis times of either 24 or 30 h revealed that 30 h yielded optimal resolution of banding patterns.

Molecular typing of clinical VRE strains has shown that VRE isolates obtained from hospitalized patients in Europe tend to be heterogeneous with respect to their PFGE fingerprints (Goossens, 1998; Willems et al., 2001). This observation is in contrast to some reports from the USA, where greater homogeneity was found (Goossens, 1998; Falk et al., 2000). The current study represents the first report of characterization and epidemiology of VRE in a Welsh hospital.

Over the sampling period, 134 VRE isolates (E. faecium, n = 89; and E. faecalis, n = 45) were obtained from a range of inanimate environmental surfaces. The predominance of E. faecium over E. faecalis is in agreement with other epidemiological analyses of VRE populations (Gambarotto et al., 2000). As VRE have been shown to survive on environmental surfaces for time-periods in excess of 5 weeks (Falk et al., 2000), there may be significant potential for their transmission to patients.

PFGE analysis of the environmental VRE isolates identified 16 pulsotypes for E. faecium and five for E. faecalis. These results indicate that a variety of VRE strain types are present within the UHW environment; this is consistent with other European hospitals. Several VRE types (E. faecium C5; E. faecalis A1, A2 and A4) were found at two or more locations in the hospital and could be indicative of local transmission within UHW. It is interesting that the outlying of patients has increased in UHW over recent years; this could play a contributory role in environmental transmission of VRE. In contrast, other VRE types (C1–4, C6–16, A3, A5 and A6) were found only at specific sites in the hospital, although localized spread of these strains was evident (e.g. strain A3 was located at several sites in a nine-bedded room in nephrology).

During the sampling, 24 clinical strains of VRE were isolated from 14 patients. PFGE analysis revealed that 11 of these clinical isolates (P3, K28, K95, I95, K30, G75, H3, K29, Q48, L94 and J24) were genetically similar to environmental isolates (C2, C9, C13, C4, C12, A6, A3 and A4). Therefore, it is likely that spread of VRE occurs between patients and the environment, although it is difficult to ascertain whether or not the environment was responsible for infecting these patients or vice versa. It was interesting to note that nine of 11 clinical VRE isolates that were similar to environmental isolates were from patients within the nephrology unit. The remaining two clinical isolates that shared profiles with environmental strains were from one patient who was attending the haematology ward and one patient in the Intensive Care Unit. The reason for the high concordance between clinical and environmental isolates in the nephrology unit is unclear. It is possible that that as there is a tendency for patients with chronic renal disease to spend longer and more frequent time-periods within the hospital setting, a greater likelihood of exposure to environmental VRE exists.

The remaining 13 clinical isolates (recovered from five patients) had pulsotypes that were distinct from those of environmental isolates. The origin of these VRE isolates is uncertain, but they may represent strains that were acquired previously from the community setting (Bonten et al., 2001). However, 11 of the clinical isolates (from four of five patients) had an identical pulsotype (C17). All strains with this pulsotype were isolated from patients on the haematology ward, although one of these patients also spent time in the Intensive Care Unit. This suggests probable patient-to-patient transfer, possibly via contact with health-care workers or patient visitors. It was also noted over this sampling period that visitors caring for relatives had frequent contact with other patients on the haematology ward. Further studies are required to confirm these possible routes of transmission.

VRE isolates were found to possess either the vanA or vanB genotype. There was a predominance of the vanA genotype (92 % of all isolates), which is consistent with the majority of findings from other countries (Bell et al., 1998; Goossens, 1998; Kawalec et al., 2000; Scagnelli et al., 2001) with the exception of Australia, where a higher incidence of vanB isolates has been reported (Bell et al., 1998). Although vanB VRE were encountered rarely in the present investigation, it was interesting to note that all vanB-positive isolates encountered were of pulsotype C4. This finding would not only support the reliability of the PFGE typing procedure, but would also further substantiate the dissemination of this particular strain and the ‘environment-to-patient’ link.

	ACKNOWLEDGEMENTS

TOP INTRODUCTION METHODS RESULTS DISCUSSION ACKNOWLEDGEMENTS REFERENCES

 
We would like to acknowledge Haru Kato (National Infectious Institute, Tokyo) and Polly Kaufmann (PHLS, Colindale, London, UK) and her colleagues for their advice regarding PFGE. We would also like to thank Gillian Beswick, Emma Trevorrow and Debbie Richards (Department of Infection Control, Cardiff and Vale NHS Trust, Cardiff) for their assistance with the collection and identification of VRE isolates.

	REFERENCES

TOP INTRODUCTION METHODS RESULTS DISCUSSION ACKNOWLEDGEMENTS REFERENCES

 

Kawalec, M., Gniadkowski, M. & Hryniewicz, W. (2000). Outbreak of vancomycin-resistant enterococci in a hospital in Gdask, Poland, due to horizontal transfer of different Tn1546-like transposon variants and clonal spread of several strains. J Clin Microbiol 38, 3317–3322.[Abstract/Free Full Text]