European Journal of Clinical Microbiology & Infectious Diseases

, Volume 31, Issue 3, pp 349–356

Prevalence of community-associated meticillin-resistant Staphylococcus aureus and Panton–Valentine leucocidin-positive S. aureus in general practice patients with skin and soft tissue infections in the northern and southern regions of The Netherlands

Authors

    • Laboratory for Infectious Diseases
    • Martini Hospital
  • M. I. A. Rijnders
    • Department of Medical MicrobiologyMaastricht University Medical Centre, School for Public Health and Primary Care (CAPHRI)
  • B. M. Roede
    • The National Institute for Public Health and the Environment
  • E. Stobberingh
    • Department of Medical MicrobiologyMaastricht University Medical Centre, School for Public Health and Primary Care (CAPHRI)
  • A. V. M. Möller
    • Laboratory for Infectious Diseases
    • Martini Hospital
Article

DOI: 10.1007/s10096-011-1316-9

Cite this article as:
Mithoe, D., Rijnders, M.I.A., Roede, B.M. et al. Eur J Clin Microbiol Infect Dis (2012) 31: 349. doi:10.1007/s10096-011-1316-9

Abstract

The purpose of this investigation was to determine the prevalence of community-associated meticillin-resistant Staphylococcus aureus (CA-MRSA) and Panton–Valentine leucocidin (PVL)-positive S. aureus in general practice (GP) patients with skin and soft tissue infections (SSTI) in the northern (Groningen and Drenthe) and southern (Limburg) regions of The Netherlands. Secondary objectives were to assess the possible risk factors for patients with SSTI caused by S. aureus and PVL-positive S. aureus using a questionnaire-based survey. From 2007 to 2008, wound and nose cultures were obtained from patients with SSTI in general practice. These swabs were analysed for the presence of S. aureus and the antibiotic susceptibility was determined. The presence of the PVL toxin gene was determined by polymerase chain reaction (PCR) and the genetic background with the use of spa typing. A survey was performed to detect risk factors for S. aureus infection and for the presence of PVL toxin.S. aureus was isolated from 219 out of 314 (70%) patients with SSTI, of which two (0.9%) patients were MRSA-positive. In 25 (11%) patients, the PVL toxin gene was found. A higher prevalence of PVL-positive S. aureus of patients with SSTI was found in the northern region compared to the south (p < 0.05). Regional differences were found in the spa types of PVL-positive S. aureus isolates, and for PVL-negative S. aureus isolates, the genetic background was similar in both regions. The prevalence of CA-MRSA in GP patients with SSTI in The Netherlands is low. Regional differences were found in the prevalence of PVL-positive S. aureus isolates from GP patients with SSTI. Household contacts having similar symptoms were found to be a risk factor for SSTI with S. aureus.

Introduction

Until recently, meticillin-resistant Staphylococcus aureus (MRSA) was commonly found in hospitals (hospital-acquired MRSA; HA-MRSA). In the past decade, there have been numerous reports worldwide of MRSA strains circulating outside the hospital setting in individuals without a previous history of hospital admissions. These strains were named community-associated (CA-MRSA) or community-onset MRSA (CO-MRSA). In the United States especially, there was a rapid increase in the incidence of CA-MRSA. In 2004, a large multi-centre study of emergency wards in 11 academic hospitals showed that 59% of patients with a skin infection were MRSA-positive [1]. In 97% of these cases, the causative MRSA belonged to pulsed field gel electrophoresis (PFGE) strain type USA300 with multilocus sequence type (MLST) ST8, of which 98% was positive for the Panton–Valentine leucocidin (PVL) toxin [1]. Clinical studies propose the exotoxin PVL to be a crucial virulence factor in necrotising infections [2, 3]. PVL is a two-component pore-forming toxin, which mainly acts on neutrophils [4]. It is expressed by only a small percentage of S. aureus wild-type isolates (2–3%) [5], but it is highly prevalent in S. aureus strains isolated from necrotising infections [2, 3]. In Europe, a similar increase in the incidence of cases with CA-MRSA was found [6]. The majority of these cases were related to one MRSA strain, e.g. MLST ST80, spa type t044, which contains the cassette chromosome SCCmec type IV and was PVL-positive. This particular strain produces the cytotoxin PVL, staphylococcal enterotoxin H (SEH), but is negative for other toxins (SEA, SEB, SEC, SED, SEE, ETA and TSST-1).

The prevalence of MRSA is generally low in The Netherlands compared to other European countries [7, 8], and differs considerably with its neighbouring countries Belgium and Germany, with a prevalence of hospital-isolated MRSA of 23.6 and 13.8%, respectively [9] . The increased prevalence in Germany is due to an increase in the prevalence of PVL-positive ST80 and the emergence of a USA300-like ST8 PVL-positive clone, most likely imported from the USA [10]. An increase in the incidence of MRSA was reported in 2003 from the northern regions of The Netherlands (Groningen and Drenthe) [11]. This increase was mainly caused by PVL-positive MRSA. In 2003, 19% and in 2004, 22% of the MRSA isolates were PVL toxin-positive and, of these, 65% belonged to the ST80 clone [11]. The national prevalence of ST80 among MRSA from 2003 to 2004 was 20% [12]. These data suggest that the prevalence of PVL toxin-positive MRSA in the north of The Netherlands is higher than the national prevalence (8%) [12, 13]. The relatively high number of CA-MRSA isolates may be ascribed to the local spread of a ST80 strain in the community. A potential source for MRSA or risk factors contributing to the high occurrence of PVL-positive MRSA in this specific region are unknown. Furthermore, there is limited information available regarding the prevalence and epidemiology of CA-MRSA and PVL toxin-positive S. aureus in The Netherlands.

Because of the epidemic potential of CA-MRSA and the risk of acquisition of additional virulence factors, together with the occurrence of meticillin resistance in PVL toxin-positive S. aureus isolates [10, 14], additional measures are needed in order to prevent and monitor the spread of these strains in the community. Therefore, this study aims to assess the prevalence of CA-MRSA and PVL toxin-positive S. aureus strains in patients with skin and soft tissue infections (SSTI) receiving medical care from general practitioners in two regions (Groningen and Drenthe in the north and Limburg in the south) in The Netherlands. Secondary objectives were to determine the prevalence of PVL toxin-positive S. aureus strains and to assess the risk factors of patients with skin infections caused by S. aureus with the use of a questionnaire-based survey.

Materials and methods

Population and study design

The study area consisted of the northern regions, Groningen and Drenthe (area of 5,639 km2; 1,058,407 inhabitants), and the southern region, Limburg (area of 2,209 km2; 1,122,702 inhabitants) of The Netherlands. In the region of Groningen and Drenthe (further referred to as “Groningen”), 522 of the 526 (99%) general practitioners were contacted to participate in this study, and 84 of the 218 (38%) general practitioners were contacted in the region of Maastricht in Limburg (henceforth, referred to as “Limburg”).

Patients were included from January 2007 to January 2008. Inclusion criteria for the study consisted of patients with SSTI that could be related to S. aureus infections, such as furuncles, carbuncles, abscesses and other local skin infections. The case definition was based upon the International Classification of Primary Care (ICPC) codes [15]. Participation in the study was on a voluntary basis. If patients were underaged, permission was first sought from their parents or caretakers.

After the patient fulfilled the inclusion criteria, agreed to participate in the study and informed consent was obtained, nose and wound cultures were taken. In addition, the general practitioner and the patient filled in the questionnaire. The questionnaire was defined de novo and designed by the National Institute for Public Health and Environment (RIVM), which also serves as the reference laboratory for MRSA in The Netherlands. The questionnaire was based on expertise in the topic and risk factors for MRSA infection as known from the literature [1, 16].

Bacterial culturing

In both regions, the nose and wound cultures were obtained using sterile cotton swabs (Amies agar gel single swab; Copan Italia S.p.A., Brescia, Italy). The swabs were processed at the Laboratory for Infectious Diseases in Groningen for the northern region and at the Maastricht University Medical Centre for the southern region according to local protocols. Putative S. aureus colonies were identified as S. aureus using Gram-stain, catalase and coagulase testing, as described previously [17].

Antimicrobial susceptibility testing

The antimicrobial susceptibility patterns were determined for all S. aureus isolates using a microbroth dilution method with Mueller–Hinton II cation-adjusted broth (Becton Dickinson, 212322, Breda, The Netherlands), according to the guidelines of the Clinical Laboratory Standards Institute (CLSI) [18]. The antibiotics tested were as follows (ranges expressed in mg/L): cefaclor, 0.06–128; cefuroxime, 0.06–128; clindamycin, 0.03–64; ciprofloxacin, 0.128–32; erythromycin, 0.03–64; gentamicin, 0.06–64; imipenem, 0.03–64; linezolid, 0.03–64; meropenem, 0.06–64; moxifloxacin, 0.12–4; oxacillin, 0.03–64; penicillin, 0.004–8; rifampin, 0.008–16; teicoplanin, 0.06–128; tetracycline, 0.03–64; sulphamethoxazole–trimethoprim, 0.015–32; and vancomycin, 0.06–128. Due to the lack of breakpoints of fusidic acid and mupirocin in the CLSI guidelines for susceptibility testing, the BSAC guideline for Standardized Disc Susceptibility Testing was used for these two antibiotics [19]. For fusidic acid, a zone diameter of 30 mm was defined as susceptible and for mupirocin, a zone diameter of 14 mm was used [20, 21].

Confirmation of meticillin resistance

In Groningen, meticillin resistance was confirmed by the penicillin-binding protein (PBP) 2’ latex agglutination test (Denka Seiken, Tokyo Japan) and, in addition, the mecA gene was identified using the Genotype MRSA test (Hain Lifescience GmbH, Nehren, Germany). In Limburg, the oxacillin-resistant isolates were analysed for the presence of the mecA gene using a real-time polymerase chain reaction (PCR) assay [22]. Confirmed mecA-positive isolates were considered as MRSA.

Molecular typing methods

The genetic background of all S. aureus isolates was determined by spa typing as previously described [23]. The spa clonal complexes (CCs) were identified with the algorithm based upon repeat patterns (BURP). spa typing, together with the BURP, yields results that are in concordance with the typing results obtained by MLST [23]. The associated MLST CCs were linked with the spa CCs through the Ridom SpaServer (http://spaserver.ridom.de). The presence of the PVL and toxic shock syndrome toxin 1 (TSST-1) genes was determined using a real-time PCR assay as previously described [24].

Statistical methods

This study aims to assess the prevalence of CA-MRSA and PVL toxin-positive S. aureus strains in general practice (GP) patients with SSTI in Groningen and Limburg. A power analysis was performed before initiation of the study to determine the sample size and, in line with this, the number of GPs required for inclusion in case of an assumptive 2% MRSA rate. The required number of participating GPs depends on the incidence of SSTI and MRSA-positive isolates per practice, and will, therefore, have a broad range.

Assuming that 10% of the total number of GPs representing 60 GP practices caring for 116,580 patients participate in the study, this would account for 1,131 to 2,051 patients with SSTI (based on incidence of 0.97–1.76%) [2527], of which 60% would be S. aureus-positive, n = 678–1,231. The number of expected MRSA isolates found should be between 0.2 and 18 (prevalence 0.03–1.5%) [28]. Based on a previous study reporting 1.6% PVL-positive S. aureus strains from the community [29], we assumed that, in this study with a selection of SSTI patients, 10% of S. aureus strains would be PVL-positive; hence, between 68–123 patients.

The questionnaires were analysed to detect risk factors for S. aureus infection and the presence of the PVL toxin gene. The two groups of patients with and without positive S. aureus wound cultures were compared using the Chi-square test or Mann–Whitney test, as appropriate. Statistical analyses were performed using SPSS v.17.0 software (SPSS Inc., Chicago, IL, USA).

Results

Patients included and diagnoses

In Groningen, 141 GPs responded, of whom 90 agreed to participate in the study. After one year, 54 of the 90 (10%) participating GPs in Groningen had included patients. Each GP included, on average, three patients. In Limburg, 84 GPs were contacted, of whom 34 agreed to participate. Eventually, 22 GPs (26%) included, on average, seven patients. During the 1-year study period, they included 314 patients with SSTI, 160 patients in Groningen and 154 patients in Limburg. GPs included, on average, four patients each, which was lower than expected.

The most frequently found diagnoses were impetigo (S84), 60/314 (19%), post-traumatic skin defects (S11), 43/314 (14%), furuncles/carbuncles (S10), 40/314 (13%), and other unclassified skin conditions (S99), 30/314 (10%).

Culture results

S. aureus was isolated from 219 out of 314 (70%) patients from Groningen and Limburg, in 193 (61%) patients from the wound swabs, and from the nose swabs only in 160 (51%) patients. The positive predictive value of a positive nose swab for wound culture was 85%. Two of the 219 patients (0.9%) with positive S. aureus culture, both of which were patients from Groningen, were MRSA-positive with PVL-positive strains. One of the MRSA strains belonged to PFGE cluster 46c (spa type t008, SCCmec IVc) and the other had PFGE type 267 (spa type t786).

In total, 25 out of 219 (11%) patients with positive S. aureus culture were positive for the PVL toxin gene; in Groningen, 18 (out of 160, 11%) compared with 7 (out of 154, 5%) in Limburg (p < 0.05). Approximately 50% of the PVL toxin-positive S. aureus were cultured from patients with furuncles and carbuncles.

Questionnaire

A questionnaire was obtained from all 314 patients included in the study. Analysis of the questionnaires showed differences in risk factors between patients with and without S. aureus infections. Patients with S. aureus infections were younger than patients who did not have a positive wound culture with S. aureus and patients with household contact with animals had significantly more positive S. aureus wound cultures. For patients with similar previous complaints, the difference was borderline significant (Table 1), but they had PVL toxin-positive isolates more frequently than patients who did not have previous complaints (Table 2). Remarkable was the difference with respect to working in a healthcare institute, where significantly fewer patients had an S. aureus-positive wound culture (p < 0.05), and none of these were PVL-positive. Both patients with positive MRSA cultures had no known risk factors for MRSA.
Table 1

Patient characteristics, risk factors and the presence of Staphylococcus aureus infection

Patient characteristics and risk factors

Wound culture

p-value

S. aureus-positive; n = 193

S. aureus-negative; n = 121

Sex (male)

95 (64)

54 (36)

0.51

Age mean (standard deviation)

27 (23)

36 (26)

<0.05a

Former complaints

85 (63)

51 (37)

0.74

Chronic disease

39 (60)

26 (40)

0.79

Works in healthcare institute

10 (40)

15 (60)

0.02*

Housemate works in healthcare institute

21 (57)

16 (43)

0.53

Others with same symptoms

39 (80)

10 (20)

<0.05*

Sport with physical contact

41 (65)

22 (35)

0.51

Professional contact with animals

7 (70)

3 (3)

057

Household contact with animals

100 (69)

46 (31)

<0.05*

Contact with animals with symptoms

7 (41)

10 (59)

0.08

Stay in hospital last year

20 (49)

21 (51)

0.07

Stay in nursing home last year

4 (57)

3 (43)

0.81

Attendance of daycare in hospital

57 (61)

37 (39)

0.84

Foreign travel last year

75 (62)

46 (38)

0.88

Antibiotic usage last year

78 (59)

55 (41)

0.38

at-test (age)

*Significant p < 0.05

Table 2

Patient characteristics and the presence of Panton–Valentine leucocidin (PVL) toxin-positive S. aureus infection

 

PVL-positive S. aureus; n = 25

PVL-negative S. aureus; n = 194

p-value

Sex (male)

12 (48)

136 (47)

0.92

Age (median, interquartile range)

29 (18;42)a

25 (8;48)a

0.83

Former complaints

15 (60)

121 (42)

0.08

Chronic disease

2 (8)

63 (22)

0.10

Works in healthcare institute

0

25 (9)

0.13

Housemate works in healthcare institute

0

37 (13)

0.06

Others with same symptoms

10 (40)

39 (14)

<0.01*

Sport with physical contact

4 (16)

59 (20)

0.60

Professional contact with animals

1 (4)

9 (3)

0.81

Household contact with animals

13 (52)

133 (46)

0.57

Contact with animals with symptoms

0

17 (6)

0.21

Stay in hospital last year

5 (20)

36 (13)

0.28

Stay in nursing home last year

1 (4)

6 (2)

0.53

Attendance of daycare in hospital

8 (32)

86 (30)

0.81

Foreign travel last year

13 (52)

108 (37)

0.15

Antibiotic usage last year

15 (60)

118 (41)

0.06

aMann–Whitney test (age)

*Significant p < 0.05

Molecular analysis

The distribution of spa CCs for all S. aureus (PVL-positive and PVL-negative) isolates analysed in both regions is shown in Fig. 1. A total of 23 different spa types were found with molecular typing of the PVL-positive S. aureus isolates (Table 3). These 23 spa types from PVL-positive S. aureus isolates were clustered into spa CCs. One cluster belonged to the spa CC122, which is related to MLST CC30 and consisted of nine isolates and seven different spa types. The other cluster, for which no founder could be identified, was associated with MLST- CC121 and consisted of three isolates with two spa types. Ten of the 23 isolates could not be clustered and were, therefore, classified as singletons. Singletons or unknown types that could not be clustered in an spa CC were not allocated in an MLST CC. One isolate could not be clustered because it contained less than four repeats. Clustering of the spa types of PVL toxin-negative S. aureus isolates showed no regional differences in the genetic background. The most commonly encountered spa CCs of all S. aureus isolates were spa CC659 and spa CC084 (Fig. 1).
https://static-content.springer.com/image/art%3A10.1007%2Fs10096-011-1316-9/MediaObjects/10096_2011_1316_Fig1_HTML.gif
Fig. 1

Distribution of spa clonal complexes (CCs) of all Staphylococcus aureus isolates in Groningen and Limburg

Table 3

spa types of PVL toxin-positive S. aureus and clinical condition

Limburg

Groningen

ICPC code

Clinical condition

spa types

Associated MLST type

ICPC code

Clinical condition

spa types

Associated MLST type

S99

Other unclassified skin conditions

t1445

CC12

Sl0

Furuncles/carbuncles

t1211

Unknown

S84

Impetigo

t363

CC30

Sl0

Furuncles/carbuncles

t276

CC30

S11

Post-traumatic skin defects

t659

CC121

Sl0

Furuncles/carbuncles

t964

CC30

S02

Itching

t084

CC15

Sl0

Furuncles/carbuncles

t342

CC30

S11

Post-traumatic skin defects

t1350

CC25

S04

Local swelling

t122

CC30

S10

Furuncles/carbuncles

t359

CC8

S09

Infection finger/toe

t1202

CC30

S84

Impetigo

t903

CC1153

Sl0

Furuncles/carbuncles

t276

CC30

    

Sl0

Furuncles/carbuncles

t159

CC121

    

S84

Impetigo

t159

CC121

    

Sl0

Furuncles/carbuncles

t005

CC22

    

S09

Infection finger/toe

t318

CC30

    

S03

Warts

t008

CC8

    

S09

Infection finger/toe

t2304

Unknown

    

Sl0

Furuncles/carbuncles

t3341

Unknown

    

S09

Infection finger/toe

t3342

Unknown

    

Sl0

Infection finger/toe

t122

CC30

    

S76

Other skin infection

t084

CC8

    

S09

Infection finger/toe

t4116

Unknown

Antibiotic use and resistance

Nearly 43% of the patients had used antibiotics in the previous year and 80% of these patients were able to be specific as to which antibiotic was used. During that year, 13% of the patients had used two or more commonly prescribed antibiotics, such as amoxicillin or amoxicillin with clavulanic acid (35%), macrolides (13%), flucloxacillin (14%) and fusidic acid (12%).

All S. aureus isolates showed susceptibility to vancomycin, ciprofloxacin and mupirocin. In total, 99% of the S. aureus isolates were susceptible to gentamicin and sulphamethoxazole–trimethoprim, and 98% of the S. aureus isolates were susceptible to oxacillin, clindamycin, cefaclor, cefuroxime, gentamicin, imipenem and meropenem. S. aureus isolates were resistant for penicillin, fusidic acid, clarithromycin and tetracycline in 76, 28, 11 and 7%, respectively.

Discussion

This study was performed because there were some indications concerning differences in the prevalence of CA-MRSA and PVL-positive S. aureus between both regions of The Netherlands [11].

In previous studies in The Netherlands, the prevalence of MRSA was 1% in hospitals [7] and <0.03% at hospital admission [25], suggesting that the prevalence of MRSA in hospitals and in the community is low, which is comparable to our findings.

The suspected increase of CA-MRSA in Groningen was not demonstrated [11] but a difference in the prevalence of PVL-positive S. aureus was found between both regions.

The prevalence of CA-MRSA in GP patients with SSTI was 0.9%. Both MRSA isolates were found in Groningen and were PVL-positive. One of these MRSA strains could have been imported by travel, since one patient was a frequent business traveller abroad. Dissemination of MRSA is possible after the introduction of MRSA clones associated with travel or through importation by the cross-border transfer of patients, particularly when patients are transferred from countries with a relatively high prevalence to a country with a low prevalence [9].

The spread of PVL-positive MRSA in the northern Netherlands was previously associated with a major MRSA clone ST80 strain, which has also been reported from Belgium and Germany [11, 30, 31].

Although this study includes locus-specific epidemiology in a country with a low MRSA rate, it has an important additional value compared to other MRSA studies because the prevalence in the community was studied instead of in a hospital population. Therefore, its relevance may be of importance for the cross-border dissemination of MRSA and PVL-positive S. aureus in neighbouring countries with a higher MRSA rate [31]. Ongoing dissemination of these strains in specific areas of the community may be explained by a carrier ship of individuals passing colonisation and infection back and forth.

Differences in patient characteristics (age) and risk factors were found between patients with and without S. aureus infections. Risk factors such as stay in hospital and contact with animals were both in agreement with existing trends in the literature data about transmission [1]. The mean age of index persons who did experience household transmission was significantly lower than the mean age of those without transmission (27 years vs. 36 years), which is in agreement with other literature data [32, 33]. Despite the absence of clear evidence, it seems most likely that the age-related risk factor for S. aureus transmission can be explained by the more crowded living conditions of households with younger patients [32, 33]. Since household contacts with the same symptoms as the index patient could serve as a reservoir for PVL-positive S. aureus, the reintroduction of S. aureus by these household contacts could lead then to recolonisation or infection of the index patient. In our study, 39% of the PVL-positive S. aureus isolates had a genetic background common to the MRSA lineage MLST CC30, which belongs to one of the five major (international) MRSA clusters [34]. These PVL-positive S. aureus isolates could serve as a potential reservoir for MRSA because of an SCCmec-susceptible genetic background and the presence of these closely related genetic strains may subsequently cause further spread of MRSA in the community.

Standard culturing of household contacts and providing eradication therapy for MRSA-positive household contacts is not included in the Dutch “Search-and-Destroy” infection control policy. This policy for hospitals and nursing homes to control the MRSA prevalence in The Netherlands has proven to be successful. However, adaptation of these guidelines to the local situation, especially in cross-border areas, remains important. The implementation of adapted guidelines could lead to a decrease of patients with infection and control of MRSA and PVL-positive S. aureus in the community [35]. Further studies in the community are needed in order to elucidate the transmission of CA-MRSA and PVL toxin-positive S. aureus, especially between household contacts.

Acknowledgements

We thank the technicians from the laboratory of medical microbiology of the Laboratory for Infectious Diseases in Groningen and the Maastricht University Medical Centre for the technical assistance and support.

Funding

This work was supported in part by a grant from the Ministry of Health, Welfare and Sports, The Netherlands.

Conflict of interest

None declared.

Ethical approval

Medical ethical approval was given by Mr. John W.P. de Vroedt, MHA, Maastricht University Medical Centre and University Hospital of Groningen. This study has been performed in accordance with the ethical standards laid down in the 1964 Declaration of Helsinki.

Copyright information

© Springer-Verlag 2011