Abstract

Introduction

Methicillin-resistant Staphylococcus aureus (MRSA) is known to be a major cause of infections worldwide in the hospital and community settings, and while hospital-associated MRSA (HA-MRSA) generally affect patients with predisposing risk factors (prolonged hospitalization, use of indwelling catheters or prior surgical procedures) [1], community-associated MRSA (CA-MRSA) generally affect healthy and younger people without such risk factors [2]. Moreover, HA-MRSA and CA-MRSA belong to distinct genetic lineages, and while HA-MRSA are mostly multidrug-resistant and carry the larger SCCmec types I, II, III, CA-MRSA isolates frequently carry smaller SCCmec elements, usually type IV and V, and are resistant to fewer classes of antimicrobials [3]. Also, CA-MRSA isolates are strongly associated with virulence factors such as Panton-Valentine leukocidin (PVL) and the arginine catabolic mobile element (ACME) which are thought to contribute to their pathogenic potential [4], [5], [6].

Nowadays, however, CA-MRSA has been increasingly identified as a cause of hospital-onset and healthcare associated infections [7], [8]. On the other hand, hospital-associated clones have been described to cause infections in the community [9], [10] suggesting that certain clones have the ability to cross barriers between hospitals and the community.

In Portugal, MRSA prevalence is one of the highest in Europe, having sharply increased in the last decade from 31.9% in 2001 to 52.2% in 2010, which contrasts with the decreasing trend in MRSA prevalence observed for other countries such as England, France and Belgium [11], [12].

On the other hand and in spite of its clinical relevance, the prevalence and epidemiology of MRSA causing infection in the Portuguese community are not known. A single study assessing the contemporary causes of skin and soft tissue infections, which are the most common clinical manifestation of CA-MRSA, showed that Portugal, among all European countries, had the highest prevalence of MRSA causing SSTI between 1998 and 2004 [13].

Additionally, MRSA colonization in healthy young people was found to be extremely low (0.24% in 1996–1998 and 0.13% in 2006–2009) [14], [15]. Furthermore, only sporadic cases of CA-MRSA infections have been reported [16], [17] in Portugal.

In this study we aimed to understand the epidemiologic phenomenon associated to the recent sharp increase in MRSA frequency and to evaluate how this has shaped the MRSA population structure in the community. For this purpose we analyzed the clonal nature of MRSA during two time periods (1993 and 2010) spanning 17 years in a major tertiary teaching hospital in Portugal and compared it with the population structure of MRSA causing skin and soft tissue infections of people attending healthcare centers in the last time period (2010/2011). We hope to understand the extension of the epidemiological links between hospital and community in a country with a high prevalence of MRSA.

Materials and Methods

Results

Epidemiological Links between the Hospital and the Community Settings

Comparing the populations of MRSA recovered in the community and hospital in 2010, it can be observed that some isolates from both settings share exactly the same molecular characteristics, namely PFGE type, SCCmec type, spa type and ST. This was the case of isolates belonging to ST22-IVh-t2357, ST22-IVh-t032, ST105-II-t002 and ST5-IVc-t535 that had the same genetic features independently of the fact that they were isolated in the hospital or in the community. In addition, five of these isolates showed an identical antimicrobial resistance profile.

Overall, our results suggest the occurrence of dissemination of typical HA-MRSA clones into the community (see Figure 1).

Open in a separate window

Figure 1

MRSA population structure in the hospital and community in Portugal.

Schematic representation of the MRSA population structure in the hospital and community settings and the possible dissemination of hospital clones into the community.

Discussion

In a major tertiary teaching hospital in Portugal, the frequency of MRSA has sharply increased from 30% in 1993 to 49% in 2010. This is a picture that mirrors the situation observed in most of the hospitals in Portugal (Faria et al., unpublished). However, the reasons lying behind this high prevalence are complex and were not explored in this study.

Our results showed that this steep increase in MRSA frequency was associated with a major clonal shift, in which the Portuguese and Iberian clones were replaced by EMRSA-15 and some less prevalent clones.

In fact, no common PFGE types, spa types (except for t008) or STs were found between the two study periods. The only CC that prevailed between the two study periods was CC8, although associated to distinct clones (the ST247-I and ST239-III variant in 1993 and ST2246-III and ST8-IV/VI in 2010). Moreover, a major change in the antibiogram profile was concomitantly observed between the two periods. Whereas in 1993 isolates were resistant to almost all antimicrobial agents tested, in 2010 the great majority of isolates was non-multidrug resistant (81.7%).

The establishment of the non-multidrug resistant clone ST22-IVh, or EMRSA-15, as the major clone in hospitals in Portugal, was previously reported [35], [36], suggesting that this clone has been in Portugal for at least 11 years. EMRSA-15 was first described in the United Kingdom, where it became the major HA-MRSA clone, together with EMRSA-16, accounting for more than 95% of bacteremia in this country [37]. It has also spread throughout Europe, being reported in Germany, Malta, Italy [38], [39], [40], and representing the major clone in hospitals in Majorca and the Czech Republic [41], [42].

The clonal replacement that occurred within the major hospital analyzed in this study, has occurred in the past in Portugal [35], and also in other parts of the world [38], [39], [43], [44], [45], [46]. Although the reasons lying behind the fade of one clone and the emergence of another are not known, it is believed that the newly introduced clone has a higher fitness than the displaced one.

EMRSA-15 carries few resistant traits, the small SCCmec type IV and has an increased capacity to form biofilm [47]. Moreover, its capacity of dissemination and invasion can be enhanced by the acquisition of additional virulence factors like ACME and PVL [48], [49]. Altogether, these characteristics may have conferred to this particular clone a higher epidemicity, growth rate and capacity for persistence that could explain its success and fitness and may also explain the increase of the MRSA frequency observed in the hospital between 1993 and 2010.

Besides EMRSA-15, we observed that two additional clones (ST105-II and ST125-IVc), belonging to CC5, accounted for a significant part of the MRSA population (22%). Previously, the New York/Japan clone was observed in the nosocomial setting in Portugal and had been suggested as the most probable next emergent MRSA clone in the nosocomial setting in this country [35]. However, this role appears to have been taken by its descendants, the ST105-II and ST125-IVc. These clones differ from ST5 by one point mutation [50] on yqil locus and have already been described as major clones in hospitals in Switzerland [43] and Spain [51] and as minor clones in Brazil [52], the USA [53] and Norway [54]. However, as far as we are concerned, this is the first time that ST105-II and ST125-IVc are reported in a Portuguese hospital. Since these clones have already proven to be capable of disseminating in the hospital and becoming dominant, it is worth watching for the replacement of EMRSA-15 by ST105-II and/or ST125-IVc.

Besides, being one of the countries with the highest prevalence of MRSA in hospitals, Portugal is also the European country with the highest prevalence of MRSA causing skin and soft tissue infections, the most common manifestation of community-associated MRSA disease [13]. In the study presented here, we documented for the first time the characterization of MRSA causing SSTI in persons attending healthcare centers in Portugal and verified that the rate of MRSA causing SSTI (≈26%) in this population was alarmingly high.

However, this study has some limitations since we cannot ignore the fact that a low number of samples were recovered, which could have influenced the MRSA rate obtained. Also, the type of population analyzed may have not been the most adequate to evaluate the prevalence of CA-MRSA since it was composed in its majority of old people (median age 71) with risk factors for hospital contact. The rate of MRSA in SSTI is much higher than that found in nasopharyngeal colonization in children in Portugal [15], but similar to the rates of MRSA in long-term care facilities (Miragaia et al., unpublished), suggesting that probably elderly, rather than children, may constitute a reservoir of MRSA in the community in Portugal.

Almost half of the MRSA isolates identified as causing SSTI in the community belonged to clone ST5-V, associated to ACME II. This clonal type was recently described as a possible emerging CA-MRSA clone in Japan [55], and might be emerging in Portugal as well.

We did not find representatives of the USA300 and ST80-IV clones, which are typical pandemic CA-MRSA clones. This clearly contrasts with the situation observed in Europe and the USA, wherein nowadays these are the most frequent clones in the community [56], [57]. This, however, may not be surprising since the prevalence of isolates belonging to epidemic CA-MRSA clones in Portugal continues to be low (<3%) (Tavares et al. unpublished).

Interestingly, the other half of MRSA found causing SSTI belonged to the major HA-MRSA clones found in the hospital setting in this study (ST22-IVh-t032/t2357, ST105-II-t002 and ST5-IVc-t535) sharing the same PFGE, spa type, SCCmec and ST. Moreover, five isolates shared the same antibiotic susceptibility profile. Altogether, these results suggest a spillover of HA-MRSA isolates into the community.

Of notice, all MRSA recovered from SSTI in this study were resistant to ciprofloxacin, which might reflect the trend of antibiotic consumption in the outpatient setting in Portugal, since quinolones are the third most prescribed antimicrobials [58] and, accordingly, ciprofloxacin (a fluoroquinolone) was prescribed to at least 50% of the patients with MRSA infection in our study (data not shown).

Also of interest was the fact that eight MRSA isolates belonging to ST5-V (CC5) and ST8-VI (CC8) were susceptible to oxacillin, in spite of the fact that they carried mecA. Sharff, K.A. et al [59], recently described two cases of infections caused by oxacillin susceptible MRSA strains belonging to these exact same clonal complexes, one of which was fatal. Since oxacillin-susceptible MRSA strains from this study carry SCCmec V and VI, which lack functional mecA regulators, the failure to detect the expression of beta-lactams resistance in vitro probably results from the fact that they have a heterogeneous profile of resistance, rather than from the lack of induction by beta-lactams. In hospitals MRSA identification relies on antimicrobial susceptibility testing only, which in cases like this, can lead to treatment failure and increased burden of disease.

The vehicles of dissemination of HA-MRSA from the hospital into the community are not firmly established and may differ in different countries. A recent study showed that public buses could function as an important reservoir of this clone in the community [60]. Nevertheless, patients themselves and healthcare workers may be the main means of dispersal. In fact, a great part of MRSA collected from SSTI were isolated from elderly people, whom for their intrinsic condition have a higher probability of having a history of hospitalization and surgery or other risk factors associated with a previous contact with healthcare-associated MRSA, and could have served as a means of spreading of MRSA from the hospital in the community.

The spillover of S. aureus from the hospital into the community observed in this study probably results from the high frequencies of MRSA observed inside the hospital setting, like was observed before for penicillin resistant S. aureus [61]. Nowadays, penicillin resistant S. aureus are endemic in the community, and is worrisome to assume that a similar situation might occur with nosocomial multidrug resistant MRSA in the future in Portugal.

Overall, our findings showed that the increase in the rate of MRSA in a major hospital in Portugal was associated to a clonal replacement and to the emergence of EMRSA-15. Moreover, we showed that in a country with a high rate of nosocomial MRSA, the population structure of MRSA in the community might mirror that found in the hospital setting.

Acknowledgments

Funding Statement

DE was supported by a fellowship from project TROCAR-FP7-HEALTH-2007-B project n° 223031 from the European Commission and project PTDC/BIA-EVF/117507/2010 from Fundação para a Ciência e Tecnologia (FCT), Portugal. NAF was supported by a grant SFRH/BPD/66514/2009 from FCT. This work was funded by project Ref. P-99911 from Fundação Calouste Gulbenkian; CONCORD-FP7-HEALTH-2007-B project n° 222718 and TROCAR-FP7-HEALTH-2007-B project n° 223031 from the European Commission. The work was supported additionally by FCT through grant #Pest-OE/EQB/LAO004/2011. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.

References