Epidemiological investigation of a foodborne outbreak in Spain associated with U.S. West Coast genotypes of Vibrio parahaemolyticus

Vibrio parahaemolyticus is a Gram-negative, halophilic bacterium that occurs naturally in estuarine environments worldwide (Kaneko and Colwell 1975). Globally, V. parahaemolyticus is the leading cause of bacterial gastroenteritis associated with the consumption of seafood produce (Powell et al. 2013). Potentially virulent strains are most frequently distinguished from likely avirulent strains by the detection of the thermostable direct (tdh) and tdh-related (trh) hemolysin genes (Bej et al. 1999; Shirai et al. 1990). V. parahaemolyticus is the leading cause of seafood-associated bacterial gastroenteritis in the United States (Anonymous 2005; DePaola et al. 2000), and is one of the most important food-borne pathogens in Asia, causing approximately half of the foodborne outbreaks in Southeast Asian countries (Martinez-Urtaza et al. 2004). The number of V. parahaemolyticus infections has increased globally during recent years (Nair et al. 2007).

Gastroenteritis caused by V. parahaemolyticus has been rarely reported in European countries (Baker-Austin et al. 2010). However, recent studies indicate that the number of V. parahaemolyticus infections also appear to be increasing in Europe (Baker-Austin et al. 2010). Currently, vibriosis is not a notifiable disease in Europe (Baker-Austin et al. 2010). Outbreaks of V. parahaemolyticus have been regularly reported since 1999, with several outbreaks affecting the Northwest region of Spain. In summer of 1999, an outbreak of V. parahaemolyticus affecting 64 persons associated with oyster consumption was detected in the city of Vigo (Martinez-Urtaza et al. 2004). All the strains isolated from stool samples were serotype O4:K11 and belonged to one single and distinctive clone endemic to Europe. A second large outbreak was detected in July 2004 in the city of A Coruña with 80 cases, all of them linked this time to pandcmic O3:K6 strains (Martinez-Urtaza et al. 2005). All clinical cases reported to date in Spain have been exclusively linked to tdh-positive and trh-negative strains (Baker-Austin et al. 2010). Conversely, large V. parahaemolyticus outbreaks have been rarely reported in Europe, with most cases attributed to sporadic cases across Europe.

In the last two decades, several largescale foodborne V. parahaemolyticus outbreaks have been reported in the Pacific Northwest region of the United States (Turner et al. 2013; Paranjpye et al. 2012). These bacteria, subsequently termed the Pacific Northwest complex (Turner et al. 2013), appear to be genetically and biochemically distinct, and carry a lower significantly lower infectious dose compared to other toxigenic V. parahaemolyticus isolates. Until 2012, infections associated with these bacteria were only reported in the Pacific Northwest region, however, in 2012, infections were subsequently reported along the NE coastline of the USA (Newton et al. 2014; Martinez-Urtaza et al. 2013). We describe here epidemiological and microbiological investigations associated with a large V. parahaemolyticus outbreak to identify the source of infections and the most probable origin of the bacteria. This outbreak is particularly noteworthy in that it represents, to our knowledge, the first report of a highly virulent and hitherto Pacific-associated Vibrio complex emerging in Europe.

Nine stool samples were collected and subjected to analysis at a local hospital (Complejo Hospitalario de Pontevedra). All the nine cultures were positive for V. parahaemolyticus and subject to further characterisation at the Centre for Environment, Fisheries and Aquaculture Science (CEFAS, UK). All strains tested positive for toxR and tlh, confirming that all isolates were V. parahaemolyticus. Seven of the nine strains were tdh and trh positive, with a single isolated strain positive for trh only and another strains being negative for both hemolysin genes. All the isolates were susceptible to doxycycline and ciprofloxacin, and resistant to ampicillin and 1st generation cephalosporins.

All recovered isolates were serotyped using commercially available V. parahaemolyticus antisera and subsequently subjected to pulsed field gel electrophoresis (PFGE) and multi locus sequence typing (MLST) analyses. With the aim of identifying a common origin of the outbreak strains, Spanish outbreak V. parahaemolyticus isolates were subsequently analysed by MLST and results were compared with data relative to clinical and environmental strains included in the MLST database (http://pubmlst.org/vparahaemolyticus/). PFGE analysis identified the Galician outbreak strains as a novel group hitherto unidentified in Europe, and different from PFGE from strains previously isolated from clinical and environmental sources in the region (Fig. 1a). Comparison of Galician strains with PFGE data of strains from a recent outbreak in New York two months before (Newton et al. 2014) showed that both outbreak strains shared indistinguishable PFGE profiles (Fig. 1b). MLST analysis revealed the existence of two different genetic variants among the outbreak strains. Allele profiles of tdh and trh positive strains matched with the ST 36, the dominant MLST type in the Pacific NW of North America. Strain tdh negative and trh positive shared the allelic profiles of six genes with ST 417, also endemic in the Pacific NW, although it showed a new allelic variant for pyrC gene according to the MLST database (Fig. 2). This finding was also supported by the serotype analysis which identified the outbreak serotypes were O4:K12 and O4:Kut, dominant groups in the Pacific NW of the USA. The other isolate identified as tdh−/trh+ was untypable by serotyping, whereas the tdh−/trh− isolate was O4:KUntypable. Both isolates showed a new combinations of allelic profiles and were assigned to the two new MLST types ST1031 (tdh−/trh+) and ST1032 (tdh−/trh−), both unrelated to the ST36.

Following the outbreak in 2012, shellfish samples (oysters and mussels) were obtained from several sites along the NW coast of France and Southwest coast of the UK as part of ad hoc testing to identify the potential emergence of ST36 strains into Europe. Samples were collected from commercial shellfish harvesting areas and tested monthly. Of the several hundred V. parahaemolyticus isolates obtained from the samples and subjected to characterisation, only one tdh+/trh+ strain was identified in the southwest of the UK in 2014, however, this strain did not either belong to the same serotype (O4:K12) or MLST type (ST36) previously associated with infections in Galicia.

Several factors make this a highly unusual and noteworthy outbreak. Firstly, this is the first instance, to our knowledge, of strains possessing both tdh+ and trh+ being implicated in an outbreak in Europe. Secondly, the strains identified from preliminary analyses represent several pathogenicity groups, including strains possessing a mixture of hemolysin genes combinations and different STs and pulsotypes. Similar strains possessing these genes have been identified as causing disease in the PNW of the USA (Paranjpye et al. 2012; Newton et al. 2014), and a highly noteworthy characteristic of these tdh+ and trh+ is a significantly higher attack rate than other variants of V. parahaemolyticus (Martinez-Urtaza et al. 2010; McLaughlin et al. 2005). Taken together, these finding suggest that a new and highly virulent strain of V. parahaemolyticus has emerged in Europe. Finally, results from subtyping analyses (including PFGE, MLST, and serotyping) additionally point out that the outbreak isolates closely related, and that these populations were detected in NY 2 months before the outbreak in Galicia (Newton et al. 2014).

Based on a retrospective epidemiological analysis of seafood produce consumed in the dinner (Table 1), shrimp showed the strongest association with cases. From an epidemiological perspective, the origin and nature of the shrimp produce (frozen wild animals from the continental shelf in Argentina) made this product an unlikely direct source of contamination. Additionally, subsequent analyses of frozen shrimp were not able to identify pathogenic V. parahaemolyticus strains. However, epidemiological investigations through questionnaires and food consumption by the different groups unambiguously identified this product as the most probable source of infection. According to the investigation, it is very plausible that pathogenic V. parahaemolyticus strains emerged via cross-contamination through the ice and water used after initial cooking. The ice employed to cool down the shrimp was in contact with local moluscan shellfish that may be the source of pathogenic strains. As the product was hot after boiling and stored at room temperature, the warm conditions may have promoted rapid bacterial growth until reaching levels likely to initiate infections. This hypothesis is further supported by the identification of anomalously warm water in the local region prior to, and during the outbreak (Fig. 3). The incursion of anomalously warm waters from the ocean in the rias has previously been associated with V. parahaemolyticus infections in Galicia (Baker-Austin et al. 2010). A noteworthy characteristic of this outbreak is that several strains were recovered from clinical sources, an unusual feature and uncommon during V. parahaemolyticus foodborne incidents. This is a striking feature of the outbreak that should be the subject of further investigation, as it implies potential cross-contamination of shellfish produce from several possible sources.

The detection in Galicia of three of the major V. parahaemolyticus outbreaks in Europe over the last 15 years stresses the importance of the existence of surveillance programs to enable the detection of V. parahaemolyticus infections. The regional government of Galicia (Xunta de Galicia) has been a strong advocate in the declaration of V. parahaemolyticus infections as notifiable from 1995 onwards, and this study underlines the growing importance of this pathogen as a serious cause of disease. We were fortunate in this instance that some strains were available for molecular and serological analysis, as in Europe there are significant gaps in laboratory testing, monitoring and reporting of such pathogens, which often precludes the capture of strains. The extension of surveillance programs to other areas in Europe with significant shellfish production and consumption may provide an important contribution to achieve a more accurate estimate of risk of Vibrio infections and the burden of this disease in Europe. The potential cross-contamination of shellfish produce using ice should be considered a risk activity in terms of Vibrio contamination. From an environmental perspective, the source of the strains and the mechanistic basis of long distance dispersion into Europe still remain unresolved. The absence of systematic programs for monitoring pathogenic Vibrio in shellfish harvesting areas in European countries represents a serious limitation to detect new local or alien pathogenic variants. In this regard, we have continued ad hoc surveillance along the Atlantic coast during 2013 and through 2014. Environmental monitoring carried out in some areas of the NW Europe to identify the presence of ST36 strains emerging in the area for the two seasons of highest risk following the outbreak (summer 2013 and summer 2014) was unsuccessful in identifying ST36 or O4:K14/O4:Kut isolates. Subsequent ad hoc collaborations with different microbiology reference laboratories in Europe since the 2012 outbreak have not identified U.S. Pacific Northwest strains residing in Europe, or causing infections. Although no further outbreaks of V. parahaemolyticus have been reported in Galicia linked to these strains, it remains unclear if these highly pathogenic variants have persisted in the region due to the absence of regular environmental monitoring of V. parahaemolyticus in the region, as they appear to have along the Eastern seaboard of the Atlantic ocean (Newton et al. 2014; Martinez-Urtaza et al. 2013). Recent data from the USA suggests that these highly pathogenic strains are possibly becoming endemic in an expanding area of the Atlantic Ocean (Newton et al. 2014), with the potential to substantially increase disease risk. Taken together, the link between these and V. parahaemolyticus strains from the PNW implicates trans-oceanic movement of these pathogens in disease emergence. A key area of future work will entail determining the precise mechanistic basis of this long distance transport as well as ascertaining the prevalence of these strains in other foodborne outbreaks.

The outbreak describe here represents a second event of illness related to the introduction of a foreign variant of V. parahaemolyticus in the same region. The first instance was reported in 2004 during an outbreak associated with the pandemic clone of V. parahaemolyticus. This situation stresses the necessity of designing and implementing new programs for enhancing the control and monitoring of seafood products in regions such as Galicia associated with the risk of introduction of foreign variants of pathogenic vibrios in natural environment and the consequent risk of dissemination into Europe. In particular, the data presented here regarding V. parahaemolyticus infections in Galicia underpins the critical relevance of including these pathogens among the notifiable diseases in regions characterised by high levels of shellfish consumption. Finally, the development of a centralised and harmonised monitoring, surveillance and reporting system in Europe would contribute to provide understanding of the real risk of Vibrio infections in Europe, as well as to understand future trends in the risks in a context of warming coastal areas.