On 21 May 2011, Germany reported an ongoing outbreak of Shiga-toxin producing Escherichia coli (STEC), serotype O104:H4. As of 27 July, 3126 cases of diarrhoeal disease caused by STEC E. coli O104:H4 (probable and confirmed), including 17 deaths linked to the outbreak in Germany and occurring in the EU (including Norway) have been reported to the European Centre for Disease and Control (ECDC). In addition, in the EU 773 cases of haemolytic uraemic syndrome (HUS) caused by this bacterium were also reported, including 29 deaths, and linked to the German outbreak. At this time of reporting, a further 119 suspect cases including 4 deaths are also associated with this outbreak. In addition, outside the EU 8 cases of STEC and 5 cases of HUS, including 1 death have been reported in the USA, Canada and Switzerland through the international health regulations (IHR), all with recent travel history to Germany. Based on the most recently reported information (last updated on 26 July 2011), the clinical onset of the last outbreak-related case in Germany was 4 July 2011.
Shortly after the onset of the outbreak in Germany, case-control studies conducted by the Robert Koch Institute (RKI) demonstrated that clinical disease was statistically significantly associated with the consumption of fresh salad vegetables. The high proportion of adult women among cases, was consistent with fresh salad vegetables as the source of infection. Later, a detailed cohort study demonstrated an association with sprouted seeds.
A tracing back and tracing forward study showed that most of the clusters, and all of the clusters for which there was sufficiently detailed food consumption data, could be attributed to consumption of sprouted seeds from one producer in Germany. Investigation of the production site showed no evidence of environmental contamination. Some employees were found to be infected, but reported not to have become ill prior to the outbreak and hence it was concluded that they were not the source of the contamination. This left the seeds used for the sprout production as the prime suspect vehicle of infection; however it was not possible to identify a single seed source since different species were used to produce the sprouted seeds, which were sold as several different mixtures.
On the 24 June, France reported to the Rapid Alert System for Food and Feed (RASFF), a cluster of patients with bloody diarrhoea, after having participated in an event in the Commune of Bègles near Bordeaux on the 8 June. At the time of issue of this report, there were 2 confirmed STEC cases and 9 cases with HUS reported to ECDC, with 4 further cases of non-HUS suspected. Eleven of these patients, 7 women and 4 men, between 31 and 64 years of age, had attended the same event in Bègles. Infection with E. coli O104:H4 was confirmed for 12 of the 15 cases. Epidemiology studies on the French outbreak also implicated sprouted seeds as the outbreak vehicle.
The phenotypic and genotypic characterisation of the E. coli O104:H4 indicated that the isolates from the French and German outbreaks were common to both incidents. Hence, it was concluded that the same strain was involved in the outbreaks both in Germany and in France, strongly indicating a common source.
A tracing back initiated to find the common food source for both outbreaks revealed fenugreek seeds to be common to the 2 outbreaks. The comparison of the back tracing information on the seeds from the French and German outbreaks led to the conclusion that a specific consignment (lot) of fenugreek seeds imported from Egypt was the most likely link between the outbreaks, although it could not be excluded that other lots imported by the same importer/exporter might be implicated.
Data concerning the trace back and trace forward were exchanged through the RASFF, allowing the Member States and European institutions to receive up to date information.
The actual cause or route of contamination of the seed has not been demonstrated. However, based on epidemiological and microbiological investigations within the EU as well as from previous sprouting seed-related outbreaks, it is likely that contamination occurred during seed production. This part of the investigation would need to extend beyond the point of EU import to include the site(s) of production.
The published data for STEC O104 are scarce as this is a very rare serogroup infecting humans in Europe and globally. According to the information reported to ECDC, there were 10 reported cases of STEC O104 infection in the EU Member States and Norway during 2004-2010 from: Austria (1 case in 2010), Belgium (1 case in 2008), Denmark (1 case in 2008), Finland (1 case in 2010), France (1 case in 2004), Norway (1 case in 2006, 3 cases in 2009), and Sweden (1 case in 2010). Moreover, a paediatric case of HUS which occurred in Italy in 2009 has now been associated with STEC O104, so giving a total of 11 cases.
Five of the 10 cases between 2004 and 2010 were related to travel outside the EU; the countries of origin of the infection being Afghanistan (2008), Egypt (2010), Tunisia (2009, 2010) and Turkey (2009). Only 3 of the STEC O104 strains isolated from these cases were of serotype O104:H4 (in Finland in 2010, in Italy in 2009, and in France in 2004). As for the outbreak strain, the STEC O104:H4 strains isolated in Italy and in Finland were both positive for genetic markers of enteroaggregative adhesion, but differed from the 2011 epidemic strain in that they were negative for extended-spectrum beta-lactamase production. The Finnish case was travel-related, with infection acquired in Egypt; whereas the Italian case had a recent history of travel to Tunisia. The origin of the source of infection for the French case was not reported.
In addition to those cases reported to ECDC, a review of the scientific literature revealed that STEC O104:H4 has been isolated twice in Germany in 2001 and once in Korea in 2005. The German isolates differed from the 2011 outbreak strain.
E. coli O104:H4 was not isolated from any batches of the suspect fenugreek seeds. The inability to demonstrate the presence of E. coli O104:H4 in the suspect seeds is not unexpected. It is possible that contaminated seeds were no longer in stock when sampling took place, or even if present were contaminated at a level which made isolation of the organism impossible. However, this does not mean that enterobacteriaceae would not have been present in seeds and sprouted seeds. Previous studies have shown enterobacteriaceae to be present on the surface of the tissue of the plants and that they can also be internalised within the plant (e.g. at primary production, through irrigating with contaminated water or application of organic fertilizer not properly treated and still containing enteric pathogens). In this regard, it is important to underline that a negative laboratory test does not prove the absence of a pathogen, which may be unevenly distributed within the food matrix, perhaps at low levels. This is particularly true when dealing with seeds: the matrix is made up of particles (seeds) which individually may become contaminated and dispersed in large lots. Similarly, favourable physico-chemical conditions to support survival or growth may not be homogeneously distributed.
The preparation of fresh sprouted seeds seldom includes a step where bacterial contamination is eliminated. Hence, food preparation of fresh sprouted seeds is based on the understanding that they are sold as ready-to-eat, i.e. safe to eat as is, or following only minimal preparation. For fresh produce, this assumes and relies on a production process which prevents contamination and an ability to detect contamination when it occurs. These conditions have proven not to be satisfied in this case. The fact that the sampling and bacteriological methods might fail to detect the presence of pathogens such as STEC O104:H4 and Salmonella spp. emphasises the importance of good production and handling practices, as the public health protection associated with such a criterion is questionable given the difficulty of tracing pathogens in seeds.