SAC Animal Health Group
The aim is to find out how frequent the most important strains of verocytotoxigenic Escherichia coli (VTEC) are in Scottish cattle closest to sale or slaughter. The aim is consistent with recommendations of both the Microbiological Safety of Food Funders Group (MSFFG) and World Health Organisation (WHO) who stated ldquo;Knowledge of the distribution of non-O157 VTEC in animals, including cattle, is limited on a global scale.
The research is to:
- Provide the proportion of Scottish finishing cattle shedding E. coli of types O26, O103, O111, O145
- Provide the proportion of Scottish beef farms with shedding present in finishing animals
- Lead to a better understanding of the association between non-O157 shedding and farm characteristics
- Bacterial strains collected in a coherent manner will be available for subsequent characterisation and typing.
Before testing cattle across Scotland it was important to find out how reliable a test based on immunomagnetic separation (IMS) was for finding the VTEC strains. (IMS is an existing method where tiny magnets are placed into cultures of the sample to attach to and separate the VTEC bacteria). There were differences in how sensitive the IMS test was for the recovery of VTEC, but no sign of differences in isolating different strains and no sign that 3 laboratory staff were obtaining different results. The specificity of the IMS test was high and that meant that very few errors would be made in wrongly identifying strains. The lower sensitivity we think is caused by the lack of a simple method to identify the non-O157 VTEC bacteria when they grow on laboratory plates. The results suggested that the reliable detection limits for non-O157 VTEC are at least 10 times higher than for E. coli O157.
The IMS method is a valuable tool in the study of non-O157 VTEC. In total 6,086 dung samples were collected from 338 farms across Scotland. We sampled all regions, by Animal Health District, and evenly across seasons so as to avoid unnecessary bias and to our knowledge this work was larger than any previously carried out. It establishes a benchmark as the first true prevalence figure for these non-O157 VTEC strains in UK.
The E. coli of types O26 and O157 were most often verocytotoxin positive (49% and 99% respectively) and these VTEC occurred on 10% and 14.7% of farms. These farms were scattered throughout Scotland and perhaps each of 3,500 Scottish farms have at least one animal carrying VTEC of the O26 or O157 types. Though, E. coli of types O103 and O145 occurred on many farms very few of these bacteria were verocytotoxin positive. Therefore, we are 95% confident that verocytotoxin positive strains of these types only occur on 0.3 to 1.6% of all farms (i.e. perhaps fewer than 250 farms in the whole of Scotland). We found no strains of E. coli O111. In examining why these strains occurred we found no sign of differences in numbers of non-O157 VTEC positive farms across the regions of Scotland but there are clear seasonal patterns with fewest positives found in the spring and most in summer and autumn. This is opposite to the results for E. coli O157 where more are found when the animals are inside buildings rather than during the summer. We found no sign of variation by Animal Health Division (AHD) or differences in occurrence of O26 E. coli that were due to the types of cattle management; whether the cattle are housed or at grass; the types of housing; the number of cattle on a farm; or whether the cattle are fed silage, hay or concentrates. Importantly, the occurrence of E. coli O157 strains at farm level is lower than that previously found in Scotland and the results show a real decrease in carriage of E. coli O157 by cattle. This change has occurred in recent years but the cause of the decline is unclear. The change is unlikely to be the result of any direct intervention as there are no known means for control of the organism in cattle.
The number of individual animals that have VTEC of types O26, O103 or O145 in their dung is low, being generally less than 2% of animals. However, there is a lot of between farm variation with some farms having most of the animals positive. This variation may simply be chance occurrence and further detailed analysis is necessary. The analysis will allow us to tell if the results are due to chance, to some farms having higher rates at which the bacteria spread, or if some farms having more introductions of the bacteria.
In Continental Europe, the most common non-O157 VTEC that cause human disease are E. coli types O111, O26, O103 and O145. These bacteria have been reported in 11, 11, 7 and 5 countries respectively. In Scotland this study has shown VTEC strains of types O103, O111 or O145 are uncommon or absent in cattle. In contrast, the type O26 VTEC are considerably more common and at farm level the 10% prevalence is close to the 14.7% found for E. coli O157. This is an important finding as O26 E. coli are the most common non-O157 VTEC found in human disease in Spain and is the first or second most common type in human non-O157 infection in Germany, Belgium, Denmark, Finland, Canada, US and Japan. In the UK, O26 human infections are unusual but four Scottish clinical cases occurred during a six-week period in June and July 2003. An FSA funded study (S11001) that collaborated with this project used a laboratory method to compare 33 human isolates of E. coli typeO26 with 152 of the isolates obtained from cattle. The 33 human isolates commonly carried the same genes as our Scottish cattle isolates and though we found no strains from the two hosts to be exactly the same the human and cattle isolates were related. Thus cattle may be a significant source for the type O26 strains and human infections could develop in Scotland to be as significant a problem as those caused by E. coli O157.
In summary we have provided a benchmark for the occurrence of non-O157 VTEC in cattle faeces in UK and found evidence that E. coli positive for the verocytotoxin gene and of typeO26 are common and widely dispersed on Scottish farms. There has been rapid emergence of new strains of type O26 in Germany so further work should be carried out to monitor the strains in Scotland. Isolates from humans and animals should be examined and studies carried out to find out the sources of human infection. A number of recommendations (p. 33) are suggested by the authors of this report.
Project Code: S01014