In a study entitled “The economic cost of major health challenges in large US swine production systems” (Holtkamp D et. al. AASV proceedings, 2007), Clostridium perfringens type A was ranked 3rd and Clostridium difficile 12th in terms of the most important health challenges in the breeding herd. Two previous studies in our laboratory had demonstrated that in piglets less than a week of age submitted to the Iowa State University Veterinary Diagnostic Laboratory with the complaint of diarrhea, Clostridium difficile and Clostridium perfringens type A were detected in 29-50% and 47-48% of the cases respectively and represent the two most commonly identified agents. For diseases believed to be of considerable importance, information on the pathogenesis, a reliable experimental model, and a gold standard for diagnostics are all lacking.
The objectives of this study were to:
1. To further characterize the effect of C. difficile infection on the colon of affected piglets
2. To refine the diagnostics for Clostridium perfringens type A enteritis in swine.
3. To evaluate competitive exclusion as a means for preventing Clostridium difficile associated disease (CDAD)
4. To evaluate an experimental human CDAD vaccine, on which a pig product can be based
Clostridium difficile
To critically evaluate the preferred sampling strategy for C. difficile diagnostics, 12 piglets, less than a week of age submitted to the diagnostic laboratory with suspect Clostridium difficile-associated disease (CDAD) were examined. Samples were collected from stomach, jejunum, ileum, cecum, 9 segments of colon, and rectum from each pig and submitted for C. difficile culture, C. difficile toxin ELISA, and microscopic examination. Data was analyzed to determine segments in which the correlation of these factors was the strongest.
Clostridium difficile was isolated from stomach, small intestine and colon. A significant difference was not identified in the rate at which C. difficile was isolated from various segments of the gastrointestinal tract. Neither C. difficile toxin nor lesions were identified in stomach or small intestine. There was a statistically significant association between C. difficile isolation, detection of C. difficile toxins, and lesions in both the cecum and colon (p < 0.05), but not in the small intestine or stomach. Statistically significant differences between large intestinal sites in regards to C. difficile isolation, lesions, or toxin ELISA results were not identified.
Study results support the following conclusions. Cecum and colon remain the preferred sites for C. difficile diagnostics. A consistently superior sampling location in these sections of large bowel was not identified. All methods (toxin, culture, microscopic lesions) were relatively similar and typically detected 66-77% of affected pigs. Sampling of more than one site from each pig improved the diagnostic sensitivity to approximately 97%.
The use of fecal swabs as a tool for antemortem testing for C. difficile was evaluated. If sufficiently sensitive, this method would be preferred because it would eliminate the need to sacrifice piglets. Toxin was detected in 78% of fecal swabs from animals positive for C. difficile toxin in colon contents. This level of sensitivity is not adequate for individual animal diagnostics, but testing of fecal swabs for C. diff toxin may be useful way to diagnose C. diff issues on a herd basis.
Results of this study indicate that cases of C. difficile infection can be missed if only one sample from the large intestine is evaluated. If two or more samples are evaluated, the sensitivity appears to be quite high (97%). A preferred sampling location in the large intestine was not identified.
Clostridium perfringens type A:
Genotyping of Clostridium perfringens type A isolates from young pigs with diarrhea indicate that this disease appears to be due to strains of Clostridium perfringens that are positive for the beta2 toxin gene. Due to frequent lack of microscopic lesions, the occurrence of C. perf type A as normal flora, and the isolation of beta2-toxin positive strains from normal animals, the current diagnostic criteria for this disease are viewed by many as equivocal. A major aim of this study was to critically assess C. perf type A diagnostics and evaluate assays for beta2 toxin as a potential tool for the diagnosis of this disease.
For the purposes of this study, a pig was considered to have C. perfringens type A enteritis if the following criteria were met: 1) gross evidence for diarrhea, 2) exclusion of other infectious causes of diarrhea (E. coli, rotavirus, TGE, C. difficile, Salmonella, C. perfringens type C, PRRSV), and 3) a heavy growth of C. perfrignes type A was isolated from mid small intestine.
The following samples were collected from 10 pigs fulfilling these criteria and from 10 apparently healthy control pigs. Swabs were collected from the following sites for semi-quantitative C. perfringens culture: stomach, duodenum, small intestine at 10 cm intervals, ileum, and rectum. Contents were collected from each segment and fixed tissue was processes for microscopic examination. Ten C. perfringens isolates from each segment of bowel were analyzed by pulse filed gel electrophoresis. Fluid from each segment of bowel was analyzed for beta2 toxin using a dot blot analysis. In those pigs in which a heavy growth of C. perfringens was isolated from mid small intestine, the typical pattern of culture and microscopic lesions were as follows.
The level of bacterial growth was significantly higher in suspect C. perf type A pigs when compared to controls (2.39 ± 0.4 vs 0.64 ± 0.31; p < 0.01). Average percent of isolates producing beta2 toxin was higher in principals than in controls (87.2% vs 53.8%). Principal pigs had higher average lesion scores throughout small intestine (average score 0.52 ± 0.22) than controls (average score 0), but variation in lesion scores from proximal to distal small intestine was not significant. Gram-positive rods were noted in both principal and control pigs, although the average score was much higher in principals (1.12 vs 0.33). Beta2 toxin was detected by dot blot analysis in 23/300 (7.7%) of suspect C. perf type A pig samples and 19/300 (6.3%) of control pig samples.
We have proposed the following pathogenesis for Clostridium perfringens type A enteritis, based on these and previous results. The stomach acts as a fermentation vessel for the proliferation of C. perfringens with continuous seeding of the intestine resulting in high levels of C. perf throughout the entire small bowel. The lack of consistent villus atrophy or necrosis would suggest that released beta2 toxin results in a secretory diarrhea.
The results of this study provide support for parts of this proposed pathogenesis. The concept that C. perfringens type A enteritis is due to quantitative and qualitative changes in the intestinal population of Clostridium perfringens in supported by the higher numbers of C. perf type A harvested throughout the intestinal tract of suspect Clostridium perfringens type A pigs and the higher percentage of organisms positive for the beta2 toxin gene. The isolation of significantly higher numbers of beta2 toxin gene positive isolates from the stomach supports the concept that these quantitative and qualitative differences begin in the stomach.
Diagnostics for C. perf type A remain a central issue. Based on genotyping results, which have associate C. perfringens type A enteritis with beta2-toxin gene positive strains of the organism, we proposed that a test measuring beta2 toxin levels in the intestine could represent a useful diagnostic tool. Unfortunately, a statistically significant increased rate of beta2 toxin detection was not identified in affected pigs compared to control pigs. It does not appear that analysis for beta2 toxin will prove to be a useful test for the diagnosis of C. perf type A enteritis.
For ubiquitous organisms, such as Clostridium perfringens, the mere isolation of the organism is not diagnostic for disease. Supporting information is needed to establish a definitive diagnosis. For many diseases, confirmatory microscopic lesions provide the necessary supportive information to establish a definitive diagnosis. In this study, sections of intestine were critically evaluated to determine whether lesions could be used to support a diagnosis of C. perf type A enteritis. Results were equivocal. Microscopic changes in small intestinal were mild, but were on average distinct from controls. Gram-positive rods were noted in both principal and control pigs, although the average score was higher in principals (1.12 vs 0.33). Unfortunately, two issues limit the usefulness of microscopic changes to confirm a diagnosis: 1) a consistently superior sampling location was not identified, and 2) lesions were not diffuse and differences typically only become apparent following examination of multiple sections of intestine. Because lesions were multifocal, mild, and randomly distributed, microscopic examination of intestinal sections did not appear to be of great value in confirming a diagnosis of C. perfringens type A enteritis.
Previous studies by our laboratory have suggested that the isolation of a moderate to heavy growth of C. perf type A from multiple segments of small intestine is the best indicator of C. perf type A enteritis. Results from this study provide further support for this notion as significantly higher levels of C. perfringens were isolated from pigs with suspect C. perfringens type A pigs enteritis than from apparently healthy controls.
Clostridium perfringens type A enteritis in neonatal swine is a disease purported to be common and economically significant in the breeding herd, yet there are gaps in many aspects of our understanding of this condition. Results of this study provide support for the concept that C. perfringens type A is a disease of neonatal swine which results from qualitative and quantitative differences of the intestinal population of C. perfringens. A semiquantitative estimate of bacterial growth in culture remains the most useful diagnostic test, while histologic lesions and beta2 toxin detection failed to be of diagnostic value.
Competitive exclusion and vaccination as a means to control C. difficile infection.
The effect of competitive exclusion, by a nontoxigenic strain of C. difficile, on preweaning performance was to be evaluated by selecting 200 gestating sows/gilts from a candidate herd with CDAD, randomize animals into two groups of 100, dosing piglets with 1 ml of spores prepared from a nontoxigenic strain, and collecting data on number of diarrhea days per group, number of pigs born live/weaned, weaning weights (individual pigs, litter totals, average), and the rate at which C. difficile and its toxins were detected in rectal swabs.
The efficacy of vaccination with an experimental CDAD vaccine was to be evaluated. An appropriate dose of vaccine was to be determined by using several different doses of a TcdA-based vaccine in pigs. Once an appropriate dose was established, a candidate farm with a high level of endemic CDAD was to be identified. 100 sows/gilts were to be randomized into 2 groups of 50. They were to be vaccinated twice (4 and 8 weeks before farrowing). Data to be collected would include diarrhea days, pigs born live/weaned, weaning weights (individual pigs, litter totals, average) for all litters. Rectal swabs were to be obtained from 5 pigs per litter at 5 days of age will be examined by bacteriologic culture and toxin testing.
These two objectives could not be completed as the lead investigator (Dr. Yaeger) accepted a new position in a different department. His new responsibilities no longer provided the necessary access to materials or the necessary time to carry out the remaining studies. Dr. Ramirez then took over the project re-directing it to develop a Clostridium difficile pig model to be utilized to further investigate the pathogenesis as well a possible intervention steps to mitigate C. difficile disease in piglets. Before this study started there were no published models for creating C. difficile disease in piglets. It was also the goal of this project to be able to create disease in piglets with an actual C. difficile isolate from pigs. After several tries, we were able to take snatched farrowed pigs and inoculate them with C. difficile isolates from actual field cases diagnosed with CDAD submitted to the Iowa State University Veterinary Diagnostic Laboratory (ISU-VDL). We were successful in achieving our goal as well as were able to identify variations in clinical, gross, and histologic lesions between isolates used. With the development of this model, in future studies we are now able to look at possible interventions with a simple and repeatable way to evaluate the process.
