Tmelab.org

AJVR—09-05-0189R—Thakur—0fig—3tab—VLS—BGM Prevalence of antimicrobial resistance
and association with toxin genes in Clostridium
difficile in commercial swine
Siddhartha Thakur, BVSc & AH, MVSc, PhD; Michelle Putnam, MS; Pamela R. Fry, DVM; Melanie Abley, MS; Wondwossen A. Gebreyes, DVM, PhD Objective—To estimate prevalence and determine association between antimicrobial
resistance and toxin gene profile of Clostridium difficile in commercial pigs at the preharvest food-safety level.
Animals—68 sows and 251 young pigs from 5 farms in North Carolina and 3 in Ohio.
Procedures—Fecal samples were collected from sows (8/farm) and matched young pigs
(32/farm) at farrowing. Cohorts were sampled again at nursery and finishing stages. Clos- tridium difficile isolates were tested for susceptibility to 6 antimicrobials. A PCR assay was used to detect genes coding for enterotoxin A (tcdA), cytotoxin B (tcdB), and binary toxin (cdtB).
Results—C difficile prevalence in young pigs at farrowing was 73% (n = 183) with sig-
nificantly higher prevalence in Ohio (87.5%) than North Carolina (64%). Clostridium difficile was isolated from 32 (47%) sows with no significant difference between the 2 regions. A single pig had a positive test result at the nursery, and no isolate was recovered at the fin- ishing farms. Resistance to ciprofloxacin was predominant in young pigs (91.3% of isolates) and sows (94%). The antimicrobial resistance profile ciprofloxacin-erythromycin-tetracycline was detected in 21.4% and 11.7% of isolates from young pigs and sows, respectively. Most isolates had positive results for tcdA (65%), tcdB (84%), and the binary toxin cdtB (77%) genes. Erythromycin resistance and tetracycline resistance were significantly associated with toxin gene profiles.
Conclusions and Clinical Relevance—The common occurrence of antimicrobial-resistant
C difficile and the significant association of toxigenic strains with antimicrobial resistance could contribute to high morbidity in farms with farrowing pigs. (Am J Vet Res 2010;71: Clostridium difficile is a common nosocomial in-
fection and has been known for many decades to bbreviAtions
cause CDAD in patients. A recent increase in the deaths Clostridium difficile–associated diarrhea among hospital patients has been attributed to the pathogenic strain NAP1/027, which is hypervirulent and has resistance to fluoroquinolones.1–4 Clostridium difficile is also an important pathogen in food animals and is responsible for causing colitis in neonatal pigs, enterocolitis in foals, typhocolitis in adult horses, and enteritis in calves.5–8 Our understanding of the epide- miologic and microbiological features of C difficile in humans has tremendously improved over the past 2 decades. It has recently been suggested that pigs and with identical ribotypes and toxinotypes detected in other food animals might serve as a source of the patho- food animals and humans.2,9–14 It is also possible that gen for humans and cause community-acquired CDAD, humans are responsible for pathogen transmission to pigs; however, this has not been reported. Few studies have been conducted to determine the epidemiologic features and potential importance of toxigenic and an- timicrobial-resistant strains of C difficile in pigs.
From the Department of Population Health and Pathobiology, College In pigs, C difficile is an important cause of neonatal of Veterinary Medicine, North Carolina State University, Raleigh, enteritis, particularly from 1 to 7 days of age in pigs NC 27606 (Thakur, Putnam); and Department of Veterinary Pre- that develop CDAD characterized by colonic edema.7,15 ventive Medicine, College of Veterinary Medicine, The Ohio State Loss of productivity is common, and affected pigs typi- University, Columbus, OH 43210 (Fry, Abley, Gebreyes).
Supported by National Pork Board grant ID NPB No. 07-044.
cally weigh 10% to 15% less at weaning than those in Address correspondence to Dr. Thakur ([email protected]).
unaffected litters.2 An important concern of swine pro- duction medicine is the development of resistance in um taurocholate, and cefoxitin were added and tubes pathogens against important classes of antimicrobials. were incubated under anaerobic conditions at 35°C for This is particularly important in toxigenic C difficile 7 days. Following incubation, the culture broths were strains that can damage the intestines and prolong treat- centrifuged (7,800 X g for 5 minutes), treated with 96% ment, leading to high morbidity and mortality rates. Al- ethanol for 50 minutes to select for spores, and plated though previous studies2,6,7,11,15 have reported C difficile on cycloserine cefoxitin fructose agara with C difficile in swine, no study has been conducted to determine selective supplementb supplemented with 7% laked and compare its prevalence in the same pigs sampled at horse blood.c The presumptive C difficile isolates were 3 stages of production and from 2 distinct geographic biochemically tested via detection of l-proline amino- regions in the United States. In addition, to our knowl- peptidase activity on discs.d The identity of the isolates edge, no reports concerning a statistical association be- was confirmed via PCR amplification of the specific tween antimicrobial resistance and the virulence pro- marker housekeeping gene, tpi, which encodes for tri- file in the C difficile population have been published. Prompted by the lack of such data, the main objective of the study reported here was to determine the preva- Antimicrobial susceptibility testing—The MIC
lence of C difficile in pigs at the preharvest food-safety was determined for a panel of 6 antimicrobials by use level. We also aimed to assess the occurrence of MDR of strips with exponential gradients of antimicrobial phenotypes and determine the statistical association concentrations, which correspond to specific MIC dilu- between toxin genes and antimicrobial resistance of tions.e Susceptibility testing was conducted on Muller C difficile in commercial pigs at different stages of pro- Hinton plates supplemented with 5% sheep bloodf fol- lowing the manufacturer’s instructions.e The antimicro- bials tested, abbreviations, dilution range, and number Materials and Methods
of breakpoints used included ampicillin (Amp; dilu- tion range, 0.016 to 256 µg/mL; 2 breakpoints), cip- Sample collection and bacterial isolation—Pig
rofloxacin (Cip; dilution range, 0.002 to 32 µg/mL; 8 fecal samples were collected from 8 farms, including 5 from North Carolina and 3 from Ohio. Thirty-two breakpoints), erythromycin (Ery; dilution range, 0.016 young pigs (1 to 7 days old) and 8 sows were sampled to 256 µg/mL; 2 breakpoints), metronidazole (Met; at every farm. Fecal samples were collected from 4 dilution range, 0.016 to 256 µg/mL; 16 breakpoints), healthy young pigs/sow if available. The probability of tetracycline (Tet; dilution range, 0.016 to 256 µg/mL; a diseased young pig surviving to the time of slaughter 4 breakpoints), and vancomycin (Van; dilution range, is low; therefore, healthy-appearing young pigs were 0.016 to 256 µg/mL; 4 breakpoints). Superscript let- selected to determine the profile of C difficile in each ter R was used with an abbreviation to indicate resis- pig as it moved from farrowing, nursery, and finishing tance to that drug (eg, VanR). The breakpoint used for farms to slaughter. Every pig was sampled 3 times; the the fluoroquinolone ciprofloxacin was 8 µg/mL.17 The first sample was taken at the farrowing unit from 8 to breakpoint values used in a previous study18 were used 10 days of birth, with subsequent sampling at nursery for the remaining antimicrobials. The MIC and MIC and finishing farms. The pigs were part of an all-in and all-out production flow and therefore were ear tagged Toxin gene detection—The DNA was extracted
for identification and subsequent sample collections from the C difficile colonies by a resin-based DNA ex- during each production phase. Samples were randomly traction kit following manufacturer’s instructions.g Am- collected from more than 8 sows and associated young plification of the housekeeping gene tpi and the toxin pigs at a few farms. This was done primarily to avoid genes including tcdA, tcdB, and cdtB coding for TcdA a decrease in sample size because of pig deaths at the (toxin A), TcdB (toxin B), and CDT (binary) toxins, nursery and finishing farms, and these 4 pigs were in- respectively, was done by use of specific primers as de- cluded in the analysis only when used to replace miss- scribed.16,19 The PCR running conditions included an ing samples caused by pig deaths. Fecal samples from initial denaturing step at 95°C for 5 minutes, followed the young pigs in the farrowing barns were collected by 30 cycles of denaturation at 95°C for 1 minute, an- with sterile loops. All samples from nursery and fin- nealing at 54°C for 1 minute, and an extension-elonga- ishing-age pigs and sows in the farrowing barns were tion step of 72°C for 1 minute. This was followed by a collected with gloved hands directly from the rectum final extension step at 72°C for 7 minutes. Amplified and transported to the laboratory in a sterilized cup at products were run on 2% agarose gel containing ethid- 4°C. Pathogen isolation for the entire study was con- ium bromide in 0.5X Tris-acetate EDTA buffer.h ducted in North Carolina, and samples originating in Ohio were shipped overnight under refrigerated con- Statistical analysis—The frequency of antimicro-
ditions. The study was reviewed and approved by the bial resistance profiles and comparison of MIC values North Carolina State University Institutional Animal between isolates from different production phases and region were evaluated by use of the χ2 test and Fisher Clostridium difficile was isolated by transferring exact 2-tailed test, when applicable, by use of a statis- 2 g of fecal sample in 10 mL of C difficile broth con- tical software package.i Significant association of anti- sisting of proteose peptone (4%), disodium hydrogen microbial resistance and virulence gene profiles for the phosphate (0.5%), potassium dihydrogen phosphate C difficile isolates was determined by use of the OR test (0.1%), magnesium sulfate (0.01%), sodium chloride with 95% CIs by use of commercially available software.j (0.2%), and fructose (0.6%). Cycloserine, 0.1% sodi- Values of P ≤ 0.05 were considered significant.
systems, than in North Carolina (14% in young pigs and 9% in sows). The MIC for the C difficile isolates C difficile prevalence—Two hundred fifty-one
from both sources to erythromycin was > 256 µg/mL young pigs and 68 sows were sampled from 5 farms in (the highest concentration tested). The MIC to tetra- North Carolina (155 young pigs and 44 sows) and 3 cycline varied from 64 µg/mL in isolates from sows to farms in Ohio (96 young pigs and 24 sows). In North 16 µg/mL in isolates from young pigs. The MIC and Carolina, 13 sows only had 3 young pigs/sow available MIC for isolates to ciprofloxacin were > 32 µg/mL for for sampling. In addition, we could not locate 8 pigs representing 2 finishing farms in North Carolina dur- ing sampling. It was assumed that these pigs died dur- Phenotyping based on antimicrobial resistance
ing their stay at the nursery farm or finishing farm or profile—Nine distinct antimicrobial resistance pro-
during transport. The overall C difficile prevalence in files were detected. Ciprofloxacin resistance was rep- young pigs was 73% (n = 183) with significantly (P < resented in 5 profiles. The antimicrobial resistance 0.001) higher prevalence in Ohio (87.5% [n = 84]) than profile CipR-EryR-TetR was the predominant pattern in North Carolina (64% [99]). The overall C difficile and was represented by 19.5% (n = 42) of the iso- prevalence in sows was 47% (n = 32) with no significant lates. Specific resistance phenotypes were found to geographic difference between North Carolina (50%) be associated with the region of sample collection. and Ohio (41.7%). All the C difficile–positive pigs were For example, the CipR-TetR pattern was found in iso- detected at the farrowing stage, except a single nursery- lates from young pigs (n = 6) and sows (34) in North age pig from North Carolina. No deaths were observed Carolina, but none of the isolates from Ohio had this when pigs were sampled at the nursery farms, and none pattern. This clear distinction of resistance pheno- of the pigs tested positive at the finishing farms.
types in the 2 geographic regions was also observed Antimicrobial susceptibility—Clostridium diffi-
for the antimicrobial resistance profile CipR-EryR- cile isolates had resistance to 4 of the 6 antimicrobials TetR, with significantly (P < 0.001) higher frequency tested (Table 1). Overall, the frequency of CipR was sig-
among young pigs (35.7%) and sows (40%) in Ohio. nificantly greater, compared with other antimicrobials, Eleven young pigs from 5 sows had matching antimi- irrespective of the source or the region, with 91.3% fre- quency among isolates from young pigs and 94% from Toxin gene profile—The PCR testing of the 215
sows. All the C difficile isolates in this study were sus- C difficile isolates for the presence of important toxin ceptible to metronidazole (MIC -to-MIC range, 0.13 genes revealed that toxigenic strains commonly oc- to 0.25 µg/mL) and vancomycin (MIC -to-MIC range, curred among the isolates. Genes for toxins A (65% 0.5 to > 0.5 µg/mL), 2 drugs of choice for treatment of of isolates) and B (84%) and the binary toxin coding C difficile in human medicine. Only 6 isolates, includ- genes (77%) were detected. Further analysis revealed 4 ing 5 from young pigs (Ohio, n = 4; North Carolina, 1) toxin gene profile combinations of the 3 toxin encoding and a single isolate from a sow (North Carolina), were genes tcdA, tcdB, and cdtB (Table 2). The predominant
susceptible to all drugs tested in the study. Resistance toxin gene profile coding for A+B+CDT+ was found in to ampicillin was detected in 2.7% (n = 5) of isolates 59% (n = 127) of the isolates. Forty-two (19.5%) iso- from young pigs only. The single isolate from the nurs- lates had the A–B+CDT+ profile, whereas 33 (15.3%) ery farm was susceptible to all the antimicrobials except isolates tested negative for all toxins tested. A signifi- ciprofloxacin. Antimicrobial resistance to tetracycline cantly (P < 0.001) higher number of C difficile isolates was detected in isolates from sows (31.3%) and young from young pigs in Ohio farms (n = 28) had that pro- pigs (46%) and to erythromycin in sows (34.4%) and file. Twenty-two young pigs representing 13 sows had matching toxin profiles. The pansusceptible C difficile A significantly (P < 0.001) higher number of iso- isolate from a sow and a single young pig from North lates from young pigs (66.7%) and sows (90%) in Ohio Carolina had the A–B+CDT+ toxin profile. The remain- had resistance to the macrolide, erythromycin, a class ing 4 pansusceptible isolates from young pigs in Ohio of antimicrobial commonly used in swine production Table 1—Frequency of antimicrobial resistance at different MICs in Clostridium difficile isolated from 183 young pigs and 32 sows.
Antimicrobial
(dilution range [µg/mL])
Breakpoint (µg/mL)
MIC range (µg/mL)
MIC (µg/mL)
Resistance
*Indicates number (%) of C difficile isolates with resistance to an antimicrobial.
Table 2—Toxin gene profiles (number [%] or proportion) in C dif- had a prevalence of 47%, which was higher than re- ficile isolated from the same pigs as in Table 1.
ported for sows and boars in a previous study (3.8%).k It is important to mention that none of the pigs tested Source Region
A+B+CDT+ A–B+CDT+ A–B–CDT– A+B+CDT–
in this study had diarrhea or signs of any illness. The Young North Carolina 59 (59.6) 25 (25.3) young pigs in this study appeared healthy, yet the prev- alence of C difficile in these pigs was high. However, during stress or illness caused by another pathogen, it 42 (19.5) 33 (15.3) 13 (6)
is possible that C difficile might contribute to the dis- A = Toxin encoded by gene TcdA. B = Toxin encoded by gene eased state of the pig. Other than a single pig that had positive results at the nursery stage, none of the pigs Two isolates from young pigs and 1 from a sow in North Carolina had positive results at the nursery and finishing stages had the A–B+CDT– toxin profile; 1 isolate from a young pig in North Carolina had the A–B–CDT+ profile.
of production. A previous studyk has also reported that C difficile is primarily clustered in young pigs, and the prevalence in nursery and finishing age pigs is signifi- Table 3—Distribution (number [%] or proportion) of C difficile iso- cantly lower. An important reason for lower prevalence lates with antimicrobial resistance to tetracycline or erythromy- among older pigs could be reduced susceptibility to the cin by virulence genes, source, and geographic region.
pathogen. Exposure to C difficile elicits an immune re- Antimicrobial
tcdB
cdtB
sponse, which could be more pronounced in adults.22,23 However, more studies are needed to determine the rea- sons for this finding. The difference in prevalence in young pigs between the 2 regions could be the diverse potential sources of C difficile or a random error caused by the limited sample size of this study.
There is little published information regarding the antimicrobial resistance profile of C difficile isolated from pigs. Antimicrobial resistance to ciprofloxacin Tet = Tetracycline. Ery = Erythromycin.
was observed at the highest concentration tested (32 µg/mL) in most of the isolates from young pigs (91.3%) and sows (94%). The results are in accordance with Association between antimicrobial resistance
other studies3,24 that have reported high frequency of and virulence gene profile—The association between
resistance to ciprofloxacin in C difficile isolates from antimicrobial resistance (TetR and EryR) and virulence different sources, including pigs and humans. It should determinants (tcdB and cdtB) in C difficile isolates re- be noted that this class of antimicrobials had not been covered from young pigs was determined (Table 3).
used in swine production in the United States for any This association was tested at 2 levels—the source and geographic location of the sample. The low number of purpose at the time of the study. Recently, the FDA ap- isolates recovered from sows precluded testing this as- proved therapeutic treatment for respiratory infections sociation in sows. A strong association was detected in swine, but all the samples in this study were collect- between TetR C difficile isolated from young pigs with ed prior to that approval. Fluoroquinolone resistance the virulence genes tcdB (OR, 9) and cdtB (OR, 4). A in C difficile is possibly attributable to the extensive use strong association was detected between EryR isolates of this class of antimicrobial in hospitals, which might from young pigs and the tcdB genes (OR, 3). However, lead to nonsynonymous mutations in the gyrase region. no significant association was detected between EryR These acquired mutations in the C difficile gyrA and gyrB and the cdtB toxin gene (OR, 1.7; 95% CI, 0.7 to 4.2) regions are stable and have resulted in the clonal ex- for the 2 sources. When isolates were stratified by geo- pansion of the resistant strains. Pigs may have acquired graphic region, North Carolina isolates had a stronger these antimicrobial-resistant strains from humans, but association between TetR and tcdB (OR, 73; 95% CI, 17 this was not investigated in the present study.
to 298) or cdtB (OR, 16.4; 95% CI, 5.5 to 49) genes. There was no indication of high MICs for metro- However, the CI for the tcdB and the cdtB genes was nidazole or vancomycin, the 2 drugs most commonly wide. The same pattern was observed in the EryR iso- used for treating infections in humans. These results lates from young pigs in North Carolina with signifi- are in accordance with previous reports.3,18,24,25 In the cant association with tcdB (OR, 12.5; 95% CI, 1.9 to 78) present study, resistance was detected to erythromycin genes. However, the EryR isolates from young pigs in in isolates from young pigs (38.3%) and sows (34.4%) Ohio were significantly more associated with the cdtB with a clear distinction based on the geographic ori- gin of isolates. The reason for this difference was not apparent because farms in both regions used erythro- Discussion
mycin in the feed for growth promotion purposes. It is possible that other factors including environment, The high prevalence (73%) of C difficile in young type of flooring, and production flow might be associ- pigs in this study was not surprising because CDAD ated with the dissemination of these C difficile strains in is a known cause of neonatal enteritis.7,15,20 Previous specific geographic locations. The bimodal distribution studies21,k,l have detected C difficile prevalence in young (2 distinct clusterings of MIC values) of erythromycin pigs from 25.9% to 49.5%, and the findings in this study susceptibility, as indicated by the MIC and MIC , are in agreement with those reports. Sows in this study among isolates from young pigs and sows in this study, has been reported in isolates from pigs and humans.24,26 The main objective of determining a significant This distribution has been attributed to the possible association between antimicrobial resistance and viru- random circulation of resistance coding determinants lence profile was to identify and devise strategies to dif- in the population and the use of tylosin as a growth ferentiate, target, and control the pathogenic species of promotor.26 The detection of specific resistance profiles C difficile on farrowing farms, thereby reducing young associated with geographic location of sampling in the pig morbidity. Significant associations were detected present study was interesting. This may indicate that between the TetR and the EryR isolates with virulence specific C difficile strains are circulating in specific loca- markers that were dependent on the geographic origin tions. But it is also important to note that we did not of the isolates. There are conflicting reports in the liter- observe a single antimicrobial resistance or toxin gene ature regarding this association between antimicrobial profile that was restricted to either North Carolina or resistance and pathogen virulence. Although no study Ohio. However, the differences between the 2 states has reported on this possible association in C difficile, a could be attributable to random error because the study recent study36 found a positive association between re- was conducted in a limited number of farms and may sistance and virulence in Escherichia coli isolated from not have external validity for making generalized infer- healthy pigs. In contrast, a study37 conducted with En- ences. The detection of resistance to important antimi- terococcus faecalis isolated from retail foods revealed crobials is concerning and should be studied in more both positive and negative associations between antimi- detail by use of representative samples.
crobial resistance and virulence determinants. It is pos- The toxin profile of the isolates characterized in this sible that antimicrobial resistance has the potential for study was interesting. The TcdA+B+CDT+ profile was the selecting virulence in bacteria, which may result in high predominant gene profile and was found in 59% of the mortality rates in animals. Other investigators have also isolates. Detection of toxins A and B in the feces has reported a similar association in E coli with the concern been used as a standard for diagnosis of C difficile infec- that antimicrobial use may contribute to persistence of tion in pigs and humans.6,27 However, the detection of virulent strains.38 It is important to mention that asso- toxin-negative C difficile strains from the feces of pigs ciation between virulence and antimicrobial resistance in the present study (15.3%) was an important find- could be dependent on the bacterial population, strain, ing and indicates the importance of pathogen isolation source, and other important factors that must be taken and not simply relying on toxin detection. This also has into account before making valid interpretations.
important implications if a particular C difficile strain is toxin negative but resistant to multiple antimicrobials.
Cycloserine-cefoxitin-fructose agar, Oxoid, Baskingstoke, To the authors’ knowledge, the present study is the Clostridium difficile selective supplement, Oxoid, Baskingstoke, first to report the TcdA–B+CDT+ toxin profile in C dif- ficile isolates from pigs. Clostridium difficile strains with Laked horse blood, Hemostat Laboratories, Dixon, Calif.
that toxin profile have been reported to cause outbreaks and clinical cases in humans and have been isolated Epsilometric test, AB Biodisk, Solna, Sweden.
with increasing frequency from infants and elderly hu- 5% sheep blood, BD Diagnostics, Franklin Lakes, NJ.
mans.28–32 The prevalence of binary toxin coding genes Chelex, InstaGene Matrix, Biorad Laboratories, Hercules, Calif.
Tris-Acetate EDTA, Fisher Scientific, Fair Lawn, NJ.
was high, with 77% (n = 166) of the isolates testing positive for the cdtB gene. Previous reports1,33 indicate Egret, version 2.0.1, Cytel Software Corp, Cambridge, Mass.
a higher prevalence of this toxin in young pigs, rang- Harvey RB, Norman KN, Scott HM, et al. Prevalence of Clostridium ing from 78.4% to 83%. Recently, binary toxin has been difficile in an integrated swine operation (abstr), in Proceedings. 9th increasingly present in strains responsible for commu- Bienn Cong Anaerobe Soc Am 2008;5. Available at: www.anaerobe.
nity-acquired CDAD in humans.34,35 The role of binary org/2008/ASA 2008 Session 14.pdf. Accessed Aug 11, 2010.
Zidaric V, Rupnik M, Avbersek J, et al. Prevalence and diversity toxin in the pathogenesis of C difficile infection and its of Clostridium difficile in poultry, pigs and calves (abstr), in Pro- role in conjunction with TcdA and TcdB toxins need ceedings. 9th Bienn Cong Anaerobe Soc Am 2008;12. Available further investigation. In the present study, the isolation at: www.anaerobe.org/2008/ASA 2008 Session 14.pdf . Accessed of C difficile isolates with similar antimicrobial resis- tance and toxin gene profiles from the sows and young pigs may imply direct transmission of C difficile from References
sows to the young pigs. It is possible that the reverse is Cookson B. Hypervirulent strains of Clostridium difficile. Post- true; young pigs might acquire the pathogen from the grad Med J 2007;83:291–295.
barn floor and infect the sows. Because the barn floor Debast SB, van Leengoed LA, Goorhuis A, et al. Clostridium dif- was not tested for the presence of C difficile, its pos- ficile PCR ribotype 078 toxinotype V found in diarrhoeal pigs sible role in pathogen transmission to either the sows identical to isolates from affected humans. Environ Microbiol or the young pigs was not determined. No C difficile Razavi B, Apisarnthanarak A, Mundy LM. Clostridium difficile: organisms were isolated from adult pigs at nursery and emergence of hyperviluence and fluoroquinolone resistance. In- finishing levels even though the sows had positive test results in the farrowing barns. A possible explanation Riley TV. Epidemic Clostridium difficile. Med J Aust 2006;185:133– for this observation is that the sows were continuously exposed to the pathogen in the farrowing environment, Baverud V, Gustafsson A, Franklin A, et al. Clostridium difficile: where they spend a lot of time. This could also explain prevalence in horses and environment, and antimicrobial sus-ceptibility. Equine Vet J 2003;35:465–471.
why the young pigs had positive results immediately Songer JG, Uzal FA. Clostridial enteric infections in pigs. J Vet Diagn Invest 2005;17:528–536.
Songer JG, Anderson MA. Clostridium difficile: an important 24. Brazier JS, Raybould R, Patel B, et al. Distribution and antimi- pathogen of food animals. Anaerobe 2006;12:1–4.
crobial susceptibility patterns of Clostridium difficile PCR ribo- Hammitt MC, Bueschel DM, Keel MK, et al. A possible role for types in English hospitals, 2007–08. Euro Surveill [serial online] Clostridium difficile in the etiology of calf enteritis. Vet Microbiol 2008;13. Available at: www.eurosurveillance.org/viewarticle.
aspx?articleid=19000. Accessed Month Day, Year.
Rodriguez-Palacios A, Staempfli HR, Duffield T, et al. Clos- 25. Schmidt C, Löffler B, Ackermann G. Antimicrobial phenotypes tridium difficile in retail ground meat, Canada. Emerg Infect Dis and molecular basis in clinical strains of Clostridium difficile. Diagn Microbiol Infect Dis 2007;59:1–5.
10. Goorhuis A, Debast SB, van Leengoed LA, et al. Clostridium dif- 26. Post KW, Songer JG. Antimicrobial susceptibility of Clostridi- ficile PCR ribotype078: an emerging strain in humans and in um difficile isolated from neonatal pigs with enteritis. Anaerobe pigs? (lett) J Clin Microbiol 2008;46:1157.
11. Keel K, Brazier JS, Post KW, et al. Prevalence of PCR ribotypes 27. Wilkins TD, Lyerly DM. Clostridium difficile testing: after 20 among Clostridium difficile isolates from pigs, calves, and other years, still challenging. J Clin Microbiol 2003;41:531–534.
species. J Clin Microbiol 2007;45:1963–1964.
28. Moncrief JS, Zheng L, Neville LM, et al. Genetic characteriza- 12. Indra A, Lassnig H, Baliko N, et al. Clostridium difficile: a new tion of toxin A-negative, toxin B-positive Clostridium difficile zoonotic agent? Wien Klin Wochenschr 2009;121:91–95.
isolates by PCR. J Clin Microbiol 2000;38:3072–3075.
13. Jhung MA, Thompson AD, Killgore GE, et al. Toxinotype V 29. Drudy D, Harnedy N, Fanning S, et al. Isolation and character- Clostridium difficile in humans and food animals. Emerg Infect ization of toxin A negative, toxin B-positive Clostridium difficile in Dublin, Ireland. Clin Microbiol Infect 2007;13:298–304.
14. Rupnik M, Widmer A, Zimmermann O, et al. Clostridium dif- 30. Drudy D, Quinn T, O’Mahony R, et al. High-level resistance to ficile toxinotype V, ribotype 078, in animals and humans (lett). moxifloxacin and gatifloxacin associated with a novel mutation J Clin Microbiol 2008;46:2146.
in gyrB in toxin-A-negative, toxin-B-positive Clostri dium dif- 15. Songer JG. Infection of neonatal swine with Clostridium difficile. ficile. J Antimicrob Chemother 2006;58:1264–1267.
Swine Health Prod 2000;8:185–189.
31. Rupnik M, Kato N, Grabnar M, et al. New types of toxin A-nega- 16. Lemee L, Dhalluin A, Testelin S, et al. Multiplex PCR target- tive, toxin B-positive strains among Clostridium difficile isolates ing tpi (triose phosphate isomerase), tcdA (Toxin A), and tcdB from Asia. J Clin Microbiol 2003;41:1118–1125.
(Toxin B) genes for toxigenic culture of Clostridium difficile. 32. Shin BM, Kuak EY, Yoo HM, et al. Multicentre study of the preva- J Clin Microbiol 2004;42:5710–5714.
lence of toxingenic Clostridium difficile in Korea: results of a retro- 17. Clinical and Laboratory Standards Institute. Methods for antimi- spective study 2000–2005. J Med Microbiol 2008;57:697–701.
crobial susceptibility testing of anaerobic bacteria: approved stan- 33. Rupnik M. Is Clostridium difficile-associated infection a po- dard. CLSI document M11–A7. Wayne, Pa: Clinical and Labora- tentially zoonotic and foodborne disease? Clin Microbiol Infect 18. Zheng L, Citron DM, Genheimer CW, et al. Molecular charac- 34. Terhes G, Urbán E, Sóki J, et al. Community-acquired Clos- terization and antimicrobial susceptibilities of extra-intestinal tridium difficile diarrhea cause by binary toxin, toxin A, and Clostridium difficile isolates. Anaerobe 2007;13:114–120.
toxin B gene-positive isolates in Hungary. J Clin Microbiol 19. Stubbs S, Rupnik M, Gibert M, et al. Production of actin-specific ADP-ribosyltransferase (binary toxin) by strains of Clostridium 35. Barbut F, Decre D, Lalande V, et al. Clinical features of Clostrid- difficile. FEMS Microbiol Lett 2000;186:307–312.
ium difficile associated diarrhoea due to binary toxin (actin-spe- 20. Post KW, Jost BH, Songer JG. Evaluation of a test for Clostridium cific ADP-ribosyltransferase)-producing strains. J Med Microbiol difficile toxins A and B for the diagnosis of neonatal swine en- teritis. J Vet Diagn Invest 2002;14:258–259.
36. Rosengren LB, Waldner CL, Reid-Smith RJ, et al. Associations 21. Alvarez-Perez S, Blanco JL, Bouza E, et al. Prevalence of Clos- between antimicrobial exposure and resistance in fecal Campy- tridium difficile in diarrhoeic and non-diarrhoeic piglets. Vet Mi- lobacter species from grow-finish pigs on-farm in Alberta and Saskatchewan, Canada. J Food Prot 2009;72:482–489.
22. Viscidi R, Laughon BE, Yolken R, et al. Serum antibody re- 37. McGowan-Spicer LL, Fedorka-Cray PJ, Frye JG, et al. Antimi- sponse to toxins A and B of Clostridium difficile. J Infect Dis crobial resistance and virulence of Enterococcus faecalis isolated from retail food. J Food Prot 2008;71:760–769.
23. Giesemann T, Guttenberg G, Aktories K. Human alpha-de- 38. Boerlin P, Travis R, Gyles CL, et al. Antimicrobial resistance and fensins inhibit Clostridium difficile toxin B. Gastroenterology virulence genes of Escherichia coli isolates from swine in On- tario. Appl Environ Microbiol 2005;71:6753–6761.
Author: Please verify missing info in ref 24.

Source: http://tmelab.org/pdf/11.pdf