Reasons for performing study: Clostridium difficile has been associated with acute colitis in mature horses. Objectives: To survey C. difficile colonisation of the alimentary tract with age, occurrence of diarrhoea and history of antibiotic therapy; and to study the occurrence and survival of C. diff i c i l e in the environment and antimicro b i a l susceptibility of isolated strains. Methods: A total of 777 horses of different breeds, age and sex were studied. Further, 598 soil samples and 434 indoor surface samples were examined. Antimicrobial susceptibility of 52 strains was investigated by Etest for 10 antibiotics. R e s u l t s : In horses that developed acute colitis during antibiotic t reatment, 18 of 43 (42%) were positive to C. diff i c i l e c u l t u re and 12 of these (28%) were positive in the cytotoxin B test. F u rt h e r m o re, C. diff i c i l e was isolated from a small number o f d i a r rhoeic mature horses (4 of 72 [6%]) with no history of antibiotic treatment, but not from 273 healthy mature horses examined or 65 horses with colic. An interesting new finding was that, in normal healthy foals age <14 days, C. diff i c i l e w a s isolated from 1/3 of foals (16 of 56 [29%]). All older foals (170) except one were negative. Seven of 16 (44%) nondiarrh o e i c foals treated with ery t h romycin or gentamicin in combination with rifampicin were also excretors of C. diff i c i l e. On studfarms, 14 of 132 (11%) outdoorsoil samples were positive for C. difficile in culture, whereas only 2 of 220 (1%) soil samples from farms with mature horses were positive for C. difficile (P = <0.001). By PCR, it was demonstrated that strains from the environment and healthy foals can serve as a potential re s e r v o i r of toxigenic C. diff i c i l e. T h e experimental study conducted here found that C. difficile survived in nature and indoors for at least 4 years in inoculated equine faeces. The susceptibility of 52 strains was investigated for 10 antibiotics and all were susceptible to metronidazole (MIC 4 mg/l) and vancomycin (MIC 2 mg/l). C o n c l u s i o n s : C. difficile is associated with acute colitis in mature horses, following antibiotic treatment. Furt h e r m o re, C. diff i c i l e was isolated from 1 in 3 normal healthy foals age <14 days. Potential relevance: Strains from healthy foals and the e n v i ronment can serve as a potential re s e r v o i r of toxigenic C. diff i c i l e .

Summary Clostridium (C.) difficile, or its cytoxin, was demonstrated in faecal samples from 10 of 25 (40%) mature horses investigated with acute colitis treated primarily with antibiotics for disorders other than diarrhoea. C. difficile was not found in faecal samples from 140 horses without signs of enteric disorders, from 21 nondiarrhoeic horses treated with antibiotics, nor from 22 horses with colitis untreated with antibiotics. Except for C. difficile neither Salmonella nor any other investigated intestinal pathogen was isolated in any of the diarrhoeic horses. The findings strongly support some earlier reports that C. difficile is associated with acute colitis in mature horses treated with antibiotics. Of the 10 horses, 4 proved positive for C. difficile both in culture and in the cytotoxin test, 4 in culture only and 2 only in the cytotoxin test. Eight of 10 horses positive for C. difficile were or had recently been hospitalised, indicating that C. difficile may be a nosocomial infection in horses. All horses positive for C. difficile were treated with β‐lactam antibiotics. The authors are grateful to A. Hellander‐Edman, B. Green and J. Skidell for fruitful cooperation and for providing specimens and to G. Holmström, G. Sigstam and H. Ljung for excellent technical assistance. This research was supported by ATG (the Swedish Horserace Totalizator Board) and the AGRIA Foundation for Research.

In Sweden, mares sometimes develop acute, often fatal, colitis when their foals are treated orally with erythromycin and rifampicin for Rhodococcus (R.) equi infection. Clostridium (C.) dificile, or its cytotoxin, was demonstrated in faecal samples from 5 of 11 (45%) mares with diarrhoea. By contrast C. di&ile was not found in the faecal flora of 12 healthy mares with foals treated for R. equi infection or in 56 healthy mares with healthy untreated foals. No other enteric pathogen was isolated from any diarrhoeic mare. Of 7 investigated treated foals, 4 had a high (1651.0, 1468.3, 273.0 and 88.8 pglg) faecal concentration of erythromycin. The dams of those 4 foals developed acute colitis, whereas the dams of 3 foals with a lower (26.3, 4.6 and 3.7 pglg) faecal erythromycin concentration remained healthy, indicating that there might have been an accidental intake of erythromycin by mares. The foals treated with antibiotics were regarded as asymptomatic carriers and potential reservoirs, as C. dificile was found in 7 of 16 foals investigated, while 56 untreated foals proved negative. The isolated C. difJicile strains proved resistant to both erythromycin (MIC >256 mg/l) and rifampicin (MIC >32 mg/l), a fact that may have favoured the growth of C. difJicile in the foal intestine. All mares found positive for C. dimile were, or had recently been, hospitalised together with their foals, indicating that C. dificile may be a nosocomial infection, in horses. The results emphasise that routine testing for C. di&ile and its cytotoxin is recommended when acute colitis occurs in mares when their foals are treated with erythromycin and rifampicin. Preventive measures in order to avoid accidental ingestion of erythromycin by mares from the treatment of their foals are suggested.

Summary In Sweden there are several reports of mares developing acute colitis while their foals were being treated orally for Rhodococcus equi pneumonia with the combination of erythromycin and rifampicin. In this study 6 adult horses were given low oral dosages of these antibiotics, singly or in combination. Within 3 days post administration of erythromycin, in one case in combination with rifampicin, 2 horses developed severe colitis (one fatal). Clostridium difficile was isolated from one of the horses, whereas no specific pathogens were isolated from the other. Both horses had typical changes in blood parameters seen in acute colitis. Clostridium difficile was also isolated from the faeces of a third horse given an even lower dosage of erythromycin in combination with rifampicin. This horse developed very mild clinical symptoms and recovered spontaneously. In the fourth horse given erythromycin only, very high numbers of Clostridium perfringens were isolated. The horses given rifampicin only did not develop any clinical symptoms and there were no major changes in their faecal flora. In conclusion, it has been demonstrated that low dosages of erythromycin ethylsuccinate can induce severe colitis in horses associated with major changes of the intestinal microflora. Clostridium difficile has been demonstrated as a potential aetiological agent in antibiotic‐induced acute colitis.

Six healthy adult horses were given repeated administrations of trimethoprim/ sulfadiazine (TMP/SDZ) intravenously (i.v.) (2.5 mg/kg TMP and 12.5 mg/kg SDZ) and orally (p.o.) as a paste (5 mg/kg TMP and 25 mg/kg SDZ). Both formulations were given twice daily for 5 days, with a 3-week interval between i.v. and oral administration. The influence of the drug combination on the intestinal microflora was examined and the plasma concentrations, pharmacokinetic parameters and plasma protein binding were determined. There were no major changes in the bacterial intestinal flora and no clinical evidence of gastrointestinal disturbances following the i.v. and oral TMP/SDZ administration. An initial reduction in the number of coliform bacteria during the treatment was notable, though with no evident difference between i.v. and oral treatment. The minimum concentration during a dose interval at steady state (Cminss), the elimination half-life (t1/2beta) and the mean residence time (MRT) were significantly greater after oral administration compared to i.v. for both TMP and SDZ. The plasma protein binding was measured to be 20% for SDZ and 35% for TMP. Oral administration of TMP/SDZ in a dose of 30 mg/kg given twice daily in the form of paste appeared as a satisfactory method for obtaining plasma levels above MIC (minimum inhibitory concentration in vitro) values during the interdosing interval.