Clostridium sordellii-associated enterocolitis in horses

Mauricio Navarro1, Akinyi Nyaoke1, Karina Fresneda1, Santiago Diab1, Janet Moore1, Dena Lyras2, Milena Awad2, Francisco Uzal1
1California Animal Health and Food Safety (CAHFS) Laboratory System, San Bernardino Branch, University of California, Davis, CA, USA. 2Department of Microbiology, Monash University, VIC, Australia

Background: Clostridium sordellii has been associated with fatal toxic shock-like syndrome in humans, gas gangrene in ruminants, pigs and horses, abomasitis in lambs, and omphalitis in foals. However, the role of this microorganism in enteric disease of animals remains poorly documented and controversial. Here we present a series of cases of C. sordellii-associated enteric disease in horses.


Methods: Seven cases were selected from the archives of the CAHFS lab based on: (1) horses received for necropsy with a post-mortem diagnosis of intestinal disease, (2) C. sordellii was isolated from the intestine and, (3) the laboratory tests for other agents known to produce intestinal disease in horses were negative. Gross and microscopic intestinal pathology of these cases was described. Formalin-fixed, paraffin-embedded (FFPE) sections of ileum and/or colon were processed for immunohistochemistry (IHC) of C. sordellii and C. perfringens. Available FFPE sections and C. sordellii isolates were also processed for PCR detection of the sdlO, tcsL and tcsH genes. In addition, small and large intestinal contents from 26 healthy Thoroughbred horses were collected and cultured anaerobically to determine the carrier rate of C. sordellii.


Results: Lesions in the small and large intestine were similar in all horses and characterized by multifocal mucosal hemorrhages, erosions, ulcers, and diffuse pseudomembrane formation. There was also transmural edema and hemorrhage. No toxins of C. perfringens or C. difficile were detected in the intestinal contents of any horse, and no C. difficile or Salmonella spp. were isolated in any case. All cases (7/7) were positive for C. sordellii IHC but negative for C. perfringens. Two available C. sordellii isolates from these cases were positive for the sdlO gene, but negative for the tcsL and tcsH genes. Three out of the five remaining cases resulted positive for the sdlO gene on FFPE samples. C. sordellii was not detected in any of the 26 control horses.


Conclusions: A presumptive C. sordellii etiology was established in the seven cases of this study. Although Koch’s postulates have not been fulfilled for C. sordellii-associated enterocolitis, the results of this study strongly suggest that this microorganism is involved in enteric disease of horses.

Comparison of the Fecal Microbiome of Healthy Horses to Horses with Antibiotic-Associated Enterocolitis

Carolyn Arnold1, Keith Chaffin1, Joerg Steiner1, Jan Suchodolski1
1Texas A&M University

Background and Aims

Antimicrobial-associated diarrhea (AAD) is a reported complication of antibiotic administration in horses.  Antibiotic administration is causes a primary dysbiosis or disruption of the microbial communities of the horse’s hindgut which results in colitis. The objective of this study was to compare the fecal microbiome of horses with AAD to that of healthy horses using 16s rRNA sequencing methods.



Fecal samples were collected from 17 adult horses with diarrhea caused by antibiotic administration. Fecal samples were also collected from 80 normal horses. DNA was extracted from naturally voided feces and 16S rRNA genes sequenced on an Illumina platform. QIIME 2.0 was used to process the sequencing data and to calculate alpha and beta diversity measures. Differences in alpha diversity measures were evaluated with a Mann Whitney U test, whereas analysis of similarities (ANOSIM) was used to detect differences between beta diversity measures. The bacterial abundances between groups were evaluated using a Kruskal-Wallis test and adjusted for multiple comparisons. Linear discriminant analysis effects size (LEfSe) was used to find bacterial taxa associated with the AAD or healthy population.




Antimicrobials associated with the development of AAD included (used alone or in combination): penicillin (procaine penicillin G or potassium penicillin), gentamicin, doxycycline, metronidazole, ceftiofur crystalline free acid, minocycline and chloramphenicol. Of the 17 horses in the AAD group, 11 survived whereas all 80 horses in the control population survived.


All measures of alpha (Chao1, p<0.00; Observed OTUs, p<0.001; Shannon, p<0.001) and beta diversity (unweighted Unifrac distance, R=0.753, p=0.001; weighted Unifrac distance, R=0.701, p=0.001) were found to be statistically different between the AAD and control groups. Both univariate and multivariate analysis found differences between AAD and control horses at the phylum level: Actinobacteria (p=0.019, q=0.047), Bacteriodetes (p=0.0004, q=0.001), Firmicutes (p=0.0239, q=0.0478), Proteobacteria (p=0.0001, q=0.0005), and Verrucomicrobia (p=0.001, q=0.005). LefSe analysis (score >3.5) indicated that Firmicutes, Verrucomicrobia, Actinobacteria and Chlorflexi were more abundantly expressed in the control population whereas Bacteriodetes, Proteobacteria and Fusobacteria were differentially expressed in the AAD group.



Horses with AAD have significantly altered fecal microbiota compared to healthy horses.

Outcome of horses with enterocolitis receiving oncotic fluid support with either plasma or hetastarch

Jamie Kopper1, Clark J Kogan1, Vanessa L Cook2, Harold Schott2
1Washignton State University 2Michigan State University

Background and Aims: Horses with enterocolitis experience severe fluid loss and compromised hemodynamic status. This fluid loss may be further complicated by hypoproteinemia resulting in decreased colloid oncotic pressure making volume resuscitation with crystalloid fluids challenging. As a consequence, colloids are commonly administered to horses with enterocolitis in an attempt to provide hemodynamic stabilization. Two types of colloids exist: natural blood derived products such as plasma and synthetic products such as hetastarch (HES). Synthetic colloids are more cost effective, do not carry risk of disease transmission and are more readily accessible because they can be stored at ambient temperature. However, safety of HES products has recently been questioned in systemically compromised human and canine patients. Currently, there are no studies assessing use of natural versus synthetic colloids in equine patients. The objective of this study was to determine whether there was an association between type of colloid administered and survival of horses with enterocolitis. We hypothesized that horses that received plasma would be more likely to survive to discharge compared to horses that received HES.

Methods: A retrospective review of medical records of horses with enterocolitis treated with plasma or HES between January 1, 2005 and December 31, 2011 was performed. Data collected included signalment, outcome, physical and clinicopathologic findings and volume and type of colloid administered.

Results: 69 horses (75%) were treated with plasma and 23 horses (25%) were treated with HES. After accounting for confounding variables, horses treated with plasma (80% survival) were more likely to survive to discharge compared to horses treated with HES (47% survival, p=0.041) despite similar disease severity at admission. The estimated odds ratio of non-survival in horses that received HES compared to horses that received plasma was 3.13 (95% confidence interval: 1.04-10.22).

Conclusions: This study provides support that use of natural colloids may be superior to treatment with synthetic colloids in horses with enterocolitis. A prospective, multi-center trial comparing outcome of critically ill equine patients treated with natural or synthetic colloids is warranted to further evaluate the safety of HES

Rate of detection of multiple pathogens in adult horses with gastrointestinal disease and association with survival in a hospitalized population

Jamie Kopper1, Jaclyn Willette2, Clark Kogan1, Christian Leutenegger3, Steven Bolin2, Harold Schott2
1Washington State University 2Michigan State University 3IDEXX laboratories

Background and Aims: The rate(s) of isolating ≥1 potential pathogen (PP) from feces of adult horses with gastrointestinal (GI) disease and impact on outcome has not been reported. Our primary aim was to determine rates of detection for single and multiple PP by multiplex PCR of fecal samples from adult horses with GI disease. A secondary aim was to determine whether detection of one or more PPs isolated from feces of adult horses hospitalized for treatment of acute enterocolitis resulted in poorer outcome (survival), compared to horses from which no PP was isolated.

Methods: Retrospective evaluation of results of multiplex PCR fecal testing for enteric disease was performed for 3,753 fecal samples submitted to IDEXX Laboratories. Fecal testing results for PPs and outcome (survival to discharge) was assessed in a cohort of 239 horses admitted to Michigan State University’s (MSU) Veterinary Medical Center for treatment of acute enterocolitis. Fecal samples from these horses were tested by multiple methods at MSU’s Veterinary Diagnostic Laboratory.

Results: One or more PPs was detected by multiplex PCR testing of 1,175/3,753 (31.3%) fecal samples with 23.5% returning one positive result, 6.2% returning two positive results, 1.3% returning three positive results and 0.3% returning four positive results. Coronavirus (5.9%) and Salmonella sp. (5.9%) were most frequently detected. Clostridium difficile toxins A and B were most frequently co-detected, however they likely represent production of multiple toxins by the same organism. No significant correlations were found between rates of DNA detection of other PPs. In horses hospitalized for treatment of enterocolitis, a single PP and >1 PP were detected in 49% and 12% of horses, respectively. Horses with no PPs detected in fecal samples were more likely to survive to discharge (88%, P=0.04) compared to horses with one PP (76%), but no difference in survival was found between horses no PPs and horses with ˃1 PP (89%).

Conclusions: Detection rates for potential enteric pathogens in horses with GI disease is low. Further, finding PPs in feces of horses with acute enterocolitis provided little prognostic information. Nevertheless, testing for potential enteric pathogens remains an important biosecurity tool.