Recurrent Actinobacillus peritonitis in an otherwise healthy Thoroughbred horse
Abstract
A Thoroughbred gelding in North America was evaluated for Actinobacillus peritonitis on three different occasions over a 4-year period. At each presentation, peritoneal fluid had an elevated nucleated cell count (220,000–550,000 cells/µL) characterised by non-degenerate neutrophils, no visible bacteria, an elevated total protein (4.6–5.5 g/dL) and bacterial culture yielding Actinobacillus spp. Actinobacillus peritonitis appears to be a regional disease occurring in Australia and less commonly in New Zealand and North America. Recurrence, other than incomplete resolution, has not been previously reported. This case highlights the classical presentation, response to therapy and excellent prognosis despite the alarmingly abnormal peritoneal fluid characteristic of Actinobacillus peritonitis and questions the role of parasite migration in the pathogenesis. Finally, this case is remarkable because Actinobacillus peritonitis was recurrent over several years in an otherwise normal horse.
Abbreviations
-
- CD
-
- cluster of differentiation
-
- LFA
-
- lymphocyte function-associated antigen
-
- MHC
-
- major histocompatibility complex
Actinobacillus is a common equine pathogen that affects foals and adult horses in multiple body systems.1–5 Primary Actinobacillus peritonitis appears to be a regional disease, occurring in Australia6–8 and less commonly in New Zealand and North America,9–12 is generally responsive to antimicrobial therapy and has an excellent long-term prognosis.6–12 The pathogenesis of Actinobacillus peritoneal infection in adult horses is unknown. Historically, strongyle migration has been implicated in the pathogenesis of both Actinobacillus peritonitis and idiopathic peritonitis,6,7,13–15 and, in one Australian study, 8 of 13 horses tested had elevated faecal egg counts.8
Recurrence of Actinobacillus peritonitis, other than incomplete resolution,8 has not been previously reported. This case from North America documents recurrence (twice) following complete resolution of Actinobacillus peritonitis over a period of approximately 4 years in the face of normal immune function and aggressive de-worming practises.
Case report
Peritonitis event no. 1
A 5-year-old Thoroughbred gelding was evaluated for mild colic and fever in April 2005. The horse had been acquired 5 months previously, at which time its teeth were examined and floated routinely, and 0.2 mg/kg ivermectin with 1 mg/kg praziquantel (Zimectrin Gold; Merial) was administered. The horse was fed hay ad libitum, stalled at night and turned out in a dry-lot paddock during the day. The horse had eaten normally the morning of presentation and by the evening was febrile (39.1°C), inappetent, dull and lethargic with mild signs of colic. Physical examination was unremarkable other than fever, mild tachypnoea (24 breaths/min), tacky mucous membranes and reduced borborygmi. Body condition score was 5/9. Rectal examination and transabdominal ultrasound were unremarkable: there was no increase in the amount of peritoneal fluid nor was there increased echogenicity of abdominal fluid. On analysis, the peritoneal fluid was orange, opaque, turbid, non-odorous and had a nucleated cell count of 550,000/µL (reference range: <5000 cells/µL)16 with a total protein concentration of 4.6 g/dL (reference range with refractometer: <2.5 g/dL),16 indicative of severe peritonitis. Microscopic examination of the fluid revealed primarily non-degenerate neutrophils and organisms were not seen with either Gram or Wright's stain. Complete blood count revealed a mild neutrophilia (8100/µL, reference range: 2700–6600/µL) and plasma chemistry was within normal limits, other than mild elevations in indirect bilirubin (4 mg/dL, reference range: 0.5–2.1 mg/dL) and total bilirubin (4.1 mg/dL, reference range: 0.8–2.2 mg/dL) and globulins in the low end of the normal range (3 g/dL, reference range: 2.4–4.4 g/dL). Initial therapy consisted of flunixin meglumine (1.1 mg/kg IV once; Banamine, Intervet/Schering-Plough Animal Health), enrofloxacin (7.5 mg/kg IV every 24 h)17 (Baytril, Bayer Animal Health) and a 60 mL/kg bolus of polyionic fluids IV. Within 2 h of presentation, the horse resumed eating and had normal physical examination parameters.
Enrofloxacin, at the previous dosage and route, and flunixin meglumine (0.5 mg/kg IV every 12 h) were continued on day 2. Transabdominal ultrasound was repeated and there were no abnormal findings other than a mild increase in anechoic peritoneal fluid. The Streptococcus equi titre was moderately positive (1:800), faecal quantitative analysis using centrifugation concentration flotation with sugar and zinc sulfate solutions was negative for eggs and larvae, and faecal culture was negative for Salmonella. The horse continued to have normal physical examination parameters and normal appetite and attitude.
On day 3, repeat peritoneal fluid analysis revealed a substantial improvement: the fluid was slightly cloudy and yellow with a nucleated cell count of 40,000/µL characterised by non-degenerate neutrophils and a total protein <2.5 g/dL. Antimicrobials were continued and flunixin was discontinued. On day 4, repeat haematology revealed resolution of the neutrophilia. The horse was discharged from the hospital and the attending veterinarian continued treatment with enrofloxacin at the same dosage and route.
Aerobic culture of the peritoneal fluid resulted in growth of Actinobacillus spp. (species not determined) and antimicrobial sensitivity testing revealed microbial resistance to penicillin and trimethoprim-sulfamethoxazole, and sensitivity to enrofloxacin. At weeks 3 and 5 after presentation, peritoneal fluid analysis was performed and revealed nucleated cell counts of 40,000 and 20,000/µL, respectively, and total protein level <2.5 g/dL. Week 7 after presentation, the peritoneal fluid was within normal limits with a nucleated cell count of 2600/µL, and total protein <2.5 g/dL. Enrofloxacin, at the same dosage and route, was continued for an additional 2 weeks (until week 9).
Peritoneal fluid analysis was confirmed to be normal 1 month after discontinuation of antimicrobials and again at 2 years after the incident of peritonitis. The horse received veterinary care for a laceration over and communicating with the right femoropatellar joint 2.5 years after the incident of peritonitis, but did not require any other additional care, other than routine preventive care. Feeding and turnout practices remained the same with the exception that, during summer months, the horse had daytime access to grass pasture. Preventive care included routine vaccination, annual oral examinations and routine teeth floating, biannual oral administration of 1 mg/kg praziquantel and 0.2 mg/kg ivermectin combination and administration of 0.2 mg/kg ivermectin (Zimectrin; Merial) every other month when not scheduled for the praziquantel/ivermectin combination.
Peritonitis event no. 2
In December 2008, 3.6 years after the initial occurrence of peritonitis, the horse was evaluated for mild colic and fever (38.9°C), with nearly identical history and clinical findings to the first incident of peritonitis. Peritoneal fluid analysis revealed severe peritonitis: the fluid was orange, opaque, turbid, non-odorous and had a nucleated cell count of 220,000/µL with a total protein concentration of 4.7 g/dL. Cytological examination of the fluid revealed suppurative peritonitis with primarily non-degenerate neutrophils. No organisms were seen with either Gram or Wright's stain. Complete blood count revealed a mild neutrophilia (7300/µL, reference range: 2700–6600/µL) with a mild left shift (100 band neutrophils/µL; normal: 0/µL) and a mild hyperfibrinogenaemia (300 mg/dL, reference range: 0–200 mg/dL). On plasma chemistry, there was a mild elevation in both indirect bilirubin (3.1 mg/dL, reference range: 0.5–2.1 mg/dL) and albumin (3.9 g/dL, reference range: 2.8–3.8 g/dL), with continued low normal globulin level (2.4 g/dL, reference range: 2.4–4.4 g/dL). Eight litres of water were administered via nasogastric tube. Flunixin meglumine (1.1 mg/kg IV once) and enrofloxacin (7.5 mg/kg IV every 24 h) were administered.
Aerobic culture of enrichment broth resulted in growth of Actinobacillus, most closely resembling A. equuli subsp. haemolyticus. Antimicrobial sensitivity testing revealed the same pattern as the previous isolate in 2005. Three weeks after the onset of the second incident of peritonitis, peritoneal fluid was within normal limits, both grossly and cytologically. Enrofloxacin was continued for an additional 10 days and then discontinued. Faecal quantitative analysis revealed an egg count of zero. The horse was given a double-dose of fenbendazole (Panacur PowerPac, Intervet/Schering-Plough Animal Health) for 5 days (10 mg/kg PO every 24 h) and was continued on the routine de-worming schedule thereafter. One month after antimicrobial discontinuation, peritoneal fluid remained within normal limits.
Peritonitis event no. 3
In March 2009, 4 months after the second incident of peritonitis, the horse was again evaluated for mild colic and fever, with nearly identical history and clinical signs to the first two peritonitis events. Peritoneal fluid analysis revealed severe peritonitis: the fluid was orange, opaque, non-odorous, turbid and had a total protein concentration of 5.5 g/dL. Cell count and cytologic examination of the fluid and haematology were not performed. Flunixin meglumine (1.1 mg/kg IV) and enrofloxacin (7.5 mg/kg IV every 24 h) were administered. Aerobic culture of enrichment broth again resulted in growth of A. equuli subsp. haemolyticus and antimicrobial sensitivity testing revealed the same pattern as before.
Eight days after onset of the third event of peritonitis, a complete blood leucocyte count was within normal limits. Although haematological results were normal, the occurrence of recurrent peritoneal infections despite appropriate treatment led to consideration of a possible underlying immunodeficiency. Therefore, basic immunological testing comprising measurement of serum immunoglobulin concentrations and peripheral blood lymphocyte phenotyping via flow cytometry was performed (Table 1).18 Radial immunodiffusion assay revealed a robust humoral response with serum immunoglobulin concentrations of 2300 mg/dL IgG (manufacturer reference range: 984–1685 mg/dL), 50 mg/dL IgM (reference range: 90–150 mg/dL), and 125 mg/dL IgA (reference range: 67–230 mg/dL; VMRD, Inc., Pullman, WA, USA).
| % positive cells | ||
|---|---|---|
| Patient | Reference | |
| Negative control | 0.7 | 0 |
| CD2+ T cells | 90.1 | 87.1 ± 3.3 |
| CD3+ T cells | 89.4 | 90.9 ± 2.6 |
| CD4+ T cells | 68.1 (2855 cells/µL) | 64.8 ± 5.9 |
| CD5+ T cells | 91.4 | 92.0 ± 3.4 |
| CD8+ T cells | 11.7 (445 cells/µL) | 18.2 ± 3.2 |
| CD4+/CD8+ ratio | 5.8 | 3.5 ± 1.0 |
| CD19-like B cells | 19.5 (741 cells/µL) | 10.2 ± 2.5 |
| CD21 B cells | 15.4 | 10.7 ± 5.2 |
| IgM B cells | 21.3 | 11.7 ± 2.1 |
| MHC class I (B and T cells) | 99.7 | 99.9 ± 0.1 |
| MHC class II (B and T cells) | 95.7 | 94.1 ± 5.1 |
| LFA-1 (B and T cells) | 99.6 | 99.8 ± 0.1 |
- CD, cluster differentiation; IgM, immunoglobulin M; MHC, major histocompatibility complex; LFA, lymphocyte function-associated antigen
Immunological testing was performed. In brief, the test used fluorescent-tagged monoclonal antibodies against distinct lymphocyte cell surface markers and flow cytometry to identify and characterise the subpopulations of lymphocytes in peripheral blood. One million isolated leukocytes were labelled with each monoclonal antibody and 10,000 events were counted in the lymphocyte-gated area by flow cytometry.
Both the phenotyping results for CD4+ and CD8+ T cell distributions and the absolute values were normal (Table 1). Slightly increased B cell distribution and activation, based on the distribution of cells expressing CD19 and CD21 molecules, indirectly suggested an active humoral response. These results were not consistent with an underlying immunodeficiency, so additional immunological tests were not performed.
Nine days after onset, peritoneal fluid was within normal limits, with a nucleated cell count of 3160 cells/µL and total protein concentration <2.5 g/dL. Enrofloxacin at the same dosage and route was continued for an additional 2 weeks. One month after antimicrobial discontinuation, peritoneal fluid remained within normal limits. Follow-up at 12 months revealed a normal clinical exam without any recent history of colic, fever or other abnormalities.
Discussion
The cytological abnormalities and gross appearance of the peritoneal fluid in this case led to a diagnosis of peritonitis. In the absence of recent surgery or abdominal wounding, peritonitis can be secondary to gastrointestinal disease, hepatitis, urogenital disease (particularly in postpartum mares), intra-abdominal abscessation, pleuritis, intra-abdominal neoplasia or septicaemia. Peritonitis secondary to gastrointestinal disease, hepatitis, urogenital disease, pleuritis or septicaemia was ruled out based on the mild clinical signs and systemic stability of the horse at each presentation, the nearly immediate response to antimicrobial therapy (1–2 days) and the presence of suppurative inflammation with primarily non-degenerate neutrophils. Peritonitis secondary to intra-abdominal abscessation or neoplasia was initially ruled out because of the lack of haematological abnormalities suggestive of chronic inflammation, the lack of chronic weight loss and the lack of abnormalities on rectal examination and transabdominal ultrasound. Additionally, most adult horses with abdominal abscesses have abnormally high plasma protein and globulin concentrations,19 which were not present in this case. Further evidence against a primary intra-abdominal abscess included the moderate positive S. equi titre (1:800), consistent with previous natural exposure or vaccination but not with intra-abdominal abscessation,20 and the unlikely exposure to Corynebacterium pseudotuberculosis. Once bacteriological culture revealed the presence of Actinobacillus, the definitive diagnosis of primary peritonitis was made. In the previously reported cases of Actinobacillus peritonitis, diagnosis was based on either positive culture results7 or characteristic clinical signs, peritoneal fluid abnormalities and rapid response to treatment, with or without identification of bacteria on peritoneal cytology or positive culture of Actinobacillus spp.8
Given the duration between each peritonitis event (3.6 years and 4 months) in the horse reported here, and the presence of grossly and cytologically normal peritoneal fluid 1 month after antimicrobial discontinuation, and again 2 years after antimicrobial discontinuation for the first event, it is highly improbable that recurrence was related to incomplete resolution of septic peritonitis. Although intra-abdominal abscessation was initially ruled out and intra-abdominal Actinobacillus abscessation has not been reported, the possibility that an intra-abdominal abscess was intermittently shedding Actinobacillus to the peritoneal cavity was considered during the second and third events of peritonitis. Yet this seems unlikely, given the normal peritoneal fluid between events of peritonitis, as abscessation generally causes at least an elevation in the peritoneal total protein because of chronic inflammation either within or adjacent to the abdominal cavity.19,21,22 Additionally, between the events of peritonitis there were no clinical signs indicative of chronic abscessation, such as weight loss, chronic colic, inappetence, lethargy or intermittent fever, nor were there haematological abnormalities consistent with chronic inflammation such as hyperfibrinogenaemia, leucocytosis, neutrophilia and hyperglobulinaemia. Although not used in this case, exploratory laparoscopy or laparotomy, or nuclear scintigraphy with radiolabelled autologous white blood cells or ciprofloxacin,23 may have helped to rule out an intra-abdominal source of inflammation and/or infection, such as an abscess.
Actinobacillus equuli is a Gram-negative pleomorphic rod that normally inhabits the equine oral cavity, pharynx and gastrointestinal tract. As an equine pathogen, it is considered opportunistic, being most commonly associated with foals with failure of passive transfer, resulting in ‘sleepy (or sleeper) foal disease’, septicaemia and joint-ill.2 Because of the commensal role of Actinobacillus and its association with disease in immunocompromised foals, the recurrence of infection in this horse, and blood globulin levels repeatedly being at the lower end of the normal range, basic immunological testing was performed to investigate an underlying humoral or cellular immunodeficiency, mediated by B and T cells, respectively.24 The most common immune problem in horses is humoral immunodeficiency,24 so serum immunoglobulin concentrations were measured. An appropriate humoral immune response to infection, characterised by high serum IgG concentration, was observed.25 The serum IgM concentration was below the reference range, but not low enough to suggest selective IgM deficiency (<25 mg/dL).26 These immunoglobulin results were suggestive of normal B cell differentiation and function in lymphoid tissues. Performing peripheral blood lymphocyte phenotyping allowed for some characterisation of cellular immunity via the T cell subpopulations, as well as additional assessment of humoral immunity via the B cell subpopulation. Results were again appropriate for a horse with an infection and did not support an underlying immunodeficiency. Additional tests for cellular immunity, including in vitro measurement of peripheral blood lymphocyte proliferation and cytokine production upon mitogenic stimulation, could have been performed, but given the overall health of the horse and the lack of other opportunistic infections, coupled with the normal results for the basic immunological testing, further testing was not pursued.
It is commonly stated that the presence of parasites, or strongyle migration, is part of the aetiopathogenesis for Actinobacillus and idiopathic peritonitis.6–8,13–15 In this horse, recurrence occurred despite aggressive larvicidal therapy and repeated negative faecal parasitology. It may be that the aetiopathogenesis of Actinobacillus peritonitis also involves foreign body migration, other than parasites, from the alimentary tract across the mucosal barrier, allowing the introduction of gastrointestinal organisms to the peritoneal cavity.1,27,28
In contrast to most other causes of peritonitis, the prognosis appears to be good to excellent for Actinobacillus peritonitis.7,8 Additionally, it appears that the syndrome is easily treatable and cultures are frequently susceptible to penicillin-based antibiotics and trimethoprim-sulfamethoxazole.8 This susceptibility pattern was not the case in the horse reported here, and antimicrobial sensitivity testing revealed resistance to both penicillin and trimethoprim-sulfamethoxazole during all three events of peritonitis. These findings may represent a regional difference of Actinobacillus8 and suggest that broad-spectrum antibiotics should be used until the results of antimicrobial sensitivity testing are available.
This case demonstrates the classical presentation, rapid response to therapy and excellent prognosis, despite the alarmingly abnormal peritoneal fluid, seen in Actinobacillus peritonitis. Recurrence of the disease is intriguing, especially given that each event seemed unrelated to the previous one, and because the horse is in North America, where the disease is less common. Finally, given the recurrence of disease, repeated negative findings on faecal parasitology and aggressive de-worming practises, this case also highlights that the aetiopathogenesis of Actinobacillus peritonitis in otherwise immunocompetent animals is incompletely understood.
