Equid Herpesvirus 2 -Associated Oral and Esophageal Ulceration in a Foal

Abstract

A case of a 1-month-old Thoroughbred foal with dysphagia, salivation, pyrexia, oral mucosal pustules, and esophageal ulceration is reported. Swabs from the ulcerated lesions yielded Equid herpesvirus 2 (EHV-2) in virus isolation assays, and histopathology of a biopsy from the esophageal lesion identified nuclear inclusions suggestive of herpesviruses. Immunohistochemical staining with antibodies specific for EHV-2 was positive for epithelial cells in the vicinity of the ulcer but not in more distant mucosa. Electron microscopic evaluation of the biopsy showed herpesviral particles in epithelial cells. The foal recovered over 5 days of supportive and gastroprotective therapy, and the esophageal ulcers healed. Serology and immunohistochemistry indicated that this foal likely had lesions associated with EHV-2 and not EHV-1, −4, or −5.

A 1-month-old Thoroughbred filly was presented to the Ontario Veterinary College, Veterinary Teaching Hospital (University of Guelph, Guelph, Ontario, Canada) with a history of acute onset of depression, profuse salivation, dysphagia, and pyrexia (40°C) of approximately 15-hr duration. The foal had been treated with intravenous (IV) flunixin meglumine a (0.5 mg/kg) and intramuscular ceftiofur sodium b (4 mg/kg) 15 hr and 3 hr before presentation. No other horses on the premises were showing abnormal clinical signs. All horses on the premises were serologically negative for Equine arteritis virus (EVA).

Six weeks before foaling, the mare had been vaccinated against tetanus, Eastern and Western equine encephalomyelitis, equine influenza, Equid herpesvirus 1 (EHV-1), and Equid herpesvirus 4 (EHV-4). During gestation at the 5th, 7th, and 9th month, the mare had been revaccinated against EHV-1 with an inactivated killed vaccine. c Gestation and parturition were unremarkable, and adequate passive transfer of immunoglobulins (Ig) to the foal was reported with the foal's serum IgG concentration greater than 8 g/l at 24-hr postpartum. There were no recent management changes on the premises, and examination of the mare and foal's stall and paddock did not reveal hazardous chemicals or sharp objects.

On presentation, the filly was in good body condition (90 kg) but depressed. The foal demonstrated abnormal tongue movements, drooling of saliva, and an increased sensitivity to palpation of the buccal and gingival mucosa. The rectal temperature and heart and respiratory rates were within normal limits. The oral mucous membranes were dark pink, and the capillary refill time was <2 sec. Neither lymphadenopathy nor other lymph node abnormalities were noted. On auscultation, gastrointestinal sounds were present in all 4 abdominal quadrants. The extremities were palpably cool, and the amplitude of the peripheral pulses appeared reduced. The animal was assessed to be approximately 5% dehydrated. Lung sounds were mildly increased bilaterally, and a slight bilateral, serous nasal discharge was present. Percussion of the thorax revealed a normal lung field. Apart from depression, no abnormalities were detected on a detailed neurological examination.

The oral cavity was thoroughly examined, and video endoscopy of the oropharynx, esophagus, and stomach was performed. The oral mucosa was dark pink with discrete and coalescing pustules visible in the mucosa of the upper gum and lip. Ulcers and erosions were visible on the buccal mucosa, and there was evidence of pharyngeal lymphoid hyperplasia. The esophagus had a large number of punctate ulcers throughout its entire length, and the esophageal mucosa was hyperemic and edematous (Fig. 1). The tongue was not affected. There was no evidence of gastric ulceration or reflux esophagitis. No abnormalities were detected in nasal chambers and nasopharynx. There was no evidence of major trauma, foreign body, or dental problems. The foal had foul-smelling, watery diarrhea.

Therapy was commenced with ranitidine d (1.5 mg/kg IV q 8 hr), sucralfate e (1 g PO q 6 hr), ketoprofen f (1.1 mg/kg IV q 24 hr), and IV boluses of lactated Ringer's solution g (120 ml/kg/24 hr). The filly tried on several occasions to nurse the mare but was unable; consequently the mare was hand milked.

Swabs and biopsies of the oral ulcers, and buccal and upper lip mucosa, were submitted for virus isolation and bacteriological culture and antibiotic sensitivity. h Fecal samples were submitted for standard anaerobic bacteriological culture, specific culture for Salmonella spp., latex agglutination test for rotavirus, and tests for the detection of Clostridium difficile and Clostridium perfringens toxins. h No bacterial pathogens or toxins were detected in feces. Acute (day of presentation) and convalescent (14 days later) blood samples were collected for serology.

Bacteriological culture of the mucosal ulcerations demonstrated colonization with Actinobacillus equuili subspecies equuili and Pasteurella aerogenes. On histological examination of the buccal mucosa, focal areas of necrosis, erosion, and ulceration of the squamous mucosa were present (Fig. 2). In the adjacent mucosa, there were focal areas of intracellular edema and vesicle formation. The edema was most prominent in the basilar areas of the mucosa. Necrosis of single and clusters of mucosal cells was also evident in other areas of the mucosa. Increased numbers of mitotic figures, and acidophilic and larger amphophilic intranuclear (cytomegalic) inclusions in basilar mucosal cells, were suggestive of herpes-like inclusions (Fig. 3). Based on the histological findings, a diagnosis of erosive stomatitis was made.

Serum antibodies to EHV-1, EHV-2 (formerly known as Equine cytomegalovirus), and EHV-4 were detected by serology. h Neutralization assays of the acute and convalescent serum samples revealed the presence of antibodies against EHV-1 and −4 at dilutions of 1:128 and 1:192, and EHV-2 at dilutions of 1:24 and 1:16, respectively. Equid herpesvirus 2 was isolated by cell culture from the oral mucosa and blood and distinguished from other equine herpesviruses, including EHV-5, with specific antisera. i Immunohistochemical staining of sections of oral mucosa for EHV-1 antigen was negative, whereas staining with a monoclonal antibody specific for EHV-2 glycoprotein B protein j ^, 9 was positive in epithelial cells surrounding the ulcer (Fig. 4).

To investigate the presence of herpesviruses in the ulcerated lesion, the coverslips were removed from the hematoxylin and eosin—stained slides, and the sections were postfixed in phosphate-buffered osmium tetroxide, dehydrated in ethanol, and infiltrated and embedded in Embed 812 k via propylene oxide, as previously described. 6 Sections were polymerized in a thin layer of Embed 812 overnight. After immersion in a liquid nitrogen bath, the sections were removed from the glass slide and areas of interest were cut out and mounted on Embed 812 dummy blocks. Ultrathin sections were cut, mounted on grids, and stained with uranyl acetate and lead citrate. The grids were examined in a transmission electron microscope l and images captured using a charge-coupled device (CCD) camera. m Intranuclear inclusions with a halo surrounding an electron dense core were identified in numerous epithelial cells and were consistent with the ultrastructural appearance of Herpesviridae (Fig. 5).

Commencing during the second day of hospitalization, the filly started to nurse normally from the mare. On day 2, IV fluid therapy was reduced to 80 ml/kg/24 hr and discontinued on day 3. By day 3, feces were of normal consistency, vital parameters were within normal limits, and there was no evidence of any respiratory disease on clinical examination.

On day 5, the oral mucosa became a normal pink color, the number of pustules decreased, and the erosions showed evidence of healing. Endoscopic evaluation of the esophagus on day 5 showed marked improvement in the degree and extent of ulceration (Fig. 6). A minor degree of ulceration was still present in the terminal esophagus. The foal made an uneventful recovery.

Equid herpesvirus 2 (family Herpesviridae, subfamily Gammaherpesvirinae, genus Rhadinovirus) is among the most widespread viruses of Equidae, with nearly 100% of horses showing evidence of previous infection. The virus has been isolated from a variety of body sites such as the respiratory and genital tract, lymph nodes, lung, spleen, buffy coat cells, and nasal and ocular swabs of both clinically healthy horses as well as those with diseases. 24 Equid herpesvirus 2 can be isolated from nasal swab filtrates and from the leukocytes of 80–90% of clinically healthy horses. 24 Similar to infection with other gammaherpesviruses, EHV-2 can persist for the lifetime of the individual. The virus probably replicates in cells of the oropharynx, and most seropositive individuals may actively shed the virus in saliva. 13

Most EHV-2 infections are either not associated with clinical signs or with signs of nonspecific illness, such as upper respiratory difficulty, poor performance, malaise, pyrexia, lymphadenopathy, granulomatous dermatitis, and immune suppression. 15,17,20 Despite its uncertain pathogenicity, EHV-2 has been proposed to be associated with keratoconjunctivitis and respiratory disease in foals. 4,7,15,17,20 Equid herpesvirus 2 has also been proposed to play a role in transactivation and possibly reactivation of EHV-1 and EHV-4, 23 as well as modulation of foal immune responses, predisposing them to Rhodococcus equi pneumonia. 16

As a cytopathogenic gammaherpesvirus, 21 EHV-2 is related to Human herpesvirus 4 (commonly known as Epstein-Barr virus). Epstein-Barr virus has been associated with several serious diseases in humans such as Burkitt's lymphoma, heterophile-positive infectious mononucleosis, nasopharyngeal carcinoma, Hodgkin's lymphoma, T-cell lymphomas, 10 non-Hodgkin's lymphoma, and oral hairy leukoplakia in patients with acquired immunodeficiency syndrome (AIDS). 8 In contrast, whether EHV-2 is pathogenic in horses has remained controversial.

In the current case, EHV-2 was associated with, and may have been the cause of, ulcerative and erosive lesions in the oral mucosa in a young Thoroughbred foal that presented with overt clinical signs. Biopsies of the affected oral mucosa had lesions and inclusions compatible with herpesvirus infection. Serum neutralization did not demonstrate a rise in the EHV-2 antibody titer. However, seropositive mares transfer immunoglobulins to colostrum, which may suppress production of antibodies against the same antigen in the offspring. 25 This was considered the most likely cause of the lack of increasing EHV-2 titers in the present case. Similarly, serum EHV-1 and EHV-4 antibodies were most probably of colostral origin. Infection with EHV-5, a herpesvirus with much antigenic similarity to EHV-2, was ruled out by EHV-2-specific serology results, i along with use of an antibody for immunohistochemistry that is nonreactive with EHV-5 glycoprotein B. 1 Bacterial colonization of the affected mucosa within the oral cavity was considered insignificant with respect to the initiation of the disease. Actinobacillus equuli subspecies equuli and P. aerogenes are commensal bacteria in the oropharynx and are considered important secondary pathogens in the lower airways of equine neonates, 18,22 but in the present case they were likely cultured as part of the oral microbial flora.

Horses are usually exposed to EHV-2 in the first 3 months of life. 4 It is possible that, similar to the Epstein-Barr virus, EHV-2 has evolved strategies to elude the immune system and to persist in the horse for life. 3 It may also have the potential to modulate host immune responses. 5 Considerable antigenic, genetic, and biological heterogeneity among EHV-2 isolates have been reported even in the individual horse. It has been suggested that variability among EHV-2 isolates may reflect variations in pathogenicity. 1,2,24 However, the association of genomic heterogeneity with pathogenesis and epidemiology of EHV-2 has yet to be determined. To the authors' knowledge, this is the first report of the actual presence and replication of EHV-2 in a lesion of a horse. Serum IgG and globulin concentrations were normal in this foal; therefore, it was concluded that the foal has adequate humoral immunity. The diarrhea may have been due to ceftiofur sodium administration before presentation, infection of intestinal epithelium by EHV-2, duodenal ulceration, parasitism, or stress. Spontaneous resolution suggested that parasitism or duodenal ulceration were less likely causes. The foal did not develop any significant secondary disease and responded favorably to basic supportive therapy.

Other viruses could also be implicated in stomatitis of horses. Vesicular stomatitis virus causes vesicular lesions on oral and nasal mucosa, external genitalia, or coronary bands, but rarely causes overt disease in horses less than 1 year of age. Typically, almost 50% of horses in an outbreak are affected. 11,12 Stomatitis was also reported as one of several clinical findings in an outbreak of EVA in Spain 14 ; but the seronegative status of the horse herd made this an unlikely cause in the present case. Jamestown Canyon virus, an isolate of California encephalitis virus, was isolated from vesicular lesions from the coronary band, tongue, and buccal mucosa from a horse that presented with anorexia and weight loss. 19 Calici- and adenoviruses could also potentially cause stomatitis, but neither has been identified in horses. 12 There were no indications to test this foal for the latter viruses. The prevalence of EHV-2 in the horse population is high. To determine the clinical importance of EHV-2 infection, extensive epidemiological investigations, in-depth case reports, and correlation of disease with genomic diversity will be required.

Figure 1.

Endoscopic view of punctate esophageal ulcers (arrows) in a 1-month-old Thoroughbred foal, day 1.

Figure 2.

Biopsy of ulcerated buccal mucosa from foal. There is disruption of the epithelial layer, hemorrhage, and necrosis. Bar = 100 μm.

Figure 3.

Higher magnification of Figure 2 showing epithelial cells with eosinophilic intranuclear inclusions (arrows). Hematoxylin and eosin stain. Bar = 15 μ.

Figure 4.

Immunohistochemical detection of Equid herpesvirus 2 antigen in epithelial cells and inclusions adjacent to a mucosal ulcer in foal. Note lack of staining in cells at the periphery (arrows). Bar = 10 μ.

Figure 5.

Electron microscopic image of the biopsy in Figure 2 shows nucleocapsids (arrows) consistent with herpesviruses in epithelial cells. 40,000X.

Figure 6.

Endoscopic view of esophagus from foal in Figure 1, day 5. Whitish material is compatible with adherent topical medication.

Acknowledgements. The authors are grateful for the assistance of Drs. S. Carman and J. DeLay, Animal Health Laboratory, University of Guelph, with virus isolation and immunohistochemical staining, respectively.