• Open Access

Association of Asinine Herpesvirus-5 with Pyogranulomatous Pneumonia in a Mare


Corresponding author: Dr Nicholas Frank, Department of Clinical Sciences, Tufts Cummings School of Veterinary Medicine, 200 Westboro Road, North Grafton, MA 01536; e-mail:nicholas.frank@tufts.edu


equine herpesvirus


asinine herpesvirus


equine multinodular pulmonary fibrosis


pyogranulomatous pneumonia


pulmonary fibrosis

An 11-year-old, 532-kg, 3 months pregnant American Quarter horse mare was referred to the University of Tennessee Veterinary Medical Center with a 1-month history of pneumonia, tachypnea, and intermittent fever (maximum, 104.0°F). Medical treatments administered before referral included gentamicin sulfate (6.6 mg/kg IM q24h) and ceftiofur sodium (4.4 mg/kg IM q24h) for 2 weeks, followed by chloramphenicol palmitate (50 mg/kg PO q8h) for 2 weeks. Flunixin meglumine (500 mg IM) also was administered when fever was detected. Fever resolved in response to treatment with chloramphenicol, but tachypnea persisted. A transtracheal wash was performed 7 days before presentation and 36 hours after withdrawal of antibiotics. No cytological abnormalities were detected and there was no growth on aerobic culture. All antimicrobial treatments had been discontinued 48 hours before presentation.

On presentation, the horse was bright and responsive with normal rectal temperature (99.6°F) and heart rate (36 beats/min). Body condition was 5/9. Tachypnea (44 breaths/min) was detected and thoracic auscultation identified bronchovesicular sounds bilaterally that were worse after application of a rebreathing bag, especially in the left hemithorax. No wheezes or crackles were detected. Moderate tachypnea (>60 breaths/min) persisted for more than 10 minutes after removal of the rebreathing bag. The mare was accompanied by her 5-month-old foal.

CBC showed neutrophilia (11.6 × 103/uL; reference range: 2.6–5.5 × 103/uL), monocytosis (0.46 × 103/uL; reference range: 0–0.35 × 103/uL), and hyperfibrinogenemia (500 mg/dL; reference range: 100–400 mg/dL). Plasma biochemistry results included hyperproteinemia (8.2 g/dL; reference range: 5.7–7.5 g/dL), hypoalbuminemia (2.6 g/dL; reference range: 2.7–3.7 g/dL) and hyperglobulinemia (5.6 g/dL; reference range: 2.5–4.6 g/dL).

Thoracic ultrasound examination identified diffuse superficial hyperechoic areas with comet tail artifacts consistent with pleural damage. Generalized multifocal small hyperechoic areas were detected in both lungs, with the left lung more severely affected. No pleural effusion was evident. Lateral radiographs of the thorax (Fig. 1) disclosed a severe diffuse reticular nodular interstitial pattern, which was more severe ventrally and bilaterally, and progressed from an interstitial to an alveolar pattern in the left ventral lung fields.

Figure 1.

Left lateral craniodorsal projection of the thorax. There is a severe diffuse reticular nodular interstitial pattern, with greater severity in the ventral aspect. The left ventral lung is more affected, progressing from an interstitial to an alveolar pattern.

Broncho-alveolar lavage was performed using a 2.4-meter endoscope with the sample collected from the cranioventral area of the left lung. Mild accumulation of mucus was observed in the trachea. Cytological examination of cytocentrifugated fluid indicated granulomatous to pyogranulomatous inflammation. Cells were predominantly macrophages, including occasional multinucleated giant cells, with low to moderate numbers of neutrophils, few small lymphocytes, and rare eosinophils. No microorganisms were observed. No growth was reported on fungal, and aerobic and anaerobic bacterial culture plates after 7 days. Feces were submitted for routine sugar and Baermann flotation tests and no parasite ova or larvae were detected.

A presumptive diagnosis of pyogranulomatous pneumonia (PP) was made and chloramphenicol palmitate was administered (50 mg/kg PO q8h) along with fluconazole (10 mg/kg PO q8h), clenbuterol (1 μg/kg PO q12h), and omeprazole (1 mg/kg PO q8h).

Neutrophil and monocyte counts were within normal reference ranges when re-evaluated on day 3, but hyperproteinemia (8.4 g/dL), hyperglobulinemia (5.9 g/dL), and hypoalbuminemia (2.5 g/dL) remained evident and hyperfibrinogenemia (600 mg/dL) was detected. There was no evidence of improvement when thoracic radiographs were repeated on day 6 and tachypnea persisted (40–44 breaths/min). However, the mare was bright, alert, and responsive and had a good appetite.

Because clinical signs of respiratory disease had not resolved, a percutaneous lung biopsy was performed to provide tissue for histopathologic examination and herpesvirus PCR analysis. Lung tissue was obtained with a 14-gauge 15.2-cm Tru-Cut biopsy instrument.1 Ultrasound imaging was used to guide placement of the biopsy instrument through the right eighth intercostal space. Three samples were collected from the lung periphery at the end of inspiration at depths of 4, 5, and 7 cm. No complications were observed. The mare was discharged the next day with a diagnosis of PP of undetermined etiology. Fluconazole and chloramphenicol were continued at home for 5 days.

Histopathologic examination of lung tissue (Fig 2) disclosed central areas of necrosis with amorphous eosinophilic material admixed with degenerate neutrophils, surrounded by macrophages, lymphocytes, and plasma cells that also were seen in the adjacent parenchyma. No organisms (fungi or bacteria) were identified with routine, Gomori methanamine silver, or Gram stains and all culture results were negative. Fibrosis was not evident and no viral inclusion bodies were detected.

Figure 2.

Hematoxylin and eosin stained section of lung tissue obtained by means of biopsy. The section consists of inflammatory infiltrate with a small amount of adjacent alveolar parenchyma (far left of field). There is a central area of necrosis with amorphous eosinophilic material admixed with degenerate neutrophils surrounded by a rim of macrophages and an outer infiltrate of lymphocytes and plasma cells. These inflammatory cells extend into the adjacent parenchyma. No infectious organisms were present with routine, Gomori methanamine silver, or Gram stains. Bar = 100 μm.

For herpesvirus PCR analysis, lung tissue was macerated in approximately 200 μL sterile PBS, and DNA was extracted from the supernatant using a commercially available kit.2 A nested consensus primer PCR was used to amplify a portion of the viral DNA polymerase3 gene.[1] The product was visualized by ultraviolet transillumination4 of a 1.4% agarose gel containing ethidium bromide. PCR product cleanup 5 was performed, and the sample was submitted for sequencing6 using the nested forward and reverse primers. The sequence was analyzed7 and examined for homology with published sequences in GenBank. The sequence was 100% homologous with Asinine Herpesvirus 5 (AsHV-5) type 1 (GenBank accession number AY054993). 8

A diagnosis of AsHV-5 was pursued by collecting additional information from the owners. The mare and her foal lived in a barn with two 15-year-old miniature donkeys (1 male, 1 female) and a 7-year-old Quarter horse gelding. All of the animals grazed on the same pasture during the day. None of the adult animals showed signs of respiratory disease, but the foal suffered from pneumonia at 3 months of age that responded to empirical treatment with rifampin/clarithromycin and chloramphenicol on the farm. To further investigate, nasal swabs were collected from all of these animals and submitted for PCR analysis.

DNA was extracted from the nasal swabs using the buccal swab spin protocol2 from the same kit, and the same protocol also was used to amplify and evaluate herpesvirus DNA from these samples. Nasal secretions from both donkeys also contained herpesvirus DNA that was 100% identical to AsHV5-1. The nasal sample from the mare contained EHV-5 DNA, and EHV-2 DNA was detected in the sample from the foal. The identity of the product from the nasal swab sample from the gelding could not be determined by direct sequencing because of sequence overlap and the presence of 2 different DNA products of almost the same size. Sequences therefore were ascertained by cloning into plasmid vectors, transducing Escherichia coli,7 and sequencing the isolated recombinant plasmids9 from 6 colonies. The sample from the gelding contained DNA from both EHV-5 and AsHV-5. All of the animals that tested positive were carrying AsHV-5 with the same single-base substitution (AsHV5-1).

It also was necessary to rule out the presence of EHV-5 in the lung tissue collected from the mare, which might not have been detected by consensus PCR. EHV-5 type-specific primers have been published previously.[2-5] However, these primers amplify portions of the gamma herpesvirus that have not been sequenced in AsHV-5. EHV-5-specific primers therefore were designed to target the region of the DNA polymerase gene that differs between the 2 viruses. Previously isolated plasmid DNA was used as control to determine the specificity and sensitivity of the reaction. Serial 10-fold dilutions were prepared and tested by nested PCR. Samples were electrophoresed, visualized, and the identity of the PCR product was confirmed by sequencing as previously described. The sensitivity of the PCR assay was approximately 100 genomic copies of EHV-5. The EHV-5-specific PCR was unable to detect the AsHV-5 DNA, even when approximately 106 copies were tested. Each of the samples tested by consensus PCR also was analyzed by EHV-5-specific PCR to confirm the ability of the test to detect viral DNA in clinical samples. Testing confirmed results from the consensus PCR; EHV-5 was not present in lung tissue collected from the mare. Virus isolation from BAL fluid also was attempted through 10 passages in rabbit kidney (RK13) cells, but negative results were obtained. Virus isolation from lung tissue was not attempted because of small sample size.

Three weeks after discharge, the referring veterinarian reported that the mare's physical examination results were unremarkable, including a normal respiratory rate at rest and no signs of dyspnea when exercising on the paddock. The mare subsequently gave birth to a healthy foal and then was sold at an auction and lost to follow-up. Despite multiple requests, follow-up radiographs were not performed because the owners had decided to sell the mare. Because follow-up information was limited, it is conceivable that radiographic evidence of pneumonia persisted and clinical signs of pneumonia would have been detectable during strenuous exercise or pulmonary function testing.

Pyogranulomatous pneumonia was diagnosed in this case of persistent respiratory disease, and AsHV-5 was detected in lung tissue collected from the affected animal. This virus also was detected in nasal secretions from 3 animals in direct contact with the mare, including 2 donkeys and 1 horse. It can only be stated that AsHV-5 was associated with PP in this case, because it is conceivable that other etiological agents initiated the disease. Other causes of pneumonia include bacterial, fungal, and parasitic infections, neoplasia, pneumotoxins, silicosis, and the recently described equine multinodular pulmonary fibrosis (EMPF).

Bacterial or fungal pneumonia were differential diagnoses in this case, but no organisms were detected on microscopic examination of BAL fluid and lung tissue, and culture results for all samples were negative. Other diagnostic tests for fungal infection, including serum antibody titers for Blastomyces, Coccidioides, Histoplasma, Aspergillus, and Cryptococcus spp. were not performed. Parasitic pneumonia caused by Dictyocaulus arnfieldi was ruled out by BAL cytology findings and Baermann fecal flotation results. There was no reported exposure to silica or other lung irritants, or to toxins.

Pulmonary fibrosis (PF), accompanied by a variable inflammatory infiltrate, is reported in cats, horses, donkeys, and humans.[6-9] Gamma herpesviruses have been associated with PF, but pathophysiological mechanisms have not been elucidated. Viruses also play a role in the development of idiopathic PF in humans and it is thought that occult viral infections contribute to the development of this condition.[6] EHV-5 recently has been associated with PF in horses,[4, 10, 11] and thus we attempted to detect herpesvirus in the biopsy specimen by PCR analysis. This type of analysis was selected because gamma herpesviruses are difficult to isolate and require multiple passages in cell culture.[5] A subsequent attempt to isolate virus from BAL fluid was unsuccessful, and virus isolation was not performed on the lung tissue because of small sample size. Because virus isolation was attempted 7 days after BAL fluid was collected and refrigerated, it is possible that viruses were inactivated.

Equine herpesviruses 1, 3, and 4 are included in the Alpha herpesvirinae subfamily, whereas EHV-2 and EHV-5 belong to the Gamma herpesvirinae subfamily. Herpesviruses can survive and persist in animals through lifelong latency, with virus replication and shedding periodically reactivated.[12] Gamma herpesviruses are widespread in the equine population, yet their clinical relevance is unclear.[2, 13] EHV-2 has been isolated from healthy horses and foals with keratoconjunctivitis, pneumonia, and pharyngitis, and from adult horses with lymphadenopathy, fever, anorexia, general malaise, and poor performance.[2, 14, 15] The epidemiology of equine herpesviruses was examined in a study of 12 mares and their foals, and EHV-2, -4, and -5 were detected in nasal swabs.[16] In the same study, only EHV-4 was consistently associated with respiratory disease whereas EHV-2 was detected in all mares and foals at 2 months of age, and only in 50% of foals at 1 month of age.

Six herpesviruses infecting donkeys have been designated asinine herpesviruses 1 to 6.[8, 17] AsHV-1 (also called EHV-6) was first isolated from vesicular and erosive lesions on the muzzle of a donkey foal, and the external genitalia and udder of its dam. In contrast, AsHV-2 (also called EHV-7) was isolated from the blood of healthy donkeys and is more closely related to EHV-2 and EHV-5; it has been detected in 8% of mules and donkeys.[18] AsHV-3 (also called EHV-8) has been isolated from healthy donkeys after treatment with corticosteroids. Asinine herpesviruses 4, 5, and 6 are gamma herpesviruses and have recently been described in donkeys with interstitial pneumonia characterized by marked syncytial cell formation.[8] These donkeys were from different herds and presented with acute, and often fatal respiratory disease. In a study, AsHV-5 was detected by PCR analysis in respiratory lavage fluid from healthy horses, although testing was not specific enough to detect the AsHV5-1 subtype previously associated with pneumonia.[1, 4]

A diagnosis of EMPF initially was considered in this case because of the radiographic findings and limited response to antibiotics and antifungals. EMPF is a nodular fibrotic lung disease of adult horses with an average age of 14.5 years (range: 4–28 years old).[4] In a recent report, 4 of 5 horses with EMPF tested positive for EHV-5[11] and the affected horses presented with fever, weight loss, cough, mild tachypnea or respiratory distress, clinicopathologic abnormalities consistent with chronic inflammation, and radiographic evidence of bronchointerstitial to nodular pulmonary disease.[4, 10] Pancytopenia (anemia, thrombocytopenia, neutropenia) also was reported in a case of EMPF, in which EHV-5 was detected in lung and bone marrow.[19]

The diagnosis of EMPF is based on histopathologic findings of well-demarcated nodular pulmonary interstitial fibrosis, mixed inflammatory cell infiltration, type II pneumocyte hyperplasia, intranuclear viral inclusion bodies, and detection of EHV-5.[10] In contrast, the key postmortem findings for AsHV-5 in donkeys include marked inflammation characterized by syncytial cell formation, and the absence of both fibrosis and intranuclear viral inclusions.[4, 8, 10, 19] In the case reported here, inflammation was the prominent finding, nodular regions were poorly demarcated, and fibrosis, intranuclear viral inclusion bodies, and syncytial cell formation were absent. It is difficult to explain why the histopathologic findings associated with AsHV-5 infection in this mare differed from those previously described in affected donkeys and why this mare developed clinical disease whereas the donkeys and gelding on the same farm remained healthy. One explanation for these findings is that another respiratory pathogen was involved in the development of pneumonia, and histopathologic lesions were caused by multiple pathogens. Rhodococcus equi was presumptively diagnosed when the foal developed pneumonia 2 months earlier, but it is conceivable that another species of bacteria affected the foal and mare at this time. Stress associated with pregnancy also might have contributed to the development of disease in this mare, compared with other exposed animals. Lesions also might have differed from previous reports of AsHV-5 infection in donkeys because of inherent differences between the species.

None of the treatments administered to this mare targeted herpesviruses or PF. Humans with idiopathic PF are treated with corticosteroids, antiviral and antifibrotic drugs, antioxidants, interferon gamma, and specific cytokine antagonists.[20] Oral acyclovir has been used to treat horses with EMPF in combination with corticosteroids, but in a report, only 2 of 5 horses improved clinically.[10] In addition, the bioavailability of acyclovir administered PO is poor in horses and it has not been established whether equine gamma herpesviruses are susceptible to this drug.[21, 22] Valacyclovir has better PO bioavailability and may be more appropriate for this reason.[23] Hart et al[19] also reported the use of pentoxifylline to inhibit PF in a horse with EPMF. Corticosteroids and antiviral drugs were not administered to this mare because she appeared to be recovering and, in the case of the corticosteroids, because of the owner's concerns about her pregnancy. Administration of a nonsteroidal antiinflammatory drug might have contributed to her recovery because inflammation was evident on histopathologic examination of lung tissue.

In summary, AsHV5-1 was isolated from the lung tissue of an adult horse with persistent tachypnea and nodular interstitial pneumonia. Equine herpesvirus 5 was not detected in this mare's lung tissue, but herpesviral DNA, with a sequence consistent with AsVH5-1, was detected. This virus is an established pathogen in donkeys and was isolated from clinically normal horses and donkeys at the same farm. Additional research is required to establish a cause and effect relationship between AsHV5-1 and pneumonia in horses.


The authors thank Dr Kimmons for providing samples and follow-up information.


  1. 1

    Temno evolution TT1820. Cardinal Health, Australia

  2. 2

    QIAamp DNA Blood Mini Kit, Qiagen, Valencia, CA

  3. 3

    Top Taq DNA Polymerase, Qiagen

  4. 4

    GDS-8000 Bioimaging System, UVP, Upland, CA

  5. 5

    ExoSAP-IT, USB (Affymetrix), Santa Clara, CA

  6. 6

    UTK Molecular Biology Resource Facility, The University of Tennessee, Knoxville, TN

  7. 7

    Lasergene, DNASTAR, Madison, WI

  8. 8

    TA cloning kit, Invitrogen, Carlsbad, CA

  9. 9

    Wizard Plus SV Minipreps, Promega, Madison, WI