Absence of equid herpesvirus-1 reactivation and viremia in hospitalized critically ill horses

Background: Equid herpesvirus‐1 (EHV‐1) reactivation and shedding can occur in latently infected, asymptomatic animals. Risk factors for reactivation include stress and illness. The risk of asymptomatic shedding in hospitalized, critically ill horses with acute abdominal disease is unknown. This information is important to assess the need for additional biosecurity protocols to prevent spread of EHV‐1 in hospitalized critically ill horses with acute abdominal disorders. Objectives: To determine the frequency of reactivation and nasal shedding of EHV‐1 in hospitalized critically ill horses. Animals: One hundred twenty‐four client‐owned horses admitted to the Veterinary Teaching Hospital with acute abdominal disorders were included in the study. Methods: Cross‐sectional study examining the risk of reactivation of EHV‐1 in horses admitted with acute, severe, gastrointestinal disease. Whole blood and nasal secretions were collected throughout hospitalization. In addition, mandibular lymph nodes were collected from 9 study horses and 26 other Michigan horses. All samples were tested for the presence of EHV‐1 nucleic acid by real‐time PCR assays targeting the glycoprotein B gene and the polymerase (ORF 30) gene. Results: One hundred and twenty‐four horses met the inclusion criteria. None of the samples were positive for EHV‐1 DNA. Conclusion and Clinical Importance: These results suggest that nasal shedding and viremia of EHV‐1 in hospitalized critically ill horses with acute abdominal disorder is extremely rare. Implementation of additional biosecurity protocols to limit aerosol spread of EHV‐1 among horses with acute abdominal disease and other hospitalized horses is not necessary.     

Intranasal vaccination of pigs against pseudorabies: absence of vaccinal virus latency and failure to prevent latency of virulent virus

The avirulent Bartha's K strain of pseudorabies virus (PRV) was used to vaccinate 8 pigs at 10 weeks of age by the intransal route (experiment 1). On postvaccination days (PVD) 63 and 91, pigs were treated with corticosteroids. Viral shedding could not be detected. Explant cultures of trigeminal ganglia and tonsils did not produce virus. Four pigs with maternal antibody were vaccinated intranasally with Bartha's (attenuated) K strain of PRV at 10 weeks of age and were challenge exposed with a virulent strain of PRV on PVD 63 (experiment 2). Corticosteroid treatment, starting on postchallenge exposure day 70 (PVD 133) resulted in viral shedding in 1 of 4 pigs. In another pig of these 4, a 2nd corticosteroid treatment was required to trigger reactivation. In both pigs, sufficient reactivated virus was excreted to infect susceptible sentinel pigs. Restriction endonuclease analysis indicated that viruses isolated from the 2 pigs after challenge exposure and corticosteroid treatment were indistinguishable from the virulent virus. Evidence was not obtained for simultaneous excretion of vaccinal and virulent virus. Of 4 pigs without maternal antibody vaccinated twice with 1 of 2 inactivated PRV vaccines, challenge exposed on PVD 84, and treated with corticosteroids on postchallenge exposure day 63 (PVD 147), 1 was latently infected, as evidenced by the shedding of PRV (experiment 3). However, its sentinel pig remained noninfected.(ABSTRACT TRUNCATED AT 250 WORDS)     

Abortogenic viruses in horses

Viral causes of abortion include equine viral arteritis (EVA) and infection with equine herpesviruses‐1 and ‐4 (EHV‐1 and EHV‐4). Transmission of equine arteritis virus (EAV) occurs through respiratory, venereal or transplacental routes. Horizontal respiratory transmission of EAV results from exposure to infective nasopharyngeal secretions from acutely infected horses. For this transmission to occur, direct and close contact between horses is necessary. Venereal infection is an efficient method of transmission, with seroconversion of 85 to 100% of seronegative mares bred to virus shedding stallions. Asymptomatic carrier stallions are the essential natural reservoir of equine arteritis virus. Equine herpesviruses‐1 and ‐4 infect a susceptible host, replicate and establish a lifelong latent infection without any associated clinical signs. Reactivation of latent infections can result from factors such as stress and intercurrent disease. The control of these diseases is by implementation of appropriate management and hygiene measures, supplemented by vaccination and, in the case of EVA, by the identification of persistently infected stallions, which can be removed from breeding or continue to be bred to if managed under controlled conditions to prevent the risk of an outbreak of the disease.     

Experimental reactivation of equine herpesvirus‐3 following corticosteroid treatment

State of latency, well known for several herpesviruses, has been proposed for equine herpesvirus‐3 (EHV‐3) and supported by epidemiological observations. No detailed assessment about reactivation, patterns of excretion and re‐excretion has been formally reported. An experimental reactivation study by corticosteroid treatment in previously naturally infected horses was therefore carried out. Two polo mares with clinical and virologically confirmed history of equine coital exanthema were injected with dexamethasone and prednisolone on 3 successive days. Clinical signs, body temperature and clinical samples for virological and serological studies were obtained daily. Mares did not show any systemic clinical signs or hyperthermia. EHV‐3 shedding, seroconversion and the presence of a small lesion were observed in one of the mares under study 2 weeks after corticosteroid treatment. The results demonstrate that this virus exhibits a latency‐reactivation behaviour similar to that of other alpha herpesviruses. Reactivation of latency may have an important bearing on the appearance of clinical signs in mares and/or stallions during the breeding season without the actual evidence of transfer from mare to stallion or vice versa.     

Detection and quantification of equine herpesvirus-1 viremia and nasal shedding by real-time polymerase chain reaction.

Equine herpesvirus-1 (EHV-1) infection is common in young horses throughout the world, resulting in respiratory disease, epidemic abortion, sporadic myelitis, or latent infections. To improve on conventional diagnostic tests for EHV-1, a real-time polymerase chain reaction (PCR) technique was developed, using primers and probes specific for the EHV-1 gB gene. Amplification efficiencies of 100% +/- 5% were obtained for DNA isolated from a plasmid, infected peripheral blood mononuclear cells (PBMCs), and nasal secretions from infected ponies. The dynamic range of the assay was 8 log10 dilutions, and the lower limit of detection was 6 DNA copies. Fifteen ponies, seronegative for EHV-1, were experimentally infected with EHV-1, and nasal samples were used to quantify shedding of virus by both virus isolation and real-time PCR analysis. Virus isolation identified nasal shedding of EHV-1 in 12/15 ponies on a total of 25 days; real-time PCR detected viral shedding in 15/15 ponies on 75 days. Viremia was quantified using PBMC DNA, subsequent to challenge infection in 3 additional ponies. Viremia was identified in 1/3 ponies on a single day by virus isolation; real-time PCR detected viremia in 3/3 ponies on 17 days. When real-time PCR was used to analyze PBMC DNA from 11 latently infected ponies (documented by nested PCR), EHV-1 was not detected. We conclude that real-time PCR is a sensitive and quantitative test for EHV-1 nasal shedding and viremia and provides a valuable tool for EHV-1 surveillance, diagnosis of clinical disease, and investigation of vaccine efficacy.     

In vivo and in vitro reactivation of latent pseudorabies virus in pigs born to vaccinated sows

Latency of pseudorabies virus (PRV) was established in 8 of 9 pigs born to 2 vaccinated sows. Pigs had high, low, or no maternal antibody titers at the time of the initial inoculation. At postinoculation months 3 to 4, latent PRV could be reactivated in vivo by the administration of large doses of corticosteroids. In most pigs, the stress-simulating treatment resulted in recrudescence of virus shedding after lag periods of 4 to 11 days. In 3 pigs, virus shedding was without clinical signs of disease, whereas clinical signs that developed in 4 pigs appeared to be due to the corticosteroid treatment, rather than to the reactivation of PRV. Pigs with a log10 neutralizing antibody titer of less than or equal to 2.55 at the onset of corticosteroid treatment had a booster response. Reactivated PRV spread to sentinel pigs housed with the inoculated pigs. Reactivation of PRV was also demonstrated in vitro. Explant cultures of trigeminal ganglia from pigs killed between postinoculation months 4 to 5 produced infectious virus. Restriction endonuclease analysis indicated that the reactivated PRV was indistinguishable from virus isolated shortly after the primary infection. Seemingly, pigs with maternal antibodies can become latently infected and therefore may be regarded as potential sources of dissemination of PRV.     

Evaluation of orally administered valacyclovir in experimentally EHV1-infected ponies

The purpose of the current study was to investigate the therapeutic efficacy of valacyclovir against EHV1 in a controlled study. Eight na?ve Shetland ponies were inoculated with 10(6.5) TCID(50) of the neuropathogenic strain 03P37. Four ponies were treated with valacyclovir at a dosage of 40mg/kg bodyweight, 3 times daily, for 5 (n=2) or 7 (n=2) consecutive days, while the other four ponies served as untreated controls. The treatment regimen started 1h before inoculation. Ponies were monitored daily for clinical signs. At 0, 1, 2, 3, 4, 5, 7, 9, 11, 14, 17 and 21 days post inoculation (d pi), a nasopharyngeal mucus sample was taken to determine viral shedding. At the same time points, blood was collected and peripheral blood mononuclear cells (PBMC) were isolated to determine viremia. During the treatment, blood samples were collected 6 times daily, i.e. just before valacyclovir administration and 1h later, to determine the concentration of acyclovir in plasma. Also a nasopharyngeal swab was taken to measure the acyclovir concentration in nasal secretion. No differences could be noticed between valacyclovir-treated and untreated ponies. The clinical signs, the viral shedding and the viremia were similar in both the groups. Plasma acyclovir concentration could be maintained above the EC(50)-value of EHV1 during 50% of the entire treatment period in valacyclovir-treated ponies. Acyclovir could be detected in nasal swabs at concentrations varying from 50% to 100% of the corresponding plasma concentration. Although sufficiently high acyclovir levels could be reached in plasma and nasal mucus, no effect was seen of the treatment with valacyclovir on clinical signs, viral shedding and viremia of EHV1-infected ponies.     

Efficacy of oral supplementation with L-lysine in cats latently infected with feline herpesvirus

OBJECTIVE: To examine the effects of orally administered L-lysine on clinical signs of feline herpesvirus type 1 (FHV-1) infection and ocular shedding of FHV-1 in latently infected cats. ANIMALS: 14 young adult, FHV-1-naive cats. PROCEDURE: Five months after primary conjunctival inoculation with FHV-1, cats were rehoused and assigned to receive 400 mg of L-lysine in food once daily for 30 days or food only. On day 15, all cats received methylprednisolone to induce viral reactivation. Clinical signs of infection were graded, and viral shedding was assessed by a polymerase chain reaction assay throughout our study. Peak and trough plasma amino acid concentrations were assessed on day 30. RESULTS: Fewer cats and eyes were affected by conjunctivitis, and onset of clinical signs of infection was delayed on average by 7 days in cats receiving L-lysine, compared with cats in the control group; however, significant differences between groups were not demonstrated. Significantly fewer viral shedding episodes were identified in the treatment group cats, compared with the control group cats, after rehousing but not following corticosteroid-induced viral reactivation. Mean plasma L-lysine concentration was significantly increased at 3 hours but not at 24 hours after L-lysine administration. Plasma arginine concentration was not significantly altered. CONCLUSIONS AND CLINICAL RELEVANCE: Once daily oral administration of 400 mg of L-lysine to cats latently infected with FHV-1 was associated with reduced viral shedding following changes in housing and husbandry but not following corticosteroid administration. This dose caused a significant but short-term increase in plasma L-lysine concentration without altering plasma arginine concentration or inducing adverse clinical effects.     

Detection and Management of an Outbreak of Equine Herpesvirus Type 1 Infection and Associated Neurological Disease in a Veterinary Teaching Hospital

Background: Because of the serious disease sequelae associated with equine herpesvirus type 1 (EHV‐1) infections, awareness and control measures used to control outbreaks are important issues for all horse populations. Objectives: Describe the occurrence and management of an outbreak of EHV‐1 infection at a veterinary hospital. Animals: Horses hospitalized at a referral veterinary hospital. Methods: A horse with myeloencephalopathy associated with EHV‐1 infection (EHM) was admitted for diagnostic evaluation and treatment under strict infection control procedures. We describe the occurrence and management of a nosocomial outbreak of EHV‐1 infections associated with admission of this patient. Results: Despite institution of rigorous biosecurity precautions at the time of admission of the index case, EHV‐1 infections spread to 6 other horses that were hospitalized at the James L. Voss Veterinary Teaching Hopsital, including 2 that served as sources of infection for horses on their home premises after discharge. Infection with EHV‐1 was confirmed by polymerase chain reaction (PCR) and by seroconversion documented by glycoprotein G ELISA. A voluntary quarantine was imposed and admissions were restricted to prevent additional horses from being exposed. Quarantine duration was abbreviated by serial testing of all horses with PCR. Conclusions and Clinical Importance: These findings illustrate the contagious disease risk that can accompany management of horses with EHM. Horses with active nasal EHV‐1 shedding should be isolated in an airspace that is separate from other horses by strictly enforced biosecurity and isolation procedures. Serial testing with PCR may be a useful adjunct to determine when the risk of transmission has been minimized.     

Nasal Shedding of Equid Herpesvirus Type 1 and Type 4 in Hospitalized,Febrile Horses

The objective of this study was to determine the prevalence of shedding of equid herpesvirus 1 (EHV-1) or EHV-4 in nasal swab samples from any febrile, hospitalized horses during a 1-year period. It was hypothesized that some fevers in horses are associated with viral replication following recrudescence of latent virus or following a horizontal viral infection prior to or during admission to a referral hospital. During the observational period, nasal swab samples were collected from 64 febrile and 10 nonfebrile hospitalized horses. Routine DNA extraction was performed, and a validated quantitative polymerase chain reaction (qPCR) assay was used to detect and quantify genomic EHV-1 and -4 DNA. Genomic DNA of EHV-4 was detected in the nasal swab specimen of 1 of 64 febrile horses. EHV-1 DNA was not detected in any of the febrile horses. Samples from all nonfebrile horses were negative for both viruses. Considering the known association between fever and shedding of EHV-1 and EHV-4, we anticipated finding a higher percentage of PCR-positive samples from febrile patients. Fevers detected were likely a result of active disease processes for which the horses were hospitalized; concurrent other diseases appeared not to affect viral recrudescence. Further studies are warranted to examine frequency and factors of EHV latency and reactivation.     

Detection of respiratory herpesviruses in foals and adult horses determined by nested multiplex PCR

A nested multiplex PCR was developed as a rapid (<12h), sensitive test for the simultaneous identification of equine herpesviruses (EHV1, EHV4, EHV2 and EHV5) in clinical samples from horses. Peripheral blood and nasal swab (NS) samples from 205 weanling Thoroughbred foals on 6 different studs over 3 consecutive seasons and from 92 adult horses without clinical signs of respiratory disease were examined using direct multiplex PCR of clinical samples (direct PCR) and conventional cell culture with differentiation of EHV in cell cultures by multiplex PCR. Multiplex PCR proved a sensitive and specific technique for the detection of EHV in cell culture and clinical samples. The technique described appeared equally sensitive as one using a single set of primers for individual EHV but reduced labour and reagent costs. Cell cultures showing cytopathic effect (CPE) were always positive for EHV on PCR. EHV were also detected by multiplex PCR in 11 samples which failed to show CPE. By a combination of multiplex PCR and cell culture or direct multiplex PCR, the presence of up to three EHV in the same sample was detected. Overall, EHV5 was detected by direct multiplex PCR of peripheral blood mononuclear cells (PBMC) and/or NS samples from 78% of foals and 47% of adult horses. Repeated sampling or cell culture in combination with multiplex PCR and with the incorporation of IL-2 in culture medium increased the sensitivity for detection of EHV in PBMC and demonstrated that EHV5 DNA could be identified in PBMC from 89% of foals and 100% of adult horses. EHV2 was identified from approximately 30% of foals, but was more frequently identified in samples from 17 foals with mild respiratory disease and was isolated infrequently from adult horses. EHV1 and EHV4 were identified uncommonly in any population in the current study.     

Equine herpesvirus 1‐associated ulcerative dermatitis in a horse

This case report describes the clinical and histopathological findings of an infection caused by equine herpesvirus‐1 (EHV‐1) in a horse showing respiratory signs and a papular, crusted and ulcerative dermatitis involving mucosae. This diagnosis was supported by real‐time PCR positive for EHV‐1 on nasal swabs and tissues.     

A survey of respiratory viruses in New Zealand horses

AIMS: To determine which viruses circulate among selected populations of New Zealand horses and whether or not viral infections were associated with development of respiratory disease.

METHODS: Nasal swabs were collected from 33 healthy horses and 52 horses with respiratory disease and tested by virus isolation and/or PCR for the presence of equine herpesviruses (EHV) and equine rhinitis viruses.

RESULTS: Herpesviruses were the only viruses detected in nasal swab samples. When both the results of nasal swab PCR and virus isolation were considered together, a total of 41/52 (79%) horses with respiratory disease and 2/32 (6%) healthy horses were positive for at least one virus. As such, rates of virus detection were significantly higher (p<0.001) in samples from horses with respiratory disease than from healthy horses. More than half of the virus-positive horses were infected with multiple viruses. Infection with EHV-5 was most common (28 horses), followed by EHV-2 (27 horses), EHV-4 (21 horses) and EHV-1 (3 horses).

CONCLUSIONS: Herpesviruses were more commonly detected in nasal swabs from horses with respiratory disease than from healthy horses suggesting their aetiological involvement in the development of clinical signs among sampled horses. Further investigation to elucidate the exact relationships between these viruses and respiratory disease in horses is warranted.

CLINICAL RELEVANCE: Equine respiratory disease has been recognised as an important cause of wastage for the equine industry worldwide. It is likely multifactorial, involving complex interactions between different microorganisms, the environment and the host. Ability to control, or minimise, the adverse effects of equine respiratory disease is critically dependent on our understanding of microbial agents involved in these interactions. The results of the present study update our knowledge on the equine respiratory viruses currently circulating among selected populations of horses in New Zealand.     

Attempted reactivation of latent bovine herpesvirus 1 infection in calves by infection with ruminant pestiviruses

Three calves, latently infected with bovine herpesvirus 1 (BHV 1), were each inoculated intranasally with 9 strains of ruminant pestivirus (BVDV). All three calves developed a biphasic pyrexia and a lymphopenia followed by a neutrophilia. They did not shed BHV 1 in their nasal secretions in the 14 days following BVDV inoculation, and their BHV 1 antibody levels remained static, as did those of 2 control calves not given BVDV. All five calves were subsequently shown to be latently infected with BHV 1 by the production of recrudescent infections following the administration of dexamethasone. BHV 1 was recovered from nasal secretions and there was a marked rise in BHV 1 antibody titres in the second week after dexamethasone administration.     

Evaluation of a nested PCR test and bacterial culture of swabs from the nasal passages and from abscesses in relation to diagnosis of Streptococcus equi infection (strangles)

REASONS FOR PERFORMING STUDY: Streptococcus equi is the cause of strangles in horses. To improve diagnostic sensitivity, development and evaluation of DNA-based methods are necessary. OBJECTIVES: To evaluate diagnostic methods and observe the pattern of bacterial shedding during natural outbreaks. METHODS: Two herds with natural outbreaks of strangles were visited over a period of 15 weeks and 323 samples originating from 35 horses investigated. The diagnostic use of a nested PCR test was evaluated using a collection of 165 isolates of Lancefield group C streptococci (species specificity) and swabs from nasal passages or from abscesses from horses infected with S. equi (diagnostic sensitivity). RESULTS: All 45 S. equi isolates tested positive in the nested PCR, whereas no amplicon was formed when testing the other 120 Lancefield group C isolates. A total of 43 samples were collected from 11 horses showing clinical signs of strangles during the study period. The diagnostic sensitivity for PCR test was 45% and 80% for samples from the nasal passages and abscesses, respectively; the corresponding diagnostic sensitivity for cultivation was 18% and 20%. The diagnostic sensitivity was significantly higher for PCR than for bacterial cultivation. Furthermore, the shedding of S. equi in 2 infected horse populations was evaluated. An intermittent shedding period of S. equi of up to 15 weeks was recorded in this part of the study. It was also shown that shedding of S. equi occurred both from horses with and without clinical signs. CONCLUSIONS AND POTENTIAL RELEVANCE: The nested PCR test represents a species-specific and -sensitive method for diagnosis of S. equi from clinical samples. It may, however, be desirable in future to develop detection methods with high diagnostic sensitivity and specificity without the potential problems inherent in nested PCR.     

Oral immunisation against classical swine fever (CSF): onset and duration of immunity

A recently developed competitive PCR for ovine herpesvirus 2 (OvHV-2) was used to examine the levels of viral DNA in nasal secretions and peripheral blood leukocytes (PBL) of lambs and adult sheep. Viral DNA first appeared in the PBL of most lambs after about 3 months of age and the levels remained relatively constant thereafter. In most of the lambs (83%, n=12), viral DNA was undetectable by PCR in nasal secretions prior to 5 months of age. A dramatic rise of OvHV-2 DNA levels in the nasal secretions occurred starting at 5-6 months of age, which peaked at approximately 7 months. The highest level recorded in lamb nasal secretions was 7.5x10(8)copies/2microg DNA which were 75,000-100,000-fold higher than the levels in PBL of the same lambs. In adult sheep (n=10), the viral DNA levels in both PBL and nasal secretions were relatively stable over the 13-month period of the study, which included a lambing season. The data strongly suggest that neonatal lambs are not an important source for the transmission of OvHV-2 to clinically susceptible species, and that the nasal cavity is an important portal for shedding of infectious OvHV-2 in sheep. Furthermore, this study failed to identify a seasonal pattern in levels of viral DNA in nasal secretions or PBL of adult sheep that would provide a basis for the traditionally held belief that clinical cases of malignant catarrhal fever are significantly associated with lambing ewes.     

P‐81 A retrospective study of equine sarcoidosis

The purpose of this retrospective study was to evaluate six cases of equine sarcoidosis for initial presenting symptoms, response to therapy and actual outcome. Dermatologists and dermatopathologists from Europe, the United States, Australia and Canada were contacted to obtain these six cases, as this is a rare disease. Signalment, clinical signs, histological findings, clinical management and outcome were determined via a questionnaire and compared to former reports. There was no age or breed predilection, and four of six horses were geldings. Age of onset ranged from 3 months to 17 years. Onset of the disease was insidious or rapid. Interestingly, in five of six cases, scaling began on the trunk (girth and shoulder). Scaling, crusting and alopecia were seen in all six horses. In one horse, clinical signs of systemic disease were reported and included intermittent fever, prescapular lymphadenopathy, depression, poor body condition and nasal discharge. Treatment included phenylbutazone, deworming agents, antibiotics, short‐term low‐dose corticosteroids, and 1–1.5 mg/kg of prednisolone. One horse showed a partial response to trimethoprim and sulfonamide, and five of six went into clinical remission with corticosteroid treatment. Five of six horses were still alive 1 year after diagnosis; one horse was diagnosed <12 months ago. Two horses are in complete remission 4 and 8 years after diagnosis. In both horses, clinical signs recurred after cessation of therapy and went into remission again with reintroduction of treatment. Both of these horses have been in remission for several years without therapy. Funding: Self‐funded.     

Antemortem detection of latent infection with neuropathogenic strains of equine herpesvirus-1 in horses

OBJECTIVE: To evaluate a technique for identifying horses latently infected with neuropathogenic strains of equine herpesvirus-1 (EHV-1). ANIMALS: 36 adult mares, 24 of which were experimentally infected as weanlings with neuropathogenic or nonneuropathogenic EHV-1. PROCEDURES: Mandibular lymph node (MLN) tissue was obtained from each horse via biopsy during general anesthesia. Purified DNA from MLNs was tested for EHV-1 DNA by use of a magnetic bead, sequencecapture, nested PCR assay. For MLNs that contained EHV-1 DNA, the 256-bp DNA fragments amplified via sequence-capture nested PCR were sequenced to determine the nucleotide at the polymorphic site that determines pathotype (ie, neuropathotype [G(2254)] or non-neuropathotype [A(2254)]). RESULTS: Latent viral DNA was detected in 26 of the 36 (72%) mares tested. Neuropathogenic and nonneuropathogenic EHV-1 genotypes were detected in the latently infected horses. In each mare previously infected with known EHV-1 pathotypes, the open reading frame 30 genotype of latent EHV-1 was identical to that of the strain that had been inoculated 4 to 5 years earlier. Latent viral DNA was detected in 10 of the 12 mares that were inoculated as weanlings with neuropathogenic strains of EHV-1. The detection rate of the sequence-capture PCR method for EHV-1 latency was double that of conventional nested or realtime PCR assays performed on the same MLN DNA preparations. CONCLUSIONS AND CLINICAL RELEVANCE: The magnetic bead, sequence-capture, nested PCR technique enabled low-threshold detection of DNA from latent neuropathogenic strains of EHV-1 in MLN specimens from live horses. The technique may be used to screen horses for latent neuropathogenic EHV-1 infection.