Materials and Methods

Animals and clinical examination—All protocols were approved by the Institutional Animal Care and Use Committee of Cornell University, and informed consent was obtained from owners of dogs prior to sample collection. Conjunctival samples were collected from 30 consecutive dogs admitted to the Cornell University Hospital for Animals Ophthalmology and Community Practice services with a diagnosis of naturally acquired idiopathic conjunctivitis. Each dog received complete physical and ophthalmic examinations including slitlamp biomicroscopy, indirect ophthalmoscopy, Schirmer tear testing, tonometry, fluorescein corneal staining, and rose bengal corneal staining. All ophthalmic examinations were performed by a board-certified veterinary ophthalmologist (ECL). Study exclusion criteria included Schirmer tear test values < 15 mm/min, intraocular pressure measurements > 25 mm Hg, corneal retention of fluorescein or rose bengal stains, and detection of any recognized etiology of conjunctivitis in the dog from historical information or clinical examination. These etiologies included, but were not limited to, frictional irritants (eg, distichiasis, ectopic cilia, and foreign bodies), abnormal eyelid conformation or function, keratoconjunctivitis sicca, trauma, uveitis, glaucoma, ocular surgery, orbital disease, and ulcerative keratitis. Additionally, dogs vaccinated within 4 weeks of admission were excluded from the study to reduce the likelihood of detecting vaccine virus.²⁰

Conjunctival samples were also collected from 30 dogs that served as the case-control population. Control dogs were selected by testing the next dog admitted to the hospital, typically for wellness examination, of similar age, breed, and sex to the dog with conjunctivitis. All control dogs received complete ophthalmic and physical examination as described for dogs with conjunctivitis to exclude the presence of ocular disease prior to conjunctival sample collection. Dogs vaccinated within 4 weeks of admission were excluded from the control population. Signalment, clinical examination findings, and the following historical and clinical information, when applicable, were recorded for each dog in the conjunctivitis and control groups: duration of clinical signs prior to admission, vaccination history, eye or eyes affected, concurrent or recent systemic diseases, current medication administration, frequency and duration of exposure to domestic dogs other than housemates in the 4 weeks prior to the onset of clinical signs, and number of dogs residing in the household.

Conjunctival sample collection—Two swab specimens were collected from the conjunctiva of each affected eye. For dogs with bilateral conjunctivitis and control dogs, swab specimens were collected from both eyes and the conjunctival samples were pooled for diagnostic testing. Sterile polyester-tipped swabs^a were moistened with sterile, preservative-free saline (0.9% NaCl) solution and gently rolled against the superior and inferior conjunctival fornices. All swab specimens were collected prior to the application of fluorescein or rose bengal stains. Swab specimens for virus isolation were collected prior to, and swab specimens for PCR assay were collected following, installation of a single drop of topical ophthalmic proparacaine. One swab specimen was placed in a sterile tube containing 3 mL of viral transport medium (ie, Leibovitz medium^b with 0.25% bovine serum albumin solution, 1% amphotericin B solution, and 1% gentamicin sulfate solution) and immediately processed for virus isolation. The second swab specimen was placed in a dry sterile tube, immediately frozen at −80°C, and stored for up to 4 weeks prior to PCR analysis.

Virus isolation—Tubes containing conjunctival swab specimens and viral transport medium were vortexed. Approximately 50% of the tube volume was inoculated onto cells and allowed to absorb for at least 1 hour with periodic agitation. Virus isolations were performed with A-72 canine cells and a laboratory-developed canine skin cell line in Eagle minimum essential medium^c with 10% fetal bovine serum,^d 5% serum replacement solution,^e 2% penicillin-streptomycin solution, 1% amphotericin B solution, and 1% gentamicin sulfate solution. Cultures were incubated at 37°C, evaluated at 24-hour intervals for cytopathic effect, subcultured every 5 to 7 days, and held for 21 days. Cell cultures with cytopathologic characteristics were evaluated with monoclonal antibodies or antisera (ie, adenovirus group–specific 2HX-2 monoclonal antibody^f or anti–CHV-1 polyclonal antiserum conjugated to fluorescein isothiocyanate^g) to confirm viral identifications.

PCR analysis—Swab specimens were placed in 0.5 mL of PBS solution for 5 minutes at room temperature (approx 20°C [68°F]) and then vortexed for 5 minutes. Swab specimens were removed, and 200 μL of the eluent was used for DNA extraction by use of a kit.^h The DNA was eluted in a 100-μL volume. During extraction of viral DNA, exogenous DNAⁱ was spiked into samples to monitor DNA extraction, and an internal control PCR reactionⁱ was performed to ensure the DNA was successfully extracted and there was no inhibition of PCR amplification. Negative control samples were included in the extraction process. For extraction of RNA, 140 μL of the eluent was removed and RNA was extracted by use of a kit.^j The RNA was eluted in a final 100-μL volume. Exogenous RNAⁱ was spiked into samples to serve as a monitor for RNA extraction and PCR amplification inhibition, and negative control samples were included in the extraction process.

Polymerase chain reaction assays for DNA viruses were performed with a proprietary master mixⁱ containing a DNA binding dye. Real-time PCR assay was performed by use of a thermal cycler.^k One-step reverse transcription–coupled PCR reaction for RNA viruses was performed with a proprietary master mixⁱ containing a DNA binding dye and the thermal cycler.^k Proprietary primersⁱ for the following target genes were used: CAV-2 hexon gene, canine distemper virus nucleocapsid gene, CHV-1 equine herpesvirus-1 tegument protein-like protein, canine parainfluenza virus nucleocapsid gene, canine respiratory coronavirus spike protein gene, influenza A virus nucleocapsid gene, and West Nile virus nucleocapsid gene. Detection limits per PCR reaction were 5 genomic copies for the CAV-2 and CHV-1 assays; 10 genomic copies for the canine distemper virus, canine parainfluenza virus, and canine respiratory coronavirus assays; and 20 genomic copies for the influenza A virus and West Nile virus assays. Positive and negative control samples for each virus were included in the thermocycling process.

Statistical analysis—The frequency of virus detection in conjunctival samples by either virus isolation or PCR methods was compared between dogs with conjunctivitis and dogs without conjunctivitis by use of the Fisher exact test. To determine the similarity of the conjunctivitis group and the case-control group, categorical variables (ie, sex, neuter status, breed, frequent exposure to domestic dogs other than housemates, and residing in a multiple-dog household) were compared between groups with the χ² test of independence or the Fisher exact test (where at least 1 cell value was < 5) and continuous variables (ie, age) were compared between groups by use of the Wilcoxon rank sum test. To determine whether differences in age, sex (male or female), neuter status (sexually intact or neutered), breed (purebred or mixed breed), duration of clinical signs prior to admission, bilaterality of clinical signs (bilateral or unilateral conjunctivitis), frequent exposure to domestic dogs other than housemates (defined as contact for ≥ 15 min/episode with an episode frequency ≥ 2/wk in the 4 weeks prior to the onset of clinical signs), and residing in a multiple-dog household were present between the dogs in the conjunctivitis group with and without viral detection in conjunctival samples by use of either virus isolation or PCR methods, the Fisher exact (categoric variables) and the Wilcoxon rank sum test (continuous variables) were used. Values of P ≤ 0.05 were considered significant for all comparisons.

Results

Viruses were detected by either virus isolation or PCR methods significantly (P = 0.01) more frequently in conjunctival samples from dogs with conjunctivitis (7/30 [23.3%]) than from dogs without conjunctivitis (0/30 [0%]). Canine herpesvirus-1 was isolated from conjunctival samples of 2 dogs and detected by use of PCR assay in samples of 5 dogs. Canine adenovirus-2 was isolated from a conjunctival sample of 1 dog and detected by use of PCR assay in samples of 2 dogs. Three conjunctival samples were positive for CHV-1 (2 samples) or CAV-2 (1 sample) by use of both virus isolation and PCR assay. One conjunctival sample was positive for CAV-2 by use of PCR assay only, and 3 conjunctival samples were positive for CHV-1 by use of PCR assay only. No conjunctival samples were positive for a virus on virus isolation and negative for a virus on PCR assays. Cell cultures from all conjunctival samples that had positive virus isolation results had cytopathological characteristics consistent with either CAV-2 or CHV-1, and all viral identifications were confirmed by use of a positive fluorescent antibody assay result. No additional viruses were detected by either virus isolation or PCR assay. Conjunctival samples were collected over a 12-month period, with approximately equal numbers of dogs included in each season of the year. Conjunctival samples with positive virus detection results were distributed equitably throughout the study period, and seasonal variations were not identified for viral detection. No significant difference was detected between dogs with conjunctivitis and case-control dogs for the frequency of the following variables: age, neuter status, sex, breed, frequent exposure to domestic dogs other than housemates, and residing in a multiple-dog household.

All dogs with conjunctivitis had varying degrees of conjunctival hyperemia, chemosis, ocular discharge, and blepharospasm. In addition to these lesions, conjunctival follicles were present in 4 dogs without viral detection in conjunctival samples. A single dog with CHV-1 detected in conjunctival samples also had scattered conjunctival petechiae. Subjectively, there were no differences in ocular lesions between dogs with viral detection and without viral detection, or between dogs with CHV-1 and CAV-2 detection, that would permit clinical differentiation of these groups. All dogs with conjunctivitis and positive viral detection results were evaluated for bilateral ocular lesions and a history of frequent exposure to dogs other than housemates in the preceding 4 weeks. This exposure included visits to boarding kennels (n = 3 dogs), veterinary hospitals (2), dog day care facilities (1), dog parks (1), dog shows (1), obedience classes (1), and canine breeding facilities (1). Two dogs with CHV-1 detected in conjunctival samples were receiving immunosuppressive medications at the time of admission, including systemic lymphoma chemotherapy and topical ophthalmic dexamethasone. Signalment, clinical, and historical findings and virologic data for the 7 dogs with conjunctivitis and positive viral detection results for conjunctival samples were summarized (Table 1).

A significant (P = 0.03) correlation was identified between viral detection and neuter status, with sexually intact dogs with conjunctivitis more likely to have positive virologic assay results. Additionally, frequent exposure to dogs outside the household was significantly (P = 0.01) associated with positive virologic assay results in the conjunctivitis group. There were no significant correlations between the following variables and virologic assay results within the conjunctivitis group: age (P = 0.08), sex (P = 1.0), breed (P = 0.37), duration of clinical signs (P = 0.54), bilaterality of clinical signs (P = 0.29), and residing in a multiple-dog household (P = 1.0).

Discussion

In the study reported here, CHV-1 and CAV-2 were detected in conjunctival samples from dogs with naturally acquired conjunctivitis, but not dogs without ocular disease. These findings suggest that CHV-1 and CAV-2 are infectious etiologic agents of conjunctivitis in dogs. Both viruses have anecdotally been described as etiologic agents of naturally acquired conjunctivitis in dogs,^21,22 but controlled studies to confirm the presence and prevalence of these viruses in conjunctival samples from dogs with and without conjunctivitis are lacking. Previous studies^23,24 investigating the role of CHV-1 and CAV-2 in spontaneous conjunctivitis in dogs were restricted to puppies with a specific subtype of conjunctivitis (ie, follicular conjunctivitis). In 1 study,²³ no evidence of infection with CHV-1 or canine adenoviruses was found in 20 dogs by use of serologic or PCR assays. In the other study,²⁴ investigators failed to detect herpesviruses in 12 dogs by use of immunohistochemistry and virus isolation. In these previous studies,^23,24 conjunctival samples were not collected from dogs with other more common forms of idiopathic conjunctivitis or from a canine population as large and diverse as was included in the present study.

Canine herpesvirus-1 is a member of the subfamily Alphaherpesvirinae and is biologically related to several herpesviruses that have been implicated in the development of conjunctivitis in other host species, including bovine herpesvirus-1, cervid herpesvirus-1, feline herpesvirus-1, herpes simplex virus-1, herpes simplex virus-2, and varicella-zoster virus.^11,12 Primary CHV-1 infection is acquired in utero, during parturition, or from mucosal contact with infected secretions.²⁵ In adult canids, primary CHV-1 infection is frequently subclinical or causes localized respiratory, genital, or mucosal disease.^26,27 Adult and immature naïve dogs experimentally infected with CHV-1 by ocular inoculation have been reported to develop self-limiting conjunctivitis and ocular viral shedding.²⁸ Following primary infection, lifelong latent CHV-1 infection is established in neurons of sensory ganglia and lymphatic tissues.²⁹ Reactivation of latent CHV-1 results in viral shedding and may be subclinical or associated with disease recrudescence, including genital mucositis and dendritic ulcerative keratitis.^30–32 Systemically administered and ophthalmic topically applied immunosuppressive medications have previously been linked to CHV-1 reactivation.^31,32 It is unclear whether dogs with conjunctivitis and CHV-1 detection in the present study represented primary or reactivated viral infection; however, it is plausible that both were included in the study population. Young dogs not receiving immunosuppressive medications and without concurrent immunomodulating systemic disease may have had primary infection. Older dogs receiving immunosuppressive medications, including systemic administration of chemotherapeutics and topical application of ophthalmic corticosteroids (the 8-year-old Labrador Retriever with lymphoma in remission and the 3-year-old Beagle with no recent history of disease, respectively; Table 1), may have developed recurrent infection and disease recrudescence. Subclinical ocular shedding of CHV-1 has been experimentally induced in dogs³¹ and occurs spontaneously with other alphaherpesviruses (eg, feline herpesvirus-1 and herpes simplex virus-1),³³ but has not been previously reported for dogs under natural conditions and was not detected in the current study.

Canine adenovirus-2 is a member of the family Adenoviridae and is primarily transmitted between dogs by oronasal contact.³⁴ It is a relatively common etiologic agent of canine infectious tracheobronchitis and replicates within the upper and lower respiratory tract mucosal epithelium.³⁴ Conjunctivitis associated with active laryngotracheitis has been reported for dogs from which CAV-2 was isolated from respiratory samples; however, attempts were not made to detect virus in ocular tissues.³⁵ Canine adenovirus-2 is biologically related to the group of adenoviruses that are considered the most common etiologic agent of viral conjunctivitis in humans.¹⁰ In humans, adenoviruses induce several distinct clinical syndromes.¹⁰ Conjunctivitis is present as a component of these clinical signs and may be an isolated lesion or associated with keratitis or respiratory tract disease. In this study, both dogs with CAV-2 detection in conjunctival samples had received a modified-live CAV-2 vaccine within 12 months of admission, suggesting that vaccinal immunity had waned or that vaccination does not prevent localized ocular infection or clinical disease in all dogs. Conversely, detection of vaccinal CAV-2 in the study dogs cannot be excluded as a possibility.²⁰

Both CHV-1 and CAV-2 are also upper respiratory tract pathogens in dogs.^26,36 In this study, 4 of 7 dogs with conjunctivitis and viral detection had a history of recent clinical signs consistent with respiratory tract infection; however, ocular abnormalities were not observed until after the resolution of respiratory disease in all dogs. This finding may reflect viral tropism for both respiratory and conjunctival epithelium, but with delayed clinical manifestation of ocular lesions. Discrepancies in CHV-1 and CAV-2 detection between virus isolation and PCR assays in the present study likely reflect the greater diagnostic sensitivity of PCR assays. Virus isolation also requires actively replicating virus for detection, so periods of positive PCR assay results and negative virus isolation results may occur with the presence of antibody neutralized virus or viral fragments or when viral viability is lost between the time of sample collection and testing. Detection of CHV-1 and CAV-2 in some study dogs by use of virus isolation confirms that actively replicating virus was present within the inflamed conjunctiva and implies that ocular secretions are capable of transmitting infection between dogs.

Risk factors identified in this study for viral detection in dogs with conjunctivitis likely result from the communicable nature of CHV-1 and CAV-2 in canine populations. Behavioral characteristics of sexually intact dogs may increase exposure to other dogs, increasing the opportunity for infection acquisition. This includes increased frequency of contacts with other dogs and increased direct contact with potentially infectious secretions. Dogs with recent and frequent exposure to other dogs in this study included dogs that had visited kennels, veterinary hospitals, dog day care, dog parks, dog shows, obedience classes, and trips for the purpose of breeding in the 4 weeks preceding the onset of detectable conjunctivitis. These environments might be important in the maintenance and dissemination of viral infections within and between canine populations.

Results of the present study suggest CHV-1 and CAV-2 may be relatively common etiologic agents of conjunctivitis in domestic dogs. Without the use of virus-specific diagnostic assays, the diagnosis would have been idiopathic or allergic conjunctivitis³⁷ for these dogs. Risk factors for viral conjunctivitis in dogs reflect increased exposure to other dogs and opportunities for contact with infectious secretions. This information may be useful in the development of targeted studies examining the role and prevalence of these viral agents in specific subpopulations of dogs with conjunctivitis. Adenoviruses and herpesviruses are susceptible to specific ophthalmic antivirals, and research to determine the safety and efficacy of these medications in dogs with viral conjunctivitis is warranted.