Descriptive epidemiologic study of disease associated with influenza virus infections during three epidemics in horses.

OBJECTIVE: To describe 3 epidemics of respiratory tract disease caused by influenza virus infections in a large population of horses. DESIGN: Cross-sectional and prospective longitudinal observational studies. ANIMALS: All horses stabled at a Thoroughbred racetrack. PROCEDURES: During a 3-year period, descriptive information was collected as horses arrived at the racetrack and throughout race meetings. Routine observations and physical examinations were used to classify horses' disease status. Cause of epidemics was established by use of serologic testing and identification of influenza virus in nasal secretions. RESULTS: An epidemic of respiratory tract disease caused by influenza virus infections was identified during each year of the study. Attack rates of infectious upper respiratory tract disease (IURD) ranged from 16 to 28%. Incidence of disease caused by influenza virus infections during racing seasons in the second and third years was 27 and 37 cases/1,000 horses/mo, respectively. Physical distributions of stall locations revealed that affected horses were stabled throughout the population; horses affected later in epidemics were often clustered around horses affected earlier. Mucopurulent nasal discharge and coughing were observed in 83 and 62% of horses with IURD, respectively. Median duration of clinical disease was 11 days. Serologic testing was the most sensitive method used to detect influenza virus infections; 76% of affected horses seroconverted to influenza virus. CONCLUSIONS AND CLINICAL RELEVANCE: Epidemics of IURD were observed annually in association with influenza virus infections. Few precautions were taken to limit spread of infection. Preventing or decreasing the likelihood of exposure and improving immunity in the population could substantially decrease risk of disease in similar populations.

Risk factors for disease associated with influenza virus infections during three epidemics in horses.

OBJECTIVE: To identify risk factors associated with respiratory tract disease in horses during 3 epidemics caused by influenza virus infections. DESIGN: Cross-sectional and prospective longitudinal observational studies. ANIMALS: 1,163 horses stabled at a Thoroughbred racetrack. PROCEDURES: Investigations were conducted during a 3-year period. An epidemic of respiratory tract disease caused by influenza virus infections was identified in each year. Routine observations and physical examinations were used to classify horses' disease status. Data were analyzed to identify factors associated with development of disease. RESULTS: Results were quite similar among the epidemics. Concentrations of serum antibodies against influenza virus and age were strongly associated with risk of disease; young horses and those with low antibody concentrations had the highest risk of disease. Calculation of population attributable fractions suggested that respiratory tract disease would have been prevented in 25% of affected horses of all horses had high serum antibody concentrations prior to exposure. However, recent history of vaccination was not associated with reduction in disease risk. Exercise ponies had greater risk of disease than racehorses, which was likely attributable to frequent horse-to-horse contact. CONCLUSIONS AND CLINICAL RELEVANCE: Particular attention should be paid to young horses, those with low serum antibody concentrations, and horses that have frequent contact with other horses when designing and implementing control programs for respiratory tract disease caused by influenza virus infections. It appears that control programs should not rely on the efficacy of commercial vaccines to substantially reduce the risk of disease caused by influenza virus infections.

Efficacy of a commercial vaccine for preventing disease caused by influenza virus infection in horses.

OBJECTIVE: To evaluate efficacy of a commercial vaccine for prevention of infectious upper respiratory tract disease (IURD) caused by equine influenza virus. DESIGN: Double-masked, randomized, controlled field trial. ANIMALS: 462 horses stabled at a Thoroughbred racetrack. PROCEDURE: Vaccine or saline solution placebo was administered 4 times in the population at 6-week intervals. The vaccine contained 3 strains of inactivated influenza virus, and inactivated equine herpesvirus type 4. Horses received 1 or 2 doses of vaccine or placebo prior to onset of a natural influenza epidemic, and were examined 5 d/wk to identify and monitor horses with IURD. Serum antibody concentrations were determined, and virus isolation was performed. RESULTS: Vaccination of horses prior to the influenza epidemic did not result in significant decrease in risk of developing respiratory tract disease. Severity of clinical disease was not different between affected vaccinated horses with IURD and controls with IURD, but median duration of clinical disease was 3 days shorter in vaccinated horses. Serum concentrations of antibodies to H3N8 influenza viruses were lower prior to initial vaccination in horses that were sick during the epidemic, and did not increase in these horses in response to vaccination. On arrival at the racetrack, young horses had lower antibody concentrations than older horses, and did not respond to vaccination as well. CONCLUSIONS AND CLINICAL RELEVANCE: Vaccination was of questionable benefit. A greater degree of protection must be obtained for influenza vaccines to be effective in protecting horses from IURD. Objective field evaluations of commercial vaccines are needed to adequately document their efficacy.

Cell-mediated cytolysis of equine herpesvirus-infected cells by leukocytes from young vaccinated horses.

The objective of this study was to determine whether the administration of modified-live equine herpesvirus (EHV-1) to young horses with residual maternal antibodies stimulated EHV-specific cytolytic responses, and whether these responses were crossreactive between EHV-1 and EHV-4. Eighteen clinically normal Belgian cross-foals were used in the study and were commingled in two adjacent pens. Skin biopsies were harvested from 16 foals within 24 h of birth and fibroblast cultures were established, expanded and cryopreserved. Beginning at approximately 10 weeks of age, 10 randomly chosen foals were inoculated on days 0, 21, and 43 of the study with a vaccine containing modified-live EHV-1. Blood mononuclear leukocytes were obtained on days 0, 32, and 50 for the assessment of EHV-specific cytolytic activity using 5 h and 18 h chromium release assays. EHV-1-specific antibodies were assessed by enzyme-linked immunosorbent assay using serum collected on days -21, 0, 32, and 50 of the study. Lymphocyte blastogenic tests and bioassays for interferon activity were conducted on day 50. After two vaccinations, mononuclear leukocytes from seven of ten vaccinated foals had cytolytic activity against autologous EHV-1 cells and leukocytes from six of ten lysed EHV-4-infected cells when tested in an 18 h assay. This activity was enhanced by exogenous interleukin 2 and was markedly reduced using target cells from unrelated horses. Cytotoxicity was not detected in a 5 h assay following in vitro stimulation of leukocytes. After three vaccinations, blood leukocytes from 6/6 vaccinated foals and 0/6 unvaccinated foals had proliferative responses EHV-1. There were no significant differences in interferon production by leukocytes from these foals. Twelve foals tested had low concentrations of (maternal) EHV-1-specific antibody prior to vaccination. Five of eight foals tested had increases in EHV-specific antibodies, while 4/4 commingled unvaccinated foals had a decrease or no change in EHV-specific antibodies. These results demonstrate cytotoxic cellular immune responses can be induced in young horses with maternal antibodies following administration of modified-live vaccine.

Nitric oxide production following in vitro stimulation of ovine pulmonary alveolar macrophages.

Production of inducible nitric oxide (NO) as measured by nitrite in supernatant from ovine pulmonary alveolar macrophage (PAM) cultures was assessed following stimulation of PAM with live cells and supernatants from Corynebacterium pseudotuberculosis and Pasteurella haemolytica cultures; purified bacterial lipopolysaccharide derived from both Escherichia coli and Pasteurella haemolytica alone and in combination with interferon-gamma or lymphocyte conditioned medium; or ovine lentivirus. PAM cultured ex vivo with no further stimulation for 24 h, 48 h or 72 h, produced low concentrations of NO that was not substantially increased following co-culture by the various additives. Assessment of NO production in PAM cultures containing P. haemolytica or supernatant from P. haemolytica cultures was complicated by production of high levels of nitrite in the bacterial cultures. Species differences in inducible NO production may affect the efficacy of clearance of bacterial infections and be responsible for inter-host differences in disease expression following infection by intracellular pathogens.

The relationship between single radial hemolysis, hemagglutination inhibition, and virus neutralization assays used to detect antibodies specific for equine influenza viruses.

Antibodies specific for equine influenza viruses are usually quantified using single radial hemolysis (SRH), hemagglutination inhibition (HI) or virus neutralization (VN). Neutralizing antibodies are thought to provide optimum protection to challenged animals. The purpose of this study was to determine the extent to which SRH and HI assays detect antibodies which neutralize equine influenza viruses. Acute and convalescent sera from 41 horses were analyzed using VN, SRH, and HI assays. These horses were present in a population of Thoroughbred racehorses during an epidemic of upper respiratory tract disease associated with influenza A/equine/Saskatoon/1/91 (H3N8), infections. Concentrations of antibodies binding to influenza A/equine/Kentucky/1/81 (H3N8), A/equine/Miami/1/63 (H3N8), and A/equine/Prague/1/56 (H7N7) were determined. Results of the VN assay were compared with results from the SRH and HI assays for acute antibody levels, changes in antibody concentrations between acute and convalescent sampling, and the occurrence of seroconversion. The correlation between assays for pre-exposure antibody levels ranged from 88% to 96%. The correlation between assays for change in antibody concentration ranged from 83% to 90% for the H3N8 viruses. This study shows that antibody concentrations specific for equine influenza virus, measured using SRH and HI assays, are highly correlated with concentrations detected using a VN assay.

The effect of changing single radial hemolysis assay method when quantifying influenza A antibodies in serum.

Single radial hemolysis (SRH) is a simple method for measuring concentrations of antibody specific for influenza A viruses. There are many differences in reported assay methods and these changes may significantly affect results. This study evaluated the effect of changing erythrocyte concentrations, influenza virus concentrations used to sensitize erythrocytes, incubation times, and incubation temperatures. These factors were shown to significantly affect results of the SRH assay. A method for standardizing the assay and expressing results as relative antibody concentrations is described. Standardized relative antibody concentrations show significantly less variability than do hemolysis zone measurements when the SRH assay method is changed. This paper demonstrates that use of this method would allow for more consistent results to be obtained within laboratories. Use of this method with quantified, internationally standardized sera would allow results of the single radial hemolysis assay to be directly compared between laboratories.

Evaluation of Directigen Flu A assay for detection of influenza antigen in nasal secretions of horses.

The Directigen Flu A assay (Becton Dickinson, Microbiology Systems, Mississauga, Ontario, Canada) is a commercially available immunoassay designed for rapid in vitro recognition of influenza A nucleoprotein. The purpose of this study was to evaluate this assay for detection of influenza virus in nasal secretions of naturally infected horses. The assay was shown to react with representative strains of influenza virus which cause disease in horses and did not react with nasal secretions from uninfected horses kept in isolation. Between 33% and 45% of nasal secretions specimens obtained from clinically diseased horses during influenza epidemics reacted positively in the assay and 95% to 98% of horses not showing signs of disease during influenza epidemics tested negative. In contrast, influenza virus was isolated from only 7% of diseased horses using conventional techniques. Diseased horses which were positive in the Directigen assay had lower pre-exposure influenza antibody concentrations and showed more clinical signs than diseased Directigen-negative horses. This evaluation demonstrates that the Directigen Flu A assay detects influenza virus in nasal secretions of infected horses and is more sensitive than virus isolation.

Cellular and antibody responses to equine herpesviruses 1 and 4 following vaccination of horses with modified-live and inactivated viruses.

The ability of monovalent and bivalent equine herpesvirus (EHV) vaccines to stimulate cellular and antibody responses to EHV-1 and EHV-4 was compared in healthy horses. Comparison of data from lymphocyte blastogenesis tests in which live viruses were used as antigens and that were conducted prior to vaccination and after 2 vaccinations revealed that horses given modified-live EHV-1 had significant increases in proliferative responses to EHV-1 (P = 0.03) and EHV-4 (P = 0.04). Responses to EHV-1 and EHV-4 in horses given the inactivated-virus bivalent vaccine were less; however, significant differences were not noticed when postvaccinal lymphocyte blastogenesis tests were compared between the groups of vaccinees. Interleukin-2 activity was not detected in leukocyte cultures from either group of vaccines following stimulation with live EHV-1 or EHV-4; however, interferon activity was found in similar cultures from both groups of vaccinees. For EHV-4, interferon activity in cultures from both groups of vaccinees was significantly (P < 0.05) greater than that in leukocyte cultures from unvaccinated controls. Both vaccines induced significant (P < 0.05) increases in serum antibodies that neutralized EHV-1 infectivity. The ELISA for EHV-1 and EHV-4 antibodies revealed that both vaccines induced significant (P < 0.05) increases (compared with preinoculation values) in antibodies reactive with these 2 types of EHV. Total serum antibody responses, as measured by ELISA, to EHV-1 and EHV-4 were significantly (P < 0.05) higher in horses that received the bivalent inactivated-virus vaccine, compared with that in horses that received monovalent vaccine. Evaluation of these data revealed that vaccination with modified-live EHV-1 can stimulate cellular and antibody responses that cross-react with EHV-4.

Effect of influenza A/equine/H3N8 virus isolate variation on the measurement of equine antibody responses.

This study has tested the effect of using homologous or heterologous equine influenza A virus isolates to evaluate serum antibody levels to influenza A virus in vaccinated and naturally-infected horses. In addition, the potential effect of antigenic selection of virus variants in egg versus tissue culture propagation systems was studied. Serum antibody levels in samples from horses recently infected with a local influenza A virus isolate (A/equine 2/Saskatoon/1/90) or recently vaccinated with a prototype isolate (A/equine 2/Miami/1/63) were assessed by hemagglutination inhibition and by single radial hemolysis using cell or egg-propagated A/equine 2/Saskatoon/1/90, A/equine 2/Miami/1/63 or A/equine 2/Fontainebleau/1/79. There were no significant differences in hemagglutination inhibition or single radial hemolysis antibody levels obtained with homologous or heterologous isolates or between viruses propagated in either eggs or cell culture. However there was a trend to higher titers in the hemagglutination inhibition assay when cell-propagated virus was used. These results suggest that antigenic variation in equine influenza A virus isolates and host-cell selection of antigenic variants during virus propagation may not be of sufficient magnitude to influence serological evaluation of antibody responses by hemagglutination inhibition or single radial hemolysis.