Assessment of antigenic difference of equine influenza virus strains by challenge study in horses

We previously reported that horse antiserum against the Japanese equine influenza vaccine virus, A/equine/La Plata/1993 (LP93) exhibited reduced cross‐neutralization against some Florida sublineage Clade (Fc) 2 viruses, for example, A/equine/Carlow/2011 (CL11). As a result, Japanese vaccine manufacturers will replace LP93 with A/equine/Yokohama/aq13/2010 (Y10, Fc2). To assess the benefit of updating the vaccine, five horses vaccinated with inactivated Y10 vaccine and five vaccinated with inactivated LP93 were challenged by exposure to a nebulized aerosol of CL11. The durations of pyrexia (≥38.5°C) and other adverse clinical symptoms experienced by the Y10 group were significantly shorter than those of the LP93 group.

LP93 vaccine. Horses were vaccinated twice, one month apart, by intramuscular injection of monovalent non-adjuvanted vaccine.

| Challenge study
Two horses in each group (horses 1 and 2 of Y10 group, horses 6 and 7 of LP93 group) were experimentally challenged with 10 9.4 50% egg infectious dose (EID 50 ) of CL11 per horse, 2 weeks after the second vaccination as previously described. 5 The remaining three horses in each group (horses 3, 4, and 5 of Y10 group, horses 8, 9, and 10 of LP93 group) were similarly challenged 4 weeks after second vaccination.
Rectal temperatures were measured daily for 14 days postchallenge, and pyrexia was defined as ≥38.5°C. 6 Nasopharyngeal swabs were collected daily for 14 days after the challenge and virus isolation conducted as previously described. 5 Virus shedding was defined as ≥10 0.7 EID 50 /200 μL. Sera were collected on the day of the primary vaccination and on the challenge day. The experimental protocols were approved by the Animal Care Committee of Equine Research Institute of Japan Racing Association.

| Serological tests
Sera were treated with trypsin-heat-potassium metaperiodate to remove non-specific inhibitors. 7 Hemagglutination inhibition (HI) and virus neutralization (VN) titers were determined as previously reported. 2,7

| Data analysis
The mean rectal temperatures were analyzed with a two-way analysis of variance and post hoc Fisher LSD test between the groups on each day.
The mean durations (days) of pyrexia and virus shedding between the groups were compared using an unpaired Student's t-test. All statistical analyses were performed with graphpad prism 6 for Windows (GraphPad Software, Inc, San Diego, CA, USA). A level of P<.05 was considered significant. When geometric mean (GM) HI and VN titers were calculated, titers at <8 were provisionally considered four in this study.

| RESULTS
Horses were seronegative by HI and VN on the day of primary vaccination. The HI and VN titers on the challenge day (Day 0) are represented in Table 1. While the GM HI titers against CL11 of sera collected from Y10 and LP93 groups were similar (27.9 and 24.3, respectively), the GM VN titers against CL11 of sera collected from the Y10 group (48.5) were approximately 4.6-fold higher than the LP93 group (10.6). Horse 5 (Y10 group) showed no antibody response after vaccination even against the homologous virus, suggesting that the horse was a poor vaccine responder. By excluding Horse 5, the GM VN titer against CL11 was 8.5 times higher for Y10 group compared with the LP93 group ( Table 1).
The mean rectal temperatures of the Y10 group were significantly lower than LP93 group for 5 days (Fig. 1), and the mean duration of pyrexia of the Y10 group (days±SD) (0.4±0.9) was significantly shorter than the LP93 group (3.0±1.9, P=.023). Whereas all the horses in LP93 group exhibited virus shedding after challenge, there was no evidence of virus shedding by two horses in the Y10 group (Table 2). However there was no significant difference in the mean durations (days±SD) of virus shedding between the two groups (P=.294).

| DISCUSSION
We previously reported that the VN titers of horse LP93 antiserum were ≥eightfold lower against Fc2 viruses with HA A144V T A B L E 1 Hemagglutination inhibition and virus neutralization titers against vaccine viruses (Y10 and LP93) and the challenge virus (CL11) on the day of challenge, Day 0 .7 (No EIV was isolated from a nasopharyngeal swab).
indicating the superior efficacy of the Y10 vaccine compared with the LP93 vaccine.
In summary, this study demonstrates the superior efficacy of the Y10 vaccine compared with the LP93 vaccine, against an Fc2 virus carrying HA A144V substitution. The increase in protection against virus challenge may be due to the higher VN titers against CL11 induced by vaccination with the Y10 vaccine. As similar HI antibody levels were induced by both vaccines, it appears that VN antibody measurement is more likely, than HI, to identify antigenic differences in influenza viruses, which could affect vaccine efficacy in horses.