Evaluation of the protection induced by avian influenza vaccines containing a 1994 Mexican H5N2 LPAI seed strain against a 2008 Egyptian H5N1 HPAI virus belonging to clade 2.2.1 by means of serological and in vivo tests

Abstract

Since 2006 Egypt has been facing an extensive epidemic of H5N1 highly pathogenic avian influenza (HPAI) with a huge number of outbreaks both in rural and intensively reared poultry areas. The use of efficacious vaccines in this country has been, and still remains, essential for the control and possible eradication of HPAI. The present study was performed to establish whether the administration of inactivated vaccines containing an H5 virus belonging to a different lineage to the Eurasian H5N1 HPAI viruses guarantees protection from clinical signs, provides significant immune response and is able to achieve a reduction of viral shedding in the face of a challenge with a contemporary H5N1 virus isolated in Egypt. Despite the genetic and antigenic differences between the vaccine strain (H5N2/Mexico) and the challenge strain (H5N1/Egypt), confirmed by molecular and serological (haemagglutination inhibition) tests, it was established that the immune response induced by these conventional vaccines is sufficient to prevent infection in the majority of birds challenged with a contemporary H5N1 Egyptian strain. The data reported in this study also indicate that there may be a low degree of correlation between haemagglutination inhibition titres, clinical protection and reduction of shedding.

Introduction

Although the use of vaccines against avian influenza (AI) viruses in birds has been discouraged over the years, the unprecedented occurrence of outbreaks caused by AI viruses in recent times has required review of this policy. A variety of products are now available on the market, ranging from inactivated conventional to live recombinant products. The general consensus on the use of vaccination is that, complying with good manufactory practice standards and properly administered, birds will be more resistant to field challenge and will exhibit reduced shedding levels in case of infection.

Since 2006 Egypt has been facing an extensive epidemic of H5N1 highly pathogenic avian influenza (HPAI) with a huge number of outbreaks both in rural and intensively reared poultry. The use of vaccines in this country has been and still remains massive, with millions of doses of vaccine used in the field. Despite the use of vaccination, the number of outbreaks in animals and in humans is increasing.

The use of heterologous inactivated vaccine–challenge virus combinations has been widely described in the literature in experimental challenges and also applied widely in the field (Terregino et al., 2007; Bos et al., 2008; Van der Goot et al., 2008; Poetri et al., 2009). These types of vaccines have several advantages, including their use in differentiating infected from vaccinated animals (DIVA) strategies. This system was developed to support the eradication programmes in the presence of several introductions of AI viruses (Capua & Alexander, 2004; Capua et al., 2004). Previous studies have shown that vaccine containing the H5N2 A/chicken/Mexico/232/94/CPA strain is protective in chickens and ducks against H5N1 HPAI viruses of the Eurasian lineage despite the low haemagglutination (HA) homology between the vaccine strain and the field virus (Swayne et al., 2006; Van der Goot et al., 2008).

The present study was performed to establish whether administration of an inactivated vaccine containing an H5 virus belonging to a different lineage to the Eurasian H5N1 was protective against clinical signs and achieved a reduction of viral shedding in the face of challenge with a 2008 H5N1 virus isolated in Egypt.

Materials and Methods

Birds

Five groups of 10 specific pathogen free (SPF) chicks, hatched in isolation from SPF eggs supplied by Charles River Laboratories® Inc., males and females, were used. Chicks were reared in isolation and were individually identified by means of a numbered wing tag. Chickens were fed with a commercial compound suitable for their age. Birds had water and feed ad libitum.

Vaccines and vaccine administration

Volvac® AI KV and Volvac® AI+ND KV are emulsified, inactivated oil vaccines prepared with A/chicken/Mexico/232/94/CPA (H5N2), which for the bivalent vaccine is in combination with the La Sota Newcastle disease (ND) virus strain. Each 0.5 ml vaccine contains a minimum titre of 10^8.5 median embryo infective dose (EID₅₀) or 256 haemagglutinating units (HAU) of AI and 10^8.2 EID₅₀ or 128 HAU of ND viruses (bivalent vaccine only). The vaccine dose for the two vaccines under study was 0.5 ml per bird, inoculated in the lower (dorsal) part of the neck by the subcutaneous route.

Experimental design

Chicks were randomly distributed in five groups (Group 1 to Group 5) of 10 birds each. Birds of Groups 1 and 2, were vaccinated with Volvac® AI and birds in Groups 4 and 5 with Volvac® AI+ND at 21 days of age. Birds in Group 3 were left as unvaccinated controls. Birds were observed daily for clinical signs throughout the study. Four weeks after vaccination, chickens from Groups 1, 2 and 3 were challenged intranasally (IN) with 0.1 ml viral suspension containing 10⁶ EID₅₀/ml (corresponding approximately to 10 100% chicken lethal doses, based on in vivo tests performed prior to the trial; data not shown) of the challenge HPAI H5N1 virus A/chicken/Egypt/1709-6/08.

Blood samples for serology studies were obtained from each chicken of each group prior to each vaccination, before challenge and every week until 4 weeks after challenge (end of the study). In order to monitor virus shedding after challenge, tracheal and cloacal swabs from each challenged bird were obtained on days 3, 7, 10, 14 and 21 post challenge. These samples were used for real-time reverse transcriptase-polymerase chain reaction (RT-PCR) analyses and at the same time they were processed for virus isolation in SPF eggs.

Challenge virus

The challenge virus was selected from 10 Egyptian isolates provided by the Animal Health Research Institute, Egypt. Viruses were isolated from rural and industrial poultry farms at the end of 2007 and beginning of 2008. All of the viruses were passaged in embryonated SPF fowl's eggs once. Viruses were then titrated in embryonated SPF fowl's eggs and the EID₅₀ was calculated according to the Reed and Muench formula (Reed & Muench, 1938). All isolates were fully sequenced and A/chicken/Egypt/1709-6/08 was selected. This virus was isolated in 2008 from a poultry farm in the Quen-Quena District, Egypt. It clustered together with other 2008 isolates but separately from the 2007 isolates. This challenge virus belongs to the phylogenetic clade 2.2.1 (World Health Organization/World Organisation for Animal Health/Food and Agriculture Organization H5N1 Evolution Working Group, 2008), and was selected because it was considered the prototype of the viruses circulating in Egypt in 2008 according to the results of the phylogenetic analysis performed using available sequences of viruses isolated in Egypt between 2006 and 2008 (data not shown).

Preliminary indirect antigenic analysis of the challenge virus was carried out by cross-haemagglutination inhibition (HI) tests using serum produced with the vaccine strain A/chicken/Mexico/232/94 (H5N2) and a selection of antigens; namely, the virus contained in the vaccine (H5N2), a reference H5N1 isolate (A/chicken/Yamaguchi/7/04), the challenge virus and an Egyptian isolate considered representative of HPAI H5N1 viruses circulating in Egypt in 2007 (A/chicken/Egypt/1709-1/2007). These HI tests showed a very low HI cross-reactivity between the vaccine virus and the other viruses tested, including the challenge virus. More specifically, the HI titres of the serum produced with the vaccine strain were 1:2048, 1:512, 1:128 and 1:16 with the vaccine virus, a reference H5N1 isolate (A/chicken/Yamaguchi/7/04), the H5N1 Egyptian strain of 2007 (A/chicken/Egypt/1709-1/2007) and the challenge virus, respectively.

The nucleotide similarity between the HA gene of the vaccine strain and the respective HA gene of the challenge virus and the viruses used for serological investigations ranged between 72.5% and 74.3%. In particular, the degree of similarity for the vaccine strain was 72.5% with A/chicken/Egypt/1709-6/2008, was 73.7% with A/chicken/Egypt/1709-1/2007 and was 74.3% with A/chicken/Yamaguchi/7/2004.

The amino acid similarity between the HA1 gene of the vaccine strain and the respective gene of the challenge virus and the viruses used for serological investigations ranged between 80% and 84%. In particular the similarity for the vaccine strain was 80% with A/chicken/Egypt/1709-6/2007, was 83% with A/chicken/Egypt/1709-1/2007 and was 84% with A/chicken/Yamaguchi/7/2004.

Sampling

Birds were observed daily for clinical signs throughout the duration of the study. Following challenge, the occurrence of clinical signs was evaluated twice a day. Samples were collected and identified individually with a study number, date of collection, an individual bird identification number and a treatment group number.

Following challenge, tracheal and cloacal swabs were processed for attempted virus isolation in SPF eggs, and were analysed by real-time RT-PCR.

After collection, tracheal and cloacal samples were placed in 1 ml isotonic phosphate-buffered saline solution (pH 7.0 to 7.4). One hundred microlitres from each sample were used for RNA extraction. RNA was analysed by real-time RT-PCR.

The remaining sample was mixed with an equal volume of phosphate-buffered saline (pH 7.0 to 7.4) containing penicillin (2000 u/ml), streptomycin (2 mg/ml), gentamicin (0.05 g/ml) and mycostatin (1000 u/ml) for virus isolation attempts.

Serology

At least 2 ml blood were collected by puncture from the wing vein at each sampling.

Sera were tested immediately after harvesting in HI tests using 4 HAU of the challenge virus and the vaccine strain as antigens (OIE, 2008b). Sera collected from the birds vaccinated with Volvac® AI+ND were also tested for NDV antibodies (OIE, 2008a) using the Ulster 2C strain as antigen. The vaccine virus was identical to the seed virus and was supplied by the company that produces the vaccine as an antigen inactivated by treatment with 0.1% formalin (0.037% formaldehyde). The other viruses used in this study were inactivated by treatment with 0.05% β-propiolactone for 2 h at 37°C.

The sera of vaccinated birds were tested by an indirect immunofluorescence (iIFAT) anti-N1 antibody detection test (Capua et al., 2003).

Virus isolation

Virus isolation was performed in SPF embryonated fowl's eggs in accordance with the OIE guidelines (OIE, 2008b). Briefly, following clarification by centrifugation at 1000 x g, the supernatant fluid of the samples collected was inoculated into the allantoic sac of at least five SPF embryonated eggs of 9 to 11 days' incubation. The eggs were incubated at 36 to 37°C for 7 days. Allantoic fluids from eggs containing dead or dying embryos, and all embryos remaining viable at the end of the first passage were tested for HA activity. Fluids yielding a negative reaction were inoculated into at least one further batch of eggs. Positive samples collected from vaccinated and unvaccinated birds were titrated in SPF embryonated fowl's eggs by inoculation of serial dilutions, and the EID₅₀ was calculated according to the Reed and Muench formula.

Real-time RT-PCR

The swabs (tracheal or cloacal) were immersed in sterile phosphate-buffered saline. A 100 µl volume of supernatant from each sample was used for RNA extraction by conventional methods (Machery-Nagel, Duren, Germany) and was analysed immediately. A fixed amount of RNA was analysed by real-time RT-PCR for type A influenza virus (Spackman et al., 2002). Samples with a threshold cycle value ≤35 were considered positive for influenza type A viral RNA based on internal validation trials.

Overall evaluation of efficacy

In order to evaluate the efficacy of the vaccine by comparing the results obtained from testing different groups (vaccinated and unvaccinated chickens) challenged with the same virus, a bird was considered infected if it showed specific signs of HPAI; that is, appeared clearly sick or died and/or exhibited seroconversion (unvaccinated birds) or a significant increase of antibody titres (vaccinated birds) after challenge, and was positive for, at least, real-time RT-PCR or virus isolation from tracheal/cloacal swabs.

Statistical analysis

Serological data generated by HI were subjected to statistical analysis to evaluate the significance of the differences observed among and within the experimental groups. The methods for the assessment of statistical significance were the Wilcoxon–Mann–Whitney rank-sum test on unmatched data and the Wilcoxon matched-pairs signed-ranks test on matched data. These tests were applied to establish whether the pre-infection and post-infection HI titres differed significantly among and between experimental groups.

In order to determine the survival rate among vaccinated and unvaccinated groups Cox's regression analysis for the calculation of the hazard ratio was used. Data were analysed with the test for proportional hazard assumption before the applications of Cox's regression analysis. Finally, the log-rank test for equality of survival functions was used to compare the survival rate among groups.

Results

Antibody response

All the Volvac®AI-vaccinated and Volvac® AI+ND-vaccinated chickens showed high serological titres when tested with the homologous antigen (H5N2). By contrast, the majority of the vaccinated birds was below the threshold of positivity (≤1:16) by HI using the inactivated challenge virus (H5N1) as antigen. There was no evidence of statistical significance between pre-challenge and post-challenge titres generated with either antigen. At the end of the experimental trial, the uninfected vaccinated groups as well as the vaccinated and infected birds showed a slight decrease in antibody titres when tested with the homologous vaccine haemagglutinating antigen (H5N2). Unexpectedly, even after infection only a few subjects showed a slight increase of HI titre with the antigen produced with the challenge virus (H5N1). Details of the HI results are presented in Tables 1 and 2.

Table 1.  Serological log₂ haemagglutination inhibition titres in vaccinated groups.

	21 days post vaccination			7 days post challenge			14 days post challenge			21 days post challenge			28 days post challenge
Bird	H5N2	H5N1	NDV	H5N2	H5N1	NDV	H5N2	H5N1	NDV	H5N2	H5N1	NDV	H5N2	H5N1	NDV
Group 1: vaccinated with Volvac® AI
1	12	2		14	2		11	4		10	4		11	3
2	11	1		14	2		10	1		11	0		11	0
3	12	0		14	0		10	0		10	0		11	0
4	11	0		14	5		9	3		9	4		11	4
5	12	0		Dead	Dead		Dead	Dead		Dead	Dead		Dead	Dead
6	11	0		12	1		8	3		9	4		9	3
7	9	0		9	1		Dead	Dead		Dead	Dead		Dead	Dead
8	12	0		14	0		9	1		10	2		11	2
9	11	0		13	0		9	0		9	1		10	2
10	12	0		14	0		10	0		10	0		10	0
GMT	11.3	0.3		13.11	1.22		9.50	1.50		9.75	1.88		10.50	1.75
Group 2: vaccinated Volvac® AI+ND
21	12	0	6	10	0	7	9	1	5	7	2	5	8	1	6
22	13	0	7	15	4	8	13	3	6	10	3	6	10	2	6
23	12	0	6	11	0	6	11	0	4	7	0	5	8	0	5
24	12	0	7	11	0	8	10	1	6	10	1	6	9	1	7
25	11	0	6	11	1	7	10	2	4	10	2	5	9	2	5
26	10	0	5	10	0	6	9	0	3	6	0	4	7	0	4
27	12	4	6	13	3	8	11	2	5	10	3	6	9	2	7
28	11	0	6	11	1	8	10	1	4	10	2	5	9	1	6
29	10	0	6	Dead	Dead	Dead	Dead	Dead	Dead	Dead	Dead	Dead	Dead	Dead	Dead
30	10	0	6	10	1	7	9	1	5	9	1	5	8	1	6
GMT	11.3	0.4	6.1	11.33	1.11	7.22	10.22	1.22	4.67	8.78	1.56	5.22	8.56	1.11	5.78
H5N2=A/chicken/Mexico/232/94 antigen; H5N1=A/chicken/Egypt/1709-6/08 antigen; NDV=Ulster 2C antigen; GMT=geometric mean titre.

Table 2.  Serological log₂ haemagglutinatination inhibition titres in vaccinated (control) groups.

	21 days post vaccination			28 days post vaccination			35 days post vaccination			42 days post vaccination			49 days post vaccination
Bird	H5N2	H5N1		H5N2	H5N1		H5N2	H5N1		H5N2	H5N1		H5N2	H5N1
Group 4: Volvac® AI vaccinated and uninfected
11	10	3		13	3		10	3		9	3		9	3
12	11	0		13	0		11	0		9	0		9	0
13	12	1		14	1		13	1		11	1		10	2
14	12	0		14	0		14	1		10	2		10	2
15	12	0		14	0		10	0		10	0		9	0
16	9	0		11	0		14	0		9	0		9	0
17	12	0		13	0		11	0		10	0		10	0
18	12	0		13	0		12	0		9	0		9	0
19	10	0		13	0		12	0		8	0		8	0
20	9	0		13	0		12	0		7	0		6	0
GMT	10.9	0.4		13.1	0.4		11.9	0.5		9.2	0.6		8.9	0.7
Group 5: Volvac® AI+ND vaccinated and uninfected
31	11	0	6	11	0	8	10	0	5	9	0	5	9	0	5
32	11	0	6	11	0	7	10	0	5	8	0	4	9	0	7
33	10	0	6	10	0	7	10	0	6	9	0	5	9	0	7
34	11	3	7	12	3	8	9	3	6	9	3	5	9	2	6
35	12	0	4	10	0	5	9	1	3	7	0	3	7	0	4
36	10	0	6	12	1	8	8	1	5	9	0	5	10	0	6
37	11	0	6	11	0	7	8	0	5	8	0	4	9	0	6
38	11	0	6	11	0	7	8	0	4	7	0	3	8	0	5
39	11	0	6	11	0	7	10	0	4	8	0	4	8	0	6
40	10	0	7	12	0	9	11	0	6	9	0	5	10	0	7
GMT	10.8	0.3	6	11.1	0.4	7.3	9.3	0.5	4.9	8.3	0.3	4.3	8.8	0.2	5.9
H5N2=A/chicken/Mexico/232/94 antigen; H5N1=A/chicken/Egypt/1709-6/08 antigen; NDV=Ulster 2C antigen; GMT=geometric mean titre.

All of the vaccinated birds exhibited negative results for the iIFAT anti-N1 antibody assay (DIVA test).

Clinical signs

Unvaccinated chickens were clinically ill after challenge showing signs of HPAI including depression, ruffled feathers, cyanosis of the combs, haemorrhages on legs and shanks, nervous signs (incoordination, seizures, tremors), dyspnoea and diarrhoea starting 48 h post infection. Birds started dying on day 3 post infection, and by day 5 all chickens were found dead.

Among vaccinated chickens, in Group 1 (Volvac® AI) three birds showed slight depression and monolateral conjunctivitis on day 4 post infection. One bird (the chicken identified as Bird 4 in Table 1) recovered in 48 h, while the other two began showing severe depression and nervous signs and then died on days 5 and 7, respectively. In some subjects, temporary unspecific signs were observed (diminished activity and mild periorbital hyperaemia) for a few days after challenge.

In Group 2 (Volvac® AI+ND) only one bird showed clinical signs of HPAI on day 4, and died on day 5 post infection, presenting severe nervous signs such as seizures and incoordination.

In three birds of this group, mild non-specific signs such as lethargy and ruffled feathers were also observed from day 4 to day 8 after infection.

Chickens of Groups 4 and 5 (vaccinated control groups) remained healthy throughout the duration of the study.

Real-time RT-PCR results

All chickens of Group 3 (control group) yielded positive results by real-time RT-PCR (with threshold cycle values ranging from 19.7 to 29.9) on day 3 post infection or on the day of death (day 5).

Among the vaccinated chickens only birds showing clear signs of illness were positive by real-time RT-PCR (with threshold cycle values ranging values from 25.9 to 32.7). All of the real-time RT-PCR positive birds died between days 4 and 7 post infection.

Virological results

All infected chickens of the control group (Group 3) were positive by virus isolation. Among the vaccinated chickens, virus isolation confirmed the results of real-time RT-PCR. In addition, all of the positive samples from the vaccinated chickens were titrated in eggs, displaying a level of viral load ranging from <10^0.5 to 10^2.6 EID₅₀/0.1 ml. The level of viral load in unvaccinated birds ranged from 10^0.5 to 10^3.7 EID₅₀/0.1 ml.

Statistical analysis

Statistical analysis on serological data showed no significant difference (P>0.05) between vaccinated experimental groups prior to challenge.

No statistically significant difference (P>0.05) was observed between pre-challenge and post-challenge (28 days post infection) HI titres in either of the vaccinated and infected groups.

The test for proportional hazard assumption indicated that all of the groups taken into account showed a proportional risk (P>0.05) of developing infection. Cox's regression analysis showed that there is a statistically significant difference in the hazard risk of developing infection between vaccinated groups and unvaccinated groups (hazard ratio<1). The values observed were 0.047 and 0.074 for the Volvac® AI-vaccinated and Volvac® AI+ND-vaccinated groups, respectively. This implies that the administration of either vaccine significantly reduced the risk of developing infection.

Finally, the log-rank test for equality of survival functions showed that there is a statistically significant difference (P<0.05) in the survival rate among vaccinated and unvaccinated chickens, while there is no statistically significant difference (P>0.05) between the two vaccines tested.

The data on the shedding were analysed by two-sample Wilcoxon rank-sum (Mann–Whitney) tests.

A statistically significant difference (P<0.05 or P<0.01) was observed in shedding levels from both tracheal and cloacal swabs between vaccinated birds and unvaccinated birds. In detail, the values were as follows—for tracheal swabs, vaccinated AI versus unvaccinated, P=0.0129 and vaccinated AI+ND versus unvaccinated, P=0.0129; and for cloacal swabs, vaccinated AI versus unvaccinated, P=0.0051 and vaccinated AI+ND versus unvaccinated, P=0.0332.

No statistically significant differences were observed in shedding levels between the two vaccines (Volvac® AI versus Volvac® AI+ND).

No significant differences were observed between shedding levels of tracheal and cloacal swabs (although overall tracheal shedding was higher than cloacal shedding).

Discussion

International organizations recommend that vaccines used for control of AI be of high quality and meet the standards of international health guidelines. The experiments performed in this evaluation comply with the OIE standards (OIE, 2008b) and the minimum requirements indicated by the European Medicine Agency (2006) for vaccines for use in birds against HPAI viruses.

According to the OIE, potency tests may rely on the measurement of challenge and assessment of morbidity (the ratio of diseased to healthy birds in the population) and quantitative reduction in challenge virus replication in respiratory (oropharyngeal or tracheal) and intestinal (cloacal) tracts.

According to the European Medicine Agency, the efficacy of the inactivated vaccine against AI should be demonstrated in laboratory conditions by a challenge model aiming at defining the onset and the duration of immunity for each of the indicated target species. A high degree of protection against mortality and clinical signs of disease and a significant reduction of excretion and transmission of the challenge virus must be the major goals.

In this experiment we established the efficacy of two commercial products in a standardized challenge model.

The number of birds and the composition of the experimental groups were in accordance with previous experiments (Swayne et al., 1997; Capua et al., 2002; Terregino et al., 2007).

The challenge dose used (10⁵ EID₅₀/0.1 ml) and the strain chosen are in line with the OIE standards for vaccine potency tests, which provide that challenged control birds must all die within the sixth day post challenge. All unvaccinated birds were positive by real-time RT-PCR and virus isolation both in the respiratory (tracheal swabs) and intestinal (cloacal swabs) tract. Even though some of the birds were found negative during the first shedding evaluation (day 3 post infection), all of the birds were already showing clear disease signs 2 days after the administration of the challenge virus, therefore the death of all subjects must be related to the direct administration of the challenge virus.

Volvac® AI+ND-vaccinated chickens (Group 2) protected 90% of challenged birds from lethal infection. The only chicken that showed severe clinical signs died within a few days. This bird shed virus from days 3 to 5 post infection (when it was found dead). All the other chickens in the group did not shed detectable RNA or virus in the days chosen for the collection of the swab samples.

In the group vaccinated with Volvac® AI (Group 1), 80% of the challenged birds were protected from lethal infection.

There was no statistically significant difference in the survival rate between the two vaccinated/challenged experimental groups.

No significant increase of serological titres was observed in vaccinated and infected surviving chickens 4 weeks after challenge. This is in accordance with other studies (Capua et al., 2002; Terregino et al., 2007) and could possibly represent a lack of replication of the challenge virus in the surviving vaccinated birds. The negative results from the DIVA test strengthen this hypothesis.

The level of detectable antibodies against H5 was surprisingly high after just one administration of both vaccines using the homologous strain of the vaccine virus in HI tests. This could be due to the high quality of the components of the vaccine batch used and presumably due to the composition of the adjuvant, which has not been divulged by the producers of the vaccine. The lower titres detected towards ND virus can be explained both by the lower concentration of the ND antigen in the bivalent vaccine with respect to AI virus and the use of an heterologous strain (Ulster 2C) to the vaccine virus as HA antigen in the HI assay. In fact, the use of the homologous La Sota antigen in the HI assay would have resulted in significantly higher titres than using the heterologous Ulster virus (Maas et al., 1998). On the other hand, for the AI viruses, the low genetic and antigenic similarity between the vaccine virus and the challenge virus is in keeping with the very low HI titres observed using the challenge virus as antigen. It must also be taken into consideration that the challenge virus emerged after the extended use of field vaccination against H5 in Egypt and for this reason it could have emerged as an escape mutant. This would explain the low levels of cross-reactivity between the challenge virus and sera obtained from vaccinated birds as measured by HI tests.

According to the definition reported above in the evaluation of efficacy section, the only birds that developed infection according to the first set of criteria were the birds that developed clinical disease and died of HPAI.

The data obtained from this study show a significant increase of the survival rate in vaccinated birds, a significant reduction in sick/dead birds, and a significant reduction in the number of birds shedding the challenge virus in vaccinated birds, which results in an overall reduction of shedding levels

Another important element for the OIE and the European Medicine Agency evaluation of the potency of vaccines against AI is the assessment of the immune response. For a good-quality vaccine, the onset of immunity should be as rapid as possible to allow for the use of the vaccine in emergency conditions. In addition, the duration of immunity induced by the vaccine should cover the economic life of the target species.

Volvac® AI and Volvac® AI+ND induced a high titred immune response in chickens vaccinated at 3 weeks of age. A high level of specific antibody against the H5 protein was seen until the end of the study period (70 days of age), as shown in Table 2.

The results comply with previous work in which the ability of genetically distant vaccines against avian influenza in preventing infection, disease, and transmission in chickens and ducks was investigated (Swayne et al., 2006; Van der Goot et al., 2008). In these papers the authors showed that despite the low level of homology between the challenge viruses (Asian HPAI H5N1) and the vaccine strain (Mexican LPAI H5N2) transmission, the mortality and disease rate were markedly reduced after vaccination.

The efficacy of vaccination even in the absence of measurable specific HI antibodies against the HA of the challenge virus could be explained by the generation of antibodies against conserved proteins (i.e. matrix proteins, nucleoprotein and non-structural proteins) that could be cross-protective in birds (Imai et al., 2007; Van der Goot et al., 2008). This may also suggest that cellular immunity plays a more important role in AI vaccination efficacy than originally thought (Khalenkov et al., 2009). Both the former and the latter have been described to play a role in protecting mammalian species such as mice (Tompkins et al., 2007; Zhirnov et al., 2007) and ferrets (Price et al., 2009).

Our findings also suggest that the HI test alone is likely to underestimate the degree of protection, since other factors, which are not measurable with this test, contribute to in vivo protection.

In conclusion, despite the genetic and antigenic differences between the vaccine strain (H5N2/Mexico) and the challenge strain (H5N1/Egypt) used in this experiment, it can be concluded that the immune response induced by a single administration of both Volvac® AI and Volvac® AI+ND vaccines in SPF birds is sufficient to prevent infection in the majority of birds challenged with a contemporary H5N1 Egyptian strain under experimental conditions.

The results also indicate that while the vaccine protects 100% of birds against disease and mortality, some birds are still susceptible to infection and might contribute to virus spread particularly in populations that have low vaccination coverage, which is often the case in countries such as Egypt and Indonesia. Furthermore, inadequate vaccination practices can lead to the selection of variants exhibiting antigenic drift.

It is known that vaccination alone has not been successful in achieving AI eradication (Capua & Marangon, 2004; EU Commission, n.d.). Indeed, it is our opinion that if vaccination is used and not managed appropriately, eradication will not be obtained and the resulting public health threat will not be removed. Experience has shown that, to be successful in controlling and ultimately eradicating AI infection, vaccination must be part of a wider control strategy that includes biosecurity and monitoring of the evolution of infection.

Acknowledgements

The present work was funded by Boehringer Ingelheim Vetmedica. The authors gratefully acknowledge Dr Mona Aly of the Animal Health Research Institute in Egypt for the provision of isolates, and Dr William Dundon and Dr Nadia Micoli for editing the manuscript.