Vaccine 24 (2006) 3184–3190

Effect of emergency FMD vaccine antigen payloadon protection, sub-clinical infection and persistence

following direct contact challenge of cattle

S.J. Cox ∗, C. Voyce, S. Parida, S.M. Reid,P.A. Hamblin, G. Hutchings, D.J. Paton, P.V. Barnett

Institute for Animal Health, Pirbright Laboratory, Ash Road, Woking, Surrey GU24 0NF, UK

Received 18 March 2005; received in revised form 17 January 2006; accepted 18 January 2006Available online 30 January 2006

Abstract

Previous work, in sheep vaccinated with emergency foot-and-mouth disease (FMD) vaccine, indicated the benefit of increasing the antigenptcewd21vwirpftp©

K

1

iIs

0d

ayload in inhibiting local virus replication and consequently persistence following an indirect aerosol challenge with a virus homologous tohe vaccine strain. The work presented here investigates this possibility further using cattle and a more severe semi-heterologous direct contacthallenge. The quantitative dynamics of virus replication and excretion in both vaccinated and non-vaccinated cattle following challenge arexamined. Two experiments were carried out each involving 20 vaccinated and 5 non-vaccinated cattle. An O1 Manisa vaccine (18 PD50)as used for the first, previously reported experiment [Cox SJ, Voyce C, Parida S, Reid SM, Hamblin PA, Paton DJ, et al. Protection againstirect contact challenge following emergency FMD vaccination of cattle and the effect on virus excretion from the oropharynx. Vaccine005;23:1106–13]. The same vaccine was used for the second experiment described in this paper except the antigen payload was increased0-fold per bovine dose, resulting in significantly higher FMD virus neutralising antibody titres prior to challenge. Twenty-one days post-accination the cattle received a 5-day direct contact challenge with FMD virus from five further non-vaccinated cattle infected 24 h earlierith O UKG 34/2001. All vaccinated cattle regardless of antigen payload were protected against clinical disease. Sub-clinical oropharyngeal

nfection was detected in animals from both experiments but the level of virus replication shortly after direct contact challenge was significantlyeduced in vaccinated animals. Cattle immunised with the 10-fold antigen payload cleared the virus more readily and consequently at 28 daysost-challenge fewer animals were persistently infected compared to the single strength vaccine. Following a severe challenge, the resultsrom both experiments show that use of emergency vaccine can prevent or decrease local virus replication and thereby dramatically reducehe amount of virus released into the environment, particularly during the early post-exposure period. Additionally, increasing the antigenayload of the vaccine may reduce sub-clinical infection, leading to fewer persistently infected virus carrier animals.

2006 Elsevier Ltd. All rights reserved.

eywords: Foot-and-mouth disease; Vaccine; Antigen payload; Direct contact challenge; Protection

. Introduction

In Europe, foot-and-mouth disease (FMD) has been erad-cated and prophylactic vaccination is no longer permitted.n order to protect the FMD free status, countries maintaintrict control on importation of live animals and animal prod-

∗ Corresponding author. Tel.: +44 1483 232441; fax: +44 1483 236430.E-mail address: sarah.cox@bbsrc.ac.uk (S.J. Cox).

ucts. However, as the 2001 epidemic showed, such measuresmay be breached resulting in rapid spread of disease throughthe susceptible population. Countries within the EuropeanUnion (EU) are required to maintain contingency plans todeal with incursions of FMD and emergency vaccination isone of the control options that should be available (Directive2003/85/EC). To meet this need, the UK competent author-ity has established a new strategic reserve of FMD antigenswhich could be formulated into vaccine at short notice. The

264-410X/$ – see front matter © 2006 Elsevier Ltd. All rights reserved.oi:10.1016/j.vaccine.2006.01.037

S.J. Cox et al. / Vaccine 24 (2006) 3184–3190 3185

serotypes/strains of vaccine included have been selected toprovide protection against a broad range of FMD viruses cir-culating in the world.

Acceptance of FMD vaccine antigens into the UK strate-gic reserve is presently determined by a cattle potency testas described in the European Pharmacopoeia Monograph [2]and OIE Manual of Standards [3], from which a PD50 valuefor the formulated vaccine is calculated based on protec-tion from clinical disease i.e. prevention of generalisationof FMDV to the feet. For a vaccine to be acceptable as an‘emergency vaccine’ it has to achieve a potency value (PD50)of ≥6. Such vaccines are often referred to as ‘high potency’.In contrast, vaccines prepared for routine, prophylactic cam-paigns need only achieve a minimum PD50 of ≥3. It is welldocumented that exposure of vaccinated ruminants to FMDvirus infection can protect against disease but still permitsub-clinical oropharyngeal infection and that such infectioncan persist. The occurrence of persistently infected “carrier”animals following vaccination creates difficulties for the re-establishment of the status of FMD-free with consequentialimpact on trade restrictions and control policies relating tovaccination [4,5]. However, the current potency test takesno account of how effective the vaccine is at reducing sub-clinical infection.

A number of previous research studies have provided evi-dence that FMD vaccination, particularly emergency FMDvvri[hcgtvfFbetodigaorwqcsis

a

to protect against a semi-heterologous direct contact chal-lenge and the effect of vaccination on virus excretion fromthe oropharynx has been reported previously [1]. Therefore,the opportunity was taken to investigate whether increas-ing the antigen payload of the O1 Manisa vaccine furtherwould effect sub-clinical infection and persistence. This arti-cle reports the follow-up study and compares the resultsgained with those of the previously reported work (Expt. 1)[1].

2. Materials and methods

2.1. Experimental animals, vaccination and challenge

The first experiment (Expt. 1) has been previouslydescribed in detail by Cox et al. [1]. The second experiment(Expt. 2) was similar to the first in all respects except thatthe antigen payload of the O1 Manisa vaccine was increased10-fold per bovine dose. All other components of the vaccineformulation remained the same as used in the first experiment.Briefly, 21 days prior to challenge, 20 cattle were vaccinatedwith this increased antigen payload O1 Manisa vaccine, pre-pared from antigen concentrate, which is being held overliquid nitrogen as part of a new UK strategic reserve A furtherfive animals remained unvaccinated as controls. A total of fivedcowcwdutflv(aa

ssconrM[

2R

ew

accines, formulated to higher potency than conventionalaccines, can have some inhibitory influence on local viruseplication and excretion in the oropharynx, thereby lim-ting transmission of disease to other susceptible animals6,7]. Although most of these studies used vaccines withigher antigen payloads than those that are incorporated intoonventional vaccines, they did not evaluate the effect of anti-en payload directly, being primarily designed to investigatehe speed of onset of protection. It is also known from aariety of studies, including potency tests that compare dif-erent vaccine doses, that increasing the antigen payload ofMD vaccines can enhance the systemic neutralising anti-ody response and improve clinical protection rates. How-ver, since divided dose regimes are often applied in potencyests, it cannot be ruled out that the difference in the amountf adjuvant in each dose, is also having some effect. Evi-ence exists from sheep studies [8], however, that sub-clinicalnfection could be prevented merely by increasing the anti-en payload. This more recent study in sheep examined thebility of three similarly formulated vaccines, which differednly in antigen payload, to decrease or inhibit local viruseplication. The study indicated that higher payload vaccinesere capable of inhibiting local virus replication and conse-uently persistence following homologous indirect aerosolhallenge. A systemic gamma interferon response was alsohown to be related to payload. The work presented herenvestigates this possibility further using cattle and a moreevere semi-heterologous direct contact challenge.

The ability of a commercially prepared emergency oildjuvanted O1 Manisa FMD vaccine with a PD50 of 18

onor cattle were infected with O UKG 34/2001 by needlehallenge and after 24 h were placed in direct contact with thether 25 animals for 5 days. After challenge, the donor cattleere removed and the vaccinated and unvaccinated recipient

attle were housed separately from one another. All animalsere examined regularly for clinical signs of FMD until 28ays post-challenge. Rectal temperatures were recorded dailyntil 10 days post-challenge. Various samples including clot-ed blood, heparinised blood and oesophageal–pharyngealuid (probang) samples were taken at regular intervals pre-accination, 3, 5, 7, 14 days post-vaccination, at challenge21 days post-vaccination) and at 2, 4, 7, 10, 14, 16/17, 21nd 28 days post-challenge and stored as described in Cox etl. [1].

The challenge virus is considered semi-heterologous,ince it is a different strain to the vaccine, but of the sameerotype. O1 Manisa virus and O UKG 34/2001 virus areategorized within the Middle-East South Asia topotypef serotype O FMD virus on the basis of their VP1 geneucleotide identity. The two viruses are also antigenicallyelated (r1 = 1) and a previous study had shown that O1

anisa vaccine could protect cattle against O UKG/20019].

.2. Virus isolation and detection of FMD viral RNA byT-PCR

Heparinised blood samples and probang samples werexamined for the presence of virus by cell culture inoculationith confirmatory ELISA as detailed in Cox et al. [1]. The

3186 S.J. Cox et al. / Vaccine 24 (2006) 3184–3190

probang samples were also tested by quantitative real-timeRT-PCR using automated programmes for total nucleic acidextraction and RT and PCR set up, similar to those describedby Cox et al. [1], in order to determine the viral RNA copynumber per ml of sample.

2.3. Serology

Serum samples collected at weekly intervals up to14 days post-challenge were examined for anti-FMDVneutralising antibodies [10] and for the presence ofantibodies to the FMDV non-structural proteins (NSP)3ABC (Ceditest FMDV-NS (Cedi-Diagnostics)) at 28 dayspost-challenge.

2.4. Detection of interferon gamma (IFN-γ)

Systemic IFN-� was measured in plasma samples from allcattle of both experiments at various time points before andafter challenge, using a double antibody sandwich ELISA, asdetailed by Barnett et al. [8].

2.5. Statistical analyses

Since each experiment involved just one treatment, andanimals used in the experiments were of a similar breed,smaiviwesbaat

cddma

3

3

dofe

ing foot and tongue lesions, which appeared between 2 and10 days, and viraemia from 2 to 7 days after introduction ofthe needle challenged cattle. However, all of the vaccinatedcattle in both experiments, regardless of antigen payload,were protected against clinical disease and none developeda detectable viraemia in samples collected at regular inter-vals up to 28 days after contact exposure with the needlechallenged cattle. Since 20 animals were vaccinated in eachexperiment the true level of protection from clinical diseasefor each treatment (i.e. neat or 10-fold antigen payload) is thesame with a lower one-sided confidence limit of 86% [11].

3.2. Local virus replication, detection of FMD viralRNA and development of antibodies againstnon-structural FMDV polyprotein 3ABC

FMDV was recovered in cell cultures from one or moreprobang samples collected from most vaccinated cattle andall unvaccinated cattle in both experiments, although fre-quency of detection, varied greatly between individual ani-mals (Table 1 and [1]). Likewise, FMDV RNA was variablydetected by quantitative RT-PCR in the probang samplestaken from individual animals of both groups and copy num-bers per ml are also shown in Table 1 and [1]. Samples foundpositive on virus isolation were not necessarily positive byRT-PCR and vice versa. Combining the results of the twovbm2tmtt2waScaEatca

sst2vep

ts

ex and age, and the conditions under which the experi-ents were carried out were kept uniform the data analyses

ssumes that any differences observed between experimentss due to the effect of the treatments and not due to anyariation in experimental conditions across experiments. Thenclusion in both experiments of five unvaccinated controls,hich were treated the same way, further allows for any

xperimental variation to be adjusted for. Additionally, thetatistical analyses of the data from these experiments areased on the assumption that observations made on onenimal are independent of observations made on anothernimal and that animals were randomly allocated to thereatments.

The statistical analyses included determination of 95%onfidence limits for the levels of protection from clinicalisease and viral persistence, ANOVA analyses to investigateifferences in FMDV RNA copy numbers between treat-ents, and logistic regression analysis [11] for virus recovery

t various days post-challenge.

. Results

.1. Development of clinical FMD and viraemia

The needle challenged donor cattle in both experimentseveloped typical FMD foot and tongue lesions within 3 daysf inoculation and were probably a potent source of virusrom day 1 [1]. The five cattle in the unvaccinated groups inach experiment all developed clinical signs of FMD includ-

irological tests, virus or viral genome was confirmed in allut three animals from the vaccinated groups of both experi-ents. In four vaccinated animals in Expt. 1 and five in Expt.oropharyngeal virus or genome was only detected whilst

he donor animals were still present, and this virus/genomeight have originated directly from the donor cattle, rather

han after virus replication in vaccinated recipient cattle. Theotal number of animals with some form of viral recovery at8 days after challenge exposure, i.e. persistently infected,as nine vaccinates and one unvaccinated control in Expt. 1

nd 2 vaccinates and two unvaccinated controls in Expt. 2.amples from one vaccinated animal (UY83) and one unvac-inated animal (UY 97—dead) were unavailable for testingt 28 days post-challenge exposure. Since nine animals inxpt. 1 and two animals in Expt. 2 were persistently infectedt 28 days post-challenge exposure the true level of persis-ence resulting from each treatment can be said, with 95%onfidence, to be included in the intervals from 23% to 68%nd from 1.2% to 32% [11] respectively for each experiment.

NSP serology results at 28 days post-challenge are alsohown in Table 1 and [1] and a correlation is evident betweeneroconversion and extent of virus replication. Thirteen ofhe 20 vaccinated animals in Expt. 1 and 15 of the 20 in Expt.

showed no evidence of antibodies against non-structuraliral protein, whereas all of the unvaccinated cattle in bothxperiments developed a NSP antibody response by 28 daysost-challenge exposure.

In order to assess whether antigen payload was affectinghe amount of sub-clinical infection and persistence (as mea-ured by virus recovery either by virus isolation or RT-PCR),

S.J. Cox et al. / Vaccine 24 (2006) 3184–3190 3187

Table 1Virus isolation (VI) and PCR results from probang samples and non-structural antibody (NSAb) results from vaccinated and unvaccinated cattle (Expt. 2)

Animal reference Days post-challenge

2 4 7 10 14 17 21 28

VI PCR VI PCR VI PCR VI PCR VI PCR VI PCR VI PCR VI PCR NSAb

VaccinatedUY72 − 0 + 0 − 3.49* IS IS − 0 IS IS − 0 − 0 +UY73 − 0 − 0 + 0 − 0 − 0 − 0 − 0 − 0 −UY74 − 0 + 4.07 − 0 − 0 − 0 − 0 − 0 − 0 −UY76 + 4.32 + 4.37 + 0 + 4.57 − 0 + 3.96 + 4.44 + 4.1 +UY77 − 0 + 0 − IS − 0 − 0 − 0 − 0 − 0 −UY78 − 0 − 0 − 0 IS IS − 0 IS IS − 0 − 0 −UY79 − 0 + 4.99 + 5.9 IS IS − 0 − 0 − 0 − 0 +UY80 − 0 + 0 + 4.71 + 3.96 − 0 − 0 + 0 − 0 −UY81 − 0 − 0 + 4.13 − 0 − 0 − 0 − 0 − 0 −UY82 + 3.86 − 0 + 3.58 − 0 − 0 − 0 − 0 − 0 −UY83 IS IS IS IS IS IS IS IS IS IS IS IS − 3.33 IS IS −UY84 + 0 + 3.95 − 0 − 0 − 0 − 0 − 0 − 0 −UY85 − 0 − 0 − 0 − 0 − 0 − 0 − 0 − 0 −UY86 − 0 − 0 + 0 − 0 − 0 − 0 − 0 − 0 −UY87 + 0 − 0 − 0 − 0 − 0 − 0 − 0 − 0 +UY88 IS IS IS IS IS IS IS IS − 0 IS IS − 0 − 0 −UY89 − 0 − 0 − 0 − 0 − 0 − 0 − 0 − 0 −UY90 + 6.14 + 4.56 + 4.61 + 3.36 + 3.99 + 4.74 + 3.77 + 0 +UY91 + 0 − 0 + 5.56 + 3.88 − 0 + 3.6 − 0 − 0 −UY92 + 3.27 + 4.61 + 4.55 − 0 − 0 − 0 − 0 − 0 −

UnvaccinatedUY93 IS IS IS IS − 6.28 IS IS IS IS − 0 IS IS − 0 +UY94 + 5.79 + 7.6 + 5.98 − 4.16 − 0 − 0 − 0 − 0 +UY95 + 5.09 + 6.32 + 6.5 − 4.68 + 4.36 + 4.4 + 5.74 + 5.33 +UY96 + 0 + 4.23 + 5.97 − 5.16 − 0 + 0 + 0 + 4.46 +UY97 + 4.58 + 3.96 + 6.23 − 3.7 − 3.43 − 0 − 0 ND ND +

IS: insufficient sample; *: viral RNA levels [log10(copies ml−1)]; 0: no viral RNA detected; +: virus or non-structural antibodies detected; −: no virus ornon-structural antibodies detected.

a comparison of how often virus was detected by either tech-nique was made. The proportion of animals from which viruswas recovered at each of the days post-challenge exposure isshown in Fig. 1. Ninety-five percent confidence limits asso-ciated with each of the proportions are also shown. On alloccasions other than 7 days post-challenge exposure viruswas less frequently recovered from cattle vaccinated withthe 10-fold antigen payload but the wide confidence bandsreflect the extent of variation between individual cattle withinthe same groups.

Fig. 1. Proportion of animals in each treatment group from which virusrecovered over time post-challenge with associated 95% confidence limits.

Additionally, an ODDSRATIO analysis was also per-formed on the data and the results are presented in Table 2. Onall occasions except at 7 days post-challenge exposure, theodds of virus recovery were higher with the single strengthantigen payload than with the 10-fold antigen payload. Thiswas shown to be significant at 14 and 28 days post-challengeexposure (P < 0.05). Although no significant difference wasidentified for virus recovery between the single strength anti-gen payload and the unvaccinated controls at any time pointpost-challenge exposure, a significant difference (P < 0.05)was shown between the 10-fold antigen payload and unvacci-nated controls at 14 and 21 days post-challenge exposure. The10-fold antigen payload also performed better at 16 and 28days post-challenge exposure although the differences werenot shown to be significant.

The mean quantities of FMD viral RNA [log10(copies ml−1)] in probang samples found positive (non-zero)on quantitative RT-PCR, from both vaccinated and unvacci-nated control cattle at different times after contact with thedonor cattle, are presented in Fig. 2. During the early periodbetween 4 and 10 days post-challenge exposure, the levels oforopharyngeal viral RNA detected in the unvaccinated cattleof both experiments was between 102 and 103 greater thanthat seen in the vaccinated cattle at the same time points.

3188 S.J. Cox et al. / Vaccine 24 (2006) 3184–3190

Table 2ODDSRATIO with 95% confidence limits for virus recovery at various days post-challenge

dpc Comparison Odds ratios Lower limit Upper limit Significance

7 (1×:10×) 0.7778 0.2133 2.8364 0.70347 (1×:control) – – –7 (10×:control) – – –

10 (1×:10×) 2.75 0.6506 11.6242 0.16910 (neat:control) – – –10 (10×:control) – – –

14 (1×:10×) 12 1.3251 108.6741 0.027114 (1×:control) 0.8333 0.1699 4.0876 0.822214 (10×:control) 0.0694 0.0063 0.7693 0.0297

16 (1×:10×) 2.8889 0.6184 13.496 0.177416 (1×:control) 1 0.2124 4.7091 116 (10×:control) 0.3462 0.0582 2.0574 0.2434

21 (1×:10×) 3.2727 0.8023 13.3499 0.098421 (1×:control) 0.4091 0.0792 2.1136 0.286121 (10×:control) 0.125 0.0214 0.7314 0.0211

28 (1×:10×) 6.9545 1.2583 38.4362 0.026228 (1×:control) 1.6364 0.3167 8.4542 0.556728 (10×:control) 0.2353 0.0313 1.768 0.1597

The zero values at 7 and 10 days post-challenge (dpc) for the odds ratios arise because virus recovery is 100% for the control group in these cases. Theproportions of animals showing virus recovery with the single strength and 10-fold antigen payloads are respectively 0.55 and 0.61 at 7 dpc and 0.50 and 0.27at 10 dpc.

Since both experiments used animals of a similar breed, sexand age, and the conditions under which the experimentswere carried out were kept uniform, the unvaccinated cat-tle data from both experiments were pooled together for thepurpose of data analyses. The differences in quantity of viralRNA detected between the vaccinated and unvaccinated cat-tle was analysed using ANOVA during this time period and at7 days post-challenge exposure, significantly less (P < 0.05)viral RNA was detected from the vaccinated animals in bothexperiments. However, these differences were not maintainedat the later time period. The viral load in the unvaccinated cat-tle peaked at 7 days, and then fell, even in persistently infectedanimals. In contrast, some persistently infected vaccinatedanimals, particularly in Expt. 1 had their highest levels ofdetectable viral RNA at 28 days after the introduction of theneedle challenged cattle. ANOVA analyses of the viral RNA

Fp

copy number data at these later time points post-challengeshowed significantly less (P < 0.05) viral RNA was recov-ered from both the unvaccinated and 10-fold antigen payloadvaccinated cattle when compared with the vaccinated cattleof Expt. 1 at 21 days post-challenge exposure. However, nosignificant difference was identified between groups at 14, 16or 28 days post-challenge exposure.

3.3. Virus neutralising antibody induction

Fig. 3 shows mean serum neutralising antibody responsesagainst O1 Manisa in vaccinated and unvaccinated cattle,up to 14 days post-challenge exposure, following vacci-nation and direct contact with infected donor cattle. At 7days post-vaccination, 9 animals in Expt. 1 and 20 in Expt.

Fs(i

ig. 2. Mean FMDV RNA copy number detected over time by RT-PCR fromositive samples.

ig. 3. Mean neutralising antibody responses in cattle vaccinated with eitheringle strength vaccine (1×) or vaccine containing 10-fold antigen payload10×). Control unvaccinated cattle for comparison after challenge are alsoncluded.

S.J. Cox et al. / Vaccine 24 (2006) 3184–3190 3189

2 had seroconverted. Neutralising antibody titres, at everypost-vaccination time point, were significantly higher for the10-fold antigen payload group (P < 0.05).

3.4. IFN-γ responses

A systemic IFN-� response related to vaccination or chal-lenge was not clearly identified in either the unvaccinatedor vaccinated cattle of both experiments although low lev-els, close to the detection limit of the assay, were identifiedin some animals, particularly vaccinated animals (results notshown).

4. Discussion

A second cattle experiment has been completed in whichthe only difference to that previously reported work [1] wasthat the antigen payload was increased 10-fold. Based on pre-vious observations our hypothesis was that this would reducethe level of sub-clinical infection and lead to a reduced rateof virus persistence. In the event, the higher payload vaccinestimulated a stronger systemic neutralising antibody responseand reduced the amount of oropharyngeal virus replicationand persistence, but the reductions observed were not alwaysstatistically significant.

pqsccpiag

iaitFytdeirmabttp

b

not always significantly so. Therefore, despite the relativelylarge number of animals used per experiment, the benefitof using higher antigen payload has not been definitivelyproved. Owing to the severity and length of the challenge,there is a possibility that the virus recovered in the first5–7 days following the start of the challenge exposurecould have been directly inhaled or ingested rather thanbeing a true indication of oropharyngeal replication. Thiscomplicates analyses and may have masked differencesbetween virus replication in the vaccinated cattle of the twoexperiments. An independent measure of virus replication isseroconversion in the non-structural antibody test, and fivecattle showed such a response Expt. 2, compared to sevenin Expt. 1.

An unexpected finding in the first experiment using thelower antigen payload, was that the levels of viral RNA inprobang samples collected from vaccinated cattle after 10days post-challenge were often higher than those in equiv-alent samples from unvaccinated animals (Fig. 2). It wasconsidered that this might have been due to chance or couldreflect a genuine difference in the local immune responsesof vaccinated and unvaccinated cattle. In the second exper-iment, with the higher antigen payload, no difference wasseen in the titres of viral RNA recovered after 10 days post-challenge from the oropharyngeal regions of vaccinated andunvaccinated cattle.

bagaatsrgvelprtcsavlmeaastaoc

As expected, the cattle vaccinated with the 10-fold antigenayload were all protected from clinical disease. Conse-uently, we can report with a high level of confidence thatuch vaccines, after a single application, provide good clini-al protection against severe semi-heterologous direct contacthallenge. Further studies, however, utilizing shorter timeeriods between vaccination and challenge, are needed todentify how quickly such protection is achieved and whethert such time points, the effect on protection of different anti-en payloads may be more pronounced.

As has been previously documented [4,5], our first exper-ment confirmed that protection from clinical disease did notlways coincide with prevention of localised, sub-clinicalnfection and that this infection can persist and give riseo carriers. In the first 10 days after challenge exposure,MDV was recovered to a variable extent from the orophar-nx of vaccinated cattle receiving both antigen payloads, buto a lesser extent than from unvaccinated cattle. This furtheremonstrates the ability of high potency FMD vaccines toither prevent or reduce viral replication at the site of primarynfection, thereby reducing the amount of infectious materialeleased to the environment from sub-clinically infected ani-als. However, since transmission from both the vaccinated

nd unvaccinated animals was not measured directly, theiological significance of this reduction remains unsubstan-iated. Clearly future studies need to be designed to measureransmission directly and are the natural progression from theresent work.

The proportion of vaccinated cattle shedding virus andecoming carriers was lower in the second experiment, but

The mean systemic neutralising antibody response foroth groups of vaccinated cattle was examined to see whetherny differences were evident as a result of the different anti-en payloads. Cattle that had received the 10-fold higherntigen payload vaccine produced antibodies more quicklynd had significantly higher titres at every post-vaccinationime point compared to the responses of the cattle given theingle strength vaccine. This suggests that systemic immuneesponses were significantly enhanced by increasing the anti-en payload. This might have improved protection againstirus generalisation following local replication, but such anffect was not observable in these experiments where even theower dose of vaccine abrogated viraemia and provided com-lete clinical protection. The fact that this enhanced immuneesponse did not prevent virus replication is at variance withhe studies of sheep, in which high antigen payload vac-ines completely inhibited virus replication. In the sheeptudy, a less severe challenge regime was used, involvingshort period of indirect contact with aerosols from FMD

irus infected pigs and further experiments could examineess severe challenge in cattle rather than the ‘worst case’

odel used here. A correlation was observed in the sheepxperiments between levels of circulating IFN-� productionnd prevention of oropharyngeal viral replication, but similarnalyses in these cattle experiments failed to substantiate aimilar correlation. A whole blood stimulation assay in whichhe ability of cells to produce IFN-� in response to vaccinentigen, is presently being assessed as an alternative meansf determining whether IFN-� is involved in reducing sub-linical infection in vaccinated cattle as a result of increased

3190 S.J. Cox et al. / Vaccine 24 (2006) 3184–3190

antigen payload [12]. It would also be desirable to comparethe oropharyngeal immune responses of the cattle given thelow and high antigen payload vaccines.

Although this study provided no conclusive evidence thatthe higher antigen payload vaccine enhanced clinical or viro-logical protection, it might contribute to a broader protectionagainst more antigenically diverse field strains and virus neu-tralisation studies utilising pooled 21 days post-vaccinationsera from both experiments and FMDV field isolates areunderway.

Previous work investigating the importance of antigenpayload [13,14] mainly focussed on the systemic immuneresponse following increases in vaccine potency, whilstpotency tests carried out according to the European Pharma-copoeia have examined the affect of vaccine dose on clinicalprotection. This study looked at sub-clinical infection in rela-tion to antigen payload but did not confirm a relationshipbetween increased antigen payload and prevention of oropha-ryngeal virus replication, although higher antigen payloadappeared to aid virus clearance and reduce persistence. Fur-ther studies involving a less severe form of challenge mightbe needed to demonstrate a more convincing dose response.Although much effort continues to be directed at the develop-ment of novel FMD vaccines, it is still likely to be some timebefore inactivated whole viral antigens are superceded andtherefore detailed studies of the efficacy and optimal dosagefc

A

Ssa

References

[1] Cox SJ, Voyce C, Parida S, Reid SM, Hamblin PA, Paton DJ, etal. Protection against direct contact challenge following emergencyFMD vaccination of cattle and the effect on virus excretion from theoropharynx. Vaccine 2005;23:1106–13.

[2] Anon. European Pharmacopoeia. Foot-and-Mouth Disease (Rumi-nants). Vaccine 1997;63:875–6.

[3] Anon. Section 2.1 List A diseases, Chapter 2.1.1: Foot-and-mouthdisease, Article 2.1.1.7 OIE terrestrial animal health code. 12th ed.;2003.

[4] Alexandersen S, Zhang Z, Donaldson AI. Aspects of the persistenceof foot-and-mouth disease virus in animals—the carrier problem.Microbes Infect 2002;4:1099–110.

[5] Sutmoller P, Barteling S, Olascoaga R, Sumption KJ. Control anderadication of foot-and-mouth disease. Virus Res 2003;91(1):101–44.

[6] Barnett PV, Carabin H. A review of emergency foot-and-mouth dis-ease (FMD) vaccines. Vaccine 2002;20:1505–14.

[7] Donaldson AI, Kitching RP. Transmission of foot-and-mouth dis-ease by vaccinated cattle following natural challenge. Res Vet Sci1989;46:9–14.

[8] Barnett PV, Keel P, Reid S, Armstrong RM, Statham RJ, Voyce C,et al. Evidence that high potency foot-and-mouth disease vaccineinhibits local virus replication and prevents the ‘carrier’ state insheep. Vaccine 2004;22(9–10):1221–32.

[9] Aggarwal N, Zhang Z, Cox S, Statham R, Alexandersen S, KitchingRP, et al. Experimental studies with foot-and-mouth disease virus,strain O, responsible for the 2001 epidemic in the United Kingdom.Vaccine 2002;20:2508–15.

[10] OIE. Manual of standards for diagnostic tests and vaccines. 4th ed.

[

[

[

[

or presently available vaccines are essential in support ofontingency planning for their emergency use.

cknowledgements

This work was supported financially by Defra, UK (ProjectE2808). We thank Malcolm Turner, Barry Collins and theirtaff for assistance with the handling and care of experimentalnimals and Savitri Abeyasekera for statistical analyses.

Paris, France: Office International des Epizooties; 2004.11] Collett D. Modelling binary data. 2nd ed. London, UK: Chapman &

Hall; 2002.12] Parida S, Oh Y, Reid SM, Cox SJ, Statham RJ, Mahapatra M, et al.

Interferon-� production in vitro from whole blood of foot-and-mouthdisease virus (FMDV) vaccinated and infected cattle after incubationwith inactivated FMDV. Vaccine, in press.

13] Rweyemamu MM, Black L, Boge A, Thorne AC, Terry G. Therelationship between the 140S antigen dose in aqueous foot-and-mouth disease vaccines and the serum antibody response of cattle.J Biol Stand 1984;12:111–20.

14] Doel TR. Optimisation of the immune response to foot-and-mouthdisease vaccines. Vaccine 1999;17:1767–71.