PAPERS & ARTICLESU

Factors affecting the serological response

of dogs and cats to rabies vaccination
K. L. MANSFIELD,, R D. BURR, D. R. SNODGRASS, R. SAYERS, A. R. FOOKS
After being vaccinated against rabies some cats and dogs fail to show an antibody titre adequate to meet
the requirements of the UK Pet Travel Scheme. To investigate this problem, the data derived from 16,073
serum samples submitted to the Veterinary Laboratories Agency for serological testing between 1999 and
2002, 1002 samples submitted to BioBest during March and April 2001, and 1264 samples associated with
one make of vaccine submitted to BioBest between June 2001 and January 2003, were analysed. The
probability of antibody titre failing to reach at least 0-5 iu/ml was analysed by logistic regression as a
function of the choice of vaccine, the interval between vaccination and sampling, the sex and age of the
animal, and its country of origin. In dogs, all these factors, except sex, had highly significant (P<0-001)
effects on the test failure rate, and in cats all the factors had a significant effect (P<0-05).
THE UK Pet Travel Scheme (PETS) was introduced in February of the last rabies vaccination, the make of vaccine and its batch
2000 as an alternative to quarantine for dogs and cats enter- number, and the microchip number of the animal. On July 3,
ing the UK from rabies-free countries. A principal component 2004, EU Regulation 998/2003 will take effect. This regulation
of the scheme is the compulsory serological testing of all lays down the animal health requirements for the non-com-
rabies-vaccinated dogs and cats by the fluorescent antibody mercial movements of pet animals both withing the European
virus neutralisation (FAVN) test approved by the Office Community and into the Community from third countries.
International des Epizooties (OIE 2000). In the UK, the sero- However, the rules of PETS will continue largely unchanged for
logical testing is carried out by two laboratories recognised by at least five years after the Regulation takes effect.
the Department for Environment, Food and Rural Affairs Data submitted to the VLA and BioBest for FAVN testing
(DEFRA), the Veterinary Laboratories Agency (VLA) - have been used to investigate the potential factors that may
Weybridge, a DEFRA agency, and BioBest, a private company. influence the response of dogs and cats to rabies vaccination.
The UK has remained free from endemic rabies since 1922, The data were analysed statistically to assess any significant
principally owing to its island status and the enforcement of factors associated with the antibody titre failing to reach at
strict quarantine laws (Fooks and others 2002). Although least 0.5 iu/ml or more. This paper describes the results of the
quarantine for six months is a highly effective system, ani- analysis.
mal welfare groups have suggested that it may have adverse
effects on an animal's health and welfare (Rochlitz and oth-
ers 1998), and the UK Government therefore considered alter- MATERIALS AND METHODS
natives to quarantine. Following the recommendations of a
committee chaired by Professor Ian Kennedy (Kennedy 1998), Source of samples
the Government introduced PETS on February 28, 2000. Data supplied with serum samples from 14,035 dogs and 2038
Provided that a cat or dog met defined criteria, it could be cats which were submitted to the VLA for FAVN testing between
imported into the UK from certain countries and rabies-free 1999 and early 2002 were used; the dogs and cats were treated
islands without quarantine. Under the statutory conditions as separate populations. Of these 16,383 samples, 16,073 were
of PETS, an animal must be vaccinated against rabies with a used for data analysis and computer modelling; samples that
licensed vaccine, and identifiable either by a microchip or failed to meet the criteria of the scheme, for example, samples
individual tattoo. with a low volume, and samples for which the interval
In the UK, three whole, inactivated vaccines are licensed for between vaccination and blood sampling was six days or less,
use. A blood sample is taken from the animal for serological were omitted.
analysis at a specified period after vaccination, the period Data supplied with 1002 sequential samples submitted to
being that considered the optimum for a strong antibody BioBest between March and April 2001 were also analysed.
response. The serological test used at the VLA and BioBest is Because only 24 samples were associated with one make of
the FAVN assay (Cliquet and others 1998), which detects rabies vaccine, the mean FAVN titre for this vaccine was calculated
Veterinary Record (2004) virus-specific neutralising antibodies as an indirect method of from samples taken from 183 cats and 1081 dogs between
154, 423-426 confirming the status of rabies vaccination in the animal. June 2001 and January 2003 (Table 1, vaccine 3).
Following OIE recommendations, the Kennedy report
K. L. Mansfield, BSc, (Kennedy 1998) advised that an animal must have a serum FAVN
R. Sayers, BSc, rabies virus-specific neutralisation titre of at least 0-5 iu/ml. The FAVN test described by Cliquet and others (1998) was
A. R. Fooks, MBA, PhD, It has been observed that a small proportion of animals have used; the test measures only neutralising antibodies and total
Rabies Research and failed to reach this threshold antibody level, and have had to antibody levels were therefore not measured.
Diagnostic Group, be re-vaccinated (Burr and Snodgrass 2002, Fooks and oth-
Veterinary Laboratories ers 2002) and re-tested to show that they have an acceptable Statistics
Agency - Weybridge, antibody titre. In order to maintain an effective level of immu- At the VLA, the probability of the antibody titre 'failing to
Woodham Lane, New nity that enables a sample to pass the FAVN test, the vaccine reach at least 0 5 iu/ml' as an outcome of vaccination was
Haw, Surrey KT15 3NB manufacturers recommend that regular booster vaccinations analysed by the separate logistic regression of the results for
P. D. Burr, PhD, MRCVS, should be administered; there is no requirement for further the dogs and cats as a function of the choice of vaccine, the
D. Snodgrass, PhD, blood testing if these boosters are given at the correct time. breed, sex, age and country of origin of the animal, and the
MRCVS, A further requirement for the scheme is that any sample interval between vaccination and blood sampling. The statis-
BioBest, Pentlands submitted to either the VLA or BioBest for rabies testing must tical package STATA (Stata Statistical Software: Release 7.0;
Science Park, Penicuik, be accompanied by a sample submission document giving StataCorp) was used for all the analyses. Likelihood ratio tests
Midlothian EH26 OPZ information including the date of blood sampling, the date were used to assess the statistical significance of the data.

The Veterinary Record, April 3, 2004 423

PAPERS & ARTICLES

The animals' ages were treated as whole years up to 12, but

older animals were grouped together. Countries with fewer
than 50 records for cats, and fewer than 100 records for dogs,
were grouped into a single'other' category.
At BioBest, the data was entered into a Microsoft Access One Ntimber of T Nuber Of One and Number Of
database, and analysed in Microsoft Excel for each vaccine Spedes Vaccine dose an l does anim two an s
and species by a two-way analysis of variance, to test for dif-
ferences between the vaccines and the number of doses. Dogs I 1 110(0) 292 131 ( 21 1122 313
2 6-1 ( 456 106 65 66(3 521
Individual mean values were then compared by t tests. 3 3-1 898 4-9 183 34(3.3) 1081
cats 1 133 4 55 1" ) 4 13-3 (4 59
Interpretation of results 2 12 (50) 79 l025 6 102(50) 85
The results were summarised in terms ofodds ratios and their 3 4.9(3.6 147 6-5 36 5-2 (3.6) 183
confidence limits. Except for'time interval between vaccina-
tion and blood sampling') the first category of each risk factor
acted as a reference level and was therefore given the nominal Country of origin
value of 1. The odds ratios for the other categories estimated At the VLA, there was a considerable variation in the risk of
the extent to which their odds of a 'test fail' as opposed to a test failure associated with the country of origin of the sam-
'test pass' were higher or lower than for the reference category. ple (dogs P<0-001; cats P=0.029). Dogs from Australia and
For all analyses P<0-05 indicated statistical significance. New Zealand had a higher chance of failure than those from
the UK. In contrast, the highest risk in cats was from samples
from Gibralter.
RESULTS
For dogs, the logistic regression analysis indicated that there DISCUSSION
were highly significant (P<0.001) effects on the test failure
rate for all the factors analysed except for the sex of the ani- After they have been vaccinated against rabies, dogs and cats
mal. For cats, the analysis indicated that there were significant are protected primarily through the generation of rabies
(P<0-05) effects on the test failure rate for all the factors. virus-specific neutralising antibodies. Cell-mediated immun-
At the VLA, 4-12 per cent of the dogs and 2-85 per cent of ity also plays a role, especially in the longevity of the immune
the cats failed the FAVN, and similar figures were recorded at response. In the UK, three inactivated vaccines are licensed for
BioBest (Table 2). the vaccination of dogs and cats that participate in PETS, and
they are all highly effective in eliciting an immune response,
Choice of vaccine and number of doses and meet the requirements for PETS. However, a single dose
At BioBest all three vaccines elicited mean antibody titres does not provide lifelong immunity, and booster doses need
greater than the threshold of 0-5 iu/ml (Table 1). For dogs to be administered at specific intervals. Specific studies ofvac-
there were significant differences between the mean titres cine failures have been reported by Cliquet and others (2000),
induced by the three vaccines (P<0-001), and there was also and it is accepted that a small proportion of animals will fail
an increase in the mean titre after two doses of vaccine to respond adequately to vaccination, the principal reason
(P<0-001). For cats, there were significant differences between being immunosuppression induced by the animal's age and
the mean titres induced by the three vaccines (P<0-00 1), but reproductive status.
there was no significant increase in the titre after two doses of The results of this study show that another statistically sig-
any of the vaccines. nificant factor is the interval between vaccination and blood
sampling. The three rabies vaccines all underwent rigorous
Interval between vaccination and blood sampling potency testing before they were licensed, and the manufac-
At the VLA the interval between vaccination and blood turers test the potency of each batch of vaccine. However,
sampling had a significant effect on the risk of an animal rather than questioning the potency of the vaccines, this study
failing the test (Fig 1). In the dogs, the risk of test failure was has investigated the ability of a vaccinated animal to pass the
least when the blood sample was taken four weeks after
vaccination, and there was a marked increase in risk after six
weeks (P<0-00 1). In the cats, the risk of a test failure was least
when the sample was taken four or five weeks after vaccina-
tion. l.;1 .0. 3-& ;;; |^ .6

At BioBest, similar results were obtained, with a significant

(P<0-01) decrease in neutralising antibody titre in animals Spedes VLA
sampled eight to nine weeks after vaccination. Dogs 578/14 5 (4.12) 44/852 (.16)
Cats 58/238 (2-85) 4/150(26)
Sex
The risk of test failure was significantly affected by the sex of
the animals for cats (P=0.028) but not for dogs (P=0-124).
Entire animals had a greater risk of failing the test (Table 3),
and the effect was particularly marked in entire male cats.
Age Risk of failure (95 per cent confidene i l)
As they became older there was a steady and significant Sex Dogs Cats
(P<0.001) increase in the risk ofboth dogs and cats failing the
test (Fig 2); animals less than a year old also had a slightly Neutered female 1 1
higher risk of failure than animals one year old. At BioBest, Uncastrated male 1-31 (0-98-1.73) 3-24 (1-43-7.35)
Castrated male 0.99 (0-69-1.43) 1.03 (0.44-2-41)
dogs less than six months old had significantly (P-0-03) lower Unneutered female 1-29 (0-97.172) 1-23 (0.48-3.13)
antibody titres, and cats more than 14 years old also had a P 0.124 0-023
significantly (P=0-04) lower titre.

424 The Veterinary Record, April 3, 2004

FIG 1: Relationship
between the interval
between rabies
vaccination and blood
sampling in weeks and
the risk of failing the
fluorescent antibody
virus neutralisation test
(determined by logistic
regression analysis) in
dogs and cats,
expressed in terms of
the odds ratios
-
-

.2
0o
a)

-
0

if2-
12 -

10-

0
8-

6-
4-
2-

1 2
-
-

3
Dogs
Cats

4 5 6
Interval (weeks)
7
F F1
8 8+

FAVN test. An animal's antibody response after iit has been vac-
cinated will peak at different times with diffe rent brands of
vaccine, and the timing of the postvaccinatic in blood test is
therefore a critical factor. Differences betwee n the vaccines
may also have been influenced by different mar]keting and data
sheet recommendations as to the required nuLmber of doses
PAPERS & ARTICLES

titres persisted for longer if the vaccine was administered

intramuscularly rather than subcutaneously. Ganiere and oth-
ers (1989) reported that dogs and cats remained immune two
years after they had been vaccinated. This apparent contra-
diction between protective efficacy and the levels of neutral-
ising antibodies suggests that although there is a rapid decline
in the titres of serum neutralising antibodies, as measured by
in vitro assays, an anamnestic, protective response against
intramuscular challenge with rabies virus persists.
The country of origin was also a significant factor in deter-
mining the number of samples that failed to reach an anti-
body titre of 0-5 iu/ml. Dogs from Australia and New Zealand,
and cats from the UK and Gibraltar, were more likely to fail the
FAVN test, but the reasons are unclear.
The haplotype of specific breeds of dog is a factor which
affects a dog's immune response to vaccination (Kennedy and
others 1999), and there is evidence that a dog's breed may
have a significant effect on its immune response to rabies vac-
cination (VLA, unpublished observations).
The reproductive status of the animal also has an effect on
its response to vaccination. The present data suggest that
unneutered animals have an increased risk of responding
(Mackay and others 2003). The present result,s show that the poorly to vaccination, and the effect was quite marked in
number of vaccine doses a dog had received alffected its anti- uncastrated male cats. It is known that gonadal steroid hor-
body titre, although no similar effect was observed in cats. The mones play an important role in modulating the immune
sample submission forms are required to conttain the date of response (Schuurs and Verheul 1990), and Rife and others
only the most recent vaccination, and reportiing the dates of (1990) showed that they can either stimulate or suppress it,
any earlier vaccinations is optional; as a rresult, animals depending upon the antigen involved. Rife and others (1990)
reported to have received one dose may in fac:t have received demonstrated that castrated mice responded better to anti-
more. This deficiency in the data would tend to obscure any genic stimulation than uncastrated ones, and had twice the
beneficial effect of repeat vaccinations. The fiact that signifi- number of T lymphocytes, and they suggested that testo-
cant increases in antibody titre were recorded ir dogs after sec- sterone may affect the immune system through the enhance-
ond vaccinations is therefore even more relevanit. If it had been ment of suppressive activity in testosterone-treated animals.
compulsory for two doses of each of the vaccin, es to be admin- Verthelyi and Klinman (2000) suggested that sex hormones
istered, the results observed might have been X rery different. may modulate the production of cytokines, and thus induce
The interval between vaccination and blo od sampling is sex-related differences in the immune response.
probably one of the most significant factors iin determining The results also showed that the ability of dogs and cats
the host's apparent response. Rabies vaccinat ion produces a to develop an adequate antibody titre after being vaccinated
typical antibody response curve, with the resi ponse decaying against rabies declined steadily with age, presumably as the
over time. Previous studies have shown that; antibody levels immune system becomes less efficient. Animals must be at
peak at about four weeks after vaccination (C lliquet and oth- least three months old when they are first vaccinated, and it
ers 2000), although the period may be slightl)y different with has been shown that animals less than one year old have a
different vaccines. After four weeks, there is a steady increase slightly increased risk of having a poor FAVN antibody titre.
in the risk of test failure, as the levels of circu] lating antibody Aghomo and others (1990) showed that young dogs can pro-
start to wane. However, with some batches of vaccine, signif- duce rabies antibodies from four weeks of age, when their titre
icantly higher FAVN titres have been reported t two weeks after of maternal antibody has waned. If the dam has been vacci-
vaccination, and as a result vaccine manufac turers are con- nated, the pups do not begin to respond to vaccination until
sidering amending data sheet recommendatic :ns on the tim- they are 10 weeks old, and they are less efficient at produc-
ing of blood sampling. ing immunoglobulins. The mechanisms for this short-lived
Coyne and others (2001) reviewed earlierr studies which suppression of the immune system are not well understood,
showed that vaccination induced long-term] protection. For but suggestions include the activation of putative suppressor
example, dogs vaccinated with a modified livre virus vaccine T cells by the maternal antibodies, and increased corticos-
were protected against virulent rabies viruis for 38 to 41 teroid levels (Banks and McGuire 1989). The interference by
months, and they were protected for 22 to 3 9 months after maternal antibodies may explain the slightly increased risk of
being vaccinated with inactivated vaccines; the protective vaccine failure in such young animals. The popularity of tak-
ing pets in and out of the UK, and the fact that a high pro-
portion of puppies bred recently are likely to be from
140 - Dogs vaccinated bitches, make these factors particularly relevant.
Cats The two most important factors in achieving a successful
FIG 2: Relationship .o 120 FAVN test result appear to be the number of doses of vaccine
between the age in co 100 and the interval between vaccination and blood testing. To
years of dogs and cats *0 guarantee that inadequately immunised animals do not par-
and their risk of failing a °2 80 ticipate in PETS, periodic blood tests and, if necessary, a
the fluorescent -' 60 booster vaccination should be considered to maintain a sat-
antibody virus isfactory rabies antibody titre. However, it is known that the
neutralisation test -0 40
(determined by logistic ('eIcn cellular immune response plays an important role in protec-
regression analysis) c: 20 tive immunity, and animals may still have a potent cellular
after rabies vaccination, 0 -F---F---F---1 response despite having a low titre of neutralising antibod-
expressed in terms of 0 2 4 6 8 10 12 ies (Zanetti and others 1998). Those workers demonstrated
the odds ratios Age (years) that certain dilutions of vaccine can induce high levels of

The Veterinary Record, April 3, 2004 425

PAPERS & ARTICLES

gamma-interferon but low titres of neutralising antibodies. COYNE, M. J., BURR, J. H. H., YULE, T. D., HARDING, M. J., TRESNAN,
Further work should include the assaying of cellular immune D. B. & MCGAVIN, D. (2001) Duration of immunity in dogs after vaccina-
response in dogs and cats that do not develop levels of anti- tion or naturally acquired infection. Veterinary Record 149, 509-515
FOOKS, A. R., MCELHINNEY, L. M., BROOKES, S. M., JOHNSON, N.,
bodies detectable by FAVN. KEENE, V., PARSONS, G. & SOLDAN, A. (2002) Rabies antibody testing and
the UK Pet Travel Scheme. Veterinary Record 150,428-430
GANIERE, J. P., ANDRE-FONTAINE, G., BLANCOU, J., ARTOIS, M. &
ACKNOWLEDGEMENTS AUBERT, A. (1989) Vaccination antirabique du chien et du chat: taux
d'anticorps et resistance a l'epreuve virulente deux ans apres l'injection de
The authors wish to thank members of the Rabies Group, VLA, rappel d'un vaccin additionne d'adjuvant. Revue de Medecine Veterinaire 140,
for technical support and advice, and also Sara Robertson (IT 281 -285
Department, VLA) for assistance in management of the rabies KENNEDY, I. (1998) Quarantine and rabies: a reappraisal. Report by the
database. This work was supported by DEFRA, UK. The VLA and Advisory Group on Quarantine to the Right Hon Nick Brown MP, Minister
BioBest acknowledge the support given by all their customers of Agriculture, Fisheries and Food. London, MAFF Publications
KENNEDY, L. J., CARTER, S. D., BARNES, A., BELL, S., BENNETT, D.,
in submitting samples for analysis. OLLIER, B. & THOMSON, W. (1999) Interbreed variation of DLA-DRBI,
DQAI alleles and haplotypes in the dog. Veterinary Immunology and
Immunopathology 69, 101- 111
References MACKAY, D., ILOTT, 1, & PAGE, K. (2003) PETS and postmarketing surveil-
AGHOMO, H. O., ODUYE, 0. 0. & RUPPRECHT, C. E. (1990) The serologi- lance of veterinary rabies vaccines authorised in the UK. Veterinary Record
cal response of young dogs to the Flury LEP strain of rabies virus vaccination. 153, 155
Veterinary Research Communications 14, 415-425 OIE (2000) Rabies. In Manual of Standards for Diagnostic Tests and Vaccines.
BANKS, K. L. & MCGUIRE, T. C. (1989) Neonatal immunology. In Veterinary 4th edn. Paris, Office International des Epizooties. pp 276-291
Clinical Immunology. Eds R. E. W. Halliwell and N. T. Gorman. Philadelphia, RIFE, S. U., MARQUEZ, M. G., ESCALANTE, A. & VELICH, T. (1990) The
W. B. Saunders. pp 193-204 effect of testosterone on the immune response 1. Mechanism of action on
BURR, P. D. & SNODGRASS, D. R. (2002) Serological response to rabies vac- antibody-forming cells. Immunological Investigatiotns 19, 259-270
cination for the pet travel scheme. Research in Veterinary Science 72 (Suppl ROCHLITZ, I., PODBERSCEK, A. L. & BROOM, D. M. (1998) Effects of quar-
A), 25 antine on cats and their owners. Veterinary Record 143, 181-185
CLIQUET, F., AUBERT, M. & SAGNE, L. (1998) Development of a fluorescent SCHUURS, A. H. W. M. & VERHEUL, H. A. M. (1990) Effects of gender and sex
antibody virus neutralisation test (FAVN) for the quantitation of rabies- steroids on the immune response. Journal of Steroid Biochemistry 35, 157-172
neutralising antibody. Journal of Immunological Methods 212, 79-87 VERTHELYI, D. & KLINMAN, D. M. (2000) Sex hormones correlate with the
CLIQUET, F., VERDIER, Y., SAGNE, J. L., SCHEREFFER, M., SELVE, M. & activity of cytokine-secreting cells in vivo. Immunology 100, 384-390
AUBERT, M. (2000) Alternative a la quarantine dans les pays ou territoires ZANETTI, C. R., DE FRANCO, M. T., VASSAO, R. C., PEREIRA, C. A. &
indemnes de rage: analyse des taux d'anticorps antirabiques chez les chiens PEREIRA, 0. A. C. (1998) Failure of protection induced by a Brazilian
et chats vaccines. Bulletini Mensuel de la Societe Veterinaire Pratique de France vaccine against Brazilian wild rabies viruses. Archives of Virology 143, 1745-
84, 29-39 1756

Radiographic findings in 16 dogs infected

with Angiostrongylus vasorum
A. K. BOAG, C. R. LAMB, P. S. CHAPMAN, A. BOSWOOD
Thoracic radiographs of 16 dogs infected naturally with Angiostrongylus vasorum showed signs of bronchial Veterinary Record (2004)
thickening, an interstitial pattern and a multifocal and/or peripheral alveolar pattern. In dogs treated with 154,426-430
fenbendazole, follow-up radiographs showed that the alveolar pattern had resolved and a mild, hazy
interstitial pattern had developed. In contrast with dogs with heartworm (Dirofilaria immitis), no pulmonary A. K. Boag, MA, VetMB,
vascular lesions were identified. DipACVIM, MRCVS,
C. R. Lamb, MA, VetMB,
Angiostrongylus vasorum is a metastrongylid nematode that Angiostrongylosis is endemic in dogs in the Republic of DipACVR, DipECVDI,
infects domestic dogs and related canids. It has an indirect life Ireland (Dodd 1973) and is generally considered to be a spo- ILTM, MRCVS,
cycle with gastropods (slugs and snails) acting as the inter- radic condition of dogs in the UK, although there is evidence P. S. Chapman, BVetMed,
mediate host (Rosen and others 1970, Bolt and others 1994). that its incidence in south-east England is increasing (Cobb and MRCVS,
Dogs become infected as a result of ingesting an intermediate Fisher 1990, Elwood 2001, Kriek 2001, Phillips 2001, Chapman A. Boswood, MA, VetMB,
host containing the L3 larvae. Once liberated in the small and others 2002). The clinical and radiographic findings in nat- DVC, DipECVIM, ILTM,
intestine, the larvae migrate through the intestinal wall to the urally infected dogs have been the subject of several case reports MRCVS,
mesenteric lymph nodes, where two further moults occur. and small case series (Dodd 1973, Drape and Giraud 1976, Department of Veterinary
Immature L5 larvae then migrate via the hepatic portal vein, Patteson and others 1987, 1993, Cobb and Fisher 1990, Koch Clinical Sciences, The
liver and caudal vena cava to the right ventricle and pul- and others 1992, Martin and others 1993, Ramsey and others Royal Veterinary College,
monary arteries. Here they reach maturity approximately 33 1996, Otter and others 1997, Gould and McInnes 1999). Hawkshead Lane,
to 35 days after the dog was infected (Rosen and others 1970, The clinical signs of the infection are diverse but commonly North Mymms,
Bolt and others 1994). include coughing, dyspnoea, collapse and bleeding diatheses Hertfordshire AL9 7TA

426 The Veterinary Record, April 3, 2004