West Nile Virus Vaccination: IgM and
IgG responses in horses after injection
in different muscles
F.J. Jonquiere1, H.M.J.F. van der Heijden2, C. van Maanen2 and M.M.
Sloet van Oldruitenborgh-Oosterbaan1*
1. Department of Equine Sciences, Faculty of Veterinary Medicine, Utrecht University,
Yalelaan 114, 3584 CM Utrecht, the Netherlands
2. Animal Health Service (GD), Arnsbergstraat 7, 7418 EZ Deventer, the Netherland
*corresponding author
Dr. M. Sloet
Department of Equine Sciences
Yalelaan 114
3584 CM Utrecht
The Netherlands
phone +31302531350 , fax +31302537970
Email address: M.Sloet@uu.nl (M.M. Sloet van Oldruitenborgh- Oosterbaan)
Abstract
West Nile Virus (WNV) may cause severe symptoms in birds, horses and humans.
WNV is not (yet) present in the Netherlands, but it is steadily approaching from south-eastern
Europe. Recently, a WNV-vaccine (Duvaxyn®-WNV) became available in Europe. It is
claimed that vaccination results solely in an IgG response making it possible to differentiate
acutely-infected horses from vaccinated horses by using an IgM-based ELISA. The aims of
the study were to investigate this claim and to evaluate whether different intramuscular
injection sites influence the immunological responses and whether any local or systemic
adverse reactions would occur.
Twenty horses, 3 to 21 years old, were vaccinated twice (day 0 and 21) in different
muscle groups. Weekly blood samples were collected over a period of 42 days and tested for
Flavivirus IgM antibodies using a WNV-IgM ELISA and for Flavivirus Ig antibodies using a
WNV blocking ELISA.
None of the horses tested positive for WNV antibodies on Day 0. No side effects were
found following any of the intramuscular injections.
Location of the intramuscular injections showed no significant effect on the
immunogenic response.
All horses showed a clear Ig (total antibody) response to the WNV vaccination, but in
two horses this response was limited. Surprisingly, ten horses gave also a (limited) positive
IgM response. This indicates that an IgM capture ELISA is not reliable to diagnose acute
WNV infections in recently vaccinated horses.
Keywords: West Nile Virus, Vaccination response, ELISA
Introduction
West Nile Virus (WNV) is a positive stranded RNA virus belonging to the
Flaviviridae. It may cause an arthropod borne disease which is endemic in Africa, North
America, South America and parts of Europe (Castillo-Olivares, 2004). Arthropod-borne
Flaviviruses are maintained in an enzootic cycle of birds and mammal hosts. In the USA the
virus has been found in over 300 species of birds (Sloet, 2009). Arthropods, such as
mosquitos, act as vectors for WNV between avian and mammalian hosts. The Culex species
are considered to be the most important group of vectors for WNV( Shirafuji, 2009), but
numerous other mosquito species can be infected as well. The WNV is also found in ticks and
Culicoides species (Sloet, 2009). These vectors can infect mammalian species with WNV,
when taking a blood meal. WNV replicates in avian species, whereas mammals like horses
and people are dead-end hosts. In those species the multiplication of the virus is so low that
arthropods are not infected when taking a blood meal (Castillo-Olivares, 2004) and the
transmission cycle ends.
The clinical symptoms in the dead-end hosts, however, can be severe. The clinical
progress of a West Nile Virus infection can vary greatly, with symptoms varying from hardly
any (asymptomatic), mild flu-like symptoms to a severe form (West Nile Fever) involving
neurological symptoms with considerable mortality rates. Its important complications are
meniningo-encephalitis and other neurological symptoms such as acute paralysis or
paresis(Chu, 2007 ). The occurrence of the severity of the disease is influenced by the age and
physical condition of the host and the virulence of the virus (van der Heijden, 2008). In horses
the most common form is the asymptomatic version while in humans less than 1% of the
infected patients develop neurological symptoms (van der Heijden, 2008). However, in the
past decennium there has been an increased incidence of the neurological form among horses
and people (van der Heijden, 2008). In case of a WNV infection in horses with a virulent
lineage 1 strain as happened during the WNV epidemic in the USA about a third of the
affected horses will develop the severe neurological form of the disease and will either
eventually die or will have to be euthanized. The most common mild symptoms in horses are
low fever, mild ataxia and lethargy. The severe symptoms are: acute or progressive ataxia,
paralysis of the nervus facialis, fasciculations, epileptical episodes and other neurological
symptoms(Castillero-Olivares, 2004).
West Nile Virus was introduced into Northern America on the east coast in 1999. It
has rapidly spread over the whole of the continent (Malan, 2004) and is now endemic.
Recently (November 2008) some cases of WNV infection have been found in Europe (Italy,
Rumania, Austria and Greece) (ECDC website). It is not unlikely that the disease will spread
over Europe in the same way as in the USA.
In the USA several WNV vaccines have become available. In 2003 a new type of
vaccine was introduced that was based on recombinant DNA technology: a canarypox virus
vector WNV vaccine: Recombitek® (Merial). The first vaccine that was available in the
USA, the West Nile Innovator® vaccine (Fort Dodge), is based on inactivated whole virus
particles combined with a potent adjuvant. Many of these vaccines have been used for years
since, with hardly any adverse reactions reported and a reportedly good clinical protection
against WNV infections. Some manufacturer’s studies even claim a 100% protection (Seino,
2007). In case of the canarypox vector WNV vaccin, field studies showed also good results,
both for safety and efficacy (Seino, 2007).
In November 2008, the first WNV vaccine for Europe, Duvaxyn WNV® from Fort
Dodge, was approved by the European Medicines Agency (EMEA) . It became available in
the Netherlands during the summer of 2009. It is assumed that vaccines often do not trigger an
IgM response that is typical for the acute field infection with such a disease. This is also the
premise of the Duvaxyn WNV vaccine. This is especially important when vaccinating against
WNV in a WNV free country. This assumption makes it possible to distinguish a horse with a
WNV field infection from one that has been (recently) vaccinated using a IgM specific WNV
ELISA. Therefore, the main goal of this study is to examine the assumption that the WNV
vaccine will trigger a strong Ig response and no IgM response. In addition the effect of the
vaccination site on IgG and IgM antibody responses is examined, as well as whether
vaccination results in any adverse reactions.
Materials and Methods
Horses
A total of twenty horses was used in the study, eighteen of these belonged to the
Faculty of Veterinary Medicine and two were privately owned. The horses ranged in age from
3 to 21 years old. For breed and gender see Table 1. The study was approved by the Ethical
Commission for Animals (DEC) of the University of Utrecht.
Vaccination
On days 0 and 21 the horses were given an intramuscular injection in 5 different
muscles locations with Duvaxyn WNV® (Fort Dodge) using a single dose of 1 mL per animal
on day 0 and day 21. On days 1, 2, 22, and 23, all horses were checked for local and systemic
reactions due to vaccination. This was done by general inspection of the horses, checking the
appetite, behaviour, and inspection and palpation of the vaccination site, paying special
attention to swelling, local temperature and tenderness at the injection site.
Blood sampling
For the measurement of the humoral responses 20 mL blood was taken from the left or
right jugular vein using a vacuum system. In total 6 sequential blood samples were collected,
starting on day 0 just before the first vaccination, on day 7, 14, 21 (just before the second
vaccination), and day 28 and day 42. Serum was collected by centrifugation and stored at 18°C until tested.
WNV ELISAs
All the samples were semi-quantitatively tested for antibodies against WNV using
either a competition/blocking ELISA (total Ig) or an IgM-capture ELISA.
The WNV IgM capture ELISA was performed by coating ELISA plates with a rabbit
polyclonal antibody directed against horse IgM. Subsequently, unbound coating sites were
blocked using Bovine Serum Albumin. After washing, serum samples diluted 1:400 were
added to duplicate wells and incubated. After further washing either recombinant WNVantigen or control antigen was added and incubated. After another washing conjugate was
added, which consisted of a monoclonal antibody directed against WNV, labelled with an
enzyme. After a final washing, a chromogenic substrate solution was added and incubated.
The enzyme reaction was stopped and the optical densities (OD's) were measured using an
ELISA reader. The difference between the OD of the WNV antigen well and the control
antigen well was related to a positive control sample. The P/N-ratio is proportional to the
concentration of Flavivirus specific IgM antibodies present in the sample. For qualitative
interpretations, a P/N-ratio higher than 2.0 is considered positive, while a P/N-ratio of 2.0 or
lower is considered negative.
The WNV competition/blocking ELISA was performed according to the
manufacturer's inctructions. Shortly, samples were diluted 1:2 and added to the wells of
ELISA plates coated with a recombinant WNV antigen. After an incubation, plates were
washed and a conjugate, consisting of a monoclonal antibody directed against WNV labelled
with an enzyme was added and incubated. Then the plates were washed again and a
chromogenic substrate solution was added. The enzyme reaction was stopped and the optical
densities (OD's) were measured using an ELISA reader. The OD's were related to a negative
control sample. The resulting S/N percentage (S/N%) is inverse proportional to the
concentration of Flavivirus specific Ig antibodies present in the sample. For qualitative
interpretations, a sample with an S/N% less than 40% is considered positive, a sample with a
S/N% ≥40 and <50% as doubtful, and a sample with an S/N% ≥ 50% as positive.
Results
Vaccination reactions
None of the vaccinated horses showed any systemic reaction nor any significant local
reaction on the injection sites. Two horses showed a small local swelling on day 22, one day
after the second (booster) vaccination. At closer inspection these were diagnosed to be insect
bites since they were not at the exact location of the injection and neither warm nor painful on
palpation.
WNV Competition ELISA
In the competition ELISA, the average S/N% of the groups of horses vaccinated
against WNV in different muscles sharply declined with time following vaccination on day 0
(Figure 1). At day 14 all group averages were below the cut-off level of the test; i.e. most
samples were still positive. The croupe site injected horses on day 14 and 21 had slightly
higher average S/N% in comparison with the horses injected in other muscles. From day 28
on, all samples were strongly positive. Individual S/N% showed that two horses (7 and 13)
had a delayed response, while the response of horse 9 remained rather low. Upon retesting,
the results proved to be reproducible (data not shown).
WNV IgM Capture ELISA
The average S/N-ratio of the horses showed an increase between day 0 (vaccination)
and day 14 (Figure 2). The average S/N-ratio exceeded the cutoff-level at day 7 and 14 (i.e.
positive results). At the second vaccination day (day 21) the average S/N-ratio had declined
below the cut-off level, but this was followed by another increase in average S/N-ratio, which
was again above the cutoff-level at day 28, after which the mean antibody levels declined
again. At day 42 the average S/N-ratio was below the cutoff-level, but still higher than at day
0.
The sequential S/N-ratios of the 10 individual horses which had one or more positive
results in the WNV IgM Capture ELISA in time, are given in Figure 3. In seven animals the
S/N-ratios are relatively low (just above the cut-off level) and temporarily. However, in three
horses (numbers 2, 6, and 16) relatively high S/N-ratio's were found. The results were
confirmed when the samples were retested (data not shown).
In most of these 10 animals the S/N-ratio's increased after the first vaccination on day 0 and
day 14, followed by a decrease. A similar, but slighter increase and decrease was seen in some
horses following the second vaccination at day 21. The exception is horse 2 that hardly
responded to the first vaccination, but showed the highest S/N-ratio's of all following the
second vaccination.
Effect of the vaccination site on IgG and IgM antibody responses
The horses were vaccinated in groups of 5 at 4 different muscle locations on the body.
This study shows that the location of the vaccination has no statistical significant effect on the
antibody response to the vaccine.
Discussion
At the start of this experiment all Dutch horses used were negative for antibodies
against WNV. This indicates that none of these horses have had recent contact with WNV, as
to be expected in a WNV free country.
None of the animals showed any adverse effects to the vaccinations. This corresponds
with the findings in the USA where thousands of horses are vaccinated every year with a
comparable vaccine (e.g. West Nile Innovator by Fort Dodge), and very little side effects are
reported (website Fort Dodge) .
Generally blocking ELISAs have several advantages when compared with indirect
ELISAs or capture ELISAs: they can be more specific by using WNV epitope instead of
Flavivirus epitope and from a technical point of view they are species independent. A
disadvantage is that in general blocking ELISAs are less sensitive than capture ELISAs1.
However this may be compensated by using a lower pre-dilution of the sera to be tested. In
this experiment the dilution factors were 1:2 (competition-ELISA) versus 1:400 (IgMELISA), which begs the question if the capture ELISA really is more sensitive in this case?
After the first and second vaccination all the horses seroconverted in the WNV
competition ELISA, indicating that they developed a humoral Ig reponse against WNV. This
corresponds with the claim of the manufacturer of the vaccine.
The most important finding of the present study is the fact that three horses showed a
definite IgM response to the vaccination, whereas another seven horses showed a short and
weak response (just above the cut-off level). This implicates that during a WNV outbreak or
WNV surveillance it may not be possible to differentiate infected horses from recently
vaccinated horses using the IgM capture ELISA. This may have implications for the
interpretation of positive IgM results in monitoring programmes such as are implemented now
by the Animal Health Service at Deventer (GD), especially when vaccination against WNV
becomes common practice. For a correct interpretation of IgM capture ELISA results, the
vaccination history of the horses should be known.
Conclusions
None of the horses in this group tested positive for WNV antibodies on Day 0.
No side effects were found following the first WNV vaccination nor the booster.
No statistically significant association was found between the specific location of the
intramuscular injection and the WNV IgM and IgG responses.
All horses demonstrated a clear Ig (total antibodies) response to the WNV vaccinations, but in
two horses this response was less pronounced.
This study shows that 10 out of the 20 horses demonstrated a positive IgM response at
some point after first and/or second vaccination, indicating that the IgM capture ELISA used
is not reliable to diagnose acute WNV infections in recently vaccinated horses.
Conflict of interest statement
Fort Dodge sponsored this study in part by supplying the vaccines used.
Acknowledgements
The authors thank Fort Dodge for partial financial support of this study.
References
Castillo-Olivares, J., Wood, J., 2004. West Nile Virus infection of horses. Veterinary
Research, nr 35, p.467-483
Chu, Jang-Hann J., Cern-Cher, S. Chiang, Ng, Mah-Lee, 2007. Immunization of
Flavivirus West Nile Recombinant Envelope Domain III Protein Induced Specific Immune
Respons and Protection against West Nile Virus Infection. The Journal of Immunology,
p. 2690-2705
El Garch, H., Minke, J.M., Rehder, J., Richard, S., Edlund Toulemonde, C., Dinic, S.,
Andreoni, C., Audonnet, J.C., Nordgren, R., Jullard, V., 2008. A West Nile Virus (WNV)
recombinant canarypox virus vaccine elicits WNV-specific neutralizing antibodies and
cell-mediated immune responses in horses. Veterinary Immunology and
Immunopathology, nr 123, p. 230-239
Long, M.T., Gibbs, E.P., Bowen, R.A., ea., 2007. Efficacy, duration, and onset of
immunogenicity of a West Nile virus vaccine, live Flavivirus chimera, in horses with a
clinical disease challenge model. Equine Veterinary Journal, nr 39 (6), p. 491-497
Malan, Anette, K., Martins, Thomas B., ea., 2004. Evaluations of Commercial West Nile
Virus Immunogloblines G (IgG) and IgM Enzyme Immunoassays Show the Value of
Continuous Validation. Journal of Clinical Microbiology, p. 727-733
Seino, K.K., Long, M.T., Gibbs, E.P.J., ea., 2007. Comparative efficacies of three
commercially available vaccines against West Nile Virus (WNV) in a short-duration
challenge trial involving an equine WNV encephalitis model. Clinical and Vaccine
Immunology, p. 1465-1471
Shirafuji, H., Kanehira, K., Kubo, T., ea., 2009, Antibody Responses Induced by
Experimental West Nile Virus Infection with or without Previous Immunization with
Inactivated Japanese Encephalitis Vaccine in horses. Journal of Veterinar y Medical
Science, nr 71 (7), p. 969-974
Sloet van Oldruitenborgh-Oosterbaan, M.M., Goehring, L.S., ea, 2009. Emerging Vectorborne Diseases bij het Paard. Tijdschrift voor Diergeneeskunde, nr 134, p. 439 - 447
Woo, Patrick, C.Y., Susanna K. Lau, 2004. Longitudinal Profile of Immunoglobulin G
(IgG), IgM, and IgA Antibodies against the Severe Acute Respiratory Syndrome (SARS)
Coronavirus Nucleocapsid Protein in Patients with Pneumonia due to the SARS
Coronavirus, Clinical Diagnostic Lab Immunology, nr 11(4): p. 665–668
Websites
European Center for Disease Prevention and Control website:
http://www.ecdc.europa.eu/en/activities/sciadvice/Lists/ECDC%20Reviews/ECDC_DispFor
m.aspx?List=512ff74f-77d4-4ad8-b6d6-
bf0f23083f30&ID=938&RootFolder=%2Fen%2Factivities%2Fsciadvice%2FLists%2FECDC
%20Reviews
Fort Dodge Website
http://www.fortdodgelivestock.com/equine/equine-westnile.htm
Merial Website
http://us.merial.com/equine/products.asp
Canadian Food Inspection Agency website
http://www.inspection.gc.ca/english/anima/vetbio/eaee/vbeaalvace.shtml
Table 1
Breed, age and gender of horses involved in the WNV vaccination study and the location
of the intramuscular injection site
nr
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
breed
Trakhener
Arab
Welsh
KWPN
Thoroughbred
KWPN
Frysian
Thoroughbred
KWPN
Frysian
KWPN
KWPN
unknown
Shetland pony
Shetland pony
Shetland pony
Shetland pony
Shetland pony
Thoroughbred
KWPN
sex
gelding
stallion
mare
mare
mare
mare
gelding
mare
mare
gelding
mare
mare
mare
gelding
gelding
gelding
gelding
gelding
gelding
gelding
age
3
20
17
11
4
16
21
4
8
9
6
9
8
9
10
10
7
12
18
16
location of injection
croupe
pectoral muscles
semitendinoseus muscles
pectoral muscles
pectoral muscles
pectoral muscles
croupe
neck
neck
semitendinoseus muscles
neck
neck
neck
croupe
croupe
croupe
semitendinoseus muscles
semitendinoseus muscles
pectoral muscles
semitendinoseus muscles
S/N
60
40
20
120
0
0
100
10
20
30
40
days post first vaccination
S/N %
80
croupe
neck
pectoral muscles
semitendinoseus muscles
60
40
20
0
0
10
20
30
40
50
days post first vaccination
Figure 1: Sequential results of the WNV competition ELISA per location of the injection sites
of horses vaccinated against WNV on day 0 and 21. croupe
Means and standard errors of the mean
(S.E.M.) are indicated. The horizontal lines indicate neck
the test cutoff (<40% positive, 40-50%
pectoral muscles
doubtful, >50% negative).
semitendinoseus muscles
5
4.0
3.5
P/N ratio
3.0
2.5
2.0
1.5
1.0
0.5
0
10
20
30
40
50
days post first vaccination
Figure 2: Sequential average results of the WNV IgM capture ELISA of horses vaccinated
against WNV on day 0 and 21. The standard error of the mean (S.E.M.) is indicated. The
horizontal line indicates the test cutoff (<2.0 negative, >2.0 positive).
P/N rati
8
6
WNV-vaccinatie paarden
WNV IgM-capture ELISA
Selectie IgM positieve paarden
4
2
14
0
12
0
P/N ratio
10
2
3
6
8
11
13
16
17
19
20
8
6
4
2
0
0
10
20
30
40
50
days post first vaccination
2
3
Figure 3: Sequential individual results of of horses vaccinated against WNV on day 0 and 21
6
that scored positive in the WNV IgM capture ELISA at some point in time. The horizontal
8
line indicates the test cutoff (<2.0
negative, >2.0 positive).
11
13
16
17
19
20
MRSA PREVALENCE IN HEALTHY HORSES
IN THE NETHERLANDS:
A FOLLOW-UP STUDY
F. Jonquiere, E. van Duijkeren, J.A. Wagenaar, M.M. Sloet van
Oldruitenborgh-Oosterbaan
Faculty of Veterinary Medicine, Utrecht University, the Netherlands
Summary
Methicillin resistant Staphylococcus aureus (MRSA) is considered to be an increasing
problem in hospitalized horses in the Netherlands. In 2008, 24 of 259 horses admitted to a
veterinary hospital tested positive (1), while in a study in 2005, 200 randomly selected
ostensibly healthy horses all tested negative for MRSA (2). This raised the question as to
whether the high percentage of MRSA-positive horses reflected an overall increase in the
population or was a product of the selection of the referral horses. The aim of the present
study was to perform a follow-up screening of the 200 horses tested previously in the 2005
study. Of the horses screened in 2005, only 48 were still accessible (11 different premises).
These horses were sampled with an additional 54 horses selected randomly at 6 other
premises. Two nasal swabs were collected from each horse and incubated in two different
selective media: one method similar to the study in 2005 (2) and one more sensitive method
that is now used as a standard and has been used in the study of the referral horses (1). All
102 horses tested negative for MRSA using both isolation methods. No MRSA was found in
the 102 healthy horses tested in the Netherlands. Further studies are needed to establish
whether specific risk factors for testing MRSA-positive can be identified in the group of
referred horses.
Introduction
Methicillin-resistant Staphylococcus aureus (MRSA) is already considered a big problem in
hospitals and it is also an emerging problem in many animal species. In recent studies
methiciline-resistant Staphylococcus aureus infections in horses is also becoming more
common. However, the epidemiology of infection and colonization is still poorly understood.
The aim of this study is to do a follow-up report on a study by Busscher et al. in 2005 on the
prevalence of methiciline-resistant Staphylococcus aureus in healthy horses in the
Netherlands. In that report two hundred healthy horses housed at 24 different locations in the
Netherlands were tested for the presence of methiciline-resistant Staphylococcus aureus in
their nostrils and on the pasterns of each forelimb. All horses tested negative for MRSA.
Van Duijkeren et al. researched the MRSA colonization rates of horses and horse personnel in
2008, the possible occurrence of nosocomial transmission and the degree of contamination of
the hospital’s interior, because of a MRSA outbreak in hospitalized horses in the Netherlands.
In this study they also researched the presence of MRSA in horses on admittance at the clinic,
before entering the hospital grounds. They found an occurrence of 24 out of 259 horses, a
percentage of 9.3%.
The discrepancy between these results prompted the question whether the high percentage of
MRSA-positive horses reflected an overall increase in the population or was a product of the
selection of the referral horses. Had the prevalence of MRSA carriers among (healthy) horses
indeed increased towards an almost 10% incidence, did this percentage merely also show that
the screening tests have become much more sensitive or was there a selection of horses as
most of these had been seen by a veterinarian before referral? The best way to investigate this
question was to do follow-up study to the one of 2005, preferably testing the same horses
again and seeing whether the presence of MRSA in these horses had also increased to almost
10%. To exclude the influence of the sensitivity of the screening test, two culturing
techniques were used: one exactly the same as the 2005 study (method 2) and one more
sensitive method that is now used as a standard and has been used in the study of the referral
horses (method 1). Two nasal swabs were collected from each horse and incubated in two
different selective media according to methods 1 and 2.
Out of the 200 horses that were tested by Busscher et al. in 2005, 48 were still accessible, at
11 different premises. These horses were sampled with an additional 54 horses selected
randomly at 6 other premises to bring the total to a 102 horses tested at a total of 17 different
locations.
Methods & Materials
The samples were taken by using two sterile, dry swabs per horse. These were slowly swiped
through both nostrils, taking care to rotate the swab along the inside of the nose, where the
mucosa started. The swabs were stored inside their own containers, in a cool, dry place and
taken back to the lab. There they were cultured using different techniques. Two culture
techniques using a broth with aztreonam and ceftizoxime were used, one with (method 1) and
another without (method 2) pre-enrichment in Mueller Hinton with 6.5% NaCl. Method 1 had
been used successfully to detect MRSA, including ST398, in meat samples (van Loo et al.,
2007) and was used in the 2008 hospital study by van Duijkeren et al. and method 2 is the
standard procedure at the University Medical Centre Utrecht for clinical as well as screening
cultures of humans. This is surprising as method 1 is considered more sensitive to screen for
MRSA. (van Duijkeren et al, 2009). This is the method that was used in the 2005 study.
Method 1 (“new method”)
Swabs were put into tubes containing 5ml Mueller Hinton Broth (MHB), containing 6.5%
NaCl. After overnight incubation at 37,8 C one ml of the pre-enrichment broth was
transferred to 9ml phenolred mannitol broth PHMB) (BioMe´ rieux, Marcy l’Etoile, France)
with 5mg/ml ceftizoxime and 75mg/ml aztreonam. This enrichment broth was also incubated
overnight at 37,8 C and 10ml of the PHMB broth was subsequently plated onto sheep blood
agar (Biotrading, Mijdrecht, The Netherlands) and brilliance MRSA agar (Oxoid,
Basingstoke, United Kingdom).
Method 2 (“old method”)
Swabs were put into a tube with 5ml tryptone soya broth (TSB) containing 4% NaCl, 1%
mannitol, 16 mg/ml phenol red, 50mg/ml aztreonam and 5mg/ml ceftizoxime. After
incubation for 48 h at 37 8C, 10ml was plated onto sheep blood agar (Biotrading, Mijdrecht,
The Netherlands) and MRSA brilliance agar (Oxoid, Basingstoke, United Kingdom).
Identification of bacteria
Suspect colonies were identified as Staphylococcus aureus (S. aureus) using standard
techniques: colony morphology, Gram stain, catalase and coagulase test (Pasteurex Staph
Plus, Bio-Rad Laboratories, Hercules, USA). Colonies suspect of being MRSA were tested by
PCR for the S. aureus specific DNA fragment (Martineau et al., 1998), the mecA gene (De
Neeling et al., 1998).
Results
All 102 horses tested negative for MRSA using both isolation methods.
Conclusion
No MRSA was found in the 102 healthy horses tested in the Netherlands. Further studies are
needed to establish whether specific risk factors for testing MRSA-positive can be identified
in the group of referred horses. Since none of the horses tested positive, no statements can be
made regarding any differences in testing methodology. Further studies will be needed to
establish the sensitivity/specificity differences between methods 1 and 2.
Discussion
It was very surprising that none of the horses tested were positive for MRSA, regardless of the
method used. This goes against the expectation of an occurrence of about 10% in horses that
is based on the 2008 study by van Duijkeren et.al. One of the factors that could be of
influence on MRSA prevalence is the fact that all of the horses that were used in the van
Duijkeren-study were tested before being admitted at the hospital. This means that some these
horses were probably not completely healthy. The MRSA bacteria can take advantage of this
situation. Another factor that is related to that fact is the use of antibiotics. Well known risk
factors for MRSA infections are: previous colonization of the horse, previous identification of
colonized horses on the farm, antimicrobial administration within 30 days, admission to the
neonatal intensive care unit, and admission to a service other than the surgical service (4).
References
1. Busscher JF et all. The prevalence of methicillin-resistant staphylococci in healthy
horses in the Netherlands. Vet. Microbiol. 113, 131-136, 2006.
2. Van Duijkeren E et all. Methicillin-resistant Staphylococcus aureus in horses and
horse personnel: an investigation of several outbreaks. Vet. Microbiol. E-pub ahead of
print Aug 8 2009.
3. Weese, J.S., Rousseau, J.,ea., “Community associated methicillin-resistant
Staphylococcus aureus in horses and humans who work with horses”, in: Journal of
American Vet. Medical Association, 2005a, 226, p. 580-583
4. Weese, J.S., Lefebre, S.L., “Risk factors for methicillin-resistant Staphylococcus
aureus colonization in horses admitted to a veterinary teaching hospital”, in: Canadian
Veterinary Journal, 2007, 48, p. 921–926