ISRAEL JOURNAL OF
VETERINARY MEDICINE
WEST NILE VIRUS ENCEPHALITIS IN HORSES
IN ISRAEL
Vol. 57 (2)
2002
S. Perl1, L. Fiette2, D. Lahav1, N. Sheichat1, C.Banet3, U.
Orgad1, Y. Stram4 and M. Malkinson3
1 Departments of Pathology, 3Avian Diseases and 4Virology, Kimron Veterinary Institute,
50250 Beit Dagan, Israel
2 Unit‫ י‬de Recherche et d'Expertise en Histotechnologie et Pathologie, Institut Pasteur, 25,
rue du Docteur Roux, 75724 Paris Cedex 15, France
Abstract
West Nile (WN) virus is a mosquito-borne flavivirus that can cause a variety of
neurological symptoms in mammals and birds. The virus is considered to be endemic in
many parts of Asia, Africa and Europe, particularly in countries bordering the
Mediterranean Sea, where outbreaks of disease occur from time to time. During August
through November 2000, an outbreak of West Nile fever affected over 400 people in Israel
with 29 fatalities. At the same time an outbreak of encephalomyelitis was diagnosed in the
local equine population. We describe here the clinical, histopathological and virological
findings of five terminally-affected equines presenting neurological signs that were
submitted between August and October 2000. The animals (two males and three female)
were 2 to 10 years old. Four of them were recumbent and all had paraplegia of various
degrees. Histologically, four had encephalitis and two of these had WN viral antigen in
sections of the brain and spinal cord with a sparse distribution of positive cells. WN virus
was isolated from three of the horses.
Introduction
West Nile virus (WNV) is a flavivirus belonging to the Japanese encephalitis-St.Louis
encephalitis serocomplex of the Flaviviridae family. The virus is transmitted by
mosquitoes, usually Culex spp., from avian reservoir hosts to susceptible mammals,
principally horses and man as well as to wild and domestic birds (1,2). WN disease is a
zoonosis and when favorable ecological and climatic conditions are in synchrony,
epidemics of encephalitis occur in humans with a case-mortality rate of up to 10%, while a
fatality rate of 40% in young domestic geese has been encountered (3).
Since its original isolation in Uganda in 1937 (4), WNV infections and fever have been
recognized in humans in numerous countries in Europe, Africa, western Asia since the
1960s (5). Equine disease was first described in Egypt (6), France during the 1960s (7)
and in Portugal in 1971(8). The disease was experimentally reproduced in equines as
encephalomyelitis (6,9,10). West Nile disease has reemerged recently in man, horses and
birds during the 1990s with mild to severe outbreaks of human encephalitis including
mortality being reported in Algeria (11), Romania (12), Czech Republic (5), Volgograd,
Russia (13) and the Congo (14). The virus was isolated for the first time in the New World
in New York City in 1999 and has since caused illness in humans and equines throughout
the eastern USA and southern Canada (15). Outbreaks affecting equines were also
described in this decade in Morocco in 1996 (16), Italy in 1998 (17), France in 2000 (18).
In Israel, outbreaks of WN fever in humans were first reported in the 1950s (19) and
most recently in 2000 (20). WNV was recently isolated from domestic geese and various
wild birds (2,3). Between August and October 2000, an unspecified number of horses from
various regions of Israel were reported by veterinary practitioners as showing neurological
signs of weakness and ataxia of the hind limbs (A.Steinman, personal communication).
We
describe here the clinical, virological and pathological findings of the disease affecting
four horses and one pony.
Materials and Methods
The first equine case was referred to us on August 23,2000 and the last case on
October 15. During this period five equines were submitted for post-mortem examination
to the Pathology Department of the Kimron Veterinary Institute. A complete necropsy was
carried out, and the brain and spinal cord from all five cases and samples of the viscera of
two horses were fixed in 10% neutral buffered formalin and processed routinely for
histopathology. Samples of brain and spinal cord were also frozen and stored at -20oC for
virological investigation.
Histological sections were stained with hematoxylin and eosin (H&E).
Immunohistochemistry was performed on 5 µm sections of selected paraffin blocks using
the standard peroxidase-anti-peroxidase method with amplification by the EnVision+
system (Dako) using paraffin sections. The primary antibody was an anti-WNV mouse
monoclonal antibody.
Virology
Brain and spinal cord were collected from necropsies of all five animals.10%
suspensions were prepared in PBS (pH 7.0), clarified and filtered through 0.22µ Millipore
filters. Six well plates containing semi-confluent Vero cell monolayers were inoculated with
filtrates. The plates were incubated at 370C in 5%CO2:95% air and inspected daily for
seven days for cytopathic effect. Two blind passages were performed.
RT-PCR
RNA was extracted using QIAamp viral RNA kit (Qiagen). RT-PCR was conducted with
primers WN132 and WN240 (21). The PCR product was 327 bp in size.
Results
Table 1: Summary of clinical, histopathological , immunohistochemical (IHC) and virological
findings in five horses
Hor
se
A
ge
No.
(y
rs)
1
2
M
/F
Clinical
signs
Duratio
Histology
IHC
n
isolati
on
of
illness
F
Ataxia,
fever,
Virus
4 days
No
lesions
3 days,
Encephali
tis
Negati
ve
+
A few
+
paresis,
recumbency
2
(pon
y)
8
F
Ataxia,
paraplegia
hind legs,
fever,
recumbenc
y
euthaniz
ed
positiv
e cells
3
10
F
Ataxia,
quadripleg
ia,
2 days,
euthaniz
ed
Very
discrete
Negati
ve
-
A few
+
encephalit
is
recumbenc
y
4
10
M
Ataxia,
7 days
quadripleg
ia,
Encephali
tis
positiv
e cells
circling,
fever
5
7
M
Ataxia,
paresis,
2 days
Encephali
tis
Negati
ve
-
recumbenc
y
Clinical findings:
The clinical findings are summarized in Table 1. All five horses were ataxic and became recumbent in the final s
horses (case nos. 1 and 2) only the hind limbs were affected while quadriplegia was present in the other three ho
illness ranged from 2 to 7 days and two horses had to be euthanized 2 and 3 days after they became recumbent.
Macroscopical findings
In one horse (case no.1) there were multifocal subcutaneous
haemorrhages in the subcutis and endocardium. In horse no.3,
a granulomatous focal lesion in the mandible was noted and
focal dark areas in the gray matter of the lumbo-sacral region of
the spinal cord were seen (Fig.1). In horse no.4 we observed a
few haemorrhagic infarcts and edema was seen in the lungs,
multifocal haemorrhages on the epicardium and endocardium,
and multifocal dark areas in the gray matter of the lumbar-sacral
region of the spinal cord. In horse no.5, only the head was
submitted for pathological examination and no pathological
changes were visible
Figure 1. Lumbar spinal cord; horse. Multif
matter.
Microscopical findings
In 4 of the 5 horses a non-suppurative, moderate to severe
polio-encephalomyelitis was observed. The lesions were located
mainly in the lumbar-sacral region of the spinal cord and in the
brain stem. Multifocal perivascular to diffuse haemorrhages
were present in some of the cases in the gray matter of the
spinal cord. (Fig.2). In other cases perivascular mononuclear
cell infiltrations were present in the gray matter. These were
composed of lymphocytes, plasmocytes and macrophages (Fig.
3). The lesions were symmetrical and bilateral and involved the
gray matter and particularly the ventral horns of the spinal cord.
Figure 2. Lumbar spinal cord. Multifocal an
gray matter. H&E. x5.
Slight inflammation and edema of the meninges and spinal cord
were also present in some horses.
No significant histopathological lesions were detected in the
lung, heart muscle, liver, spleen and kidney of any of the
horses.
Viral antigen:
We found viral antigen in two cases (Nos.2 and 4). Positive
cells were very sparse and associated with foci of inflammation
in the gray matter. They had the morphology of pyramidal
neurons in the ventral horns (Fig.4). In addition, we saw labeled
cellular debris in the neuropil or microglial cells (Fig.5).
Virological findings
Three WNV isolates were made from the brain and spinal cord
(horse no. 1, 2 and 4). The isolates were identified by RT-PCR
on Vero cell supernatants and mouse brain (Banet, personal
communication ). Partial sequence analysis of the E gene
showed that the isolates were 99.8% homologous with WNVISR98 isolated from a goose (22).
Figure 3. Lumbar spinal cord; severe lymph
H&E. x20.
Figure 4. Spinal cord: Immunoperoxidase-staining of WN viral
antigen in the cytoplasm of a neuron and multifocal glial cell
proliferation (gliosis). x20.
Figure 5. Spinal cord: Immunoperoxidase-s
glial cells. x40.
.
Discussion
In the five equine cases described here, the clinical and histopathological findings are
similar to those that have been described previously in spontaneous or experimental WNV
infection of horses and ponies. In the majority of affected horses, WNV infection usually
causes a sub-clinical illness, associated with the appearance of specific serum antibodies.
Only a small number of infected horses develop acute neurological signs varying from
20% to 40% in different studies, with a mortality rate of about 45% (1). Ataxia, weakness,
paresis, paraplegia or tetraplegia are the most consistent signs; while fever is not
(9,11,17,18).
Lesions induced by WNV are limited to the central nervous system. They consist of a
moderate to severe meningo-encephalitis associated with haemorrhages. They are
preferentially observed in the brain stem and ventral horns of the lumbo-sacral spinal cord,
whereas the brain and cerebellum show few changes. The preferential location and
microscopic appearance has to be considered in the differential diagnosis of equine viral
encephalitides, and it is essential to examine the spinal cord where WNV is suspected.
West Nile disease has then to be confirmed by appropriate tests, e.g. MAC-ELISA and
RT-PCR. As described previously in cases from Italy and New York (23,24), viral antigen
is detected in the gray matter in morphologically normal or degenerate neurons or fibers,
and in glial cells. In our cases, immunohistochemistry was positive only in two cases so
this method is not very sensitive. In our series, we used IHC and virus isolation in parallel.
Three cases were positive by both methods, in one case (No.1), the virus was isolated but
lesions were not found in the samples, while in horse no.4 we observed inflammatory
lesions similar to the other cases but virus isolation was negative. Considering the low
level of antigen that was found, immunohistochemistry does not seem to be the most
reliable method to confirm the diagnosis of WNV
References
1.