Vet. Pathol.

26:191-194 (1989)

A Necrotizing Meningoencephalitis of Pug Dogs

D. R. CORDYAND T. A. HOLLIDAY
Department of Pathology and Department of Surgery, School of Veterinary Medicine,
University of California-Davis, Davis, CA

Abstract. Clinical and pathologic features of a sporadic, necrotizing meningoencephalitis affecting adolescent
and mature pug dogs are described. Many of the affected animals were closely related. Acute and chronic forms
occur, with clinical signs reflecting the pathologic affinity of the disease for the cerebral hemispheres. No etiologic
agent has been identified. The extensive necrosis and affinity for the cerebral hemispheres are similar to alpha-
type herpesvirus encephalitides of other species.

For more than 20 years we have observed in pug
dogs a unique non-suppurative meningoencephalitis
associated with extensive cerebral necrosis. Except for
one brief account5 the disease has not been described.
Because of its unusual nature and because of some
confusion about the disease, we present this retro-
spective study. Included are some speculations on
etiology, based on lesions.
Case Histories
The study material comprised 17 naturally occurring cases
of this disease in pug dogs from northern California (all from
within 160 km of Sacramento). All dogs were patients of our
Veterinary Medical Teaching Hospital. Six were males and
11 were females. They ranged in age from 6 months to 7
years (two aged 6 months, five aged 1 year, seven aged 2 to
3 years, and three aged 5 to 7 years). At least 11 of the dogs
were born in the same kennel, and their pedigrees revealed
numerous common ancestors, including one male that ap-
peared at least once in the three generations preceding each
affected dog. Pedigree data on the other six dogs were in-
complete and not contributory.
Ten dogs showed signs for 2 weeks or less, one for a month,
and six for 4 to 6 months. Thus, over half were acute and
about a third quite chronic. The most common clinical sign
was generalized seizures; these occurred in 16 of the 17 dogs.
In some cases, seizures were the initial sign ofillness reported
by the owners; in others, seizures began after variable periods
of lethargy and/or ataxia. Some form of depressed conscious-
ness, progressing from lethargy early in the disease to coma
terminally, occurred in all dogs. Other signs seen frequently
included walking in circles, head-pressing, blindness with
normal pupillary light reflexes, cervical rigidity (resistance to
passive movement of the head and neck), and opisthotonus.
Hemograms of 12 dogs did not reveal significant abnor-
malities. Various blood chemistries, including liver enzymes,
plasma proteins, and serum electrolytes, were done in seven
cases but did not reveal abnormalities. No abnormalities
were found in the urine of six dogs examined by routine
urinalysis.
191
Cerebrospinal fluid (CSF) was collected from the cerebel-
lomedullary cistern of 16 of the dogs. Analysis of 12 of these
specimens included total and differential cell counts and total
protein determination (Table 1). In all 12 specimens the total
white blood cells were increased (mean = 374 ? 178.2), and
the predominant cell type was the small lymphocyte, which
constituted from 71% to 989'0 of the nucleated cells (Table
1). Complete examinations were not done on the CSF of the
other four dogs, but elevated cell counts were found in three,
and the predominance of small lymphocytes was noted in
all. CSF of four dogs was cultured aerobically and anaerobi-
cally with negative results.
Nine of the animals died naturally; eight were killed when
they became severely ill. Tissue specimens were fixed in neu-
tral buffered formalin, embedded in paraffin, and cut at 6
Fm. Hematoxylin and eosin was the routine stain; selected
slides also were stained with Luxol fast blue-periodic acid-
Schiff-hematoxylin or Bodian methods.
Results
Marked pulmonary congestion and alveolar edema
were the chief extraneural lesions in these dogs. Ap-
pearing in all that died and in some acutely ill cases
that were killed, these lesions are thought to reflect the
mode of death rather than the primary disease.
Leptomeningeal hyperemia and occasional pete-
chiae were the chief gross neural lesions in acute cases.
Moderate localized swelling of the cerebrum was pres-
ent in a few acute cases, resulting in some asymmetry.
Grossly apparent yellow or tan depressed cortical foci,
tiny cortical cavities, and subcortical areas of gelati-
nous softening were reported in four chronic cases.
Microscopic study disclosed a non-suppurative
meningoencephalitis characterized by its marked nec-
rotizing nature and striking affinity for the cerebral
hemispheres. In the hemispheres, the inflammatory
and necrotic changes were severe and extensive, in-
volving both gray and white substance. Any part of the
cortex, including the hippocampal formation, might
192 Cordy and Holliday

Table 1. Cerebrospinal fluid cytology and protein of 12 pug dogs with necrotizing meningoencephalitis.
Small Large PMNst Unclassified
RBCS Total Protein
WBC* Lymphocytes Mononuclear Cells
Dog
(/d) (Yo)
Cells (%) (Yo)
(Yo)
(/PI) (mg/dl)

1 398 88 11 1 0 692 200

3 400 85 4 6 5 2 190.3
4 520 95 5 0 0 6 70.2
6 380 91 8 1 0 380 58.3
7 71 93 5 2 0 3 143
8 580 94 4 2 0 35 98.6
9 174 94 6 0 0 286 60.6
10 50 1 95 4 1 0 105 95.5
11 480 98 1 1 0 22 171
12 94 96 0 4 0 48 -
15 630 71 25 4 0 410 228
16 260 80 10 10 0 1,400 150
Means 374 90.0 6.9 2.7 0.4 282.4 122.1
SD 178.2 7.5 6.3 2.8 1.4 397.0 66.0
* White blood cells.
t Polymorphonuclear neutrophils.
$ Erythrocytes.

be affected. Cortical lesions were bilateral, asymmet-
rical, and often confluent over large areas. Subcortical
white substance was most often involved; paraventric-
ular white substance least. Lesions extended into the
corpus callosum in nearly half the dogs, and in several
the internal capsule was involved.
The leptomeningitis was multifocal, widespread, and
of greatly varying severity (Fig. 1). Characterized by
perivascular accumulations of lymphocytes, plasma
cells, and macrophages, the meningitis decreased
sharply in severity caudal to the tentonum. The heavi-
est cellular infiltrations were in cerebral sulci and in
the longitudinal and transverse fissures.. Occasional
substantial accumulations were seen basally over the
thalamus and in the cerebellopontineangle. Otherwise,
foci of meningitis caudal to the transverse fissure were
small and scattered. Foci diminished progressively
caudally until over the spinal cord less than half the
dogs had any section showing meningitis.
In the cerebral cortex, extensive neuronal necrosis
(Fig. 2) often occurred without concurrent evidence of
inflammation. The usual cortical reaction was one of
neuronal and glial necrosis, edema, hyperemia, occa-
sional petechiation, and diffuse and perivascular ac-
cumulation of mononuclear cells. A few neutrophils
appeared among the mononuclear cells where necrosis
was acute and severe. Hypertrophy and hyperplasia of
endothelium was often prominent.
In a few cases with longer clinical histories there were
pale, cell-poor cortical areas showing neuronal deple-
tion and astrocytic hypertrophy. Gemistocytes were
frequent in such sites. It is assumed these represented
areas of repair following selective necrosis.
Lesions in cerebral white substance were edema, glial
necrosis, loss of myelin staining, and diffuse and pen-
vascular infiltration of mononuclear cells. Axons
seemed to survive somewhat longer than myelin
sheaths. With severe glial and axonal destruction, areas
of malacia with eventual cavitation developed (Fig. 3).
All dogs had extensive cerebral necrosis, varying in
stage from selective necrosis of neurons and glia with-
out obvious neuropil disruption through progressive
neuropil breakdown and massive macrophage accu-
mulation (malacia) to the ultimate formation of cav-
ities containing only fluid and a few macrophages. There
was generally good correlation between observed clin-
ical course and stage of necrotic lesions. Nine acute
cases and the single subacute case showed only selec-
tive cell necrosis. The other acute case reported signs
of 10 days duration and had several small subcortical
cavities. All six chronic cases had malacia, cavitation,
or both. Rarely, a dog with chronic lesions also showed
a few areas of more recent selective necrosis.
Four of the chronic cases also had necrotic foci in
nearby non-hemispheric sites. One dog with a necrotic
focus in one internal capsule showed involvement of
the adjacent globus pallidus. Another animal had bi-
lateral necrosis of the columns of the fornix, presum-
ably related to hippocampal damage. Subpial malacia
of the rostra1 colliculi, associated with severe overlying
meningitis in the transverse fissure, was noted in two
dogs.
Caudal to the cerebral hemispheres there was a strik-
ing diminution in extent and severity of the lesions.
Necrosis was seen only in the four chronic cases men-
tioned above, where the lesions seemed to be related
 Necrotizing Meningoen.cephalitis of Pug Dogs
Fig. 1. Non-suppurative leptomeningitis. Some inflam-
matory cells and neuronal necrosis in underlying cortex. HE.
Fig. 2. Necrotic neurons in deeper strata of cortex. HE.
to primary cerebral or meningeal inflammation. A little
diffuse inflammation was seen occasionally in the basal
ganglia, extending from the overlying lateral cortex,
and in the ventral thalamus from overlying meningitis.
Otherwise, only a few scattered perivascular cuffs were
seen in non-cerebral areas. Subpial sites were most
common. Frequently, cuffs were found in the paren-
chyma near the roots of the tela choroidea of the third
and fourth ventricles. A few tiny cuffs with diminished
staining of adjacent myelin were noted in the foliar
white substance of the cerebellum. Myelitis was prac-
tically absent; a rare cuffed vessel extending centrally
from a focus of meningitis was seen in only two dogs.
Two of three trigeminal ganglia available for ex-
amination had numerous perivascular mononuclear
accumulations. This was associated with radiculitis in
one. Inflammation of the roots of other cranial and
193
Fig. 3. Cavity in cerebral white substance showing glio-
vascular trabeculae and a few remaining macrophages. HE.
spinal nerves was not seen, although meningitis over-
lay several.
Unequivocal inclusion bodies were not found in any
of the brains. One chronic case had homogeneous aci-
dophilic nuclear inclusions in epithelial cells of the
renal convoluted tubules. The record of this dog noted,
without details, the isolation of a “herpes-like’’ virus
from the brain.
Discussion
This meningoencephalitis is of sporadic nature and
affects adolescent or adult dogs. The dogs could be
classified into one of two clinical groups: an acute group
showing overt signs for 2 weeks or less, and a chronic
group, affected for several months. Important clinical
features are the neurological signs referable almost ex-
clusively to the cerebrum and meninges, i.e., depressed
consciousness, seizures, amaurosis, ataxia. The cer-
vical rigidity seen in some cases reflects the meningitis
found hist~logically.~~~ The cerebrospinal fluid (CSF)
changes are an important clinical diagnostic feature.
The preponderance of small lymphocytes in the CSF,
seen in all dogs reported here, is unusual in granu-
lomatous meningoencephalomyelitis,2 toxoplasmosis,
or canine d i ~ t e m p e r . ~ , ~
The unique pathologic features of the meningoen-
cephalitis are its non-suppurative, necrotizing nature
and its strong predilection for the cerebral hemi-
spheres. Selective necrosis often occurred without con-
current inflammation; therefore, it may be the initial
lesion. Whether meningeal infection occurs first, with
subsequent parenchymal extension, is uncertain. A he-
matogenic or CSF spread does not appear likely from
the topography of the lesions. The affinity for the cere-
I94 Cordy and Holliday
brum may reflect an intrinsic vulnerability of the cere-
brum, or it may be related to a portal of entry or site
of latency. Lesions in chronically affected animals show
little necrosis and appear to represent residual damage
and reparative response.
The cause is unknown. Those few canine encepha-
litides which occasionally show some necrosis can be
readily eliminated. In canine distemper encephalitis
there is frequently surface-related demyelination of
brainstem and cerebellum. Necrosis, which is rare in
distemper, is reported as present chiefly in rhinence-
phalic structures.6J0Necrosis in the pugs was not lim-
ited to rhinencephalon but could be found in any part
of the cortex. Canine granulomatous meningoencepha-
litis has some selective cell necrosis but no massive
m a l a ~ i aAlthough
.~ De Lahunta reported granuloma-
tous changes in his pug cases,5none of our cases showed
such lesions. This may be a matter of differing ter-
minology. Toxoplasmosis often has small necrotic foci,
but these are vessel-related and randomly 10cated.~
Seizures due to a primary inflammatory disease could
have induced or augmented the necrosis, but the extent
of white substance necrosis would appear to make this
mechanism unlikely. Toxic or metabolic causes also
seem improbable in what is primarily an inflammatory
entity. The necrosis was neither infarctive nor laminar
as in cardiac arrest.
The extensive necrosis and the affinity for the ce-
rebral hemispheres are features suggestive of alpha-
type herpesvirus encephalitides of man and other an-
imals.1,7,11,12It may be that the pug cases are recurrences
of latent infections following an initial neonatal infec-
tion with canine herpesvirus type 1. I 3 The isolation of
a herpes-like virus from one of the early cases is tan-
talizing, but the isolate was not retained. Failure to
find clear-cut nuclear inclusions does not negate the
role of a herpesvirus infection since inclusions are often
sparse or absent in such diseases. It takes some days
for inclusions to appear, and virus may no longer be
present in chronic cases. A search for herpesvirus in
future cases is warranted.
This disease seems limited to pugs and possibly to
certain pug lineages. It is unknown if cases in the east-
em United States5 were related to the northern Cali-
fornia dogs. Eleven of the latter were closely related,
and it is possible that others were related also, as there
Request reprints from Dr. T. A. Holliday, Department of Surgery, School of Veterinary Medicine, University of California-
Davis, Davis, CA 956 16 (USA).
are relatively few pug breeders in the area. Since the
population at risk did not suffer an unusual incidence
of opportunistic infections, we do not consider a he-
reditary immune deficiency of T or B cell type likely.
Perhaps there was some genetic predisposition which
lowered resistance to primary infections or facilitated
latency and reactivation8
Acknowledgements
The authors acknowledge the contributions of the many
referring veterinarians, clinicians, and pathologists initially
involved in the cases reported here.
References
1 Adams H, Miller D: Herpes simplex encephalitis: a clin-
ical and pathological analysis of 22 cases. Postgrad Med
J 49~393-397, 1973
2 Bailey CS, Higgins RJ: Characteristics of cerebrospinal
fluid associated with canine granulomatous meningoen-
cephalitis: a retrospective study. J Am Vet Med Assoc
1 8 8 ~18-421,
4 1986
3 Braund KG: Clinical Syndromes in Veterinary Neurol-
ogy. Williams and Wilkins, Baltimore, 1986
4 Cordy D R: Canine granulomatous meningoencephalo-
myelitis. Vet Pathol 16:325-333, 1979
5 De Lahunta A: Veterinary Neuroanatomy and Clinical
Neurology, 2nd ed. WB Saunders, Philadelphia, 1983
6 Finnie J W, Hooper PT: Polioencephalomalacia in dogs
with distemper encephalitis. Aust Vet J 61:407408, 1984
7 Hill BD, Hill MWM, Chung YS, Whittle RJ: Menin-
goencephalitis in calves due to bovine herpesvirus type
1 infection. Aust Vet J 61:242-243, 1984
8 Hill T J: Herpes simplex virus latency. In: The Herpes-
viruses, ed. Roizman B, vol. 3, pp, 174-240. Plenum
Press, New York, London, 1985
9 Koestner A, Cole CR: Neuropathology of canine toxo-
plasmosis. Am J Vet Res 21:831-844, 1960
10 Lisiak J A, Vandevelde M: Polioencephalomalacia as-
sociated with canine distemper virus infection. Vet Pa-
tho1 16~650-660, 1979
1 1 Nunoya T, Tajima M, Kuwahara H: Pseudorabies virus
infection in piglets accompanying with the lesion of bi-
lateral encephalomalacia. Jpn J Vet Sci 47: 165-169, 1985
12 Olander HJ, Saunders JR, Gustafson DP, Jones R K
Pathologic findings in swine affected with a virulent strain
of Aujeszky’s virus. Pathol Vet 3:64-82, 1966
13 Percy DH, Olander HJ, Carmichael LE: Encephalitis
in the newborn pup due to a canine herpesvirus. Pathol
Vet 5:135-145, 1968