Necrotizing Meningoencephalitis of Maltese Dogs

Transcription

1 Vet Pathol 32: (1995) Necrotizing Meningoencephalitis of Maltese Dogs I. H. STALK, B. CHADWICK, B. DAYRELL-HART, B. A. SUMMERS, AND T. J. VAN WINKLE Laboratory of Pathology (IHS, TVW) and Department of Clinical Studies (BD), School of Veterinary Medicine, University of Pennsylvania, Philadelphia, PA; Department of Pathology, School of Veterinary Studies, Murdoch University, Perth, WA, Australia (BC); and Department of Pathology, New York State College of Veterinary Medicine, Cornell University, Ithaca, NY (BAS) Abstract. The clinical and pathologic features of five young Maltese dogs with a necrotizing meningoencephalitis were studied and compared with published reports of the necrotizing meningoencephalitis of Pug dogs. The ages of the Maltese dogs ranged from 9 months to 4 years. Four dogs were male, and one was female. The dogs had a history of seizures with or without other neurologic signs for 3 days to 20 weeks prior to death. Cerebrospinal fluid examination in three dogs revealed a pleocytosis and elevated levels of protein. At necropsy, the cerebrum was asymmetrically swollen in four dogs, with a loss of distinction between the gray and white matter and mild to moderate asymmetrical dilation of the lateral ventricles. Histologically, there was extensive necrosis and nonsuppurative inflammation of the cerebral gray and white matter, overlying meninges, and adjacent thalamus and hippocampus. The 4-year-old dog had the longest duration of clinical signs and had little inflammation but extensive atrophy of affected areas, with astrocytosis. The clinical course and pathologic changes in these Maltese dogs are indistinguishable from those in reported cases of necrotizing meningoencephalitis of Pug dogs, indicating that this lesion is probably not unique to Pug dogs. Key words: Dogs; encephalitis; Maltese dogs; meningoencephalitis. Meningoencephalitis with necrosis of the neuroparenchyma occurs in dogs secondary to a variety of infectious agents, including protozoa, viruses, and fungi; however, in many dogs the cause of the meningoencephalitis is unknown. We report here on five Maltese dogs that had a necrotizing meningoencephalitis of unknown etiology that resembles the necrotizing meningoencephalitis of Pug dogs. Materials and Methods Three Maltese dogs (Nos. 1-3) were clinically evaluated at the Veterinary Hospital of the University of Pennsylvania (VHUP). One dog (No. 4) was evaluated by a private practitioner and one dog (No. 5) was seen at the Murdoch University Veterinary Hospital (MUVH) in Western Australia. The signalment and clinical findings of these five animals are summarized in Table 1. Dog No. 1 was a 4-year-old male that was presented to the referring veterinarian because of seizures. The dog had been vaccinated 6 months previously with a distempedhepatitis/leptospirosis/parvovirus/parainfluenza combination vaccine. At presentation, the dog circled to the right and wandered aimlessly. The dog was hospitalized and treated with diazepam, phenobarbital, and dexamethasone. Complete blood count (CBC), serum chemistry panel, and urinalysis (UA) results were within normal limits. The dog improved slightly and was discharged after 2 days. There was gradual improvement over 2 months, but then the dog had 230 more seizures and was returned to the veterinarian for further examination. At that time, an initial cerebrospinal fluid (CSF') evaluation revealed a high white blood cell (WBC) count (85 WBC/mm3) and protein content (33 mddl). The differential cell count was 50 neutrophils, 25 lymphocytes, and 10 monocytes. Despite continued treatment with phenobarbital, dexamethasone, and amoxicillin, the clinical signs progressed and the dog was referred to the VHUP. On physical examination, the dog had ataxia, hind-limb paresis (which was worse on the left), postural deficits in both hind limbs and the left forelimb, and a left menace deficit with normal pupillary light response. A second CSF evaluation was within normal limits. Computed tomography (CT) scans revealed a radiolucent, non-contrast-enhancing area in the left cerebrum and dilation of the ventricular system. Magnetic resonance imaging (MRI) revealed high-signal regions in the right parietal and left frontal areas as well as asymmetrically dilated lateral ventricles. The radiographic interpretation was that an ischemic or inflammatory condition existed. Despite therapy, seizure activity continued with increasing frequency and the dog was euthanatized 5 months after the initial seizure. Dog No. 2 was a 16-month-old male who had been vaccinated for distemper and rabies during the first year of life. The dog was presented to the referring veterinarian with a 2-3-week history ofprogressive vision loss and a recent onset of seizure activity. Serum chemistry and UA evaluation produced no significant findings, and the dog was treated with prednisone. The dog improved only slightly and was referred to the VHUP. At presentation, the dog circled to the left,

2 Vet Pathol 323, 1995 Meningoencephalitis of Maltese Dogs 23 1 had right postural deficits, and lacked a menace response bilaterally but had normal pupillary light responses. An increased number of white blood cells were seen in the CSF (50 WBC/mm3), but a differential cell count was not done. A CT scan revealed asymmetrical dilation of the lateral ventricles and dilation of the third and fourth ventricles and aqueduct. There was a contrast-enhancing area in the left parietal lobe, which was suggestive of necrosis or edema. MRI findings were similar, with ventricular system dilation and a high-signal area in periventricular white matter of the left parietal region. Differential diagnoses included neoplasia, inflammation, and hemorrhage. A clinical diagnosis of granulomatous meningoencephalitis was made. The dog was treated with prednisolone and chloramphenicol, but clinical signs progressed. A second CSF evaluation 3 weeks after the first revealed elevated white blood cells (59 WBC/mm3) and protein (126 mg/dl). A differential cell count was not done. The dog was euthanatized 4 weeks after presentation to the VHUP. Dog. No. 3 was a 2-year-old male that had never been vaccinated. After its first seizure, the dog was presented to the VHUP for evaluation. CBC, serum chemistry, and UA results were unremarkable. The dog was normal for the next 36 hours but then had a second seizure. The dog was treated with Valium and phenobarbital but continually paddled and was later euthanatized. Dog No. 4 was a 9-month-old female that had been vaccinated for coronavirus and distemper virus 3 weeks prior to presentation to a local veterinarian because of a seizure episode. The dog had bradycardia and tachypnea and was slightly dull, but the neurologic examination was otherwise normal. Over the next few days, the dog had episodes of weakness, recumbency, dullness, drooling, and focal facial seizures that alternated with periods when the dog was bright, alert, and ambulatory. One period of opisthotonos was noted on the fourth day. The dog was treated with diazepam, phenobarbital, dexamethasone, and chloramphenicol and despite clinical improvement was found dead 4 days after presentation. Dog No. 5 was a 13-month-old male. This dog had received two distemper/hepatitis/parvovirus vaccinations prior to 4 months of age and a follow-up parvovirus vaccination at 4 months of age. The dog was presented to the MUVH with an acute onset of seizure activity. Seizures, which were not related to external stimuli, recurred over a 3-week period, with an interictal phase of 4-5 days. The dog was generally normal during the interictal period, although one neurologic examination revealed slight hind-limb ataxia, mild bilateral hind-limb proprioceptive deficit and a behavioral change described as increasing nervousness or mild hysteria, particularly when handled. CSF analysis revealed marked leukocytosis (247 WBC/mm3) with elevated protein (1 1 1 mg/dl). The differential cell count was 217 lymphocytes, 27 monocytes, and 3 neutrophils. CBC, serum chemistry, and UA were unremarkable except for elevated alkaline phosphatase (540 U/liter; normal = U/liter). A clinical diagnosis of granulomatous meningoencephalitis was made, and therapy with prednisolone was begun. The dog responded poorly and was euthanatized 1 week later. All five dogs were necropsied. Dog Nos. 1,2, 3, and 5 were Table 1. Signalment and clinical findings for five Maltese dogs with necrotizing meningoencephalitis. csl- ~ ~~ Duration of Dog No. Age Sex Clinical signs WBC/ Protein mm3 (mg/dl) 1 4 years M 20 weeks year M 6 weeks It QNS 59$ 126$ 3 2 years M 3 days NE NE 4 9 months F 4 days NE NE 5 13 months M 4 weeks * Cerebrospinal fluid. Normal values: 0-5 WBC/mm3; 0-25 mg protein/dl. QNS = quantity not sufficient to test. NE = not evaluated. t Second sample, 15 days later. 11 Second sample, 21 days later. necropsied by the pathology sections of the institution at which they were clinically evaluated. Tissue samples of major organs were fixed in 10% neutral buffered formalin with the exception of brains, which were fixed in 50% formalin. The spinal cord was also removed from dog Nos. 1 and 5. The private practitioner performed the necropsy on dog No. 4, fixed the brain and liver in 10% neutral buffered formalin, and submitted these tissues to the Laboratory of Pathology at the University of Pennsylvania for histologic evaluation. For all five dogs, representative sections of central nervous system and other organs were processed routinely for histopathology and stained with hematoxylin and eosin (HE). In dog No. 2, only the brain was examined histologically. Sections of brain from dog Nos. 1,2, and 3 were also stained with Lux01 fast blue-holmes. One section of brain from dog Nos. 1, 2, 3, and 5 was stained for glial fibrillary acid protein (GFAP) in an immunocytochemical procedure using polyclonal antibody (1 : 3,000 dilution) (Dako Corp., Carpinteria, CA) and streptavidin-biotin. Sections of brain from all five dogs were stained for canine distemper virus antigen using a monoclonal antibody to the nucleocapsid protein of canine distemper virus,5 diluted 1 : 1,000, in a streptavidin-biotin procedure. Results Significant gross and histologic lesions were limited to the central nervous system. Gross lesions In dog No. 1, the left cerebral hemisphere was pale gray and swollen. The right hemisphere was firm, slightly shrunken, and white dorsally with congested meninges laterally and ventrally. On cut surface, there was a loss of distinction between gray and white matter in the rostra1 and middle left cerebrum. The right lateral ventricle was moderately dilated, and the right cerebrum was thinner than the left because of severe atrophy of the white matter. No gross lesions were noted in dog No. 2. In dog No. 3, the lateral ventricles were mildly di-

3 232 Stalk, Chadwick, Dayrell-Hart, Summers, and Van Winkle Vet Pathol Fig. 1. Brain, cross section at level of septal nuclei; dog No. 4. The right side of the brain is swollen. There is a soft area with a loss of distinction between gray and white matter in the right cerebrum extending from the coronal sulcus (arrow) ventrally to the lateral aspect of the septal nucleus (arrowhead) and medially to the internal capsule. Bar = 0.5 cm. lated, and the cerebellum and brain stem were compressed. On cut surface, portions of the cerebrum were soft and gray, and there was loss of distinction between gray and white matter. Changes were most severe in the ventral cerebrum. In dog No. 4, the right piriform and right frontal lobes were 'Oft and tan with defined sulci' On cut surface, there was a soft area in the right ventral cerebrum with loss of distinction between gray and white matter, whereas in the right dorsolateral cerebrum the white matter was slightly widened and poorly demarcated from adjacent gray matter (Fig. 1). In dog No. 5, the lateral ventricles were moderately dilated. Most of the cerebral gyri were flattened. On cut surface, there were multifocal to confluent areas of discoloration with loss of distinction between gray and white matter. These changes were most severe in the occipital and ventral cerebrum. Histologic lesions In dog No. 1, the left lateral ventricle was mildly dilated and the right was moderately dilated. The entire right cerebral cortex from the frontal lobe through the occipital lobe, excluding the piriform cortex, was markedly thinner than the left, with replacement of normal parenchyma by numerous astrocytes (Fig. 2). Several cavitations up to 1 mm in diameter were present at the junction between gray and white matter. Perivascular cuffs of lymphocytes were present in these sclerotic areas, in the adjacent parenchyma, in the nonsclerotic cerebral cortex bilaterally, and in the right Fig. 2. Right dorsal cerebrum, base of endomarginal sulcus (*); dog No. 1. There is marked atrophy of the cerebral gray (G) and white (W) matter with marked astrocytosis. HE. Bar = ~m. mesencephalon. The sclerotic areas were poorly demarcated from adjacent tissue by a zone of astrocytic gliosis, which extended into the right thalamus and left medial corona radiata. Astrocytic gliosis was also present in the corpus callosum and hippocampus bilaterally and in the caudal ventral cerebrum. Endothelial cell hypertrophy was present in the right ventral cerebrum. The meninges overlying the sclerotic areas were normal, but adjacent meninges had mild infiltrates of lymphocytes and macrophages. Small numbers of perivascular lymphocytes and plasma cells were also seen within the spinal cord. GFAP staining of a section of left rostra1 cerebrum demonstrated large numbers of astrocytic fibers in the submeningeal cortex. A few leptomeningeal vessels of the spinal cord had mild perivascular infiltrates of lymphocytes. In dog No. 2, the dorsal and lateral frontal lobe, the adjacent internal capsule, the right side of the thalamus, and the right and left hippocampus were affected. There were extensive areas of necrosis in both white and gray matter with infiltrates of macrophages and

4 Vet Pathol 32:3, 1995 Meningoencephalitis of Maltese Dogs 233 Fig. 3. Cerebrum, endomarginal gyrus; dog No. 2. There is a marked focal infiltration of lymphocytes in the meninges and multifocal infiltration around vessels in the parenchyma. HE. Bar = 750 pm. lymphocytes and with perivascular cuffs of lymphocytes (Fig. 3). Scattered neutrophils were also present. Within affected areas, vessels had plump endothelial cells. Astrocytosis was present both within and peripheral to the lesions. The leptomeninges had mild infiltrates of lymphocytes, macrophages, and plasma cells. Staining with Luxol fast blue-holmes demonstrated decreased myelin in affected white matter. GFAP stain demonstrated increased numbers of astrocytes with prominent fibers in white matter predominantly adjacent to the necrotic areas. Severely sclerotic areas alternated with areas of active inflammation and necrosis in some sections of the cerebral cortex. Lesions in dog No. 3 were limited to the neocortex and were similar to those of dog No. 2, except astrocytic sclerosis was not present. The leptomeninges overlying the most severely affected areas had more severe infiltrations of lymphocytes and macrophages (Fig. 4) than were seen in dog No. 2. Lesions in dog No. 4 were similar to those of dog No. 3 and were Fig. 4. Cerebrum, ectosylvian gyrus; dog No. 3. There are extensive infiltrates of lymphocytes and macrophages and scattered neutrophils in the meninges, perivascularly, and in the neuropil of the cortex. HE. Bar = 40 pm. most severe in the right ventral lateral cerebral cortex, including the piriform cortex, and extended from the frontal lobe caudally through the temporal lobe. Scattered lymphocyte foci were in the contralateral cerebral cortex. In addition, in dog No. 4 at the periphery of necrotic areas there were ischemic neurons characterized by cytoplasmic hypereosinophilia and nuclear pyknosis. Lesions in dog No. 5 were similar to those of dog Nos. 3 and 4. They were predominantly in the cerebrum and most severe in the gray matter. Rare perivascular cuffs of lymphocytes were noted in the mesencephalon and spinal cord. Small numbers of lymphocytes and macrophages were present in the leptomeninges of the mesencephalon and cerebellum. All five cases were negative for canine distemper viral antigen by immunohistochemical methods. Discussion The clinical and pathologic findings in these five Maltese dogs are similar to those in reported cases of

5 234 Stalk, Chadwick, Dayrell-Hart, Summers, and Van Winkle Vet Pathol 32:3, 1995 necrotizing meningoencephalitis of Pug dog~l.~ and to those of Pug dogs with necrotizing meningoencephalitis seen at our institutions. The lesions are quite distinctive and do not resemble those of any other canine neurologic disease. The definitive diagnosis of this disease is based on histologic findings. The neurologic signs (predominantly seizure activity) and CSF findings are compatible with cerebral disease but are nonspecific. Similar signs and clinical findings may be present in a variety of inflammatory, necrotizing, and neoplastic conditions. Several of these dogs had a clinical diagnosis of granulomatous meningoencephalitis (GME) based on the clinical signs and CSF results. There were no lesions in any other organs that could be correlated with the necrotizing meningoencephalitis. Two dogs (Nos. 1 and 2) were evaluated with MRI and CT scans. The location of the lesions detected generally correlated with those seen at necropsy. In dog No. 2, the scans had shown dilated third and fourth ventricles, which were considered normal at necropsy. Results of these ancillary studies were insufficiently conclusive to permit a definitive diagnosis, but results were compatible with necrotizing meningoencephalitis. Based on our findings, when MRI or CT scans suggest a focal inflammatory lesion, necrotizing encephalitis should be considered as a diagnosis. The normal CSF findings (in one of two samples) in dog No. 1 when the dog was reevaluated later in the course of the disease may have been a reflection of lack of active necrosis or meningeal involvement at the time of examination or the result of glucocorticoid therapy. The etiology of this necrotizing meningoencephalitis is unknown. Some reported cases of necrotizing meningoencephalitis in Pug dogs occurred in related animals, suggesting that there may be a genetic component.ix3 The predilection for purebred animals might also suggest a genetic component. The relationships among the dogs in this study is unknown, but the three dogs examined at the VHUP were obtained from breeders in three different states. Four of these Maltese dogs were males; however, the sample size is small and sex predisposition is uncertain. A sex predilection has not been noted in Pug dogs. Four of these dogs were 2 years old or younger. A similar age predilection has been noted in Pug dogs.' The vaccination history for these dogs is incomplete but highly variable. One dog had never been vaccinated; others were adequately vaccinated. No correlation could be made between vaccination status and the necrotizing meningoencephalitis. The distribution of lesions is not consistent with a vascular etiology. Although lesions were often localized, they occasionally were continuous across the midline and extended beyond the distribution of a single vascular bed. No vascular lesions have been reported in Pug dogs, and none were noted in these Maltese dogs or in Pug dogs we have examined. The dog with the longest clinical history (20 weeks, dog No. 1) also had the most extensive atrophy and astrocytic sclerosis, whereas the dogs with short clinical courses (3 and 4 days, dog Nos. 3 and 4) did not have extensive sclerosis. In dog No. 2, the clinical signs began 6 weeks prior to death, and areas of necrosis and inflammation alternated with areas of sclerosis. The predilection of this disease for the cerebrum is an important feature, which has also been noted in Pug dogs with necrotizing meningoencephalitis. A predilection for cerebellar and cerebral white matter has been noted with focal GME,3 but most inflammatory diseases of the nervous system are not site specific, and many cases of GME involve the brain stem and spinal cord extensively. Inflammation of the trigeminal ganglia was noted in two Pug dogs,' but this site was not evaluated in these Maltese dogs. The lesions in these Maltese dogs and in Pug dogs previously examined by us and reported by others are somewhat suggestive of a viral etiology. Authors of one study proposed that a herpesvirus might be involved because of some similarities between necrotizing meningoencephalitis of Pug dogs and human herpes simplex type 1 (HSV-1) encephalitis.' In adults with HSV- 1 infection, lesions are localized to the basal frontal-temporal regions and are often unilateral but do not have extensive deep white matter necrosis.2 In llamas infected with equine herpesvirus- 1, lesions were most severe in the rostra1 cerebrum, and neuronal changes were very mild caudal to the level of the thalamus. These llamas had a vasculitis with edema and multifocal hemorrhage and intranuclear inclusion bod- ie~.~ Attempts at virus isolation in Pug dogs with necrotizing meningoencephalitis have been unsu~cessful.~ Our attempt to identify herpesvirus antigen within the brains of these dogs was unsuccessful because of difficulties in establishing an immunocytochemical procedure for canine herpesvirus in formalin-fixed central nervous system tissues. Canine distemper virus (CDV) antigen was not identified in the brains of these dogs. However, this result does not exclude the possibility of a restricted CDV infection that would not be detected by the methods used. Clinically and pathologically, this necrotizing meningoencephalitis in Maltese dogs closely resembles necrotizing meningoencephalitis of Pug dogs, and these two diseases probably are a single entity. We have also seen similar lesions in a few dogs of other breeds. Recently, we examined a 2-year-old castrated male Shih Tzu dog with clinical signs and lesions similar to those of necrotizing meningoencephalitis of Pug dogs. The brain of this dog was negative for CDV using immu-

6 Vet Pathol 32:3, 1995 Meningoencephalitis of Maltese Dogs 235 nohistochemical methods. A nonsuppurative necrotizing encephalitis in Yorkshire terriers has recently been reported.6 In those dogs, although the distribution Of lesions is different from that Seen in Pug and Maltese dogs, the lesions themselves are similar. Evidence of CDV and canine herpesvirus infection was also not identified in the Yorkshire terriers. Instances of comparable lesions in other breeds suggest that the entity ofnecrotizing meningoencephalitis may be more widespread among dogs than previously thought. Acknowledgements Clinical diagnostic testing, including computer assisted tomography and magnetic resonance imaging studies, were supported by the Charing Cross Research Foundation and Estherena and Donald Schotland. We thank Ms. Juli Burns and Mr. James Hayden for valuable technical assistance. References Cordy DR, Holliday TA: A necrotizing meningoencephalitis of pug dogs. Vet Pathol 26: , Davis RL, Robertson DM: Textbook of Neuropathology, 2nd ed. Williams and Wilfins, Baltimore, MD, delahunta A: Veterinary Neuroanatomy and Clinical Neurology, 2nd ed. WB Saunders Co., Philadelphia, PA, House JA, Greg DA, Lubroth J, Dubovi EJ, Torres A: Experimental equine herpesvirus- 1 infection in llamas (Lama,duma). J Vet Diam Invest 3: Orvell c, SheshberadaranH, Norrby E: Preparation and characterization of monoclonal antibodies directed against four structural components of canine distemper virus. J Gen Virol 66:443456, Tipold A, Fatzer R, Jaggy A, Zurbriggen A, Vandevelde M: Necrotizing encephalitis in Yorkshire temers. J Small Anim Pract 34: , 1993 Request reprints from Dr. T. J. Van Winkle, School of Veterinary Medicine, University of Pennsylvania, 3800 Spruce Street, Philadelphia, PA (USA).