• Open Access

Necrotizing Cerebellitis and Cerebellar Atrophy Caused by Neospora caninum Infection: Magnetic Resonance Imaging and Clinicopathologic Findings in Seven Dogs


Corresponding author: Laurent Garosi, DVM, Dip ECVN, MRCVS, Davies Veterinary Specialists, Manor Farm Business Park, Higham Gobion, Hertfordshire SG5 3HR, UK; e-mail: lsg@vetspecialists.co.uk.


Background: Adult dogs with neosporosis can develop a variety of neurologic signs. No area of predilection within the nervous system so far has been identified in adult dogs.

Objectives: To document neosporosis as a cause of progressive cerebellar ataxia and cerebellar atrophy in dogs.

Animals: Seven client-owned dogs.

Methods: Retrospective, descriptive study.

Results: Age at diagnosis ranged from 1 year 6 months to 9 years 11 months. Neuroanatomic localization indicated cerebellar and brainstem disease in 6 dogs and a central vestibular lesion in 1 dog. In all 7 dogs, there was moderate to marked bilaterally symmetrical cerebellar atrophy, with the atrophied cerebellum being surrounded by a region of T2-weighted hyperintense and T1-weighted hypointense signal. Cerebrospinal fluid (CSF) analysis in all but 1 dog showed mononuclear pleocytosis and high protein concentration. Polymerase chain reaction testing for Neospora caninum performed on the CSF was positive in 4/5 dogs tested and there was a high titer of serum antibodies to N. caninum (≥ 1 : 800) in all 6 dogs tested. Postmortem examination in 1 dog confirmed cerebellar atrophy and multifocal nonsuppurative encephalitis with areas of malacia and leptomeningitis. All of the remaining 6 dogs were treated with some combination of clindamycin, trimethoprim, sulfadiazine, and pyrimethamine. Two dogs were euthanized because of deterioration or relapse of neurologic signs, but treatment of the remaining 4 dogs resulted in improvement (3 dogs) or resolution (1 dog) of neurologic signs.

Conclusions and Clinical importance: Neosporosis is an important cause of progressive cerebellar ataxia and cerebellar atrophy in adult dogs.


magnetic resonance imaging


polymerase chain reaction


cerebrospinal fluid


fluid-attenuated inversion recovery

Neosporosis is a polysystemic protozoal disease first described in a litter of Boxer dogs in Norway in 1984.1 Naturally occurring infection with Neospora caninum is reported in many species, including cattle, sheep, goats, horses, dogs, and foxes, with experimentally induced infection being reported in other species.2,3 Clinical disease does not occur in humans,2 but serological evidence of human exposure to N. caninum exists.4

Dogs serve as both intermediate and definitive hosts of N. caninum.5–7 Animals (including dogs) with clinical neosporosis act as intermediate hosts, in which replication occurs by asexual reproduction to produce tachyzoites and cyst-forming tissue bradyzoites.3 Natural transmission of N. caninum in dogs is not fully understood. Transplacental infection is well recognized1,8–10 and is believed to be the most common route of transmission, but postnatal infection must also occur to maintain infection at the seroprevalence rates reported in canine populations.5,11 Moreover, higher seroprevalences have been documented in older than in younger dogs.12 Subclinically infected bitches can transmit the parasite to their offspring, and repeated transplacental infection of successive litters can occur despite isolation.11,13

Neosporosis can cause fatal disease in dogs of all ages, but puppies usually are affected more severely than older dogs, with most cases reported in dogs <6 months of age.3,5,8,10,14 Inflammation of the muscles and nerve roots of the hindlimbs of affected puppies causes progressive paraparesis, muscle atrophy, and loss of patellar reflexes.14 These signs often progress to pelvic limb paralysis with characteristic rigid pelvic limb hyperextension because of a combination of radiculitis and myositis, the latter causing fibrous contracture of the quadriceps and gracilis muscles.5,14,15,16 Puppies also may develop pneumonia, hepatitis, myonecrosis, encephalitis, and myocarditis (which can lead to sudden death).2,3,10,14,17 Adult animals with neosporosis can develop a variety of neurologic signs, most commonly a multifocal central nervous system (CNS) syndrome, including paresis, paralysis, ataxia, head tilt, and seizures.5,10,15,18 Myositis, myocarditis, hepatitis, dermatitis, ocular lesions, interstitial pneumonia, and pancreatitis also have been reported in adult dogs.15

In this report, we describe meningoencephalitis predominantly affecting the cerebellum in 7 dogs and characteristic magnetic resonance imaging (MRI) findings that should alert the clinician to the possibility of neosporosis.

Case Reports

Seven adult dogs with an age range of 1 year 6 months to 9 years 11 months were presented for evaluation of progressive generalized ataxia. The duration of neurologic signs before presentation ranged from 5 weeks to 12 months.

Dog 1, a 5-year-old male neutered West Highland White Terrier presented with a 3-month history of progressive ataxia. Neurologic examination revealed depressed mental status, ataxia affecting all 4 limbs and an intention tremor. Conscious proprioception was impaired in the left pelvic limb and hopping responses were exaggerated in all 4 limbs. Cranial nerve examination revealed bilaterally decreased oculocephalic reflexes and decreased nasal sensation on the right side. The neurologic examination suggested a multifocal CNS disorder (mainly cerebellar and brainstem).

Dog 2, a 9-year-old female neutered West Highland White Terrier was evaluated for a 12-month history of an initially slowly progressive ataxia which had become more rapidly progressive over the previous 10 days, during which time a head tilt also had become apparent. Neurologic examination revealed normal mental status, a mild right-sided head tilt and moderate ataxia affecting all 4 limbs. Postural reactions were abnormal on the left side and the remainder of the neurologic examination was normal. The neuroanatomic localization was a left central vestibular lesion with paradoxical head tilt.

Dog 3, a 7-year-old female neutered Greyhound presented with a 2-month history of progressive ataxia in all 4 limbs, with a rapid deterioration during the preceding 5 days. Neurologic examination revealed normal mental status and marked cerebellar ataxia, with hypermetria affecting all 4 limbs, but more severe on the left side. Postural reactions were slightly impaired in the right thoracic and pelvic limbs, and mild neck discomfort was detected on dorsoventral flexion and lateral flexion. Cranial nerve examination revealed an absent menace response in the left eye with normal palpebral reflex and visual placing. Multifocal neurolocalization involving mainly the cerebellum (likely left-sided) and brainstem (likely right-sided) was suspected.

Dog 4, a 10-year-old male neutered Dachshund, was evaluated for a 5-week history of ataxia, initially affecting just the pelvic limbs but later also the thoracic limbs. Neurologic examination revealed depressed mental status, tetraparesis and head tremor. Postural reactions were decreased in all 4 limbs, particularly on the left side. Segmental spinal reflexes were normal in the thoracic limbs but the patellar reflexes were decreased bilaterally and the withdrawal reflex was weak in the left pelvic limb; the perineal reflex also was decreased on the left side. Cranial nerve evaluation was normal aside from anisocoria, with the right pupil smaller than the left, and decreased pupillary light reflex in the left eye. In addition, facial sensation was decreased on the right, bilateral temporal muscle atrophy was present, and spinal hyperesthesia was detected on palpation of the neck. A multifocal neuroanatomic localization involving the cerebellum, brainstem, and L4-S3 spinal segments was suspected.

Dog 5, a 4-year-old female neutered Labrador Retriever, presented with a 6-month history of progressive ataxia affecting all 4 limbs. Neurologic examination revealed marked cerebellar ataxia, which affected the pelvic limbs more severely. Conscious proprioception was abnormal in both pelvic limbs and in the right thoracic limb; hopping was abnormal in all 4 limbs. Segmental spinal reflexes and cranial nerve evaluation were normal. Neuroanatomic localization was consistent with a multifocal lesion (cerebellum and brainstem).

Dog 6, a 9-year-old male neutered Labrador Retriever, presented with a 6-week history of pelvic limb ataxia that had recently progressed to involve the thoracic limbs. Episodes of staring vacantly at walls also were reported. Neurologic examination showed marked cerebellar ataxia, affecting the pelvic limbs more severely. Conscious proprioception was slightly impaired in the thoracic limbs and markedly impaired in the pelvic limbs; hopping responses were abnormal in all 4 limbs. Segmental spinal reflexes were intact in all 4 limbs, but palpebral reflexes and menace responses were impaired bilaterally, and jaw strength was decreased. Atrophy of the masticatory muscles and bilateral facial paresis also were observed. These findings indicated a lesion affecting the caudal brainstem and cerebellum.

Dog 7 was an 18-month-old female Labrador Retriever presented for slowly progressive pelvic limb ataxia over the preceding 7 months. Neurologic examination revealed marked truncal and limb ataxia, with a jerky pelvic limb gait. Postural responses were impaired in the pelvic limbs, particularly on the right side, and all spinal reflexes were intact. All cranial nerve reflexes and responses were normal, but head and ocular tremors were noted. These findings indicated a lesion affecting the brainstem and cerebellum.

Based on neuroanatomic localization suggesting cerebellar and brainstem lesions in most dogs, the main differential diagnoses included a late onset multisystem neuronal degeneration or neuronal storage disease, inflammatory or infectious CNS disease, primary or metastatic brain tumor, or caudal fossa cyst.

Materials and Methods


All dogs underwent MRI of the brain. Dog 4 additionally had MRI of the caudal lumbar spine. Several different anesthetic protocols were used for MRI, but the majority of dogs were induced with propofol administered IV to effect and maintained on isoflurane gas and oxygen. Images were acquired on MRI units at different centers (1.5 T Signa MRI, General Electric Medical Systems, Milwaukee, WI [dogs 1 and 3]; 0.4 T Aperto MRI, Hitachi, Tokyo, Japan [dogs 2 and 5]; 1.5 T Gyroscan ACS-NT, Philips Medical System, Eindhoven, the Netherlands [dog 4]; 1 T Siemens Magnetom Harmony, Erlangen, Germany [dog 6]; and, 0.2 T Esaote VetMR scanner, Genova, Italy [dog 7]). Images included transverse and sagittal T2-weighted, transverse T1-weighted, transverse T1-weighted with contrast, and transverse fluid-attenuated inversion recovery (FLAIR) sequences. Additional sequences were obtained as requested by the attending neurologist and radiologist. All images were interpreted by a board-certified veterinary radiologist, a board-certified veterinary neurologist or both.

Additional Diagnostic Tests

Cerebrospinal fluid (CSF) was collected from all 7 dogs by cerebellomedullary cisternal puncture during general anesthesia. Samples were submitted for white and red blood cell counts, protein concentrations, cytology in all dogs, and polymerase chain reaction (PCR) testing for N. caninum and Toxoplasma gondii in 5 dogs. In addition, blood samples were evaluated for serum antibody titers to N. caninum by an indirect fluorescence antibody test in 6 of 7 dogs. Dog 7 had a blood smear examined with periodic acid-Schiff staining for the presence of cells with inclusions that might indicate lysosomal storage disease; this dog was not tested for serum antibodies against N. caninum nor was a PCR test for N. caninum or T. gondii done on CSF. Dog 7 was euthanized 2 weeks after initial presentation and underwent a postmortem examination.


Information regarding outcome was obtained by at least 1 reexamination in 4 dogs, by telephone conversations with the owner and the referring veterinary surgeon in 1 dog, and by telephone conversation with and a letter from the referring veterinary surgeon in the remaining dog.



MRI showed moderate (4 dogs) to marked (3 dogs) bilaterally symmetrical cerebellar atrophy (Fig 1). On FLAIR and T2-weighted sequences, the contrast between cerebellar grey and white matter was lost in 3 dogs. In addition, 3 dogs (dogs 2, 3, and 4) showed FLAIR hyperintensities compared with CSF signal within the cerebellar white matter. In all dogs, the atrophied cerebellum was surrounded by a thick T2-weighted hyperintense and T1-weighted hypointense signal, making the cerebellar sulci prominent. In 6 dogs, the region surrounding the atrophied cerebellum remained hyperintense on FLAIR sequences (Fig 2) but in dog 7, there was suppression of the T2-weighted signal on FLAIR sequences. Contrast enhancement of the affected meninges only was detected in 2 dogs (dogs 1 and 6), which also demonstrated contrast enhancement affecting the meninges surrounding the brainstem (Fig 3). In addition to cerebellar atrophy, other lesions were detected in 3 dogs. Three dogs had mild (2 dogs) or marked (1 dog) heterogenous T2-weighted hyperintensities and contrast enhancement of the temporalis and masseter muscles, which appeared moderately atrophied (dogs 4, 5, and 6) (Fig 3). One dog had bilaterally symmetrical T2-weighted and FLAIR hyperintensities in the corona radiata of the occipital lobes (dog 2) (Fig 4). MRI scan of the lumbosacral spine in dog 4 was normal.

Figure 1.

 Sagittal and dorsal T2-weighted images (0.4 T) of 9-year-old West Highland White Terrier. There is marked bilaterally symmetrical cerebellar atrophy and loss of contrast between cerebellar grey and white matter.

Figure 2.

 Transverse T2-weighted (A), T1-weighted (B), and FLAIR (C) images (0.4 T) of the caudal fossa of 9-year-old West Highland White Terrier. The diffusely atrophied cerebellum is surrounded by a thick T2-weighted hyperintense and T1-weighted hypointense signal that only partially suppresses on FLAIR sequences (arrows). Additionally, there is bilaterally symmetrical hyperintensity within the cerebellar white matter on the FLAIR sequence (arrow heads).

Figure 3.

 Transverse (A) and dorsal (B) T1-weighted images (0.4 T) after paramagnetic contrast injection of 9-year-old Labrador. There is mild diffuse atrophy of the cerebellum and contrast enhancement of the meninges surrounding the cerebellum and brainstem (arrows). Additionally, there is moderate atrophy and marked contrast enhancement of the temporalis and masseter muscles (stars).

Figure 4.

 Dorsal FLAIR images (0.4 T) of the parieto-occipital lobe of 9-year-old West Highland White Terrier. There is bilaterally symmetrical hyperintensity within the corona radiata of the occipital lobes.

Additional Diagnostic Tests Results

CSF analysis was performed in all 7 dogs. One dog had normal CSF (dog 5), whereas all of the other 6 dogs had mild to marked mononuclear pleocytosis (12–300 nucleated cells/μL; reference range, <5 nucleated cells/μL). Moderately or markedly high protein concentrations were present in 4 of these dogs (0.6–2.9 g/L; reference range, <0.3 g/L). No protozoal organism was detected in the CSF by light microscopy. PCR testing of CSF for N. caninum was positive in 4 dogs (dogs 1, 3, 4 and 5) of 5 dogs tested (negative in dog 2). PCR testing for T. gondii was negative in all dogs tested.

The presence of high titers of serum antibodies (≥1 : 800) for N. caninum suggested infection with this protozoal agent in 6 dogs tested.

Serum creatine kinase activity was substantially increased in 2 out of 3 dogs that had heterogenous T2-weighted hyperintensities, marked contrast enhancement, and moderate atrophy of the temporalis and masseter muscles.


Dog 7 was treated with prednisolone (1 mg/kg PO q24h for 5 days then 0.5 mg/kg PO q24h) and continued to deteriorate over the next 2 weeks, at which point it was euthanized. The other 6 dogs that had high titers of serum antibodies for N. caninum were treated with either a combination of clindamycin and trimethoprim/sulfadiazine PO at the recommended dosage for 6–10 weeks (dogs 1–5) or trimethoprim/sulfadiazine and pyrimethamine for 3.5 months (dog 6). Dog 6 was supplemented with folinic acid until the end of the 3.5 months of pyrimethamine administration; afterward it continued to receive trimethoprim/sulfadiazine. At the time of writing, 2 additional dogs (dogs 2 and 3) had been euthanized at 3 and 7 months after presentation because of deterioration or relapse of neurological signs. Postmortem evaluation was declined in these 2 dogs. Treatment of the remaining 4 dogs had resulted in improvement of neurologic signs in 3 of them (dogs 1, 4, and 6), but clinical signs did not fully resolve over the follow-up period of between 7 and 14 months. Dog 1 underwent repeated MRI of the brain and CSF analysis 6 weeks after starting treatment. Although the degree of meningeal enhancement was considered to be decreased, the cerebellar changes remained identical to that observed on the initial MRI. Repeated CSF analysis was normal and repeated PCR testing for N. caninum on CSF was negative. Dog 4 remained mildly ataxic and tetraparetic on re-examination 8 weeks after starting clindamycin and trimethoprim/sulfadiazine treatment. In dog 6, hypermetria resolved and behavior normalized over a period of 5 months. Bilateral facial paresis persisted in this dog, but masticatory muscle mass increased and head conformation normalized. Dog 5 was clinically normal on reexamination 9 weeks after initiation of clindamycin and trimethoprim/sulfadiazine. Repeated CSF analyses were normal and repeated N. caninum PCR on CSF was negative. Antiprotozoal treatment was stopped and the dog remained normal in the 6-month follow-up period.

Postmortem examination of dog 7 revealed a small cerebellum relative to the size of the cerebrum; the cerebellar gyri were small and the sulci dark red. Microscopically, multiple foci of nonsuppurative inflammation and malacia were scattered throughout the cerebellum, with thickening and mononuclear cell infiltration of the meninges. Lesions were seen predominantly in the superficial (molecular) layer of the cerebellum, with malacia being most severe toward the tips of the cerebellar gyri (Fig 5). Lesions were also present in the grey matter of the cerebellar nuclei, especially near the 4th ventricle. Protozoal tissue cysts were observed near some of the lesions and were identified as intensely staining structures with relatively thin capsules. They were most common in the molecular layer of the cerebellum (Fig 5C–F). Similar lesions were present in other regions of the brain, including the medulla, thalamus, and cerebral cortex. In the forebrain, lesions were seen most commonly in superficial grey matter, often within 2 mm of the pia mater. The tissue adjacent to the subventricular zone also was severely affected, but the corpus callosum and other white matter structures appeared relatively spared. In the caudal medulla, white matter was severely affected, but grey matter contained less extensive lesions. Tissue immunostaining with indirect fluorescent antibody test using a rabbit polyclonal antibody specific to Neospora was positive.

Figure 5.

 Histopathology of case 7, an 18-month-old Labrador Retriever. (A and B) There is widespread malacia of the molecular layer (ML) and depletion of the granular cell layer (GL) of the cerebellum, with thickening and cellular infiltration of the overlying meninges (*). (C) Photomicrograph showing 3 protozoal tissue cysts in the molecular layer of the cerebellum, the predominant site at which these were found. (D) High power image of the boxed region indicated in C, showing 2 of the cysts (arrows). (E) Protozoa occasionally were seen in the cerebellar white matter (WM). (F) High power images of the boxed region indicated in E, highlighting the tissue cysts (arrow). Hematoxylin and Eosin.


These case reports document a syndrome of cerebellar dysfunction associated with N. caninum infection in adult dogs. Affected animals demonstrated progressive cerebellar ataxia and had neuroanatomic localization indicative of cerebellar and brainstem lesions. MRI identified a bilaterally and symmetrically atrophied cerebellum, with thickened overlying meninges, and diagnostic tests confirmed exposure to N. caninum. Adult dogs in 3 previous case reports had cerebellar dysfunction associated with N. canimum infection,16,19,20 as did 3 of the dogs described by Barber and Trees,11 which had head tremors, hypermetria and evidence of brainstem dysfunction.14 These studies together suggest that cerebellar dysfunction and atrophy is more common than previously realized with N. caninum infection in adult dogs.

Neosporosis can affect the nervous system at any age, but for unknown reasons, it has a predilection for the lumbosacral spinal nerve roots of young dogs. This predilection causes an ascending paralysis of the limbs, which often leads to arthrogryposis.10,14,21,22 Older dogs, which appear to be less commonly affected, typically have histopathologic evidence of multifocal CNS involvement, polymyositis, or both.14 In contrast to young dogs, no area of predilection has been identified so far in adult dogs with neosporosis, and neurologic signs of encephalomyelitis are considered to vary depending on the distribution of N. caninum within the CNS and other tissues.22 To the authors' knowledge, there are only 4 other individual reports of cerebellar necrosis in association with N. caninum infection16,19,20,23 and 3 of those reports describe adult dogs.16,19,20

The first of these 3 reports describes a 3.5-year-old male Golden Retriever with a 28-month history of slowly progressive ataxia, hypermetria and head tremor.19 Neuroanatomic diagnosis indicated diffuse cerebellar disease, but this dog also had substantial bilateral atrophy of masticatory muscles and mild bilateral atrophy of scapular and gluteal muscles. No imaging study of the brain was performed, and the dog was euthanized without treatment, because of the severity of signs and guarded prognosis. Postmortem examination revealed marked symmetrical atrophy of the cerebellum with its surface covered by a thick grey membrane obscuring the atrophied folia. Microscopic examination of the CNS revealed severe atrophy of most of the vermal and hemispheric folia, nonsuppurative inflammation of the leptomeninges and remaining parenchyma, and a few small cysts of protozoal organisms that were identified as N. caninum by immunohistochemistry.

The 2nd report describes a 4-year-old male Alaskan Malamute with a 2-month history of slowly progressive pelvic limb weakness.16 The neuroanatomic diagnosis was a multifocal neurological disease (cerebellar and brainstem) and histopathologic evaluation showed cerebellar atrophy associated with N. caninum. The organism also was present in other parts of the nervous system and muscle.

The 3rd report described a 14-year-old Labrador Retriever with head tremor, marked generalized ataxia and abnormal postural reactions on all 4 limbs. Histopathologic evaluation revealed lesions restricted to the cerebellar leptomeninges and cortex, which appeared necrotic and atrophic. N. caninum tachyzoites were identified by immunohistochemistry.20

In the present report, 3/7 affected dogs were Labrador Retrievers and 2/7 were West Highland White Terriers, perhaps representing the fact that both breeds are popular in the United Kingdom. However, all 3 of the dogs with cerebellar dysfunction described by Barber and Trees11 were adult Labrador Retrievers, as was 1 of the 3 other adult N. caninum cases described previously.20 In total, 7/13 documented cases of cerebellar dysfunction associated with N. caninum infection in adult dogs have occurred in this breed, which raises the possibility of a breed predisposition. This predisposition might occur through predilection to infection, or through a predisposition to cerebellar dysfunction associated with infection. In an earlier study, 8/27 dogs of all ages affected with N. canimum were Labrador Retrievers and 7 were Boxers, suggesting but not confirming a breed predisposition.14 This study, and another by the same authors, also measured antibody titers against N. caninum in dams and littermates of affected dogs, but it was not possible to determine whether Labradors were more likely to have relatives that were infected, as compared with other breeds.11,14 These studies also highlighted that not all littermates or offspring of seropositive individuals are themselves seropositive, and of those that are, not all develop clinical signs of disease. This observation indicates that neosporosis in adult dogs might arise either through a new infection or through reactivation of a latent infection acquired years previously. In the current study, the dam and littermates of the affected dogs were not available for examination, and one can only speculate about whether the disease represented recently acquired infection or reactivated quiescent infection during the course of another disease process. Although the possibility of N. caninum detection being incidental in the face of another disease process causing similar CSF changes cannot be excluded, the therapeutic response observed with antiprotozoal treatment suggests that neosporosis accounted for the neurologic signs.

The reason for the tendency of N. caninum to cause cerebellar dysfunction currently is not clear. Barber and Trees11 failed to identify a predilection for specific sites within the neuraxis, although only 6 dogs were examined and only 1 was an adult.22 Among the adult cases of cerebellar disease, Jackson and colleagues found numerous cerebellar protozoa, whereas Cantile found relatively few, despite severe malacia of the cerebellar molecular layer.19,20 These findings suggest that tissue response does not necessarily correlate with parasite burden and raises the possibility of an immunological component to the condition. Confirmation would require more extensive histopathologic study of a large number of affected animals. The present study also is interesting because of the large variation in duration of signs. Although 2 dogs reportedly had developed signs over ≤6 weeks, the other 5 dogs had progression of clinical signs over 2, 3, 6, 7, and 12 months before presentation. A previously reported case of neosporosis with progressive cerebellar signs which underwent necropsy had shown progression of clinical signs over 28 months.19 We hypothesize that the variation in duration and rate of progression of clinical signs may reflect differences in individual immune responses, an effect that may be influenced by age, route of infection, or strain of N. caninum.

In the dogs of this report, MRI data confirmed the neuroanatomic localization of the lesions (mostly cerebellar and brainstem). The presence of bilaterally symmetrical cerebellar atrophy and loss of contrast between cerebellar grey and white matter on MRI was a common finding in these 7 dogs. Subjectively, dogs with longest duration of signs before MRI had most marked cerebellar atrophy. In addition, the cerebellum was surrounded by a prominent layer, which was T2-hyperintense and T1-hypointense. In 6 dogs, this signal was considered to be due to thickened meninges as opposed to CSF based on the absence of signal suppression using FLAIR sequences. Although postmortem examinations were not available in these 6 dogs, these findings closely resembled the cerebellar lesions seen on histopathology in a previously reported case of cerebellar neosporosis in adult dogs.19 In this case, the authors described a cerebellum that was markedly decreased in size and covered by a thick grey membrane. Microscopic examination revealed multifocal nonsuppurative encephalitis, affecting mainly the grey matter, leptomeninges and subarachnoid space. Neuronal necrosis, nodular microgliosis and perivascular cuffing were particularly severe in the cerebellum, where extensive necrosis affected both cerebellar hemispheres, and other parts of the brain were less severely affected. Several tissue cysts (bradyzoites) were observed mostly in the grey matter of the cerebrum and in both the grey and white matter of the cerebellum, and were confirmed as N. caninum by immunohistochemistry.19 The only dog in the present study to undergo postmortem examination had similar histopathologic changes. Interestingly, we observed a predominance of lesions in subpial and periventricular regions of the neuraxis, similar to that described by Cantile and Arispici.20 Also in our study, grey matter was more severely affected in the brain but white matter was more affected in the cervical spinal cord, perhaps also reflecting the subpial distribution.

Contrast enhancement of the affected cerebellar meninges was only detected in 2 dogs, which also had demonstrated contrast enhancement of the meninges surrounding the brainstem. Both of these dogs (dogs 4 and 6) had the shortest duration of clinical signs before presentation (5 and 6 weeks, respectively) and relatively mild cerebellar atrophy. Contrast enhancement was absent in the remaining 4 dogs, which had a longer duration of signs before MRI (2–12 months). The chronicity of the lesions in these dogs may represent a less aggressive inflammatory response, which in turn may account for the lack of contrast enhancement.

MRI lesions also were detected in the masticatory muscles in 3 dogs and in the corona radiata of the occipital lobe in 1 dog. No histopathology was available to confirm whether these lesions were related to neosporosis or another disease process. Distribution of N. caninum within the nervous system and other tissues previously was described in 6 young dogs with clinical neosporosis.22 Although found most consistently in the cerebrum, parasites were distributed throughout the CNS, both in grey and white matter and in nerve roots.22 Tachyzoites also were commonly found in striated muscle, particularly in the masseter and temporalis muscles, and the heterogenous T2-weighted hyperintensities, marked contrast enhancement and moderate atrophy of the temporalis and masseter muscles that we observed in 4 dogs were considered strongly suggestive of myositis.24 Creatine kinase activity also was substantially increased in 2 of these 3 dogs, supporting the presence of myositis. That N. caninum was responsible for the myositis is further supported by the findings of Jackson et al,19 who reported myositis of the masticatory muscles in addition to necrotizing cerebellitis on histopathology.

The nature of the bilaterally symmetrical T2-weighted and FLAIR hyperintensities in the corona radiata of the occipital lobes observed in dog 2 is unknown. This location and distribution of lesions have not been reported previously in cases of neosporosis. Although there is no evidence to exclude neosporosis as the cause of these changes, bilaterally symmetrical brain lesions usually are more suggestive of metabolic, nutritional, or toxic pathologic processes. Unfortunately, repeated MRI was not performed on that dog to assess if these white matter changes resolved with the improvement in neurologic signs observed after antiprotozoal treatment. Postmortem examination was also not carried out to establish the exact nature of these white matter changes.

There are several reports of successful treatment of adult and pediatric neosporosis;11,14,25,26 however, the prognosis generally is considered poor. Current treatment protocols largely are extrapolated from recommendations for treatment of canine toxoplasmosis, in turn largely taken from studies in humans. Treatment is more likely to be successful in young animals if started early in the course of the disease, before muscle fibrosis develops; animals that develop pelvic limb hyperextension tend not to respond to treatment.12,14 Generally, treatment is aimed at controlling active inflammation and preventing further progression of clinical signs rather than achieving complete resolution. Clindamycin, trimethoprim/sulfonamide, and pyrimethamine, alone or in combination, have been used effectively to treat neosporosis.10,11,14,15,25 Experimental studies in dogs have shown that corticosteroids worsen clinical neosporosis and consequently they should be withheld.10 At present, there is no treatment effective against tissue cysts.2 In our study, there was no clear correlation between the degree of cerebellar atrophy or duration of signs before treatment and failure to respond to antiprotozoal treatment.

This case series shows that neosporosis should be considered in adult dogs with progressive cerebellar ataxia and MRI evidence of cerebellar atrophy especially in dogs with concomitant masticatory muscle involvement.