MAGNETIC RESONANCE IMAGING CHARACTERISTICS IN FOUR DOGS WITH CENTRAL NERVOUS SYSTEM NEOSPOROSIS

Authors


  • Funding sources: not applicable

  • Previous presentations or abstracts: not applicable

Abstract

Neosporosis is a polysystemic disease that can affect dogs of any age and can cause inflammation of the central nervous system. Antemortem diagnosis can be challenging, as clinical and conventional laboratory test findings are often nonspecific. A previous report described cerebellar lesions in brain MRI studies of seven dogs and proposed that these may be characteristic for central nervous system Neosporosis. The purpose of this retrospective study was to describe MRI characteristics in another group of dogs with confirmed central nervous system neosporosis and compare them with the previous report. The hospital's database was searched for dogs with confirmed central nervous system neosporosis and four observers recorded findings from each dog's MRI studies. A total of four dogs met inclusion criteria. Neurologic examination was indicative of a forebrain and cerebellar lesion in dog 2 and multifocal central nervous system disease in dogs 1, 3, and 4. Magnetic resonance imaging showed mild bilateral and symmetrical cerebellar atrophy in three of four dogs (dogs 2, 3, 4), intramedullary spinal cord changes in two dogs (dogs 3, 4) and a mesencephalic and metencephalic lesion in one dog (dog 2). Multifocal brain lesions were recognized in two dogs (dogs 1, 4) and were present in the thalamus, lentiform nucleus, centrum semiovale, internal capsule, brainstem and cortical gray matter of the frontal, parietal or temporal lobe. Findings indicated that central nervous system neosporosis may be characterized by multifocal MRI lesions as well as cerebellar involvement in dogs.

Introduction

Central nervous system infection by Neosporum caninum has been reported to cause ataxia, paresis, paralysis, head tilt, head tremors, and seizures in dogs.[1-4] Neospora caninum is an obligate intracellular protozoan parasite of the phylum Apicomplexa.[7] Exposure to Neospora has been reported in a variety of domestic and wild animals, however, viable organisms could only be isolated from a few hosts (cattle, sheep, water buffalo, dog, horse, bison, white-tailed deer).[8, 9] Neosporosis primarily affects cattle and dogs and is a major cause for abortions in cattle worldwide with associated economic losses.[8, 10] Dogs can serve as intermediate or definitive hosts and vertical as well as horizontal routes of transmission have been proposed.[8-10] Dogs of any age can be affected by this polysystemic disease, which can cause inflammation of the brain, spinal cord, meninges, nerve radices, muscles, myocardium, liver, lungs, skin, and pancreas.[1, 3], [4], [11-14] As nonsuppurative inflammation and malacia may be present throughout the central nervous system, multifocal lesions affecting the gray and white matter or cerebellar atrophy can be found histologically.[3, 6], [12-15] Antemortem diagnosis may be difficult, necessitating a variety of confirmatory test results such as increased antibody titres for Neospora caninum in the serum or cerebrospinal fluid, microscopic identification of the protozoal agent in the cerebrospinal fluid, detection of pathogen specific deoxyribonucleic acid from the cerebrospinal fluid or other organs or histological evidence of infectious stages of Neospora caninum in muscle biopsies.[3, 5], [13], [16-18]

To the authors’ knowledge, only one published study has described brain MRI characteristics of neosporosis in dogs. [6] In this study of seven dogs, the lesions were predominantly confined to the cerebellum. Neurologic examination in six of the seven dogs with neosporosis suggested cerebellar and brainstem localizations and one dog localized to the central vestibular system. Moderate to marked bilateral and symmetrical cerebellar atrophy was described as a characteristic feature in all seven dogs. This was confirmed with histopathology in one dog. Loss of cerebellar gray and white matter distinction, contrast enhancement of affected meninges and T2-weighted hyperintensity and contrast enhancement of the temporalis and masseter muscles was reported in some of the patients. Bilateral symmetrical hyperintensity on T2-weighted and FLAIR images was seen in the corona radiata of the occipital lobe in one dog.[6] Similar MR imaging findings are also briefly mentioned in a case report where Neospora caninum tachyzoites are detected in cerebrospinal fluid and in a single dog involved in a study about cerebrospinal fluid eosinophilia.[2, 19] The dog with Neospora tachyzoites in the cerebrospinal fluid showed neurological abnormalities consistent with a diffuse or multifocal central nervous system disease with suspected multifocal forebrain and brainstem involvement. MR imaging showed leptomeningeal thickening and contrast enhancement as well as a cortical–subcortical parietotemporal lesion.[2] The neurologic examination in the dog with cerebrospinal fluid eosinophilia due to Neosporosis was not specified and MR imaging showed a T2-weighted hyperintensity of the cerebellar vermis and one cerebellar hemisphere with contrast enhancement on T1-weighted images on the consecutive cerebellar side.[19]

The purpose of this retrospective study was to describe MRI, clinical and outcome characteristics in another group of dogs with confirmed central nervous system Neosporosis and compare these with previously published characteristics.

Material and Methods

The clinical data base of the Queen Mother Hospital for Small Animals, Royal Veterinary College, University of London was searched for dogs diagnosed with Neosporosis. The study was approved by and conducted according to the guidelines of the Institutional Ethic Committee. Only dogs with confirmed central nervous system Neosporosis and MRI lesions in the central nervous system that were attributed to this infection were included. A confirmed diagnosis of central nervous system Neospora caninum infection was based on the presence of deoxyribonucleic acid of this protozoal agent via polymerase chain reaction in the cerebrospinal fluid and increased antibody titres on serology. Included dogs also had to have available results from a complete blood count, serum biochemistry and cerebrospinal fluid analysis. Signalment, history, general physical and neurological examination findings, laboratory and MR imaging results, treatment and outcome data were recorded for each included dog. All MR images were reviewed by a board certified radiologist, two board certified neurologists and one neurologist in training. Imaging results were recorded based on consensus opinion of all four observers.

Results

Four dogs met the inclusion criteria. All dogs had undergone general physical and neurological examination by a board certified neurologist and resident in training between 2011 and 2013. Magnetic resonance imaging of the central nervous system had been acquired using the same 1.5 T magnet (Gyroscan NT intera, Philips, Eindhoven, Netherlands). Sagittal and transverse T2-weighted, transverse fluid-attenuated inversion-recovery (FLAIR) and T1-weighted pre- and post contrast MR images were obtained for all brain scans. Additional sequences acquired in some of the dogs included a T1-weighted inversion recovery (IR) and a T2-weighted fast field echo (T2*). Magnetic resonance imaging of the spine included sagittal and transverse T2-weighted and T1-weighted pre- and post contrast images (Appendix).

Dog 1, a 9-year, 5-month-old, female neutered Greyhound presented with a 2-month history of progressive tetraparesis and intermittent knuckling of the right thoracic limb. The patient lived in a suburban area with another healthy dog, with exposure to red foxes but no access to farms. Serology from the referring veterinary surgeon showed high antibody titres (>1600) for Neospora caninum. General physical examination was unremarkable. The neurological examination findings were mild obtundation, tetraparesis, and generalized ataxia, proprioceptive deficits in all four limbs with the right side being more affected than the left, leaning and falling forwards and to the right side, decreased lateral visual field, and positional ventromedial strabismus in the right eye. A multifocal brain disease with a left prosencephalic component was suspected. A complete blood count was unremarkable and serum biochemistry showed markedly elevated creatine kinase (3033 U/l, reference range: 61–394 U/l) and mildly elevated total bilirubin (4.0 umol/l, reference range: 0–2.4 umol/l) and alanine aminotransferase (317 U/l, reference range: 13–88 U/l). Magnetic resonance imaging of the brain showed multifocal intra-axial, poorly demarcated lesions in the thalamus, internal capsule, centrum semiovale, and in the gray matter of the parietal and temporal lobes. The lesions were hyperintense on T2-weighted and iso-hypointense on T1-weighted and on IR images with mild peripheral enhancement post gadolinium (Gadovist, Berkshire, United Kingdom) administration. The T2-hyperintensity was only suppressed in the most central portions of the lesions on FLAIR images indicating perilesional edema (Fig. 1). Cisternal cerebrospinal fluid analysis showed macrophagic inflammation with a normal total nucleated cell count and normal protein concentration. Polymerase chain reaction testing of cerebrospinal fluid for Neospora caninum was positive and negative for Canine distemper virus and Toxoplasma gondii. Treatment was initiated with clindamycin (Antirobe, Kent, United Kingdom) 20 mg/kg and trimethoprim-sulfonamide (Trimacare, York, United Kingdom) 15 mg/kg twice daily for 6 weeks. The patient was re-examined 6 weeks, 5 and 8 months after diagnosis and showed continuous neurologic improvement without complete resolution of clinical signs. Clindamycin is being continued at the same dose and the patient is alive at the time of writing the manuscript. Dog 2, a 2-year, 7-month-old male entire Labrador presented with a 1.5-year history of progressive and compulsive circling to the right side. The patient was the only animal in the household and was obtained from a farm. No information was available about the littermates. General physical examination was unremarkable. The neurological examination findings were compulsive circling to the right side, a subtle right-sided head tilt, and mildly hypermetric gait in the thoracic limbs. A multifocal (forebrain lesion with cerebellar involvement) brain lesion was suspected. A complete blood count, urine analysis, and a bile acid stimulation test was unremarkable and serum biochemistry revealed moderately elevated creatine kinase (1507 U/l, reference range: 61–394 U/l) and mildly elevated alanine aminotransferase (159 U/l, reference range: 13–88). Magnetic resonance imaging of the brain showed mild bilateral, symmetric cerebellar atrophy with widening of the dorsal folia and an extensive intra-axial, poorly demarcated mesencephalic and metencephalic midline lesion extending from the ventral aspect of the thalamus to the level of the pyramids (Fig. 2). The lesion was hyperintense on T2-weighted and T2* images and iso-hypointense on T1-weighted and IR with little obvious contrast uptake after gadolinium administration. FLAIR images indicated perilesional edema as only the central portion of the T2-hyperintensity was nulled (Fig. 3). Cisternal cerebrospinal fluid analysis revealed a mixed cell pleocytosis (total nucleated cell count: 48 mm3, reference range: <5 mm3) with mildly elevated protein concentration (0.34 g/l, reference range: <0.25 g/l). Polymerase chain reaction testing of cerebrospinal fluid was positive for Neospora caninum and serology showed increased antibody titres (>1600) using an immunofluorescence antibody assay. Toxoplasma gondii serology was negative. Treatment was initiated with oral clindamycin 20 mg/kg twice daily, trimethoprim-sulfonamide 15 mg/kg twice daily and pyrimethamine (Daraprim, Bad Oldesloe, Germany) 0.9 mg/kg once daily and was unfortunately discontinued by the owner after 3, 8, and 4 weeks, respectively. The patient represented 4 months after diagnosis (2 months after stopping oral medication) because of worsening of clinical signs. Neurological examination was similar to the previous examination but revealed more marked obtundation. Clindamycin and trimethoprim-sulfonamide were restarted at the same dose and the patient was still under therapy while writing the manuscript.

Figure 1.

Dog 1. Four transverse MR images at the level of the thalamus displaying an intra-axial, poorly delineated thalamic lesion which extends into the internal capsule and causes slight compression of the left lateral ventricle. The lesion is mainly hyperintense on T2 (A) and iso-hypointense on T1-weighted images (C) with incomplete peripheral rim enhancement post gadolinium administration (D). Only the most central portion of the T2-hyperintensity is suppressed on FLAIR indicating perilesional edema (B).

Figure 2.

Dog 2. Mid-sagittal T2-weighted MR image of the brain showing mild cerebellar atrophy and an intra-axial, hyperintense lesion with ill-defined margins affecting the mesencephalon and metencephalon.

Figure 3.

Dog 2. Four transverse MR images at the level of the midbrain reveal an intra-axial and diffuse lesion ventral to the mesencephalic aquaeduct involving the entire mesencephalic tegmentum. The lesion appears hyperintense on T2 (A) and iso-hypointense on T1-weighted images (C) with little obvious contrast enhancement post gadolinium administration (D). The FLAIR image indicates perilesional edema as only the central portion of the T2-hyperintensity is nulled (B) .

Dog 3, a 5-month-old male entire Labrador presented with a 2-week history of pelvic limb ataxia which progressed to nonambulatory tetraparesis, head tremors, and incontinence. The patient lived with another unrelated dog in a suburban area. No information exists about the littermates. Physical examination was considered unremarkable. Neurological examination showed thoracic limb paresis and pelvic limb plegia with decreased to absent proprioception in all four limbs. Patellar reflexes were absent and flexor-withdrawal reflexes were reduced in both pelvic and the left thoracic limb. Reduced muscle tone was evident in both pelvic and increased muscle tone in the left thoracic limb. The patient also displayed mild intention tremors. A multifocal central nervous system disease mainly affecting C1-T2 and L4-S3 spinal cord segments as well as the cerebellum was suspected. A complete blood count was unremarkable and serum biochemistry revealed moderately elevated creatine kinase (897 U/l, reference range: 61–394 U/l) and mildly elevated alanine aminotransferase (95 U/l, reference range: 13–88 U/l). MR imaging of the entire central nervous system demonstrated mild bilateral and symmetrical cerebellar atrophy with widening of the dorsal folia (Fig. 4) and multifocal, poorly delineated, intramedullary spinal cord lesions at the level of C5 and C6 vertebral bodies affecting both ventral horns (Fig. 5) as well as the left lateral and ventral funiculi. These spinal lesions were hyperintense on T2-weighted and isointense on T1-weighted images with no apparent enhancement following contrast administration. Cisternal and lumbar cerebrospinal fluid analysis revealed a predominantly neutrophilic but mixed pleocytosis (total nucleated cell count cisternal: 137 mm3, reference range: <5 mm3; total nucleated cell count lumbar: 168 mm3, reference range: <10 mm3) with markedly elevated protein concentrations (cisternal: 0.94 g/l, reference range: <0.25 g/l; lumbar: 1.73 g/l, reference range: <0.5 g/l). Polymerase chain reaction testing of cerebrospinal fluid for Neospora caninum was positive and serology showed high antibody titers (>1600) using an immunofluorescence antibody assay. Polymerase chain reaction testing of cerebrospinal fluid for Toxoplasma gondii and serum antibodies were negative. Polymerase chain reaction testing of cerebrospinal fluid and urine for canine distemper virus was negative whereas serology was positive at 1 in 20 by enzyme-linked immunosorbent assay. No aerobic or anaerobic bacterial isolates could be detected in the urine or cerebrospinal fluid. Treatment was initiated with clindamycin 20 mg/kg and trimethoprim-sulfonamide 15 mg/kg twice daily. The patient was hospitalized for a total course of 5 months for ongoing medical treatment and physio- and hydrotherapy with intermittent periods at home with his owners. The patient showed slow but continuous neurological improvement within this period of time. He regained good movement in both his pelvic and the right thoracic limb and was ambulatory with little harness support. His left thoracic limb remained contracted. Clindamycin was administered for a total of 4 months while trimethoprim-sulfonamide was continued. The patient unfortunately was lost to follow up after his hospitalization.

Figure 4.

Dog 3. Sagittal T2 weighted MR image showing cerebellar atrophy (arrow).

Figure 5.

Dog 3. Three transverse MR images at the level of C5 vertebral body showing intramedullary changes mainly affecting both ventral horns which are hyperintense (arrows) on T2 (A) and isointense on T1 weighted images (B) with no contrast uptake post gadolinium administration (C).

Dog 4, a 3-year-old male neutered Cavalier King Charles Spaniel presented with a several weeks history of lethargy and intermittent head tilt. No further information was available about any other household pets. Physical examination revealed a systolic heart murmur consistent with previously diagnosed mitral valve disease. The neurologic examination revealed normal mentation, proprioceptive deficits in the right thoracic and right pelvic limb, an intermittent right sided head tilt, decreased menace responses bilaterally and mild pain on cervical and lumbar spinal palpation. A multifocal central nervous system disease was suspected. A complete blood count showed moderate leuco- (23.60 10e9/l, reference range: 6–17.1 10e9/l) and lymphocytosis (12.74 10e9/l, reference range: 1–4.8 10e9/l) and serum biochemistry revealed moderately elevated creatine kinase (1729 U/l, reference range: 61–394 U/l) and mildly increased alanine aminotransferase (135 U/l, reference range: 13–88 U/l). MR imaging of the entire central nervous system revealed mild bilateral and symmetrical cerebellar atrophy with widening of the dorsal folia and multifocal intra-axial lesions affecting the thalamus, lentiform nucleus, centrum semiovale, mesencephalon, metencephalon, subcortical gray matter of the frontal, parietal, and temporal lobes (Fig. 6) and an intramedullary spinal cord lesion mainly affecting the right dorsal horn at the level of C2-C3 vertebral bodies. The lesions were hyperintense on T2-weighted and FLAIR images and iso-hypointense on T1 with minimal contrast enhancement after gadolinium administration. Cisternal cerebrospinal fluid analysis revealed a mixed cell pleocytosis with lymphocytic predominance (total nucleated cell count: 282 mm3, reference range: <5 mm3) and a moderately elevated protein concentration (0.42 g/l, reference range: 0.25 g/l). Polymerase chain reaction testing of cerebrospinal fluid was positive for Neospora caninum and serology showed increased antibody titres (>800). Treatment was initiated with oral clindamycin 12.5 mg/kg and prednisolone 1 mg/kg twice daily whilst awaiting results. The patient deteriorated neurologically within 1 week and became nonambulatory tetraparetic with swaying head movements. Prednisolone was discontinued and the patient was treated with trimethoprim-sulfonamide 15 mg/kg and clindamycin 20 mg/kg twice daily. The patient was re-examined 1 month after diagnosis and showed good neurological improvement. Only mild generalized ataxia and a subtle right sided head tilt was evident. Trimethoprim-sulfonamide was administered for a total of 3 months and clindamycin for 5 months. Serial re-examinations at the referring practice mentioned a remaining head tilt but otherwise stable neurological status and Neospora serology remained negative 3, 6, and 10 months after diagnosis. The patient developed marked mandibular lymph node enlargement, hepato- and splenomegaly, pyrexia and nonregenerative anemia, and leucocytosis with atypical lymphocytes 10 months after diagnosis. Leukemia was suspected by the referring veterinary surgeon and further investigations were declined by the owner. The patient died shortly afterwards. No postmortem examination was carried out.

Figure 6.

Dog 4. Four T2 weighted transverse MR images showing multifocal intra-axial and poorly delineated lesions in the thalamus (A: arrow), in the ectosylvian gyri (A: arrow head and dotted arrow) and sylvian gyrus (B: arrow head), in the medial geniculate body (B: arrow) and the metencephalon (D: arrow). Mild cerebellar atrophy with loss of gray and white matter distinction (C: arrows, D: arrowhead).

Discussion

In this study we identified a small population of dogs diagnosed with Neospora caninum that had central nervous system abnormalities on MR imaging that did not primarily affect the cerebellum. One of these dogs had no clinical or imaging abnormalities attributable to the cerebellum at all. Neurologic examination in the dogs of our study was indicative of a forebrain and cerebellar lesion in dog 2 and multifocal central nervous system disease in dogs 1, 3, and 4. Magnetic resonance imaging showed mild bilateral and symmetrical cerebellar atrophy in three of four dogs (dogs 2, 3, 4), intramedullary spinal cord changes in two dogs (dogs 3, 4) and a mesencephalic and metencephalic lesion in one dog (dog 2). Multifocal brain lesions were recognized in two dogs (dogs 1, 4) and were present in the thalamus, lentiform nucleus, centrum semiovale, internal capsule, brainstem, and cortical gray matter of the frontal, parietal or temporal lobe. It is likely the MR imaging abnormalities of the dogs in this study represent regions of the central nervous system with inflammatory infiltrates or malacia. These histopathological findings have been reported in dogs with Neosporosis previously. [3, 6], [12-15] Why cerebellar atrophy has only been mild or absent in these dogs and other lesions in the brain or spinal cord more predominant is unknown.

In this study, histopathology was not available in any of the dogs, however, increased serum antibody titres for Neospora caninum and the presence of protozoal deoxyribonucleic acid in the cerebrospinal fluid is indicative for active infection with this parasite. An additional infection with canine distemper virus was considered improbable in dog 3 as a weak positive serology in conjunction with a negative polymerase chain reaction result on both urine and cerebrospinal fluid most likely indicates previous vaccination.

Although all dogs stabilized or improved following antiprotozoal therapy, none have shown complete resolution of their clinical signs at this time. Treatment mainly affects the active dividing form of the parasite (tachyzoites), with little effect on the tissue encysted bradyzoites.[20, 21] Inflammation and necrosis in response to the parasite rather than the persisting tissue cyst is likely to be the reason for ongoing neurological abnormalities.[10, 22] Information about treatment is limited and has mainly been extrapolated from treatment of Toxoplasma infection in humans. There are no substantiated studies investigating the type and duration of optimal antiprotozoal therapy for the treatment of canine Neosporosis. Administration of antiprotozoal medication (clindamycin and trimethoprim-sulfonamide) for several months has proven effective for the patients in this study. The positive response to treatment in dogs 1, 3, and 4, and the deterioration of neurological signs following immunosuppression and cessation of treatment in dogs 4 and 2, respectively, supports and infectious cause for the neurological impairment.

In conclusion, findings from the current study indicated that Neosporosis should be considered as an important differential diagnosis for multifocal or focal intramedullary and intra-axial lesions affecting gray and white matter with or without cerebellar involvement detected on MR imaging.

ACKNOWLEDGMENTS

The authors would like to thank Joe Fenn, Steven De Decker, and Chris Scudder for their case contribution.

Appendix

All images were obtained in a 2D mode.

BRAIN:

1.1 SAGITTAL T2-weighted images: TE: 110, TR: 3104–4212, slice thickness: 3.5, interslice gap: 3.5, interslice gap: 3.85, flip angle: 90, number of acquisitions: 4–11, manufacturer specific name: TSE

1.2 TRANSVERSE T2-weighted images: TE: 110, TR: 3992–5986, slice thickness: 2.56–3.5 interslice gap: 3.56–4.5, flip angle: 90, number of acquisitions: 5–22

1.3 TRANSVERSE FLAIR images: TE: 120, TR:6000, inversion time: 2000, slice thickness: 2.77–3.5 interslice gap: 3.56–4.5, flip angle: 90, number of acquisitions: 7–24

1.4 TRANSVERSE T1-weighted images: TE: 15, TR: 450–487, slice thickness: 2.7–3.5 interslice gap: 3.56–4.5, flip angle: 69, number of acquisitions: 6–23

1.5 TRANSVERSE T1-weighted Inversion recovery: TE: 15, TR: 3435, inversion time: 400, slice thickness: 3.5 interslice gap: 4.5, flip angle: 90, number of acquisitions: 13-14

1.6 TRANSVERSE T2*: TE: 23, TR: 807, slice thickness: 3.75, interslice gap: 4.5, flip angle: 18, number of acquisitions: 14, manufacturer specific name: T2-weighted FFE

SPINE:

1.7 SAGITTAL T2-weighted images: TE: 110–120, TR: 1999–3144, slice thickness: 1.75–3.5, interslice gap: 2.05–3.85, flip angle: 90, number of acquisitions: 7–21

1.8 TRANSVERSE T2-weighted images: TE: 120, TR: 3675–3680, slice thickness: 2.5–3, interslice gap: 2.8–3.3, flip angle: 90, number of acquisitions: 10–13

1.9 SAGITTAL T1-weighted images: TE: 8, TR: 400, slice thickness: 1.75–2, interslice gap: 2.05–2.3, flip angle: 90, number of acquisitions: 9–22

1.10 TRANSVERSE T1-weighted images: TE: 8, TR: 430–484, slice thickness: 2.5–3, interslice gap: 2.8–3.3, flip angle: 90, number of acquisitions: 11–30

Ancillary