• Open Access

Clinical and Magnetic Resonance Imaging Features of Central Nervous System Blastomycosis in 4 Dogs

Authors


  • The original MRI for Case 3 was performed at the Veterinary Specialty Center in Buffalo, Grove, IL, and the images in Figure 3 are from this examination. All work for the other 3 cases was done at University of Wisconsin-School of Veterinary Medicine, Madison, WI. This paper has not been previously presented at a meeting. It has been accepted as an abstract scheduled for oral presentation at the 2010 ACVIM Forum in Anaheim, CA.

Corresponding author: Helena Rylander DVM, Dip. ACVIM (Neurology), Department of Medical Sciences, School of Veterinary Medicine, University of Wisconsin-School of Veterinary Medicine, 2015 Linden Drive, Madison, WI 53706; e-mail: rylander@svm.vetmed.wisc.edu.

Abbreviations:
ABLC

amphotericin B lipid complex

ALP

alkaline phosphatase

ALT

alanine aminotransferase

CNS

central nervous system

CSF

cerebrospinal fluid

CT

computed tomography

EIA

enzyme immunoassay

IFAT

immunoflourescence antibody titer

MRI

magnetic resonance imaging

OS

left eye

OU

both eyes

SAMe

S-adenosyl-methionine

Case 1

A 4-year-old female spayed Golden Retriever was evaluated after generalized seizures for 2 days and a 1-month history of decreased appetite, conjunctivitis in the left eye (OS), and pyoderma. Therapy with oral cephalexin (25.6 mg/kg PO q12h), neopolydex ophthalmic solution (OS q6h), and oral phenobarbital (1.6 mg/kg PO q12h) did not improve clinical signs. Physical examination revealed multifocal nodular skin lesions, granulomatous fundic lesions in both eyes (OU), and uveitis OS. Neurologic examination abnormalities included an absent menace response and direct pupillary light reflex OS, delayed conscious proprioception in all limbs and generalized ataxia. Neuroanatomic localization was forebrain.

CBC and serum biochemistry profile results were within reference intervals. On thoracic radiographs, a bilateral interstitial pulmonary pattern was present in the caudal dorsal lung fields with a round, soft tissue opacity (2.5 × 2 cm) in the left caudal lung.

Magnetic resonance imaging (MRI) of the brain was performed with a 1.0 T magnet.a Images included T2-weighted (T2W), fluid-attenuated inversion recovery (FLAIR), and T1-weighted (T1W) images before and after intravenous contrast administration (0.1 mmol/kg gadodiamide).b Multifocal intra- and extra-axial lesions causing a mild left midline shift were identified in the right cerebrum, thalamus, and midbrain). Lesions were also present in the left ocular posterior chamber and temporalis muscles adjacent to affected brain region. Brain lesions were hypointense to grey matter on T1W images and mildly hyperintense with marked perilesional edema on T2W and FLAIR images. After contrast administration, there was enhancement of lesions and meninges (Fig 1). Radiological diagnosis was multifocal brain disease with secondary meningitis and left ocular granuloma or retinal hemorrhage.

Figure 1.

 (A) T2-weighted transverse magnetic resonance image from dog 1. There are hyperintense lesions in the right temporal-piriform lobes (arrow) and the right temporalis muscle (arrowhead). (B) Transverse T1-weighted image. The cerebral lesion is hypointense to grey matter and the muscle lesion is isointense to normal muscle. (C) Transverse T1-weighted postcontrast image. Note the strong, homogenous contrast enhancement in the right cerebral and temporal muscle lesions (arrows) and meningeal enhancement (arrowhead). The hyperintensity on T2-weighted images (A) is larger than the area contrast enhancing, suggesting perilesional edema.

Cisternal cerebrospinal fluid (CSF) analysis showed 15 white blood cells/μL (reference range, ≤ 5), 2 red blood cells/μL, and 25.9 mg/dL protein (reference range, ≤25). The differential cell count was 30% neutrophils, 60% lymphocytes, and 10% macrophages. CSF immunoflourescence antibody titer (IFAT) for Neospora caninum was 10 (reference, ≥10 positive).

Microscopic examination of an aspirate of the pulmonary mass showed pyogranulomatous inflammation and Blastomyces dermatitidis organisms. Disseminated blastomycosis was presumptively diagnosed and the dog was treated with fluconazole (5 mg/kg PO q12h); phenobarbital (2.5 mg/kg PO q12h), cephalexin (20 mg/kg PO q12h), and neopolydex (OU q8h) were continued.

The dog represented for examination because of acute obtundation 2 weeks later. Inflammation secondary to death of Blastomyces organisms in the central nervous system (CNS) was suspected. After therapy of dexamethasone sodium phosphate (0.07 mg/kg IV once) and prednisone (0.25 mg/kg PO q12h), the dog's clinical signs rapidly improved. Clindamycin (10 mg/kg PO q12h) was added based on a N. caninum serum titer of 400 (reference, ≥100 indicates exposure).

One month later, neurologic examination was normal except for absent menace response and direct pupillary light reflex OS. On MRI examination there was resolution of the muscle granuloma, persistence of intracranial granulomas and parenchymal edema, and lateral ventriculomegaly. Cisternal CSF analysis was within reference intervals. Prednisone therapy was tapered and discontinued.

The dog had recurrence of ataxia and conscious proprioception deficits 10 weeks later. On MRI, there was resolution of 2 forebrain granulomas; however, a new granuloma adjacent to the tentorium cerebelli was noted. A CBC and cisternal CSF analysis were unremarkable. Serum biochemistry profile showed increased alanine aminotransferase (ALT, 492 U/L; reference range, 14–151 U/L), bilirubin (0.7 mg/dL; reference range, 0.1–0.5 mg/dL), cholesterol (728 mg/dL; reference range, 98–300 mg/dL), and hypoalbuminemia (2.2 g/dL; reference range 2.6–4 g/dL). Phenobarbital was tapered and potassium bromide (40 mg/kg PO q24h), S-adenosyl-methionine (SAMe, 16 8mg/kg PO q24h), and vitamin K1 (1.25 mg/kg PO q12h) were initiated; clindamycin was discontinued.

One week later, the dog was obtunded with nonambulatory tetraparesis and respiratory distress. Serum biochemistry profile showed increased ALT (716 U/L), alkaline phosphatase (ALP, 5013 U/L; reference range, 12–289), cholesterol (770 U/L), and hypoalbuminemia (2.1 g/dL). Euthanasia was elected because of acute dyspnea and poor prognosis.

Postmortem histopathologic diagnoses were pulmonary artery thrombosis, severe chronic degenerative hepatopathy, right ocular posterior uveitis, severe chronic retinal detachment with retinitis and Blastomyces organisms OS, and marked meningitis surrounding both optic nerves. There was mild multifocal interstitial edema in the midbrain and neuronal cytoplasmic vacuolation in the cerebellum. Blastomyces organisms or inflammation were not seen in the brain; periodic acid Schiff and Gomori methenamine silver staining were not performed.

Case 2

A 6-year-old female spayed German Shorthaired Pointer was evaluated for 1 day of generalized seizures and 1 week of lethargy, panting, anorexia, and ataxia. The dog was treated with oral daily prednisone (dose unknown) for atypical hypoadrenocorticism. Dexamethasone sodium phosphate (0.1 mg/kg IV) was administered by the referring veterinarian the morning of presentation. The dog was pyretic (39.5°C) and abnormalities on neurologic examination included obtundation, anisocoria, ataxia, and absent conscious proprioception in all limbs. Neuroanatomic localization was forebrain.

Administration of mannitol (1 g/kg IV once) resulted in improved mental status. A CBC was unremarkable. Abnormalities on serum biochemistry profile included increased calcium (11.8 mg/dL; reference range, 9.5–11.2 mg/dL), phosphorus (6.9 mg/dL; reference range, 2.6–6.2 mg/dL), ALP (606 U/L), creatine kinase (374 U/L; reference range, 7–203 U/L), ALT (109 U/L), cholesterol (338 mg/dL), and low blood urea nitrogen (7 mg/dL; reference range, 8–25 mg/dL). Thoracic radiographs were unremarkable.

A brain MRI, including T1W before and after contrast, and T2W images was performed. A large mass in the left nasal cavity extending through the cribiform plate into the left fronto-olfactory area was present (Fig 2). Multiple intraaxial lesions were present in the left olfactory bulb and frontal lobe with a marked right midline shift. Lesions were hypointense to grey matter on T1W images and hyperintense on T2W images with moderate perilesional edema. The majority of lesions enhanced strongly and homogeneously after contrast administration; 2 smaller areas ring-enhanced and meningeal enhancement was present. Radiological diagnosis was a mass in the caudoventral nasal cavity with extension into the brain.

Figure 2.

 T1-weighted postcontrast sagittal magnetic resonance image from dog 2. Lesions in the nasal cavity and frontal cortex enhance strongly and homogenously with an area of ring enhancement (arrow).

Cytologic examination of the intranasal portion of the mass showed pyogranulomatous inflammation and B. dermatitidis organisms. Euthanasia was elected and postmortem examination declined.

Case 3

An 8-year-old male neutered Samoyed was evaluated for previously diagnosed disseminated blastomycosis and hepatopathy. Eight weeks prior, the dog was pyretic (40.6°C), coughing, and had mucopurulent nasal discharge. Bronchointerstitial pulmonary infiltrates were present on thoracic radiographs. Two weeks later, the dog had 2 generalized seizures and MRI of the brain was performed.

The brain MRI included T1W before and after contrast, T2W, and FLAIR images. There was an extra-axial left frontal lobe lesion adjacent to the falx cerebri, an extra-axial left occipital lesion, a marked right midline shift, and a right cheek lesion (Fig 3). Brain lesions were hypointense to grey matter and the cheek lesion was isointense to surrounding muscle on T1W images. On T2W and FLAIR images, brain lesions were hyperintense peripherally and hypointense centrally with marked perilesional edema and the cheek lesion was hyperintense. There was strong homogeneous contrast enhancement of occipital lobe and cheek lesions; the frontal lobe lesion ring-enhanced. Extensive meningeal enhancement was noted. Radiological diagnosis was multifocal brain disease. The cheek lesion was not described in the initial radiology report. Cisternal CSF analysis showed 4 white blood cells/μL, 1 red blood cell/μL, and 156 mg/dL protein. A differential was not provided; however, an increased percentage of nondegenerate neutrophils was noted. Examination of CSF for N. caninum (IFAT), Blastomyces serology (agar gel immunodiffusion), Cryptococcus antigen, and bacterial culture were negative. Infectious or neoplastic disease was suspected and the dog was treated with prednisone (0.3 mg/kg PO q12h), clindamycin (10 mg/kg PO q12h), phenobarbital (2.9 mg/kg PO q12h), and fluconazole (6 mg/kg PO q12h).

Figure 3.

 (A) T2-weighted transverse magnetic resonance image from dog 3 at the level of the occipital cortex. Marked edema is present in the white matter of the left occipital cortex; the edema obscures the occipital mass lesion. (B) T1-weighted postcontrast sagittal image. Note the uniform enhancement of the lesion in the frontal cortex with an area of ring enhancement. Diffuse enhancement in the occipital cortex (arrow) and meningeal enhancements (arrowhead) are present.

The cheek lesion ruptured 1 week later and cytology showed pyogranulomatous inflammation with B. dermatitidis organisms. Skin lesions erupted and cytology showed neutrophilic inflammation without fungal organisms. Ciprofloxacin (14.8 mg/kg PO q12h) was administered when lesions failed to resolve.

On presentation to the Veterinary Medical Teaching Hospital, physical examination showed mild mucopurulent nasal discharge, sterterous breathing, and previously mentioned skin lesions. Neurologic examination and thoracic radiographs were normal. A hyperechoic liver with multiple hypoechoic nodules was present on abdominal ultrasound. A CBC showed leukocytosis (22,300 cells/μL; reference range, 5,700–16,300 cells/μL). Serum biochemistry profile showed increased bilirubin (4.5 mg/dL), ALT (637 U/L), and ALP (4060 U/L). Canine pancreatic specific lipase was >1000 μg/L (reference range, ≤200 μg/L). Leptospira spp. serology was negative. Blastomyces urine antigen enzyme immunoassay (EIA) was 26.19 (reference range, <1.0).c Therapy with SAMe (6 mg/kg PO q24h) and milk thistle (dose unknown) were initiated. The fluconazole dose was reduced (6 mg/kg PO q24h); all other medications were discontinued.

Three months later, all blood values and cisternal CSF analysis were within normal range. Blastomyces CSF antigen EIA was 0.69. On MRI examination 6 months later, brain lesions were markedly reduced in size and there was resolution of perilesional edema (Fig 4). The fluconazole dosage was increased (6 mg/kg PO q12h) after liver values normalized.

Figure 4.

 T1-weighted sagittal postcontrast image from dog 3 acquired 1 year after original magnetic resonance imaging examination (Fig 3). The previously described frontal and occipital lobe lesions have resolved and no enhancement is present in the brain.

One year later, there was complete resolution of lesions with mild cortical atrophy on MRI examination (Fig 4). Neurologic evaluation has remained normal, 3 months off all medication.

Case 4

A 7-month-old male intact English Springer Spaniel was evaluated for right thoracic limb lameness and cervical pain that progressed to nonambulatory hemiparesis over 10 days. Therapy with amoxicillin and carprofen was unsuccessful; an injection of dexamethasone sodium phosphate (dosage unknown) temporarily alleviated pain. Physical examination showed muscle atrophy in the right thoracic limb, anterior uveitis OU, and granulomas with subretinal infiltrates OU on fundic exam. Abnormalities on neurologic examination included nonambulatory right hemiparesis, right pelvic and thoracic limb absent conscious proprioception, reduced right thoracic limb withdrawal reflex, and pain on palpation of the caudal cervical spine. Neuroanatomic diagnosis was C6-T2 myelopathy.

A CBC, serum biochemistry profile, and urinalysis were normal. A generalized nodular interstitial pattern was present on thoracic radiographs.

A cervical MRI including T1W before and after contrast, fat saturation T2W, and gradient echo images was performed. There was an intramedullary mass representing 75% of the cross sectional area at the level of the spinal cord over the C5-6 intervertebral disc space (Fig 5). Compared with adjacent spinal cord, the mass was hypointense centrally with a hyperintense rim on T2W images and isointense on T1W images. There was strong homogenous contrast enhancement of the mass. Spinal cord edema was noted cranial and caudal to the lesion. Radiological diagnosis was an intramedullary spinal cord lesion and differential diagnoses were neoplasia or a fungal granuloma.

Figure 5.

 (A) T2-weighted fat saturation sagittal image from dog 4. The mass in the spinal cord over the C5-6 intervertebral disc space is hypointense centrally with a hyperintense rim (arrow). Note the perilesional edema cranial and caudal to the mass (arrowheads). (B) Sagittal T1-weighted postcontrast image. There is strong, uniform contrast enhancement of the mass.

Cisternal CSF analysis showed 8 white blood cells/μL, 105 red blood cells/μL, and 58.7 mg/dL protein. The differential count was 12% neutrophils, 21% lymphocytes, and 67% mononuclear cells.

A fine needle aspirate from a pulmonary mass showed pyogranulomatous inflammation and B. dermatitidis organisms. Disseminated blastomycosis was suspected and therapy was initiated with fluconazole (10 mg/kg PO q12h), amphotericin B lipid complex (ABLC, 10 mg/kg IV cumulative dose), carprofen (2.5 mg/kg PO q12h), morphine (1 mg/kg SC as needed), and neopolydex ointment (OU q12h). Two days later, the dog became obtunded, febrile (40.1°C), and experienced increasing cervical pain. Inflammation secondary to death of Blastomyces organisms was suspected; prednisone (0.25 mg/kg PO q12h) and tramadol (2.5 mg/kg PO q12h) were administered and carprofen and morphine were discontinued. Casts were seen on urinalysis after 7 days and ABLC was discontinued. The dog became ambulatory, nonpainful, and was discharged on fluconazole, prednisone, and neopolydex ointment.

One month later, neurologic examination was normal, retinal lesions improved, and there was partial resolution of lung lesions on thoracic radiographs. Bilateral optic neuritis developed 6 months later. ABLC was administered (2 mg/kg IV q3 days), but after 4 treatments, casts were noted in the urine and dosing was reduced (1 mg/kg IV, q4 days). Serum biochemistry profile showed increased ALT (345 U/L) and ALP (298 U/L); treatment with SAMe was initiated (11.3 mg/kg PO q12h). Six weeks later, ocular lesions appeared inactive and ALT decreased (194 U/L); ABLC was discontinued because of casts in the urine (15 mg/kg IV total dose).

Physical and neurologic examinations have remained normal except for a short period with mild ambulatory right hemiparesis 3 months later, which resolved without any change in treatment. At 2 months and 20 months after diagnosis, cisternal CSF analyses were normal. On MRI examination, 20 months after original presentation the spinal lesion reduced in size and perilesional edema resolved (Fig 6). The dog has been clinically normal 3 years without any medications.

Figure 6.

 Sagittal T2-weighted image from dog 4 acquired 20 months after original magnetic resonance imaging examination (Fig 5). The mass in the spinal cord over the C5-6 intervertebral disc space is reduced in size and there is resolution of perilesional edema.

Discussion

All dogs in the present series had abnormal neurologic examinations reflecting the site of disease. Thus, no specific neurologic signs are classic for blastomycosis as lesions can form anywhere in the CNS. One dog (case 2) presented only for neurologic signs, which is rare; 3 dogs also had signs of disease in the lungs (cases 1, 3, 4), eyes (cases 1, 4), and skin (cases 1, 3). The dog with only CNS signs had increased total calcium. Granulomatous disease is a known cause of hypercalcemia.1 However, ionized calcium is the biologically active calcium fraction but was not evaluated in this case. Total serum calcium does not correlate well with ionized calcium and ionized hypercalcemia was found to be present in only 5.3% of dogs with blastomycosis in a recent study.2

Premortem diagnosis of CNS blastomycosis is challenging as specimens are often difficult to obtain by biopsy and attempts to isolate the organism from CSF are rarely successful.1,3 Agar gel immunodiffusion against the A-antigen of B. dermatitidis is the most commonly used serologic test; however, it does not provide definitive diagnosis and has low sensitivity ranging from 17.4 to 41%.1,4,5 A recently available EIA test for detection of B. dermatitidis galactomannan antigen in urine and serum has sensitivities of 93.5 and 87.0%, respectively, and antigen concentrations decrease with treatment.5

Blastomycosis in the 4 dogs reported herein was diagnosed by cytologic examination of the intranasal portion of an intracranial mass (case 2), pulmonary masses (cases 1 and 4), and a skin lesion (case 3). Cytology was not performed on all CNS lesions; disseminated blastomycosis with involvement of the brain or spinal cord was presumed based on neurologic signs, MRI findings, CSF analysis, and response to treatment (cases 3, 4). Surgical biopsy can be performed for diagnosis of CNS lesions and to provide therapeutic decompression of spinal cord granulomas. Surgery was not an option in case 4 because of the intramedullary location of the spinal lesion.

The MRI characteristics in dogs with CNS blastomycosis have not previously been reported. In 1 report of a dog with intracranial blastomycosis, computed tomography (CT) imaging failed to show intraparenchymal pyogranulomas that were seen histopathologically.6 Strong contrast enhancement of a mass and severe mass effect were seen on CT of a cat diagnosed with cerebral blastomycosis.7 A case of intracranial blastomycosis in a human was described on MRI as isointense on T1W images, hyperintense on T2W, and strongly contrast enhancing.8

The MRI appearance of the lesions had many similarities. Lesions were hypointense or isointense on T1W images and hyperintense on T2W images. The spinal cord lesion had an area of central hypointensity. Areas of T2W hypointensity may occur because of the macrophage contents of hemosiderin and dense population of organisms.9 Moderate to marked perilesional edema was present in all cases. Most lesions contrast enhanced strongly and homogeneously. A smaller number of lesions ring-enhanced; this feature is reported on MRI in humans with fungal CNS infections secondary to abscess formation.9 Meningeal enhancement was observed in all dogs with intracranial lesions.

The majority of intracranial lesions were extra-axial, broad-based, and in the frontal lobe. Inhalation of Blastomyces spores may lead to direct invasion of the brain through the cribiform plate, resulting in a predilection for the olfactory bulb and frontal lobe. Alternatively, hematogenous entry of Blastomyces organisms into the CNS may result in intra- or extra-axial lesions and meningitis. Extra-axial location, meningeal enhancement, strong contrast enhancement, and cystic components are MRI features used to identify meningiomas.10 Fungal granulomas should be a differential diagnosis for masses resembling meningiomas on MRI examination.

CSF findings in dogs with CNS blastomycosis are rarely reported. One report in the veterinary literature describes neutrophilic pleocytosis and Blastomyces organisms on cytologic examination in a dog with neurologic disease.11 Three dogs had CSF analysis at the time of diagnosis; 2 had an increased total nucleated cell count and all had an increased percentage of nondegenerate neutrophils and increased protein concentration. Blastomyces organisms were not detected in the CSF in any of the dogs. Neutrophilic and eosinophilic pleocytosis with increased protein have been described in the CSF of dogs with Cryptococcus infection.12,13 Neutrophilic pleocytosis has also been described on CSF cytology in humans with blastomycotic meningitis.14 Culturing of the CSF, though specific, requires prolonged growing time and large sample volumes; reported sensitivity in humans with CNS Cryptococcus is 50–80%.15

Three dogs underwent serial MRI examinations and CSF analyses. In case 3, resolution of intracranial disease was seen on MRI after 1 year of fluconazole therapy. In case 1, intracranial granulomas persisted despite normal CSF cytology after 1 month of treatment. The spinal granuloma in case 4 reduced in size, but had not resolved after 20 months of therapy despite normal CSF analysis 2 months into treatment. Serial MRI examinations may support the response to medical therapy since neurologic examination and CSF analysis may not identify persistent CNS infection. However, MRI examination cannot differentiate whether static lesions are active or sterile. Patient response and assessment of decreasing antigen concentrations may be more useful in documenting recovery.

CNS blastomycosis in dogs carries a guarded prognosis.1,16 Two dogs in this case series were alive 1.5 and 5 years after diagnosis. Medical therapy was not attempted in 1 dog (case 2) and failed after 17 weeks in 1 dog (case 1). Amphotericin B is an effective drug for the therapy of blastomycosis; however, it is nephrotoxic, requires intravenous administration, and has been replaced by itraconazole as the treatment of choice in non-life-threatening cases without CNS involvement.1,16,17 Fluconazole penetrates the blood-brain barrier more readily than intraconazole and may be more successful in treating CNS blastomycosis.7,18,19 Adverse effects of fluconazole in veterinary medicine include dose-dependent anorexia, increased liver enzymes, and hepatic toxicity.20

In the dog that underwent postmortem examination (case 1), B. dermatitidis organisms and inflammation were found in the left eye but not the brain and staining for detection of organisms was not performed. Seventeen weeks of therapy elapsed between MRI and necropsy examination, which may have made organism detection more difficult. It is possible that cerebral fungal infection resolved during therapy or that representative samples of affected brain tissue were not analyzed histopathologically.

Footnotes

aSigna Advantage 1.0. T, GE Medical Systems, Milwaukee, WI

b Omniscan, GE Healthcare, Princeton, NJ

c Mira Vista, Indianapolis, IN

Acknowledgment

The study was not supported by a grant.

Ancillary