Abdominal Cryptococcosis in Two Dogs: Diagnosis and Medical Management
Canine cryptococcosis cases are typically reported as neurologic, disseminated, or both. There have been few reports of other parenchymal organ involvement. Dogs infected with Cryptococcus spp. are likely to develop central nervous system involvement, and those that are severely affected are treated aggressively with surgery and/or amphotericin B. This report describes two cases of canine abdominal cryptococcosis: one boxer with primary alimentary cryptococcosis alone and one miniature schnauzer with pancreatic and disseminated cryptococcosis. The boxer is unique in that the dog suffered from primary alimentary cryptococcosis without dissemination, secondary anemia due to gastrointestinal losses, and is the second case to have Cryptococcus spp. identified on fecal examination as part of the diagnostic workup. Unlike previous reports, surgery was not performed in either case, and both dogs were treated with fluconazole alone. Currently, both dogs are free from clinical signs, and Cryptococcus spp. antigen titers are negative at 17 and 15 mo after initial presentation. These cases suggest fluconazole may be effective therapy alone for canine abdominal cryptococcosis, negating the need for high-risk therapy options such as surgery and/or amphotericin B in some cases.
Introduction
Cryptococcosis is a fungal infection with worldwide distribution.1 Cryptococcus spp. gain entry into the body primarily through inhalation. Uncommonly, other routes of entry may include alimentary or cutaneous inoculation.2 The majority of canine cryptococcosis cases are reported as neurologic, disseminated, or nasal infections. There have been rare reports of other parenchymal organ involvement, most of which were also disseminated.3–5
Therapy for cryptococcosis includes triazole antifungal drugs. Fluconazole is a desirable treatment option given its ability to penetrate the central nervous system (CNS), eye, and urinary tract.1,2 Management of either CNS or severe disseminated cryptococcosis may also include amphotericin B and/or flucytosine.1,6 A treatment protocol for abdominal cryptococcosis has not been established because there have been limited cases reported in the literature. Surgical debulking has been recommended, but long-term success rates have yet to be documented.1,2 The following cases describe diagnosis and successful management of atypical canine cryptococcosis with triazole therapy alone.
Case Report
Case 1
A 3 yr old spayed female boxer weighing 32 kg presented in June 2011 with a 1 wk history of hematochezia and a 2 day history of pallor with two collapsing episodes. At the time, the boxer resided in New Jersey but did spend 1 yr in Pennsylvania after adoption (1 yr prior to presentation). Extended travel history was unknown because the dog was adopted from a Pennsylvania animal shelter as an adult.
Remarkable physical examination findings included fever and a palpable abdominal mass. A complete blood count (CBC) revealed a microcytic, hypochromic anemia (mean corpuscular volume, 48.7 fL; reference range, 62.7–75.5 fL; mean corpuscular hemoglobin concentration, 30.5 g/dL; reference range, 32.2–36.3 g/dL; hematocrit; 17.9%; reference range, 40.3–60.3%). Serum biochemical analysis revealed mild hypoalbuminemia (albumin, 2.4 g/dL; reference range, 2.5–3.7 g/dL). Three-view thoracic radiographs were within normal limits. A sterile urine sample was collected via cystocentesis. Urinalysis results were within normal limits, and a urine culture was negative. Abdominal ultrasonography performed by a board-certified radiologist revealed a 5 cm long segment of hypoechoic jejunum with circumferential thickening (1.46 cm), loss of layering, hyperechoic surrounding mesentery, and an enlarged (3.7 cm) hypoechoic mesenteric lymph node (Table 1). All follow-up abdominal ultrasonography examinations were performed by the same radiologist.
Cytologic examinations of feces and fine-needle aspirates of the jejunum and mesenteric lymph nodes demonstrated fungal organisms most suggestive of Cryptococcus spp. A serum Cryptococcus spp. antigen titera was positive at 1:31 (reference range, negative).
The dog was diagnosed with primary alimentary cryptococcosis and secondary iron deficiency anemia that was likely due to jejunal ulceration. No additional organ involvement was identified on thoracic radiographs, urinalysis, and fundic examination; therefore, cryptococcosis was not considered to be widely disseminated, but that hypothesis could not be definitively ruled out antemortem. The dog received 300 mL of packed red blood cells, IV fluid therapy, and a single iron dextranb injection (10 mg/kg intramuscularly) in hospital. Therapy options offered included surgical debulking with medical management and medical management alone (triazole drugs ± amphotericin B). The owner elected triazolec therapy alone, which was initiated in hospital but then discontinued after 24 hr. Due to financial constraints and improved CNS penetration, given the propensity for neurologic disease in dogs, the originally selected triazole was changed to a generic triazoled (12.5 mg/kg per os [PO] q 48 hr, divided). The dog improved on the generic triazole, proton pump inhibitore (1 mg/kg PO q 24 hr), cytoprotective agentf (31.25 mg/kg PO q 8 hr), and H2 antagonistg (1 mg/kg PO q24hr) and was discharged two days later.
The dog was re-examined 1 mo later, and abdominal ultrasonography showed decreased length of affected jejunum (length, 4–5 cm; thickness, 1.7 cm) and a smaller mesenteric lymph node (3 cm); however, an affected segment of duodenum (length, 6.5 cm; thickness, 1.5 cm) with suspected ulceration was identified. There was also a 3.5-fold increase in serum Cryptococcus spp. antigen titer at 1:107.
Surgical debulking and/or the addition of amphotericin B were again declined, and the genertic triazole dose was increased to 18.75 mg/kg PO q 24 hr, divided. That change in dose was associated with intermittent vomiting and diarrhea. The owner reduced the triazole dose for 7 days prior to the 2 mo recheck at which time the serum Cryptococcus spp. antigen titer had increased to 1:298. At that time, the higher dose of triazole (18.75 mg/kg PO q 24 hr, divided) was reinstituted.
At re-evaluation, 6 mo after diagnosis, the vomiting and diarrhea had stopped and serum Cryptococcus spp. antigen titer had decreased to 1:15. At 7 mo post diagnosis, repeat abdominal ultrasound showed significant reduction of the previously identified lesions (jejunal thickness, 8.45 mm; duodenal thickness, 9.5 mm; mesenteric lymph node, 1.3 cm). The Cryptococcus spp. antigen titer was negative (i.e., 0). At that time, the owner independently decreased the dose of triazole to 12.5 mg/kg PO q 24 hr, divided. Recheck Cryptococcus spp. antigen titer at 9 mo was again positive at 1:32, and the 18.75 mg/kg PO q 24 hr (divided) dose of triazole was reinstituted. At the 15 mo recheck the Cryptococcus spp. antigen titer returned to negative.
Recheck abdominal ultrasonography 17 mo after presentation revealed marked improvement of the previously identified lesions (affected jejunal length, 4.3 cm; jejunal thickness, 6.1 mm; duodenum normal; mesenteric lymph node, 6.7 mm). It remained possible that the residual jejunal changes represented either scar tissue or active infection. A fecal cytology was performed and did not show any evidence of infection. As of November 2012, 17 mo since initiation of traizole therapy, the dog remained free of clinical signs and the Cryptococcus spp. antigen titer remained negative. The dog continued to receive triazole (18.75 mg/kg PO q 24 hr, divided).
Case 2
An 8 yr old, castrated male miniature schnauzer weighing 9.8 kg presented in August 2011 for acute onset of seizures and inappetence. The dog resided in Massachusetts and had a history of diabetes mellitus, pancreatitis, and diabetic cataracts oculus uterque with enucleation of the left eye prior to presentation. The dog previously presented to the ophthalmology service in March of 2011 for bilateral diabetic cataracts and nonglaucomatous severe uveitis with iris bombe oculus sinister. He underwent phacoemulsification oculus dexter (OD) and enucleation oculus sinister on March 15, 2011. Histopathology of the left eye revealed pyogranulomatous endophthalmitis with hyphema, lens capsule rupture, anterior lens luxation, and anterior and posterior synechiae. The intraocular inflammatory process was suspected to be secondary to abrupt leakage of lens protein as no organisms were observed in sections stained with either hematoxylin and eosin or with periodic acid-Schiff.
Remarkable physical examination findings included an altered mentation and mild to moderate ataxia consistent with a postictal state, cranial organomegaly and pain, and an enucleated left eye with a normal right eye (normal fundic examination OD with history of previous phacoemulsification OD March 15, 2011). A CBC revealed leukocytosis (16 × 109/L; reference range, 3.1–14.4 × 109/L) with 1+ toxic change and anemia (hematocrit, 29.4%; reference range, 40.3–60.3%). Red blood cell parameters were not reported due to lipemia. A serum biochemical analysis revealed an elevated alkaline phosphatase (6.9 ukat/L ; reference range, 0.33 −2.59 ukat/L), hypercholesterolemia (12.04 mmol/L mg/dL; reference range, 3.32–8.21 mg/dL), hypertriglyceridemia (18.44 mmol/L; reference range, 0.33–1.88 mmol/L), and hyperglycemia (28.19 mmol/L; reference range, 3.61–6.22 mmol/L). Pancreatitis was suspected based on the presence of abdominal pain, inappetence and CBC/serum biochemical analysis findings. Budding yeast were identified on urinalysis collected via catheterization. Abdominal ultrasonography revealed a 2 cm mass in the distal right limb of the pancreas. Brain MRI (T1- and T2-weighted axial and T1-weighted dorsal/axial/sagittal views with and without contrast) showed exaggerated meningeal enhancement consistent with inflammatory disease. Three-view thoracic radiographs revealed mild cardiomegaly (Table 2). All diagnostic imaging was performed by a board-certified radiologist.
A fine-needle aspirate of the pancreatic mass was performed and cytological evaluation was consistent with Cryptococcus spp. and pyogranulomatous inflammation. A Cryptococcus spp. antigen titer on cerebrospinal fluid and was positive at 1:39. Additional cerebrospinal fluid findings included mixed neutrophilic/eosinophilic/lymphocytic inflammation with a protein content of 60 g/dL (reference range, <30 g/dL). The dog was diagnosed with pancreatic cryptococcosis with neurologic dissemination and treated with corticosteroidh, triazolei (5 mg/kg PO q 12 hr), sulfonamide anticonvulsantj, and anticonvulsantk. No further seizures were observed.
Three months after discharge, the dog remained free of seizures. Repeat abdominal ultrasonography showed the pancreatic granuloma to be unchanged. Recheck Cryptococcus spp. antigen titer on serum was positive at 1:21. Six months after presentation, abdominal ultrasonography showed resolution of the previously identified pancreatic granuloma, and serum Cryptococcus spp. antigen titer was negative. As of November 2012, 15 months since initiation of therapy with triazole, the dog remained free of clinical signs, and the dog continued to receive triazole (5 mg/kg PO q 12 hr).
Discussion
This report describes two cases of canine abdominal cryptococcosis treated withfluconazole alone. One of those cases was a rare case of alimentary cryptococcosis without dissemination.
A previous publication described two dogs with abdominal cryptococcosis.7 One was a border collie infected with Cryptococcus neoformans/gattii (C. neoformans/gattii), and a giant schnauzer infected with C. neoformans/gattii. Both dogs in that report underwent surgical excision of the affected intestine and were initially managed with amphotericin B/fluconazole and were later switched to itraconazole/flucytosine due to limited availability of fluconazole. The border collie was hospitalized for 21 days and died acutely 4 days after discharge. A necropsy was not permitted. The giant schnauzer developed progressive neurologic disease that responded to phenobarbital and prednisone. Antifungal therapy was discontinued after 248 days, and the dog made a full recovery (In conversation with Malik, R. September 2011). In both cases, surgical resection was thought to be an integral part of therapy.7
Another case described a boxer with abdominal and disseminated cryptococcosis with gastrointestinal bleeding.8 C. neoformans var. grubii was identified in the mesenteric lymph nodes, colon, and by fecal cytology. That dog was treated with itraconazole again. Nonetheless, neurologic signs developed after 10 days, and the dog was euthanized due to progressive disease approximately 10–12 mo later (In conversation with Barger, A.M. September 2011).8
Additional reports have described abdominal cryptococcosis in dogs and cats.3,9,10 Unfortunately, those reports do not describe individual cases, making it unclear whether they had either primary alimentary or disseminated disease. Therapy was not described. In addition, a Doberman pinscher with gastric cryptococcosis caused by C. neoformans has also been described, but was euthanized before treatment.5
In the case of the boxer described in the current report, fecal examination revealed Cryptococcus spp. making it the second report utilizing fecal cytology as a means of identifying cryptococcosis.8 Additionally, this is the second case to describe anemia secondary to suspected gastrointestinal ulceration.8 This was also the second boxer described with abdominal cryptococcosis.8 Previous reports describe histiocytic ulcerative colitis and enteroinvasive Escherichia coli, protothecal colitis, and leproid granuloma in boxers.11–14 A possible explanation for abdominal cryptococcosis in boxers may lie in an underlying breed-specific immunodeficiency predisposing them to alimentary infections. More research is needed considering other reports have also shown overrepresentation of cryptococcosis in American cocker spaniels, Doberman pinschers, border collies, and German shepherd dogs.3,9
Unlike previous reports, the two cases described herein did not undergo surgery and were successfully treated with fluconazole alone. Previous reports have shown limited success with itraconazole, flucytosine and amphotericin B, and many developed neurologic disease in the face of treatment. Previous reports also recommend surgery as an integral part of therapy. The two cases in this report suggest that fluconazole should be strongly considered for the management of cryptococcosis.1,2
Fluconazole dosing is reported as a wide range.1,15 The boxer described above was initially treated with 6.25 mg/kg PO q 12 hr. At that dose, the Cryptococcus spp. antigen titer increased and additional lesions were identified. It is possible the titer increased secondary to the initiation of therapy because the antigen titer cannot differentiate between live and dead organisms.1,2 A higher dose (18.75 mg/kg PO q 24 hr, divided) proved to be effective in lowering the titer and reducing the size of the lesions. After increasing the fluconazole dose, the antigen titer decreased from 1:298 to 1:15 (>19-fold decrease) and ultimately became negative. The miniature schnauzer's antigen titer became negative at a dose of 5 mg/kg PO q 12 hr.
Limitations of this report include the lack of fungal culture and minimum inhibitor concentration testing and relatively short follow up. It would be ideal to biotype the Cryptococcus spp., because response to therapy may be dependent on the species causing infection.10 It is likely that both dogs were infected with C. neoformans given their excellent response to fluconazole. It is possible that cases of C. gattii would not respond as successfully. Finally, fluconazole is a fungistatic drug; therefore, longer follow up, including follow up after treatment has been discontinued, is needed to verify complete response. That said, both reported dogs are free of clinical signs and have negative antigen titers at 17 and 15 mo after initial presentation.
Conclusion
Although abdominal crytpococcosis is rare, it should be considered a differential diagnosis, especially in cases demonstrating fungal organisms in fecal samples. To the author's knowledge the described cases are the only cases of canine abdominal cryptococcosis successfully treated with a triazole agent alone, therefore negating the need for more high risk therapy. While monitoring is required, triazole therapy should be considered prior to surgical intervention.
Contributor Notes
L. Tangeman's present affiliation is Veterinary Specialty and Emergency Center, Thousand Oaks, CA.
