Systemic Mycosis in Three Dogs from Nonendemic Regions
Three dogs were examined for clinical signs ultimately attributed to systemic fungal infections. One dog was evaluated for chronic, ulcerated dermal lesions and lymphadenomegaly; one dog was examined for acute onset of unilateral blepharospasm; and one dog had diarrhea and hematochezia. Two of the dogs were diagnosed with blastomycosis (one with disseminated disease and the other with the disease localized to the left eye). The third dog was diagnosed with disseminated histoplasmosis. None of the dogs originated from, or had traveled to, typical regions endemic for these fungal diseases. All diagnoses were established from histopathology and either polymerase chain reaction (PCR) or cytology and culture. The two dogs diagnosed with blastomycosis were treated with either itraconazole or ketoconazole with apparent resolution of the infections. The dog with ocular involvement had an enucleation prior to beginning therapy. The dog diagnosed with histoplasmosis was euthanized without treatment. In patients with characteristic clinical features, systemic fungal infections should still be considered as differential diagnoses regardless of their travel history.
Introduction
Blastomyces is considered endemic to the Mississippi and Ohio river valleys, Missouri, the mid-Atlantic states, and parts of Canada (Ontario, Manitoba, and southern Saskatchewan).1–3 Histoplasmosis is found predominantly in the Missouri, Mississippi, and Ohio river valleys.2,4 Sporadic cases of histoplasmosis have also been reported in Canada (New Brunswick and Ontario), Australia, and southern California.5,6 The Pacific Northwest has not historically been known for geographic exposure to systemic fungal infections. Recently, however, Cryptococcus gatii, typically found in the subtropic and tropic regions of the world, was documented to be endemic to the Pacific Northwest.7,8 It was first identified in 1999 in British Columbia and again in 2003 in Oregon, Washington, Idaho, and California. Initial identification was made in wild animals and veterinary patients; additional clusters of human and veterinary cases, most with no apparent travel history, were recognized in the years that followed.7,8 This case report describes two cases of blastomycosis and a solitary case of histoplasmosis in dogs that neither traveled to nor resided in areas outside of the Pacific Northwest.
Case Report
Case 1
An 8 yr old male mixed-breed rottweiler weighing 27 kg from Port Clements, BC, Canada was examined for a 5 mo history of swelling of the left distal pelvic limb and ulceration on the plantar aspect of the interdigital region. Additionally, the owner reported a 2 mo history of an ulcerated mucocutaneous lesion on the dog’s left upper lip as well as a 1 day history of tachypnea and peripheral lymphadenomegaly. The dog was obtained as a juvenile at approximately 6–12 mo of age from a Prince George, BC animal shelter in Canada, but had not traveled out of the region since being rescued. Prior to referral, a tapering dose of prednisone (2 mg/kg per os [PO] q 12 hr for 5 days then 2 mg/kg PO q 24 hr for 16 days) was prescribed with initial marked improvement in the lesions noted. After a decrease in the prednisone dose, the dermal lesions reappeared and a left popliteal lymphadenomegaly developed. Difloxacina (7.4 mg/kg PO q 24 hr) was prescribed 4 days prior to referral with no improvement noted.
On referral, the dog was alert and responsive with a rectal temperature of 40°C, a heart rate of 120 beats/min, and panting. The dog was non-weight-bearing on the left pelvic limb. There was ulceration of the plantar interdigital space and swelling distal to the hock. A mild, serosanguinous discharge originated from the interdigital space. A slightly swollen, ulcerated lesion was located at the mucocutaneous junction of the left upper lip. The left mandibular and left popliteal lymph nodes were moderately enlarged and firm. No obvious fundic lesions were appreciated by direct ophthalmoscopy performed in a dark examination room; however, the entire fundus could not be completely visualized.
Abnormalities of a complete blood count (CBC) and serum biochemistry included lymphopenia (0.55 × 103/μL; reference range, 1.1–5.1 × 103/μL), eosinophilia (0.3 × 103/μL; reference range, < 0.1 × 103/μL), thrombocytosis (478 × 103/μL; reference range, 157–394 × 103/μL), elevated mean platelet volume (12.7 fL; reference range, 6.5–9.9 fL), elevated alanine aminotransferase (133 U/L; reference range, 21–67 U/L), elevated alkaline phosphatase (485 U/L; reference range, 14–72 U/L), and hyperglobulinemia (4.2 mg/dL; reference range, 2.7–3.8 mg/dL).3
On thoracic radiographs, there were numerous, variably sized, soft-tissue nodules throughout the pulmonary parenchyma with no evidence of intrathoracic lymphadenomegaly. On radiographs of the left hind foot there was marked soft-tissue swelling extending from the midmetatarsal region to the distal phalanges. Abdominal ultrasound did not reveal any additional lesions. Fine-needle aspirates of the left mandibular lymph node, left popliteal lymph node, lung parenchyma, and impression smears of the left upper lip and left interdigital lesions revealed numerous refractile, spherical, basophilic structures with a thick capsule consistent with a fungal infection. There was also a population of spindle-shaped cells that exhibited moderate anisocytosis and anisokaryosis, presumably secondary to inflammation (Figure 1). Fungal agar gel immunodiffusion (AGID) serology for antibodies to Aspergillus, Blastomyces, Coccidioides immitis, and Histoplasma capsulatum was negative. Cryptococcus latex agglutination was also negative. Aerobic culture was negative for growth. Results of fungal culture of aspirates from the lymph nodes, polymerase chain reaction (PCR), and sequence analysis of PCR products were all consistent with B. dermatitidis. Itraconazoleb (11 mg/kg PO q 24 hr) and carprofenc (2 mg/kg PO q 12 hr) were initiated. After 2 mo of treatment, the dog’s dermal lesions were improving and the dog was breathing normally. The dog was treated for 6 mo with itraconazole and had an infrequent cough 1.5 yr after receiving therapy. No further follow-up diagnostics were performed.
Citation: Journal of the American Animal Hospital Association 48, 6; 10.5326/JAAHA-MS-5799
Case 2
An 11 yr old castrated male golden retriever weighing 43.6 kg from Wenatchee, WA was examined for acute onset of blepharospasm and conjunctivitis of the left eye. The dog had no history of travel out of Washington State. Examination revealed moderate corneal edema, blepharospasm, conjunctival hyperemia, and aqueous flare. Fluoresceind staining was negative and no abnormalities were seen on direct ophthalmoscopy of the left eye, which was performed without pupillary dilation. No abnormalities were detected in the right eye. Prednisone acetate ophthalmic dropse were prescribed to be administered to the left eye 1 drop q 12 hr. Carprofenc (1.7 mg/kg PO q 12 hr) was also prescribed. The condition of the dog’s left eye continued to deteriorate, and 3 wk after initial examination, the dog developed crusted, circular, nonpruritic patches of alopecia on the caudodorsal trunk. Re-evaluation of the left eye revealed severe anterior chamber aqueous flare, corneal edema, circumferential neovascularization starting at the limbus and extending 3–4 mm into the cornea, hyphema, hypopyon, and an intraocular pressure of 51 mm Hg by applanation tonometry. Fluoresceinc staining of both corneas was negative. Abnormalities on CBC and serum biochemistry included eosinopenia (0.08 × 103/μL; reference range, 0.10–1.49 × 103/μL), hypercholesterolemia (337 mg/dL; reference range, 110–320 mg/dL), and hyperkalemia (6.1 mmol/L; reference range, 3.5–5.8 mmol/L). Referral to an ophthalmologist was recommended, and the frequency of administration of prednisone acetate ophthalmic drops was increased to 1 drop q 6 hr. Latanoprost ophthalmic solutionf (1 drop to the left eye q 12 hr) was also prescribed. Latanoprost, a prostaglandin analog, was used to address the elevated intraocular pressure. Retrospectively, this medication was not an ideal choice due to the presence of uveitis, but was unlikely to have affected the outcome of this case given the final diagnosis.
Hairs from the skin lesions were cultured on dermatophyte test medium. The culture was positive 10 days later, but the organism was not submitted for identification. Miconazole topical sprayg was applied q 12 hr. Ultrasonography of the left eye performed by the ophthalmologist revealed retinal detachment and intravitreal infiltrates and hemorrhage. The eye was enucleated and submitted for histopathology, which revealed pyogranulomatous inflammation and organisms consistent with B. dermatitidis. Results of immunohistochemistry (IHC) were consistent with a diagnosis of blastomycosis. On reevaluation 5.5 wk after initial presentation, the skin lesions had improved minimally. At that time, skin biopsies, thoracic radiographs, and blastomycosis serology were performed. Right lateral and ventrodorsal radiograph projections of the thorax were normal. The skin biopsies were consistent with dermatophytosis and blastomycosis serology was negative. Ketoconazoleh (9.2 mg/kg PO q 12 hr) was prescribed. Three months after starting therapy, the dog was reported to be doing well. This case was subsequently lost to follow-up.
Case 3
A 1 yr old spayed female Yorkshire terrier weighing 3.4 kg from Lewiston, ID was referred with a 2 mo history of hematochezia, melena, diarrhea, and lethargy. The dog did not have a travel history outside of the Lewiston area. Two months prior to referral, the dog was treated with metronidazole (18.3 mg/kg PO q 12 hr) by the referring veterinarian for hematochezia with no improvement. The dog began experiencing trembling episodes 1 mo after initial presentation and was treated with tramadol (3.7 mg/kg PO q 12 hr) and methocarbamol (18.3 mg/kg PO q 12 hr) for lumbosacral pain identified on physical examination. The dog initially improved, but its condition deteriorated 4 days later. A probiotic was prescribed at that time, and the dog was examined again 11 days later for progressing diarrhea and hematochezia. Abnormalities on a CBC and serum biochemistry performed by the referring veterinarian included leukocytosis (30.67 × 109/L; reference range, 6–17 × 109/L) characterized by a granulocytosis (26.71 × 109/L; reference range, 3–12 × 109/L), monocytosis (1.67 × 109/L; reference range, 0.2–1.5 × 109/L), anemia (hematocrit was 11.39%; reference range, 37–55%), thrombocytopenia (16 × 109/L; reference range, 200–500 × 109/L), hypoalbuminemia (2.1 g/dL; reference range, 2.5–4.4 g/dL), elevated alkaline phosphatase (279 U/L; reference range, 20–150 U/L), hyperbilirubinemia (0.9 mg/dL; reference range, 0.1–0.6 mg/dL), hypocalcemia (6.9 mg/dL; reference range, 8.6–11.8 mg/dL), hyponatremia (133 mmol/L; reference range, 138–160 mmol/L), and hyperkalemia (7.2 mmol/L; reference range, 3.7–5.8 mmol/L). Hypoadrenocorticism was suspected and an adrenocorticotropic hormone (ACTH) stimulation test was performed. The dog was administered dexamethasone sodium phosphate (1.2 mg/kg intramuscularly) and then prednisone (0.4 mg/kg PO q 12 hr) while awaiting test results. Pre- and post-ACTH cortisol values were normal (pre-ACTH cortisol was 3.4 μg/dL; reference range, 1.0–4.5 μg/dL, and post-ACTH cortisol was 13.0 μg/dL; reference range, 5.5–20 μg/dL). Within 24 hr, the hematocrit had decreased to 8.3%. The dog continued to experience hematochezia, diarrhea, anorexia, and remained depressed. A whole blood transfusion (17.6 mL/kg) was performed and prednisolone sodium succinate (29.4 mg/kg IV), cefovecin sodiumi (7 mg/kg subcutaneously), cimetidine (unknown dose), and IV fluids (58.8 mL/kg subcutaneously) were administered. The next day, the dog was brighter, ate a small amount of food, and the hematocrit was 31%. The dog deteriorated 2 days following the transfusion, at which point the hematocrit was 24.8%. The hyperkalemia resolved, but the bilirubinemia (1.5 mg/dL) increased and serum albumin decreased to 1.3 g/dL. The dog was then referred to the internal medicine service.
On physical examination, the dog was laterally recumbent and minimally responsive with a temperature of 38.2°C, a heart rate of 140 beats/min, and a respiratory rate of 70 breaths/min with increased effort. Thoracic auscultation was normal. The dog’s mandibular and superficial cervical lymph nodes were mildly enlarged bilaterally. The abdomen was mildly distended with no fluid wave appreciated, and melena was present on the perineum. Systolic blood pressure was 30 mm Hg and the hematocrit was 17%. A hetastarchj bolus (5 mL/kg IV), isotonic crystalloid fluidsj (90 mL/kg IV), and packed red blood cells (17.6 mL/kg IV) were administered. Examination of a blood film revealed 1–2 μm round to oval yeast-like organisms within neutrophils and monocytes (Figure 2). The owner elected humane euthanasia.
Citation: Journal of the American Animal Hospital Association 48, 6; 10.5326/JAAHA-MS-5799
Necropsy revealed intrahistiocytic fungal yeast consistent morphologically with Histoplasma spp. within the liver, lung, kidney, spleen, lymph nodes, intestine, and in the vasculature of the brain and heart (Figures 3A, B). PCR and sequence analysis of PCR products confirmed a diagnosis of H. capsulatum.
Citation: Journal of the American Animal Hospital Association 48, 6; 10.5326/JAAHA-MS-5799
Discussion
To the authors’ knowledge, there have been no known cases of either canine blastomycosis or histoplasmosis reported in dogs residing in, and with no travel history outside of, the Pacific Northwest. A solitary case report of a human with tibial osteomyelitis secondary to Blastomyces has been reported from British Columbia, but the individual in that report did not reside in the area from where the dog described in case 1 originated and the human patient had traveled to Ontario (known to be endemic for B. dermatitidis) on several occasions.9 An element in previous descriptions of the pathogenesis of these infections is that the organisms inhabit soil near bodies of water.1–5,10 All of the dogs described in this report lived near bodies of water, keeping with the suggested environmental requirements for these particular fungal organisms.
A definitive diagnosis of fungal infections is often made by identifying organisms in affected tissues.2–4 Tissue and blood samples from all three dogs harbored fungal organisms, aiding in the diagnosis. In cases with a high index of suspicion for fungal infection and for which identification of the fungal organism is not possible (or is unsuccessful), consideration can be given to serologic testing. The available serologic tests for blastomycosis are AGID and radioimmunoassay.2,3,11,12 The AGID detects antibodies against the A antigen of B. dermatitidis and has a reported sensitivity of 41–90% and a specificity of 96–100%.2,3 The AGID titer can be negative early in the disease and titers are variable after treatment. The radioimmunoassay detects antibodies against the WI-1 antigen of B. dermatitidis and has a reported sensitivity of 92% and a specificity of 100%, but is not available commercially.2,3,12 Further, because, titers remain positive for months to years following treatment, neither of the tests can be used to monitor response to therapy.2,3,12 Reliability of serologic testing for histoplasmosis is poor, yielding frequent false negative and positive results.2,4 All of the dogs in this series demonstrate the limitations of antibody-based serologic testing for these infections.
Recently, a new antigen-based enzyme immunoassay (EIA) has been developed for the diagnosis of blastomycosis and histoplasmosis.11 The test detects fungal cell wall galactomannan.11 The sensitivity of the urine antigen EIA is reported to be 93.5–96% for blastomycosis.11 Cross-reactivity between Blastomyces and Histoplasma occurs with this test.11 It has been reported that 96% of patients diagnosed with histoplasmosis tested positive with the blastomycosis urine antigen EIA.11 Finally, the antigen EIA level decreases with treatment making this test appropriate for monitoring response to therapy.11
Alternative supportive testing that can be performed either on cultures or formalin-fixed tissue samples are PCR and IHC. PCR sequencing was performed in two of the three fungal cases described herein. PCR can be used on either formalin-fixed tissue or cultures to assist in identifying the type of fungal organism present. Unfortunately, tissue blocks for performing PCR on case 2 were no longer available when preparing this report. Downfalls of PCR that can alter results are contamination, insufficient fungal DNA present in the sample for amplification, destruction of DNA secondary to formalin fixation, and inhibitory substances in submitted samples.13 Contamination would produce false positive results, whereas the others would produce false negative results.13 Further, the presence of nucleic acid alone, without other consistent clinical features, is not conclusive evidence of clinical infection. IHC can also be performed on formalin-fixed tissue to identify various fungal organisms. IHC was performed on samples from the dog in case 2. The specificity of IHC varies with each type of mycosis, and a great deal of cross-reactivity can occur among Aspergillus, Blastomyces, Coccidioides, and Histoplasma.14 Despite the cross-reactivity of IHC assays, when results are interpreted in conjunction with the clinical presentation and characteristic morphologic appearance of each organism (cytologically or histopathologically), positive IHC or PCR results give confidence to the diagnosis.13
A differential diagnosis considered for the dog in case 1, aside from disseminated fungal disease, was metastatic neoplasia. Atypical spindle cells noted on cytology (combined with the results of the thoracic radiographs, the age of the dog, and lack of travel history to endemic fungal regions) could have easily caused a disseminated fungal infection to be disregarded as a possible differential diagnosis. Had fungal organisms not been identified, the cytologic findings could have falsely suggested a diagnosis of neoplasia. In that case, the presence of spindle-shaped cells with moderate anisocytosis and anisokaryosis, in conjunction with the identification of fungal organisms on cytology, was attributed to inflammation. A similar scenario could apply to the other cases with more common differentials taking priority over a mycotic infection. Ocular trauma, bacterial infection, immune-mediated disease, and neoplasia were considered in case 2 and inflammatory bowel disease, immune-mediated disease, hypoadrenocorticism, and neoplasia were considered in case 3. Had the organisms not been identified on either histologic or cytologic samples, erroneous diagnoses could have delayed appropriate treatment and potentially resulted in the death of cases 1 and 2.
Conclusion
Blastomycosis and histoplasmosis should be considered as differential diagnoses in patients with compatible clinical features even when seen in nonendemic regions. Failure to consider systemic fungal infections as differentials resulting in either euthanasia or death of the patient prior to definitive diagnosis, combined with few states requiring reporting of various nonendemic mycotic infections, could contribute to the underrecognition of these infections in traditionally nonendemic regions. The differentials considered for a given case reflect integration of signalment, clinical signs, physical examination, diagnostic data, and travel history. Travel history is often used to help raise or lower the index of suspicion for various infectious diseases, but as illustrated by these cases, travel history may not always provide a reliable basis on which to eliminate differentials. With the extent of travel among people and their pets and changes in the environment due to human development and natural shifts in climate, diseases are shifting from their endemic regions and becoming more widely geographically distributed. Clinicians need to be mindful of these changes and maintain a broad appreciation of differentials for the problems their patients present with.
Photomicrograph of a fine-needle aspirate of the left popliteal lymph node containing six yeast organisms measuring approximately 8–15 μm in diameter with a thick, double-contoured cell wall and a basophilic granular center admixed with abundant strewn nuclear material and cellular debris. Broad-based budding was observed (inset) compatible with Blastomyces dermatitidis. Pyogranulomatous inflammation was observed in other areas of the slide. Wright’s-Giemsa stain, original magnification ×1,000.
A peripheral blood smear revealed a left shift with the majority of monocytes and fewer neutrophils containing round yeast organisms measuring 2–4 μm in diameter with a basophilic center and a thin surrounding halo compatible with Histoplasma capsulatum. Low numbers of extracellular yeast organisms were observed. Wright’s-Giemsa stain, original magnification ×1,000.
A: Photomicrograph of the Yorkshire terrier’s body of the spleen obtained on necropsy featuring a splenic artery and adjacent red pulp. The arterial lumen is packed with activated monocytes containing numerous, round to ovoid, 2–4 μm diameter H. capsulatum yeast organisms with a basophilic center surrounded by a clear halo. The adjacent red pulp is infiltrated by plasma cells and macrophages containing similar intracytoplasmic yeast organisms. Hematoxylin and eosin stain, bar = 25 μm. Inset: A replicate section stained with periodic acid-Schiff fungal stain that highlights the intravascular phagocytic cells containing magenta intracytoplasmic yeast organisms. Bar = 10 μm. B: Photomicrograph of the junction of the colonic mucosa, muscularis mucosa, and submucosa of the Yorkshire terrier. The lamina propria and submucosa are infiltrated by macrophages containing intracytoplasmic, 2–4 μm, magenta, H. capsulatum yeast organisms. Extracellular yeast organisms are also present. Goblet cells containing large, globular, magenta staining material are interspersed among the crypt epithelium. Periodic acid-Schiff stain, bar = 25 μm.
Contributor Notes
C. Pratt's present affiliation is Department of Veterinary Pathobiology, University of Missouri, Columbia, MO.
C. Pratt’s updated credentials since article acceptance are MS, DVM, DACVIM.
