Dermatitis, cellulitis, and osteomyelitis caused by Aspergillus nidulans in a horse with pituitary pars intermedia dysfunction

Abstract

Clinical and histologic examination of a 12-y-old client-owned Quarter Horse gelding with pituitary pars intermedia dysfunction revealed dermatitis, cellulitis, and osteomyelitis caused by Aspergillus nidulans, confirmed by a PCR assay. This novel presentation of a fungal disease in a horse was characterized by aggressive local invasion and failure to respond to all medical therapy attempted over a 1-y period. Treatments included systemic and topical antifungals, anti-inflammatories, and use of cellular matrices. Surgical excision was not attempted but should be strongly considered early in the disease process in similar cases if clean margins can be achieved. Postmortem findings were of locally aggressive disease with no dissemination.

Equine aspergilloses are opportunistic infections caused by species of the genus Aspergillus (A. flavus, A. fumigatus, A. nidulans, A. niger). These fungi inhabit and multiply in the environment, including stored grain, hay, leaves, compost piles, and decaying vegetation. ¹ Aspergillus spores can cause severe infections, mainly in immunocompromised horses, including those with endocrinopathies, such as pituitary pars intermedia dysfunction (PPID), ³ a history of prolonged administration of corticosteroids or antibiotics, or a debilitating disease. ¹⁷ Clinical signs vary according to the localization of the infection but may include weight loss, fever, lymphadenopathy, bronchopneumonia, guttural pouch infection, sinusitis, or rhinitis.^2,5,13 Upper respiratory infections can respond to medical and surgical management. In contrast, lower airway infections tend to have a poor prognosis because they fail to respond to treatment and are often extremely difficult to diagnose antemortem.^2,18 Aspergillus spp.–associated nodular skin diseases in the horse have been reported,^9,12 as well as a case of osteomyelitis in a joint post-steroid injection, ¹⁶ but localized invasive dermatitis, cellulitis, and osteomyelitis have not been reported in the horse, to our knowledge. We found no cases similar to our case in a search of PubMed and Google using the following search terms: equine, aspergillus, cellulitis, osteomyelitis, and dermatitis.

A 12-y-old, 600-kg, Quarter Horse gelding used for roping was referred with a 1-mo history of progressive, coalescing, soft, hairless, pruritic nodules of 0.5–3-cm diameter on the right medial gaskin. This horse had no known travel history outside the Sonoran Valley, AZ, USA. The gelding was kept on a dry lot, and no other horses were reportedly displaying similar signs. Diet consisted of Bermuda hay, a mineral mix, and a joint supplement. Histopathology results (Idexx) revealed a fungal granuloma.

On initial evaluation in January 2021 (Fig. 1A), the gelding had no overt clinical indicators of endocrinopathy, history of steroid administration, or evidence of debilitating disease. The gelding was in ideal body condition (5 of 9 Henneke score) ⁷ with heavy muscling; no other health issues were observed or reported. The horse walked and trotted soundly in a straight line on firm, even ground. A fine-needle aspirate (FNA) was obtained from the lesion and was reported cytologically as an infection by a dichotomous branching hyphal and yeast-form fungus. Differential diagnoses included cutaneous pythiosis (or another oomycosis), basidiobolomycosis, coccidioidomycosis, or conidiobolomycosis. The FNA was submitted for PCR assay (Texas A&M Veterinary Medical Diagnostic Laboratory, College Station, TX, USA) to identify the specific fungus involved. The gelding was empirically treated with compounded fluconazole (Covetrus; 14 mg/kg PO loading dose, then 5 mg/kg PO q24h) and compounded potassium iodide (Covetrus; 1 mg/kg PO q24h) until the fungal etiology could be determined. ¹⁷

Figure 1.

Multinodular dermatitis and cellulitis caused by Aspergillus nidulans infection in the right medial gaskin of a Quarter Horse. A. Firm hemorrhagic nodules present in the medial gaskin in January 2021. Bar = 2 cm. B. Nodules are enlarged and ulcerated with cellular crusts in March 2021. Bar = 2 cm. C. Irregular raised skin nodules in the medial gaskin; the underlying tibial cortex is expanded with pyogranulomatous osteomyelitis and bony proliferation in a March 2021 radiograph. Bar = 2 cm. D. Ulcerated raised skin nodules composed of coalescing pyogranulomatous dermatitis and cellulitis in May 2021. Bar = 2 cm.

In February 2021, the PCR assay result matched the Aspergillus nidulans clade with 100% identity. A panfungal PCR assay targeting the internal transcribed spacer (ITS) region and agarose gel electrophoresis yielded one band of DNA at ~350 bp. DNA was purified from the gel and sequenced. The resulting 346-bp contig sequence was analyzed with NCBI BLAST (https://blast.ncbi.nlm.nih.gov/Blast.cgi). The sample sent for analysis contained abundant fungal hyphae 4–6 µm in diameter that were septate and had acute-angle dichotomous branching. Within the center of the mycetomas, some of the hyphae had bulbous swellings of ~9 µm diameter. The treatment protocol was subsequently amended; fluconazole and potassium iodide were ceased and replaced with compounded itraconazole solution (Rood & Riddle; 6 mg/kg PO q24h); a compounded mixture of 1.5% ketoconazole and 0.96% rifampin (Phyco-Fixer; Rood & Riddle) was applied topically to the lesions daily. Phenylbutazone (VetOne; 2 mg/kg PO q12–24h) was also administered as required to treat pain and reduce inflammation. Surgical debulking was discussed, but medical management was elected.

In March 2021 (Fig. 1B), a thyrotropin-releasing hormone (TRH) stimulation test was performed with 1 mg TRH IV to assess pituitary function because a primary immunosuppressive disease was suspected given that the disease progressed in the face of reasonable medical treatment. The gelding’s baseline adrenocorticotropic hormone (ACTH) concentration was 3.96 pmol/L (18 pg/mL), which is equivocal (RI: <3.30 pmol/L [<15 pg/mL]); PPID is likely if ACTH is >8.8 pmol/L (>40 pg/mL) and the 10-min post-stimulation concentration was 100.1 pmol/L (455 pg/mL). When performed between January and May, a post-stimulation concentration >44 pmol/L (>200 pg/mL) is consistent with a diagnosis of PPID. The gelding was treated with pergolide mesylate (Prascend, Boehringer Ingelheim; 1 mg PO q24h). Given a newly noted grade 3 lameness (AAEP lameness scale), radiographs (dorsoplantar, lateral, dorsolateral-plantaromedial oblique; Fig. 1C) of the tibia were obtained and showed marked thickening and an irregular surface of the outer mid-diaphyseal cortex on the dorsal medial aspect of the tibia subjacent to the dermal nodules. Within the tibial cortex, there were areas of proliferation and lucency consistent with osteomyelitis. Surgical debulking of the skin nodules was again discussed, but not pursued. By May 2021 (Fig. 1D), all medications, except for pergolide daily and occasional phenylbutazone, were discontinued because the owner saw no response and was no longer willing to subject the horse to treatment.

In June 2021, the lesion was injected with EquusCell StemWrap D (AniCell Biotech) per the owner’s request. The client was provided anecdotal evidence from both the equine community and the purveying company of the antifungal and immune-boosting actions of EquusCell while providing a biological scaffold for healing. No peer-reviewed literature substantiated this claim. A single vial was reconstituted with sterile saline to a final volume of 8 mL; 2 mL was injected at the edges of the lesion subcutaneously at the 12, 3, 6, and 9 o’clock positions. A second vial was reconstituted with sterile saline to a final volume of 3 mL and injected immediately superficial to the site of osteomyelitis. A month later, the intralesional injections were repeated. Animotion Plus stem cells (AniCell Biotech) were also administered systemically (1 vial in 1 L of sterile saline intravenously). These medications were administered under the observation of the company’s representative and per their instructions.

By August 2021, the lesion showed no improvement radiographically or clinically, and palliative care was pursued until the owner appreciated the diminished quality of life. Over several months, the lesions gradually became persistently open and draining, necessitating euthanasia in December because of marked cellulitis and lymphedema. Until that time, the horse had a normal attitude, movement at the walk, and appetite. The gelding was euthanized following sedation with 300 mg of xylazine intravenously with 23.4 g of pentobarbital and 3 g of phenytoin (VetOne) and submitted for postmortem examination.

On gross postmortem examination, a 12 × 8 × 5-cm, ulcerated, multinodular, hemorrhagic, dark-purple to black, firm, immobile, cutaneous mass was present on the right medial gaskin with overlying serocellular crusting. On cut section, the mass was expanded by tan to dark-red to black, firm foci. The underlying tibial bone had a focally raised, irregular, thickened cortex with osteolysis (Fig. 2A). On gross examination of the brain, the pituitary gland was mildly enlarged and irregular. The remainder of the gross examination was within normal limits. Representative tissue samples were placed in 10% neutral-buffered formalin, processed routinely, and stained with H&E for light microscopic examination.

Figure 2.

Dermatitis, cellulitis, and osteomyelitis caused by Aspergillus nidulans in a Quarter Horse. A. The cortex of the right tibia is expanded by pyogranulomatous osteomyelitis, osteolysis, and bony proliferation (arrows). Bar = 2 cm. B. The cutaneous fungal granulomas are characterized by large numbers of neutrophils, epithelioid macrophages, multinucleate giant cells, lymphocytes, and plasma cells. H&E. Bar = 500 μm. C. A mycetoma is characterized by central clusters of fungal hyphae surrounded by large numbers of neutrophils, epithelioid macrophages, lymphocytes, plasma cells, and fibroplasia. H&E. Bar = 100 μm. D. Fungal hyphae are 5–15 μm in diameter, are non-pigmented, poorly septated, irregularly branched, and non–parallel-sided with bulbous swellings. Periodic acid–Schiff stain. Bar = 50 μm.

Histologically, the right medial gaskin mass consisted of multifocal, severe, pyogranulomatous dermatitis and cellulitis. Pyogranulomas were composed of many neutrophils, epithelioid macrophages, multinucleate giant cells, lymphocytes, plasma cells, and variably sized mycetomas with frequent intralesional clusters of fungal hyphae (Fig. 2B, 2C). Hyphae were 5–15-µm wide, non-parallel, non-pigmented, and poorly septated, with bulbous morphology and irregular branching (Fig. 2D). The cortex of the underlying right tibia was expanded by pyogranulomatous osteomyelitis with mycetomas and bony proliferation. The fungal hyphae within the bony lesions were identical to those in the overlying dermatitis and cellulitis lesions. The fungal granulomas were limited to the right medial gaskin and underlying tibial cortex and did not involve any other tissues. The pituitary pars intermedia had nodular expansion of up to 10 times normal diameter, with moderate compression of the pars distalis and the pars nervosa, consistent with the source of PPID. No other organs or systems were affected by fungal granulomas.

Fresh tissue from the fungal granuloma was submitted to the microbiology laboratory at the College of Veterinary Medicine, Midwestern University (Glendale, AZ, USA) and cultured on Sabouraud dextrose agar with gentamicin and chloramphenicol, inhibitory mold agar with gentamicin and chloramphenicol, and Sabouraud dextrose agar Emmons at 26°C until fungi were visible. Aspergillus section Flavi was cultured (MALDI-TOF MS score of 2.20). The fresh tissue was submitted for further fungal identification to the Molecular Fungal Identification Laboratory at the College of Veterinary Medicine, University of Florida (Gainesville, FL, USA) for fungal PCR assays. A panfungal PCR assay targeting the ITS region and agarose gel electrophoresis yielded 2 DNA bands of ~500 and 300 bp. The long subunit region (LSU) was also amplified, and agarose gel electrophoresis yielded 1 band of DNA of ~1,000 bp. In BLAST analysis of the resulting contig sequence, the sequence matched Aspergillus nidulans with 100% confidence.

Aspergillus spp. cause a range of clinical syndromes in horses, including guttural pouch disease, pneumonia, and skin infections, with variable responses to treatment.^{1 –5,9,12,13,19} In guttural pouch mycosis in horses, A. fumigatus, A. nidulans, A. niger, and A. versicolor are frequently implicated.^4,5,13 Some horses with aspergillosis of the guttural pouch are reported to respond to topical itraconazole, but the curative treatment is to remove the blood supply by embolization of the internal carotid artery.^4,5 Salpingopharyngostomy and topical antifungal application have also been reported. ²¹ Choice of treatment depends on location of the fungal plaque, risk of fatal hemorrhage, and ultimately, clinician discretion.

In 29 horses with pulmonary aspergillosis that were presented to a teaching hospital over a 25-y period, none survived, and in only 2 was fungal infection suspected antemortem. ¹⁹ In some of these cases, the disease was attributed to fungal translocation across a compromised gastrointestinal system and not direct inhalation of fungal material. ¹⁹ Treatment of pulmonary fungal granulomas is often unsuccessful because lesions are surgically inaccessible and unresponsive to systemic antifungals. ¹⁸

In contrast to respiratory infections, dermal lesions are rarely reported as a site of Aspergillus spp. infection in the horse. A. sydowii caused superficial dermatitis in a horse that responded to topical econazole. ¹⁴ In another report, A. versicolor was the cause of a dermal granuloma that extended into the subcutaneous tissue and persisted despite local surgical resection. ⁹ Osteomyelitis caused by A. fumigatus was reported in a joint that had been injected with steroids 2 mo before diagnosis ¹⁶ ; that horse survived with systemic administration of itraconazole, but was no longer athletic. ¹⁶

Collectively, the number of published reports is too few to make generalizations about the most appropriate or effective therapy other than 1) the virulence of the opportunist Aspergillus spp. seems variable and 2) the underlying immune function of the individual plays a role in recovery. Antifungal susceptibility testing would have been beneficial when designing our treatment strategy. To date, antifungal use in veterinary fields is extrapolated from papers discussing pharmacologic effect, ¹⁷ anecdotal experience, and from human medicine in which routine antifungal susceptibility testing only began in the past 20 y. ¹⁰ Recently, publications became available on the susceptibility of dermatophytes in horses.^8,20 In the future, this option may be available to identify appropriate therapy sooner and potentially improve the outcome of horses infected by fungal organisms.

A. fumigatus is the most common fungal agent causing disease in humans, followed by A. nidulans. ¹⁵ Infection in humans can be extrapulmonary and chronic, with a tendency to disseminate contiguously across anatomical planes similar to observations in our equine case. ¹⁵ In humans, invasive aspergillosis is a significant cause of morbidity and mortality in immunosuppressed individuals. ¹⁵ Similarly, the disease in horses is more common in equids with a suppressed immune system. Indeed, there are several reports of aspergillosis in horses with PPID or in horses with a history of corticosteroid administration.^3,19 Our case had no history of recent steroid administration but did have PPID. In horses with PPID, neutrophils have been shown to have reduced superoxide production, migration, chemotaxis, and adhesion, which results in an increase in opportunistic infections from 11% to 35%. ¹¹ We theorize that the infection in our case stemmed from an unnoticed puncture wound to the gaskin with subsequent environmental contamination; the infection was allowed to establish because of immunodeficiency from undiagnosed PPID.

Aspergillosis in humans is frequently refractory to standard antifungal therapy, either topically or systemically.^4,15 Similar to cases reported in humans and horses, our case did not respond to medical management in the form of local and systemic antifungal treatment. A cure in this horse could have potentially been achieved with total surgical resection, given that postmortem examination revealed that this infection was locally aggressive rather than disseminated from another primary site. This course of action was declined because the resection would have left a large wound down and into the bone, taking months to heal, assuming clean margins could have been obtained. Still, surgical resection should be strongly considered early during treatment for similar cases because a response to medical management alone appears poor. This horse was only on appropriate systemic antifungal therapy based on identification for 6–8 wk. A longer course after the PPID was controlled might have changed the course of disease.

In our case, ancillary immune function testing, such as basal ACTH, repeat TRH stimulation, and upper and lower airway evaluation might have been pursued to elucidate the clinical course and the immune status of the gelding over the year-long course of disease. TRH stimulation is the most sensitive and specific way to diagnose PPID in the horse between January and May. ¹¹ We could attribute no response (negative or positive) to the AniCell Biotech products used in our case. Initially, once-daily oral fluconazole and potassium iodide were selected empirically as antifungals given their excellent penetration, spectrum, and pharmacokinetics and pharmacodynamics in the horse. ¹⁷ Once the fungus was identified, oral itraconazole was selected because it has the best-reported activity against Aspergillus spp. ¹⁷ Itraconazole is lipophilic and has a high affinity for the lung, brain, kidneys, and skin. ¹⁷ Phyco-Fixer was used topically because it has generational anecdotal use for cutaneous pythiosis, which is caused by an invasive oomycete. ⁶ The treatments utilized might have offered some palliation, as evidenced by the progression following cessation, but offered no resolution.

Our case did not have any upper or lower respiratory signs, as commonly cited in the literature on equine aspergillosis.^1,2,3,18,19 Our case was a cutaneous and presumably opportunistic form.^9,12 Immunocompromise should be considered early in the course of disease given that it contributes to the establishment of infection.^3,15 Fungal infections in horses vary from the superficial dermis (e.g., Malassezia spp.) to pervasive systemic infections (e.g., aspergillosis, candidiasis, coccidioidomycosis, blastomycosis, cryptococcosis, pneumocystosis). ² Given the Arizona location, one of the original differential diagnoses was coccidioidomycosis caused by Coccidioides immitis or C. posadasii. ¹⁴ Accurate diagnosis of fungal infections is critical for selection of appropriate antifungal medication ¹⁰ ; fungal DNA is ubiquitous, therefore a panfungal PCR assay should be utilized with caution, and the diagnosis benefits from simultaneous culture and histopathology, as in our case. As modalities become more available, fungal susceptibility testing should be pursued, which may result in an appreciable response to systemic antifungal medication. ¹⁰

Our case had no systemic dissemination, indicating that aggressive surgical debridement might have been successful. Based on our experience with this gelding, we suggest ruling out dissemination before pursuing total resection of the lesion, if even possible based on location and margins. Rapid investigation and determination of etiology should be considered essential in similar cases to achieve any level of treatment success.