A case of Candida albicans-induced fungal keratitis in a Chihuahua dog – with a focus on optical coherence tomographic features

Abstract

A 4-year-old Chihuahua dog was referred for bilateral corneal ulcers. Slightly raised white fluorescein-positive plaque-like corneal lesions in both eyes appeared as intense hyperreflective areas with posterior shadowing on optical coherence tomography (OCT). Based on corneal cytology and culture, Candida albicans-induced fungal keratitis was diagnosed. Despite treatment, on OCT, endothelial plaques, increased stromal infiltration thickness, vertical shapes of the ulcer edge, and necrotic stromal space were judged to be aggravation of the disease, and surgery was performed. Conjunctival grafting surgery with topical 1% voriconazole effectively resolved fungal keratitis. OCT can provide detailed and objective information related to the disease prognosis.

INTRODUCTION

Fungal keratitis in dogs occurs much less frequently than in horses [1,2] because the living environment and eye anatomy of dogs make them less exposed to potential fungal pathogens and less vulnerable to corneal trauma [3]. Nevertheless, canine fungal keratitis has been increasingly reported in recent veterinary literature [3].

Various causative agents of canine fungal keratitis have been reported, with Aspergillus spp. being the most common. Acremonium, Alternaria, Candida, Cephalosporium, Cladosporium, Curvularia, Fusarium, Hormographiella, Malassezia, Penicillium, Pseudallescharia, and Scedosporium spp. have also been occasionally reported [1,3].

A differential diagnosis between fungal and other forms of infectious keratitis, especially in the early stages, and an objective assessment of disease severity is sometimes difficult. These problems lead to delayed or improper treatment and disastrous outcomes such as corneal perforation, endophthalmitis, and permanent blindness [4]. Therefore, timely and definite diagnosis is important for disease management and better clinical prognoses [4].

Standard diagnostic methods include sample collection, microscopic evaluation, and microbial culture following clinical presentations suggestive of fungal infection [1,2,3]. Advanced diagnostic imaging modalities, such as in vivo confocal microscopy and optical coherence tomography (OCT), provide clinicians with detailed structural features of the cornea in various pathological conditions non-invasively and in real time [5,6]. The usefulness of in vivo confocal microscopy for diagnosing fungal keratitis and monitoring treatment response has been reported in humans and dogs [2,4]. Some reports have described OCT features and evaluated their clinical relevance in the diagnosis and prognosis of human fungal keratitis [7,8,9]. However, the utility of OCT in canine fungal keratitis has not yet been reported. This case report aimed to describe the clinical aspects of a dog with Candida albicans-induced fungal keratitis, with a focus on OCT observations.

CASE PRESENTATION

A 4-year-old castrated male Chihuahua dog was diagnosed with a corneal ulcer immediately after falling onto sandy ground a month ago, and the lesion appeared to be well treated with topical antibacterial agents in the referring hospital. Two weeks later, topical 0.1% dexamethasone eye drops were prescribed for 1 month to alleviate corneal scars. However, the dog was eventually referred to our clinic for ophthalmic evaluation, with a 1-day history of blepharospasm and redness in both eyes. At the first presentation to our clinic, slit-lamp biomicroscopy revealed white raised axial corneal epithelial plaques with fluorescein dye retention and stromal infiltrates in both eyes (Fig. 1A and B). Spectral-domain OCT (iVue 100, Optovue Inc., USA) showed corneal epithelial plaques as superficial intense hyperreflective areas with posterior shadowing, which precluded detailed evaluation of the structural integrity of the deep cornea (Fig. 1A and B). In the region adjacent to the epithelial plaques, stromal infiltrates were shown as ill-defined hyperreflective areas (Fig. 1A and B). In the right eye, an endothelial plaque was shown as a well-defined hyperreflective area beneath the endothelium, and the boundary between the corneal endothelial surface and plaques was discernible (Fig. 1A). In contrast, it was not observed in the left eye (Fig. 1B). The corneal thickness of the lesion was approximately 840 μm in the right eye and 730 μm in the left eye, and the thickness of the stromal infiltration ranged from approximately 380 to 390 μm in both eyes (Fig. 1A and B).

Fig. 1. Photographs of slit-lamp biomicroscopy and corneal OCT at the first presentation. (A) OCT image of the right eye showed epithelial plaques as an intense hyperreflective area (white arrow) generating posterior shadowing. A clear boundary between the corneal endothelial surface and plaque is shown (white arrowheads). (B) OCT image of the left eye showed epithelial plaques (white arrow) but no endothelial plaque. (C and D) Cytological photographs from a corneal scape smear stained with Diff-Quik. Fungal pseudohyphae and numerous round to oval shapes of yeast are depicted. The red arrows indicate a constriction at the pseudohyphae septum. 100× objective.

OCT, optical coherence tomography.

Corneal scrapes were taken from the ulcer edge for cytology using Diff Quik staining. Cytological evaluation showed fungal pseudohyphae and numerous yeasts, suggesting fungal infection (Fig. 1C and D). Medical treatment was initiated with topical 0.5% moxifloxacin (Vigamox, Alcon Singapore Manufacturing Pte. Ltd., Singapore) every 2 h, oral doxycycline 5.0 mg/kg twice a day (Doxycycline Hyclate, Kukje Pharm), and meloxicam 0.1 mg/kg once a day (Meloxifen^®, Kukje Pharm., Korea). Additionally, commercially available voriconazole powder (Vorico Injection 200 mg, Chong Kun Dang Pharmaceutical Corp., Korea) for intravenous injection was diluted with sterile water to prepare a 10 mg/mL (1%) solution [10], and topical off-label administration of the 1% voriconazole solution was initiated every 2 h.

Despite these treatments, the area of the epithelial plaques in the right eye increased and protruded further forward 2 days after the first examination (Fig. 2A). However, clinically significant changes were not found in the left eye, except for a slight increase in the total and stromal infiltration thickness (Fig. 2B). Four days after the initial examination, corneal perforation and iris prolapse were observed in the right eye. On the same day, a 360-degree conjunctival grafting surgery was performed in the right eye. Oral itraconazole 5.0 mg/kg twice daily (Ashicona; Ashish Life Science Pvt. Ltd., India) was added to the previous treatment regimen.

Fig. 2. Photographs of slit-lamp biomicroscopy and corneal OCT of both eyes at 2 days and 6 days after the first examination. (A) The area of the epithelial plaques in the right eye increased and protruded further forward 2 days after the first examination, and the corneal structures beneath the epithelial plaques were not visible in the OCT image. (B) Significant OCT changes were not found in the left eye, except for a slight increase in the total and stromal infiltration thickness 2 days after the first examination. (C and D) Six days after the first examination, detailed corneal structures were seen in the OCT images of the left eye after the epithelial plaques were peeled off, including necrotic stromal spaces (white star mark), a vertical shape of the ulcer edge (yellow dotted arrow), and a clear boundary between the corneal endothelial surface and plaques (white arrowhead).

OCT, optical coherence tomography.

Six days after the first examination, diffuse corneal edema and severe hypopyon were revealed, and the epithelial plaques were peeled off from the corneal surface in the left eye, exposing the remaining corneal structure (Fig. 2C and D). OCT revealed necrotic spaces as areas lacking any reflectivity, and the remaining corneal thickness ranged from approximately 100 to 330 μm (Fig. 2C and D). The ulcer bed was shown as ill-defined margins with a vertical or undermined shape of the ulcer edge in the OCT images (Fig. 2D). Endothelial plaques were observed as a hyperreflective zone beneath the corneal endothelium, which was not observed in the previous OCT examination (Fig. 2C and D). We also performed conjunctival pedicle grafting surgery in the left eye, considering the risk of corneal perforation due to the rapid and profound progression of corneal stromal loss. Corneal samples from the ulcer beds were submitted to a referral laboratory for fungal and bacterial culture tests. The referral laboratory reported that C. albicans isolate grew on Sabouraud dextrose and blood agar; however, the bacterial culture was negative. Corneal cytology revealed no fungal material 7 days after the graft surgery. Ocular manifestations, such as hypopyon, corneal ulcer, and corneal edema, resolved dramatically, and corneal transparency improved, except at the conjunctival graft site. At the last follow-up examination, 12 months after surgery, the fibrous tissue had largely receded, and the corneal scar had diminished; furthermore, the conjunctival grafts provided sufficient thickness and structural integrity (Fig. 3A and B). The vision of both eyes was well maintained without complications, while the grafts were placed in the axial cornea and were slightly pigmented (Fig. 3A and B).

Fig. 3. (A) Photographs of slit biomicroscopy at 12 months postoperatively. Axial corneal scar and pigmentation are shown, and fundus reflection is partially visible on retroillumination. (B) Optical coherence tomography images of both eyes at 12 months postoperatively.

DISCUSSION

Although OCT cannot completely replace classical diagnostic methods, such as slit-lamp biomicroscopy, to perform clinical examinations and monitor disease progression, it can provide more detailed and objective clinical information, as described in our case [8,9]. Some studies have reported quantitative and qualitative OCT features of the anterior segment, including the cornea and anterior chamber, in humans with infectious keratitis and have evaluated the relationship between these OCT findings and their clinical relevance [7,8,9]. Assessment of changes in total corneal and stromal infiltrate thickness demonstrated an association between a decrease in these thicknesses and favorable clinical outcomes in infectious keratitis [7,9]. In our case, the total corneal and infiltrate thickness in areas adjacent to the epithelial plaques increased slightly along with clinical deterioration. The detailed shapes of the ulcer edge, as well as the dimensions of a lesion, can also be analyzed using OCT. A previous study on human fungal keratitis stated that the ulcer edge could be classified into 3 types depending on the cross-sectional shape: sloping, vertical, and undermined edges [9]. Wound edge characteristics are well known to provide etiological and prognostic clues in human medicine [9,11]. Although sloping edges have a greater chance of healing, vertical or punched-out edges are considered difficult to heal and require surgical intervention for appropriate treatment [9,11,12]. A previous human study evaluating the usefulness of spectral-domain OCT in objective assessment and monitoring of patients with fungal keratitis reported that a sloping edge was observed in most cases. However, a subgroup analysis was not performed to compare the differences in the shapes of the ulcer edges between perforated and non-perforated eyes [9]. In our case, OCT showed vertical edges of the ulcerative lesion after the epithelial plaque in the left eye had peeled off. Because other clinical signs revealing a poor prognosis coexist, further studies are required to explore the association between ulcer edge shape and prognosis. Endothelial plaques can be interpreted as indicators of severe corneal infection with microbes and invasion of microbial elements into the anterior chamber [8]. Furthermore, a previous study investigated the characteristics of the boundary between the corneal endothelial surface and plaque according to causative agents and analyzed the association with disease severity [8]. It is sometimes difficult to perform a detailed assessment using slit-lamp biomicroscopy [8].

An irregular and unclear endothelial surface is more frequently observed in fungal keratitis than in bacterial and viral keratitis, and patients with unclear boundaries may require more intensive treatment such as surgical intervention [8]. In the current case, the boundary between the corneal endothelial surface and the plaques was clearly identified in both eyes (Fig. 1A). However, contrary to the results of a previous study [8], medical treatment with topical 1% voriconazole and systemic itraconazole was unsuccessful, and surgical intervention was eventually necessary. This contradiction should be evaluated in a large-scale study. Necrotic spaces were reported to be highly likely to lead to corneal perforation among the risk factors identified in a previous study [9]. In our case, necrotic spaces were observed in the OCT images after the epithelial plaques were peeled off. Concurrent with stromal necrosis, clinical deterioration manifested as dense stromal infiltrates, hypopyon, and prominent retrocorneal plaques, necessitating surgical intervention.

In general, corneal trauma is considered the predominant predisposing factor for fungal keratitis in humans, although the prevalence of fungal species causing keratitis varies depending on geography, climate, and socioeconomic factors [4]. In contrast, a history of trauma has not been reported as frequently as expected in cases of canine fungal keratitis cases [1,3]. This difference may be attributed to owners’ unawareness of minor corneal trauma [3]. An unusual aspect of our case was that a catastrophic clinical course occurred one month after evident trauma. This clinical peculiarity may be partly explained by several reported effects of corticosteroids on fungal keratitis [13]. With regard to this, another identifiable predisposing factor in our case was the use of topical corticosteroids over a month before admission to our clinic. Candida species as an opportunistic pathogen can cause a corneal infection in the immunocompromised host [4,13]. Specifically, the previous use of topical or systemic corticosteroids is the most common risk factor in humans and dogs [3,13,14,15]. Topical steroid use can conceal clinical signs suspicious for infectious keratitis owing to their anti-inflammatory effects, especially in the early stages of fungal keratitis, thereby delaying proper diagnosis and treatment [13]. Therefore, a more detailed and objective evaluation of corneal status, which may be delivered via advanced diagnostic imaging modalities, is crucial for accurate diagnosis and better prognosis.

In conclusion, although limited to the fungal form of infectious keratitis, the present case report describes detailed OCT features of the cornea observable in canine fungal keratitis, including the depth of stromal infiltration, shapes of the ulcer edge, stromal integrity, and aspects of the boundary between the endothelial surface and plaques. To the best of our knowledge, this is the first report of OCT features in a plaque-forming C. albicans-infected cornea. More in vivo details, which high-resolution OCT can provide, would be helpful in determining the severity of the disease and the timing of surgical intervention in similar future cases. It would be worthwhile to conduct further studies to evaluate the relationship between various OCT parameters and their clinical relevance in infectious keratitis of various microbial origins.