Clinical Characteristics and Outcomes of Rare Fungal Keratitis Caused by Verticillium dahliae

Purpose To observe clinical characteristics and treatment outcomes of fungal keratitis cause by Verticillium dahliae. Methods Clinical data of 7 patients diagnosed as fungal keratitis cause by V. dahliae were retrospectively analyzed. The clinical manifestations, mycology, in vitro antifungal susceptibility, treatment regimens and prognoses of the patients were evaluated. Results All 7 patients were farm worker, of which 5 cases were caused by plant trauma. The corneal ulcer had a round shape and a relatively limited range with the diameters mainly in the range of 2-7 mm. The stromal infiltration was mild, and had no pseudopodia, mossiness or endothelial plaques. Intact hyphaes were detected in corneal scrapings and confocal microscopy, isolates were identified by morphology and by sequencing the internal transcribed spacer region of ribosomal DNA. In vitro antifungal susceptibility testing showed that the most sensitive antifungal drug was Amphotericin B. In the 6 patients with an ulcer less than 2/3 of the corneal thickness, the ulcer healed after 18 days of antifungal treatment only in one eye. The other five patients underwent corneal ulcer debridement or conjunctival flap covering surgery. The remaining one patient with ulcer depth more than 2/3 of the corneal thickness underwent lamellar keratoplasty. Conclusion Fungal keratitis caused by V. dahliae has typical signs of a mild inflammatory response, and is not sensitive to antifungal drugs. It is recommended that patients undergo corneal ulcer debridement as soon as possible to promote rapid healing of the ulcers.


Introduction
Verticillium dahliae, a fungus in the phylum Ascomycota and the genus Verticillium, is a soil-borne pathogen that infects plant roots by forming hyphopodia; it causes severe verticillium wilt diseases, which lead to enormous financial losses in the production of cotton and other field crops [1][2][3]. Our clinical work finds that trauma caused by plants, such as scratches from cotton branches, may lead to fungal keratitis due to the infection of human corneal tissues with V. dahliae; such ulcers have unique biological characteristics, such as a limited ulcer area and mild stromal infiltration, but with slow healing and easy prolongation. A retrospective study was conducted to evaluate the clinical characteristics and treatment outcomes of fungal keratitis cause by V. dahliae, hoping to provide valuable information regarding early diagnosis and timely antifungal treatments in these circumstances.

Patients
This study was approved by the Institutional Review Board of Shandong Eye Hospital and adhered to the tenets of the Declaration of Helsinki. Clinical data of 7 patients diagnosed as fungal keratitis cause by V. dahliae from December 2018 to December 2020 in our hospital were retrospectively analyzed. The diagnostic criteria included [4][5]: (1) corneal scraping examination revealed fungal presence in smears; (2) In vivo confocal microscopy (HRT3; Heidelberg Engineering, Dossenheim, Germany) revealed reflective hyphae structures; (3) fungal pathogen culture showed that the colonies grew slowly and were white, powdery or fluffy; (4) the fungi were identified as Verticillium dahliae by DNA sequences.
The medical history included symptoms associated with the patient's complaint, the cause of the disease (e.g. foreign body scratching or plant trauma), the onset time , medication, and changes in disease condition. Best corrected visual acuity (BCVA), intraocular pressure (IOP), clinical characteristics were recorded. In vivo confocal microscopy was used to observe hyphal morphology. RTVue optical coherence tomography (OCT; Optovue, Fremont, California, USA) was used to clarify the depth of corneal ulcers. Amphotericin B, fluconazole, itraconazole, voriconazole, posaconazole, anidulafungin, caspofungin, micafungin and 5-fluorocytosine (provided by Shandong Boke Biological Co., Ltd.) were tested for vitro antifungal susceptibility, and the minimum inhibitory concentration (MIC) values were reported [6].

Demographics and clinical Characteristics
The patients were 3 males and 4 females, aged 47-74 years (mean, 57±9.0 years). All 7 patients were farm worker. Five cases were caused by plant trauma, of which 3 cases scratched by cotton branches and 2 cases by corn leaves, and the other two cases were without inducement. None of the 7 cases had a history of topical steroid use.
Foreign body sensation, red eye, photophobia, tearing, and decreased visual acuity in the affected eye were noted by all 7 patients. The onset time, and duration of hospitalization ranged from 7 to 20 days (mean, 13.8 ± 2.2 days), and 14 to 30 days (mean, 17.4 ± 1.8 days), respectively.
The corneal ulcer were all located in the center of the cornea and had a round shape and a relatively limited range. The boundary between the ulcers and the surrounding normal cornea was relatively clear. There were no typical manifestations, such as pseudopodia, mossiness, satellite ulcers, or endothelial plaques, and only one patient (with a ulcer approximately 7 mm in diameter) was complicated with a 3-mm hypopyon (Fig.1). The diameter of the ulcers was mainly in the range of 2-7 mm, with diameters less than 3 mm in 2 cases, between 3 and 6 mm in 4 cases, and more than 6 mm in 1 case. RTVue OCT showed that the ulcer was less than 1/3 of the corneal thickness in 5 patients, approximately half of the corneal thickness in 1 patient, and approximately 2/3 of the corneal thickness in 1 patient.
Intact separate and branched hyphae were detected in the 7 specimens (with 10% potassium hydroxide smears or calcofluor white staining). The hyphae were thin, with relevantly uniform diameters. No chlamydospores were detected (Fig.2).
Scanning of the ulcer area of the 7 patients using confocal microscopy showed the distribution of hyphae-like structures. The hyphae were moderately reflective, upright, and slender, and some had bamboo-like changes. Their diameters were mainly 2-3 μm, there were many branches, and the alignment was disordered. The angles between the hyphae were small, mostly acute. The field of vision with densely distributed hyphae was clean, and some spore-like structures were visible (Fig.3).

Mycology
Seven strains of filamentous fungi were isolated from the 7 specimens submitted for examination, with a positive culture rate of 100%. The colonies grew slowly and were white, powdery, or fluffy, and the back was colorless or light orange. Microscopically, the conidiophores were slender and solitary or had multilayered branches with spiral growth (verticillated) and sharp ends. The angles between the conidiophores and the hyphae were mostly acute, and most of the phialides had no basal septum. The conidia were single-celled and transparent, with smooth walls and an elliptical to oval shape, and grew singly or in clusters at the end of the phialide (cephalophores) (Fig.4).

Treatment and Outcomes
All seven patients were hospitalized and received 0.2 mg/ml intravenous drip once a day, and topically polyene (5% natamycin eye drops or 0.25% amphotericin B eye drops) and imidazole (0.5% fluconazole eye drops or 10mg/ml voriconazole eye drops).
Surgical treatment was used when drug therapy was shown to be ineffective after approximately one week. Ulcer debridement was performed when the ulcer depth was ≤1/3 corneal thickness. Once the depth reached 1/2 corneal thickness, conjunctival flap could be combined. Lamellar keratoplasty (LKP) was chosen when the ulcer depth was more than 2/3 the corneal stroma but not reach the corneal endothelium.
The ulcer in one patient healed after 18 days of antifungal treatment, and the best corrected visual acuity improved by 4 lines compared to the pretreatment assessment.
Four patients underwent corneal ulcer debridement combined with intrastromal injection of 10 mg/ml voriconazole, followed by antifungal treatment, and the ulcer healed after an average time of 7.5 days. One patient underwent conjunctival flap covering surgery followed by antifungal treatment, and the ulcer healed after 8 days.
The corneal ulcer depth of 1 patient was more than 2/3 of the corneal thickness, and the condition did not improve after 1 week of medical treatment. The patient underwent lamellar keratoplasty, with no infection detected at the postoperative follow-up visits (Fig.1).

Discussion
Fungal keratitis is the leading cause of infectious corneal disease worldwide, especially in some developing countries with warm and wet climates [7][8][9]. More than 100 species have been reported as pathogens of fungal keratitis, Fusarium species are the most commonly isolated pathogens, followed by Alternaria and Aspergillus species, in Shandong Province, China [10][11]. In the recent years, the incidence of uncommon fungal keratitis caused by rare species with diverse morphology has greatly increased [12][13][14][15][16][17][18][19]. The current study described the first confirmed series of fungal keratitis cause by V. dahliae with detailed descriptions of clinical characteristics and treatment outcomes.
V. dahliae, known as soil-borne pathogens that causes vascular wilt diseases in a wide range of plant hosts [1][2][3]. Therefore, agricultural environments and trauma caused by plants are the most common factors associated with keratitis caused by V.
dahliae. The onset of keratitis caused by V. dahliae is slow, ranging from 7 to 20 days (mean: 13.8 ± 2.2 days). The most pronounced eye signs were that a moderate inflammatory response at the ulcer and the absence of typical manifestations, such as pseudopodia, mossiness or immune ring. The ulcer diameters were small, and 4 patients in this study had ulcers with diameters in the range of 3-6mm. The range of the ulcers was relatively limited, and the boundary with the surrounding normal cornea was relatively clear. The above characteristics were consistent with the slow growth of colonies and the limited colony expansion found in fungal culture.
Additionally, confocal microscopy examinations showed that the hyphae of V. dahliae were moderately reflective and thin, their diameter was mainly 2-3μm, and their thickness was uniform; the angles between the hyphae were small, mostly acute; and chlamydospore-like structures were rarely detected, which can be used for preliminary differentiation from Fusarium, Aspergillus, etc. [20][21].
In vitro antifungal susceptibility testing showed that Amphotericin B was the optimal option for treating keratitis caused by V. dahliae (MIC 2->8). Unfortunately, Amphotericin B eye drops are not commercially available in China at present, which greatly limits the clinical application of Amphotericin B. On the other hand, keratitis caused by V. dahliae is less sensitive to commonly used antifungal drugs, such as fluconazole (MIC>256) and voriconazole (MIC 4->8). Therefore, in this study, one patient who received drug treatment had a prolonged condition for 18 days before the ulcer healed completely. Conversely, in the 5 patients who underwent corneal ulcer debridement and conjunctival flap covering surgery, the ulcers healed rapidly, at 7-8 days after surgery. We believe the reason for the different healing durations is that the corneal ulcers caused by V. dahliae infection were small, and the depth of the infiltration was shallow (the depth of the corneal ulcer in 6 patients was less than 1/3 of the corneal thickness), which is a perfect indication for corneal ulcer resection.
Corneal ulcer debridement removes the superficial fungal infection of the cornea, which shortens the ulcer healing time and reduces the disease course [22].
In summary, keratitis caused by V. dahliae has typical signs of a mild inflammatory response, small ulcers, a limited ulcer range, and shallow infiltration and is not sensitive to antifungal drugs. Therefore, the disease is prone to prolongation and heals slowly. It is recommended that patients undergo corneal ulcer debridement as soon as possible to promote rapid healing of the ulcers.