We appreciate the interest in and comments about our article. Diagnosis of infectious endophthalmitis is established by identifying an effective pathogen after obtaining intraocular specimens for culture and microscopic study. Postkeratitis fungal endophthalmitis tends to be more localized with the fungal mass, and invasion of fungal filaments into the vitreous may be uncommon. Therefore, fungal endophthalmitis from keratitis is diagnosed mainly based on positive culture results from an anterior chamber tap. Occasionally, fungus may invade the posterior chamber, and culture of vitreous samples should be performed in case of suspected vitreous invasion.
To identify the pathogen, intraocular specimens should be inoculated promptly directly onto culture media. In addition to culture, biopsy may be necessary in case of apparent microbial infection when repeated cultures demonstrate negative results. The specimen generally is split in 2 pieces, so that tissue can be evaluated by both histopathology and microbiology. Histologic examination provides information about infectious microorganisms and their particular morphologic patterns, such as aspergillosis and mucormycosis.
In our experience, repeated intracameral injections of voriconazole (100 μg/0.1 mL) may not be toxic to the corneal endothelial cells. First, voriconazole (100 μg/0.1 mL) as an additive for donor cornea tissue has been shown to cause no human corneal endothelial cytotoxicity. An in vitro study has shown that no corneal endothelial toxicity was detected after 30 days of treatment with 250 μg/mL voriconazole. Second, in the current series, injections of 100 μg/0.1 mL voriconazole into the anterior chamber once daily were shown to be efficient in eradicating fungi, and no adverse effects were noted in the follow-up period. Third, with the normal volume of the aqueous humor in humans assumed to be 0.5 mL, the injected dose of 100 μg/0.1 mL in the anterior chamber resulted in an initial aqueous concentration of 166.7 μg/mL. Furthermore, voriconazole in the anterior chamber is eliminated rapidly. In a previous study, the aqueous humor voriconazole concentration in rabbit eyes showed a short half-life of 22 minutes.
We also agree with their comments that both topically administered voriconazole and therapeutic keratoplasty surgery can be used in the treatment of fungal endophthalmitis. Topical voriconazole has been shown to achieve good corneal penetration. The mean aqueous concentration of voriconazole in rabbit eyes was maintained in a range from 19.97 to 23.70 μg/mL after the maintenance dose. In our study, topical voriconazole was administered for only a short period, and the treatment was discontinued because it seemed to do little to improve the keratitis. The excellent aqueous penetration of topical voriconazole may make it a useful adjunct in the treatment of fungal endophthalmitis in the anterior chamber.
Therapeutic penetrating keratoplasty effectively can remove or debulk the infected cornea and the fungi infiltrating in anterior chamber, whereas antifungal agents administered by topical use and intracameral injection is necessary to eradicate the residual fungi in the anterior segment. Postoperative regimens included 5% topical natamycin every half hour, 200 mg systemic ketoconazole twice daily, and intracameral voriconazole injections. Use of topical corticosteroids was usually delayed by several weeks.