We read with great interest the article titled “Intracameral Voriconazole Injection in the Treatment of Fungal Endophthalmitis Resulting from Keratitis,” by Ying-Cheng Shen and associates, which reported that intracameral voriconazole proved to be effective in the management of patients with fungal keratitis and endophthalmitis. We would like to share our observations regarding the article.

The authors have not defined endophthalmitis and cultured anterior chamber aspirates from the area of endothelial exudate. Endophthalmitis is defined as an intraocular inflammation predominantly involving the vitreous cavity and the anterior chamber of the eye. Obtaining a vitreous sample is important because infection can disseminate into the vitreous cavity without clinical evidence of the same.

Fungal filaments can penetrate the cornea and multiply in the anterior chamber. In our experience, presence of corneal infiltrates or of endothelial or anterior chamber exudates does not indicate involvement of the vitreous cavity and endophthalmitis. In this study, there is no mention about the status of the vitreous cavity. If vitreous samples were cultured, it would have confirmed the diagnosis of endophthalmitis.

The dose of intracameral voriconazole used in this study was 100 μg/0.1 ml (1000 μg/1 mL). The safety of this dose is debatable. Repeated intracameral injections of voriconazole may be toxic to the corneal endothelial cells if the concentration exceeds 250 μg/mL.

Two patients also received topical voriconazole. Topical voriconazole has been shown to have a very good penetration with a good aqueous concentration. It therefore is unclear whether the topical or the intracameral drug helped resolve the infiltrate.

The standard technique of identifying fungus involves examination of the conidiophores after sporulation with lactophenol cotton blue. The authors did not described the technique used, and histopathologic examination would not provide information on the identity of the organism, because all filamentous fungi look the same.

There is a paucity of clinical data in the Table and the Results. The following factors would have a bearing on the outcome and ought to have been mentioned: duration of patient symptoms, associated systemic illnesses such as diabetes, duration of medical therapy before injections, time taken to resolution after injection for each case, and treatment regimen after keratoplasty.

Both clinical images were of patients who had undergone penetrating keratoplasty, in which cases the bulk of the infection would have been removed during the surgery and the resolution could not be attributed completely to the drug. The authors also should have provided a clinical photograph of a patient with corneal infiltrate that resolved with scarring.

In conclusion, it is possibly inappropriate to conclude from this study that intracameral voriconazole may be used for fungal endophthalmitis resulting from keratitis when the diagnosis of endophthalmitis has not been defined well. Although intracameral voriconazole seems to be a promising method of treatment for intracameral invasion of fungus from the cornea, prospective studies are warranted.