We read with great interest the article by Shen and associates describing intracameral voriconazole injection in the treatment of fungal endophthalmitis resulting from keratitis. The authors proposed that intracameral voriconazole injection may be an effective treatment for fungal endophthalmitis contiguously spreading from keratitis. We appreciate this report by authors, which provided us with valuable information on new, challenging treatment methods for endophthalmitis spreading from fungal keratitis. Although the study was simple and concise, we have some concerns regarding the study they presented.

The final visual outcomes ranged from hand movements to no light perception in Patients 1, 2, and 5. We are concerned to know whether the authors ruled out possible posterior segment involvement by using ocular echography in these patients. Ocular echography is a useful instrument to evaluate posterior segment in endophthalmitis, especially in eyes with opaque media such as fungal keratitis. Similar to classification of bacterial endogenous endophthalmitis reported by Greenwald and associates, most cases of fungal endophthalmitis resulting from keratitis are the types of anterior focal or diffuse forms with or without posterior focal endophthalmitis. In such types of fungal endophthalmitis, intracameral voriconazole injection may be an effective treatment; moreover, repeated injections of intracameral voriconazole also may prevent organisms from spreading to the vitreous cavity. However, in fungal keratitis with posterior diffuse or panophthalmitis, intracameral voriconazole injection may be ineffective because of a low concentration of intravitreal voriconazole. Therefore, intravitreal voriconazole with or without pars plana vitrectomy should be performed in severe cases with posterior diffuse involvement. The authors should clarify whether the poor visual outcomes are related to posterior segment involvement.

First, in the Methods, the authors should provide the results of vitreous cultures and lens status for the patients. Intraocular specimens were obtained only from aqueous humor; however, results of vitreous cultures may indicate whether intravitreal antibiotics should be administered, especially in mildly anterior focal vitreous inflammation. Second, the lens also plays an important role for preventing organisms from invading the posterior segment. Therefore, the authors should illustrate whether eyes with aphakia and pseudophakia have a higher incidence of posterior segment involvement and poor visual outcomes than those with phakia.

In the Discussion, we would like to point out a minor error in the authors’ assumption of the normal volume of the anterior chamber. In fact, the normal volume of the anterior chamber in humans is 0.2 mL, rather than 0.5 mL, as reported in this study. Therefore, the injected dose of 100 μg/0.1 mL should result in an anterior chamber concentration of 333 μg/mL or more after aqueous tapping for cultures. The peak levels achieved in the anterior chamber thus were more than 300 times the minimum inhibitory concentration of voriconazole to Candida and Aspergillus species.

Final visual outcomes are related to posterior segment involvement in patients with endophthalmitis. If the authors provide more supplementary data such as echographic findings, results of vitreous cultures, and lens status, this study is a valuable reference for clinical practices.