We read with great interest the article by Lingappan and associates describing endogenous fungal endophthalmitis. Their study provided us with valuable information on endogenous fungal endophthalmitis, which we appreciate. However, although their study was comprehensive, we have some concerns regarding the presented findings.

First, the most important element of the diagnosis is maintaining a high index of suspicion for endogenous fungal endophthalmitis. For patients with an identified systemic fungal infection and ocular inflammation symptoms consistent with intraocular fungal infection, a clinical diagnosis generally can be made without intraocular specimens. However, when a systemic fungal infection is not apparent clinically or is undiagnosed in a patient suspected of having an intraocular fungal infection, a definitive diagnosis often requires analysis of intraocular specimens. Unless the initial aqueous or vitreous paracentesis culture results are negative, the required vitrectomy specimens are likely to yield high positive culture results for mild vitreous inflammation. Additionally, polymerase chain reaction testing is a rapid and a highly sensitive method for identifying fungal infections.

Second, in the authors’ series, only 11 patients (21%) demonstrated positive culture results from nonocular specimens, including 6 patients with fungemia. We specifically are interested in the primary infection sources, such as intravenous drug abuse or wound infection. However, Lingappan and associates discussed only the risk factors; they did not explore the primary infection sources and organisms. Thus, we pose the following questions: How many patients could be identified with the primary infectious sources? What causes the low positive culture results in nonocular specimens? Before ocular signs were identified, how many patients received previous systemic antifungal treatment? Did yeasts and molds show different risk factors and primary infection sources?

Third, although bilateral diffuse posterior inflammation developed in the patients, both eyes typically have varying severity of choroidoretinal infiltration and vitreous inflammation. Large posterior pole choroidoretinal infiltration typically presents a high risk of retinal detachment, even without pars plana vitrectomy, and a poor visual outcome. For eyes with mild vitreous seeding, repeated intravitreal injections of antifungal agents typically resolve Candida endophthalmitis.

Finally, the treatment of endogenous fungal endophthalmitis depends on the causative organisms and the extent of ocular involvement. In cases of aggressive aspergillus endophthalmitis, early surgery is recommended. For Candida endophthalmitis, surgery is recommended in cases with moderate to severe involvement. When the infection is limited to the choroidoretinal area, systemic treatment alone may be curative. When vitreous involvement occurs, a vitrectomy with injections of antifungal agents can be considered. We also have observed a higher incidence of retinal detachment in eyes with endogenous fungal endophthalmitis after vitrectomy. The choroidoretinal lesion with adjacent vitreous infiltration has a significant adhesion among the vitreous, retina, and choroid. Furthermore, in some patients, rhegmatogenous or tractional retinal detachment, or both, can develop during follow-up. Although intravitreal corticosteroids remain a controversial contraindication for endogenous fungal endophthalmitis, intravitreal corticosteroids may be used as an adjuvant with antifungal agents to decrease inflammation after poor responses to previous intravitreal antifungal injections. In our clinical experience, most eyes demonstrated no sequential complications and had superior visual outcomes.