We thank Dr Kocatürk and associates for their thoughtful correspondence regarding our manuscript.

First, inflammation plays a critical role in the development of corneal neovascularization; therefore, inflammation received careful consideration in the development of our clinical study. Because we utilized a clinic-based population, it was not possible to standardize baseline levels of ocular surface inflammation. Most of the study patients were on one or more medications that could directly or indirectly affect ocular surface inflammation. In order to minimize the potential confounding effects of inflammation, only patients with clinically stable corneal neovascularization were considered for inclusion. Additionally, we excluded patients who had recently undergone a change in dose or frequency of topical anti-inflammatory therapy. Although punctal plugs may lead to a slight increase in the local bioavailability of these maintenance medications, it is unlikely that this caused a measurable response in 3 weeks. Eighteen out of 20 total patients (90%) completed 24 weeks of follow-up. Fifteen out of the 18 remaining patients (83%) exhibited decreases in neovascular area and vessel caliber. Although there were several nonresponders, topical bevacizumab proved to be effective in a majority of cases.

Second, we have reported that topical bevacizumab does not penetrate through the epithelium of healthy murine corneas. However, we also reported that bevacizumab was detectable at high levels in the corneal stroma of mice with experimental corneal neovascularization. Corneal neovascularization or inflammation significantly diminishes the integrity of epithelial barrier function. Regardless, the variable penetration of bevacizumab may explain, in part, the variable responses to topical bevacizumab treatment.

Third, the concentrations of topical bevacizumab used in the clinical studies published to date have ranged from 0.5%-1.25%. We chose to use a relatively high concentration for a short period of time because longer treatment durations have been associated with localized adverse events (eg, corneal epitheliopathy). As noted in our report, we observed neither local nor systemic adverse events. The best duration, concentration, and frequency of topical bevacizumab treatment are not yet known, as we emphasized in our report.

Finally, numerous studies have found that vascular endothelial growth factor (VEGF) plays a central role in the development of new blood vessels. The optimal time for antiangiogenic intervention would be during the course of new vessel formation. However, normal blood vessels also depend on VEGF for survival, and systemic VEGF inhibition has been shown to cause vessel regression in otherwise quiescent tissues. In the eye, retinal pigment epithelium–derived VEGF has been shown to play an important role in the maintenance of the choriocapillaris. Our findings are in concordance with these studies regarding the role of VEGF in the maintenance of stable blood vessels.