We congratulate Cheng and associates on their article “Short-term topical bevacizumab in the treatment of stable corneal neovascularization.” A couple of concerns were raised in our minds after careful reading the manuscript.

First, 12 of 20 patients were still on other medications, such as topical steroids and topical cyclosporine. These medications also decreased the neovascularization and inflammatory processes (eg, herpetic simplex virus keratitis, herpetic zoster ophthalmicus, penetrating keratoplasty, and limbal stem cell deficiency). Three patients did not respond to topical bevacizumab treatment and 2 patients were followed for only 6 weeks. The remaining patient numbers are not sufficient to make a correct decision.

Second, Dastjerdi and associates reported that bevacizumab (topical) could not penetrate intact corneal epithelium in mice. Corneal neovascularization is located at the stromal or deep stromal levels of the cornea. Punctal plugs may also enhance the concentration of other medications in the eye. Thus, bioavailability and bioefficacy of other medications (steroids and cyclosporine) may also increase, helping to reduce inflammatory processes and neovascularization.

Third, Koenig and associates reported that 5 mg/mL bevacizumab eye drops inhibited corneal neovascularization and decreased the vessel diameter. Why did the authors decide to use 10 mg/mL bevacizumab? Did they see any side effects (corneal epithelial defects, inhibition of wound healing) with higher doses? Dose responses might be done for different levels of neovascularization.

Finally, Chang and associates reported that bevacizumab is only effective against actively growing blood vessels and vascular endothelial growth factor is not needed by established vessels for proliferation. Bevacizumab is possibly less effective in eyes with large areas of well-established corneal neovascularization. This also explains why corneas with smaller neovascular areas showed greater responses to the mentioned treatment.