Perilimbal bevacizumab injection for interface neovascularization after deep anterior lamellar keratoplasty

To evaluate the effect of perilimbal bevacizumab injection on neovascularization at the interface of donor and recipient tissues after deep anterior lamellar keratoplasty (DALK).

This interventional case series included 4 patients with interface neovascularization after DALK (3 patients with advanced keratoconus and 1 patient with scarring because of herpetic keratitis). Bevacizumab (2.5 mg/0.1 mL) was injected subconjunctivally near the limbus adjacent to the abnormal blood vessels. Depending on regression of new vessels, injections were repeated. If neovascularization occurred in more than 1 quadrant, injections were administered at 2 sites (2.5 mg for each injection). Total corneal neovascularization, longest neovascular length, and degree of neovascular circumference were evaluated by digital slit lamp corneal photography before injection and 1 week, 1, 3, and 6 months afterward.

The mean follow-up period was 14 ± 7.52 months from the first bevacizumab injection. Corneal neovascularization, on average,was 33.70% ± 7.8% (SD) of the total corneal surface before the injections, which decreased to 4.83% ± 3.3% (P = 0.005, t test) at the last follow-up examination. Mean degree of neovascularized circumference decreased from 165 degrees before injection to 77.5 degrees (P = 0.023) at the last follow-up examination. Mean longest neovascularized length decreased from 6.714 mm before injection to 1.497 mm (P = 0.021) at the last follow-up examination. At the end of the study, all patients had clear grafts with no sign of graft failure. Mean best-corrected visual acuity before injection was 20/100, which improved to 20/30 at the last examination.

This small case series suggests that perilimbal bevacizumab injection is safe and effective for partial regression of new blood vessels at the interface of donor and recipient tissues after DALK and may effectively prevent graft rejection.