Bevacizumab in High-Risk Corneal Transplantation: A Pilot Multicenter Prospective Randomized Control Trial

Abstract

Purpose: To determine the efficacy of local (subconjunctival and topical) bevacizumab (Avastin) treatment in patients undergoing vascularized high-risk corneal transplantation.

Design: Pilot, prospective, randomized, double-blind, placebo-controlled clinical trial conducted at 5 clinical centers in the United States, India, and Brazil.

Participants: Patients aged > 18 years undergoing high-risk penetrating keratoplasty, defined as corneal neovascularization (NV) in 1 or more quadrants ≥2 mm from the limbus or extension of corneal NV to the graft-host junction in a previously failed graft.

Methods: Patients were randomized to receive subconjunctival bevacizumab (2.5 mg/0.1 ml) or placebo at the time of surgery, followed by topical bevacizumab (10 mg/ml) or topical placebo, administered 4 times per day for 4 weeks.

Main outcome measure: The 52-week endothelial immune rejection rate.

Results: Ninety-two patients were randomized to receive bevacizumab (n = 48) or control (n = 44). The 52-week endothelial rejection rate was 10% in the bevacizumab group and 19% in the control group (P = 0.20). Post hoc, extended follow-up at the lead study site showed an endothelial rejection rate of 3% in the bevacizumab group and 38% in the control group (P = 0.003). Treatment with bevacizumab was found to have a hazard ratio of 0.15 (95% confidence interval, 0.03-0.65, P = 0.01) in a post hoc Cox regression analysis.

Conclusions: In patients undergoing vascularized high-risk corneal transplantation, there was no statistically significant difference in the rate of endothelial rejection at 1 year in the bevacizumab treatment group compared with the control group. This study may have been underpowered to detect a difference between treatment groups, and taken together, our data suggest that, in the current trial design, bevacizumab has a positive but not (yet) significant effect on endothelial rejection.

Keywords: Angiogenesis; Corneal transplantation; Neovascularization; Penetrating keratoplasty; Vascular endothelial growth factor.

Figure 1.

Study design and patient flow. *Patients who were randomized but withdrew from the study before surgery or were lost to follow-up before surgery. These patients did not undergo surgery and did not receive study medication or control and were excluded from last observation carried forward analyses. ^†Patients were lost to follow-up at week 8 (n = 1) and week 26 (n = 3). ^‡Patients were lost to follow-up at weeks 4 (n = 1), 26 (n = 1), and 39 (n = 1). ^§One patient developed infectious keratitis, was considered to have graft failure, and was withdrawn from the study to undergo keratoprosthesis surgery. LFU = lost to follow-up; QID = 4 times per day.

Figure 2.

The 52-week endothelial rejection and graft failure. A, The 52-week endothelial rejection survival rate was 90% in the bevacizumab group versus 81% (P = 0.20) in the control group. B, The survival rate for overall graft failure was 85% in the bevacizumab treatment group and 74% in the control group (P = 0.16). Solid line: bevacizumab treatment group. Dashed line: control treatment group.

Figure 3.

Endothelial rejection and graft failure in first versus repeat corneal transplants. A, In patients undergoing a first-time graft, the 52-week endothelial rejection survival rate was 86% in the bevacizumab group and 80% in the control group (P = 0.58). The survival rate for overall graft failure was also 86% in the bevacizumab group and 80% in the control group. B, In patients undergoing a repeat corneal transplant, the 52-week endothelial rejection survival rate was 92% in patients treated with bevacizumab compared with 81% in patients treated with control (P = 0.08). The 52-week overall graft failure survival rate was 84% in the bevacizumab group versus 70% in the placebo control group (P = 0.18). Solid line: bevacizumab treatment group. Dashed line: control treatment group.

Figure 4.

Endothelial rejection and graft failure by extent of corneal neovascularization (NV). A, In patients with 3 or more clock-hours of corneal NV at baseline, endothelial rejection survival was 90% in patients receiving bevacizumab and 78% in patients receiving control (P = 0.12). Graft failure survival was 85% in patients receiving bevacizumab and 72% in patients receiving control (P = 0.14). B, In patients with 6 or more clock-hours of corneal NV at baseline, the endothelial rejection survival rate was 88% in the bevacizumab group versus 77% in the control group (P = 0.17), whereas the overall graft failure survival rate was 83% in the bevacizumab group versus 70% in the control group (P = 0.17). Solid line: bevacizumab treatment group. Dashed line: control treatment group. NV = neovascularization.

Figure 5.

Endothelial rejection and graft failure by extent of corneal neovascularization (NV) in repeat corneal transplants. A, In patients with 3 or more clock-hours of corneal NV at baseline, endothelial rejection survival was 92% in patients receiving bevacizumab and 77% in patients receiving control (P = 0.13). Graft failure survival was 84% in patients receiving bevacizumab and 68% in patients receiving control (P = 0.17). B, In patients with 6 or more clock-hours of corneal NV at baseline, the endothelial rejection survival rate was 91% in the bevacizumab group versus 78% in the control group (P = 0.21), whereas the overall graft failure survival rate was 82% in the bevacizumab group versus 67% in the control group (P = 0.21). Solid line: bevacizumab treatment group. Dashed line: control treatment group.

Figure 6.

Extended follow-up of endothelial rejection and graft failure at lead study site. A, Graft survival (endothelial rejection) was 97% in the bevacizumab group and 62% in the control group (P = 0.003). B, In terms of graft failure, survival was 34% in patients receiving bevacizumab and 39% in patients receiving control (P = 0.55). The median follow-up times in the bevacizumab and control groups were 207 and 176 weeks, respectively. Solid line: bevacizumab treatment group (n = 29). Dashed line: control treatment group (n = 30).

Figure 7.

Clinical characteristics after transplantation. A, Graft endothelial cell counts at weeks 26 and 52. B, Central corneal thickness at weeks 4 and 52. C, Change in the extent of corneal neovascularization (NV) from baseline to week 52 in clock-hours of involvement (**P = 0.006). D, Change in the length (in millimeters) of the longest corneal blood vessel from baseline to week 52. Black bars = bevacizumab. White bars = control.