Intravitreal bevacizumab at the end of diabetic vitrectomy for prevention of postoperative vitreous hemorrhage: a comparative study

Abstract

Objective: To study the role of intravitreal injection of bevacizumab (Avastin) at the end of vitrectomy for prevention of postoperative vitreous hemorrhage in diabetic eye disease.

Material and method: The authors conducted a retrospective, comparative, nonrandomized study comparing patients with diabetic eye disease who received a 1.25 mg bevacizumab injection at the end of vitrectomy to a group with diabetic eye disease who underwent vitrectomy but did not receive this injection. For statistical analysis, each patient was assigned to one of four groups according to the hemostatic modalities used (group 1, none; group 2, only long acting gas; group 3, only intraoperative intravitreal bevacizumab; group 4, both long acting gas and intraoperative intravitreal bevacizumab). The primary outcome measure was the incidence of early and late postoperative vitreous hemorrhage (POVH). The secondary outcome measure was visual acuity (VA) at 1 and 6 months.

Results: The present study included 87 eyes from 78 patients. The 87 eyes were categorized in the four groups described above. The incidence of early postoperative VH was lowest in group 3 (50%), followed by group 1 and group 4 (130%) and group 2 (16%) (p = 0.49). No eyes in group 1 or group 4 had late postoperative VH. Group 3 had the lowest incidence of VA below 20/200 (11%) in the first postoperative month, followed by group 1 (22%), group 4 (33%) and group 2 (50%) (p = 0.44). Group 3 also had the best visual recovery (VA above 20/50) with 38% at the end of six months compared with 13% in group 1, 7% in group 4 and no eyes in group 2 (p = 0.03).

Conclusion: Patients with intravitreal injection of 1.25 mg bevacizumab at the end of diabetic vitrectomy had the lowest incidence of early postoperative vitreous hemorrhage with no statistical significance. However they had significantly the best visual recovery at the end of six months.