Traction retinal detachment (TRD) and combined traction and rhegmatogenous retinal detachment (T+RRD) are among the most challenging vitreoretinal surgical scenarios. The surgical management of these cases is complex because these eyes have very thin ischemic retinas, often with massive neovascularization. Intra- and postoperative complications are frequent and include intraoperative bleeding, iatrogenic breaks, postoperative bleeding, and redetachments. Of these, intraoperative bleeding that decreases visualization while removing fibrovascular tissue can hinder the completion of a surgical intervention.
Reports of the use of intravitreal bevacizumab (Avastin, Genentech) for the reduction of neovascularization in proliferative diabetic retinopathy have demonstrated significant regression in the number of new vessels.1,2 Reducing the vascularity of neovascular fibrovascular tissue can potentially reduce bleeding intraoperatively and thus facilitate the vitrectomy procedure. To this end, we chose to investigate the potential benefit of the use of preoperative bevacizumab in the management of these most complex cases.
METHODS
This is a prospective study of eyes undergoing diabetic vitrectomy for traction or traction and rhegmatogenous retinal detachment pretreated with 1.25 mg intravitreal bevacizumab 3 to 8 days prior to pars plana vitrectomy. Patients underwent 25-, 23-, or 20-gauge surgery at the discretion of the surgeon. All procedures were performed by a single surgeon. Fifty-five eyes of 55 patients were included. Thirty-four eyes had traction retinal detachment (group 1), and 21 eyes had combined traction and rhegmatogenous retinal detachment (group 2). At baseline and at all subsequent visits, BCVA, slit-lamp examination, intraocular pressure, and indirect ophthalmoscopy were performed. Optical coherence tomography and fluorescein angiography were performed as needed. Mean follow-up was 13 months (range, 6-24 months).
BASELINE CHARACTERISTICS
Group 1 (TRD). In group 1, there were 34 eyes, mean patient age was 49 years (range, 32–68), and 59% of patients (n=20) were male. Mean duration of diagnosed diabetes was 18 years (range, 5–30). Sixty-seven percent of patients had systemic hypertension, 53% renal disease, 20% cardiovascular disease, and 9% had no light perception in the fellow eye. Seventy-six percent had previous laser photocoagulation, and 3% had iris neovascularization.
Group 2 (T+RRD). In group 2, there were 21 eyes, mean patient age was 48 years (range, 21–69), and 80% percent of patients were male. Mean years of diagnosed diabetes was 20 years (range, 5–40). Hypertension was present in 86%, renal disease in 71% and cardiovascular illness in 52%. In this group, 23% were blind in the other eye and 38% had a T+RRD. Eighty-one percent of eyes had previous laser photocoagulation, 24% had a previous vitrectomy, and 28% had iris neovascularization.
RESULTS
Group 1. Mean preoperative best corrected visual acuity (BCVA) was 2/200 (range, 20/200–light perception). Mean final BCVA was 20/100 (range, 20/25–no light perception) At final BCVA, 80% of eyes had improved two or more lines of vision, 11% had remained unchanged, and 9% had decreased vision.
The vitrectomy method used included 18 eyes for which 23-gauge vitrectomy was performed, nine eyes for which 20-gauge vitrectomy was used, and seven for which 25-gauge vitretcomy was used. Gas tamponade was used in all cases, 13 with air, five with SF6, 12 with C2F6 and four with C3F8. Mean surgical time was 40 minutes (range, 20–110).
Group 2. Mean preoperative BCVA was hand motion (range 20/400–light perception), and mean final BCVA was 20/300 (range, 20/30–no light perception). At final BCVA, 57% had improved vision, 24% were unchanged, and 19% had decreased vision.
The vitrectomy method used included 11 eyes for which 23-gauge vitrectomy was used and 10 eyes for which 20-gauge vitrectomy was used. The tamponading agent used included C2F6 in 10 eyes, C3F8 in six eyes, and 1,000 centostoke silicone oil in five eyes. Mean surgical time was 51 minutes (range, 25–105).
COMPLICATIONS
Group 1. The most common complication was postoperative residual hemorrhage, which occurred in nine eyes (26%). In four of these eyes, the hemorrhage cleared within 1 month. Rebleeding occurred in four eyes (12%), glaucoma in one eye, neovascular glaucoma in one and progression of cataract in seven. In all, 19 postoperative procedures were required. These included eight in-office fluid-air exchanges for nonclearing hemorrhage, four Nd:YAG capsulotomies to the periphery to aid hemorrhage clearing in pseudophakic eyes, and seven phacoemulsifications.
Results were compared with a cohort of consecutive patients treated by the same surgeon with vitrectomy without bevacizumab. Only 70% of eyes without pretreatment with bevacizumab improved vision, compared with 80% in the bevacizumab-treated eyes. Residual hemorrhage was present in 70% of eyes without bevacizumab, compared with 26% of eyes with pretreatment with bevacizumab. Required postoperative interventions were fewer in the bevacizumab group.
Group 2. Postoperative residual hemorrhage was present in three eyes (14%), and clearing occurred within 6 weeks. One eye had rebleeding, two eyes had redetachments, three had neovascular glaucoma, one had glaucoma, and six eyes had cataract progression. In all, 15 postoperative procedures were performed, including one in-office fluid-air exchange, two Nd:YAG capsulotomies for hemorrhage clearing, two reinjections of bevacizumab, four reoperations, and six phacoemulsifications.
Comparison was made with a cohort of consecutive patients treated by the same surgeon using vitrectomy without bevacizumab. Only 25% of eyes without bevacizumab pretreatment improved vision compared with 57% in the bevacizumab-pretreated eyes. Residual hemorrhage was present in 52% of eyes without bevacizumab compared with 14% of eyes with pretreatment with bevacizumab. Required postoperative interventions were fewer in the bevacizumab group.
DISCUSSION
Since the report by Avery et al1 regarding the use of bevacizumab to elicit rapid regression of retinal neovascularization in eyes with proliferative diabetic retinopathy, its use in the management of different aspects of this disease has expanded. Several reports of the preoperative use of bevacizumab to facilitate surgery and improve outcomes in diabetics have been reported showing a significant benefit.3,4 A report by Romano et al3 showed reduced postoperative hemorrhage in eyes undergoing vitrectomy for vitreous hemorrhage pretreated with bevacizumab. Rizzo et al4 demonstrated improved visual and anatomic outcomes in eyes undergoing vitrectomy pretreated with bevacizumab. In this study, we chose to evaluate the benefit of pretreatment with bevacizumab of eyes undergoing vitrectomy for complex complications of diabetic retinopathy. The benefit appeared to be greater in the more complex eyes (ie, the T+RRD eyes) many of which had anterior segment neovascularization.
Among the eyes with TRD 80% of pretreated eyes improved BCVA versus 70% not pretreated. Twenty-six percent of pretreated eyes had some residual vitreous hemorrhage vs 70% that were not pretreated. The difference was greater in the combined T+RRD eyes, in which twice as many eyes improved BCVA in the pretreated group (57% vs 25%). A 3.7 times greater incidence of residual hemorrhage was seen in nonpretreated eyes (52% vs 14%).
CONCLUSION
Preoperative injection of bevacizumab in eyes with TRD and T+RRD appears to facilitate surgery, reduce surgical times, and improve visual outcomes. Fewer postoperative complications were seen, and fewer postoperative procedures were required. A greater benefit was seen in eyes with T+RRD. Finally, prospective, randomized studies are needed to further evaluate this treatment modality.
María H. Berrocal, MD, is an Assistant Professor at the University of Puerto Rico School of Medicine in San Juan. Dr. Berrocal is a member of the Alcon Speakers' Bureau. She can be reached at +1 787 725 9315; or via e-mail at mariahberrocal@hotmail.com.
_1773249222.png?auto=compress,format&w=75)
