Intravitreal bevacizumab versus panretinal photocoagulation for treatment‐naïve proliferative and severe nonproliferative diabetic retinopathy

Abstract

Editor, Panretinal photocoagulation (PRP) is the treatment of choice for proliferative diabetic retinopathy (PDR) and severe nonproliferative diabetic retinopathy (NPDR) (Early Treatment Diabetic Retinopathy Study Research Group 1991). PRP has limitations, including vision loss, visual field defects and nyctalopia. Macular oedema can occur or worsen after PRP (McDonald & Schatz 1985). Media opacities such as cataract and vitreous haemorrhage occasionally preclude adequate laser therapy. An alternative treatment could ameliorate these difficulties. Intravitreal bevacizumab combined with PRP resulted in better regression of Neovascularisation (NV) and less vision loss compared with PRP alone (Mason et al. 2008; Tonello et al. 2008). We investigated intravitreal bevacizumab injections alone without prior PRP in a prospective, paired-eye, randomized pilot study. Potential benefits of intravitreal bevacizumab include preventing macular oedema, treating both PDR and macular oedema concurrently prior to performing PRP, and temporising until vitreous haemorrhage cleared to allow adequate PRP. Patients with type 2 diabetes mellitus and symmetric untreated severe NPDR or PDR without macular oedema or prior intraocular surgery were included. Informed consent was obtained which clearly explained that the standard treatment for their condition was PRP. The patients were notified the eye treated with intravitreal bevacizumab probably would require PRP at the end of the study. Best corrected visual acuity (BCVA), ophthalmologic examination, macular OCT, fundus photography and fluorescein angiography were performed at baseline and serially. Patients were evaluated every 2 months for 12 months total follow-up. The right eye was randomly assigned to treatment with PRP or intravitreal bevacizumab, and the left eye received the other treatment. Eyes randomized to bevacizumab 2.5 mg (0.1 ml) (Avastin, Genentech, San Francisco, CA, USA) were injected every 2 months during the study. Fellow eyes randomized to PRP were treated in two sessions, with a third session only if there was angiographic evidence of neovascular activity at month 4. Fifteen patients were enrolled. Unfortunately, five patients were lost to follow up between the baseline exam and the 2-month visit, and we were unable to ascertain the reasons these patients left. Ten patients (20 eyes) were followed for 12 months. Seven were women and the average age was 53 ± 9 years. Five patients had PDR and five had severe NPDR. Baseline BCVA and central macular thickness (CMT) were not statistically different between groups (p = 0.67 and p = 0.38, Mann–Whitney U test). None of 10 eyes with severe NPDR developed PDR. In 10 eyes with PDR, neovascular leakage completely resolved in four of five eyes treated with bevacizumab and in one of five eyes treated with PRP (p = 0.11, chi-square mid-P exact). Final CMT was significantly less in eyes treated with bevacizumab compared to those treated with PRP, 197 ± 17 μm versus 243 ± 49 μm, respectively (p = 0.012). BCVA (logMAR) pretreatment was −0.12 ± 0.22 and −0.14 ± 0.23 compared with −0.14 ± 0.19 and −0.17± 0.10 post-treatment (p = 0.52 and 0.64, bevacizumab and PRP groups, respectively). None of 10 eyes treated with bevacizumab injections every 2 months developed a complication. Two eyes treated with PRP developed vitreous haemorrhage. Three of 10 eyes treated with PRP developed macular oedema compared with none in the bevacizumab group. As this was a fellow eye controlled study, the effects of glycaemic and blood pressure control would equally impact both eyes of each patient. Intravitreal bevacizumb 2.5 mg injections every 2 months effectively controlled NV and resulted in a thinner CMT than PRP at 1 year. There were no differences in mean BCVA, which was expected given the good baseline vision and the small sample size. In a total of 60 bevacizumab injections, there were no complications, such as endophthalmitis or tractional retinal detachment. The main limitations of this study are the small sample size and the large proportion of patients lost to follow up. Despite the small sample size, the difference in final CMT achieved statistical significance. In summary, repeated intravitreal bevacizumab injections controlled NV in treatment-naïve eyes with PDR and severe NPDR for 1 year, but this finding needs to be confirmed by a larger study. PRP remains the gold-standard treatment for these eyes. In the future, longer-lasting anti-VEGF agents may be able to avoid the complications associated with PRP. Intravitreal bevacizumab may be a useful temporising measure in eyes with macular oedema or vitreous haemorrhage until adequate PRP can be performed, but further study is required.