Ischaemic change after intravitreal bevacizumab (Avastin®) injection for macular oedema secondary to non‐ischaemic central retinal vein occlusion

Abstract

Editor, A 65-year-old male patient with a 12-year history of diabetes presented with decreased visual acuity in his right eye of 3 months duration. Best-corrected visual acuity (BCVA) was 0.4 in the right eye (RE) and 1.0 in the left eye (LE). A slit-lamp examination showed grade two bilateral nucleosclerosis with no further abnormalities of the anterior segment. Fundus examination and fluorescein angiography showed flame-shaped retinal haemorrhages in four quadrants and optocillary shunt vessels. These ocular findings were consistent with non-ischaemic central retinal vein occlusion (CRVO). Optical coherence tomography (OCT) showed submacular fluid and macular elevation with a central macular thickness of 533 μm in the RE (Fig. 1). An intravitreal off-label bevacizumab injection was recommended. The patient was informed about the treatment beforehand and provided written informed consent. The injection was performed in a sterile setting and prophylactic topical antibiotics and carbonic anhydrase inhibitor were given for 1 week pre- and post-injection. The intravitreal dosage of bevacizumab was 1.25 mg (0.05 ml). Repeat examinations of visual acuity, intraocular pressure and fundus examinations were scheduled at 1 day, 1 week and 1 month after injection. Three weeks after injection, the patient reported a deterioration of visual acuity in the RE. His visual acuity in the RE was 0.02 and a fundus examination showed disc oedema with extensive flame-shaped retinal haemorrhages in the four retinal quadrants (Fig. 2). A fluorescein angiogram of the RE demonstrated severe non-perfusion including the fovea with blocked hypofluorescence. OCT imaging showed a central macular thickness of 560 μm (Fig. 2). Six months after the injection, his visual acuity in the RE was 0.02 and panretinal photocoagulation was performed. Fundus photography (A), fluorescein angiography (B) and optical coherence tomography (OCT) (C) images at initial presentation. (A) Fundus photograph showing multiple intraretinal haemorrhages and dilated retinal veins with optociliary shunt vessels. (B) Fluorescein angiograph in the late phase showing blocked hypofluorescence with patelloid pattern of dye leakage in the fovea in the right eye. (C) A vertical OCT image of the right eye showing cystoid macular oedema with a central macular thickness of 533 μm. Fundus photography (A), fluorescein angiography (B) and optical coherence tomography (OCT) (C) images 3 weeks after intravitreal bevacizumab administration. (A) Fundus photograph showing extensive flame-shaped retinal haemorrhages and disc swelling. (B) Fluorescein angiograph in the late phase showing severe non-perfusion and dye leakage at the optic disc. (C) Vertical OCT image showing subretinal fluid with a central macular thickness of 560 μm. Bevacizumab represents a new treatment option for early intervention against the formation of macular oedema in various disease conditions. Several short-term results have been published on intravitreal bevacizumab for macular oedema secondary to retinal vein occlusion with CRVO (Iturralde et al. 2006; Spandau et al. 2006; Pai et al. 2007). These reports suggest that intravitreal bevacizumab may reduce macular oedema and increase visual acuity in CRVO. But because a physiological level of vascular endothelial growth factor (VEGF) may be necessary for retinal homeostasis, precautions may have to be taken to avoid the possible negative effects of a complete blockade of VEGF, particularly in primarily ischaemic diseases. However, no report has been issued regarding complications associated with intravitreal bevacizumab except for endophthalmitis. Here, we present a case of non-ischaemic CRVO conversion to ischaemic CRVO after injecting bevacizumab intravitreally. The formation of a new blood flow balance is presumably supported by the formation of collateral vessels with a new drainage route. Spandau et al. (2006) suggested the possibility of complete blockage of VEGF by intravitreal bevacizumab. Although 11–34% of non-ischaemic CRVO cases convert to ischaemic CRVO over a natural course of 3–5 years, we suggest that intravitreal bevacizumab injection affected the outcome in this case (TCVOS 1997). This is because ischaemic CRVO occurred 3 weeks after the injection, when bevacizumab has its peak effect and relatively good initial visual acuity and perfusion status. There were no other patient factors that might cause the ischaemic conversion of CRVO except diabetes. The causes and mechanisms of treatment failure associated with bevacizumab administration must be elucidated further in vitro and by clinical studies. Furthermore, the possibility that bevacizumab has a negative long-term effect on collateral vessel formation because of its anti-VEGF action requires investigation.