INTRODUCTION

Ranibizumab (LucentisTM, Genentech Inc, San Francisco, US) is a humanized antigen-binding fragment designed to inhibit all iso­forms and active degradation products of vascular endothelial growth factor A (VEGF-A) (Gaudreault et al. 2007). Ranibizumab  (Rosenfeld et al. 2006) and other anti-VEGF agents such as bevacizumab (AvastinTM, Genetech Inc, San Francisco) and pegapta­nib sodium (MacugenTM) (Gragoudas et al. 2004) have been shown to reduce neovas­cularisation in age-related macular degene­ration (AMD) when injected intra-vitreally.

Diabetic vitrectomy is complicated by poor visualisation by the surgeon as a result of haemorrhage due to active neovascularisa­tion. Post-operative vitreous haemorrhage is a further complication, resulting in de­creased visual acuity as a result of bleeding particularly from optic disc neovascularisa­tion, which cannot be directly cauterised.

A number of case series utilizing intra­ocular injection of bevacizumab, a drug designed for systemic anti-cancer use as an “off-label” adjunct to reduce retinal neovas­cularisation prior to diabetic vitrectomy have been published(Yeoh et al. 2008). Bevaci­zumab has been shown to induce rapid regression of retinal and iris neovasculariza­tion in proliferative diabetic retinopathy (PDR)(Avery et al. 2006).

Therefore, the aim of our study was to in­vestigate the safety and benefit of giving an intraocular injection of ranibizumab, a drug designed and licensed for intraocular injec­tion for AMD, prior to diabetic vitrectomy to reduce the intraoperative and post-operative bleeding in the eye. There has been no such study to date.

MATERIALS AND METHODS

A retrospective study was performed at the Universiti Kebangsaan Malaysia Medical Centre (UKMMC). The medical records of 16 diabetic patients were re­viewed. These eyes received 20 consec­utive ranibizumab injections from July 2007 to February 2009 prior to vitrec­tomy.

All patients were known diabetics diag­nosed by a physician and were on treat­ment with including oral hypoglycaemics and/or insulin injections. Each study eye had known or suspected diabetic retino­pathy. Pre-operative intravitreal ranibi­zumab was considered for all patients who had never received any prior laser photocoagulation or in whom laser pho­tocoagulation could not be completed due to significant areas of retinal de­tachment. It was also considered for pa­tients with clinically significant macular oedema secondary to vitreo-macular traction. This included patients with vi­treous haemorrhage, tractional retinal detachment (TRD) with or without rheg­matogenous component, vitreomacular traction (VMT) and epiretinal membrane (ERM) secondary to known or suspected proliferative diabetic retinopathy (PDR) with active neovascularisation. When the fundus could not be viewed, B scan ul­trasonography was used to confirm the diagnosis.

Patients with prior history of cerebro­vascular accident (CVA) or known vas­cular occlusion of any kind were ex­cluded from injection with of ranibizumab. In­jection was only scheduled after the rou­tine pre-operative investigations which included blood pressure reading, full blood count, renal profile, fasting blood glucose level, echocardiogram and chest X ray were acceptable for general or lo­cal anaesthesia as the injection was con­sidered a commitment towards operation.

Informed consent was obtained from each patient. This included counseling on the risk of infection, cataract and retinal detachment. Each injection was per­formed in a clean treatment room in the eye clinic outpatient setting according to strict aseptic technique. Anaesthesia for the procedure was topical with guttae proparacaine hydrochloride (AlcaineTM) and subconjunctival anaesthesia with 2%
lignocaine (XylocaineTM). Subsequently, 5% povidone iodine lid scrub and douche was performed prior to sterile draping. All patients received ranibizumab, which was aspirated by the assistant into a new 30G insulin needle after swabbing each bottle cap with 100% alcohol wipe.

Each injection of 0.05 or 0.1 ml of rani­bizumab was given 4 mm from the lim­bus in phakic patients and 3.5 mm in pseudophakic patients while directing the injection into the centre of the vitreous cavity. Five percent iodine was then re-instilled onto the ocular surface. The pa­tient then received guttae ciprofloxacin eyedrops at 2 hourly intervals for one week. They were given instructions on symptoms and signs of endophthalmitis and were reviewed the following day and on the day of admission for surgery.

Demographical data (age, gender and ethnicity) were recorded. Dates of intra­vitreal injections, volume of drug injected, surgical indications, operation record of intraoperative bleeding and duration of surgery were noted. Three cases in which anti-VEGF had been injected at any time in the past were excluded from further analysis. The pre-operative visual acuity prior to injection and on admission was noted and the post-operative visual acuity was also noted at one week, one month, 3 months, 6 months after the op­eration and at last review. The clarity of ocular media was noted from the dia­grams and comments in the operative notes and this was correlated with the visual acuity.

Post-operative haemorrhage was graded as “nil”, if all retina details were visible; “mild” if any part of the fundus was slightly obscured with blood vessels, optic disc and /or laser marks visible; “moderate” if any part of the fundus was totally obscured and “severe”, if the en­tire fundus was totally obscured. The fundus under gas was used for assess­ment in gas-injected eyes. Any complica­tion of the intravitreal injection was noted.

RESULTS

Fourteen eyes of fourteen patients re­ceived intravitreal injection of 0.05 (0.5 mg) to 0.1 ml (1mg) of ranibizumab be­tween 3 to 8 days prior to vitrectomy by two surgeons (Table 1). Mean duration of injection prior to surgery was 4.4 days. One patient received 1 mg of ranibizu­mab. One patient received injection prior to PPV in both eyes. Patient ages ranged from 46 to 72 yearswith a mean of 57.7 years. There were 5 males and 11 fe­males. Twelve patients were Malay, 3 were Chinese and 1 was Indian.Twelve were right eyes (Table 1).

There were no complications locally or systemically post-injection. Indications for vitrectomy were retinal detachment in 11 eyes of whom 3 were combined TRD and rhegmatogenous RD (RRD) with macula detached and 8 were tractional detach­ments involving the macula. Three trac­tional detachments were in combination with vitreous haemorrhage while there were eleven pre-operative vitreous hae­morrhages. One eye had vitreo-macular traction syndrome secondary to PDR. One eye had re-bled 3 years after pre­vious vitrectomy for TRD. Pre-operative visual acuity (VA) ranged from 6/36 to perception of light with the majority (19 eyes) with pre-operative vision of 6/60 or worse.

Pars planar vitrectomy (PPV) was per­formed in all cases under local (n=17) or general anaesthesia (n=3) by 2 surge­ons. 20G vitrectomy was performed pri­marily in 15 cases and 23G in 5 cases depending on instrument availability and surgeon preference. Panretinal laser photocoagulation (PRP) was topped up or given to the ora in all cases. Intraocu­lar tamponade of silicone oil 1300Cs was used in 5 cases and intraocular gas tam­ponade was used in 7 cases. Of those under gas tamponade, sulphur hexafluo­ride (SF6) was used in 5 cases, and oc­tafluoropropane (C3F8) in 2 cases (Table 2).

All eyes had minimal to moderate intra­operative bleeding. The amount of intra­operative bleeding did not correlate with duration of injection prior to vitrectomy but more with the age of the patient and activity of neovascularisation. There was no documented case of progression of tractional detachment to RRD during the study period although fibrosis of neovas­cularisation was consistently observed to various degrees in patients with view of the retina. Mild progression of the TRD not resulting in RRD accompanied fibro­sis in 3 observable cases (Table 3).

Post-operative VH in eyes without tam­ponade or with subtotal gas tamponade ranged from nil (n=1), mild (n=12) to moderate (n=1) with complete clearance 4 to 90 days post-operatively (mean 21 days). There was one case of slightly increased post-operative VH at 1 week, which cleared by 1 month during the study period.

Silicone oil filled eyes had nil (n=1) to moderate localized preretinal haemorr­hages behind the oil (n=3) to severe post-operative haemorrhage  (n=1), which cleared after an average of 44.8 days including one which took over 3 months to clear with oil pushed into the anterior chamber. We believe the pa­tient’s uncontrolled hypertension was re­sponsible for this outcome as haemosta­sis was achieved at the end of the oper­ation. Two silicone filled eyes had very minimal haemorrhages and clear fundal view. Duration of follow-up ranged from 1 month (n=19) to 14 months with 18 eyes (90%) having 3 or more months of follow-up. There were 2 patients who defaulted follow-up after 1 and 3 months. (Table 3)

In general, post operative VH cleared in mean time of 17.6 days for 23G vitrec­tomy and 30.1 days (n=10) for 20G vi­trectomy (Table 3). VA had either re­mained unchanged (n=2) (10%), improved (n=14) (70%) or worsened (n=4) with best corrected VA at last review ranging from 2/60 or better (n=15) to no perception of light (n=1). Two eyes achieved 6/12 or better vision (10%). Two eyes gained significant vision from hand motions or counting fingers to as much as 6/9 vision. Ten eyes (50%) achieved 6/36 or better vision. Mean gain in visual acuity was 1.8 Snellen lines. However, 60% of 23G eyes gained at least 3 Snellen lines compared to 46% of 20G eyes.

Two eyes (10%) developed retinal de­tachment post-operatively. One patient developed total retinal detachment one-month post PPV with gas injection. He underwent further operation and retinal reattachment surgery. Another case de­veloped retinal detachment following cat­aract extraction and silicone oil removal. There was no emergence or exacerba­tion of rubeosis in the any of the eyes at 6 months postoperatively.

DISCUSSION

Diabetic retinopathy (DR) is the leading cause of blindness among people of working age worldwide, leading to signifi­cant loss of quality of life. During the first two decades of disease onset, nearly all Type 1 diabetic patients and over 60% of type 2 diabetics have retinopathy. In the Wisconsin Epidemiologic Study of Dia­betic Retinopathy (WESDR), 3.6% of younger-onset patients (type 1 diabetes) and 1.6% of older-onset patients (type 2