The specific aims of the study are to test the following hypotheses:

• That there is a difference in change in visual acuity resulting from treatment with intravitreal bevacizumab compared with dexamethasone implant in eyes with advanced macular oedema
• That there is a difference in degree of resolution of macular oedema resulting from treatment with intravitreal bevacizumab compared with dexamethasone implant in eyes with advanced macular oedema
• That both intravitreal bevacizumab and dexamethasone implants have a manageable and acceptable safety profile in eyes with diabetic macular oedema

Diabetic retinopathy is a common cause of severe loss of vision and the most common cause of blindness in individuals between the ages of 20 and 65 years in developed countries. Swelling of the central retina, or "macular oedema", is the commonest cause of visual loss in diabetic retinopathy.

Diabetic macular oedema (DMO) is treated with laser photocoagulation of areas of leak in the macula according to established guidelines which take into account the extent of the leak and its proximity to the centre of the macula, the "fovea". This treatment does not always work, however, and is inherently destructive.

New drugs have become available which appear to reduce the risk of loss of vision in eyes with advanced diabetic macular oedema for which further laser treatment is unlikely to be beneficial. Intravitreal injection of slow-release steroid formulations such as Ozurdex™, a slow release formulation of dexamethasone, has been proposed as a new modality to treat clinically significant DMO. We have recently conducted randomised clinical trials which have demonstrated that treatment with intravitreal triamcinolone (IVTA) leads to reduction of DMO and improved vision in these eyes. Another class of drugs, inhibitors of Vascular Endothelial Growth Factor (VEGF) such as bevacizumab (Avastin®), also appear efficacious.

While both drugs appear to reduce macular oedema and improve vision in the short term, they may have differences which could guide how they are best used. Around 1/3 of eyes that receive dexamethasone may develop elevated intraocular pressure and cataract, both of which are manageable but may complicate the picture. Anti-VEGF drugs do not have these local adverse events, however they must be given more frequently (4-6 weekly vs 4-6 monthly for Ozurdex™) and it is suspected they may have a neurotoxic effect on the retina. Some authorities suspect that anti-VEGF treatment may be associated with a small increased risk of having a stroke or heart attack during treatment, even when they are injected into the eye. This has not been proven with a related drug, ranibizumab, but it is still possible that it may occur with bevacizumab.

This will be a, 2 year, phase II, prospective, multicentre, randomised, single-masked clinical trial of sustained release intravitreal dexamethasone (Ozurdex™) versus intravitreal injections of bevacizumab (Avastin®) for diabetic foveal oedema that persists or recurs despite previous laser treatment, or for which the investigator believes laser treatment is unlikely to be helpful.

Inclusion Criteria:

• Age >= 18 years
• Diagnosis of diabetes mellitus types 1 or 2
• Diabetic macular oedema affecting the fovea in one or both eyes (phakic or pseudophakic) for which laser treatment is unlikely to be helpful in the opinion of the centre chief investigator
• Best corrected visual acuity of 17-72 letters (6/12 -6/120)
• Retinal thickness > 250 micron in central 1mm subfield on Stratus (time domain) OCT and 300 on Spectral domain OCT
• Previous macular laser treatment, or the investigator believes laser treatment is unlikely to be helpful
• Intraocular pressure <22mmHg
• Women of childbearing potential must have a negative urine pregnancy test at the screening visit and prior to treatment. A woman is considered of childbearing potential unless she is postmenopausal and without menses for 12 months or is surgically sterilised
• Written informed consent has been obtained.

Exclusion Criteria:

• Known allergy to Ozurdex, Avastin or agents used in the study
• Women who are pregnant, nursing, or planning a pregnancy, or who are of childbearing potential and not using reliable means of contraception
• Glaucoma which is uncontrolled or is controlled but with more than one medication or with only one medication and with glaucomatous field defects
• Loss of vision due to other causes (e.g. age related macular degeneration, myopic macular degeneration, retinal vein occlusion)
• Macular oedema due to other causes
• An ocular condition that would prevent visual acuity improvement despite resolution of oedema (such as foveal atrophy or substantial premacular fibrosis)
• Treatment with IVTA within the last 6 months or peribulbar TA within the last 3 months or bevacizumab within the last 2 months.
• Cataract surgery within the last 6 months
• Retinal laser treatment within the last 3 months
• History of herpes virus infection in study eye
• Media opacity including cataract that already precludes adequate macular photography and laser treatment, or cataract that is likely to require surgery within 2 years
• Known allergies to dexamethasone or bevacizumab
• Patient is already receiving systemic steroid treatment > 5mg prednisolone daily or equivalent)
• Intercurrent severe disease such as septicemia, any condition which would affect follow-up or photographic documentation (e.g. geographical, psycho-social)
• History of chronic renal failure requiring dialysis or renal transplant
• Blood pressure >180/110
• Patient has a condition or is in a situation that in the investigator's opinion may put the patient at significant risk, may confound the study results, or may interfere significantly with the patient's participation in the study