This study will enrol patients with diabetes who have already elected to undergo pars plana vitrectomy (eye surgery) to remove persistent vitreous haemorrhage (a complication of severe diabetic eye disease in which blood fills the inner cavity of the eye, obscuring the vision and preventing treatment to stop the bleeding). Those in the treatment arm will have an intravitreal injection of ranibizumab (Lucentis) at the same dose used for the treatment of neovascular (wet) age-related macular degeneration (a disease that has some features in common with diabetic eye disease).

It is hypothesised that this will promote clearance of the vitreous haemorrhage and that this, in turn, may mean that some patients do not need to proceed to vitrectomy.

The eye is designed like a camera, with a lens at the front and a film at the back. Using this analogy, the retina is the film in the camera and the vitreous is the space between the lens and the retina. Light focused from the lens falls on the retina and creates an image of the outside world which is then transmitted to the brain.

Within the retina there are small blood vessels which supply it with nutrition. Sometimes in eye disease such as diabetes these blood vessels can grow abnormal new branches (stimulated by growth factors such as VEGF) which leave the retina and grow into the vitreous space. Because these new vessels are very fragile they have a tendency to get damaged and bleed. When they bleed into the vitreous space it creates a large collection of blood (haemorrhage) which prevents the focused light from the lens reaching the retina. This dramatically reduces the vision in this eye for the patient and prevents healthcare practitioners from examining the eye and giving further treatment for the cause of the bleed. This often means an eye operation is required to clear the blood from the vitreous space, to allow the patient to see and to allow further treatment to prevent worsening of the eye disease.

Diabetic retinopathy is the name given to the damage caused by diabetes to the retina in patients with this condition. It is the leading cause of vision loss in this group of patients who are mostly of working age. Patients with diabetes require regular eye examinations to look out for signs of these abnormal new vessels which can bleed into the vitreous. In the most severe form of diabetic eye disease (proliferative diabetic retinopathy) laser therapy is used to treat the back of the eye to try and prevent any further progression of these new proliferating blood vessels. However if a blood vessel has bled into the vitreous space, it prevents examination of the retina and the ability to provide any further preventative treatment for the patient. It can often take many months for these haemorrhages to clear, in which time the patient is left without any useful vision in that eye and the underlying disease is continuing to worsen. It is in these difficult situations an eye operation (vitrectomy) is required to remove the blood.

Small, uncontrolled studies with intravitreal (injection into the vitreous space of the eye) bevacizumab (Avastin) suggest that this agent enhances clearance of diabetic vitreous haemorrhage(1,2,3). Bevacizumab is a monoclonal antibody licensed for the treatment of metastatic cancer. It works by inhibiting vascular endothelial growth factor (VEGF), a chemical mediator that promotes new blood vessel growth within tumours. It has been used off-label (without a licence)to treat neovascular age-related macular degeneration (nAMD - a common eye disease with loss of vision), and other conditions characterised by intraocular new blood vessel growth (neovascularisation).

Ranibizumab (Lucentis) is another monoclonal antibody to VEGF: it was directly cleaved from Bevacizumab to produce a lower molecular weight molecular that is more appropriate for use inside the eye. Whereas bevacizumab has been used off-label to treat nAMD, ranibizumab has been the subject of large Phase III controlled clinical trials (4,5) with robust safety analysis, and is licensed for the treatment of nAMD in Europe, the USA, and most countries. Neither bevacizumab nor ranibizumab are licensed for the treatment of diabetic retinopathy.

If ranibizumab produces a similar effect to pilot studies of bevacizumab in diabetic vitreous haemorrhage,(1,2,3) then it has the potential to speed visual recovery and reduce the likelihood of patients requiring surgery, with the attendant costs, patient inconvenience, and operative risks. Whilst bevacizumab is a cheaper alternative to ranibizumab, it does not yet have such extensive safety data. Further, bevacizumab has to be prepared by the treating clinician or intermediate pharmacies in a dosage appropriate for intraocular use, whereas ranibizumab is provided by the manufacturer in a pre-packaged sterile syringe containing the dose designed for intraocular administration. The relative cost disparity is likely to reduce in the coming years, and a single intravitreal injection is a very small proportion of the total treatment costs for proliferative diabetic retinopathy, which consumes considerable health care resources, particularly if an eye operation is required.

This study will enrol patients who have already decided to undergo an eye operation (pars plana vitrectomy) to clear a persistent diabetic vitreous haemorrhage. Those in the treatment arm will have an intravitreal injection of ranibizumab at the same dose used for nAMD. It is hypothesised that this will promote clearance of the vitreous haemorrhage and that this, in turn, may mean that some patients do not need to proceed to vitrectomy.

Inclusion Criteria:

1. Adults (male or female over 18) with Type 1 or Type 2 diabetes mellitus
2. Grade 2-4 fundus obscuring diabetic vitreous haemorrhage of at least 2 months duration prior to screening in the study eye.
3. Subjects who have elected to undergo a therapeutic pars plana vitrectomy to clear persistent diabetic vitreous haemorrhage
4. Best corrected visual acuity from 40 letters (using 4 metre ETDRS visual acuity score) to perception of light in the study eye
5. Patients able and willing to give written and witnessed informed consent.

Exclusion Criteria:

1. The presence of tractional retinal elevation in the study eye, as detected by B mode ocular ultrasound or fundus biomicroscopy.
2. Other (non-diabetic) cause of vitreous haemorrhage
3. Other (non-diabetic) retinal vasculopathy in the study eye
4. Subjects who were listed for vitrectomy for recurrent vitreous haemorrhage alone, and not for persistent vitreous haemorrhage
5. Subjects whose planned vitrectomy was to have been combined with cataract surgery
6. Prior vitrectomy in the study eye
7. Visual acuity worse than 6/96 in the non study eye
8. Aphakia in the study eye
9. Pregnant (urine dipstick confirmed) or lactating women (women of childbearing potential should be advised to use appropriate contraception for three months following eye injection
10. Those with systemic or ocular contraindications to ranibizumab therapy
11. Sickle cell disease. Those with sickle trait may be included if there is no evidence of retinopathy in the non study eye.
12. Patients who have had an intravitreal injection of any therapeutic agent in the study eye
13. Subjects with active concomitant disease in the study eye, including uveitis and infection
14. Subjects with inadequate pupil dilation in the study eye, or other cause of significantly impaired fundus view
15. Subjects with potentially visually significant cataract in the study eye
16. Subjects who have undergone intraocular surgery in the study eye less than 6 months prior to screening with the exception of cataract surgery, which must have been at least 2 months prior to screening
17. Subjects who have commenced medications that target haemostasis within 3 months of screening, including antithrombotic, antiplatelet and anticoagulant therapy, or who are likely to commence or alter such medications during the course of the study. Subjects who have commenced treatment with these agents at least 3 months prior to screening, and who are stable on treatment, are eligible for inclusion.
18. Current participation in another drug or device clinical trial, or participation in such a clinical trial within the last year
19. Patients unable or unwilling to give informed consent
20. Patients unable or unwilling to return for follow up over 12 months
21. Any other condition or situation that, in the opinion of the investigator, may prevent the patient from complying with the study protocol