• The severity of intraoperative and postoperative preretinal hemorrhage [ Time Frame: Six months ] [ Designated as safety issue: No ]
  
• Reabsorption time of blood around the disc area [ Time Frame: Six months ] [ Designated as safety issue: No ]
  
• The changes of visual acuity [ Time Frame: Six months ] [ Designated as safety issue: Yes ]
  

From January 2007 to June 2007, consecutive patients undergoing primary pars plana vitrectomy with silicone oil infusion for complications of proliferative diabetic retinopathy will be recruited for the prospective study. The selection criteria are: 1) anticoagulant therapy has not been used prior to surgery or during post-operative follow-up period; 2) no medical history of blood diseases associated with abnormal blood coagulation is present.

Active PDR is defined as visible large new vessels within the proliferative tissue with fresh preretinal and/or vitreous hemorrhage. Decision of silicone oil infusion will be made before surgery when potential creation of multiple breaks or incomplete traction release is anticipated during surgery due to severe vitreoretinal adhesion. The morphological criteria set for silicone oil infusion are: severe active fibrovascular proliferation with broad vitreous attachment around the disc, arcade and extending to the periphery in at least 2 quadrants (≥grade 5 in Eliott's grading system);8 presence of macular-off traction or combined traction and rhegmatogenous retinal detachment.

Individual recruited patient will be randomly assigned to one of the two groups: group 1 will receive intravitreal injection of 1.25 mg of avastin (0.05 ml) 7 to 9 days before vitrectomy; group 2 will not receive avastin pretreatment. Standard 3 port pars plana vitrectomy will be performed followed by silicone oil (5000 CS) infusion. A total of 30 cases (15 in each group) will be recruited.

After surgery, patients will be kept in a prone position overnight, then allowed to lie on either side during sleep thereafter, but maintained a head-down position during waking hours for 2 weeks. Ophthalmological examinations will be performed in the first 4 days after surgery, then weekly for 4 weeks, biweekly for 1 month, and then monthly for at least 3 months. The preoperative, intraoperative, and postoperative data will be collected for each patient. These demographics and clinical findings include age, gender, study eye, types and duration of diabetes mellitus, systemic diseases such as hypertension, renal insufficiency (24 hours creatinine clearance estimated by Cockcroft and Gault equation), degree of intraoperative bleeding, duration of the surgery, combined lens extraction, and the use of scleral buckle. Data regarding the extent of preretinal blood in the first postoperative day; time, duration, frequency and treatment of recurrent vitreous hemorrhage; and the duration of postoperative follow-up will also be compiled. Results of ophthalmological examinations, including best corrected visual acuity, intraocular pressure, and lens status will be recorded.

Intraoperative bleeding will be graded in 3 levels: grade 1 is defined as minor bleeding stopped either spontaneously or by transient bottle elevation; grade 2 is defined as moderate bleeding resulting in broad sheaths of clots requiring endodiathermy to the bleeding sites to stop the bleeding; grade 3 is defined as thick clot formation covering half or more of the posterior pole or interfering with the surgical plane. Postoperative preretinal blood will be separated into 3 grades: isolated clots with total area less than 10 disc area and without involvement of the posterior pole (grade1); broad sheaths of clots with total area more than 10 disc area without involving the posterior pole (grade2); broad sheaths of clots with total area more than 10 disc area and with involvement of the posterior pole (grade3). Any noticeable increase of preretinal blood will be defined as recurrent hemorrhage.

The severity of intraoperative bleeding, the extent of immediate postoperative preretinal blood, reabsorption time of blood around the disc area, total reabsorption time of preretinal blood, the rate and treatment of recurrent vitreous hemorrhage, and the change of best-corrected visual acuity will be compared between groups 1 and 2. Visual acuity will be graded into three levels: low (≤1 meter counting fingers), moderate (>1 meter counting fingers, but < 20/200), and good (≥ 20/200).