A Phase 2 Evaluation of Anti-VEGF Therapy for Diabetic Macular Edema: Bevacizumab (Avastin)

Transcription

1 Diabetic Retinopathy Clinical Research Network A Phase 2 Evaluation of Anti-VEGF Therapy for Diabetic Macular Edema: Bevacizumab (Avastin) Version1.2 April 27, 2006 Avastin Protocol V1.2.doc

2 Contact Information Coordinating Center Jaeb Center for Health Research Amberly Drive, Suite 350 Tampa, FL Phone: Fax: Director: Roy W. Beck, M.D., Ph.D. Associate Director: Kimberly McLeod, M.P.H. Protocol Chair Ingrid U. Scott, M.D., M.P.H. Pennsylvania State University College of Medicine 500 University Drive, MC H097 Hershey, PA Phone: (717) Fax: (717) Avastin Protocol V1.2.doc

3 Table of Contents Chapter 1. Background Information and Study Synopsis Background and Rationale Background Information on Diabetic Macular Edema Rationale for Anti-VEGF Treatment Bevacizumab Animal Studies Potential Adverse Effects of Bevacizumab Summary of Study Rationale Study Objectives and Hypotheses Study Design and Synopsis of Protocol General Considerations Chapter 2. Subject Eligibility and Enrollment Identifying Eligible Subjects and Obtaining Informed Consent Subject Eligibility Criteria Subject-level Criteria Study Eye Criteria Fellow Eye Criteria Screening Evaluation and Baseline Testing Historical Information Baseline Testing Procedures Enrollment/Randomization of Eligible Patients Chapter 3. Macular Laser Photocoagulation Introduction Photocoagulation Technique Chapter 4. Bevacizumab Treatment Introduction Bevacizumab Intravitreal Injection Technique Sham Injection Technique Miscellaneous Considerations Delay in Giving 6-week Injection Deferral of Retreatment Chapter 5. Follow-up Visits and Additional Treatments Visit Schedule Testing Procedures Treatment During Week Period Eyes in All Groups with Substantial Improvement Eyes Not Meeting Deferral Criteria at 12 Weeks Bevacizumab Groups Laser Only Group Chapter 6. Miscellaneous Considerations in Follow-up Endophthalmitis Surgery for Vitreous Hemorrhage and Other Complications of Diabetic Retinopathy Laser Scatter (Panretinal) Photocoagulation (PRP) Macular Photocoagulation Treatment of Macular Edema in Nonstudy Eye Diabetes Management Subject Withdrawal and Losses to Follow-up Avastin Protocol V1.2.doc

4 Discontinuation of Study Contact Information Provided to the Coordinating Center Subject Reimbursement Chapter 7. Adverse Events Definition Recording of Adverse Events Reporting Serious or Unexpected Adverse Events Data and Safety Monitoring Committee Review of Adverse Events Risks Potential Adverse Effects of Bevacizumab Potential Adverse Effects of Intravitreal Injection Chapter 8. Statistical Methods Sample Size Analysis Plan Primary Analysis OCT Outcome Principles for Analysis Visual Acuity Outcome Additional Analysis Objectives Analysis Techniques to Assess Treatment and Dose Effect Analysis Techniques to Assess Duration Effect Safety Analysis Plan Confidence Interval for Safety Parameters Interim Treatment Group Analyses Additional Tabulations and Analyses Appendix I References Avastin Protocol V1.2.doc

5 CHAPTER 1. BACKGROUND INFORMATION AND STUDY SYNOPSIS 1.1 Background and Rationale Background Information on Diabetic Macular Edema Diabetic retinopathy is a major cause of visual impairment in the United States. 1-3 Diabetic macular edema (DME) is a manifestation of diabetic retinopathy that produces loss of central vision. Data from the Wisconsin Epidemiologic Study of Diabetic Retinopathy (WESDR) estimate that after 15 years of known diabetes, the prevalence of diabetic macular edema is approximately 20% in patients with type 1 diabetes mellitus (DM), 25% in patients with type 2 DM who are taking insulin, and 14% in patients with type 2 DM who do not take insulin. 1 In a review of three early studies concerning the natural history of diabetic macular edema, Ferris and Patz found that 53% of 135 eyes with diabetic macular edema, presumably all involving the center of the macula, lost two or more lines of visual acuity over a two year period. 4 In the Early Treatment Diabetic Retinopathy Study (ETDRS), 33% of 221 untreated eyes available for follow-up at the 3-year visit, all with edema involving the center of the macula at baseline, had experienced a 15 or more letter decrease in visual acuity score (equivalent to a doubling of the visual angle, e.g., 20/25 to 20/50, and termed moderate visual loss ). 5 In the ETDRS, focal photocoagulation (direct treatment to microaneurysms and grid treatment to diffuse edema) of eyes with clinically significant macular edema (CSME) reduced the risk of moderate visual loss by approximately 50% (from 24% to 12%, three years after initiation of treatment). 6 Therefore, 12% of treated eyes developed moderate visual loss in spite of treatment. Furthermore, approximately 40% of treated eyes that had retinal thickening involving the center of the macula at baseline still had thickening involving the center at 12 months, as did 25% of treated eyes at 36 months. 7 Although several treatment modalities are currently under investigation, the only demonstrated means to reduce the risk of vision loss from diabetic macular edema are laser photocoagulation, as demonstrated by the ETDRS, intensive glycemic control, as demonstrated by the Diabetes Control and Complications Trial (DCCT) 8 and the United Kingdom Prospective Diabetes Study (UKPDS) 9 and blood pressure control, as demonstrated by the UKPDS. 9 In the DCCT, intensive glucose control reduced the risk of onset of diabetic macular edema by 23% compared with conventional treatment. Long-term follow-up of patients in the DCCT show a sustained effect of intensive glucose control, with a 58% risk reduction in the development of diabetic macular edema for the DCCT patients followed in the Epidemiology of Diabetes Interventions and Complications Study. 10 The frequency of an unsatisfactory outcome with respect to proportion with vision improvement following laser photocoagulation in some eyes with diabetic macular edema has prompted interest in other treatment modalities. One such treatment is pars plana vitrectomy These studies suggest that vitreomacular traction, or the vitreous itself, may play a role in increased retinal vascular permeability. Removal of the vitreous or relief of mechanical traction with vitrectomy and membrane stripping may be followed by substantial resolution of macular edema and corresponding improvement in visual acuity. However, this treatment may be applicable only to a specific subset of eyes with diabetic macular edema that have a component of vitreomacular traction contributing to the edema. It also requires a complex surgical intervention with its inherent risks, recovery time, and expense. Other treatment modalities such as pharmacologic therapy with oral protein kinase C inhibitors and use of intravitreal corticosteroids are under investigation. The use of antibodies Avastin Protocol V1.2.doc 1-1

6 targeted at vascular endothelial growth factor (VEGF), such as in the current study, is another treatment modality that has generated considerable interest, and is currently being investigated in phase 3 trials of choroidal neovascularization in age-related macular degeneration (with pegaptanib or ranibizumab) or diabetic macular edema (with pegaptanib) Rationale for Anti-VEGF Treatment Diabetic macular edema results from abnormal leakage of macromolecules, such as lipoproteins, from retinal capillaries into the extravascular space followed by an oncotic influx of water into the extravascular space. 4 Abnormalities in the retinal pigment epithelium may also cause or contribute to diabetic macular edema. These abnormalities may allow increased fluid from the choriocapillaries to enter the retina or they may decrease the normal efflux of fluid from the retina to the choriocapillaris. 4 The mechanism of breakdown of the blood retina barrier at the level of the retinal capillaries and the retinal pigment epithelium may be due to changes to tight junction proteins such as occludin. 17 The increase in retinal capillary permeability and subsequent retinal edema may be the result of a breakdown of the blood retina barrier mediated in part by VEGF, a 45 kd glycoprotein. 18 Aiello, et al, demonstrated, in an in vivo model, that VEGF can increase vascular permeability. 18 Fifteen eyes of 15 albino Sprague-Dawley rats received an intravitreal injection of VEGF. The effect of intravitreal administration of VEGF on retinal vascular permeability was assessed by vitreous fluorophotometry. In all 15 eyes receiving an intravitreal injection of VEGF, a statistically significant increase in vitreous fluorescein leakage was recorded. In contrast, control eyes, which were fellow eyes injected with vehicle alone, did not demonstrate a statistically significant increase in vitreous fluorescein leakage. Vitreous fluorescein leakage in eyes injected with VEGF attained a maximum of 227% of control levels. Antonetti, et al, demonstrated that VEGF may regulate vessel permeability by increasing phosphorylation of tight junction proteins such as occludin and zonula occluden Sprague-Dawley rats were given intravitreal injections of VEGF and changes in tight junction proteins were observed through Western blot analysis. Treatment with alkaline phosphatase revealed that these changes were caused by a change in phosphorylation of tight junction proteins. This model provides, at the molecular level, a potential mechanism for VEGFmediated vascular permeability in the eye. Similarly, in human non-ocular disease states such as ascites, VEGF has been characterized as a potent vascular permeability factor (VPF). 20 The normal human retina contains little VEGF; however, hypoxia causes upregulation of VEGF production. 21 Vinores et al, using immunohistochemical staining for VEGF, demonstrated that increased VEGF staining was found in retinal neurons and retinal pigment epithelium in human eyes with diabetic retinopathy. 21 As the above discussion suggests, attenuation of the effects of VEGF provides a rationale for treatment of macular edema associated with diabetic retinopathy. Several different anti-vegf drugs exist, including pegaptanib (Macugen, Eyetech Pharmaceuticals), ranibizumab (Lucentis, Genentech, Inc.), and bevacizumab (Avastin, Genentech, Inc.). The drugs other than bevacizumab are summarized below and described in more detail in a separate Background and Rationale for Anti-VEGF Therapies document. Pegaptanib and ranibizumab have had more extensive testing for age-related macular degeneration (AMD) than for diabetic macular edema. Only pegaptanib has been studied in a phase 2 trial for diabetic macular edema, and results were promising with respect to decreasing edema compared with no treatment. Avastin Protocol V1.2.doc 1-2

7 Pegaptanib (Macugen, Eyetech Pharmaceuticals) Pegaptanib is an aptamer consisting of a pegylated modified oligonucleotide which binds to extracellular VEGF isoform 165 (the predominant isoform). Pegaptanib is approved for the treatment of neovascular age-related macular degeneration and is currently being studied for the treatment of macular edema associated with diabetic retinopathy and retinal vein occlusion. Significant adverse events associated with intravitreal pegaptanib administration that have been reported appear to be related to the injection procedure rather than a drug effect and include endophthalmitis in 1.3%, traumatic lens injury in 0.7%, and retinal detachment in 0.6%. Ranibizumab (Lucentis, Genentech, Inc.) Ranibizumab is a humanized monoclonal antibody fragment which, like pegaptanib, competitively inhibits VEGF in the extracellular space. Unlike pegaptanib, however, ranibizumab is designed to block all isoforms of VEGF-A and is currently being studied for the treatment of neovascular AMD Serious ocular adverse events reported to occur more frequently in patients treated with ranibizumab compared with a control group include uveitis (<1%) possibly related to a drug effect rather than the injection procedure and endophthalmitis (<1%). No imbalance in serious non-ocular adverse events was observed Bevacizumab Bevacizumab, a full-length recombinant humanized monoclonal antibody which binds to and neutralizes all VEGF-A isoforms, is currently approved as a systemic therapy by the Food and Drug Administration (FDA) for the treatment of metastatic colorectal cancer. It has been shown that combining bevacizumab with chemotherapy improves survival in metastatic colorectal cancer 22, 23 compared with chemotherapy alone. When it was observed that fluorescein-conjugated bevacizumab leaked from laser-induced choroidal neovascularization in a cynomolgus monkey after systemic administration, it was thought that systemic therapy in patients with age-related macular degeneration would also cause leakage of bevacizumab that would inhibit extracellular VEGF and improve vision. 24 This led to the initiation of an open-label prospective clinical study, the Systemic Bevacizumab for Neovascular AMD (SANA) Study. Data on the first 9 patients enrolled in this study through the first 12 weeks of follow-up have been published. 25 Patients were treated with a baseline infusion of bevacizumab (5 mg/kg), followed by one or two additional doses given at 2- week intervals. In the study eyes, significant improvements in visual acuity were observed within 1 week of treatment, and by 12 weeks, the median and mean visual acuity letter scores increased by 8 letters (p=0.011) and 12 letters (p=0.008), respectively. The median and mean central retinal thickness (CRT) measurements decreased by 157 um (p=0.008) and 177 um (p=0.001), respectively. In the fellow eyes at 12 weeks, median and mean visual acuity improvement was 27 letters (p=0.018) and 16 letters (p=0.012), respectively, and the median and mean CRT decreased by 59 um (p=0.028) and 92 um (p=0.06). Fluorescein angiography demonstrated a marked reduction or an absence of leakage of choroidal neovascularization in all study eyes. No serious ocular or systemic adverse events were observed. By 6 weeks, the only adverse event observed was a mild elevation of systolic blood pressure, controlled by changing or initiating antihypertensive medication; by 12 weeks, the elevation of systolic blood pressure was no longer significant. Intravitreal injections of bevacizumab have been given off-label for age-related macular degeneration and diabetic macular edema. Case reports of the intravitreal use of bevacizumab have been published. In one case, a 63-year-old woman with predominantly classic subfoveal choroidal neovascularization associated with age-related macular degeneration in her left eye, progressing Avastin Protocol V1.2.doc 1-3

8 despite treatment with photodynamic therapy/intravitreal triamcinolone acetonide and pegaptanib, was administered a single 1.0 mg intravitreal injection of bevacizumab (0.04 ml of commercially available bevacizumab at a concentration of 25 mg/ml). Visual acuity prior to injection was 20/125. One week following injection, the patient reported resolution of visual distortion, and optical coherence tomography (OCT) demonstrated resolution of subretinal fluid. At 4 weeks, visual acuity remained stable, and the subretinal fluid had not recurred. 26 In another case, a 68-yearold man with macular edema associated with central retinal vein occlusion was administered a 1.0 mg intravitreal injection of bevacizumab in his left eye. Visual acuity prior to injection was 20/200; one week after the injection, acuity improved to 20/50 and OCT demonstrated resolution of the macular edema. Four weeks post-injection, visual acuity was 20/60 and OCT showed continued resolution of the macular edema. In both of these cases, systemic blood pressure of the patients remained stable. 26 The authors note that, unlike chronic high-dose (5 mg/kg) intravenous infusion of bevacizumab for colorectal cancer (which is associated with risks of hypertension and thromboembolic events), a 1 mg intravitreal dose ( fold less drug) is unlikely to cause systemic side effects. Several recently published short-term studies have reported that intravitreal injections of bevacizumab were associated with decreased macular edema and improved visual acuity with no corresponding safety concerns. In a retrospective study of 16 eyes with macular edema associated with central retinal vein occlusion, treatment with intravitreal injections of bevacizumab (1.25 mg in 0.05 ml) was associated with a decrease in mean central macular thickness from 887 microns at baseline to 372 microns at 1 month (p<0.001) and was associated with an improvement in mean visual acuity from 20/600 at baseline to 20/200 at 1 month (p<0.001) and 20/138 at a mean of 3 months post-injection (p<0.001). 27 In the study, eyes received a mean of 2.8 intravitreal injections of bevacizumab during the study period (for a total of 45 injections administered during the study) and no adverse events were observed during the mean 3-month follow-up period. 27 Avery et al reported a series of 81 eyes with neovascular age-related macular degeneration which received intravitreal bevacizumab (1.25 mg in 0.05 ml) on a monthly basis until macular edema, subretinal fluid (SRF), and/or pigment epithelial detachment (PED) resolved. Fifty-five percent of eyes had a reduction of >10% of baseline retinal thickness at 1 week post-injection. At 4 weeks post-injection, 30 of 81 eyes demonstrated complete resolution of retinal edema, SRF, and PED; of the 51 eyes with 8 weeks of follow-up, 25 demonstrated complete resolution of retinal thickening, SRF, and PED. At 1, 4, 8, and 12 weeks, the mean retinal thickness of the central 1 mm was decreased by 61, 92, 89, and 67 um (p< for 1, 4, and 8 weeks and p<0.01 for 12 weeks). At 4 and 8 weeks, mean Snellen visual acuity improved from 20/200 to 20/125 (p<0.0001) and median vision improved from 20/200 to 20/80 - at 4 weeks and from 20/200 to 20/80 at 8 weeks. No significant ocular or systemic adverse effects were observed. 28 In a study of 9 eyes treated with intravitreal bevacizumab (1.25 mg in 0.05 ml) for neovascular age-related macular degeneration, Maturi et al reported that all four eyes which underwent multifocal electroretinography (mf-erg) testing before and 1 month after bevacizumab treatment demonstrated improvement in the mf-erg macular function responses; 29 in the study, all five eyes which underwent Ganzfeld electroretinography (G- ERG) testing before and at 1 week and 1 month after bevacizumab treatment demonstrated no significant changes in ERG response. No ocular or systemic adverse events were reported in this short-term study. 29 Numerous other case reports and case series have been presented at meetings, and off label bevacizumab use has become widespread despite the lack of controlled, prospective studies demonstrating its efficacy and safety. We estimate that several thousand patients have now been Avastin Protocol V1.2.doc 1-4

9 treated with intravitreal bevacizumab for age-related macular degeneration or diabetic macular edema. A survey was conducted of DRCRnet investigators with regard to intravitreal bevacizumab use for diabetic macular edema. Twenty-nine investigators responded; 20 indicated that they had used intravitreal bevacizumab, with some patients having demonstrated a marked beneficial response and others demonstrating less of a response. A dose of 1.25 mg has been most widely used, with 2.5 mg also having been used. There have been a sufficient number of well documented cases of reduction in macular edema after intravitreal bevacizumab injection without corresponding safety concerns that we believe a formal phase 1 study prior to proceeding with a phase 2 study is unnecessary. Given the rapidity with which intravitreal bevacizumab is being incorporated into clinical practice among retinal specialists, delaying a phase 2 study would also be disadvantageous as patient recruitment into clinical trials would become increasingly difficult Animal Studies An experimental study was performed to investigate potential intraocular toxicity associated with intravitreal bevacizumab injection. Ten rabbits were injected intravitreally with 2.5 mg/0.1 ml bevacizumab; fellow eyes served as control eyes and received intravitreal injections of 0.1 ml of saline. Electroretinogram (ERG) measurements were performed at 3 hours, 3 days, 1 week, 2 weeks, and 4 weeks post-injection. Visual evoked potential (VEP) measurements were performed at 4 weeks post-injection. Histological examination of treated and control retinae were performed at 4 weeks post-injection. No electrophysiologic evidence of intraocular toxicity was observed in this small study. Full-thickness retinal penetration of drug was present at 24 hours and was essentially absent at 4 weeks. There was no clinical evidence of inflammation following the injection. 30 In another study, retinal toxicity was evaluated in rabbits following intravitreal injections of four varying concentrations of bevacizumab. Four concentrations of bevacizumab were employed in this study: 500 µg/0.1 ml, 1.0 mg/0.1 ml, 2.5 mg/0.1 ml, and 5.0 mg/0.2 ml. Each concentration was injected intravitreally into one eye of each of three rabbits and 0.1mL of sterile balanced saline solution was injected into the contralateral eyes. ERG measurements were performed before the drug treatment and at 14 days post-injection. No significant decrease in amplitude was observed in any group. Only one eye injected with the highest concentration (5 mg/0.2ml) demonstrated a decrease of 11% when compared to baseline; one eye of the control group demonstrated a decrease of 13%. Light microscopy did not demonstrate any retinal toxicity in any eyes. One eye in the 5 mg dose group had some inflammatory cells in the vitreous. There was no clinical evidence of postinjection complications such as cataract, vitreous hemorrhage, or retinal detachment Potential Adverse Effects of Bevacizumab Although no serious ocular or systemic adverse events were identified in the SANA Study or in the large case series of intravitreal bevacizumab by Avery et al, patients with advanced metastatic colorectal cancer receiving concomitant chemotherapy and systemic bevacizumab had an increased risk of potentially fatal thromboembolic events compared with cancer patients receiving chemotherapy alone according to an FDA warning letter in August This information was then updated in January 2005 with a revised package insert containing warnings regarding thromboembolic events This risk was not obvious from the phase 3 clinical trials that led to bevacizumab s approval by the FDA and became apparent only after Genentech, Inc. performed a meta-analysis on all clinical trial results using bevacizumab. The overall risk of thromboembolic Avastin Protocol V1.2.doc 1-5

10 events was approximately twice as high in patients receiving infusions of 5-fluorouracil-based chemotherapy plus bevacizumab, with an estimated overall rate of up to 4.4%. It should be noted that, as a precaution in the SANA Study, patients with a history of cerebrovascular accident, transient ischemic attack, myocardial infarction, angina, or any other thromboembolic disease were excluded. 34 Other systemic side effects which have been reported in cancer studies using bevacizumab include hypertension, epistaxis, hemopytsis, and proteinuria. 35 An additional potential risk associated with bevacizumab therapy is delayed wound healing after surgery. Finally, since maintenance of the choriocapillaris in a normal eye is believed to require VEGF produced by the retinal pigment epithelium, 36 anti-vegf therapy may have deleterious intraocular effects. Unlike the chronic high-dose (5 mg/kg) intravenous infusion of bevacizumab used in colorectal cancer therapy, a 2.5 mg intravitreal dose (approximately 200 fold less drug) or a 1.25 mg dose is unlikely to cause systemic side effects. Furthermore, review with the DRCR Network investigators have failed to identify reports of blood pressure or heart problems when bevacizumab has been given in the eye, although these patients were not monitored methodically or specifically for these side effects Summary of Study Rationale Increased VEGF levels have been demonstrated in the retina and vitreous of human eyes with diabetic retinopathy. VEGF, also known as vascular permeability factor, has been demonstrated to increase vessel permeability by increasing the phosphorylation of tight junction proteins, and has been shown to increase retinal vascular permeability in in vivo models. Anti-VEGF therapy, therefore, may represent a useful therapeutic modality which targets the underlying pathogenesis of diabetic macular edema. Bevacizumab is currently approved for the treatment of metastatic colorectal cancer, and published case reports and widespread clinical use have suggested its efficacy in the treatment of neovascular age-related macular degeneration and macular edema associated with diabetes and central retinal vein occlusion. To date, no evidence of ocular inflammation or other adverse events has been noted in association with intravitreal injection of bevacizumab. However, a study has not been conducted to evaluate its efficacy and safety. In view of the widespread use of bevacizumab, such a study is important to conduct. From a public health perspective, an intravitreal bevacizumab study is also important to conduct because of the relatively low cost of the bevacizumab drug. As noted earlier, bevacizumab is marketed for systemic use for colon cancer. The dose used in the eye is a fraction of the systemic dose and costs $25 to $50 per dose. The two doses of bevacizumab being evaluated in this study will be 1.25 mg, which is the dose that has most commonly been used in clinical practice, and 2.5 mg, which has also been used though less commonly. A lower dose than 1.25 mg would create difficulties with dilution and the accuracy of injection of a small volume. The optimal interval for the bevacizumab doses is not known. Six weeks has been selected for this study as it is not believed that the effect will last longer than this. Retinal thickening and visual acuity will be measured at 3 and 6 weeks to provide the requisite information to judge the duration of effect. Avastin Protocol V1.2.doc 1-6

11 There is expected to be a beneficial cumulative effect of multiple doses. A total of two doses, spaced 6 weeks apart, was selected for the study with the primary outcome 3 weeks after the second dose. The decision as to whether to proceed to a phase 3 trial will be based on the observation of a substantial reduction in retinal thickening in the bevacizumab-treated eyes compared with the lasertreated eyes and at least a suggestion of benefit on visual acuity, plus a safety profile of minimal risk. 1.2 Study Objectives and Hypotheses To assess the dose and dose interval related effects of intravitreal administered bevacizumab on central retinal thickness and visual acuity in subjects with diabetic macular edema (DME). To assess the effect of intravitreal bevacizumab combined with macular photocoagulation in DME. To assess the safety of intravitreal bevacizumab in subjects with DME. This phase 2 study is being conducted (1) to determine whether the conduct of a phase 3 trial has merit and (2) to provide information needed to design a phase 3 trial. The study is not designed to establish the efficacy of bevacizumab in the treatment of diabetic macular edema. 1.3 Study Design and Synopsis of Protocol A. Study Design Phase 2 randomized, multi-center clinical trial. B. Major Eligibility Criteria Age >=18 years. Study eye with best corrected E-ETDRS acuity >= 24 letter score (20/320 or better) and <=78 letter score (20/32 or worse). Study eye with center-involved DME present on clinical exam and OCT with central subfield thickness >=275 microns. C. Sample Size: 100 subjects with one study eye; a minimum of 20 eyes in each of the 5 treatment groups defined below. D. Treatment Groups Randomized equally to the following 5 groups stratified by visual acuity: 1) Laser photocoagulation at baseline If edema present at 12 weeks, can be treated with 2 intravitreal injections of 1.25 mg bevacizumab spaced 6 weeks apart 2) 1.25 mg intravitreal injection of bevacizumab at baseline and 6 weeks 3) 2.5 mg intravitreal injection of bevacizumab at baseline and 6 weeks 4) 1.25 mg intravitreal injection of bevacizumab at baseline (sham injection at 6 weeks) 5) 1.25 mg intravitreal injection of bevacizumab at baseline, laser photocoagulation at 3 weeks, and intravitreal injection of 1.25 mg bevacizumab at 6 weeks During the first 12 weeks, no other treatment for DME is to be given in the study eye. Avastin Protocol V1.2.doc 1-7

12 During weeks 13-24, treatment depends on the response to the treatment given during the first 12 weeks. If treatment deferral criteria are met (see section 5.3), no additional treatment for DME is given. If treatment deferral criteria are not met, the subject is considered to have met the efficacy endpoint for this phase and then can be treated at the discretion of the investigator. After 24 weeks, follow-up is for safety, and treatment is at investigator discretion. Follow up continues for each subject until 70 weeks, which will be at least 1 year after the last protocolspecified dose of bevacizumab. Avastin Protocol V1.2.doc 1-8

13 E. Schedule of Study Visits and Examination Procedures Baseline* Baseline +4 day (2-7d) 3w +/- 3d 6w* +/- 3d 6w +4 day (2-7d) E-ETDRS visual X a X X X X X X X X X X acuity b 9w +/- 3d 12w +/- 3d 18w +/- 1w 24w +/-1w 41w +/- 2w 70w +/- 2w OCT c X a X X X X X X X d X d Fundus Photos (7F) X Eye Exam X X X X X X X X X X X Blood pressure X X X X X X X X X HbA1c Notes X 1. A medical history will be elicited at baseline and an updated history at each visit. Concomitant medications will be recorded at baseline and updated at each visit. Adverse events will be recorded at each visit. 2. Testing is on both eyes at each visit, except for OCT and photos which are only performed on the study eye. *injection of study drug at baseline and 6 weeks. a = If completed more than 8 days prior to randomization, must be repeated on day of randomization. b= measurement of best-corrected visual acuity includes protocol refraction at baseline and at 9 weeks, and at other visits if there is a 10 or more letter decrease from baseline; otherwise, refraction from most recent visit is used. E- ETDRS refers to electronic ETDRS testing using the Electronic Visual Acuity Tester that has been validated against 4-meter chart ETDRS testing. 37 c= OCT3, software version 4 (or later). d= only if performed as part of usual care. 1.4 General Considerations The study is being conducted in compliance with the policies described in the DRCRnet Policies document, with the ethical principles that have their origin in the Declaration of Helsinki, with the protocol described herein, and with the standards of Good Clinical Practice. The bevacizumab package insert provides details about the properties of the study drug and its formulation. The DRCRnet Procedures Manuals (Visual Acuity-Refraction Testing Procedures Manual, Photography and OCT Testing Procedures Manual, and Study Procedures Manual) provide details of the examination procedures and intravitreal injection procedure. Data will be collected directly in electronic case report forms, which will be considered the source data. There is no restriction on the number of subjects to be enrolled by any one site. Avastin Protocol V1.2.doc 1-9

14 CHAPTER 2. SUBJECT ELIGIBILITY AND ENROLLMENT 2.1 Identifying Eligible Subjects and Obtaining Informed Consent A minimum of 100 subjects are expected to be enrolled with a goal to enroll an appropriate representation of minorities. As the enrollment goal approaches, sites will be notified of an end date for recruitment. Subjects who have signed an informed consent form can be randomized up until the end date, which means the recruitment goal might be exceeded. Potential eligibility will be assessed as part of a routine-care examination. Prior to completing any procedures or collecting any data that are not part of usual care, written informed consent will be obtained. For subjects who are considered potentially eligible for the study based on a routine-care exam, the study protocol will be discussed with the subject by a study investigator and clinic coordinator. The subject will be given the Informed Consent Form to read. Subjects will be encouraged to discuss the study with family members and their personal physician(s) before deciding whether to participate in the study. Consent may be given in two stages (if approved by the IRB). The initial stage will provide consent to complete any of the screening procedures needed to assess eligibility that have not already been performed as part of a usual-care exam. The second stage will be obtained prior to randomization and will be for participation in the study. A single consent form will have two signature/date lines for the subject: one for the subject to give consent for the completion of the screening procedures and one for the subject to give consent for the randomized trial. Subjects will be provided with a copy of the signed Informed Consent Form. Once a subject is randomized, that subject will be counted regardless of whether the assigned treatment is received or not. Thus, the investigator must not proceed to randomize a subject until he/she is convinced that the subject is eligible and will accept assignment to any one of the five treatment groups. 2.2 Subject Eligibility Criteria Subject-level Criteria Inclusion To be eligible, the following inclusion criteria (1-5) must be met: 1. Age >= 18 years Subjects <18 years old are not being included because DME is so rare in this age group that the diagnosis of DME may be questionable. 2. Diagnosis of diabetes mellitus (type 1 or type 2) Any one of the following will be considered to be sufficient evidence that diabetes is present: Current regular use of insulin for the treatment of diabetes Current regular use of oral anti-hyperglycemia agents for the treatment of diabetes Documented diabetes by ADA and/or WHO criteria (see Procedures Manual for definitions) 3. At least one eye meets the study eye criteria listed in section Fellow eye meets criteria in section Avastin Protocol V1.2.doc 2-1

15 Able and willing to provide informed consent. Exclusion A subject is not eligible if any of the following exclusion criteria (6-15) are present: 6. Significant renal disease, defined as a history of chronic renal failure requiring dialysis or kidney transplant. 7. A condition that, in the opinion of the investigator, would preclude participation in the study (e.g., unstable medical status including blood pressure, cardiovascular disease, and glycemic control). Subjects in poor glycemic control who, within the last 4 months, initiated intensive insulin treatment (a pump or multiple daily injections) or plan to do so in the next 4 months should not be enrolled. 8. Participation in an investigational trial within 30 days of randomization that involved treatment with any drug that has not received regulatory approval at the time of study entry. Note: subjects cannot receive another investigational drug while participating in the study during the first 24 weeks. 9. Known allergy to any component of the study drug. 10. Blood pressure > 180/110 (systolic above 180 OR diastolic above 110). If blood pressure is brought below 180/110 by anti-hypertensive treatment, subject can become eligible. 11. Major surgery within 28 days prior to randomization or major surgery planned during the next 6 months. Major surgery is defined as a surgical procedure that is more extensive than fine needle biopsy/aspiration, placement of a central venous access device, removal/biopsy of a skin lesion, or placement of a peripheral venous catheter. 12. Myocardial infarction, other cardiac event requiring hospitalization, stroke, transient ischemic attack, or treatment for acute congestive heart failure within 6 months prior to randomization. 13. Systemic anti-vegf or pro-vegf treatment within 3 months prior to randomization. These drugs cannot be used during the first 6 months of the study. 14. For women of child-bearing potential: pregnant or lactating or intending to become pregnant within the next 6 months. Female subjects should be questioned about the potential for pregnancy. Investigator judgment is used to determine when a pregnancy test is needed. 15. Subject is expecting to move out of the area of the clinical center to an area not covered by another clinical center during the first 6 months of the study Study Eye Criteria The subject must have one eye meeting all of the inclusion criteria (a-e) and none of the exclusion criteria (f-r) listed below. Subjects can have only one study eye. If both eyes are eligible, the study eye will be selected by the investigator and subject. Avastin Protocol V1.2.doc 2-2

16 The eligibility criteria for a study eye are as follows: Inclusion a. Best corrected E-ETDRS visual acuity letter score of >= 24 (i.e., 20/320 or better) and <= 78 (i.e., 20/32 or worse) within 8 days of randomization. b. On clinical exam, definite retinal thickening due to diabetic macular edema involving the center of the macula. If the DME consists of circumscribed, focal leakage that the investigator believes should be treated with laser and not other treatments, the subject should not be randomized. c. OCT central subfield >=275 microns within 8 days of randomization. d. Media clarity, pupillary dilation, and subject cooperation sufficient for adequate fundus photographs. e. If prior macular photocoagulation has been performed, the investigator believes that the study eye may possibly benefit from additional photocoagulation. Exclusions The following exclusions apply to the study eye only (i.e., they may be present for the nonstudy eye): f. Macular edema is considered to be due to a cause other than diabetic macular edema. An eye should not be considered eligible if: (1) the macular edema is considered to be related to cataract extraction or (2) clinical exam and/or OCT suggest that vitreoretinal interface abnormalities disease (e.g., a taut posterior hyaloid or epiretinal membrane) is the primary cause of the macular edema. g. An ocular condition is present such that, in the opinion of the investigator, visual acuity would not improve from resolution of macular edema (e.g., foveal atrophy, pigmentary changes, dense subfoveal hard exudates, nonretinal condition). h. An ocular condition is present (other than diabetes) that, in the opinion of the investigator, might affect macular edema or alter visual acuity during the course of the study (e.g., vein occlusion, uveitis or other ocular inflammatory disease, neovascular glaucoma, Irvine-Gass Syndrome, etc.). i. Substantial cataract that, in the opinion of the investigator, is likely to be decreasing visual acuity by 3 lines or more (i.e., cataract would be reducing acuity to 20/40 or worse if eye was otherwise normal). j. History of treatment for DME at any time in the past 3 months (such as focal/grid macular photocoagulation, intravitreal or peribulbar corticosteroids, anti-vegf drugs, or any other treatment). If the investigator believes that there may still be a substantial effect 3 months after prior treatment (e.g., dose of intravitreal triamcinolone higher than 4 mg), the eye should not be included. k. History of panretinal scatter photocoagulation (PRP) within 4 months prior to randomization. l. Anticipated need for PRP in the 6 months following randomization. Avastin Protocol V1.2.doc 2-3

17 m. History of prior pars plana vitrectomy. n. History of major ocular surgery (including cataract extraction, scleral buckle, any intraocular surgery, etc.) within prior 6 months or anticipated within the next 6 months following randomization. o. History of YAG capsulotomy performed within 2 months prior to randomization. p. Aphakia. q. Uncontrolled glaucoma (in investigator s judgment). r. Exam evidence of external ocular infection, including conjunctivitis, chalazion, or significant blepharitis Fellow Eye Criteria The fellow eye must meet the following criteria: a. Best corrected E-ETDRS visual acuity letter score >= 19 (i.e., 20/400 or better). b. No anti-vegf treatment within the past 3 months and no expectation of such treatment in next 3 months. 2.3 Screening Evaluation and Baseline Testing Historical Information A history will be elicited from the subject and extracted from available medical records. Data to be collected will include: age, gender, ethnicity and race, diabetes history and current management, other medical conditions, medications being used, and ocular diseases, surgeries, and treatment Baseline Testing Procedures The following procedures are needed to assess eligibility and/or to serve as baseline measures for the study. If a procedure has been performed (using the study technique and by study certified personnel) as part of usual care, it does not need to be repeated specifically for the study if it was performed within the defined time windows specified below. The testing procedures are detailed in the DRCRnet Procedures Manuals (Visual Acuity- Refraction Testing Procedures Manual, Photography Testing Procedures Manual, and Study Procedures Manual). Visual acuity testing, ocular exam, fundus photography, and OCT will be performed by DRCR.net certified personnel. The fundus photographs and OCTof the study eye will be sent to the Fundus Photograph Reading Center for grading, but subject eligibility is determined by the site (i.e., subjects deemed eligible by the investigator will be randomized without need for Reading Center confirmation). In some cases, assessment of eligibility and the baseline treatment (laser or bevacizumab injection, which must be given on day of randomization) will require at least two visits. For this reason, maximum time windows from the completion of each procedure to the day of randomization have been established. 1. Electronic-ETDRS visual acuity testing at 3 meters using the Electronic Visual Acuity Tester (including protocol refraction) in each eye. (within 8 days prior to randomization) This testing procedure has been validated against 4-meter ETDRS chart testing. 37 Avastin Protocol V1.2.doc 2-4

18 OCT on study eye. (within 8 days prior to randomization) 3. Ocular examination on each eye including slit lamp, measurement of intraocular pressure, and dilated fundus examination. (within 21 days prior to randomization) 4. ETDRS protocol 7-standard field stereoscopic fundus photography on study eye (fields 1M, 2, 3M, 4, 5, 6, 7, reflex). (within 21 days prior to randomization) 5. Measurement of blood pressure. 6. Laboratory Testing- HbA1c. HbA1c does not need to be repeated if available in the prior 3 months. If not available at the time of randomization, the patient may be enrolled but the test must be obtained within 3 weeks after randomization. 2.4 Enrollment/Randomization of Eligible Patients 1. Prior to randomization, the subject s understanding of the trial, willingness to accept the assigned treatment group, and commitment to the follow-up schedule should be reconfirmed. 2. Treatment (laser photocoagulation or intravitreal bevacizumab injection) must be given on the day of randomization; therefore, a patient should not be randomized until this is possible. 3. Randomization is completed on the DRCRnet website. 4. The randomization schedule will be stratified by visual acuity (>= 66 letters/ <= 65 letters) equally to one of the 5 treatment groups. 1) Laser photocoagulation at baseline (referred to as laser only group) If edema present at 12 weeks, can be treated with 2 intravitreal injections of 1.25 mg bevacizumab spaced 6 weeks apart 2) 1.25 mg intravitreal injection of bevacizumab at baseline and 6 weeks 3) 2.5 mg intravitreal injection of bevacizumab at baseline and 6 weeks 4) 1.25 mg intravitreal injection of bevacizumab at baseline only and sham injection at 6 weeks 5) 1.25 mg intravitreal injection of bevacizumab at baseline, macular laser at 3 weeks, and 1.25 mg bevacizumab at 6 weeks (referred to as bevacizumab+laser group). Avastin Protocol V1.2.doc 2-5

19 CHAPTER 3. MACULAR LASER PHOTOCOAGULATION 3.1 Introduction Eyes assigned to the laser only group will be treated with focal laser using the modified-etdrs technique. The laser photocoagulation will be given on the day of randomization. Eyes assigned to the bevacizumab+laser group will be treated with focal laser using the modified-etdrs technique. The laser photocoagulation will be given at the three-week visit. 3.2 Photocoagulation Technique The laser treatment session should generally be completed in a single sitting. The photocoagulation treatment technique, as described below, is a modification of the ETDRS technique and is the treatment approach that is commonly used in clinical practice. Posttreatment photographs (field 2 stereo) may be requested on selected subjects by the Reading Center. Burn Characteristic Focal Treatment Change in MA Color with Focal Treatment Burn Size for Focal Treatment Burn Duration for Focal Treatment Grid Treatment Area Considered for Grid Treatment Burn Size for Grid Treatment Burn Duration for Grid Treatment Burn Intensity for Grid Treatment Burn Separation for Grid Treatment Focal / Grid Photocoagulation (Modified-ETDRS technique) Focally treat all leaking microaneurysms in areas of retinal thickening between 500 and 3000 microns from the center of the macula (but not within 500 microns of disc) Not required, but at least a mild gray-white burn should be evident beneath all microaneurysms 50 microns 0.05 to 0.1 sec Applied to all areas with diffuse leakage or nonperfusion within area described below for treatment 500 to 3000 microns superiorly, nasally and inferiorly from center of macula 500 to 3500 microns temporally from macular center No burns are placed within 500 microns of disc 50 microns 0.05 to 0.1 sec Barely visible (light gray) 2 visible burn widths apart 679 Wavelength (Grid and Focal Treatment) MA = microaneurysm Green to yellow wavelengths Avastin Protocol V1.2.doc 3-1