VISION 2004.

Methods	Included trials: 2 concurrent RCTs (EOP 1003; EOP 1004) Number randomized (total and per group): 1208 participants randomly assigned to study treatment: 297 in 0.3 mg pegaptanib group, 305 in 1.0 mg pegaptanib group, 302 in 3.0 mg pegaptanib group, and 304 in sham injection group Exclusions after randomization: 22 total participants: 18 participants did not receive at least 1 injection, and 4 participants were not included in efficacy analyses because "sufficiently standardized assessment of visual acuity was not completed at baseline" Number analyzed (total and per group): 1186 participants at 1 year: 294 in 0.3 mg pegaptanib group, 300 in 1.0 mg pegaptanib group, 296 in 3.0 mg pegaptanib group, and 296 in sham injection group Unit of analysis: individuals (1 study eye per participant) Losses to follow‐up: 101 at 1 year: 23 in 0.3 mg pegaptanib group, 25 in 1.0 mg pegaptanib group, 32 in 3.0 mg pegaptanib group, and 21 in sham injection group Compliance: approximately 90% of participants completed the study Intention‐to‐treat analysis: no; 22 participants enrolled and randomized were not included in analysis Reported power calculation: yes; sample of 244 participants in each group for power of 95%; at least 270 participants were recruited for each group, assuming 10% would have missing data Study design comment: at 54 weeks, participants were re‐randomized; those in pegaptanib groups were randomized to discontinue treatment or continue with same dose; those in sham group were randomized to 1 of 5 groups: discontinue sham injections, continue with sham injections, or receive injections with 0.3, 1.0, or 3.0 mg pegaptanib
Participants	Country: USA, Canada, Austria, Belgium, Czech Republic, Denmark, France, Germany, Hungary, Israel, Italy, the Netherlands, Poland, Portugal, Spain, Switzerland, UK, Brazil, Chile, Colombia, and Australia (117 study centers) Age: mean age in EOP 1003 was 77 years, and in EOP 1004 was 75 years Gender (per cent): 696/1190 (58%) women and 494/1190 (42%) men (based on those receiving at least 1 study treatment) Inclusion criteria: age 50 years or older; subfoveal CNV lesion secondary to AMD; BCVA of 20/40 to 20/320 in study eye and 20/800 or better in fellow eye; all angiographic subtypes of total lesion size up to and including 12 disc areas Exclusion criteria: subretinal hemorrhage in study eye ≥ 50% of lesion area; < 50% of lesion with active CNV; > 1 previous PDT treatment; PDT treatment < 8 weeks or > 13 weeks before baseline visit; IOP > 23 mmHg; without clear ocular media; inadequate pupillary dilation for stereoscopic fundus photography; atrophy > 25% of total lesion area or subfoveal scarring in study eye; history of previous subfoveal thermal laser therapy or previous or concomitant therapy with any investigational therapy to treat AMD; need for cataract surgery within 2 years; other potential causes of CNV such as myopia; ocular histoplasmosis syndrome, angioid streaks, choroidal rupture, or multifocal choroiditis; any intraocular surgery within 3 months or extrafoveal/juxtafoveal laser within 2 weeks of study entry; previous posterior vitrectomy or scleral buckling surgery; presence of retinal pigment epithelial tears or rips; participants with diabetic retinopathy, severe cardiac disease, myocardial infarction within 6 months, ventricular tachyarrhythmia requiring ongoing treatment, unstable angina, peripheral vascular disease, stroke within 12 months, acute ocular or periocular infection, previous therapeutic radiation to the eye, head, or neck; treatment with any investigational agent within 60 days; allergies to fluorescein dye or to components of the pegaptanib formulation Equivalence of baseline characteristics: yes Diagnoses in participants: 306/1190 (26%) had predominantly classic CNV; 426/1190 (36%) had minimally classic CNV; and 458/1190 (38%) had occult with no classic CNV
Interventions	Intervention 1: 0.3 mg pegaptanib intravitreal injection every 6 weeks; at 54 weeks re‐randomization to continue or discontinue treatment Intervention 2: 1.0 mg pegaptanib intravitreal injection every 6 weeks; at 54 weeks re‐randomization to continue or discontinue treatment Intervention 3: 3.0 mg pegaptanib intravitreal injection every 6 weeks; at 54 weeks re‐randomization to continue or discontinue treatment Intervention 4: sham injection every 6 weeks; at 54 weeks re‐randomization to continue sham injections, discontinue sham injections, or receive treatment with 1 of 3 pegaptanib doses (0.3, 1.0, or 3.0 mg) Length of follow‐up Planned: 54 weeks after first randomization; 48 weeks after re‐randomization Actual: 54 weeks after first randomization; 48 weeks after re‐randomization; up to 4 years for safety outcomes
Outcomes	Primary outcome, as defined: proportion of participants losing fewer than 15 letters of VA between baseline and 54 weeks Secondary outcomes, as defined: proportion of participants maintaining or gaining ≤ 0, 5, 10, or 15 letters, or losing 30 or more letters; mean changes in VA at 6‐week intervals from baseline to week 54; proportion with VA 20/200 or worse at week 54; changes in size of lesion, size of CNV, and size of leakage at weeks 30 and 54 as measured by color fundus photography and fluorescein angiography Adverse events Intervals at which outcomes assessed: 6‐week intervals from baseline to week 54; 6‐week intervals from week 54 to week 102; color fundus photography and fluorescein angiography done at baseline, and at weeks 30, 54, 78, and 102
Notes	Full study name: VEGF Inhibition Study in Ocular Neovascularization Trial registration: NCT00021736; NCT00321997 Funding sources: Eyetech Pharmaceuticals, Inc., New York, and Pfizer Inc., New York, USA Declarations of interest: various authors reported that they had served as a paid consultant for Eyetech Pharmaceuticals and Neovista; had received royalty income from Coherent, the manufacturer of a laser used in photodynamic therapy; and/or were employees of and shareholders in Eyetech Pharmaceuticals Study period: not reported Reported subgroup analyses: none Contacting study investigators: trial authors contacted and contributed information for this review
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	"Patients were allocated in each trial to one of four treatment arms (sham or 0.3 mg, 1 mg, or 3 mg pegaptanib) by a dynamic procedure using a stochastic treatment allocation algorithm based on the variance method to minimize imbalances simultaneously for study center, angiographic lesion subtype and previous treatment with PDT"
Allocation concealment (selection bias)	Low risk	"The study coordinator randomized the patient by going on‐line to IDDI (an independent statistics/CRO) and answering eligibility and stratification questions. In response they were instructed which code on the treatment pack to use. Only when it was opened [sic] immediately prior to use would the injecting physician know whether it was active (but not which dose) or sham" (email communication with Eyetech, dated July 11, 2005)
Masking of participants (performance bias)	Low risk	"To maintain masking of the patients, the patients receiving sham injections and those receiving the study medication were treated identically, with the exception of scleral penetration. All patients (including those receiving sham injection) underwent an ocular antisepsis procedure and received injected subconjunctival anesthetic. The patients receiving sham injections had an identical syringe ‐ but without a needle ‐ pressed against the eye wall to mimic the active doses that were injected through the pars plana into the vitreous cavity. The injection technique precluded the patient from seeing the syringe"
Masking of study personnel (performance bias)	Low risk	"To maintain masking of the investigators, the study ophthalmologist responsible for patient care and for the assessments did not administer the injection"
Masking of outcome assessment (detection bias)	Low risk	"In all cases, a separate, certified visual‐acuity examiner masked to the treatment assignment and to previous measurements of visual acuity assessed distance visual acuity"
Incomplete outcome data (attrition bias) All outcomes	Low risk	"For all efficacy analyses, patients were evaluated in the treatment group to which they were randomly assigned. Several analyses of the primary efficacy endpoint that accounted for missing data were also conducted." At 54 weeks, 18 participants were excluded because they had not received at least 1 study treatment; 4 participants were excluded "because a sufficiently standardized assessment of visual acuity was not completed at baseline"; and missing data for about 10% of the study population were imputed using the last observation carried forward method
Selective reporting (reporting bias)	Low risk	Primary and secondary outcomes for week 54 (first year) were reported; visual acuity outcomes defined for the first year were not reported in second year outcomes
Other bias	Unclear risk	Sponsored by Eyetech Pharmaceuticals and Pfizer. Study chair and some others involved in the trials were paid consultants, employees, and/or shareholders of Eyetech Pharmaceuticals

Study acronyms: see Table 3.

AMD: age‐related macular degeneration.
 BCVA: best‐corrected visual acuity.
 CMT: central macular thickness.
 CNV: choroidal neovascularization.
 CRO: clinical research organization.
 CRT: central retinal thickness.
 ETDRS: Early Treatment Diabetic Retinopathy Study.
 FA: fluorescein angiogram.
 FDA: Food and Drug Administration.
 FMD: XXX.
 IOP: intraocular pressure.
 IVRS: interactive voice response system.
 IVTA: intravitreal triamcinolone acetonide.
 MI: myocardial infarction.
 NEI VFQ‐25: National Eye Institute‐Visual Functioning Questionnaire‐25.
 NHS: UK National Health Service.
 OCT: optical coherence tomography.
 PDT: photodynamic therapy.
 PRN: pro re nata.
 RAP: XXX.
 RCT: randomized controlled trial.
 RPE: retinal pigment epithelium.
 SD‐OCT: spectral domain optical coherence tomography.
 VA: visual acuity.
 VEGF: vascular endothelial growth factor.