Table 1 Characteristics of included studies.

Study	Intervention	Participants
VISION study year 15	(1) 0.3 mg pegaptanib (n = 297)†	Target population:
VISION year 26,7Two concurrent multicentre RCTs (1003 and 1004)Primary outcome: proportion losing <15 letters at week 54.Length of follow‐up: 54 weeks, plus 48 weeks after re‐randomisation.MARINA8Multicentre RCTPrimary outcomes: proportion losing <15 letters at 12 months; safety and tolerabilityLength of follow‐up:24 monthsANCHOR9Multicentre RCTPrimary outcome: proportion losing <15 lettersLength of follow‐up: 24 months (ongoing)FOCUS10Multicentre RCT (phase I/II)Primary outcome: proportion losing <15 lettersLength of follow‐up: 24 months (ongoing)	(2) Sham injection (n = 304)Frequency: injections every 6 weeks, total of 9 treatments.Patients re‐randomised after 54 weeks†Year 1–Year 20.3 mg: 0.3 mg n = 133 discontinue n = 132Sham: 0.3 mg n = 53 sham n = 53 discontinue n = 54(1) 0.5 mg ranibizumab (n = 240)†(2) Sham injection (n = 238)Frequency: monthly injections(1) 0.5 mg ranibizumab + sham PDT (n = 140)†(2) Sham injection + verteporfin PDT (n = 143)Frequency: injections given monthly, PDT every 3 months if needed(1) 0.5 mg ranibizumab‡ + PDT (n = 106)(2) Sham injection + verteporfin PDT (n = 56)Frequency: injections given every 30 days, starting on day 7; verteporfin PDT 7 days prior to initial study drug administrationSubjects were allowed to receive further verteporfin PDT if deemed necessary by the investigator at any of the evaluation visits.	All angiographic subtypes of lesionsAngiographic subtype of lesion at baselinePC: (1) 24%, (2) 26%MC: (1) 38%, (2) 34%ONC: (1) 38%, (2) 40%Angiographic subtype of lesion for re‐randomised groups in year 2 not reportedTarget population:Occult CNV or minimally classic CNVAngiographic subtype of lesion at baselinePC: (1) 0%, (2) 0%*MC: (1) 37.9%, (2) 36.6%*ONC: (1) 62.1%, (2) 63.4%*Missing data: (1) 0, (2) 0.4%Target population:Predominantly classic lesionsAngiographic subtype of lesion at baselinePC: (1) 96.4%, (2) 98.6%MC: (1) 3.6%, (2) 1.4%ONC: (1) 0, (2) 0Target population:Predominantly classic lesionsAngiographic subtype of lesion at baselinePC: (1) 65.7%, (2) 66.1%,MC: (1) 30.5%, (2) 26.8%ONC: (1) 1.9%, (2) 7.1%Unclassified: (1) 1.9%, (2) 0%

PC, predominantly classic (⩾50% classic); MC, minimally classic (<50% classic); ONC, occult with no classic.

*Patients in sham injection group total n = 239 rather than 238 in original paper⁸.

†See full papers for data on other dose groups.

‡A lyophilised formulation of ranibizumab was used for the first 12 months.