Amendment 1: The study design was updated to reflect changes in timing of selected procedures, additional data collection in the nonstudy eye (3-field confocal scanning laser ophthalmoscopy (cSLO) exams), and the addition of blood samples for modelling of disease progression by gene expression analysis. In addition, roflumilast was added as a prohibited medication and clarifications were made throughout the protocol.

Amendment 2: Updates were made to the inclusion and exclusion criteria (specifically the addition of inclusion criterion #6 requiring the presence of drusen as assessed with fundus photography and/or spectral-domain optical coherence tomography (SD-OCT) at Screening and the modification of exclusion criterion #14 to exclude participants with a history of glaucoma who might had had visual field deficits). The study design was updated to reflect the addition of corneal curvature assessment, the removal of blood draw for pharmacokinetic (PK) at Months 9 and 24, and the requirement that complete ophthalmic examinations include: external examination of the eye and adnexa, screening for eyelid/pupil responsiveness, slit-lamp biomicroscopy, indirect ophthalmoscopy, dilated fundus examination, and intraocular pressure (IOP) assessment. The following other efficacy measure was added: “GA lesion area growth in the fellow eye, as assessed with FAF and quantified by the central reading centre (CRC),” and scotopic microperimetry was changed to scotopic/mesopic microperimetry.

Amendment 3: The number of sites was revised from approximately 15 to 20 to approximately 15 to 30 and inclusion criteria were updated to reflect looser restrictions for participant eligibility in order to facilitate participant recruitment. In addition, visit windows for corneal curvature assessments were adjusted for participant convenience, and other minor clarifications were made.

Amendment 4: The maximum lesion size was increased from 12.5 mm^2 to 18mm^2. The visit window for Screening Visit 2 was changed from -19 to -2 days to -20 to -2 days for participant and investigator convenience. In addition, several corrections and clarifications were made.

Amendment 5: Several revisions were made to reflect that microperimetry would only be conducted at selected sites in participants who qualified and consented to the procedure in order to ease participant recruitment; loss of retinal sensitivity was removed from the clinical hypotheses and retinal sensitivity was changed to the other efficacy endpoint, rather than a secondary efficacy endpoint. In addition, a possible interim analysis of the first 50% of participants that completed Month 18 was added in order to facilitate earlier project planning.

Amendment 6: To facilitate participant recruitment, the number of sites was increased from approximately 30 to approximately 40 and the inclusion criterion requiring that the study eye have a BCVA better than or equal to 55 letters (20/80 Snellen equivalent) was changed to require a BCVA better than or equal to 45 letters (20/125 Snellen equivalent). A second interim analysis was planned to be carried out when 100% of participants had completed Month 18. Updates were also made to reflect that standard and low luminance BCVA would be assessed in both eyes, rather than the study eye only, at Baseline and Months 12 and 24, and that these BCVA assessments would be done prior to pupil dilation, as pupil dilation temporarily reduces visual acuity. In addition, cSLO quantitative fundus autofluorescence (Qfaf) (central field) was changed to be an assessment performed in participants who participated in the microperimetry procedure only, and in the study eye only, rather than both eyes.

Amendment 7: Updates were made to reflect the following changes to efficacy measures and analyses: area of reticular drusen was changed to presence or absence of reticular drusen (given difficulty of measuring the area), the primary efficacy analysis method was changed from an analysis of covariance to an MMRM in order to attain greater statistical power, near-IR measurement was changed from 7-field to 3-field, genotyping was changed from analysis of specific changes to genome-wide variations, use of a Goldmann applanation tonometer for IOP measurements was changed to recommended instead of required, last observation carried forward analyses were removed, per-protocol population was removed, and it was specified that analyses performed on the mITT population would be based on the randomised treatment. In addition, the interim analysis that was planned to be carried out when 100% of participants completed Month 18 was dropped from the planned analyses. Several clarifications and corrections were also made throughout the protocol.