To estimate the incidence, size, and growth rate of geographic atrophy (GA) during 5 years of follow-up among participants in the Comparison of Age-Related Macular Degeneration Treatments Trials (CATT).

Cohort within a clinical trial.

Participants included in CATT.

A total of 1185 CATT participants were randomly assigned to ranibizumab or bevacizumab treatment and to 3 treatment regimens. Participants were released from protocol treatment at 2 years and examined at approximately 5 years (N = 647). Two masked graders assessed the presence and size of GA in digital color photographs (CPs) and fluorescein angiograms (FAs) taken at baseline and years 1, 2, and 5. Cox proportional hazard models were used to identify risk factors for incidence of GA. Annual change in the square root of the total area of GA was the measure of growth. Multivariate linear mixed models including baseline demographic, treatment, and ocular characteristics on CP/FA and optical coherence tomography (OCT) as candidate risk factors were used to estimate adjusted growth rates, standard errors (SEs), and 95% confidence intervals (CIs).

Geographic atrophy incidence and growth rate.

Among the 1011 participants who did not have GA at baseline and had follow-up images gradable for GA, the cumulative incidence was 12% at 1 year, 17% at 2 years, and 38% at 5 years. At baseline, older age, hypercholesterolemia, worse visual acuity, larger choroidal neovascularization (CNV) area, retinal angiomatous proliferation (RAP) lesion, GA in the fellow eye, and intraretinal fluid were associated with a higher risk of incident GA. Thicker subretinal tissue complex and presence of subretinal fluid were associated with less GA development. The overall GA growth rate was 0.33 mm/year (SE, 0.02 mm/year). Eyes treated with ranibizumab in the first 2 years of the clinical trial had a higher growth rate than eyes treated with bevacizumab (adjusted growth rate, 0.38 vs. 0.28 mm/year; P = 0.009). Geographic atrophy in the fellow eye, hemorrhage, and absence of sub–retinal pigment epithelium fluid at baseline were associated with a higher growth rate.

Development of GA is common 5 years after initiating therapy. Several risk factors identified at 2 years of follow-up persist at 5 years of follow-up.