Incidence and Progression of Geographic Atrophy Observations from a Population-based Cohort

Purpose: To examine early age-related macular degeneration (AMD) lesion characteristics and risk factors associated with the long-term development and progression of geographic atrophy (GA). Design: Population-based cohort. Participants: Of 3654 participants aged >= 49 years in the Blue Mountains Eye Study, 75.8%, 76.7%, and 56.1% of survivors attended the 5-, 10-, and 15-year follow-up examinations, respectively. Methods: Retinal photographs were taken at each visit. Incident GA was confirmed using a side-by-side grading method. Computer planimetry was used to measure the area involved by GA. Fast and slow/normal progression rates were defined as GA area enlargement by >= 2 and <2 mm(2)/year, respectively. Incident GA was estimated using the Kaplan-Meier product-limit method. Early AMD lesion characteristics were assessed for association with GA incidence using eye-specific data and generalized estimating equation models adjusting for age, current smoking, and presence of risk alleles of the complement factor H (CFH) or age-related maculopathy susceptibility 2 (ARMS2) genes, genotyped or imputed using genome-wide scan data. Main Outcome Measures: Incidence and progression of GA. Results: By excluding 41 subjects with GA at baseline, of 2503 participants at risk of GA, incident pure GA (without coexisting neovascular AMD lesions) was confirmed in 57 participants, with a 15-year incidence of 3.6%. Baseline early AMD lesion characteristics associated with GA incidence included drusen type (soft indistinct: odds ratio [OR], 59.0; 95% confidence interval [CI], 20.4-171.0; reticular drusen: OR, 13.9; 95% CI, 4.0-47.6); drusen location within a 500-mm radius of the fovea (OR, 15.1; 95% CI, 7.4-30.8); drusen area greater than 375 mm in diameter (OR, 10.1; 95% CI, 4.0-25.6); presence of retinal pigment epithelial depigmentation (OR, 9.0; 95% CI, 4.1-19.8); or hyperpigmentation (OR, 12.0; 95% CI, 6.1-23.5), referenced to subjects with no or hard drusen only. Fast progression was more frequent among current smokers at baseline, subjects with the CFH or ARMS2 risk genotypes, and pseudophakic eyes. Conclusions: Early AMD lesion characteristics (type, location, area involved) were strongly associated with higher long-term risk of developing GA independent of age, smoking, and AMD genetic susceptibility from the CFH or ARMS2 genes. Known AMD risk factors also were more frequently present among quickly progressing GA cases.