Abstract
PURPOSE: To examine the change in size and location of pure geographic atrophy (GA). DESIGN: Population-based cohort study. METHODS: Ninety-five persons with GA either at baseline or at the one of the three five-year follow-up examinations, or both, were identified. Using computer-assisted software, the lesion area and greatest linear dimension (GLD) were calculated. Thirty-two persons (53 multiple eye-visit pairs) were seen at multiple visits five years apart with GA in the same eye to evaluate changes in total area and GLD. RESULTS: At the first occasion when pure GA was identified (n = 95), 45% had a single GA lesion, 18% had multifocal GA lesions, and 37% had a merged GA lesion. Of 53 eyes with multiple visits, the overall increase in atrophy was 6.4 mm(2) over a five-year period. The atrophy progressed to involve the foveal center in 47% of 19 eyes, and there was a mean decrease of 17 letters read correctly. Eyes with multifocal GA were most likely to have the area of atrophy increase (mean, 12 mm(2)), to have atrophy progress to the foveal center (83%), and to have a decrease in vision (mean, 22 letters), whereas eyes with a single GA lesion were least likely to have the area of atrophy increase (mean, 2 mm(2)), to have the lesion progress to the foveal center (22%), and to have a decrease in vision (mean, 10 letters). CONCLUSIONS: These are the first population-based data describing the five-year change in eyes with pure GA. Information on progression of GA will be useful for clinical trials of new interventions for GA.