Abstract
Purpose: :
To determine from morphological data the risk of developing a choroidal neovascular membrane in an eye with dry AMD.
Methods: :
We used data derived from 946 eyes of 522 subjects with dry AMD who were subjects in two large dry AMD clinical trials, the Age-Related Eye Disease Study (AREDS) and Prophylactic Treatment of Age-Related Macular Degeneration Study (PTAMD). Fundus photographs from all eyes were digitized and read by trained readers using a drusen analysis algorithm, where the presence or absence of pigment abnormalities in geographic atrophy was also noted. We assessed risk based on these variables: ETDRS visual acuity, presence or absence of pigment abnormalities or geographic atrophy, number of small, medium, and large drusen, total drusen area, and whether or not the fellow eye had wet AMD (previously affected). We assessed risk using generalized estimating equations (GEE) logistic regression and survival analysis with frailty.
Results: :
If a patient had any pigment, the estimated odds for an event increase by a factor of about 2.46 (p=0.027, 95% CI: 1.10 to 5.47. The estimated odds for an event increased by a factor of about 7.13 (p<0.0001, 95% CI: 3.45 to 14.73) if the fellow eye was affected. A one standard deviation increase in the total drusen area of 0.39 mm2 increased the odds for an event by about a factor of about 1.30 (p=0.043, 95% CI: 1.01 to 1.68). For each one letter increase in baseline ETDRS visual acuity the odds for an event decreased by about 2.7% or by a factor of about 0.973 (p = 0.015, CI: 0.951 to 0.995). Age was not statistically significant with an odds ratio of 0.96 (p = 0.103, CI: 0.92 to 1.01).
Conclusions: :
The risk of developing wet AMD in an eye is greatly influenced by the presence of a previous event in the fellow eye and the presence of pigment in the given eye. Decease in visual acuity and increase in drusen area also increase the probability of a CNV event.
Keywords: drusen • choroid: neovascularization • visual acuity