Objectives
To characterize the progression of eye fundus changes by multimodal imaging and to identify morphological and/or functional predictors of conversion in wet AMD in the fellow eyes of patients with exudative AMD.
Sample Description and Methods
Single-center, prospective, observational, longitudinal 2-year study. Sixty-two patients were enrolled in the study, with diagnosis of neovascular age-related macular degeneration in one eye and age-related maculopathy in the fellow eye. Each patient underwent a detailed ocular and medical history, a complete ophthalmologic examination with color fundus photography, fluorescein angiography, indocyanine green angiography and fundus autofluorescence imaging at baseline and repeated at six-month intervals. CFP images were graded resorting to the RetmarkerAMD (Critical Health SA) software, a computer assisted grading system based on the International Classification Grading System guidelines (Bird, Bressler et al., 1995).
Results
Fifty-three patients completed the two-year follow-up period, eighteen of which converted to the exudative form in the fellow eye during the study period. No measurable evolution was found in consecutive fundus autofluorescence and indocyanine green angiography images. Total drusen area in CFP images presented no significant increase between each visit and the subsequent (p=0.100), but significant increase was found between the first and the last (p=0.048). IGC images: 23% of patients presented a normal exam; observations have shown that the hot spots or areas and the hypofluorescent spots or areas where either attributable to drusens or drusenoid pigmentary epitelium dettachments or did not correspond to any lesion visible in the CFP images; the sensitivity of the overall exam result was 88% and the specificity of presence of early hot spots or areas was 97%. FAF
Abstract
Multimodal Image: Assessment of Risk Factors for the Development of Choroidal Neovascularisation in the Fellow Eye
images: the distribution of patterns for the cases that converted and those that did not is significantly different. CFP images: drusen area was significantly higher in the group of patients which did not convert when considering total area (p=0.012), sub-field 4 (p=0.020) and sub-field 5 (p=0.039); two main groups were identified by hierarchical cluster analysis, with Ward's linkage, when considering total area of drusens of the last visit available; principal component analysis identified total area as the most characterizing feature; no significant differences were found between the groups in what concerns the areas of drusen with specific sizes (C1: p=0.28; C2: p=0.11; C3: p=0.31); no correlation was found between age or total drusen area and conversion, and these two variables are not correlated.
Conclusion
The percentage of patients which converted during the study (34%) exceeds the expected rate for this time period. In what concerns evolution, further study including a longer follow-up period is suggested as it was only apparent for total drusen area calculated in CFP images between the first and last visits available. FAF images: FAF may be an interesting exam for predicting AMD progression but further studies are necessary to pinpoint the characterizing patterns. ICG images: the most characterizing features for conversion were the overall exam result and the presence of early hot spots or areas. CFP images: in this sample conversion does not occur for patients with the higher extension of lesions; no relation was found between conversion and patient’s age or drusen of specific sizes.