Abstract
Purpose: :
To evaluate the significance of autofluorescence patterns in determining risk status in early age–related macular degeneration (AMD).
Methods: :
117 eyes from 117 consecutive patients recruited to the UK–MRC AMD collection at Moorfields Eye Hospital were used. Patients had either drusen in both eyes (D/D), the fellow eyes of patients with unilateral exudative AMD (D/CNV) and the fellow eyes of patients with unilateral geographic atrophy (D/GA). Two graders then graded the colour images according to the Rotterdam grading system and the autofluorescence (AF) images using the proposed international AF classification system by Bindewald et al. Colour and AF images were evaluated to determine any predominance in the high risk fellow eye (F/E) subgroup relative to the D/D group.
Results: :
A high proportion of eyes in the F/E groups were found to have minimal changes on colour photography. When the distribution of the AF patterns was analysed, category ‘speckled pattern’ was found to be more prevalent in the D/D subgroup (60% D/D, 20% D/CNV and 20% D/GA subgroups), ‘focal increased pattern’ was more prevalent in the D/CNV subgroup (33% D/D, 61% D/CNV and 5.6% D/GA subgroups) and ‘reticular’ was more prevalent when both F/E subgroups were combined (21% D/D, 79% FE). Statistical significance was limited by the sample size. The colour and AF images for each patient were then compared to identify any relationship between the Bindewald and Rotterdam grading systems. AF images graded as ‘minimal change’ decreased linearly across the Rotterdam grades of increasing severity (OR=0.70; 95% ci 0.6 to 0.8 P=0.001) and AF images graded as ‘focal increase’ and ‘speckled’ increased along the Rotterdam grades of increasing severity, although they did not reach statistical significance (OR=1.3 95% ci. 0.99 to 1.8; P=0.05 and OR=1.3 95% ci 0.96 to 1.9; P=0.13 respectively). There was a significant increase in ‘reticular pattern’ across the Rotterdam grades of increasing severity (OR=1.4 95% ci 1.10 to 1.3 P=0.02.)
Conclusions: :
The presence of a large number of eyes with no or minimal changes on colour imaging in the high risk F/E subgroup suggests that colour images may be misleading in predicting AMD risk. 'Reticular’ changes on AF images may be more useful in signifying a higher risk. AF imaging is a potentially useful in identifying risk status in AMD when used in conjunction with colour images. These results need to be validated with a larger series if available.
Keywords: age-related macular degeneration • choroid: neovascularization • retina