Abstract

P 245

Comparison of different imaging modalities for staging of AMD in clinical trials

Tina Ristau1, Nils F. Mokwa1, Pearse A. Keane2, Srinivas R. Sadda2, Bernd Kirchhof1, Sandra Liakopoulos1
1Zentrum für Augenheilkunde, Universitätskliniken Köln, Köln, 2Doheny Retina Institute, Los Angeles, CA, United States

Objective
To compare color fundus photography (FP), fluorescein angiography (FA) and spectral domain OCT (SDOCT) for staging of age-related macular degeneration (AMD).
Methods
Color FPs, FAs, and SDOCT volumes (20°x15°) composed of 37 parallel B-scans (Spectralis, Heidelberg Engineering) from 29 eyes diagnosed with AMD and 21 controls were randomly collected. Control patients were required to show no signs for AMD, but other retinal pathology was allowed. Images were graded by one certified reading center grader. During image analysis, the grader was masked to all other images and grading results of the patient. For all images, the presence of large drusen (>125µm), atrophy of the retinal pigment epithelium (RPE), and any subretinal or subRPE tissue was noted. Additionally, signs of active choroidal neovascular membranes (CNV) such as hemorrhage, intra or subretinal fluid were recorded. Eyes were divided into 5 stages: 1-no AMD, 2-early AMD, 3-intermediate AMD, 4-geographic atrophy and 5-neovascular AMD. A feature was considered present, if it was visible on at least one image. The sensitivities of FP, FA and SDOCT images for the detection of the various features and stages were calculated.
Results
The neovascular form of AMD was diagnosed in 20 eyes and geographic atrophy in 1 eye. FP showed the highest sensitivity for detection of AMD (28/29=97%) and large drusen (10/14=71%), and the lowest sensitivity for detection of neovascular AMD (16/20=80%), any signs for active CNV (10/21=48%), RPE atrophy (7/13=54%) and any tissue (16/24=67%). FA showed the highest sensitivity for detection of the neovascular form of AMD (20/20=100%), RPE atrophy (7/13=77%), any tissue (24/24=100%) and any signs of active CNV (18/21=86%); and the lowest sensitivity for detection of large drusen (8/14=57%). SDOCT showed the highest sensitivity for detection of large drusen (10/14=71%), any tissue (24/24=100%) and any signs of active CNV (18/21=86%); and the lowest sensitivity for detection of AMD (26/29=90%) and RPE atrophy (7/13=54%). The 1 case with geographic atrophy was detected by all imaging modalities.
Conclusions
The presence of drusen and RPE changes as signs for AMD were best appreciated on FPs. FA and SDOCT images were more sensitive for the detection of the neovascular form of AMD and the signs of CNV activity. Further studies are needed to define the role of each of these imaging modalities for the staging of AMD in clinical trials.

Copyright: 2009 Porstmann Kongresse GmbH. All rights reserved. Impressum, Disclaimer.