Abstract
Purpose: :
To determine the frequency and causes of screening failures based on fundus autofluorescence (FAF) images in a multicenter, international, randomized clinical controlled trial (GATE-Study) for geographic atrophy (GA) in late stage AMD.
Methods: :
For eligibility screening, confocal scanning laser ophthalmoscopy (cSLO) images (Heidelberg Retina Angiograph/Spectralis, Heidelberg Engineering, Germany) were recorded and submitted by clinical sites according to study protocol. Eligibility was determined by a central reading center from central near-infrared- (NIR, 810nm) and blue reflectance (BR, 488nm), and 3-field fundus autofluorescence (FAF, exc: 488nm, em: 500-700nm) images. Eligibility required: 1. high quality images 2. GA secondary to AMD 3. a well-demarcated lesion 4. GA lesion total atrophy size <20mm² (ca. 8 disc areas [DA]) and one single lesion of at least 1.25mm² (0.5 DA), quantified by image analysis software (RegionFinder, Heidelberg Engineering, Germany) 5. increased FAF adjacent to GA.
Results: :
From Apr. to Nov. 2009, screening failures occurred in 270 of 1093 (24.7%) subjects from 51 sites (mean submission: 21.4 subjects per site, range 1-54). Ineligible lesion size was the most common screening failure (no single lesion >1.25mm² in 77 eyes (28.5%), lesion size exceeding >20mm² in 36 eyes (13.3%)), followed by insufficient image quality in 57 eyes (21.1%), no well-demarcated lesion in 47 eyes, reflecting concurrent fibrosis/CNV (17.4%) and GA not secondary to AMD (e.g. macular dystrophies) in 42 eyes (15.6%). Eleven (4.1%) were excluded without increased FAF adjacent to GA. Screening failures of 24 sites with more than 20 submissions varied from 3.7 to 61.5%.
Conclusions: :
Screening for GA eligibility based on lesion criteria as determined by cSLO FAF imaging is possible in large-scale interventional clinical trials. With this screening, a study population can be identified that has accurately quantified areas of GA, abnormal FAF signals in the perilesional area, an absence of poorly defined margins indicative of subretinal fibrosis and/or regressed CNV, and has GA due to late-stage dry AMD as differentiated from other causes e.g. hereditary macular dystrophies. Accurate identification of a defined study population will facilitate comparisons between control and treatment groups in interventional GA trials.
Clinical Trial: :
www.clinicaltrials.gov NCT00890097
Keywords: age-related macular degeneration • imaging methods (CT, FA, ICG, MRI, OCT, RTA, SLO, ultrasound) • imaging/image analysis: clinical