Factors impacting the assessment of geographic atrophy on ultra‑widefield fundus autofluorescence imaging

Purpose To determine how often geographic atrophy (GA) lesions cannot be reliably delineated on ultrawidefield (UWF) green-light fundus autofluorescence (GAF) and to identify factors associated with poor lesion demarcation. Methods Retrospective cross-sectional study. 88 patients (138 eyes) with GA due to non-exudative age-related macular degeneration (AMD). Consecutive retina clinic subjects with GA secondary to AMD and UWF-GAF, UWF pseudocolor, and OCT imaging at the same visit were included in this study. Eyes with AMD due to other causes, evidence of exudative AMD, or images with overall poor quality were excluded. Two independent masked graders classified the UWF-GAF images as gradable (clearly delineated GA borders amenable to measurement) or ungradable (indistinct borders), and discrepancies were adjudicated by the senior investigator. Choroidal thickness (CT) was measured on central horizontal B-scan at the foveal center and at points 1.5 mm nasal and temporal to the center. Main Outcome Measures: Frequency of ungradable UWF-GAF images and choroidal thickness measurements at the foveal, nasal, and temporal locations. Results UWF-GAF images were deemed ungradable for GA size in 7/138 eyes (5.1%; 95% CI, 2.5–10.1). Inter-reader agreement for gradability was excellent (κ = 0.867; p < 0.001). Ungradable eyes demonstrated a markedly thinner choroid at all measurement locations including the foveal center (74.71 ± 37.04 μm vs. 177.41 ± 91.16 μm; p = 0.0002), nasal (58.14 ± 14.73 μm vs. 132.70 ± 88.28 μm ; p = 0.0013), and temporal (85.29 ± 34.54 μm vs. 177.40 ± 87.07 μm; p = 0.018) points. Conclusion In retina practices that routinely obtain Optos UWF-GAF imaging, approximately 5% of GA lesions may not be measurable on UWF GAF imaging, with a very thin choroid being a common feature among such cases. Green fluorescence from the unmasked sclera is the likely explanation for this phenomenon, though such GA lesions lend themselves to reliable quantification using the companion UWF pseudocolor image. These findings have implications for the monitoring of GA lesions in clinical practice in the era of GA therapeutics.