Assessing Reliability and Agreement in Topographic Measurement of Reticular Pseudodrusen Area: A Multimodal Imaging Approach

Purpose: This study aims to assess the reliability and agreement of fundus autofluorescence (FAF), near-infrared reflectance (NIR), and a combination of NIR and dense structural optical coherence tomography (OCT) scans (OCT+NIR) in delineating the areas of reticular pseudodrusen (RPD) in eyes affected by age-related macular degeneration (AMD). Methods: This was a single-center, cross-sectional study. Patients with non-advanced AMD exhibiting signs of RPD on multimodal imaging were enrolled. Two independent masked graders manually delineated the margins of the area occupied by RPD using the three distinct imaging techniques. Results: The study included 60 eyes from 51 patients, with a mean age of 81.5 (±7.1) years. The intraclass correlation coefficient between the two graders across all imaging modalities was 0.96 for FAF, 0.92 for NIR, and 0.98 for OCT+NIR. The narrowest limits of agreement were observed with OCT+NIR (-4.38 +5.17 mm2). Foveal involvement and age were significantly correlated with larger RPD area (p=0.036 and p=0.019 respectively). Pairwise comparisons of square root-transformed RPD areas indicated that FAF detected a significantly larger RPD area compared to other methods (p<0.001). Conclusions: These findings validate OCT+NIR as a reliable approach for measuring RPD areas, potentially serving as a critical biomarker for AMD in future clinical trials.