To compare the ability of spectral domain optical coherence tomography (SD-OCT), blue light fundus autofluorescence (FAF), and near-infrared fundus autofluorescence (NIR-FAF) to evaluate foveal involvement in geographic atrophy as a result of age-related macular degeneration.Methods:
All consecutive patients with geographic atrophy underwent FAF (excitation λ = 488 nm; emission λ > 500 nm), NIR-FAF (excitation λ = 787 nm; emission λ > 800 nm), and simultaneous SD-OCT scanning (Spectralis HRA + OCT; Heidelberg Engineering). Two readers independently graded foveal involvement on FAF, NIR-FAF, and SD-OCT and measured the width of foveal sparing. In eyes with an intergrader agreement of foveal sparing by at least one among FAF, NIR-FAF, and SD-OCT, microperimetry (Spectral OCT/SLO; OPKO-OTI) was analyzed.Results:
A total of 158 eyes (83 patients; 53 women, 30 men, mean age 69.2 ± 4.8 years) with geographic atrophy were included. Spectral domain OCT showed the highest intergrader agreement of foveal involvement (k = k′ = 0.8, P = 0.001 vs. k = k′ = 0.7, P = 0.01 for NIR-FAF and k = k′ = 0.5, P = 0.01 for FAF). In 74 eyes (46.8%) foveal sparing was present according to interobserver agreement. Width of the foveal sparing was larger on SD-OCT than on NIR-FAF and FAF (1,334 ± 943 μm vs. 1,228 ± 912 μm, P < 0.001 and 1,201 ± 922 μm, P < 0.001, respectively). Retinal fixation was predominantly central and stable in 97.3% of eyes with foveal sparing.Conclusion:
Spectral domain OCT is an appropriate imaging modality for evaluating the presence and extent of foveal sparing, followed by NIR-FAF and FAF.