Abstract
Purpose :
Geographic atrophy (RORA) is a debilitating manifestation of advanced age-related macular degeneration (AMD), defined by the degeneration of retinal pigment epithelium (RPE) and photoreceptors (PR). Optical coherence tomography (OCT) has become an invaluable tool for imaging retinal structures. We present a comprehensive analysis of retinal layer thickness surrounding RORA lesions using OCT A-scans.
Methods :
OCT scans (N=252) from 36 eyes from 18 participants were sampled from the LEAD trial. All sampled participants developed complete RPE and outer retinal atrophy (RORA) lesions. The MZ, EZ+OPR+IZ, RPE, and CC+CS were automatically segmented and the thickness was measured in staggered distance bands from the lesion edge and drusen material edge (0mm to 1mm distance at 0.1mm intervals), excluding lesion areas for drusen and vice-versa. 500,000 A-scans were sampled at random from the cohort for each analysis. Linear mixed-effects models were used to compare the layer thickness to distance from the lesion edge, with patient ID used as a random effect, controlled from distance from fovea.
Results :
The four measured biomarkers behave differently in function of the distance to the atrophic lesion and drusen material edges. MZ and EZ+OPR+IZ increase in thickness further from the RORA lesion edge, showing a steeper decline closer to the lesion edge (all bins p<0.01). Whilst distance to the drusen material edge show similar characteristics close to the legion edge for EZ+OPR+IZ, MZ thickness is more constant (p>0.01).
RPE is the thinnest close to the lesion edge but is the thickest at around 0.2mm from the lesion, after which there is a decline (all bins p<0.01). From the drusen material edge, RPE is thinnest the furthest from the lesion (all bins, p<0.01). CC+CS was the thickest furthest from the lesion edge, while it was the thinnest furthest from the drusen material edge (all bins, p<0.01).
Conclusions :
Our approach allowed the enface examination of layer thickness alterations in response to the proximity to the RORA lesion. We showed a distinct spatial distribution of PR, RPE, and CC+CS thickness loss outside the lesion boundaries. This behavior might be related to further RORA progression or lesion spreading. This study contributes to the growing knowledge concerning RORA and establishes a quantitative framework for characterizing structural changes surrounding lesions.
This abstract was presented at the 2024 ARVO Annual Meeting, held in Seattle, WA, May 5-9, 2024.