Progression dynamics of early versus advanced atrophic lesions in non-neovascular AMD using AI-based OCT analysis

Abstract

Purpose : Atrophic lesion growth rates can vary during the progression of non-neovascular age-related macular degeneration (AMD) which might have potential influence on the use of complement inhibition. The purpose of this study is to investigate the progression of atrophic lesions as early as they arise in order to set a standard for the optimal time of first intervention.

Methods : Patients which had recently converted to atrophic late AMD were included and prospectively imaged using optical coherence tomography (OCT). The respective total areas of external limiting membrane (ELM) disruption and retinal pigment epithelial (RPE) loss on longitudinal OCT volumes were quantified using artificial intelligence (AI) algorithms. Freshly converted eyes were compared to eyes with advanced disease status using mixed effect models.

Results : Eighty-two eyes from 53 patients were included. The first cohort consisted of 30 eyes from 25 patients with a recent conversion to complete retinal pigment epithelium and outer retinal atrophy (cRORA) within 6 months. The second cohort consisted of 52 eyes from 28 patients with a conversion to cRORA at least two years prior to the first study visit. The study period encompassed two years with scheduled visits every 6 months. Mean (95%CI) square root progression of recent atrophy was 123.92 µm [76.88 – 170.86] and 126.08 µm [80.18 – 171.94] per year for ELM disruption and RPE loss, respectively. Mean square root progression of advanced lesions was 220.2 µm [187.46 – 252,68] and 219.2 µm [187 – 251.06] per year for ELM disruption and RPE loss, respectively. A significant difference regarding progression of ELM disruption (p=0.001) and RPE loss (p=0.001) between converted and advanced lesions was found.

Conclusions : Data on early disease progression in atrophic AMD is rare. Atrophic lesions from recently converted patients have a slower growth rate in comparison to patients in an advanced disease stage. These findings might play a crucial role in planning future treatments using complement inhibition. Individual disease monitoring in early AMD supported by AI-quantifications is necessary to capture the ideal point of care in non-neovascular AMD.

This abstract was presented at the 2022 ARVO Annual Meeting, held in Denver, CO, May 1-4, 2022, and virtually.

 

Figure 1. Progression of atrophic lesion growth (Red: Advanced lesions, baseline at least 2 years after conversion; Blue: Recent lesions, baseline after conversion). A) Growth of ELM disruption B) Growth of RPE loss