Abstract
Purpose :
Age related macular degeneration (AMD) is a prevalent chronic ocular disease without reliable and accurate prognostic markers. Possible prognostic indicators of progression from intermediate AMD (iAMD) to either choroidal neovascularisation (CNV;“wet”), geographic atrophy (GA;“dry”) or remaining as iAMD are currently unknown and were investigated.
Methods :
Spectral Domain Optical Coherence Tomography (SD-OCT) images were selected from an electronic register of patients who attended AMD clinics from 2009-2014 at Moorfields Eye Hospital who were known to be on treatment for CNV in the fellow eye. 1671 individuals were assessed for eligibility from prior work studying iAMD and CNV development.1423 were excluded. A proportion developed GA. 248 patients developed either CNV or remained as iAMD. A random selection was then included across all groups resulting in those remaining as iAMD (n=23), progressed to CNV (n=21) or developed GA (n=16). Subjects with macular pathology other than drusen were excluded, resulting in n=60. Topcon SD-OCT images were segmented and all automated segmentation errors were manually corrected. Volume thickness measurements were then correlated with the different patient groups
Results :
There was a significant difference between the iAMD and CNV groups outer nuclear layer volumes (0.22mm3±0.04mm3 vs 0.179mm3± 0.05mm3 [p=0.0004]). This was found to be higher in GA subjects relative to CNV subjects (0.197mm3± 0.05mm3 vs 0.179mm3±0.05mm3 [p=0.34]). Between those that remained as iAMD the CNV groups the photoreceptor layer volume was lower in the CNV group (0.078mm3±0.013mm3 vs 0.077mm3±0.031mm3 [p=0.04]). The difference between the CNV and GA groups retinal pigment epithelium volume was trending towards statistical significance (0.063mm3±0.015mm3 vs 0.059mm3±0.012mm3 [p=0.06]). Drusen count was higher in CNV group versus the iAMD group (16.33±10.15 vs 10.17±8.95 [p=0.02]), as was the mean count in the CNV versus GA group (16.33±10.15 vs 9.30±9.77 [p=0.02]). Mean drusen volume of iAMD versus CNV groups (0.13mm3±0.14mm3 vs 0.23mm3±0.21mm3 [p=0.03]) and CNV versus GA groups (0.23mm3±0.21mm3 vs 0.13mm3±0.18mm3 [p=0.02]) were significant.
Conclusions :
Our pilot study has demonstrated interesting trends, which may allow better patient counselling and allow for more accurate future trial design.
This is an abstract that was submitted for the 2018 ARVO Annual Meeting, held in Honolulu, Hawaii, April 29 - May 3, 2018.