Outer retinal band curvature - a novel quantitative biomarker for prognosticating late age-related macular degeneration (AMD)

Abstract

Purpose : Current methods of assessing outer retinal integrity in AMD are subjective, which can lead to inconsistent interpretation and limit its clinical application. Sinuosity is a novel, quantitative biomarker adapted from environmental science that objectively measures retinal pigment epithelium (RPE) curvature to capture alterations from pathological lesions. Sinuosity has been shown to screen for intermediate AMD with excellent performance. This study assessed whether it could predict conversion to late AMD.

Methods : Consecutive patients with intermediate AMD aged >50 years attending the Centre for Eye Health, Sydney, Australia on >1 occasion with Spectralis OCT (Heidelberg Engineering, Germany) volume scans of quality scores >15 were included. Bruch’s membrane (BM) and RPE length was measured using custom MATLAB code and sinuosity calculated as RPE:BM length (Fig 1). RPE integrity was assessed as disrupted or continuous in the central 1mm of B-scans. Drusen volume in the central 5mm was extracted using proprietary software. The prognostic ability of sinuosity, RPE integrity and drusen volume at baseline for predicting late AMD was calculated as odds ratios (ORs).

Results : In total, 17 converters and 43 non-converters with mean follow-up times of 43 and 53 months (p=0.25) were included. Sinuosity was greater for converters than non-converters (mean difference=0.039, p=0.003). RPE integrity was also significantly different between groups (χ²<0.01), but drusen volume was not (p=0.08). The adjusted odds of developing late AMD for RPE integrity, drusen volume>0.03mm³, sinuosity>1.05, sinuosity>1.075 and per 0.1 increase in sinuosity at baseline was 12.6 [3.0-54.0], 7.6 [2.0-29.4], 4.7 [1.4-15.6], 6.8[1.7-28.0] and 7.1 [1.6-30.9] respectively.

Conclusions : The odds of developing late AMD was ~7 times higher per 10% increase in RPE sinuosity, which was comparable to drusen volume thresholds currently used in clinical practice. RPE sinuosity may be advantageous to existing, subjective outer retinal biomarkers as it is quantitative, amenable to automation and generates measurements that are independent of physiological variations in retinal thickness.

This abstract was presented at the 2024 ARVO Annual Meeting, held in Seattle, WA, May 5-9, 2024.

 

Figure 1: (i) River sinuosity calculated as length of A -B (blue line) divided by C-D (red line). (ii) Foveal OCT B-scan of stable and (iii) converting iAMD eye at baseline with RPE (light blue) and BM (dark red) segmentation lines.