Purchase this article with an account.
Xiangrong Kong, Alexander Ho, Beatriz Munoz, Sheila K West, Rupert W Strauss, Anamika Jha, Ann Margret Ervin, Janet Cheetham, Michael S Ip, Hendrik P Scholl; Using retinal structural measurements derived from optical coherence tomography as clinical endpoints for Stargardt disease - reproducibility assessment and lessons learned from the ProgStar study. Invest. Ophthalmol. Vis. Sci. 2019;60(9):3945. doi: https://doi.org/.
Download citation file:
© ARVO (1962-2015); The Authors (2016-present)
To assess the reproducibility of retinal structural measurements from optical coherence tomography (OCT) in ABCA4-related Stargardt disease (STGD1) to inform choices of outcome measures for future intervention trials.
The Progression of atrophy secondary to Stargardt disease (ProgStar) study enrolled 259 STGD1 patients from 9 sites in the USA and Europe. Images were acquired using Heidelberg Spectralis OCT. The study reading center (RC) used semiautomatic segmentation with manual adjustments to detect the retinal layers of the retinal pigment epithelium (RPE), outer segments, inner segments (IS), outer nuclear layer (ONL), inner retina, and total retina. The segmented images were overlaid to an Early Treatment of Diabetic Retinopathy Study (ETDRS) grid. For each layer, the thickness and intact area of the ETDRS central subfield, inner ring and outer ring were recorded, respectively. Reproducibility of measurements was assessed by regrading 30 independent images by a different set of RC graders, using Bland-Altman plots and statistics including intraclass correlation (ICC) and relative absolute difference (RAD).
The median age of the 30 subjects was 32 years (47% female, 13% non-white). Median best corrected visual acuity was 0.78 LogMAR. Measurements of the ETDRS central subfield had low ICC and/or large RAD in all layers. The outer ring region was not fully captured in some images (Fig.1). The inner ring had good reproducibility for the intact area in the IS (ICC=0.99, RAD=4%), thicknesses of the ONL (ICC=0.93, RAD=6%) and the total retina (ICC=0.99, RAD=1%). Qualitative reviews of images indicated that relatively large grading discrepancy resulted from confounding pathologies such as intraretinal cystoid spaces, outer retinal tubulations, hyperreflective foci, and RPE atrophy (Fig.2).
The complex morphology in STGD1 makes outer retina segmentation challenging. Inner ring measurements including the intact area of IS (i.e. a measure of the integrity of the ellipsoid zone), the ONL and total retinal thickness had good reproducibility and may provide reliable measures to track disease progression.
This abstract was presented at the 2019 ARVO Annual Meeting, held in Vancouver, Canada, April 28 - May 2, 2019.
An example of incomplete scan coverage of the ETDRS outer ring.
An example of an ambiguous case where determination of preserved RPE vs. debris was different between gradings.
This PDF is available to Subscribers Only