Purpose: To identify valid and reproducible methods for quantifying anatomic outcome measures for eyes with choroideremia (CHM) in clinical trials.
Design: Reliability analysis study.
Methods: In this multicenter study, patients with confirmed genetic diagnosis of CHM were enrolled. All cases underwent spectral-domain optical coherence tomography (SDOCT) and fundus autofluorescence (FAF) imaging. Two graders independently delineated boundaries of preserved autofluorescence (PAF) and preserved ellipsoid zone (EZ) on FAF and OCT images, respectively. The results of the 2 independent gradings of both FAF and OCT images were compared to assess the reproducibility of the grading methods.
Results: A total of 148 eyes from 75 cases were included. In 21% of eyes PAF and in 43% of eyes preserved EZ had extended beyond the image capture area. After exclusion of these eyes and low-quality images, 114 FAF and 77 OCT images were graded. The mean PAF areas from 2 independent gradings were 3.720 ± 3.340 mm2 and 3.692 ± 3.253 mm2, respectively. Intraclass correlation coefficient (ICC) for these gradings was 0.996. The mean preserved EZ areas from 2 independent gradings were 2.746 ± 2.319 mm2 and 2.858 ± 2.446 mm2, respectively. ICC for these gradings was 0.991.
Conclusions: Quantifying preserved retinal pigment epithelium and EZ areas on FAF and OCT images, respectively, in CHM patients is highly reproducible. These variables would be potential anatomic outcome measures for CHM clinical trials and could be studied and tracked longitudinally in choroideremia.
Copyright © 2017 Elsevier Inc. All rights reserved.