Longitudinal reproducibility of spectral domain optical coherence tomography in children with physiologic cupping and stable glaucoma.
PURPOSE: To determine whether Spectralis (Heidelberg, Germany) spectral domain optical coherence tomography (SD-OCT) measurements are reproducible over time in children with physiologic cupping and stable glaucoma. METHODS: Subjects were identified from a subset of participants in an earlier retrospective study conducted by our group and included children (<18 years of age) with physiologic cupping and stable primary congenital glaucoma (PCG) having had at least 2 SD-OCTs over a period of more than 1 between April 2010 and September 2015. Thicknesses of average peripapillary retinal nerve fiber layer (pRNFL) and six individual sectors and volumes of three segmented retinal layers and total retina were measured. Spectralis review software was used for segmentation. Intraclass correlation coefficients (ICC) and coefficient of variation (COV) were calculated. RESULTS: A total of 35 eyes of 35 children were included: 15 eyes had physiologic cupping; 20 eyes, PCG. Mean ages at initial SD-OCT were 11.2 ± 3.3 years and 9.7 ± 3.3, respectively; mean intervals between first and last imaging were 2.2 ± 1.1 and 3.0 ± 1.4 years, respectively. ICCs across three visits for both groups for average and sectoral pRNFL thicknesses were 0.887-0.997 and for segmented retinal volumes were 0.806-0.993. ICCs for total retinal volume for physiologic cupping and PCG were 0.993 and 0.954, respectively. COVs for average pRNFL thickness were 0.9% and 1.7%, respectively. For all other measurements, COVs ranged from 0.3% to 5.4%. CONCLUSIONS: Reproducibility of longitudinal SD-OCT measurements for average pRNFL thickness in children with stable glaucoma over about 2 years is comparable to short-term reproducibility (COV) in normal children (1.16%) and normal and glaucoma adults (1.62%-3.4%).
Xu, L; Freedman, SF; Silverstein, E; Muir, K; El-Dairi, M
Pubmed Central ID
Electronic International Standard Serial Number (EISSN)
Digital Object Identifier (DOI)