Purchase this article with an account.
Y. Ouyang, S. R. Sadda, A. C. Walsh; The Axial Location of Cystoid Spaces in Retinal Vein Occlusion Using 3d-oct. Invest. Ophthalmol. Vis. Sci. 2010;51(13):3560.
Download citation file:
© ARVO (1962-2015); The Authors (2016-present)
Histopathologic studies have demonstrated cystoid macular edema (CME) frequently involves the outer plexiform layer(OPL). Using new information about the true boundary between the outer nuclear layer (ONL) and the OPL, the purpose of this study was to analyze the frequency distribution of cystoid changes in each retinal layer using 3D-OCT.
Retina clinic patients diagnosed with RVO who underwent 3D-OCT imaging (3D-OCT-1000, Topcon Corporation) between May 2006 and Nov. 2009 were retrospectively reviewed. Eyes with previous treatment or other retinal diseases which could potentially cause CME, including diabetic retinopathy, age-related macular degeneration, and hypertensive retinopathy, were excluded. The earliest 3D-OCT scan (512 Ascans x 128 Bscans) deemed to be high enough quality for analysis from each eye was included in this cohort. Two graders evaluated each set of images independently. CME in each retinal layer was graded as present, questionable, absent or can't grade based on the presence of hyporeflective, intraretinal, cystoid spaces on 3D-OCT.
Out of 458 total visits for 81 patients, 39 visits from 39 eyes of 37 patients met the inclusion criteria for this study (19 BRVO, 12 CRVO, 8 HRVO). Definite CME was present in the ONL in 23%, OPL in 67%, inner nuclear layer (INL) in 67%, inner plexiform layer (IPL) in 15%, ganglion cell layer (GCL) in 41% and the nerve fiber layer (NFL) in 0%. If questionable grades were included as positive, CME rates increased to 41% in ONL, 72% in OPL, 74% in INL, 46% in IPL, 54% in GCL, and 5% in NFL. There was no statistically significant difference in the frequency of CME in the OPL and INL although the occurrence of CME in these layers was significantly different from all other layers. Subretinal fluid (SRF) was graded as definite in 15% of the cases and questionable in another 15% of cases. The presence of SRF correlated most strongly with CME in the ONL (p=.001) and was not significantly correlated to CME in layers anterior to the INL. Cohen's kappa coefficient for intergrader CME assessments was nearly perfect for the OPL (0.934) and showed substantial agreement for INL (0.766), GCL (0.766) and all layers grades including SRF (0.649).
In this study, CME was found most commonly and with equal frequency in the OPL and INL. CME was found significantly less frequently (p<.001) in the hyperreflective layers (IPL and NFL) than in other layers suggesting that tissue orientation may play a role in CME formation. CME was also found less frequently in the ONL than in other studies, possibly due to our refined definition of the ONL. The presence of SRF was most associated with CME in the ONL.
This PDF is available to Subscribers Only