Epiretinal membrane as a source of errors during the measurement of peripapillary nerve fibre thickness using spectral-domain optical coherence tomography (SD-OCT).

Abstract

We aimed to examine the extent to which measurement errors in the determination of retinal nerve fibre layer (RNFL) using spectral-domain optical coherence tomography (SD-OCT) occur in cases of epiretinal membrane and whether systematic deviations are found in the values obtained. A macular scan and a circumpapillary scan were performed on 97 eyes of 97 patients using SD-OCT. Group 1 comprised 53 patients with epiretinal membrane at an age of 70 ± 4.8years (median ± average absolute deviation). Group 2 consisted of 44 patients without any macular pathologies (median age 70 ± 5.8years). Differences in the thickness of the RNFL and segmentation errors in the detection of the RNFL were recorded quantitatively in both groups and checked for statistical significance using non-parametric tests. The median central retinal thickness in Group 1 was 357 ± 79μm (median ± average absolute deviation), and in Group 2 it was 270 ± 11μm (p < 0.001). The result of the quadrant-by-quadrant measurement of the average RNFL in Group 1 and Group 2, respectively, was: temporal 88 ± 17 and 73 ± 9μm, inferior 121 ± 17 and 118 ± 15μm, nasal 87 ± 15 and 89 ± 14μm and superior 115 ± 15 and 114 ± 9μm. Temporally, the difference was statistically significant (p < 0.001). Segmentation errors of the RNFL were found in 19 of 53 eyes (35.8%) in Group 1 and in no eyes (p < 0.001) in Group 2. In eyes with epiretinal membrane, measuring errors in the SD-OCT occur significantly more frequently than in eyes without any retinal pathologies. If epiretinal membrane and glaucoma are present simultaneously, the results of the automated RNFL measurement using SD-OCT should be critically scrutinised, even if no papillary changes are visible clinically.