Purchase this article with an account.
William Howard Swanson, Brett J King, Douglas G. Horner; Development of Normative Reference Values for En Face RNFL Thickness. Invest. Ophthalmol. Vis. Sci. 2017;58(8):1321. doi: https://doi.org/.
Download citation file:
© ARVO (1962-2015); The Authors (2016-present)
En face imaging of the retinal nerve fiber layer (RNFL) with optical coherence tomography (OCT) makes it possible to assess RNFL thickness across the retina. We performed an exploratory analysis of between-subject variability in RNFL thickness from the optic disc to the fovea.
A Spectralis (Heidelberg Engineering) was used to image and auto-segment RNFL for both eyes of 30 healthy people ages 20-31 (mean ± SD = 25±3) yr, one eye each of 25 healthy people ages 49-82 (66±8) yr and one eye each of 30 patients with glaucoma ages 49-87 (69±8) yr. Analysis methods were developed for sixty-eight 2° square boxes around the line connecting the fovea and the optic disc using data from the left eyes (OS) of the 20-31 yr group. The Coefficient of Variation (CoV) for each box was computed and mean CoVs were used to develop reference ranges. Then the methods were applied to the right eyes (OD) and correlations performed between eyes. Next, the analysis was used to derive reference ranges for the 40-85 yr group, and finally these were used to flag suspect retinal regions in the patients with glaucoma.
The 20-31 yr group had mean CoV 0.18 OS and 0.19 OD, and thicknesses had strong between-eye correlation (R2 = 75%). With a reference range of ±2 CoVs, 1.8% of boxes fell below the reference range for OS and 2.7% for OD. Most of these locations were from three people who had the three lowest values for mean RNFL thickness for both eyes; two had arcades farther from the fovea than average, and one had arcades closer to the fovea than average. Then the reference range was increased to ±2.7 CoVs, so that no more than 2 boxes were below the reference range for any eye in this group. For the 49-82 yr group the mean thickness was 8% smaller, the mean CoV was 0.23, and using ± 2.7 CoVs no more than 2 boxes fell below the reference range for any eye. Twenty-one of 30 patients with glaucoma had more than 2 boxes that fell below the reference range, and half had 7 to 45 boxes.
The strong correlation between eyes was evidence that CoV reflects between-subject variability rather than effects of segmentation artifacts. Most of the patients with glaucoma had RNFL regions below the 49-82 yr reference range. Normal between-subject variability in the location of the arcades limited the ability of RNFL thickness measures to identify defects. Reduction in effects of this variability on thickness indices is needed.
This is an abstract that was submitted for the 2017 ARVO Annual Meeting, held in Baltimore, MD, May 7-11, 2017.
This PDF is available to Subscribers Only