Purchase this article with an account.
P.G. Schlottmann, J.M. Shewry, F.C. Ikeji, E.T. White, D.F. Garway–Heath; Correlation between Retinal Nerve Fibre Layer Thickness and Visual Field Sensitivity at Different Retinal Eccentricities . Invest. Ophthalmol. Vis. Sci. 2004;45(13):5522.
Download citation file:
© ARVO (1962-2015); The Authors (2016-present)
Purpose:Published histological data suggest greater ganglion cell loss is required in the central, compared to peripheral, retina for similar degrees of visual field (VF) loss. We evaluate the strength and pattern of correlation between VF sensitivity and retinal nerve fibre layer thickness (RNFLT), measured by scanning laser polarimetry (SLP) and optical coherence tomography (OCT) at various retinal eccentricities. Methods: 52 eyes randomly selected from 24 normal subjects, 17 ocular hypertensive and 11 glaucoma patients were scanned using OCT 3 and GDxVCC. Diagnosis was based on repeatable SITA 24–2 VF defects and not on optic nerve head appearance. RNFLT measurements were made along a circle centred on the optic nerve head of default radii 3.4 mm for OCT and 3.2 mm for GDx. The circle was divided into twelve 30–degrees sectors from raw data obtained from each device. Each sector was correlated with each point of the 24–2 SITA visual field according to a correspondence map (Garway–Heath et al., Ophthalmology 2000). Linear correlations were calculated between decibel (dB) differential light sensitivity (DLS) and thickness using a micron scale and a logarithmic transformation of microns to dB. Correlations were calculated for various degrees of VF eccentricity and the structure/function slope value recorded. Results: Correlations were better when using the logarithmic transformation of RNFLT for all degrees of eccentricity. Worst R2 values were found for both OCT and GDx at 4.2º eccentricity (0.01 and 0.00 respectively). Best R2 values were found for both OCT and GDx at 15.3º and 27.2º eccentricity (0.36 and 0.16; 0.28 and 0.30, respectively). Structure/function slopes were < 1 within 10º eccentricity, between 1 and 2 for 10º to 25º eccentricity and between 3 and 4 at 27º eccentricity. Conclusions: Correlation between DLS and RNFLT improves when both DLS and RNFLT are measured in the same scale (in this case, dB). Correlations generally improved with greater eccentricity. The structure–function slope gets steeper with greater eccentricity, indicating greater RNFLT loss per unit loss of function is required in central, comapred to peripheral, retina.
This PDF is available to Subscribers Only