Enhanced Correlation Between Retinal Nerve Fiber Layer Thickness (RNFLT) and the Spectral Domain Optical Coherence Tomography (SDOCT) derived Neuroretinal Rim Parameter "Bruch’s Membrane Opening Minimum Rim Area" (BMO-MRA) in Human Ocular Hypertension and Glaucoma

Abstract

 

Purpose:

 

To assess the correlations between two SDOCT optic nerve head (ONH) rim area parameters, Heidelberg Retinal Tomography Rim Area (HRT-RA) and several SDOCT RNFLT and RNFL crossectional area (RNFLA) parameters in 119 ocular hypertension and glaucoma patients.

 

Methods:

 

Participants underwent HRT and SDOCT (Heidelberg Spectralis, 870nm) ONH and RNFL imaging (48 high-resolution, ONH centered, radial B-scans and one circle scan) on the same day. Data from one eye per patient was chosen randomly. For each SDOCT ONH data set, every 2nd (24) radial B-scan was manually delineated and the following parameters were quantified: 1) BMO minimum rim area (BMO-MRA); and 2) BMO-horizontal rim area (BMO-HRA). RNFLT and RNFLA were calculated from the circle scan, from a circle 1700μm from the BMO centroid or 950μm from BMO. Random effects models were constructed to assess the correlations between BMO-MRA, BMO-HRA, HRT-RA and each measure of RNFLT/RNFLA.

 

Results:

 

The mean rim area was 1.16 ± 0.31mm² using SDOCT BMO-MRA, 1.23 ± 0.35 using SDOCT BMO-HRA and 1.54 ± 0.32 mm² using HRT. The correlation between the SDOCT BMO-MRA and HRT rim area measures was 0.50. The correlations of SDOCT BMO-MRA with all RNFL measures ranged from 0.66 to 0.72 and were significantly larger (p<0.001 in all cases, Steiger’s Z₂* statistic) compared with the correlations achieved using SDOCT BMO-HRW (0.38 to 0.50). BMO-MRA correlations were also larger (p<0.05 in all cases, Steiger’s Z₂* statistic) than those achieved using HRT-RA (0.20 to 0.29). There were no significant differences between the correlations achieved using different SDOCT measures of RNFLT/RNFLA.

 

Conclusions:

 

Our data suggest BMO-MRA is a better predictor of retinal nerve fiber layer status than HRT-RA and BMO-HRA. Current ONH rim assessment techniques (HRT-RA and BMO-HRA) lack a solid anatomical basis for two reasons: 1) the clinical disc margin (HRT-RA) is not a reliable outer border of rim tissue due to clinically invisible BM extensions [Reis, et al, Ophthalmology. In Press]; and 2) the measurement that is made is not perpendicular to the greatest number of axons.  

 

Keywords: retina • imaging/image analysis: clinical • optic nerve