Longitudinal neuroretinal rim and retinal nerve fibre layer change in glaucoma patients and healthy controls: Factors influencing rate estimates

Abstract

 

Purpose

 

To determine longitudinal rates of change of disc margin-based rim area (DMRA), Bruch’s membrane opening-based minimum rim width (BMO-MRW) and peripapillary retinal nerve fibre layer thickness (RNFLT) in glaucoma patients and healthy controls.

 

Methods

 

We analyzed data of one eye of 187 treated glaucoma patients and 32 controls followed for a median of 4 (range: 2-6) years, tested at 6-month intervals. Confocal scanning laser tomography (Heidelberg Retina Tomograph) was used to measure global DMRA, while spectral domain OCT (Spectralis) radial and circle scans centred on the optic nerve head were used to measure global mean BMO-MRW and RNFLT, respectively. Applanation tonometry was performed on each visit. Individual rates of change were estimated with ordinary least squares regression (OLSR) and linear mixed effects modeling was used to estimate average rates and evaluate the influence of baseline age, baseline measurement, disc size (for DMRA) or BMO area (for BMO-MRW and RNFL), image quality, intraocular pressure (IOP) and presence of glaucoma.

 

Results

 

There was considerable variation in OLSR individual rates of change (Fig. 1). Control subjects had significant reduction of BMO-MRW (mean, -1.66 µm/y, p<0.01) and RNFLT (mean, -0.41 µm/y, p=0.02), but not DMRA (mean, 0.13x10^-2 mm²/y, p=0.67). Baseline BMO-MRW and RNFLT significantly influenced the rates of change, with thicker baseline values relating to faster reduction (p=0.04 for both). Baseline age significantly influenced the rate of only BMO-MRW change, with older age related to less pronounced reductions (p<0.01). The IOP at each visit significantly influenced BMO-MRW and DMRA measurements at the corresponding visit, with lower IOP related to higher measures (p<0.01 for both), However, this effect was not observed for RNFLT (p=0.84; Fig. 2). After adjusting for covariates, the mean difference in rates between glaucoma patients and controls was -1.36 µm/y (p=0.04) for BMO-MRW rate, -0.35 µm/y (p=0.14) for RNFLT rate, and -0.65x10^-2 mm²/y (p=0.14) for DMRA.

 

Conclusions

 

Longitudinal change of BMO-MRW, RNFLT and DMRA can be significantly influenced by age, baseline measurement, IOP and the presence of glaucoma, with each covariate exerting a distinctive influence. These effects must be considered in the assessment of glaucomatous progression.