Correlations Between the Individual Risk for Glaucoma and RNFL and Optic Disc Morphometrical Evaluations in Ocular Hypertensive Patients.

Abstract

To assess the correlation between the 5-year risk of developing primary open-angle glaucoma (POAG) [on the basis of the Ocular Hypertension Treatment Study (OHTS)-European Glaucoma Prevention Study (EGPS) model] and optic nerve head and retinal nerve fiber layer (RNFL) indices as evaluated by optical coherence tomography (Stratus OCT), scanning laser polarimetry (GDx ECC), and confocal laser ophthalmoscopy [Heidelberg Retinal Tomograph (HRT) III] in ocular hypertensive (OHT) individuals. Sixty-eight OHT individuals with the untreated intraocular pressure in both eyes between 22 and 32 mm Hg, an ophthalmoscopically normal optic disc with a cup/disc ratio asymmetry between the 2 eyes of ≤0.2, and a visual field with glaucoma hemifield test within normal limits, and pattern SD within P<0.05 were recruited. A comprehensive glaucoma workup and examination by means of OCT, GDx ECC, and HRT III were performed in all of the patients. The probability of developing POAG in 5 years was calculated by means of the OHTS-EGPS model (risk calculator score-RS), which has been developed by the OHTS-EGPS Collaborative Study Group by taking into account the age and the mean values of the central corneal thickness, the intraocular pressure, the pattern SD, and the cup/disc ratio collected in the 2 eyes. Correlations between the individual RS and OCT, GDx, and HRT parameters were then assessed by regression analysis (exponential and linear), with a statistically significant level of P<0.01. The RS of the 68 individuals ranged between 1.5% and 28.6%, with a mean of 10.2% (SD 6.4). Only OCT RNFL-related parameters showed a statistically significant correlation with the RS. The average RNFL thickness had an exponential regression r of 0.21 (P<0.0001) and a linear regression r of 0.14 (P=0.0012); the superior average RNFL thickness had an exponential regression r of 0.26 (P<0.0001) and a linear regression r of 0.23 (P<0.0001); the nasal average RNFL thickness had an exponential regression r of only 0.12 (P=0.0031). Neither GDx RNFL-related parameters nor HRT disc-related parameters showed statistically significant correlations with the RC score. This explorative study showed that the individual risk to develop POAG within 5 years in OHT individuals is significantly correlated with OCT RNFL parameters, but not with GDx RNFL parameters or HRT disc parameters. OCT parameters could explain about 20% to 25% of the variability in individual RS. The RS seems to take into account the subclinical RNFL thinning shown by the OCT.