Abstract
Purpose: :
To determine whether optical coherence tomography (OCT) anatomic measurements are useful in predicting the development of glaucomatous visual field (VF) defects.
Methods: :
We analyzed the data from the glaucoma suspects and pre-perimetric glaucoma patients group (GSPPG) enrolled in the multi-center longitudinal Advanced Imaging for Glaucoma Study (www.AIGStudy.net). Both time-domain (TD) and spectral-domain (SD) OCT were used to measurement thickness profiles of peripapillary retinal nerve fiber (RNFL) and ganglion cell complex (GCC) layers. Standard automated perimetry was used to assess VF every 6 months. Conversion to perimetric glaucoma was defined when pattern standard deviation (PSD) or glaucoma hemifield test fell outside of normal limits (p<0.05 and p<0.01, respectively) on 3 consecutive tests. A Cox proportional hazard model was used to calculate the hazard ratios (HR) for risk factors; results were adjusted for correlation between eyes from the same individual. A multivariate model was fitted for each of the OCT parameters with age and PSD.
Results: :
The analysis included 492 eyes (229 participants), among which 341 had pre-perimetric glaucoma (optic nerve head or RNFL defects), 129 had ocular hypertension, and 22 had glaucoma in the fellow eye. Average follow-up was 38 months. In the cohort, 34 eyes had visual field conversion. In multivariate models that controlled for baseline age and PSD, the overall RNFL thickness was found to be a significant risk factor with TD-OCT (HR=1.66 per 10 micron thinner, p=0.004) and SD-OCT (HR=1.68, per 10 micron thinner, p=0.002). The average GCC thickness measured by SD-OCT was also a risk factor (HR=1.23, p=0.02 per 10 micron thinner). TD and SD-OCT RNFL overall thickness classified based on a 1% cut-off of normative databases resulted in HR’s of 3.8 (p<0.001) and 3.6 (p<0.001), respectively. A similarly defined abnormal classification on overall GCC thickness measured by SD-OCT had a HR of 2.2 (p=0.03).
Conclusions: :
RNFL and GCC thickness measured by OCT at baseline visit are predictive of VF conversion among glaucoma suspects and pre-perimetric glaucoma patients after controlling for age and disease severity in multivariate analysis. Abnormal classification based on normative data can be used to evaluate patient risk.
Keywords: imaging/image analysis: clinical • ganglion cells • nerve fiber layer