Baseline Risk Factors for Event and Trend-based Visual Field Glaucoma Progression using Fourier-Domain Optical Coherence Tomography in the Advance Imaging for Glaucoma Study

Abstract

Purpose: To determine whether optical coherence tomography anatomic measurements are useful in predicting the development of glaucomatous visual field progression.

Methods: We analyzed the data from perimetric glaucoma (PG) patients enrolled in the multi-center longitudinal Advanced Imaging for Glaucoma Study. Fourier-domain optical coherence tomography (FD-OCT) was used to measure disc variables and thickness profiles of retinal nerve fiber layer (RNFL) and ganglion cell complex (GCC). Standard automated perimetry was used to assess visual field. Subjects were followed every 6 months. Event-based visual field (VF) glaucoma progression was defined as significant worsening of ≥ 3 test locations on pattern deviation plot (three completely filled black triangles) repeated on three consecutive follow-up VF’s by Humphrey Glaucoma Progression Analysis software. The trend-based VF progression was defined as significant negative VFI slope shown in the Guided Progression Analysis software. Endpoint was reached when either type of progression was observed. Cox proportional hazard model was used to calculate the hazard ratios (HR) of the baseline risk factors. A multivariate model was fitted for each of the parameters with baseline pattern standard deviation (PSD); correlation from eyes of the same subjects was accounted for.

Results: The analysis included 302 eyes (204 participants) with average age of 61.8 years, among which 123 (60%) are female and 21 (10%) are African Americans. Average PSD at baseline was 5.7 while Average mean deviation (MD) was -4.8. Average follow-up time was 51 months. In the cohort, 48 eyes had event-based progression, 64 had trend-based progression and 29 had both. The most significant risk factors for VF progression was the GCC focal loss volume (FLV) (HR = 1.09, per 1% higher, p =0.0003). Other independent risks included baseline GCC global loss volume (GLV) (HR = 1.03, per 1% higher, p =0.03) and central corneal thickness (HR = 1.15 per 20 μm thinner, p = 0.016). None of the disc variables was a significant risk factor.

Conclusions: Glaucoma patients with higher baseline GCC FLV/GLV are more likely to have worsening of visual field defects, even after controlling for disease severity (VF PSD) in multivariate analysis, thus providing glaucoma patients in care with a better prediction for progression.

Keywords: 464 clinical (human) or epidemiologic studies: risk factor assessment • 758 visual fields • 531 ganglion cells