Abstract
Purpose :
There is evidence that baseline retinal nerve fiber layer (RNFL) or macular measurements derived from spectral domain optical coherence tomography (SD-OCT) may forecast glaucoma deterioration. We tested the hypothesis that baseline RNFL or macular ganglion cell-inner plexiform layer (GCIPL) thickness can predict visual field (VF) progression in a cohort of established glaucoma patients.
Methods :
171 eyes of 95 glaucoma patients with good-quality baseline RNFL and macular SD-OCT images (Optic Disc and Macular Cubes 200x200, Cirrus HD-OCT), more than 2 years of follow-up and 5 or more VFs were selected from the clinical database and ongoing studies at Stein Eye Institute’s Glaucoma Division. Guided Progression Analysis (GPA) software was used to define time to progression with pointwise event analysis according to Early Manifest Glaucoma Trial criteria. Trend analysis of Visual Field Index (VFI) was used to define VF progression at the end of follow-up (p <5% for VFI slope over time). Logistic and Cox’s proportional hazard regression were used to test the hypothesis and adjust for other baseline prognostic factors for VF progression.
Results :
Median (IQR) age and VF mean deviation at baseline were 68 (61-72) years and –2.9 (–6.2 to –1.1) dB. Median (IQR) follow-up time was 54 (44-65) months. Seventeen and 39 eyes were determined to be progressing at the final follow-up based on GPA and VFI, respectively. In logistic models using GPA event outcomes, thinner CCT (OR=1.02 per μm, p =0.001), higher baseline IOP (OR=1.3 per mmHg, p=0.009), and female gender (OR=7.5, p=0.02) were associated with higher risk of progression. In logistic models using a significantly negative VFI slope as the outcome, thinner baseline RNFL (OR=1.04 per μm, p=0.022) was the only predictor of progression. Thinner CCT (p<0.001), female gender (p=0.001), younger age (p=0.004), and higher baseline IOP (p=0.008) predicted VF progression on Cox’s proportional hazard model. The baseline average or sectoral GCIPL thickness was a predictor for VF progression in none of the explored models (p >0.15).
Conclusions :
Thinner CCT, female gender, and higher baseline IOP were consistent predictors of subsequent VF progression in time-to-event models. Average baseline RNFL predicted VF worsening when a binary VFI outcome was used. Baseline global or regional macular GCIPL thickness was not a prognostic factor for VF progression.
This is an abstract that was submitted for the 2017 ARVO Annual Meeting, held in Baltimore, MD, May 7-11, 2017.