Abstract
OBJECTIVES: To investigate whether the rate of circumpapillary retinal nerve fibre layer (cpRNFL) change during initial follow-up can predict event progression detected by guided progression analysis (GPA) over an extended follow-up. METHODS: Primary open-angle glaucoma (POAG) and glaucoma suspect patients with a minimum of seven optical coherence tomography (OCT) scans were analysed. The cpRNFL slopes were calculated for an initial follow-up of 2, 3, 4, and 5 visits (6 months apart for each visit) using a linear mixed-effects model. After the initial follow-up, the event date was then marked as the date of the first OCT test that indicated likely progression using GPA software. The cumulative risk of progression in slow (>-1.0 µm/year) vs moderate-and-fast (≤-1.0 µm/year) initial OCT slopes was compared using Kaplan-Meier survival analysis and log-rank test for the optimal scenarios. RESULTS: The study included 472 eyes (340 POAG and 132 glaucoma suspects) from 307 patients. The optimal scenario for predicting GPA progression over the ensuing 3 years included the initial 4 OCT sessions (the adjusted-HR = 1.76 (95% CI: 1.37-2.26) per initial 1 μm/faster; P < 0.001. For predicting progression over the next 5 years, the optimal scenario included the initial 5 OCT sessions that occurred for the first 2 years. The adjusted-HR was 1.65 (95% CI: 1.16-2.33) per initial 1 μm/faster; P = 0.005. CONCLUSIONS: The rate of initial cpRNFL change observed within half the monitoring period may effectively predict the subsequent structural progression and can be used by clinicians to predict the course of glaucoma when monitoring patients.