Discrepancy between optic disc and nerve fiber layer assessment and optical coherence tomography in detecting glaucomatous progression.

PURPOSE: To compare the outcomes of Cirrus spectral-domain optical coherence tomography (OCT) and optic disc/retinal nerve fiber layer (RNFL) photographic assessment in detecting glaucomatous progression.
METHODS: Two-hundred twenty-six eyes of 130 glaucoma patients (mean follow-up: 2.5 years) with at least 5 OCT examinations were included. Eyes were classified into one of four groups (diffuse RNFL defect; localized RNFL defect; no RNFL defect; unidentifiable RNFL status) based on baseline RNFL photographs. After performing the entire series of optic disc/RNFL photographic assessments, the eyes were classified into one of three groups: stable, progressed, and undetermined. Progression was divided into one of four categories (optic disc rim thinning; widening RNFL defect; deepening RNFL defect; new disc hemorrhage). OCT progression was determined using guided progression analysis (GPA) software.
RESULTS: One-hundred thirty-nine eyes had diffuse RNFL defects, 34 eyes had localized RNFL defects, 42 eyes had no RNFL defects, and 11 eyes had unidentifiable RNFL at baseline. Forty-six eyes showed at least one category of progression upon expert assessment of optic disc/RNFL photographs, while OCT GPA detected progression in 35 eyes. Among the 34 eyes in which progression was observed in photographs only, 15 showed a new disc hemorrhage, 12 presented deepening of an RNFL defect, 10 showed optic disc rim change, and 6 had widening of an RNFL defect. Among the 23 eyes processed only by OCT GPA, 18 had a diffuse RNFL defect at baseline.
CONCLUSION: OCT GPA was more sensitive in eyes with a diffuse RNFL defect whereas photographic assessment was better for detecting optic disc hemorrhage and deepening of an RNFL defect when evaluating structural progression.