Macular Ganglion Cell Layer Assessment to Detect Glaucomatous Central Visual Field Progression

2016 
PURPOSE: To investigate the use of ganglion cell inner plexiform layer (GC-IPL) thickness, as measured by spectral domain optical coherence tomography, to detect central visual field (VF) progression. METHODS: This study included 384 eyes from 384 patients (219 preperimetric and 165 perimetric glaucomatous eyes; average follow-up, 4.3 years). Photographic assessment of retinal nerve fiber layer (RNFL) and serial VF analysis were performed to detect glaucoma progression in the central (within 10°) area. Study inclusion required at least five serial spectral domain optical coherence tomography exams at different visits. The long-term test-retest variability of average GC-IPL thicknesses was calculated in 110 stable preperimetric glaucomatous eyes. The sensitivity and specificity of GC-IPL measurements for the detection of central VF progression were calculated in an event-based analysis using the calculated variability as a cut-off and were compared with those of central RNFL photographic assessment. RESULTS: The intersession test-retest variability, defined as the 95% confidence interval, was 1.76 µm for average GC-IPL thickness. The sensitivity and specificity of the average GC-IPL thickness for detecting central VF progression were 60.7% and 78.9%, respectively. Among six sectors, the inferonasal GC-IPL sector showed the highest sensitivity (53.6%). The sensitivity of the ≥1 sector GC-IPL to detect central VF progression was significantly higher than that of central RNFL photographic progression (p = 0.013). Other GC-IPL parameters showed comparable sensitivity and specificity to detect central VF progression compared with RNFL photographic progression. CONCLUSIONS: Serial GC-IPL measurements show comparable performance in the detection of central glaucomatous VF progression to RNFL photographic assessment.
    • Correction
    • Source
    • Cite
    • Save
    • Machine Reading By IdeaReader
    23
    References
    2
    Citations
    NaN
    KQI
    []