Detection of Primary Angle Closure Glaucoma Progression by Optical Coherence Tomography.

PURPOSE

To compare the role of spectral-domain optical coherence tomography (SD-OCT) in regard to retinal nerve fiber layer (RNFL) and ganglion cell complex (GCC) assessment in the detection of primary angle-closure glaucoma (PACG) and primary open-angle glaucoma (POAG) progression.

MATERIALS AND METHODS

In the prospective study, 131 subjects with PACG and POAG were examined during 72 months with follow-up visits every 6 months. Visual field (VF) progression was detected using the Guided Progression Analysis (GPA) of the Humphrey visual field analyzer and structural change using SD-OCT while a significant negative trend for the RNFL and GCC was gauged. The diagnostic accuracy of RNFL and GCC thinning in the detection of glaucoma progression was compared between PACG and POAG eyes using the Kaplan-Meier method with the calculation of the log-rank test.

RESULTS

Progression was detected in 57% of eyes with POAG and 59% of eyes with PACG. The rate of thinning of RNFL (-2.95±1.85 μm/y) and GCC (-3.22±2.96 μm/y) was significantly higher in PACG progression eyes compared with POAG [-1.64±2.00 μm/y (P=0.018) and -1.74±2.05 μm/y (P=0.046), respectively]. The progression was associated with initial pattern standard deviation in both glaucoma subtypes, while only in PACG-with long-term intraocular pressure fluctuations (cutoff >5.2 mm Hg) and lens thickness (cutoff >4.92 mm), and only in POAG-with initial focal loss volume of GCC (cutoff >1.5%).In PACG, the rate of the visual function deterioration correlated with GCC thinning rate (r=0.330, P=0.027), but not with the RNFL thinning rate (r=-0.010, P=0.79), while in POAG, it was significant for both RNFL thinning (r=0.296, P=0.039) and GCC thinning (r=0.359, P=0.011). In PACG patients with progressive GCC thinning, functional progression was detected earlier (log-rank test P≤0.001) than in patients with progressive RNFL thinning (log-rank test P=0.457), while for POAG, these results were P=0.012 and ≤0.001 for GCC and RNFL thinning, respectively.

CONCLUSIONS

SD-OCT plays an important role in detecting PACG progression. In contrast to POAG, GCC thinning predicted functional loss better than RNFL thinning in PACG.