Reduction of posterior pole retinal thickness in glaucoma detected using the Retinal Thickness Analyzer.

OBJECTIVES To report evidence of changes in posterior pole retinal thickness in glaucoma and preperimetric glaucoma. DESIGN Cross-sectional study. PARTICIPANTS One hundred forty participants (140 eyes): 31 normal subjects, 33 with ocular hypertension, 43 with primary open-angle glaucoma (POAG), and 33 with preperimetric glaucoma. INTERVENTION Posterior pole retinal thickness measurements using the Retinal Thickness Analyzer (RTA; Talia Technology Ltd., Neve-Ilan, Israel). MAIN OUTCOME MEASURES The mean foveola average thickness (VAV), foveola minimum thickness (VMI), perifoveal average thickness (PFAV), perifoveal minimum thickness (PFMI), posterior pole average thickness (PPAV), and posterior pole minimum thickness (PPMI) were compared among the groups. The correlation between RTA values and visual field values was evaluated. The area under the receiver operator characteristic (AROC) curve, the optimal cut values, and the predictability values were estimated statistically between normal and the other groups using RTA values. RESULTS Results from 91 eyes (65%) were included. Perifoveal average thickness values in preperimetric glaucoma (190.4+/-15.6 microm; mean+/-standard deviation) and POAG (185.7+/-16.8 microm) were significantly less than in normal controls (211.4+/-16.8 microm) and in eyes with ocular hypertension (211.6+/-15.9 microm). Perifoveal minimum thickness, PPAV, and PPMI were significantly less in preperimetric glaucoma and in POAG than in normal controls; VAV and VMI were not significantly different among groups. Perifoveal average thickness, PFMI, PPAV, and PPMI significantly correlated with visual field mean defect (R = -0.38, -0.42, -0.36, and -0.41, respectively) and loss of variance (R = -0.29, -0.30, -0.26, and -0.17, respectively). The AROC using PFAV, PFMI, PPAV, and PPMI was significantly larger than 0.5 between normal and preperimetric glaucoma (0.83, 0.92, 0.82, and 0.87, respectively) and between normal and POAG eyes (0.86, 0.97, 0.83, and 0.93, respectively). When using the PFMI, the sensitivities at the minimum specificity cutoffs of 90% were highest in comparisons between normal and preperimetric glaucoma (sensitivity, 75%) and between normal and POAG eyes (sensitivity, 95%). By the logistic regression analysis, PFAV, PFMI, PPAV, and PPMI distinguished between normal and preperimetric glaucoma (R(2) = 0.30, 0.49, 0.27, and 0.37, respectively) and between normal and POAG (R(2) = 0.37, 0.59, 0.31, and 0.51, respectively). CONCLUSIONS Posterior pole retinal thickness decreases in early- and moderate-stage glaucoma. Reduction of perifoveal retinal thickness is correlated with visual field loss. In vivo measurements of posterior pole retinal thickness may help distinguish between normal and glaucomatous eyes.