Structural evaluation of preperimetric and perimetric glaucoma

PURPOSE: To evaluate diagnostic ability of macular ganglion cell layer–inner plexiform layer (GCL-IPL) for detection of preperimetric glaucoma (PPG) and perimetric glaucoma and comparison with peripapillary RNFL. METHODS: Three hundred and thirty seven eyes of 190 patients were enrolled (127 normals, 70 PPG, 140 perimetric glaucoma). Each patient underwent detailed ocular evaluation, standard automated perimetry, and spectral domain optical coherence tomography. Diagnostic abilities of GCL-IPL and RNFL parameters were determined. Data were compared using one-way analysis of variance, Pearson's Chi-square test, and area under the curve (AUC). RESULTS: After adjusting for age, gender, and signal strength, all GCL-IPL and RNFL parameters except mean thickness and disc area differed significantly. Among GCL-IPL thicknesses, inferotemporal had the highest AUC (0.865) for classifying perimetric glaucoma from normals, inferior (0.746) for PPG from normals, and inferotemporal (0.750) for perimetric glaucoma from PPG. When using RNFL, inferior thickness had the highest AUC (0.922) in discriminating POAG from normal, while the same parameter had lower AUC (0.813) in discriminating PPG from normal. The average thickness had maximum AUC (0.775) for discriminating POAG from PPG. For discriminating perimetric glaucoma and normals, inferotemporal GCL-IPL had the highest strength (sensitivity 81.43% and specificity 77.96%), slightly lower than inferior RNFL thickness (sensitivity 87.85% and specificity 84.26%). The same parameters were sensitive in discriminating perimetric glaucoma from PPG (87.14% and 92.85%, respectively). However, their specificities were poor (56.43% both). CONCLUSION: RNFL had better diagnostic ability, when compared with GCL-IPL for detecting PPG and perimetric glaucoma. However, difference was small and may not be clinically relevant.