Diagnostic use of macular layer analysis by SD-OCT in primary open angle glaucoma

Purpose. - The purpose of this study is to evaluate the diagnostic ability of segmentation of the various internal macular layers by spectral domain optical coherence tomography (Cirrus HD-OCT, Carl Zeiss Meditec (CZM), Dublin, CA, USA) and to compare it to that of the peripapillary retinal nerve fiber layer (cpRNFL) in primary open angle glaucoma (POAG). Materials and methods. - This study included 252 eyes diagnosed with primary open angle glaucoma (POAG) (164 early POAG, 44 moderate POAG and 44 advanced POAG) and 223 eyes of control subjects. All patients underwent visual field testing (Humphrey Field Analyser, SITA-Standard 24-2, CZM), and SD-OCT imaging (Cirrus HD-OCT) of the macular and optic nerve head regions (ganglion cell analysis (GCA), macular cube 200 x 200; optic disc cube 200 x 200). OCT macular scans were segmented into macular nerve fiber layer (mNFL), ganglion cell layer with inner plexiform layer (GCIPL), outer retinal layers, and ganglion cell complex (GCC) (mNFL + GCIPL). Glaucoma discriminating ability was assessed using the area under the receiver operator characteristic curve (AUC) for all macular parameters and mean circumpapillary retinal nerve fibre layer (cpRNFL). Results. - For the entire POAG population of this study, the minimum GCIPL index provided greater diagnostic ability than the other parameters studied, with a statistically significant difference in their AUC (minimum GCIPL [0.887], mean GCIPL [0.865], GCC [0.863], cpRNFL [0.823], mean mNFL [0.786] and minimum mNFL [0.742]). The results were similar in the early POAG group but without any statistically significant difference with the cpRNFL parameter. In the moderate POAG group, the diagnostic ability was similar for all indices, whereas in the advanced POAG group, minimum GCIPL and GCC gave the largest AUC indices. Discussion and conclusion. - The minimum macular GCIPL is a new index obtained with the GCA algorithm of the Cirrus HD-OCT. It appears to have an excellent ability to detect glaucoma at every stage and demonstrates performance comparable to that of the cpRNFL parameter, in combination with which it may provide important complementary information for clinical practice. (C) 2013 Elsevier Masson SAS. All rights reserved.