Abstract: Objective To determine the diagnostic performance of macular ganglion cell-inner plexiform layer (GCIPL) thickness to discriminate early glaucoma form normal eyes. Design Evaluation of diagnostic technology. Participants Early glaucoma patients (30 cases, 30 eyes) and healthy controls (56 cases, 56 eyes). Methods All subjects were underwent macular scanning and peripapillaryretinal nerve fiber layer (RNFL) scanning using the Cirrus high-definition optical coherence (HD-OCT) ganglion cell analysis (GCA) algorithm (Carl Zeiss Meditec, Dublin, CA). The GCA algorithm was used to detect the macular GCIPL and to measure the thickness of the overall average, minimum, superotemporal, superior, superonasal, inferonasal, inferior, and inferotemporal GCIPL. The area under the receiver operating characteristic curve (AUC) was used to assess the ability to discriminate early glaucomatous eyes form normal eyes of each testing parameter. Main Outcome Measures The thickness of GCIPL and RNFL, and their AUCs. Results There were statistical differences in all measurement parameters (GCIPL and RNFL) between normal and glaucoma patients (all P< 0.01). The average GCIPL thickness of normal and glaucomatous eyes were (85.43 ± 5.27) μm and ( 69.30 ± 7.71) μm, and the average RNFL thickness of normal eyes and glaucomatous eye were (100.98±7.98) μm and (78.80 ± 10.38) μm (P<0.001). The largest AUC of all GCIPL parameters were the minimum and inferotemporal sector (both were 0.985), followed by the average (0.971), inferior sector (0.941), superotemporal sector (0.934), and superonasal sector (0.907). The largest AUC of all RNFL parameters was the average (0.990). Conclusion The macular GCIPL parameter was comparable to the peripapillary RNFL parameter, performs good in discriminating normal and glaucomatous eyes, and GCA algorithm may be a valuable tool for the diagnosis of early glaucoma.

Key words: ganglion cell-inner plexiform layer, retinal nerve fiber layer, optical coherence tomography, glaucoma/diagnosis