Comparing measurements of retinal nerve fiber layer thickness obtained on scanning laser polarimetry with fixed and variable corneal compensator

PURPOSE. To compare retinal nerve fiber layer (RNFL) thickness measurements obtained on scanning laser polarimetry (SLP) with commercially available instruments coupled with fixed (FCC) and variable corneal compensator (VCC). METHODS. Forty-two eyes of 42 patients underwent a complete ophthalmologic evaluation and achromatic automated perimetry (24-2 program, SITA standard strategy). Nineteen eyes were healthy (average mean deviation: -0.12 dB ± 2.26) and 23 glaucomatous (average mean deviation: -4.92 dB ± 6.49). All patients underwent SLP with both FCC and VCC. Adequate compensation of corneal birefringence on FCC-SLP was checked acquiring macular retardation map (MRM). RNFL thickness was evaluated considering superior and inferior maximum (SM, IM), average thickness and ellipse average (AT, EA), and superior and inferior average (SA, IA). Mean values (±SD) for each parameter measured by the two polarimeters were compared and linear regression calculated. The ability of each parameter to discriminate between normal and glaucomatous eyes was evaluated on both polarimeters calculating area under ROC curve. RESULTS. A significant linear correlation for all parameters was noted (r range: 0.65-0.78). VCC produced slightly higher thickness values than FCC, both in normal and glaucomatous eyes. On both polarimeters, area under ROC curve for all parameters discriminated adequately healthy from glaucomatous eyes (range: 0.68-0.81). CONCLUSIONS. In a highly comparable and selected group of normal and glaucomatous eyes, FCC-SLP and VCC-SLP showed considerable concordance in measuring peripapillary RN-FL thickness, both for sectorial and global parameters. Proper corneal birefringence compensation provided separation of normal from glaucomatous eyes on both polarimeters.