Effect of uncompensated corneal polarization on the detection of localized retinal nerve fiber layer defects.

PURPOSE

To investigate the effect of corneal polarization on the detection of retinal nerve fiber layer (RNFL) defects in an artificial eye using scanning laser polarimetry.

METHODS

The model eye was composed of a film with a birefringence magnitude of 80 nm (single-pass retardance) in front of a plastic disc with a radial axis of birefringence, which was partially covered with a strip of black paper. The plastic disc, black paper and the film simulated the RNFL, the RNFL defect and the cornea, respectively. The model eye was located at 25 cm from the lens of the Nerve Fiber Analyzer II (NFA II) or GDx with variable corneal compensator (GDx-VCC). Using NFA II, the retardation of the laser beam was measured with/without the artificial cornea in differently simulated corneal birefringence axes that are 15 degrees nasally downward, and 15 degrees nasally upward. Using a GDx, the model eye with/without cornea was measured with a fixed corneal compensator (GDx-FCC) or GDx-VCC.

RESULTS

Using NFA II or GDx-FCC, there were three areas in the RNFL thickness profile where an RNFL defect was detected, undetected and reversed. The mean thickness of the RNFL defects was not different from the mean thickness of RNFL in one model eye using NFA II (p = 0.3). Using GDx-VCC, all RNFL defects were detected in all model eye settings.

CONCLUSIONS

The influence of incorrect compensation of the anterior segment birefringence is significant. With an FCC, there were particular locations in the retina where the RNFL defects were not detected. The GDx-VCC is a promising tool in glaucoma diagnosis because it detects RNFL defects in any corneal setting.