Can a Portable Flash Visual Evoked Potential (VEP) Device Identify Chiasmal Decussation Anomalies in Albinism?

Abstract

Background: Visual evoked potentials (VEPs) are used to detect chiasmal misrouting associated with albinism. However, VEPs are only performed in specialist centres and typically have long waiting lists. The portable electrophysiology device RETeval^® shows promise as a clinical screening tool across a range of ophthalmic conditions. Here, we explore its utility in detecting chiasmal abnormalities associated with albinism. Methods: Flash VEPs were recorded on the RETeval^® and by standard ISCEV techniques for 27 patients with suspected albinism and 40 control patients as part of routine appointments. We retrospectively investigated the agreeability between the two methods. The amplitude/latency of the main component was measured for standard VEPs whilst a correlation value of interhemispheric difference was calculated for the RETeval^® data. Results: We demonstrate a significant difference between albinism patients and controls (p < 0.001) with respect to the interhemispheric difference identified by the RETeval^®. By applying a threshold of 0.001865 to the correlation value, the RETeval^® detected chiasmal misrouting in all 27 patients with albinism and had 97% agreeability to standard testing. Conclusions: This study shows the potential of using the RETeval^® as a clinical tool for the diagnosis of chiasmal anomalies in albinism. The RETeval^® has significant time/cost savings which could hasten diagnoses.

Keywords: OCA; VEP; albinism; chiasmal misrouting; decussation defects.

Figure 1

Flash VEP waveforms for normal, chiasmal, and post-chiasmal defects. Schematic sketch of predicted single electrode waveforms for monocular flash VEP. (A) Normal crossing waveforms for right eye (left panel) and left eye (right panel) are close to 0 at P2. (B) Enhanced crossing, as would be seen in albinism, causes a positive interhemispheric difference when stimulating the left eye, and a negative difference when stimulating the right eye. (C) A chiasmal compression lesion causes a waveform opposite to that seen in albinism where there is a negative interhemispheric difference when stimulating the left eye and a positive difference when stimulating the right eye. (D) A left post-chiasmal defect causes positive interhemispheric differences when stimulating both the left and the right eye. (E) A right post-chiasmal defect causes negative interhemispheric differences when stimulating both the left and the right eye.

Figure 2

Example RETeval^® set up on a baby. The permission was obtained from the parent to use the photo.

Figure 3

Using the RETeval^® to detect misrouting across the chiasm as seen in albinism. (A) Representative waveforms of recordings using the RETeval^® for control patients (left panel) and albinism patients (right panel). (B) Grouped averages for control and albinism patients were generated where the main line is the mean and the shaded area shows SD. (C) Comparison of the RETeval^® score for control and albinism patients. The RETeval^® score was generated by calculating a Pearson’s correlation value for each patient to compare the interhemispheric differences from the left eye and the right eye. An unpaired t-test was performed to calculate the statistical significance between the RETeval^® scores for each group (**** shows p < 0.0001). (D) An ROC curve for detecting chiasmal crossing anomalies using the RETeval^®. The red line represents the ROC curve for a random guess whilst the grey line is the curve for our data. The area under the curve is 0.9972, with a confidence interval of 0.9903 to 1.000 showing the device to be highly sensitive and specific in detecting misrouting across the optic chiasm. Using the ROC curve, a cut-off value of 0.001865 gives 100% sensitivity and 97% specificity.

Figure 4

Using the RETeval^® to detect misrouting is not age dependent. (A) Comparison of the RETeval^® score for controls and albinism across the age groups, with mean and SD shown for each group. Two-way ANOVAs were performed using GraphPad Prism (**** denotes p < 0.0001). (B) Correlation of the RETeval score with age for controls and albinism patients. Linear regression analysis was performed and it found that age did not affect the RETeval^® score for either group (controls: R² = 0.001501 p = 0.8148; albinism: R² = 0.0223, p = 0.4482). (C) Correlation of the RETeval^® score with age for the albinism patients split by age group. Again, linear regression analysis showed that age did not affect the RETeval^® score for either age group (0–7: R² = 0.0003628, p = 0.9347; 8+: R² = 0.1697, p = 0.3584).