Latencies of the Multifocal Visual Evoked Potential and the Diagnosis of Optic Neuritis and Ischemic Optic Neuropathy

Abstract

Purpose:: Optic neuritis (ON) and ischemic optic neuropathy (ION) can usually be distinguished on clinical grounds, but the overlap in signs and symptoms sometimes renders this decision impossible. Here, we examine the use of the latency of the multifocal visual evoked potential (mfVEP) to differentiate between ON and ION.

Methods:: Thirty-four patients with ON, 5 with bilateral disease, (39 eyes) and 32 patients with ION, 2 with bilateral disease, (34 eyes) were enrolled in the study and compared with 50 normal controls (100 eyes). Shortly after presentation to a neuro-ophthalmologist for vision loss, monocular mfVEPs were obtained from both eyes using a 60-sector, pattern-reversal dartboard array, 44.5 deg in diameter with VERIS software (EDI, San Mateo). Recording electrodes were placed at the inion (I) and I+4 cm, and at two lateral locations up 1 cm and over 4 cm from I. Responses were analyzed with custom software. [1] For monocular analysis, latencies compared to a template of 100 normals were determined for each of the 60 sectors that met a minimal SNR. [1,2] For interocular analysis, relative latencies were determined by comparing the responses from the two eyes. [1] Each of the sectors was also classified as having normal or abnormal latency based on 1% and 5% confidence intervals derived from the normal database.

Results:: ION patients had an average monocular mean latency of 3.7 ms, compared to 14.5 ms for ON and 0.5 ms for control subjects, but there was overlap in the mean latencies for all three groups, especially in the case of the ON and ION patients. Using interocular analysis, controls had an average mean latency difference of 0.9 ms, compared to 3.7 ms for ION and 13.2 ms for ON, again with overlap between the groups. The best differentiation between the ON and ION was achieved, using cluster analysis. With ON, 64% of affected eyes had 3 adjacent sectors with abnormal latencies within a 1% confidence interval compared to 4% of control eyes and 3% of ION eyes. The sensitivity and specificity of this analysis can be varied by using different cluster criteria.

Conclusions:: While there is overlap in the latencies seen with the mfVEP in ON and ION, the mfVEP with a cluster analysis can be a useful supplement to the diagnosis of these diseases. 1. Hood et al Doc Ophthal (2004) 2. Fortune et al Doc Ophthal (2004)

Keywords: electrophysiology: clinical • neuro-ophthalmology: diagnosis • neuro-ophthalmology: optic nerve