Comparison of pattern visual-evoked potentials to perimetry in the detection of visual loss in children with optic pathway gliomas

Abstract

Purpose: We sought to compare visual evoked potentials (VEPs) with standard visual field testing in children with visual pathway gliomas.

Methods: Fifteen of 40 children with visual pathway gliomas verified on magnetic resonance imaging scan who cooperated with Goldmann visual field (GVF) and 3-channel VEPs were studied. GVFs were obtained in 25 eyes with adequate vision. VEP amplitudes, latencies, and signal-to-noise ratios (SNRs) were compared with control subjects. Four of the patients (5 eyes) also had Humphrey visual field testing.

Results: Twenty-two of 25 eyes had a field defect, 15 eyes showed a relative or absolute hemianopia, 7 eyes showed a central or generalized depression, and 3 eyes were normal. In hemianopic eyes, 87% showed a depression (GVF) or reduced sensitivity (Humphrey field) in the opposite hemifield. VEP amplitudes and SNRs, normally largest at the midline electrode, were significantly reduced in all eyes with visual field loss. By comparison, lateral electrodes showed significantly lower amplitudes and SNRs in patients and controls. Interhemispheric VEP asymmetry (>2:1 ratio) was seen in 67% of patients with hemianopia and 53% of controls.

Conclusions: Reduction of amplitude and SNR at the midline VEP electrode was a sensitive indicator of visual field loss. Interhemispheric VEP asymmetry was not reliable in detection of a hemianopic field defect. VEPs can be a reliable and objective alternative for the detection of visual loss due to optic pathway glioma in children who are intolerant to visual field testing. We recommend the test protocol include pattern-onset and check reversal stimuli of at least one high and one low spatial frequency.