Abstracts

Rationale: Vigabatrin (VGB) attributed visual-field loss (VFL) occurs in 30 - 50% of patients. The standard for assessing VFL is using perimetry. However, it is estimated that 10-20% of adults with epilepsy are unable to perform reliable perimetry and it is an unsuitable tool to use in young children. It has been suggested that perimetry is not sensitive enough to detect early changes in the retina induced by VGB and patients may have significant retinal pathology before VFL is detected. This has previously been shown by demonstrating retinal nerve fibre layer (RNFL) thinning in VGB-exposed patients (Invest Ophthalmol Vis Sci 2006 47:917-24). We have used optical coherence tomography (OCT) to explore a relationship between structural changes in the retina and VFL. From our knowledge of the topographic relationship between peripapillary RNFL changes and specific localised visual-field defects, we understand that the peripheral retina and visual field is largely represented by retinal nerve fibres entering the superior and inferior aspects of the optic disc. Using this alongside our awareness of the well characterised peripheral, concentric VFL seen in VGB-exposed patients, we have chosen to explore whether changes in peripapillary RNFL thickness in the superior and inferior poles shows a relationship with VFL. Methods: 36 VGB-exposed patients were recruited, 8 were excluded, 4 because of other visual-field or eye pathology and 4 were unable to perform perimetry. Data from 28 patients was used in the analysis. All patients were assessed using Goldmann perimetry and OCT fast RNFL thickness scan. Mean radial degrees (MRD) were used to quantify the visual field. Severe VFL was described as < 30° MRD, 30-50° was described as mild-moderate constriction and MRDs >50° were described as normal. The RNFL thickness for each patient was analysed based on the RNFL Thickness Average Analysis Report provided by the OCT software. The absolute average RNFL thickness for each of four 90° peripapillary sectors (superior, nasal, inferior and temporal) was analysed as well as a description of which percentile (≤ 1st, 5th or 95th percentile) the absolute value fell into, based on the manufactures generic database of several hundred age-corrected normal values. For this analysis, values which fell into the 1st percentile were classed as demonstrating abnormal thinning. Results: 5 patients had normal visual fields, 13 mild/moderate VFL and 10 had severe VFL. 90% of patients with severe visual field loss showed thinning of the RNFL in at least one quadrant, 46% with mild/moderate VFL showed RNFL thinning. No patients with normal visual fields showed thinning. A Pearson correlation showed that there was a positive correlation between MRD and RNFL thickness in the superior and inferior sectors (r = 0.882, p < 0.001). Conclusions: RNFL thickness using OCT may provide a suitable tool for assessment of visual function in patients taking VGB. Average RNFL thickness in the peripapillary superior and inferior sectors correlates with the degree of VFL.