Vigabatrin-associated visual field constriction in a longitudinal series. Reversibility suggested after drug withdrawal

Purpose: To evaluate through a longitudinal study the effects on visual fields of long-term vigabatrin medication in patients with partial epilepsy and to discuss visual field screening strategies. Methods: A total of 26 patients aged 14-68 years with a mean history of vigabatrin medication of 8.5 years (range 2-14 years) were followed by manual kinetic Goldmann perimetry (objects IV,4 and I,4) for 6-26 months (mean value 12.3 months). At time zero and at follow-up, each patient was assigned a 'pooled' averaged value, as a linear percentage of normal isopter position, for the two objects as tested nasally and temporally in the five most horizontal meridians on the Goldmann chart. Twelve eyes from nine adults (age 24-60 years) served as controls. Results: Constrictions were recorded in 24 of 26 patients at baseline. Averaged isopters ranged from 8% to 96% of the controls' averaged isopter positions. Median values of 71.5% and 60.5% for large and small objects, respectively, indicated that the smaller object was more sensitive to visual field constriction. There was no difference in the degree of constriction between nasal and temporal hemifields. Significant improvement in the visual field (mean gain 13.6% units) was seen in the eight patients who underwent full drug withdrawal. No similar improvement was seen in the 12 patients still on full dose or the six with reduced intake. Conclusions: Most Danish patients on long-term vigabatrin medication have suffered some visual field loss. Contrary to most clinical evidence so far, the present follow-up study indicates some reversibility of visual field loss after drug withdrawal. Kinetic Goldmann perimetry appears to be a fair alternative to computerized static perimetry techniques for screening and following vigabatrin-treated patients.