Statistics from Altmetric.com
Peripheral visual field defects in patients taking vigabatrin seem to arise in a proportion of patients. Typically a concentric narrowing of the field is seen with preservation of central acuity and colour vision. Case definition remains poorly defined and prevalence figures ranging from as low as 0.14% to as high as 39% have appeared.1 2 The high figures reported by Lawden et al, this issue, pp 716–722,3 reflect a definition which includes any bilateral field defect that could not be explained in terms of other retinal or neurological pathology regardless of severity of symptomaticity. This clearly highlights the problem but leaves more work to be done.
Visual field testing alone does not give any insights into pathogenesis as both cortical and retinal effects could cause the same result. The ocular examination is usually normal, although optic disc pallor is sometimes seen. However subtle electrophysiological abnormalities which include electro-oculogram light rise reductions, loss of oscillatory potentials on the Ganzfield electroretinogram and focal ERG changes do point to a retinal origin of the visual loss.4 The inhibitory neurotransmitter GABA is found in retinal ganglion and amacrine cells and it is a plausible hypothesis that anticonvulsants which irreversibly inhibit GABA amino-transferase such as vigabatrin will impair retinal function.
The natural history of this effect is a concern. In an audit from our own department we found five cases with major field defects (no responses to the Goldmann 1/4e isopter outside 25 degrees nasally and 40 degrees temporally) and in only three of these did the visual field improve after withdrawal or reduction of the medication. This seems to be consistent with anecdotal reports elsewhere.
It has been suggested that epileptic patients taking other anticonvulsant drugs might show similar field abnormalities if they were tested in the same way. Lawden et alhave dispelled this by showing that 16 controls taking a range of other anticonvulsant drugs all had normal fields.
Prescribers and patients now face a range of practical difficulties. Firstly, there are patients whose seizure control is good enough for them to apply for a driving licence, but whose visual fields may disqualify them. Then there are other patients who have mild field defects of no great clinical relevance. They need simple reassurance that there is no reason to change their medication on visual grounds. A further group are young children and adults of cognitive age of less than 9 years who cannot reliably perform perimetry testing. Prescribers should beware false positive field abnormalities and appreciate the overall limitations of perimetry here.5 Unfortunately standard clinical electrophysiology on its own will not provide an alternative in practice as both electro-oculography and electroretinography performed to the reliability standards of the International Society for the Electrophysiology of Vision are not widely available. Cognitively impaired people in whom there are no occupational consequences from having reduced fields may not be helped by referrals to ophthalmology clinics to diagnose field defects which are too subtle to detect by simple confrontation techniques, especially when the benefits of good seizure control may outweigh the risks of mild field loss. Clinical guidelines for the visual assessment of these patients will need to be developed jointly by neurologists and ophthalmologists according to local circumstances. In the meantime, more data on the prevalence of both mild and severe field defects and on the natural history will be helpful.