Many more patients may be affected than were found in study
EDITOR--We note that Wilton at al report an increase in the number of confirmed cases of visual field defects attributed to vigabatrin in their observational cohort study.[1] However, both of their studies were based on questionnaires. Since many of the visual field defects observed in epileptic patients treated with vigabatrin are asymptomatic and the field loss may begin in the periphery, as in glaucoma,[2] we suspect that many more patients have visual field defects than is implied by these data.
We recently presented data on 14 epileptic patients treated with vigabatrin (cases) to the International Society for Clinical Electrophysiology of Vision. We compared their visual fields and electrophysiological results with those of 10 epileptic patients taking other drugs (controls). Most of the cases had some degree of field loss, mostly mild constriction, but in three of them changes were grossly abnormal. None of the controls had field loss.
Ten of the 14 patients taking vigabatrin had a reduced Arden ratio in at least one eye. Nine of the 10 controls had a normal ratio, and the mean ratio was significantly different in the two groups. Various aspects of the electroretinogram were abnormal in at least half of the cases in at least one eye. Mean photopic b-wave amplitudes were significantly reduced in cases compared with controls, as were mean oscillatory potential amplitudes and latencies. Only one control had slightly reduced photopic b-wave amplitude in one eye.
Only two patients taking vigabatrin had symptoms. We continue to collect data on epileptic patients treated with other anti-convulsants as well as vigabatrin, but, even with these numbers, results have been significantly different between the groups. We have found that changes persist long after stopping treatment with the drug. Other groups agree with us that long term treatment with vigabatrin may predispose to irreversible visual changes.[3] The manufacturer suggests that visual field loss may occur in a third of subjects, but we suspect it is even more common. Patients taking vigabatrin must have their visual field monitored; in cases in which this is difficult, electrophysiological results may help. Prescribers should note that special considerations apply to children (see letter from the Vigabatrin Study Group).
I F Comaish specialist registrar in ophthalmology Comaish@aol.com
C Gorman specialist registrar in ophthalmology
N R Galloway consultant ophthalmologist University Hospital, Queen's Medical Centre, Nottingham NG7 2UH
Competing interests: None declared.
[1] Wilton LV, Stephens MOB, Mann RD. Visual field defect associated with vigabatrin: observational cohort study. BMJ 1999;319:1165-6. (30 October.)
[2] Kalviainen R, Nousiainen I, Mantyjarvi M, Nikoskelainen E, Partanen J, Partanen K, et al. Vigabatrin, a gabaergic antiepileptic drug, causes concentric visual field defects. Neurology. 1999;53:922-6.
[3] Miller NR, Johnson MA, Paul SR, Girkin CA, Perry JD, Endres M, et al. Visual dysfunction in patients receiving vigabatrin: clinical and electrophysiologic findings. Neurology 1999;53:2082-7.
Method of estimating prevalence was inappropriate
EDITOR--I was surprised by the low rate of visual field defects associated with vigabatrin found by Wilton et al in their cohort of 7228 patients (0.8%),[1] especially when the manufacturer estimates the prevalence to be between 10% and 30%.[2]
On closer examination, I note that Wilton et al used a questionnaire sent to the general practitioners of patients taking vigabatrin as their main source of data to identify patients with visual field problems. This method is completely inappropriate because, as with glaucoma, the visual field defects associated with vigabatrin are usually totally asymptomatic until the very late stages when the field defects start to encroach on fixation. Thus most patients with visual field defects would be missed if they were to be identified by the presence of symptoms. This is. why recent guidance from the manufacturer, Hoechst Marion Roussel, suggests that patients taking vigabatrin should have their visual fields assessed every six months whenever possible.[2]
Moreover, in a case-control study on vigabatrin and associated field defects that I and others have recently submitted for publication we found a significant relation between total dose of vigabatrin and associated visual field defects. Visual field defects seemed to be especially noticeable after an accumulated dose of 2.5 kg.
Kaykhosrov Manuchehri specialist registrar in ophthalmology Birmingham and Midland Eye Centre, City Hospital, Birmingham B18 7QU harrj@dircon.co.uk
Competing interests: None declared.
[1] Wilton LV, Stephens MOB, Mann RD. Visual field defect associated with vigabatrin: observational cohort study. BMJ 1999;319:1165-6. (30 October.)
[2] Hoechst Marion Roussel. Important information on pharmacovigilance, 27 August 1999. (After review by Committee for Proprietary Medicinal Products.)
Means of selecting patients was misleading
EDITOR--Wilton et al reported an incidence of risk of visual field defects probably or possibly related to vigabatrin of 7.0 per 1000 patients.[1] The study entailed sending a questionnaire to the treating general practitioner and selecting patients referred for visual problems to ophthalmologists.
According to my and others' experience,[2 3] this means of selection might be misleading because most patients with visual field defects associated with vigabatrin are symptom-free. Furthermore, no consensus has yet been reached on the best way of detecting visual field defects early. The manufacturer suggests static perimetry as a standard method,[4] but some uncertainty still exists about how to examine the periphery of the visual field and evaluate the results.
I have examined several patients receiving vigabatrin in whom standard examination tests in the central 30 [degrees] of the visual field showed normal results, although they had defects in the periphery. Furthermore, the deleterious effect of vigabatrin might be overlooked in patients with visual field defects from other causes.
I recently examined the visual field of a 37 year old woman with a complete left sided hemianopsia due to a cerebral haemorrhage at birth. She had taken vigabatrin since 1992 and had been referred for visual field testing, although none of her symptoms had worsened. The test showed a complete left homonymous hemianopsia. However, in addition, a nasal defect was found in the periphery of the visual field in her left eye that might have been associated with vigabatrin treatment. On the right side, the possible loss could not be evaluated owing to overlap with the pre-existing defect in the nasal hemifield. This observation supports the recent recommendation from the manufacturer to consider other antiepilepsy drugs in patients having visual field defects before starting treatment.[4]
In my experience, visual field defects possibly related to vigabatrin treatment are encountered more frequently than the results from the study of Wilton et al imply. In the light of the recent study showing visual field defects in as many as 40% of patients taking vigabatrin,[3] awareness about this possible side effect needs to be increased among doctors and patients. Furthermore, clinical guidance on the examination of visual fields and evaluation of the results is urgently needed.
Anna Midelfart professor Department of Ophthalmology, Faculty of Medicine, Norwegian University of Science and Technology, 7005 Trondheim, Norway anna.midelfart@medisin.ntnu.no
Competing interests: None declared.
[1] Wilton LV, Stephens MOB, Mann RD. Visual field defect associated with vigabatrin: observational cohort study. BMJ 1999;319:1165-6. (30 October.)
[2] Krauss G L, Johnson M A, Miller N R. Vigabatrin-associated retinal cone system dysfunction. Electroretinogram and ophthalmologic findings. Neurology 1998; 50: 614-18.
[3] Kalviainen R, Nousiainen I, Mantyjarvi M, Nikoskelainen E, Partanen J, Partanen K, et al. Vigabatrin, a gabaergic antiepileptic drug, causes concentric visual field defects. Neurology 1999;53:922-6.
[4] Hoechst Marion Roussel. Viktig informasjon vedrorende terapeutisk bredde. Synsfeltinnskrenkning hos pasienter behandlet med vigabatrin. (Letter to Norwegian Physicians, Nordic Area Medical Department, Oslo, Norway, 25 October 1999.)
Author's reply
EDITOR--We thank Comaish et al, Manucheri, and Midelfart for the information they provide on the subject of visual field defects in patients treated with vigabatrin. Our initial study was conducted between 1991 and 1994 to monitor the safety of vigabatrin soon after it was marketed by using an observational cohort technique, prescription event monitoring. With this method previously unrecognised adverse events may be identified in large cohorts of patients--10 033 in our study, among the first patients in England to be prescribed vigabatrin. The signal that visual field defects may be a long latency adverse reaction associated with vigabatrin was identified in this study, confirming the findings of Eke et al, who published the first report of this signal.[1] The manufacturer was notified of our findings in 1995.
Our method was not designed to address the question of the incidence of asymptomatic visual field defects, and we agree with both Comaish et al and Midelfart that visual field defects in patients taking vigabatrin may be more common than the number of such cases identified in our cohort. We do not agree with Midelfart that our selection might be misleading: we did not select the patients as ours was an observational study. Neither do we see eye to eye with Manucheri that our method was completely inappropriate; he is addressing a different question from that which captured our attention. In fact, prescription event monitoring not only was successful in strengthening this safety signal but provided some indication of the incidence of detected visual field defects in a large cohort of patients being treated with vigabatrin in the community at this time (1998).
During 1997 and 1998 letters were published in the BMJ on visual field defects and vigabatrin, some of which recommended regular visual field testing.[2-4] Only a small proportion ([is less than] 2%) of the 4741 patients in our follow up study were referred for visual field testing. Of these 89 cases, 36 (40%) had objective evidence of a visual field defect and 30 (34%) were considered to be probably or possibly associated with vigabatrin use. Our study was conducted during 1998, before the latest guidance from Hoescht Marion Roussel was available.[5]
To understand fully clinically important adverse drug reactions, a range of scientific approaches is usually necessary, including community based observational studies. Evaluation of the data requires understanding of the strengths and weaknesses of each method.
L V Wilton senior research fellow Drug Safety Research Unit, Southampton SO31 1AA
Competing interests: None declared.
[1] Eke T, Talbot JF, Lawden MC. Severe persistent visual field constriction associated with vigabatrin. BMJ 1997; 314:180-1.
[2] Wong ICK, Mawer GE, Sander JWAS. Severe persistent visual field constriction associated with vigabatrin. BMJ 1997;314:1693-4.
[3] Blackwell N, Hayllar J, Kelly G. Severe persistent visual field constriction associated with vigabatrin. BMJ 1997; 314:1694.
[4] Rao GP, Ah Fat F, Kyle G, Leach JP, Chadwick DW, Batterbury M. Study is needed of visual field defects associated with any long term antiepileptic drug. BMJ 1998;317:206.
[5] Hoescht Marion Rousell. Important information on pharmacovigilance, 27 August 1999. (After review by Committee for Proprietary Medicinal Products.
COPYRIGHT 2000 British Medical Association
COPYRIGHT 2000 Gale Group