Gabapentin may cause reversible visual field constrictionBMJ 2006; 332 doi: https://doi.org/10.1136/bmj.332.7551.1193 (Published 18 May 2006) Cite this as: BMJ 2006;332:1193
All rapid responses
The vagaries of perimetry a more likely explanation for apparent field loss associated with gabapentin
Bekkelund and colleagues  report a case of apparent progressive
visual field loss attributed to gabapentin in a 52 year-old woman.
Following withdrawal of gabapentin, the visual field returned to normality
after approximately five years from the initial presentation of symptoms,
attributed to the drug, of transient disturbed vision.
It is regrettable, but understandable, that perimetry was not
undertaken prior to treatment with gabapentin. However, the evidence
presented by Bekkelund and colleagues for visual field loss is, for
several reasons, far from convincing.
The authors do not represent the visual field in the most appropriate
format for interpretation and they do not provide any indication as to the
reliability of the response from the patient during each examination. The
use of the grayscale to represent visual field loss is of limited value in
that the levels of gray are neither age- nor eccentricity-compensated and
this omission can give rise, particularly in the elderly, to apparent
abnormality or to a seemingly more severe defect. Conversely, the
grayscale can either mask, or underestimate field loss, close to fixation,
particularly in younger age groups. It would have been considerably more
informative if Bekkelund and colleagues had presented the outcome at each
stimulus location, within each of the fields, in terms of the probability
of the difference between the measured value and the age-corrected normal
value lying within the statistically normal range.  Such analysis,
readily available from the perimeter printout, also separates generalized
loss from localized loss. The results of each visual examination should
also have been accompanied by an indication of the reliability of the
response from the patient i.e. in terms of the number of incorrect
responses to the false-positive, false-negative and fixation loss catch
Unfortunately, Bekkelund and colleagues did not repeat the visual
field examination at key points over the follow-up i.e. when likely
artefactual contamination of the visual field result could have been
present and/or when the fields exhibited an apparent worsening, as at
stages (a) and (b) in the accompanying figure to their commentary. The
visual field examination is a subjective and difficult task which is
heavily dependent upon the understanding, co-operation and attention of
the patient. The results of any visual field examination exhibit
physiological variability within- and between-examinations. [3-4] The
outcome of the examination can also be influenced by artefacts such as the
learning effect, whereby the measured result improves with repeated
examination within- and between-eyes, within- and between-visits generally
upto the third visit [5-6] but also beyond; the fatigue effect, whereby
the measured result declines over the course of the examination, at any
visit, and is more pronounced for the second eye examined; [7-8] and the
degree of certainty which the patient adopts in defining a ‘seen’
response, a liberal criterion leads to an overestimation of the ‘true’
result whereas a conservative criterion leads to an underestimation of the
‘true’ result.  Fields similar to those illustrated in (b) of the
authors’ figure frequently result from a fatigue effect. Delineation of
the presence of field loss, or of progressive loss, from the variability
in response and from any potential artefact can be very difficult and, at
least in the case of field loss arising from normal tension glaucoma,
generally requires up to three repeat examinations to confirm
deterioration.  Apparent gross defects such as that illustrated in (b)
of the authors’ figure can result in profound visual disability and should
be validated by simple confrontation testing and by observing whether the
patient exhibits impaired mobility in the unfamiliar environment of the
Any apparent abnormality of the field should be evaluated in the
context of the spatial arrangement, i.e. the pattern of loss,
characteristic of the given condition under investigation. The authors
consider the apparent visual field abnormality arising from gabapentin to
be of the type associated with the GABAergic anti-epileptic drug
vigabatrin. However, the appearance of the fields illustrated by Bekkelund
and colleagues does not resemble the field loss attributed to vigabatrin,
namely, a bilateral, symmetrical, peripheral constriction that, by static
perimetry within 30° eccentricity, preferentially affects the nasal part
of the visual field and, which in all but the most severe cases, is
associated with some sparing of the temporal field. [11-12] The field loss
associated with vigabatrin is also accompanied by normal or near normal
visual acuity. To date, whilst either reviewing the fields or conducting
the visual field examination, itself, of over 700 patients exposed to
vigabatrin, I have not found field loss of the type illustrated by
Bekkelund and colleagues to be associated with vigabatrin. Moreover, the
field loss associated with vigabatrin does not improve on withdrawal from
the drug. [13-14]
The authors also failed to find any supporting evidence for the
apparent field loss in terms of electrophysiological abnormality; the
field loss attributed to vigabatrin is accompanied by abnormalities of the
electroretinogram which are indicative of, at least, damage to the cone
The authors did not apply any of the clinical in vivo techniques for
imaging of the retina which could have provided confirmation of the field
loss; the field loss associated with vigabatrin is mostly, if not
exclusively, of retinal origin and is accompanied by a thinning of the
retinal nerve fibre layer. 
Visual field loss does not seem to be a class effect of drugs acting
upon the GABAergic neurotransmitter system; no credible examples of field
loss arising from tiagabine have been reported. Gabapentine is associated
with modest increases in human brain GABA levels. Although the
pharmacology of gabapentin is not well understood, it is unlikely to
resemble either that of vigabatrin or of tiagabine. 
The case presented by Bekkelund and colleagues emphasises the
importance of ensuring that the appearance of the initial fields and/or a
change in the appearance of a subsequent field are representative of the
‘true’ appearance and do not arise from an under performance by the
patient or from artefactual contamination. All unexpected results should
always be confirmed at a repeat examination on another day. ‘True’ field
loss should also be consistent with other clinical findings. Such an
approach, of course, applies to any condition under investigation not just
that of a potential drug toxicity.
1. Bekkeland SI, Lilleng H, Tonseth S. Gabapentin may cause
reversible visual field constriction. British Medical Journal
2. Heijl A, Lindgren G, Olsson J. A package for the statistical
analysis of visual fields. Documenta Ophthalmologica Proceedings Series
3. Flammer J, Drance SM, Zulauf M. Differential light threshold,
short- and long-term fluctuation in patients with glaucoma, normal
controls, and patients with suspected glaucoma. Archives of Ophthalmology
4. Heijl A, Lindgren G, Olsson J. Normal variability of static
perimetric threshold values across the central visual field. Archives of
5. Wood JM, Wild JM, Hussey MK, Crews SJ. Serial examination of the
normal visual field using Octopus automated projection perimetry: evidence
for a learning effect. Acta Ophthalmologica 1987; 65:326-33.
6. Heijl A, Lindgren G, Olsson J. The effect of perimetric experience
in normal subjects. Archives of Ophthalmology 1989;107: 81-6.
7 Heijl A. Time changes of contrast thresholds during automatic
perimetry. Acta Ophthalmologica. 1977;55:696-708.
8 Hudson C, Wild JM, O'Neill EC. Fatigue effects during a single
session of automated static threshold perimetry. Investigative
Ophthalmology and Visual Science. 1994;35:268-80.
9. Kutzko KE, Brito CF, Wall M. The effect of instructions on
conventional automated perimetry. Investigative Ophthalmology and Visual
10. Schulzer M. Errors in the diagnosis of visual field progression
in normal-tension glaucoma. Ophthalmology 1994;101:1589-94.
11. Wild JM, Martinez C, Reinshagen G, Harding GFA. Characteristics
of a unique visual field defect attributed to vigabatrin. Epilepsia 1999;
12. Lawden MC, Eke T, Degg C, Harding GFA, Wild JM. Visual field
defects associated with vigabatrin therapy. Journal of Neurology,
Neurosurgery, and Psychiatry 1999; 67: 716-22.
13. Johnson MA, Krauss GL, Miller NR, Medura M, Paul SR Visual
function loss from vigabatrin - Effect of stopping the drug Neurology
14. Nousiainen I, Mantyjarvi M, Kalviainen R. No reversion in
vigabatrin-associated visual field defects. Ophthalmology 2001; 57:1916-
15. Krauss GL, Johnson MA, Miller NR. Vigabatrin-associated retinal
cone system dysfunction. Electroretinogram and ophthalmic findings.
Neurology 1998; 50:614-8.
16. Harding GFA, Wild JM, Robertson KA, Rietbrock S, Martinez C.
Separating the retinal electrophysiological effects of vigabatrin:
treatment versus field loss. Neurology. 2000; 55:347-52.
17. Wild JM, Robson CR, Jones A, Cunliffe IA, Smith PEM. Detecting
vigabatrin toxicity by imaging of the retinal nerve fibre layer.
Investigative Ophthalmology and Visual Science. 2006; 47:917-24.
18. Sills GJ. The mechanisms of action of gabapentin and pregabalin.
Current Opinion in Pharmacology 2006;6:108-113.
JMW has received honoraria and travel expenses from the manufacturers of vigabatrin.
Competing interests: No competing interests
We read with interest the case report of Bekkelund et al but have a
few points to raise.
Field loss with GABA transaminase inhibitors is a well recognised
phenomenon however the vast majority of case series' have shown that it is
The field loss seen in these patients is normally (but not entirely)
a binasal field loss - arcuate scotoma is very unusual.
Investigation of these patients with GDx VCC scanning laser
polarimetry (Laser Diagnostic Technologies Inc., San Diego, CA USA)
reveals a remarkable loss of retinal nerve fiber layer, with the longer
nerve fibers being affected primarily (hence the binasal appearance of the
scotomata). The advent of this investigation, we feel, is going to
supercede regular perimetry investigations as it is non-invasive, quick,
relatively easy to perform and can be used in children and those patients
with learning difficulties - both groups of which are prevalent in the
population taking vigabatrin. The physical loss of nerve fiber layer is
permenant and cannot be a transient effect.
We feel that the progression and improvement of the fields in the
case report can be attributed to the well recognised learning curve of
field testing - that is to say the more fields a patient performs the
better they seem to become. Before any field test is analysed the
reliablity parameters must be detailed.
Competing interests: No competing interests
I was interested to read Bekkelund et al’s account of an apparent
‘visual field constriction’ related to use of gabapentin (1).
Unfortunately, the authors seem to have overlooked the most common cause
of this type of visual field defect: artefact. Automated testing will
often give this pattern of ‘field defect’ in subjects who have difficulty
concentrating, have trouble following instructions, or are simply learning
to do the test. Usually, this type of artefact is apparent from the
‘reliability indices’ on the printout from the Humphrey Field Analyser
(2). Unfortunately, these data were edited out of Bekkelund’s field
plots, but this learning/inattention artefact can be present even if the
‘reliability indices’ appear to be good. It is common for this type of
‘field defect’ to appear variable and then resolve with repeated testing,
as the subject learns how to carry out the test. A similar report, of
‘reversibility of vigabatrin-induced visual field defect’ is also most
likely to represent learning/inattention artefact (3). Clinicians should
be aware of the limitations of automated testing when screening for visual
1. Bekkelund SI, Lilleng H, Tonseth S. Gabapentin may cause
reversible visual field constriction. Br Med J 2006; 332:1193.
2. Heijl A, Patella VM. Essential Perimetry: The Field Analyzer
Primer, 3rd ed. Dublin, CA: Carl Zeiss Meditec Inc; 2002.
3. Versino M, Veggiotti P. Reversibility of vigabatrin-induced
visual-field defect. Lancet 1999 Aug 7;354(9177):486.
TE was first author of the original paper on vigabatrin and visual field defects (reference 1 in Bekkelund's paper), and has received honoraria and travel expenses from the manufacturers of vigabatrin. He has been a paid adviser on medico-legal issues related to vigabatrin and visual fields.
Competing interests: No competing interests