Abstract
Vigabatrin is considered the drug of choice for infantile spasms and simple and complex partial epilepsy in childhood. Its mechanism of action relies on the irreversible inhibition of γ-aminobutyric acid (GABA) transaminase. Since June 1997 several articles have been published reporting visual field constriction in adult patients on vigabatrin therapy. Recently, 7 pediatric patients, 1 on vigabatrin monotherapy and 6 on add-on therapy with visual field constriction have been described. We have observed 30 pediatric patients with epilepsy (14 boys and 16 girls), ages ranging from 4 to 20 years (mean: 11 years and 2 months) treated with vigabatrin for infantile spasms, simple and complex partial epilepsy, who had never complained of ophthalmologic disturbances. Twenty-one patients underwent complete routine ophthalmologic examination (fundus oculi, visual acuity, intraocular pressure, and visual field tests); 9 children (<6 years old) underwent only fundus examination, because collaboration was lacking. We report on 4 children showing constriction of visual field, prevailing in nasal hemifield. In 1 child, visual abnormalities were stable even 10 months after vigabatrin discontinuation, while in another a greater improvement was observed 5 months after discontinuation. The possible mechanisms have been discussed and the cone dysfunction, connected with GABA augmentation in the outer retina, has been outlined. We suggest a possible protocol to control visual abnormalities in epileptic children.
Vigabatrin has been licensed as the drug of choice for infantile spasms and, as an add-on therapy, for other types of childhood epilepsies, mainly simple and complex partial epilepsy. Vigabatrin is a synthetic derivative of γ-aminobutyric acid (GABA), which exerts its antiepileptic action by irreversibly inhibiting GABA transaminase, the enzyme responsible for its catabolism. As a consequence there is an increase, in a dose-dependent manner, of the concentration of GABA, at a synapse level, in the brain and in the retina.1,2 In the retina GABA is located in horizontal, interplexiform, and amacrine cells.2 Although mainly found in the inner retina, GABA has also been implicated as a regulator of cone synaptogenesis located in the outer retina of neonatal rabbits.2 Experimental studies of neurotoxicity indicate that, in some animal species, neuronal damage is present and seems reversible on discontinuation of vigabatrin treatment. This toxicity is characterized by intramyelinic edema and brain vacuolation occurring in rodents and dogs.2 Long-term studies using clinical examinations, magnetic resonance imaging (MRI) and multimodal-evoked potentials showed normal features suggesting that vigabatrin is safe in human beings. However visual abnormalities have been documented.2
Since June 1997 several articles have been published reporting visual field constriction in adult patients on vigabatrin therapy, with an incidence up to 28%.3–8 Recently, 7 pediatric patients, 1 on vigabatrin monotherapy and 6 on add-on therapy, with visual field constriction, have been described.9,10 We report on 4 additional children.
MATERIALS AND METHODS
We have observed 30 pediatric patients with epilepsy (14 boys and 16 girls) ages ranging from 4 to 20 years (mean: 11 years and 2 months) treated with vigabatrin for simple and complex partial epilepsy, who had not complained of any ophthalmologic disturbances. Twenty-one patients underwent complete routine ophthalmologic examination (fundus oculi, visual acuity, intraocular pressure, and visual field tests); 9 children (<6 years old) underwent only fundus examination, because of lack of collaboration. Four patients of the study group underwent visual field tests, in a period ranging from 6 months to 36 months after vigabatrin discontinuation.
Visual field tests using Goldmann perimetry were first performed on the children as outpatients by other ophthalmologists; the second visual field tests were conducted in our ophthalmologic department to evaluate any modification of the previous results by means of computerized static perimetry, and the data obtained were analyzed by the same ophthalmologist (B.B.).
The complete routine ophthalmologic examination consisted of visual acuity determination and slit-lamp examination with and without vital stain (fluoresceine). Intraocular pressure was measured using a Goldmann applanation tonometer. Anterior segment biomicroscopy and applanation tonometry were performed before mydriasis that was obtained by local instillation of cyclopentholate 1% (1gtt × 3 in 45′). Retinal biomicroscopy, indirect ophthalmoscopic examination of retina, and visual acuity determination were performed in cycloplegic mydriasis. Two perimetric procedures were conducted: 1. Goldmann kinetic perimetry and 2. automatic static perimetry (Humphrey Field Analyzer [HFA], Allergan, 32–2, full threshold, single-field analysis). The blind spot detection and the peripheral, medium and central isopter were tested in kinetic perimetry. In static perimetry, the single-field analysis was performed using the Statpac program which provided, in addition to the examination parameters and an evaluation of the patients' cooperation, a printout of the absolute sensitivity values, including double measurements and gray scale printout. The additional information was grouped into 3 blocks: total deviation, pattern deviation, and the calculation of global indices. Test reliability indices, including fixation losses, false-positive errors, false-negative errors and questions asked, were good in all the examined patients.
RESULTS
In 21 of the 30 examined patients the following results were found: 16 patients had fundus oculi and perimetry with normal results; 1 patient revealed right shaded optic disk and normal visual field examination; 4 patients showed narrowness of visual field, prevailing as nasal hemifield.
Patient 1
An 18-year-old boy previously neurologically and psychologically normal, started to have complex partial epilepsy at 13 years of age; he was treated with sodium valproate (15 mg/kg/day) for 2 years; vigabatrin (40 mg/kg/day) was added at the age of 15 years, because of reappearance of the seizures. Cerebral MRI, performed at 16 years, was normal. At 18 years of age ophthalmological reexamination showed normal visual acuity, intraocular pressure, and fundus oculi. The first visual field examination, performed 4 years after vigabatrin treatment, revealed bilateral severe concentric constriction prevailing in the left eye; visual field was reevaluated 2 months later, during vigabatrin discontinuation, confirming the severity of the abnormalities previously described (Fig 1). At present the drug has been stopped.
Patient 1. Gray scale and sensitivity map (dB) in a Statpac (Allergan) single-field analysis on the HFA, program 30–2 in a 19-year-old boy treated with vigabatrin. Bilateral concentric absolute scotoma more pronounced in the right than in the left eye prevailing in nasal hemifield.
Patient 2
A 10-year-old previously healthy girl, had, at the age of 5 years and 8 months, complex partial seizures. Neurologic and psychologic evaluations were normal. Vigabatrin was started with a dosage of 50 mg/kg/day. Cerebral MRI, performed 2 months after vigabatrin treatment had started, was also normal. After 3 years, carbamazepine was added because of poor seizure control; soon after vigabatrin was tapered over the following months and then stopped. Ophthalmologic examination showed normal visual acuity, intraocular pressure, and fundus oculi. The girl underwent a visual field test 6 months after vigabatrin discontinuation, and the examination revealed bilateral peripheral visual field constriction. These data were also confirmed by the second visual field test, performed 4 months later, 10 months after drug discontinuation (Fig 2).
Patient 2. Gray scale and sensitivity map (dB) in a Statpac (Allergan) single-field analysis on the HFA, program 30–2 in a 9-year-old girl 10 months after vigabatrin discontinuation. Bilateral peripheral concentric visual field defect more severe in the left than in the right eye prevailing in nasal hemifield.
Patient 3
A 12-year-old boy had been treated elsewhere with phenobarbital at 8 months of age because of convulsions. At the age of 3 years he started to have complex partial epilepsy treated first with carbamazepine (20 mg/kg/day), then 1 year later with vigabatrin (70 mg/kg/day) as add-on therapy because of poor seizure control. We examined the boy at 10 years of age. Neurologic and psychologic evaluations were normal; MRI of the brain was also normal. Ophthalmologic examination showed normal visual acuity, intraocular pressure and fundus oculi. Visual field examination, performed 8 years after vigabatrin treatment, showed bilateral peripheral visual field constriction and left enlargement of blind spot, which were confirmed by a second visual field test, performed 4 months after drug discontinuation.
Patient 4
A 12-year-old healthy boy with complex partial epilepsy, which began when he was 10 years and 6 months old, was treated with vigabatrin 50 mg/kg/day for 2 years and 3 months. Neurologic and psychologic evaluations were normal; cerebral MRI was also normal. Ophthalmologic examination showed normal visual acuity, intraocular pressure, and fundus oculi. The visual field test, performed 2 years after starting vigabatrin, showed bilateral peripheral constriction more pronounced in the left eye (Fig 3). Discontinuation of vigabatrin was tapered over the following months and then stopped. A second visual field test, performed 1 month after drug discontinuation, did not show any modification. A third visual field examination, performed 5 months after treatment was stopped, revealed a significant improvement (Fig 4).
Patient 4. Gray scale and sensitivity map (dB) in a Statpac single-field analysis on the HFA, program 30-2 in a 12-year-old boy. Bilateral peripheral concentric visual field defect more severe in the left than in the right eye prevailing in nasal hemifield.
Patient 4. Gray scale and sensitivity map (dB) in a Statpac single-field analysis on the HFA, program 30-2 in a 12-year-old boy 5 months after vigabatrin discontinuation. Significant improvement of visual field defect.
DISCUSSION
Vigabatrin is an aminobutyrate-aminotransferase antagonist and acts as a GABA analog; it is a transmitter working at different sites of the postreceptoral brain and retina, and it also modulates the cone synaptogenesis in both the outer and inner retina of neonatal rabbits.2 The association between visual field constriction and vigabatrin treatment in adult epileptic patients has been signaled.3–8 Recently visual field abnormalities have also been detected in 7 pediatric patients undergoing vigabatrin monotherapy or add-on therapy.9,10
In the examined group of 21 children with epilepsy subjectively asymptomatic of visual complaints, a complete routine ophthalmologic examination was performed twice at different intervals, to identify possible retinal damage; we report on 4 patients selected among this population. Patient 1 showed the more severe visual field constriction, patient 2 had the longest follow-up and the peripheral restriction was persistent 10 months after vigabatrin was withdrawn. Patient 3 showed, in addition to the visual field limitation, unilateral enlargement of the blind spot, confirmed 4 months later at visual field test control. Patient 4, treated with vigabatrin monotherapy, showed bilateral peripheral visual field constriction still persistent at the reevaluation performed 1 month after vigabatrin discontinuation, nevertheless the patient demonstrated greater improvement at the examination performed 5 months later. The presence of the enlargement of the blind spot has been attributed to different antiepileptic drugs. Nevertheless in patient 3 the monolateral appearance has been considered the expression of an anatomic variant.
The mechanism by which the drug produces this subtle retinal involvement is still unknown and it has not been established whether retinal involvement decreases after drug discontinuation.11 To our knowledge the follow-up has been performed in an adult patient,7 in 3 of the 5 recently reported children10 and in our 4 patients. Three of the 4 examined patients showed persistence of retinal damage up to 2 years and 6 months after vigabatrin discontinuation, except for patient 4 in whom a significant improvement was observed.
Whether the start of vigabatrin treatment at an early age is an important factor in the severity of visual field involvement or not cannot be completely established. From our data it appears that the most severe damage was not observed in the youngest patient (patient 2); moreover, retinal involvement did not correlate with the longest period of treatment (patient 3).
Among different etiologic causes, disseminated intravascular coagulation has been suggested as a consequence of prolonged and recurrent convulsions.8 In our patients, this mechanism can be ruled out because seizures rarely recurred.
The density of retinal ganglion cells has been considered a critical factor.12 The possible mechanism for cone system dysfunction has been speculated and the augmentation of GABA by means of vigabatrin with subsequent abnormal integration of receptive fields has been hypothesized. This could yield clinically significant photopic visual loss in the relatively low-density cone population in the peripheral retina.12 GABA mediates lateral inhibition in the retina and that may result in restriction of the visual field. Moreover, GABA plays a role in the regulation of the horizontal cell coupling, and its accumulation can be found in amacrine cells.7 Electroretinogram (ERG) has been used to measure the cone function that has resulted as being abnormal in vigabatrin-treated patients.2 These changes have also been considered a normal response that may happen in vigabatrin-treated patients because of a physiologic response to an elevated level of GABA.13–15 Nevertheless, different antiepileptic drugs (carbamazepine and phenytoin) have been reported to modify ERG response.2
The observation that visual field defect is also present in patients not treated with vigabatrin raises the hypothesis of a possible GABA-mimetic mechanism of other drugs.10,16 In addition, the visual field constriction may also result from other conditions such as opaqueness of dioptric lenses (corneal leucoma and cataract), severe visual defects, glaucoma, inflammatory, vascular and degenerative diseases of retina, tracts, and optic nerves.
A single patient with concentric visual field restriction, on valproate and lamotrigine treatment for absence epilepsy, was reported.10 In such a type of seizure, it is difficult to exclude the presence of clinical or subclinical manifestations, which can be triggered by light while performing visual field tests. The possible role of epilepsy in the appearance of retinal cone dysfunction needs further study.17
Vigabatrin is a successful drug in the treatment of infantile spasms in patients with tuberous sclerosis, simple and complex partial epilepsy. Nevertheless, when prescribing vigabatrin the fundamental issue is the evaluation of the risk/benefit ratio. A possible protocol could plan a visual field test before starting vigabatrin in order, not only to avoid the deterioration of a preexisting visual deficit, but also to obtain a baseline that could be helpful for the evaluation of subsequent modification during vigabatrin treatment. The visual field assessment should be repeated 3 to 4 times a year. Because the Humphrey technique is able to reveal the dysfunction of peripheral retina only when >30% of retinal fibers is already involved,18 an additional complete ophthalmologic evaluation is recommended.
Footnotes
- Received September 29, 1999.
- Accepted January 18, 2000.
Reprint requests to (P.I.) Pediatric Neurology, Department of Pediatrics, La Sapienza University, Viale Regina Elena 324, 00161 Roma, Italy. E-mail: iannetti{at}iol.it
- GABA =
- γ-aminobutyric acid •
- MRI =
- magnetic resonance imaging •
- HFA =
- Humphrey Field Analyzer •
- ERG =
- electroretinogram
REFERENCES
- Copyright © 2000 American Academy of Pediatrics
In this issue
