SEARCH

SEARCH BY CITATION

Keywords:

  • vigabatrin;
  • visual field defect;
  • adverse events;
  • vision testing;
  • ocular coherence tomography;
  • taurine

Abstract

  1. Top of page
  2. Abstract
  3. Introduction
  4. History of vigabatrin-associated pVFDs
  5. Vision testing for vigabatrin-treated patients
  6. Potential confounding factors
  7. Mechanism of vigabatrin-associated pVFDs and taurine link
  8. Role of taurine in the retina
  9. Patient registry implemented to monitor vision function in vigabatrin-treated patients
  10. Conclusions
  11. Conflicts of interest
  12. Acknowledgments
  13. References

Plant GT, Sergott RC. Understanding and interpreting vision safety issues with vigabatrin therapy. Acta Neurol Scand: 2011: 124 (Suppl. 192): 57–71. © 2011 John Wiley & Sons A/S.

Vigabatrin is an irreversible inhibitor of γ-aminobutyric acid (GABA) transaminase. It is effective as adjunctive therapy for adult patients with refractory complex partial seizures (rCPS) who have inadequately responded to several alternative treatments and as monotherapy for children aged 1 month to 2 years with infantile spasms. The well-documented safety profile of vigabatrin includes risk of retinopathy characterized by irreversible, bilateral, concentric peripheral visual field constriction. Thus, monitoring of visual function to understand the occurrence and manage the potential consequences of peripheral visual field defects (pVFDs) is now required for all patients who receive vigabatrin. However, screening for pVFDs for patients with epilepsy was conducted only after the association between vigabatrin and pVFDs was established. We examined the potential association between pVFDs and epilepsy in vigabatrin-naïve patients and attempted to identify confounding factors (e.g., concomitant medications, method of vision assessment) to more accurately delineate the prevalence of pVFDs directly associated with vigabatrin. Results of a prospective cohort study as well as several case series and case reports suggest that bilateral visual field constriction is not restricted to patients exposed to vigabatrin but has also been detected, although much less frequently, in vigabatrin-naïve patients with epilepsy, including those who received treatment with other GABAergic antiepileptic therapy. We also reviewed published data suggesting an association between vigabatrin-associated retinal toxicity and taurine deficiency, as well as the potential role of taurine in the prevention of this retinopathy.


Introduction

  1. Top of page
  2. Abstract
  3. Introduction
  4. History of vigabatrin-associated pVFDs
  5. Vision testing for vigabatrin-treated patients
  6. Potential confounding factors
  7. Mechanism of vigabatrin-associated pVFDs and taurine link
  8. Role of taurine in the retina
  9. Patient registry implemented to monitor vision function in vigabatrin-treated patients
  10. Conclusions
  11. Conflicts of interest
  12. Acknowledgments
  13. References

Vigabatrin, an irreversible inhibitor of γ-aminobutyric acid (GABA) transaminase, is effective as adjunctive therapy for adult patients with refractory complex partial seizures (rCPS) who have responded inadequately to several alternative treatments and as monotherapy for pediatric patients aged 1 month to 2 years with infantile spasms (IS) (1–5). Vigabatrin is available in more than 50 countries, with initial approval in 1989 in the UK and more recent approval in 2009 in the United States (6).

Partial-onset seizures occur in >60% of patients with epilepsy and are resistant to pharmacotherapy, including combination regimens, in approximately one-third of cases (7). In clinical trials of adults with rCPS, the addition of vigabatrin (3–6 g/day) resulted in a significant decrease in the frequency and severity of seizures (≥ 50% seizure reduction) in 43–51% of patients and resulted in complete seizure freedom in 5–12% of patients (6). Generally, response to vigabatrin therapy for rCPS was observed within the first 12 weeks of treatment. See the articles in this supplement by Faught (3) and by Ben-Menachem & Sander (4) for a more detailed discussion of clinical trials of vigabatrin for rCPS.

Infantile spasms represent <4% of childhood epilepsies and are associated with a high mortality rate. Adrenocorticotropic hormone controls spasms in many patients but is associated with a high incidence of serious adverse effects (8). In clinical trials of patients with IS, treatment with vigabatrin resulted in rapid cessation of seizures (within approximately 14 days or less) in 36–54% of patients (8–11) and has been reported to induce complete absence of spasms at 12–14 months’ follow-up in 76% of infants (9, 12). See the article in this supplement by Carmant (5) for a more detailed discussion of clinical trials of vigabatrin for IS.

Despite its proven efficacy, vigabatrin has been limited in use by the associated risk of retinopathy, characterized by irreversible, bilateral, and concentric peripheral visual field constriction (13). Screening for peripheral visual field defects (pVFDs) was generally not performed in patients with epilepsy before the discovery of vigabatrin-associated pVFDs, which was first reported in 1997, 9 years after licensed availability in many countries (14).

Therefore, it is useful to examine whether there is any association between pVFDs and epilepsy for vigabatrin-naïve patients and to identify any potential confounding factors such as concomitant medications and method of vision assessment. Such information would be helpful to more accurately delineate the prevalence of pVFDs directly associated with vigabatrin. We will also examine possible mechanisms of pVFDs associated with vigabatrin and the potential role of taurine in prevention of this adverse effect.

History of vigabatrin-associated pVFDs

  1. Top of page
  2. Abstract
  3. Introduction
  4. History of vigabatrin-associated pVFDs
  5. Vision testing for vigabatrin-treated patients
  6. Potential confounding factors
  7. Mechanism of vigabatrin-associated pVFDs and taurine link
  8. Role of taurine in the retina
  9. Patient registry implemented to monitor vision function in vigabatrin-treated patients
  10. Conclusions
  11. Conflicts of interest
  12. Acknowledgments
  13. References

Peripheral visual field defects associated with vigabatrin are characterized by a bilateral, concentric peripheral constriction of the visual field. The defect ranges from mild to marked severity and generally appears on standard visual field plots to be more pronounced nasally (Fig. 1). However, it should be noted that the temporal field has a greater peripheral extent than does the nasal field. Thus, if the visual field is constricted by the same amount throughout the entire circumference, the loss will encroach more into the central field nasally than temporally. The defect has a slow onset, with the majority of cases in children and adults having been reported after ≥ 1 year of treatment (13). The earliest well-documented case of vigabatrin-associated pVFD reported in the literature began after 4 months of treatment (16). In a multinational study that included 393 evaluable vigabatrin-exposed patients, the earliest detection of a confirmed pVFD was after 9 months of treatment in adults (mean duration of vigabatrin exposure of 4.8 years) and after 11 months in children (mean duration of vigabatrin exposure of 5.5 years; Lundbeck Inc., Deerfield, IL, USA, data on file). In infants, characteristic electrophysiologic abnormalities associated with vigabatrin-associated pVFDs have been detected as early as 3 months after starting therapy but usually develop after 1 year of therapy (13).

image

Figure 1.  Typical examples of bilateral visual field defects from (A) patient exposed to vigabatrin, (B) patient exposed to alternative γ-aminobutyric acid (GABA)-ergic antiepileptic drug, and (C) patient never exposed to a GABAergic drug. Reprinted from Epilepsy & Behavior, Vol. 16, Gonzalez P, Sills GJ, Parks S, et al. (15), Binasal visual field defects are not specific to vigabatrin, p521–526, Copyright 2009, with permission from Elsevier.

Download figure to PowerPoint

The first published report of a pVFD associated with vigabatrin appeared in 1997. Eke et al. (14) reported three cases of severe persistent peripheral visual field loss in patients who had been receiving vigabatrin for >2 years. This was 9 years after licensing of vigabatrin and approximately 15 years after the first clinical investigation of vigabatrin in humans (13, 14, 17). The delay in recognition of this retinopathy has been attributed to the fact that pVFDs are generally asymptomatic with mild to moderate field losses, except in a small subset of patients (<5%) for whom the defect can be severe enough to interfere with daily activities (13, 18–21), as well as the fact that there has been no motivation to routinely measure visual fields in patients with epilepsy, or indeed to carry out any form of ophthalmologic examination.

Subsequent to the first published case reports in 1997, several reports have described pVFDs in vigabatrin-treated patients (18, 22–26). The reported wide-ranging prevalence of 14–92% was based on the results of systematic retrospective (18, 25–32) and prospective (33) analyses. One study also reported decreased visual acuity, ranging from 20/25 to 20/60, in association with pVFDs in some patients, but no pretreatment baseline visual acuities were recorded for these patients, refractions were not performed, and several patients had mild cataracts (34). Potential reasons for the variable prevalence rates include differences in the extent of drug exposure, variability in perimetric methods, unblinded assessments, relatively small sizes of the cohorts, retrospective nature of many studies, and sampling bias arising from the asymptomatic nature of the defect (19). The better evidence in the literature places the clinical study prevalence range at 30–50% (2, 33–37). Other sources (e.g., registry data and post-marketing surveillance) may indicate much lower rates overall.

Maguire et al. (38) performed a systematic review of observational studies in patients with partial epilepsy to ascertain the magnitude of vigabatrin-associated visual field loss (VAVFL) and to identify clinical predictors of risk. This review included 32 studies in which 1,678 patients were exposed to vigabatrin and 406 were not. A total of 738 (44%) vigabatrin-exposed patients had visual field loss compared with 30 (7%) vigabatrin-naïve patients. A random-effects estimate for the percentage of vigabatrin-exposed adults with pVFDs was 52% (95% confidence interval [CI]: 46–59%). The percentage was lower for children: 34% (95% CI: 25–42%). In the 16 studies with a control group, the relative risk for field loss in vigabatrin-exposed patients was 4.0 (95% CI: 2.9–5.5). Prevalence of pVFDs was greater for patients with greater mean cumulative dosage and older age.

The systematic review (38) of observational studies by Maguire showed that visual field loss occurs in approximately one-half of adults and one-third of children exposed to vigabatrin for the treatment of partial epilepsy. In the controlled studies, more conservative estimates of one-third of adults and less than one-fifth of children with VAVFL were identified. Therefore, it is probable that some patients in the uncontrolled studies had visual loss that was unrelated to vigabatrin treatment. For infants with IS receiving vigabatrin, studies have established a pVFD incidence of approximately 16% and prevalence of approximately 40% based on the highly sensitive electroretinogram (ERG) 30-Hz flicker amplitude and an incidence and prevalence of approximately 4% and 20%, respectively, based on cone B-wave abnormalities (35, 39–42).

Wild et al. (36) conducted a prospective comparative, open-label, multinational study (Study 4020) in 735 children (≥9 years) and adults with refractory partial epilepsy to further investigate the prevalence of vigabatrin-associated pVFDs, better characterize the defect, and determine risk factors. A preliminary analysis of 432 evaluable patients demonstrated that the incidence of pVFDs was greatest in those exposed to vigabatrin for ≥ 6 months who continued to receive the drug at study entry (Group 1) compared with those previously treated with vigabatrin for ≥ 6 months who had discontinued the drug for ≥ 6 months prior to study entry (Group 2) (19).

The final analysis and results are based on a conclusive outcome to the visual field examination of 524 patients, 386 of whom had been exposed to vigabatrin (36). As in the preliminary analysis (19), the incidence of pVFDs was greatest in Group 1 (26.3% of patients 8–12 years old, 43.3% of patients >12 years old) compared with Group 2 (14.9% and 24.5% for these age groups, respectively). In a control group of 186 patients who were never exposed to vigabatrin, one case (0.5%) of visual field loss was reported. Risk of developing a pVFD increased significantly with treatment duration (odds ratio [OR]: ≤ 15.2 [95% CI: 4.4–51.7] for duration of 10 or more years), and with mean daily dosage (OR: ≤ 26.4 [95% CI: 2.4–291.7] for >3 g), and male sex (OR: 2.51 [95% CI: 1.5–4.1]).

Sergott et al. (43) conducted an analysis of data from a subset of patients in Study 4020 who underwent Goldmann kinetic perimetry (= 341) to affirm the findings of that study (which was based primarily on static perimetry) through a more objective, quantitative methodology. Of vigabatrin-exposed patients (= 258), 17% had moderate pVFDs (30–60° monocular temporal field retained) and 2% had severe defects (<30° monocular temporal field retained), compared with a rate of 4% moderate and 1% severe pVFDs for vigabatrin-naïve patients (Figs 2 and 3). The severity categories chosen for this analysis were intended to reflect the degree of resulting functional limitation (e.g., the limit for having a driver license in many jurisdictions [60° of monocular temporal field] was chosen as the dividing line between mild and moderate severity) (43).

image

Figure 2.  Severity of visual field defect at last kinetic perimetry in (A) vigabatrin-exposed and (B) vigabatrin-naïve patients. Unimpaired: >80° monocular temporal field retained; mild: 60–80° monocular temporal field retained; moderate: 30–60° monocular temporal field retained; severe: <30° monocular temporal field retained. Measurements are of largest isopter tested at final Goldmann perimetry. Reprinted from Epilepsy & Behavior, Vol. 92, Sergott RC, Bittman RM, Christen EM, et al. (43), Vigabatrin-induced peripheral field defects in patients with refractory partial epilepsy, p170–176, Copyright (2010), with permission from Elsevier.

Download figure to PowerPoint

image

Figure 3.  Patients exposed to vigabatrin experienced a mean reduction in their visual fields to 71° (normal = 90°) in Study 4020. Vigabatrin: 2008 update,Willmore LJ, Abelson MB, Ben-Menachem E, et al. (6) Copyright © 2009 Epilepsia. Reproduced with permission of John Wiley & Sons, Inc.

Download figure to PowerPoint

Studies of possible risk factors for vigabatrin-associated pVFDs are characterized by widely varying results because of several factors: retrospective nature of studies, small numbers of patients studied, lack of patient symptoms, long latency until condition was recognized, and variety of visual testing methods used (19). Several (but not all) studies have shown an association with male sex (1.5- to 2-fold greater risk vs female patients) (18, 19), increased dosage and duration of therapy (16, 19, 31, 44, 45), and cumulative exposure (16, 46). However, some patients have long-term exposure (>10 years) without developing a pVFD (47).

In an extensive review of studies of the frequency of visual field constriction and cumulative vigabatrin dosage, Lawden (21) concluded that despite wide variation, particularly at lesser cumulative dosages, there was rough agreement between studies that the frequency of visual field constriction increased with increasing cumulative vigabatrin dosage, that pVFDs were rare (though not unknown) at cumulative vigabatrin exposure of <1 kg, and that cumulative vigabatrin exposure of >5 kg did not seem to be associated with further increase in risk. The interim analysis of Study 4020 indicated that treatment with vigabatrin for more than 5 years was associated with 14.2-times the risk of developing a pVFD than treatment for less than 1 year (19). Based on the prevalence, distal appearance, lack of progression (in most cases), and lack of remission of pVFDs after discontinuing vigabatrin and the fact that many patients never develop vigabatrin-related pVFDs despite long-term exposure at high dosages, several investigators have proposed that the pathophysiology of the injury is an idiosyncratic adverse effect as opposed to a strict dosage- or duration-dependent toxicity (30, 35, 48, 49).

Studies evaluating reversibility, improvement, and progression of vigabatrin-associated pVFDs have yielded conflicting results arising from lack of agreement on specific criteria used to classify pVFDs and difficulty in differentiating real change from perimetry test variability. On average, progression of a pVFD is <2° per year from the temporal visual field and <1° per year in the nasal field (2). Overall, available data suggest that pVFDs do not typically progress or regress to a clinically significant degree after discontinuation of therapy (13).

Several small prospective studies (with up to 18 months of follow-up) suggest that most patients with detectable pVFDs who elect to continue therapy with vigabatrin do not experience progressive worsening of the defect (49, 50). A recent case report described a 36-year-old woman receiving vigabatrin 3 g/day as add-on therapy for partial epilepsy in whom serial assessments with Goldmann kinetic perimetry demonstrated rapid deterioration in visual fields between assessments 1 year apart after 10 years of vigabatrin therapy and earlier stable visual fields (51). Of note, a few reports have suggested some improvement in pVFDs, especially in children (27, 52, 53), although this may represent a learning effect or variability in test measurement as patients become more accustomed to repeated visual field testing (54). Indeed, such an improvement may bring into question whether the initial pVFD had been associated with vigabatrin therapy at all (13).

Notably, pVFDs aside, vigabatrin is generally well-tolerated, with an adverse event profile similar to that of other antiepileptic drugs (AEDs) (2, 55). Add-on vigabatrin was well-tolerated or extremely well-tolerated by 72.4% of patients in a multicenter, 36-week, double-blind, randomized study in patients with refractory partial epilepsy (56). The most common adverse effects were headache, fatigue, dizziness, and drowsiness. Similarly, vigabatrin was well-tolerated and had a better adverse event profile compared with carbamazepine in an open-label, long-term comparative study of vigabatrin vs carbamazepine in children with newly diagnosed partial seizures (57). The most frequent adverse events were irritability/excitability in 15.8% of patients and weight gain in 26.3% of patients. See the article in this supplement by Walker & Kälviäinen (58) for a more detailed discussion of non-visual adverse events with vigabatrin.

Vision testing for vigabatrin-treated patients

  1. Top of page
  2. Abstract
  3. Introduction
  4. History of vigabatrin-associated pVFDs
  5. Vision testing for vigabatrin-treated patients
  6. Potential confounding factors
  7. Mechanism of vigabatrin-associated pVFDs and taurine link
  8. Role of taurine in the retina
  9. Patient registry implemented to monitor vision function in vigabatrin-treated patients
  10. Conclusions
  11. Conflicts of interest
  12. Acknowledgments
  13. References

Early detection of a vigabatrin-associated pVFD through periodic visual function testing during treatment is necessary to continually assess the benefit–risk balance of vigabatrin treatment. This approach should improve early detection of a pVFD when it is asymptomatic to allow for appropriate treatment modifications to minimize retinal damage (13).

Visual field testing in patients with epilepsy can be challenging, however, because of behavioral and cognitive limitations and the fact that perimetry used to diagnose pVFDs requires active patient cooperation. Approximately 25% of patients cannot be tested by perimetric techniques (19, 59, 60). Experience from two Finnish studies indicates that up to 25% of initial positive results of visual field testing in children change to negative when perimetry is repeated (18, 61). In addition, perimetric methodology (static vs kinetic) differs between centers, inconclusive investigations are frequent, and reproducibility is poor (13, 17). Recent data suggest that up to 20% of perimetry tests for a vigabatrin-associated pVFD may produce false-positive results. These confounding factors associated with perimetric testing are one reason for the considerable variation in reported VAVFL prevalence between studies (range, 14–92%) (13, 17).

Alternative methods are available if patients have difficulties in reliably performing perimetry (i.e., confrontation testing, electrophysiologic testing, and/or imaging of the peripapillary retinal nerve fiber layer [RNFL] and macular anatomy and thickness with optical coherence tomography [OCT]). Electrophysiologic results can be affected by the specific recording techniques used, presence and severity of pVFDs, whether patients are receiving vigabatrin at the time of recording, and concomitant treatments (some of which are known to also affect the ERG) (21). See the article by Sergott & Westall (62) in this supplement for a more detailed discussion of vision testing methodologies.

A promising technique for identification of vigabatrin-associated visual abnormalities involves imaging of the retinal nerve fiber layer using OCT. The characteristic pattern of the visual defect consists of thinning of the nasal quadrant but sparing of the temporal quadrant — a finding that may precede visual field loss. Using OCT to measure the thickness of the RNFL, Lawthom et al. (60) found this pattern in all 11 patients with confirmed vigabatrin-associated field deficits, four of 15 vigabatrin-treated patients with normal fields, and none of the 13 controls (patients with epilepsy treated with non-GABAergic antiepileptic monotherapy, all of whom had normal fields). In a study of 125 vigabatrin-exposed patients with epilepsy for whom both visual field data (through Goldmann kinetic perimetry) and OCT data were available, Clayton et al. (63) observed a strong association between RNFL thickness and visual field size (Fig. 4). Abnormal RNFL thinning was most frequently seen in the superior and inferior quadrants (29.7% and 31.0% of patients, respectively), followed by the nasal quadrant (12.5%) and the temporal quadrant (9.2%). Compared with a group of 90 healthy controls, the average RNFL thickness and the thickness of each of the 90° quadrants were significantly thinner in the vigabatrin-exposed patients.

image

Figure 4.  Optical coherence tomography (OCT) report of retinal nerve fiber layer (RNFL) thickness taken from two vigabatrin-exposed individuals. (A) OCT report from a vigabatrin-exposed patient with normal visual fields showing normal RNFL thickness with an average RNFL thickness of 91 μm. The average RNFL thickness and RNFL thickness in each of the 90° quadrants are shown in green as they fall within the ≤95th to ≥5th percentile of the normal distribution percentiles provided by the manufacturer’s inbuilt database. (B) OCT report from a patient with severe vigabatrin-associated visual field loss showing RNFL thinning with an average RNFL thickness of 55 μm. The average RNFL thickness and the RNFL thickness in the superior, nasal, and inferior quadrants are shown in red as they fall below the 1st percentile of the normal distribution percentiles provided by the manufacturer’s inbuilt database. T, temporal; S, superior; N, nasal; I, inferior; TSNIT, temporal, superior, nasal, inferior, temporal; SUP, superior; TEMP, temporal; NAS, nasal; INF, inferior. Retinal nerve fiber layer thickness in vigabatrin-exposed patients, Clayton LM, Dévilé M, Punte T, et al. (63) Copyright © 2011 Annals of Neurology. Reproduced with permission of John Wiley & Sons, Inc.

Download figure to PowerPoint

This pattern of nerve fiber layer thinning is in accordance with the pattern of VAVFL reviewed earlier. It took some time for clinicians to appreciate that nerve fiber layer thinning was a potential marker of VAVFL, because the pattern of nerve fiber loss is the converse of what is usually observed with neurologic diseases, for which there is a tendency for the central visual field projection to be more affected (e.g., multiple sclerosis and inherited optic neuropathies) (64–66). This central field projection in the retina is often referred to as the papillomacular bundle, but this term is ambiguous and the anatomically more precise term centrocecal projection may be preferred. Thus, clinicians are accustomed to observing nerve fiber layer thinning on funduscopy predominantly affecting the central retina and the retinal region between the macula and the optic disk. In cases of VAVFL, this situation is reversed, and the thinning affects the projections entering the optic disk superiorly and inferiorly in the arcuate bundles. These are the thickest portions of the RNFL, and, therefore, relatively greater loss may be necessary in this instance, rather than when the central projection is affected, before the change becomes clinically apparent. OCT can detect changes long before they are clinically apparent.

Moreover, this pattern of RNFL thinning is exactly what would be expected if retinal ganglion cells subserving the peripheral retina were being lost in a concentric manner. An exception according to the study of Clayton et al. (63) seems to be the projection entering the optic disk nasally. These are the optic nerve fibers of the sparse retinal ganglion cells subserving a region of the peripheral visual field extending temporally from the blind spot in a horizontal “V-shaped” sector.

A prospective, open-label, Phase IV study (ClinicalTrials.gov identifier: NCT01278173) is underway to evaluate changes in visual fields by means of automated static perimetry and changes in retinal structure by means of spectral domain optical coherence tomography (SD–OCT) in patients with rCPS for whom vigabatrin treatment is planned (67). The study is expected to validate SD–OCT as a potentially more robust method for monitoring retinal changes and for providing earlier detection of vision loss in vigabatrin-treated patients. The study will enroll approximately 80 patients aged ≥ 18 years at approximately 25 centers in the United States with access to Humphrey static perimetry and SD–OCT. See the article by Sergott & Westall (62) in this supplement for a more detailed discussion of this trial.

Potential confounding factors

  1. Top of page
  2. Abstract
  3. Introduction
  4. History of vigabatrin-associated pVFDs
  5. Vision testing for vigabatrin-treated patients
  6. Potential confounding factors
  7. Mechanism of vigabatrin-associated pVFDs and taurine link
  8. Role of taurine in the retina
  9. Patient registry implemented to monitor vision function in vigabatrin-treated patients
  10. Conclusions
  11. Conflicts of interest
  12. Acknowledgments
  13. References

As noted earlier, the more conservative estimate of incidence of visual field loss from the controlled studies compared with the observational studies included in the systematic review by Maguire et al. (38) suggests the probability that some patients in the uncontrolled studies had visual loss unrelated to vigabatrin.

This probability is supported by the findings of a prospective cohort study by Gonzalez et al. (15), in which bilateral visual field constriction (by static perimetry) was not restricted to patients exposed to vigabatrin but also was detected in vigabatrin-naïve patients who received treatment with other GABAergic AED therapy. A total of 204 adults with focal-onset epilepsy were grouped based on AED exposure: patients who had been receiving vigabatrin therapy for at least 1 year (Group 1), patients who had previous exposure to vigabatrin for at least 1 year but had not received it for at least 2 years (Group 2), vigabatrin-naïve patients who were receiving treatment with other GABAergic AEDs (Group 3), and vigabatrin-naïve patients receiving non-GABAergic AEDs (Group 4). Bilateral visual field constriction was observed in 59% of patients currently receiving vigabatrin, 43% of patients with previous vigabatrin exposure, and 30% of patients with no exposure to vigabatrin (Groups 3 and 4 combined; Fig. 5). Assessment of retinal function by wide-field multifocal ERG revealed abnormal responses for 48% of current vigabatrin-treated patients and 22% of prior vigabatrin-treated patients but for no patients without prior vigabatrin exposure. The investigators speculated that two effects on the visual system could be at play: (i) the pathologic effect on retinal function (toxicity) and (ii) the physiologic effect of dampening of the visual system owing to the inhibitory nature of AEDs, which may lead to the development of a pVFD. They observed that AED-associated binasal field defects in patients with epilepsy appear to be compounded by a toxic effect of vigabatrin on retinal function, which would explain the increased frequency of pVFDs in patients exposed to vigabatrin (15).

image

Figure 5.  Percentages of participants exhibiting visual field constriction as determined by Humphrey static perimetry (open bars) and retinal dysfunction as determined by wide-field multifocal electroretinography (closed bars) in individual patient groups. Group 1 = patients who had been receiving vigabatrin therapy for at least 1 year; Group 2 = patients who had previous exposure to vigabatrin for at least 1 year, but had not received it for at least 2 years; Group 3 = vigabatrin-naïve patients who were receiving treatment with other γ-aminobutyric acid (GABA)-ergic antiepileptic medications; and Group 4 = vigabatrin-naïve patients receiving non-GABAergic antiepileptic medications. Reprinted from Epilepsy & Behavior, Vol. 16, Gonzalez P, Sills GJ, Parks S, et al. (15), Binasal visual field defects are not specific to vigabatrin, p521–526, Copyright 2009, with permission from Elsevier.

Download figure to PowerPoint

In a review of estimates of vigabatrin-associated pVFDs from a number of studies, Lawden (21) calculated an overall visual field constriction prevalence of 34% (265 of 790) in patients exposed to vigabatrin compared with a prevalence of 4% (10 of 249) in patients receiving other AEDs (Table 1). Several published case series and case reports also describe pVFDs associated with other AEDs. Arndt et al. (45) published a case series of 52 patients who were treated with vigabatrin and either sodium valproate (VPA) or carbamazepine and examined with automated kinetic perimetry, static perimetry, electrooculogram, and ERG. Patients treated with vigabatrin and VPA had more severe visual field constriction, as assessed by both static and kinetic perimetry, compared with patients treated with vigabatrin and carbamazepine. The investigators could not identify a particular degree of VPA-related additive retinal toxicity by the electrophysiologic tests as electrooculogram and ERG results did not differ significantly between the two groups.

Table 1.   Published estimates of vigabatrin-associated pVFDs
 Methods of measurementMale, n/N (%)Female, n/N (%)Both, n/N (%)Symptomatic constriction, n/N (%)VFDs in controls on other antiepileptic drugs, n/N (%)
  1. pVFDs, peripheral visual field defects.

  2. Lawden M (21). Vigabatrin-associated visual field constriction: a review. Optometry Pract. 2006;7:1–6. Reprinted with permission. Copyright © 2006 The College of Optometrists.

  3. aSeries based upon patients evaluated for epilepsy surgery.

  4. bData from Kälviäinen et al. (31).

Arndt et al. (68)Static + kinetic8/9 (89)3/10 (30)11/19 (58)2/19 (11)
Lawden et al. (24)Static5/10 (50)7/15 (47)12/25 (48)3/25 (12)0/16 (0)
Daneshvar et al. (69)Static12/41 (29)4/41 (10)
Wild et al. (26)Static + kinetic17/45 (38)12/54 (22)29/99 (29)0/42 (0)
Manuchehri et al. (44)Static8/12 (67)3/8 (38)11/20 (55)0/20 (0)1/11 (9)
Midelfart et al. (70)Static9/9 (100)6/9 (67)15/18 (83)0/5 (0)
Hardus et al. (28)aStatic + kinetic15/53 (28)5/65 (8)20/118 (17)0/39 (0)
Mauri-Llerda et al. (71)Static6/10 (60)2/10 (20)
Toggweiler & Wieser (72)Kinetic9/15 (60)1/12 (8)
Malmgren et al. (16)aKinetic19/99 (19)0/99 (0)5/55 (9)
Nousiainen et al. (73)Kinetic11/25 (44)13/35 (37)24/60 (40)0/18 (0)b
Schmitz et al. (33)Static + kinetic13/29 (45)3/31 (10)
Jensen et al. (74)Kinetic2/5 (40)1/5 (20)3/10 (30)1/10 (10)0/10 (0)
Van der Torren et al. (75)Static + kinetic19/29 (66)
Newman et al. (30)Kinetic6/46 (13)14/54 (26)20/100 (20)0/10 (0)
Nicolson et al. (25)Static42/98 (43)1/98 (1)
Total, purely staticStatic22/31 (71)16/32 (50)98/212 (46)1/32 (3)
Total, purely kineticKinetic19/76 (25)28/94 (30)75/284 (26)6/105 (6)
Total, all methodsBoth81/214 (38)64/255 (25)265/790 (34)13/322 (4)10/249 (4)
Total, excluding surgical seriesBoth66/161 (41)59/190 (31)226/573 (39)13/223 (6)5/155 (3)

Data from a retrospective observational study of children 1 month to 18 years of age who received vigabatrin as monotherapy or add-on therapy were analyzed. Of 446 children identified during a 10-year period, 160 had adequate ERG data to determine the presence of toxicity (vigabatrin monotherapy, = 73; vigabatrin add-on therapy, = 87) (76). Retinal toxicity was detected in 18 of 160 patients (11%). Of these 18 patients, 14 (77%) were in the vigabatrin add-on therapy group and four (22%) were receiving vigabatrin monotherapy, indicating a greater percentage of toxicity in children receiving vigabatrin in combination with other AEDs instead of as monotherapy.

One published case report described a woman with a pVFD similar to those described with vigabatrin but who was treated with VPA and then carbamazepine without vigabatrin (77). Investigators reported the possible presence of an inherited metabolic disorder for which visual loss can occur and theorized about the possibility of a genetic predisposition in some patients to the manifestation of adverse effects of AEDs. Another case report described a female patient who experienced a bilateral concentric pVFD during long-term therapy with VPA that was confirmed by manual and automated perimetry, with ERG indicating reduced B waves (78). One year after discontinuing VPA and starting lamotrigine, follow-up examination indicated less severe concentric visual field loss and normal ERG. Lorenz and Kuck (79) reported the case of a woman who experienced bilateral reduced visual acuity and restricted visual fields associated with prolonged toxic blood concentrations of phenytoin in the context of a hereditary defect in phenytoin metabolism. Kaufman et al. (80) described a male patient treated with adjunctive tiagabine (in addition to carbamazepine and several non-AED medications) for bipolar disorder in whom a mild pVFD (predominantly peripheral nasal constriction) was detected on examination after 13 and 18 months of tiagabine. The pVFD appeared to have been reversible upon discontinuation of the drug. However, no baseline perimetry or ERG measurements were available for this patient.

Epilepsy-related surgery is another potential confounding factor in considering the association of pVFDs with vigabatrin. For example, in a prospective study of 105 consecutive patients who had an anterior temporal lobe resection for epilepsy, 16 patients (15.2%) had postoperative pVFDs that were not present preoperatively (81).

An additional potential confounder to consider is the presence of structural abnormalities at baseline. For example, OCT in an infant with IS demonstrated bilateral foveal hypoplasia, a structural abnormality that could create a field defect at baseline (R. C. Sergott, unpublished data).

Mechanism of vigabatrin-associated pVFDs and taurine link

  1. Top of page
  2. Abstract
  3. Introduction
  4. History of vigabatrin-associated pVFDs
  5. Vision testing for vigabatrin-treated patients
  6. Potential confounding factors
  7. Mechanism of vigabatrin-associated pVFDs and taurine link
  8. Role of taurine in the retina
  9. Patient registry implemented to monitor vision function in vigabatrin-treated patients
  10. Conclusions
  11. Conflicts of interest
  12. Acknowledgments
  13. References

The mechanism for developing a vigabatrin-associated pVFD is only partially understood. In the brain, vigabatrin increases GABA concentrations (thereby reducing seizure activity) by inhibiting GABA-transaminase (GABA-T), the enzyme responsible for catabolism of GABA (82). In animal studies, vigabatrin has been shown to accumulate in the retina in significantly greater concentrations relative to other tissues and is associated with accumulation of GABA in the retina (83). Therefore, accumulation of vigabatrin in the retina may be responsible for visual field constriction. Specifically, vigabatrin treatment is associated with a dosage-related decrease in GABA-T activity and increase in GABA concentrations that are more pronounced in the retina (GABA-T reduced to 22% of control; GABA elevated to 260% of control) than in any region of the brain (83). In addition, vigabatrin concentrations were 5- to 18.5-fold greater in the retina than in the brain in several animal studies (83, 84).

Other animal studies have reported disruption of the outer nuclear layer of the retina after long-term vigabatrin treatment (68, 85). Both the amacrine cells of the inner retina (which support retinal cone function) and the Müller cells of the peripheral retina have been implicated as potential sites of retinotoxicity associated with vigabatrin (23, 86).

Although GABA is physiologically an inhibitory transmitter, it appears to become excitatory and potentially excitotoxic in damaged neurons and when present in excessive concentrations (21). Based on ERG testing in affected adults and children and structural analysis of the RNFL, several studies have demonstrated a correlation between atrophy (or thinning) of the RNFL and vigabatrin-associated pVFDs (87, 88). This is consistent with findings using OCT (60, 63). The defect appears to arise from injury to both the retinal photoreceptors in the outer retina and the retinal ganglion cells and their axons in the inner retina (39, 60, 89). In a prospective cross-sectional observational study, Wild et al. (87) reported a significant reduction in mean RNFL thickness in all 11 vigabatrin-treated patients with pVFDs, four of 16 of those without pVFDs, and none of 13 patients with epilepsy who were never exposed to vigabatrin. ERG studies of visual field loss associated with vigabatrin have disclosed an abnormality of the B wave, which is generated from the inner retina (cone system). Oscillatory potential responses were also lost, suggestive of amacrine cell dysfunction (90).

Role of taurine in the retina

  1. Top of page
  2. Abstract
  3. Introduction
  4. History of vigabatrin-associated pVFDs
  5. Vision testing for vigabatrin-treated patients
  6. Potential confounding factors
  7. Mechanism of vigabatrin-associated pVFDs and taurine link
  8. Role of taurine in the retina
  9. Patient registry implemented to monitor vision function in vigabatrin-treated patients
  10. Conclusions
  11. Conflicts of interest
  12. Acknowledgments
  13. References

The biologic importance of taurine in the retina has been recognized since the observation by Schmidt and colleagues (91) nearly 40 years ago that taurine deficiency in cats caused retinal degeneration. In a mouse model, disruption of the taurine transporter gene resulted in severe, progressive retinal photoreceptor cell degeneration, suggesting that taurine is critical for normal retinal development and function (92).

Early work by Alm and Tornquist (93) demonstrated that carrier systems for amino acids exist in the blood–retina barrier similar to those in the blood–brain barrier, the only difference being a separate, saturable uptake of taurine in the retina. More recently, investigators found that the taurine transporter at the blood–retina barrier also transports GABA as a substrate with a lesser affinity than taurine (94).

A study of 21 children and 23 adults demonstrated that prolonged taurine-free parenteral nutrition resulted in retinal degeneration that was reversed with taurine supplementation (95). Recent work suggests that taurine promotes the generation of rod photoreceptor cells from retinal progenitor cells via an effect on both GABAA and glycine receptors (96). The investigators added a series of pharmacologic agents to newborn mouse retinal cultures, which were dissociated after 10 days and quantified in terms of presence of rod photoreceptors. Taurine-treated cultures produced 2.7-fold more rod photoreceptors compared with controls, and the combination of glycine and GABA induced a 2.5-fold increase, whereas addition of GABA or glycine alone to saturating concentrations of taurine did not induce more rod photoreceptors over taurine alone (Fig. 6). The addition of strychnine, a glycine receptor antagonist, or bicuculline, a GABA receptor antagonist, inhibited the rod-promoting effect of taurine.

image

Figure 6.  Percentage of total retinal cells expressing rhodopsin after exposure to pharmacologic agents. Newborn mouse retinae were dissociated and cultured in collagen gels with serum-free defined medium. Pharmacologic agents were added and cells were cultured for 10 days. Cultures were then dissociated, immunostained with an anti-rhodopsin antibody (RHO4D2) that labels only rod photoreceptors, and quantified y-axes represent the fold change in the percentage of rod photoreceptors relative to control cultures. Strychnine, a specific antagonist of glycine receptors, picrotoxin, an antagonist of GABAA receptors and homomeric glycine receptors, and bicuculline, an antagonist of GABA receptors, were added to cultures on Days 0, 3, and 6. Graph represents data from three to five independent experiments. Significance tests were performed comparing control vs agonists (black asterisks) and taurine-treated vs taurine-plus antagonists (blue asterisks). *< 0.05. **< 0.01. ***< 0.001. Reprinted from Neuron, Vol. 42, Young TL, Cepko CL, A role for ligand-gated ion channels in rod photoreceptor development, p867–879, Copyright 2004, with permission from Elsevier.

Download figure to PowerPoint

A recent study in albino rodents demonstrated an association between vigabatrin-associated retinal toxicity and taurine deficiency (97). In this study, vigabatrin-treated rats exposed to cycles of 12 h of light and 12 h of darkness exhibited more pronounced retinal lesions than those kept in darkness alone. Taurine concentrations were 67% lesser in vigabatrin-treated animals compared with controls, and taurine concentrations correlated with loss of visual functions (e.g., ERG amplitudes; Fig. 7), suggesting that vigabatrin induces a deficiency in taurine that results in retinal phototoxicity. Dietary taurine supplementation in the rodents did not reverse existing retinal changes but reduced the development of retinal damage.

image

Figure 7.  Partial prevention of vigabatrin-induced retinal toxicity in rats by taurine supplementation. (A–F) Retinal sections showing that vigabatrin (VGB)-elicited retinal lesions are less extensive in a rat with taurine supplementation (C, F, VGB + taurine) than without (B, E, VGB), but still greater than in a control animal (A, D). These sections were stained with 4′-6-diamidino-2-phenylindole (DAPI, blue in A–F) and immuno-labeled with antibodies directed against Goα (red in A–C), GFAP (green in A–C), and cone arrestin (red in D–F). Photoreceptor nuclei displaced above the outer nuclear layer (ONL) are observed in both groups of VGB-treated rats treated with or without taurine supplementation (B, C), but not in control animals (A). GFAP-positive processes extending vertically throughout the retina were observed in VGB-treated rats (B), but not in control animals (A); greater degrees of GFAP staining were also observed for VGB-treated rats receiving taurine supplementation (C) than those observed in control animals (A). Similarly, there were clearly fewer cone arrestin-positive photoreceptors in the VGB-treated rats receiving morning injections (E) than in control animals (D), with a smaller decrease in cone arrestin-positive photoreceptors in VGB-treated rats receiving evening injections (F). Quantification of photopic ERG amplitude (G), length of retinal areas with displaced photoreceptor (PR) nuclei (H), density of cone inner/outer segments (I), and areas with increased GFAP expression (J) in control rats (standard error of the mean [SEM], n = 6), in the VGB-treated animals with or without taurine supplementation (VGB, n = 7; VGB + taurine, n = 7, SEM). The scale bar represents 50 μm (inner plexiform layer [IPL]). Taurine deficiency is a cause of vigabatrin-induced retinal phototoxicity, Jammoul F, Wang Q, Nabboul R, et al. Copyright © 2009 Annals of Neurology. Reproduced with permission of John Wiley & Sons, Inc.

Download figure to PowerPoint

To assess the potential relevance of these findings to humans, the authors retrospectively reviewed data for six vigabatrin-treated infants with IS. Five had low or undetectable taurine concentrations, including one who had a normal concentration prior to vigabatrin treatment (97). The investigators suggested that a taurine-rich diet and reduced sunlight exposure may significantly reduce the extent of vigabatrin-associated retinal lesions and proposed future studies to investigate whether co-treatment with taurine and vigabatrin can preserve the antiepileptic efficacy of vigabatrin without inducing visual defects (98).

The mechanism by which vigabatrin reduces taurine concentrations is not clear. Vigabatrin-elicited increases in GABA may be involved given that GABA is a substrate and thus competitive inhibitor of the taurine transporter (99). Alternatively, vigabatrin may have a direct effect on taurine uptake and release (97). Interestingly, taurine exhibits antiepileptic properties (97).

In a separate study, the same investigators showed that in neonatal rats treated with vigabatrin from postnatal days 4 to 29, vigabatrin triggers not only cone photoreceptor damage and disorganization of the photoreceptor layer and gliosis but also retinal ganglion cell loss (98). As in their previous study, taurine supplementation partially prevented the retinal lesions, particularly the retinal ganglion cell loss.

Patient registry implemented to monitor vision function in vigabatrin-treated patients

  1. Top of page
  2. Abstract
  3. Introduction
  4. History of vigabatrin-associated pVFDs
  5. Vision testing for vigabatrin-treated patients
  6. Potential confounding factors
  7. Mechanism of vigabatrin-associated pVFDs and taurine link
  8. Role of taurine in the retina
  9. Patient registry implemented to monitor vision function in vigabatrin-treated patients
  10. Conclusions
  11. Conflicts of interest
  12. Acknowledgments
  13. References

A comprehensive Risk Evaluation and Mitigation Strategy (REMS) was implemented in August 2009 in conjunction with Food and Drug Administration approval of vigabatrin. The REMS is administered through the Lundbeck Inc. Support, Help And Resources for Epilepsy (SHARE) program. The aim of REMS is to decrease the risk of vigabatrin-associated vision loss while providing benefit–risk analyses for appropriate patient populations. Vigabatrin REMS included implementation of a patient registry to assess the incidence, prevalence, time to onset, progression, and severity of vision loss in vigabatrin-treated patients (100). All US patients treated with vigabatrin are required to enroll in the registry. Vision assessments are required at baseline (≤ 4 weeks after therapy initiation), every 3 months during therapy, and 3–6 months after discontinuation of therapy. As of February 1, 2011, a total of 2,473 patients (1,500 with IS, 846 with rCPS, and 120 with other diagnoses) had enrolled in the registry. Kaplan-Meier analysis of time in registry indicated that 83% of all enrolled patients with rCPS remained in the registry beyond 3 months and that 97% of patients with IS remained in the registry beyond 1 month. The ongoing registry will ensure compliance with regular vision monitoring and benefit–risk assessments.

Conclusions

  1. Top of page
  2. Abstract
  3. Introduction
  4. History of vigabatrin-associated pVFDs
  5. Vision testing for vigabatrin-treated patients
  6. Potential confounding factors
  7. Mechanism of vigabatrin-associated pVFDs and taurine link
  8. Role of taurine in the retina
  9. Patient registry implemented to monitor vision function in vigabatrin-treated patients
  10. Conclusions
  11. Conflicts of interest
  12. Acknowledgments
  13. References

Vigabatrin meets an important unmet need in the treatment of rCPS and uncontrolled IS. Given the severe consequences of these conditions, vigabatrin is an appropriate option for those who achieve substantial clinical benefit. Because response to treatment typically occurs much more rapidly (i.e., within 12 weeks for CPS and 1–2 weeks for IS) than the onset of a vigabatrin-associated pVFD, there is little risk associated with initiating the drug to test efficacy (13). With careful monitoring of peripheral vision, the majority of patients for whom vigabatrin is effective can continue the medication with only minimal risk that their peripheral vision will be significantly compromised.

Screening for pVFD was not conducted for patients with epilepsy prior to the association of pVFD with vigabatrin treatment. Results of a prospective cohort study as well as several case series and case reports suggest that bilateral visual field constriction is not restricted to patients exposed to vigabatrin but has also been detected, although less frequently, in vigabatrin-naïve patients with epilepsy who received treatment with other GABAergic AEDs (15). This indicates that pVFDs may be attributable to factors other than vigabatrin therapy (e.g., background disease, comorbidities).

Areas for further research include the functional impact of VAVFL, the further development of retinal imaging techniques, particularly for detecting field defects in very young children, and further research into the potential use of taurine to ameliorate vigabatrin-associated pVFDs.

Acknowledgments

  1. Top of page
  2. Abstract
  3. Introduction
  4. History of vigabatrin-associated pVFDs
  5. Vision testing for vigabatrin-treated patients
  6. Potential confounding factors
  7. Mechanism of vigabatrin-associated pVFDs and taurine link
  8. Role of taurine in the retina
  9. Patient registry implemented to monitor vision function in vigabatrin-treated patients
  10. Conclusions
  11. Conflicts of interest
  12. Acknowledgments
  13. References

Medical writing and editorial assistance were provided by Angela Cimmino, PharmD, BCPS, and Robin L. Stromberg, PhD, of Arbor Communications, Inc. (Ann Arbor, MI, USA), and Michael A. Nissen, ELS, of Lundbeck Inc. (Deerfield, IL, USA). This support was funded by Lundbeck.

References

  1. Top of page
  2. Abstract
  3. Introduction
  4. History of vigabatrin-associated pVFDs
  5. Vision testing for vigabatrin-treated patients
  6. Potential confounding factors
  7. Mechanism of vigabatrin-associated pVFDs and taurine link
  8. Role of taurine in the retina
  9. Patient registry implemented to monitor vision function in vigabatrin-treated patients
  10. Conclusions
  11. Conflicts of interest
  12. Acknowledgments
  13. References