J. M. Pellock, MD, Division of Child Neurology, Department of Neurology, Virginia Commonwealth University, Children’s Pavilion, 1001 East Marshall Street, Richmond, VA 23298, USA e-mail: firstname.lastname@example.org
Vigabatrin is an effective and well-tolerated antiepileptic drug (AED) for the treatment of refractory complex partial seizures (rCPS) and infantile spasms (IS), but its benefits must be evaluated in conjunction with its risk of retinopathy with the development of peripheral visual field defects (pVFDs). Vigabatrin should be considered for rCPS if a patient has failed appropriate trials of other AEDs or is not a suitable candidate for other AEDs, is not an optimal surgical candidate, and continues to experience debilitating effects from seizures. Vigabatrin is indicated as monotherapy for pediatric patients with IS. Its efficacy in achieving improved seizure control should be apparent within 12 weeks in patients with rCPS and within 2–4 weeks after attaining appropriate dosage for patients with IS. Because 12 weeks is well less than the known time of onset of visual defects, the risk of developing pVFDs may be minimized by discontinuing vigabatrin early during the course of therapy for patients with inadequate response. Appropriate vision screening is recommended at baseline, every 3 months during continued vigabatrin treatment, and at 3–6 months after discontinuation (if therapy has spanned more than a few months). If a pVFD is detected at any point and the decision is made to discontinue therapy, the pVFD is not likely to progress after discontinuation of vigabatrin. Although some patients will be at risk of retinopathy, vigabatrin is an appropriate treatment option for patients who achieve substantial clinical benefit, especially given the severe consequences of rCPS and uncontrolled IS. While retinopathy with the development of pVFDs is a serious adverse event, it is not life-threatening and its risk can be effectively managed.
Approximately 30% of patients with complex partial seizures (CPS) continue to experience intractable seizures despite therapy with antiepileptic drugs (AEDs) (1). Complex partial seizures are associated with significantly greater risks compared with the general population, including greater risk of bodily injury (2) or sudden unexpected death (3). In addition, mortality rates for patients with refractory CPS (rCPS) are much greater than mortality rates for age-matched healthy individuals and are four- to seven-times greater than mortality rates for patients whose seizures are controlled (2, 3).
Of patients with infantile spasms (IS), 70–90% exhibit learning difficulties or significant cognitive impairment, and psychiatric symptoms occur in 20–40% of adults with a history of IS (4). The reported premature death rate for patients with a history of IS ranges from 2% to 31%, with 61% of deaths occurring at or before 10 years of age (4–6). However, the death rate remains elevated into adolescence and adulthood compared with the general population, as evidenced by a population-based cohort of 245 children with a diagnosis of epilepsy followed prospectively for 40 years (7). The 24% death rate in this cohort was three-times greater than the expected age- and sex-adjusted mortality rate for the general population. The median age at death was 23 years (range: 1–50 years), and the majority of deaths (51 of 60 [85%]) were patients who were not in remission at time of death.
Vigabatrin is an irreversible inhibitor of γ-aminobutyric acid (GABA) transaminase that increases concentrations of GABA, a major inhibitory neurotransmitter in the brain. The full mechanism of action of vigabatrin is reviewed by Ben-Menachem in a separate article in this supplement (8).
Vigabatrin is effective as adjunctive therapy for adult patients with rCPS who have responded inadequately to several alternative treatments and as monotherapy for pediatric patients aged 1 month to 2 years with IS (9–23). In clinical trials in patients with rCPS, the addition of vigabatrin 3 or 6 g/day resulted in ≥ 50% seizure reduction in 43–51% of patients and complete seizure freedom in 5–12% of patients (24). Generally, response to therapy was observed within the first 12 weeks of treatment. In clinical trials in patients with IS, treatment with vigabatrin 100–150 mg/kg/day resulted in rapid cessation of seizures (within approximately 14 days or less) in 36–54% of patients (9, 14, 25, 26) and has been reported to induce complete absence of spasms at 12–14 months’ follow-up in 76% of infants (14, 27). The articles in this supplement by Faught (28), Ben-Menachem and Sander (29), and Carmant (30) present a more detailed discussion of clinical trials of vigabatrin.
Despite proven efficacy, the use of vigabatrin has been limited by the associated risk of retinopathy, characterized by irreversible, bilateral, and concentric peripheral visual field constriction (31). Vigabatrin is otherwise well-tolerated compared with other AEDs and is not associated with known life-threatening adverse effects, following more than 21 years of worldwide use. Given the severe consequences of rCPS and uncontrolled IS and the demonstrated efficacy of vigabatrin for these indications, a systematic consideration of the potential benefits and risks of vigabatrin for individual patients should help guide appropriate selection of patients for this therapy.
Vision defects with vigabatrin
Peripheral visual field defects (pVFDs) associated with vigabatrin are characterized by a bilateral, concentric peripheral constriction of the visual field. The defects range from mild to moderate or, more infrequently, marked severity, and are generally more pronounced nasally (31). Mild-to-moderate pVFDs are generally asymptomatic.
Multiple studies have evaluated the incidence, prevalence, pathophysiology, risk factors, and likely outcomes associated with vigabatrin-induced pVFDs. The wide variation in prevalence estimates and potential risk factors is likely a result of the diversity of perimetry techniques and definitions of pVFDs employed, as well as differences in extent of drug exposure, small cohort sizes, retrospective methodology of many of the studies, and lack of patient symptoms (32). In a review of 11 studies, Kälviäinen & Nousiainen (33) reported that pVFDs were detected in 169 of 528 patients (32%) treated with vigabatrin. Based on a review of 22 studies, Kinirons et al. (34) reported a pVFD prevalence range of 19–92% for vigabatrin-treated adults with CPS and up to 31% for vigabatrin-treated infants with IS. Most reported estimates for vigabatrin-associated pVFDs are in the range of 30–50% in adults, depending on dosage and duration of drug exposure, and approximately 20% in children (35, 36).
A recent systematic review of observational studies in patients with partial epilepsy to determine the magnitude of vigabatrin-associated visual field loss (VAVFL) included 32 studies in which 1,678 patients were exposed to vigabatrin and 406 were not (36). A total of 738 (44%) vigabatrin-exposed patients had visual field loss compared with 30 (7%) vigabatrin-naïve patients. The random-effects estimates for the percentages of vigabatrin-exposed patients with pVFDs were 52% (95% confidence interval: 46–59%) for adults and 34% (95% confidence interval: 25–42%) for children. In a subanalysis of the 16 trials with a control group, more conservative estimates of one-third of adults and less than one-fifth of children with VAVFL were identified, suggesting that some patients in the uncontrolled studies had visual loss unrelated to vigabatrin treatment.
A prospective comparative, open-label, multinational study (Study 4020) of 735 children (≥ 9 years) and adults with refractory partial epilepsy reported vigabatrin-associated pVFDs in 26.3% of 8- to 12-year-olds and in 43.3% of patients > 12 years of age who were exposed to vigabatrin for ≥ 6 months prior to study entry and continued treatment (35). For the group of patients previously treated with vigabatrin for ≥ 6 months who had discontinued the drug for ≥ 6 months, pVFDs were reported in 14.9% of 8- to 12-year-olds and in 24.5% of patients > 12 years of age.
Some evidence suggests the risk of developing vigabatrin-associated pVFDs increases with increased dosage and duration of therapy (32, 37–40) and with cumulative exposure (38, 41). However, some patients have long-term exposure (> 10 years) without developing pVFDs (42), suggesting that dosage and other yet-to-be-identified factors may play a role.
Although based on a small number of patients studied, recent data suggest that the risk of vigabatrin-associated visual field abnormalities may be lesser for children with IS who were treated with vigabatrin in infancy compared with those who received the drug later in childhood (43). Of 16 children treated with vigabatrin for IS in infancy (mean age of therapy initiation: 7.6 months; mean duration of therapy: 21 months), 15 had normal visual fields determined by Goldmann kinetic perimetry performed at age 6–12 years, and one (6%) had mild vigabatrin-associated pVFDs (duration of treatment: 19 months; cumulative dosage: 572 g).
A more recent study at the University of Toronto detected retinal toxicity during a 10-year period in 18 of 160 patients (11%) 1 month to 18 years of age treated with vigabatrin for IS (in the infant population) or refractory partial onset seizures in the older children (44). Retinal toxicity was determined based on sustained 30-Hz flicker amplitude abnormalities, an electroretinography (ERG) parameter associated with pVFDs. Toxicity was strictly defined as ERG reduction from baseline on at least two consecutive occasions, and ERG reduction was defined as reduction greater than normal limits (44). Of the 18 patients with retinal toxicity, 14 (78%) were treated with additional AEDs and the other four received vigabatrin as monotherapy. Data on medication dosages, duration of treatment, and extent of pVFDs were not provided but may be part of a follow-up assessment. The authors found a significantly greater percentage of toxicity in the group of children treated with vigabatrin plus additional AEDs, indicating that vigabatrin may not be acting alone with respect to potential retinopathy. Other agents (e.g., GABAergic drugs such as valproic acid) and other factors (e.g., underlying seizure etiology) may play roles (44). The study also provides further evidence that the prevalence of retinal pathology among vigabatrin-treated patients may be far less for younger patients.
The development of vigabatrin-associated pVFDs is gradual, with the majority of cases in children and adults having been reported after ≥ 1 year of treatment (31). For adults, the earliest reported finding of an abnormal field examination was after 4 months of treatment (38). For children, the earliest reported onset of an abnormal field examination occurred after 11 months of treatment (mean time to onset: 5.5 years) (24). The earliest sustained onset of a vigabatrin-induced retinal defect in an infant was 3.1 months in a longitudinal study of children (24). On average, the progression of pVFDs associated with vigabatrin is < 2 degrees per year from the temporal visual field and < 1 degree per year in the nasal field (45). Overall, available data suggest that pVFDs do not typically progress or regress to a clinically significant degree after discontinuation of therapy (31). Vigabatrin has not been associated with central visual acuity changes in several multinational studies (46). However, one study reported decreased visual acuity, ranging from 20/25 to 20/60 in one or both eyes in 12 of 32 patients treated with vigabatrin (47).
The limited information available regarding the vision prognosis for patients with detectable VFD who elect to continue therapy with vigabatrin suggests that, for the majority of patients, visual fields remain stable for long periods. Paul et al. (48) investigated this issue in 15 patients who, after having received vigabatrin for at least 2 years, were assessed at quarterly intervals while continuing therapy for 1 year. Initially, nine patients had normal fields and six had constricted fields. Thirteen showed no significant change in field extent over the year, one showed some initial improvement (probably a practice effect), and one showed apparent progressive worsening of field extent, moving from the normal group to the impaired group during that time. In another cohort of 16 patients with vigabatrin-associated pVFDs who elected to continue vigabatrin for at least 18 months (range: 18–43 months), Best and Acheson (49) found evidence of progression for only one patient. Therefore, if a vigabatrin-induced pVFD is diagnosed definitively, the severity of the pVFD and the associated impact, or lack thereof, on various aspects of quality of life and functioning must be weighed against the therapeutic benefit of ongoing therapy for the individual patient (e.g., decrease in or freedom from seizures, improved quality of life) in determining an appropriate treatment decision (Table 1) (31). Of note, 95% of adults with rCPS cannot drive as a result of their epilepsy, whereas only severe pVFDs would affect a patient’s ability to drive.
Table 1. Risk–benefit analysis by severity of pVFDs
Severity of pVFD based on intact binocular visual field
√, yes, able to perform; X, no, unable to perform; ±, may or may not be able to perform.
Severe (< 60°)
Develop adaptive strategies, such as scanning
The article by Plant and Sergott in this supplement provides a further discussion of the epidemiology, pathophysiology, and outcomes associated with vigabatrin-associated pVFDs (50).
Given the potential risk of developing an irreversible pVFD with vigabatrin therapy, appropriate vision screening at baseline and during therapy is imperative to minimize the risk of vision loss and to ensure that the benefit–risk considerations regarding initial and ongoing therapy in individual patients are fully informed. Recommended screening tests should be based on a patient’s developmental age and ability or inability to perform perimetry (31). Many patients with epilepsy or IS have behavioral and cognitive limitations to visual field testing. Confrontation testing can be performed with patients of all ages (although less reliably for children ≤ 2 years old) and provides an adequate qualitative assessment to identify patients who may have pVFDs and require further testing. Perimetry usually can be performed with patients with a developmental age > 9 years. Kinetic perimetry (e.g., Goldmann perimetry or automated kinetic perimetry) is preferred over automated static perimetry, because the latter may produce variable results in patients with epilepsy. ERG is usually the preferred electrophysiologic test for children suspected of having vigabatrin-associated pVFDs who are unable to perform visual field testing. However, ERG testing requires sedation in infants and young children, thus requiring a benefit–risk judgment for each patient.
All patients should undergo vision testing at baseline, and vision monitoring should be performed every 3 months during vigabatrin therapy and at 3–6 months after discontinuation of treatment (31). For patients with an approximate developmental age of 9 years or greater, perimetry testing should be used if possible. For infants, children, and those not able to perform perimetry, ERG and/or optical coherence tomography (for patients aged ≥ 6 years) may be considered. If testing indicates pVFDs at any point, additional confirmatory testing should be completed and the benefit and risks of further treatment should be evaluated.
Both the detailed review of visual testing methodologies, algorithms, and recommendations by Sergott and Westall (51) and the discussion of vigabatrin vision safety issues by Plant and Sergott (50) in this supplement present more thorough discussion of these issues.
Magnetic resonance imaging changes with vigabatrin in children
Intramyelinic edema (IME) has been reported in rodents and dogs treated with vigabatrin (52, 53) but was not found in a review of data from 350,000 patient-years of vigabatrin exposure in humans (54). In 2006, magnetic resonance imaging (MRI) signal changes consistent with IME were reported in infants with IS treated with vigabatrin (55). Subsequent studies have confirmed MRI signal changes associated with vigabatrin in infants (56–58). In a retrospective analysis of infants (≤ 24 months old) with IS, the prevalence of prespecified MRI abnormalities was 21.5% (17 of 79) in vigabatrin-treated patients and 4.3% (four of 93) in vigabatrin-naïve patients (P < 0.001) (57). Of the nine vigabatrin-treated patients with a prespecified MRI abnormality who also had at least one subsequent MRI evaluation, MRI abnormalities resolved for six patients (two [33.3%] during continued vigabatrin treatment and four [66.7%] after discontinuation of vigabatrin). These abnormalities were observed in both white and gray matter, including the thalamus, columns of the fornix, hypothalamus, cerebellum, corpus callosum, and optic tracts (57).
MRI abnormalities tend to peak after 3–6 months of exposure to vigabatrin and most resolve, even with continued use of the drug (56, 57). Because vigabatrin-associated MRI changes have not been observed for older children or adults, investigators have hypothesized that such changes are associated with developmental changes in myelination in infants or an unknown, underlying metabolic condition (56, 57). The clinical significance, if any, of these MRI changes is not yet understood (57).
Recently, Horton et al. (59) reported the first case of vigabatrin-induced IME in a 9-month-old patient treated for IS who had a history of cerebral palsy secondary to hypoxic-ischemic brain injury following premature birth (29 weeks’ gestation). The child’s 3-month hospital stay after birth was further complicated by bronchiolitis and necrotizing enterocolitis. At 7 months of age, IS developed and was initially treated with topiramate. A brain MRI at that time revealed slightly delayed myelination of the white matter, slightly enlarged lateral ventricles, reduced thalamus volume, and thinning of the corpus callosum. At 9 months of age, vigabatrin was added, because of increased daily seizure episodes, at a dosage of 175 mg (25 mg/kg) twice daily that was increased to 50 mg/kg twice daily the next day. Vigabatrin therapy was stopped 3 days later when the child developed fever, oxygen desaturation, and decreased consciousness. It was restarted in the intensive care unit 9 days later, and a dosage of 175 mg twice daily was continued for 11 days until the child’s death from bronchopneumonia at 10 months of age. Neuropathologic examination on autopsy revealed white matter vacuolation consistent with IME. Walker and Kälviäinen address the issue of vigabatrin-associated MRI changes in a review of non-vision adverse events in this supplement (60).
Other vigabatrin-related adverse events in adults and children
Vigabatrin is generally well-tolerated, with an adverse event profile similar to that of other AEDs (45). As noted previously, non-vision adverse events are described in more detail in a separate review article by Walker and Kälviäinen (60) in this supplement. In clinical trials, the most common treatment-emergent adverse events were central nervous system–related, including drowsiness, dizziness, headache, and fatigue with adjunctive vigabatrin for adults with rCPS and sedation, somnolence, and irritability with vigabatrin monotherapy for infants with IS (61). Vigabatrin has little effect on cognitive function, which is an important concern in patients with rCPS (62–64). It is one of several AEDs associated with depression, which is typically mild and responds to dosage reduction or slow taper (65). Overall, with the exception of pVFDs, vigabatrin has an adverse event profile similar to that of many of the newer generation AEDs, but without the life-threatening adverse events that accompany several of the other newer generation AEDs (e.g., Stevens-Johnson syndrome with lamotrigine, oxcarbazepine, and zonisamide; aplastic anemia and fulminant hepatic failure with felbamate) (66). Many clinical studies, including the pivotal randomized controlled trials in rCPS and IS, demonstrate that vigabatrin is generally well-tolerated, with an overall adverse event profile similar to that of other AEDs. In addition, as with any AED therapy, the benefit of improved seizure control must be balanced against the unknown potential risk associated with abnormal MRI changes (60).
Types of seizures that may worsen with vigabatrin
Several AEDs, including vigabatrin, have been associated with the paradoxical effect of inducing seizures (67). In the context of inappropriate choice of antiepileptic medication, myoclonic seizures can be provoked by vigabatrin, lamotrigine, and gabapentin (68–70). Exacerbation of seizures owing to antiepileptic medications, including vigabatrin, has been reported for patients with Lennox-Gastaut syndrome (70, 71). For 194 children with uncontrolled seizures (of various types), vigabatrin as add-on therapy was associated with increased seizure frequency in 10% of children older than 2 years, mostly affecting generalized seizures. Increased atypical absence and tonic seizures were observed in patients with non-progressive myoclonic epilepsy and with Lennox-Gastaut syndrome (71).
Refractory CPS is a severe debilitating disease that can severely affect quality of life and possibly lead to early death (2). It is in this context that the benefit–risk ratio of vigabatrin treatment must be considered. Vigabatrin should be considered for rCPS if a patient has failed appropriate trials of and/or is not a suitable candidate for other AEDs, is not an optimal surgical candidate, and continues to experience debilitating effects from seizures. An example of such a patient is an adult with rCPS who has responded inadequately to several treatments, which may include both oral agents and vagus nerve stimulation, and who is not a candidate for surgical resection. Table 2 lists general recommendations for patient selection and monitoring (45). Appropriate vision screening should be employed at baseline, throughout vigabatrin therapy, and after discontinuation of treatment.
Table 2. Recommendations for use of vigabatrin in adults with complex partial seizures
Reprinted from the Journal of Neuropsychiatric Disease and Treatment, Vol. 5, Waterhouse EJ, Mims KN, Gowda SN. Treatment of refractory complex partial seizures: role of vigabatrin, p505-515, Copyright 2009, with permission from Dove Medical Press Ltd.
The patient has failed adequate trials of multiple antiepileptic drugs or therapies (i.e., neurostimulation).
The patient is not a candidate for resective epilepsy surgery.
The patient and/or guardian understand the potential risks of treatment, give consent for treatment, and will be compliant with follow-up testing.
The patient or caregiver should keep a seizure calendar at baseline and during treatment, to facilitate assessment of efficacy.
Baseline visual field testing must be performed prior to starting vigabatrin. Patients with pre-existing visual field defects should not receive vigabatrin.
After 12 weeks of treatment, seizure response to vigabatrin should be assessed. If there has been no significant improvement, vigabatrin should be discontinued
If meaningful improvement in seizures has occurred with vigabatrin treatment, treatment may be continued with formal testing of visual fields or retinal function every 3–6 months.
If there is evidence of visual impairment, the risks and benefits of vigabatrin treatment should be reconsidered in light of the individual’s circumstances.
The efficacy of vigabatrin in achieving improved seizure control is typically evident within 2–4 weeks after attaining appropriate dosage for patients with IS and within 12 weeks for patients with rCPS (9, 12–15, 61). Therefore, clinical response can be evaluated early during the course of therapy, well before the average time of onset of visual defects (25, 47) (Fig. 1). If substantial improvement in CPS has not been achieved by 12 weeks of therapy, discontinuation of vigabatrin at this point should minimize the risk of developing pVFDs (46). If vigabatrin treatment is effective, then regular vision monitoring every 3 months and regular evaluation of clinical benefit will facilitate ongoing, informed benefit–risk decision-making by the patient and clinician (46). If a pVFD is detected at any point and the decision is made to discontinue therapy, evidence suggests that the pVFD is not likely to progress to a clinically significant extent after discontinuation of the drug (31).
Controlled studies concerning the optimal duration of vigabatrin therapy for IS are lacking (43, 72). Limited data suggest that vigabatrin treatment can be safely withdrawn for children who have been seizure-free for 6 months (72). However, there is some evidence that children with focal cortical dysplasia and tuberous sclerosis may relapse after discontinuation of therapy and become refractory to treatment (73). In patients with IS, clinical response (or lack thereof) to vigabatrin therapy may be realized earlier than the 12-week time point, at which clinical evaluation of efficacy is suggested (24, 72). As noted earlier, the efficacy of vigabatrin in achieving improved seizure control in patients with IS is typically evident within 2–4 weeks (9, 13, 14, 61). Therefore, a US consensus report regarding IS recommended assessment of treatment at or before 2 weeks of dosage titration of vigabatrin (72). Treatment modification early in the course of therapy is recommended because longer treatment trials (i.e., > 3 months) are not likely to be effective and may be associated with increased risk of developing pVFDs.
Studies are also lacking regarding the most effective course of action if spasms recur after discontinuing vigabatrin. Unanswered questions include whether to restart vigabatrin treatment or to use an alternative therapy; whether a patient’s history of exposure and/or response to adrenocorticotropic hormone would affect the choice or outcome of therapy; and whether, with a second course of vigabatrin, the additional exposure would provide a cumulative increased risk for developing pVFDs.
Patient registry implemented to monitor vision function in vigabatrin-treated patients
A comprehensive Risk Evaluation and Mitigation Strategy (REMS) was implemented in August 2009 in conjunction with Food and Drug Administration approval of vigabatrin. REMS is administered through the Lundbeck Inc. Support, Help And Resources for Epilepsy (SHARE) program. The goal of REMS is to decrease the risk of vigabatrin-associated vision loss while providing benefit–risk analyses for appropriate patient populations. Vigabatrin REMS includes implementation of a patient registry to assess the incidence, prevalence, time to onset, progression, and severity of vision loss in vigabatrin-treated patients (74). 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. Table 3 summarizes the other components of the REMS program. As of February 1, 2011, a total of 2473 patients (1500 with IS, 846 with rCPS, and 120 with other diagnoses) had enrolled in the registry. A total of 83% of all enrolled patients with rCPS remained in the registry beyond 3 months, and 97% of patients with IS remained in the registry beyond 1 month according to Kaplan-Meier analysis. The ongoing registry will ensure compliance with regular vision monitoring and benefit–risk assessments.
Table 3. Vigabatrin REMS administered through the SHARE program
Vigabatrin is very effective in the treatment of rCPS and IS. It is associated with benefits as well as risks — the most notable risk being retinopathy with the development of pVFDs. While retinopathy with pVFDs is a serious adverse event, it is not life-threatening and its risk can be effectively managed. To minimize the risk of vision loss, a physician’s benefit–risk analysis of a therapy trial with vigabatrin should take into consideration the general time frame of onset of efficacy vs that of risk of onset of pVFDs. The benefit–risk analysis for continued therapy with vigabatrin for patients who respond to vigabatrin treatment is an ongoing process that must include appropriate vision monitoring at regular intervals to minimize the risk of pVFDs. The decision to continue vigabatrin therapy should be made on an individual basis. Continuation of therapy is an appropriate option if the benefits of treatment continue to outweigh the risk of vision loss.
Conflicts of interest
John M. Pellock is a paid member of the Sabril Registry steering committee for Lundbeck Inc. He has also served on scientific advisory boards for Lundbeck, and as a consultant and research investigator.
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.