Vigabatrin therapy for refractory complex partial seizures: review of major European trials

Authors


E. Ben-Menachem, MD, PhD, Department of Clinical Neuroscience, Institute for Clinical Neuroscience and Physiology, Sahlgrenska Academy, Göteborg University, 413 45 Göteborg, Sweden
e-mail: Elinor.Ben-Menachem@neuro.gu.se

Abstract

Ben-Menachem E, Sander JW. Vigabatrin therapy for refractory complex partial seizures: review of major European trials.
Acta Neurol Scand: 2011: 124 (Suppl. 192): 16–28.
© 2011 John Wiley & Sons A/S.

Complex partial seizures (CPS) are a form of localization-related seizures associated with serious comorbidities and risks. CPS can be difficult to treat and may remain refractory to treatment with antiepileptic drugs (AEDs). Refractory CPS (rCPS) can be hazardous because of the potential for severe dysfunction and bodily harm, sometimes with fatal consequences. Control of seizure activity is critical to the clinical management of CPS. Vigabatrin is a unique AED approved in both Europe and the United States as adjunctive therapy for adult patients with rCPS who have responded inadequately to several alternative treatments. This review focuses on appropriately controlled studies of vigabatrin conducted in Europe. Several double-blind studies randomized those with rCPS to treatment with vigabatrin vs placebo, and two evaluated durability of response to long-term, open-label vigabatrin. Endpoints included seizure frequency, treatment satisfaction, and adverse events (AEs). Efficacy outcomes demonstrated that vigabatrin add-on therapy significantly reduced the frequency of seizures. Long-term studies indicated durability of response and tolerability of vigabatrin therapy for up to several years. Treatment satisfaction data indicated a preference for vigabatrin vs placebo for both physicians and study participants. Vigabatrin was well-tolerated with generally mild AEs considered common to AEDs. Vision effects were not formally monitored in these studies. In European trials, vigabatrin was efficacious as adjunctive therapy for rCPS.

Introduction

Complex partial seizures (CPS) are a form of localization-related epilepsy that causes those who suffer from them to experience abnormal consciousness. Despite the availability of several antiepileptic drugs (AEDs) in many combinations, more than 30% of those with CPS continue to experience refractory partial seizures (1). Refractory complex partial seizures (rCPS) are associated with serious comorbid conditions and risks, including greater risk of bodily injury (2), shortened lifespan (2, 3), and sudden unexpected death (4–7). Injury rates for individuals with rCPS can be substantial, from one per 20- to one per 3-person-years (2). Mortality rates for these patients are increased up to seven-fold compared with the general population (2, 4). Neuropsychologic and psychiatric complications include decreased quality of life, clinical depression, and an increased risk of suicide (2).

Because the health-related consequences of rCPS are serious and can be severe, therapeutic efforts to prevent and control of seizure activity are essential. If seizure control cannot be achieved with existing AED regimens, add-on therapies that may improve seizure control are needed.

Vigabatrin is a unique AED used as add-on therapy for patients with rCPS. Available in Europe starting in the 1980s, vigabatrin was the first novel AED introduced since valproate in the 1970s (8). Vigabatrin is now available in more than 50 countries (9). Clinical studies conducted in Europe during the 1980s demonstrated vigabatrin to be an efficacious and well-tolerated therapeutic agent in treating intractable seizures, including long-term treatment of chronic rCPS (8).

A structural analog of γ-aminobutyric acid (GABA), vigabatrin, is γ-vinyl GABA and is thought to act as an enzyme-activated irreversible inhibitor of GABA transaminase, the enzyme responsible for the GABA metabolism. The downstream effect of vigabatrin action is a marked increase in the concentration of the inhibitory neurotransmitter GABA in the central nervous system (CNS), which in turn decreases seizure activity (9, 10). The neurochemistry and mechanism of action of vigabatrin are described in greater detail in a separate review (11).

Vigabatrin was approved in several European countries in 1989 as adjunctive therapy for rCPS in adults. It is also approved in Europe as primary treatment of infantile spasms (IS) in children aged 2 years and younger. This indication is discussed in a separate review (12). Vigabatrin was approved in the United States in 2009 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. US indications are discussed in separate reviews (12, 13). The current Summary of Product Characteristics (SmPC) (14) indicates (via a special warning) that because of the risk of visual field defect, “vigabatrin should only be used after a careful assessment of the balance of benefits and risk compared with alternatives.” Further, the SmPC notes that “vigabatrin is not recommended for use in patients with any pre-existing clinically significant visual field defect,” and that “patients should undergo systematic screening examination when starting vigabatrin and at regular intervals for detection of visual field defects. Visual field testing should continue at 6 month intervals for the whole duration of treatment” (14).

The results of several well-controlled European clinical studies, including several long-term, open-label, follow-up extension studies, on vigabatrin efficacy for those with rCPS are the focus of this review. Vision effects were not formally monitored in these studies.

Review of well-controlled studies in Europe

Seven studies of vigabatrin add-on therapy for rCPS conducted in Europe that included well-controlled study designs are described later (15–21). A summary of the study designs and key outcomes for each of the double-blind, placebo-controlled studies is provided in Table 1. Outcomes for key efficacy and safety endpoints for each study are provided. Results for a key measure of AED effectiveness, ≥ 50% decrease in seizure frequency, are provided for each double-blind, placebo-controlled study in Fig. 1. Summary data from open-label extensions of double-blind studies are also provided (21, 22).

Table 1.   Summary of double-blind, placebo-controlled study designs
 Study
Rimmer & Richens (15) (= 24)Gram et al. (16) (= 21)Loiseau et al. (17) (= 23)Tartara et al. (18) (= 23)Tassinari et al. (19) (= 31)McKee et al. (20) (= 24)Grünewald et al. (21) (= 45)
  1. AEDs, antiepileptic drugs.

  2. aDuration of double-blind treatment only. This does not include the open-label extension study that followed.

  3. bDosage was dependent on patient’s weight.

  4. cThose randomized to receive vigabatrin were administered 2 g/day for 6 weeks followed by 3 g/day for an additional 6 weeks within each treatment period.

  5. dNumber of participants who entered the study provided first (with the number who completed the study given parenthetically).

Study designCrossoverCrossoverCrossoverCrossoverCrossoverCrossoverParallel group
Duration of treatment, weeks91297a121220a
Dosage, g/day3332 or 3b2 or 3b2–3c3
Concomitant AEDs, N≤3≤3≤3≤2≤4≤2≤3
Participants enrolled (completed) study, Nd24 (21)21 (18)23 (19)23 (20)31 (30)24 (21)45 (45)
Efficacy and safety outcomes50% seizure decreases, adverse effects50% seizure decreases, adverse effects50% seizure decreases, adverse effects50% seizure decreases, adverse effects50% seizure decreases, adverse effects, withdrawal from study50% seizure decreases, adverse effects50% seizure decreases, neuropsychological testing, adverse effects, withdrawal from study
Figure 1.

 Percentage decrease in seizure frequency. The percentages of patients who had achieved ≥75% or ≥50% decreases in seizure frequency during vigabatrin therapy are provided for each of the double-blind, placebo-controlled studies (15–21). The denominators represent the number who completed the respective study and who were thus included in efficacy analysis. For studies with open-label extensions (18, 21), only outcomes from the double-blind period of the study are provided.

Double-blind, placebo-controlled crossover study of vigabatrin for rCPS: Rimmer & Richens (15)

The randomized, double-blind, placebo-controlled crossover study enrolled 24 participants aged 16–61 years with poorly controlled epilepsy, primarily rCPS with or without secondary generalization (Table 2) (15). Study participants with a mean duration of epilepsy of 21 years were having an average of 5.8 seizures per week despite receiving ≤3 concomitant AEDs (most were receiving two). The study was 22 weeks and included a 4-week baseline period to assess seizure frequency and two treatment periods of 9 weeks each, in which vigabatrin 3 g/day (given as 1.5 g twice daily) or matching placebo was administered without tapering or washout between periods (e.g., treatments were switched abruptly). Seizure frequencies and types were recorded in patient diaries, as were adverse events (AEs). Electroencephalograms (EEGs) were recorded at baseline and during treatments. Compliance was measured by capsule counts. Laboratory analyses were conducted for vigabatrin and concomitant AED serum concentrations during treatment periods, and routine blood chemistry and hematologic analyses were performed. At the end of the study, participant treatment preference and observer preference (based on seizure frequency and severity, AEs, and overall impression of treatments) were obtained.

Table 2.   Baseline patient characteristics
Baseline variableStudy
Rimmer & Richens (15) (= 24)Gram et al. (16) (= 21)Loiseau et al. (17) (= 23)Tartara et al. (18) (= 23)Tassinari et al. (19) (= 31)McKee et al. (20) (= 24)Grünewald et al., (21) (= 45)
  1. AEDs, antiepileptic drugs; NA, not available.

  2. Note: Number of participants (N) for each study represents the number for whom baseline data were collected at study entry.

  3. aData on mean or median age or duration of epilepsy were not reported for some studies.

  4. bPercentages of participants with specific seizure types were not reported. However, the study’s authors reported inclusion of mainly complex partial seizures, with or without secondary generalization.

  5. cSome data were not available for those who discontinued the studies. Hence, in some instances, the sum of n-values may not equal N.

  6. dSome were counted more than once as having had multiple seizure types based on data collection methods. Hence, the sum of n-values is greater than N.

  7. eCategories are exclusive (i.e., patients are counted only once in a single category describing the largest number of AEDs concomitantly used).

  8. fPercentages of those receiving one up to a maximum of three AEDs were not reported. However, the study’s authors reported that the median was two concomitant AEDs.

Age, yrs, mean (range)33 (16–61)NA (17–63)aNA (10–58)aNA (17–50)a28.9 (10–58)NA (17–53)aNA (15–61)a
Male, n (%)9 (37.5)11 (52.4)7 (30.4)13 (56.5)16 (51.6)8 (33.3)24 (53.3)
Epilepsy duration, years, mean (range)21 (6–45)26 (8–47)13 (2–40)NA (2–42)aNA (2–42)aNA (4–43)aNAa
Seizure type, n (%)
 Complex partialNAb21 (100.0)19 (82.6)17 (73.9)15 (50.0)c22 (91.7)44 (97.8)d
 Simple partialNAb0 (0.0)0 (0.0)3 (13.0)0 (0.0)c0 (0.0)35 (77.8)d
 Atonic, complex partialNAb0 (0.0)0 (0.0)0 (0.0)8 (26.7)c0 (0.0)0 (0.0)
 Tonic, tonic–clonicNAb0 (0.0)3 (13.0)0 (0.0)0 (0.0)c2 (8.3)0 (0.0)
 OtherNAb0 (0.0)1 (4.3)3 (13.0)7 (23.3)c0 (0.0)14 (31.1)d
Concomitant AEDs, n (%)e
 Monotherapy7 (29.1)NAf2 (8.7)3 (13.0)2 (6.7)11 (45.8)26 (57.8)
 ≥2 AEDs13 (54.1)NAf14 (61.0)20 (87.0)9 (30.0)13 (54.2)18 (40.0)
 ≥3 AEDs4 (16.7)NAf6 (26.1)0 (0.0)18 (60.0)0 (0.0)1 (2.2)
 ≥4 AEDs0 (0.0)0 (0.0)0 (0.0)0 (0.0)2 (6.7)0 (0.0)0 (0.0)

Results —  Three of 24 participants withdrew from the study. One withdrew because of acute confusion during the first week of vigabatrin therapy, and two withdrew because of protocol non-compliance. The 21 individuals who completed the study experienced a significantly lesser mean number of seizures overall on vigabatrin therapy (3.54 seizures) compared with placebo (6.23 seizures) and vs baseline (6.14 seizures; < 0.001). Twenty of 21 study participants (95%) had fewer seizures during vigabatrin therapy than during placebo treatment. Of these, 14 (70.0%) had a ≥50% decrease in seizure frequency, and 6 (30.0%) had a ≥75% decrease in seizures frequency (Fig. 1).

When only complex partial seizures were considered, vigabatrin was also associated with a significant decrease in seizure frequency (< 0.001). Analysis of only tonic–clonic seizures (occurring in 11 of 21 patients) also demonstrated a significant decrease in seizure frequency on vigabatrin therapy (< 0.05). The effect of vigabatrin on seizure reduction was rapid, within the first week of therapy, and was sustained throughout the 9-week treatment period. Only one patient had a marked increase in seizure frequency upon discontinuation of vigabatrin, presenting with an organic confusion state. Otherwise, in general, discontinuing vigabatrin did not significantly impact seizure frequency or tolerability.

Vigabatrin therapy was associated with more AEs vs placebo overall (Table 3). Drowsiness and mood changes occurred more frequently with vigabatrin vs placebo (P < 0.01). Drowsiness tended to resolve over time.

Table 3.   Incidence of commonly reported adverse events
Adverse eventaStudy
Rimmer & Richens (15) (= 21)Gram et al. (16) (= 18)Loiseau et al. (17) (= 19)Tartara et al. (18) (= 21)Tassinari et al. (19) (= 31)McKee et al. (20) (= 24)Grünewald et al. (21) (= 45b)
PBOVGBPBOVGBPBOVGBPBOVGBPBOVGBPBOcVGBcPBOVGB
  1. PBO, placebo; VGB, vigabatrin.

  2. Data are reported as number of study participants who experienced the adverse event during the VGB or PBO therapy (patients may have been counted twice if the same event was recorded during each therapy in a crossover study). Data include those who dropped out of the respective study or had a dosage decrease because of one or more adverse events.

  3. aSome adverse events were not reported for the studies. Although the events may have occurred, they were not included among commonly reported events. Therefore, ‘–’ in this table means not reported.

  4. bAdverse events reported for the double-blind phase of the study only.

  5. cData shown here represent combined numbers for 2 and 3 g/day dosing arms of the study.

Drowsiness2724171018
Fatigue/tiredness43201011034
Dizziness1112032313
Irritability1210210134
Ataxia01010201
Headache1112132241
Depression140224
Constipation111115
Weight gain1342
Impaired memory1102
Diplopia111123
No adverse events1461091210

The majority of participants selected vigabatrin as their treatment preference (16 of 21 patients, 76.2%). Similarly, observers favored vigabatrin (19 of 21 individuals, 90.5%).

Antiepileptic drugs concentrations remained unchanged during the study except for phenytoin, which had significantly lesser concentrations during vigabatrin treatment for 10 participants (mean, 40 μmol/l) vs placebo (mean, 59 μg/ml) and compared with baseline (mean, 51 μmol/l; P < 0.05). No significant changes were noted in laboratory investigations or EEGs.

Conclusions —  Vigabatrin demonstrated rapid and sustained efficacy in reducing seizure frequency for those with rCPS and was generally well-tolerated. The most common AEs, drowsiness and mood changes, tended to subside over time. Withdrawal seizures that occurred following discontinuation of vigabatrin suggested that tapering may be helpful.

Double-blind, placebo-controlled crossover study of vigabatrin for rCPS: Gram et al. (16)

This was a randomized, double-blind, placebo-controlled crossover study that enrolled 21 participants aged 17–63 years with rCPS with or without secondary generalization (Table 2) (16). Enrolled individuals had a median duration of epilepsy of 26 years and were experiencing ≥4 seizures per month despite treatment with ≤3 concomitant AEDs (median, two AEDs). The 28-week study included two treatment periods of 12 weeks each, during which vigabatrin 3 g/day (given as 1.5 g twice daily) or matching placebo was administered with no taper or washout between periods (e.g., treatments were abruptly crossed). Only data collected in the last 8 weeks of each treatment period were analyzed to avoid carryover effects. After the second treatment period, all participants entered a 4-week single-blind placebo period. Seizure frequency and type were reported by patients. AEs and weights recorded were reported at clinic visits. Evoked potentials were recorded after each treatment period and routine laboratory assessments (chemistry, hematology) were performed during the study. At study conclusion, study participant and investigator treatment preferences were recorded.

Results —  Of the 21 enrolled, 18 completed the study. One was unable to accurately record seizure frequency, and two discontinued because of AEs. The median number of seizures was significantly less with vigabatrin therapy (nine seizures; range, 1–49) than that of placebo (16 seizures; range, 0–37; < 0.05). The median number of CPS also decreased significantly with vigabatrin (eight seizures; range, 0–47) compared with placebo (14 seizures; range, 0–36; < 0.05). Of the 18 participants who completed the study, eight (44.4%) had a ≥50% decrease, and three (16.7%) had a ≥75% decrease in seizure frequency with vigabatrin therapy (Fig. 1). Moreover, two had a moderate and one had a marked increase in seizure frequency on vigabatrin compared with placebo.

Similar numbers of AEs were reported for vigabatrin and placebo (Table 3). The most common AE was fatigue, reported in approximately equal numbers for both treatments. No significant weight changes occurred with vigabatrin therapy (> 0.10). Two individuals discontinued the study because of AEs (fatigue, dizziness, ataxia, and vomiting) within several days of starting vigabatrin treatment. Treatment was stopped spontaneously by one participant, and symptoms subsequently subsided. Upon restarting vigabatrin, the symptoms reoccurred.

Eight of 18 participants (44.4%) favored vigabatrin treatment, but the majority (55.6%) were unable to select a preference. The high percentage unable to choose a preferred treatment may have been influenced by memory impairment experienced as an AE. Investigators preferred treatment with vigabatrin for 12 of 18 participants (66.7%), preferred placebo for two others (11.1%), and had no preference for four individuals (22.2%).

Serum concentrations of vigabatrin revealed no correlation with efficacy or AEs, and there were no significant changes in AED concentrations during either treatment. Evoked potentials and laboratory investigations did not show changes correlating with vigabatrin therapy.

Conclusions —  Vigabatrin decreased seizure frequency and was generally well-tolerated in those with rCPS. Approximately half of participants in this study experienced no AEs, and there was no significant difference in the occurrence of AEs between vigabatrin and placebo treatments for the remaining patients. Two discontinued the study because of AEs during vigabatrin therapy. One of these had received vigabatrin 2 g/day in a previous single-blind study, with no unwanted AEs. After discontinuing the 3-g/day dosage for the present study, the patient was restarted on 2 g/day and was able to tolerate the lower dosage. In addition, the investigators suggested that both of the discontinued patients may have experienced lesser tolerability to vigabatrin because of their advanced ages (62 and 63 years). Because vigabatrin is excreted largely unmetabolized in the urine, elderly individuals with decreased renal clearance abilities may be more susceptible to toxicity. No vigabatrin and AED interactions were noted in this study, in contrast to the previous double-blind trial by Rimmer & Richens (15).

Double-blind, placebo-controlled crossover study of vigabatrin for rCPS: Loiseau et al. (17)

Twenty-three study participants aged 10–60 years who had rCPS with or without secondary generalization or generalized tonic or tonic–clonic seizures were enrolled in a randomized, double-blind, placebo-controlled study with a crossover design (Table 2) (17). Participants with a mean duration of epilepsy of 13.4 years were having ≥1 seizure per week while treated with ≤3 concomitant AEDs (most patients were receiving two AEDs). The duration of the study was 25 weeks. A 5-week observation period preceded each treatment period, during which only pre-existing AEDs were given at constant dosages. Treatment periods were 10 weeks each. For the first week of each treatment period, participants received vigabatrin 1.5 g/day or placebo (given as vigabatrin 1.5 g or placebo, plus placebo, twice daily). Dosages were increased to 3 g/day (given as two doses of vigabatrin 1.5 g or placebo) of vigabatrin or matching placebo during the remaining 9 weeks. The final 5 weeks of the study comprised a single-blind, placebo, add-on treatment period designed to control for potential carryover effects for those who had received vigabatrin in the second double-blind crossover treatment period. Seizure frequencies and types were recorded in participants’ calendars, and AEs were reported to the clinic. At the end of each study period, personal preference (based on perceived efficacy and comfort) and a double-blind neurologist’s global efficacy rating (based on seizure frequency and severity and AEs) were obtained. AED plasma concentrations were monitored at study visits, and routine serum chemistry and hematology laboratory tests were conducted.

Results —  Nineteen of the 23 enrolled participants completed the study. Three discontinued the study because of AEs and one patient discontinued because of non-quantifiable seizure frequency. Vigabatrin was associated with a significant decrease in seizure frequency during the 9-week treatment with vigabatrin 3 g/day or placebo. The decrease in seizure frequency was significantly greater with vigabatrin (6.33 seizures) than with double-blind placebo (10.7 seizures) or single-blind placebo (9.72 seizures; < 0.01). No period effect or treatment by period interaction was noted. Thus, there were no observed carryover or withdrawal effects from vigabatrin therapy. Of the 19 who completed the study, 11 (57.9%) had a ≥50% decrease in seizure frequency (Fig. 1). Two had a 25–50% decrease in seizure frequency, four had unchanged seizure frequency, and two had an increase in the occurrence of seizures. The two participants with tonic–clonic seizures had 64% and 100% decreases in seizure frequency, respectively. Decreases in seizure frequency were generally observed within the first 2 weeks of treatment with vigabatrin 3 g/day, and four patients had decreased seizure activity within the first week of therapy (with vigabatrin 1.5 g/day).

Similar numbers of participants reported AEs during vigabatrin and placebo treatments (Table 3), and most AEs were considered mild. Nine experienced AEs during double-blind vigabatrin therapy, seven had AEs during double-blind placebo treatment, and three had AEs in the single-blind placebo period. The most common AE during vigabatrin therapy was drowsiness. One individual required a dosage decrease because of an AE, and three participants were discontinued from the study because of AEs. One of these had increased seizure frequency after crossover from vigabatrin therapy to placebo. The other two were withdrawn from the study because of intolerance of vigabatrin after 12 and 45 days of treatment, respectively.

Most study participants (14 of 19, 73.7%) preferred vigabatrin over placebo. Three indicated a preference for placebo, and two did not indicate a preference. Neurologists rated efficacy as greater during vigabatrin therapy vs placebo for 15 individuals (78.9%; < 0.05). One received equal ratings for vigabatrin and placebo treatments, two had better efficacy ratings during placebo therapy, and one had no rating.

Mean plasma concentrations of AEDs demonstrated no significant differences between treatment periods, and no within-patient trends were observed. Overall, no significant changes were noted in laboratory investigations. One participant, however, had decreased hemoglobin without concomitant changes in red blood cell count during vigabatrin therapy. This was considered to be either a laboratory error or a non-serious condition.

Conclusions —  Vigabatrin 3 g/day was generally well-tolerated and efficacious in decreasing seizure frequency for those with rCPS. Some patients achieved decreases in seizure frequency during the first week of treatment with the lower dosage of vigabatrin (1.5 g/day). AEs were mostly mild and similar between treatments. No effects of vigabatrin on AED concentrations were noted in this study.

Double-blind, placebo-controlled crossover study of vigabatrin for rCPS: Tartara et al. (18)

This study was a randomized, double-blind, placebo-controlled trial with a crossover design. It enrolled 23 individuals aged 16–65 years with drug-resistant epilepsy. Those who entered the study had only one seizure type and presented with rCPS with or without secondary generalization, simple partial seizures with secondary generalization, atonic seizures, or absence of seizures (Table 2) (18). Study participants had epilepsy for 2–42 years and were experiencing ≥1 seizure per week despite treatment with one or two concomitant AEDs (most patients were receiving two AEDs). The study was 14 weeks, with two treatment periods of 7 weeks each, in which vigabatrin 2 or 3 g/day (given as 1 or 1.5 g twice daily) or matching placebo was administered without taper or washout in between (e.g., treatments were crossed abruptly). Dosing was based on weight, with patients weighing ≤ 65 kg receiving vigabatrin 2 g/day and patients weighing >65 kg receiving 3 g/day. Seizure frequencies and types were recorded in participants’ calendars. EEGs and evoked potentials were recorded at baseline and the end of each treatment period. Electrocardiograms and serum chemistry, hematology, and urine laboratory tests were performed during the study. Concomitant AED serum concentrations were measured at each study visit. At the end of the study, treatment preference and observer preference (based on seizure frequency and severity, AEs, and overall impression of the treatments) were obtained.

Results —  Of the 23 enrolled, 20 completed the study. Three discontinued the study. One patient discontinued because of a missed study visit, one because of non-compliance with seizure frequency recording, and one because of several AEs. Compared with placebo, vigabatrin was associated with a significant decrease in total seizure frequency (2.2 vs 3.8 seizures per week, < 0.01) and partial seizure frequency (2.0 vs. 3.7 seizures per week, < 0.01). Of the 20 participants who completed the study, 12 (60.0%) had a >50% decrease in seizure frequency, and four (20.0%) had a >75% decrease in seizures frequency (Fig. 1). Three had a 25–50% decrease in seizure frequency, one had a <25% decrease, and four had an increase in seizure frequency. Ten of 17 study participants (58.8%) with partial seizures had a >50% decrease in seizure frequency.

Most AEs reported were mild and transient. The most common AE during vigabatrin therapy was drowsiness (Table 3). One individual discontinued the study because of several AEs (ataxia, dysarthria, vertigo, and headache) that occurred after 4 weeks of treatment with vigabatrin 3 g/day. Several positive effects were recorded. “Improved sense of well-being” was reported by six participants during vigabatrin therapy and by one during placebo. In addition, two receiving vigabatrin reported “feeling more alert” vs none during placebo treatment.

Antiepileptic drug concentrations were unaffected during the study with the exception of decreased mean serum phenytoin concentration for one participant during vigabatrin therapy (6.2 μg/ml) compared with study baseline (7.4 μg/ml) and double-blind placebo treatment (8.9 μg/ml). EEGs, electrocardiograms, and laboratory investigations demonstrated no significant changes.

Conclusions —  Vigabatrin was efficacious as an add-on therapy in reducing the frequency of seizures by >50% in the majority of those enrolled and by >75% in a subset of those participants. Most (15 of 20, 75%) who completed the study and were analyzed for seizure frequency were receiving the lower dosage of vigabatrin (2 g/day) based on the weight-adjusted dosing. AEs were generally mild and expected for vigabatrin therapy. Several events suggested positive effects of vigabatrin. This may have been related to improved seizure control or some independent effect of vigabatrin. Patients who completed this study were eligible to continue in a long-term, open-label, follow-up study, described later.

Long-term, open-label study of responder patients: Tartara et al. (22)

This open-label extension (22) was a long-term follow-up study enrolling those who had completed the preceding double-blind, placebo-controlled study (17) or another single-blind study of vigabatrin therapy. Eligible participants were responders, defined as having a ≥50% decrease in seizure frequency, or clinically important improvement in quality of life consequent to decreased seizure frequency or severity. Thirteen who had completed the preceding double-blind study continued in the extension study, as well as 12 who had completed a single-blind study. The single-blind study included a 6-month treatment period with vigabatrin 2–4 g/day (with dosages adjusted per clinical response). In total, 25 participants aged 18–58 years (median, 31 years) with CPS (= 20), simple partial seizures (= 2), atonic seizures (= 1), tonic–clonic seizures (= 1), or absence seizures (= 1) were enrolled and continued long-term therapy for up to 47 months.

The extension study included 13 men and 12 women. All were receiving one (= 5) or two (= 20) concomitant AEDs. Participants were maintained on open-label vigabatrin 2–4 g/day, with dosage adjustments per clinical response. The median duration of vigabatrin therapy for all patients in the long-term extension study was 22 months (range, 1–47 months). For the 24 individuals who entered the open-label extension directly without discontinuation of initial vigabatrin therapy, this includes the initial treatment period. Evaluations were every 3 months at clinic visits and included seizure frequency recorded in participants’ diaries; AEs; serum chemistry, hematology, and urine laboratory tests; and evoked potentials performed every 6 months.

Results —  Eight discontinued the study. One withdrew because of ataxia (which resolved after discontinuation of vigabatrin); one, after 16 months because of psychotic symptoms; three, after 6, 14, and 18 months, respectively, because of increased seizure activity; and three, after 4, 8, and 15 months, respectively, for reasons unrelated to treatment. The duration of treatment for all those who remained in the study was 14–47 months (median, 22 months). Within the first 2 months of open-label vigabatrin therapy, patients had a >50% decrease in seizure frequency compared with previous blinded placebo treatment, with this effect sustained for the duration of the study. Maintenance of efficacy for 2 years was influenced by bias, as several individuals dropped out. However, analysis of data for those who had received vigabatrin for ≥14 consecutive months confirmed the treatment effect.

The most common AEs reported were generally mild and included weight gain (3–11 kg) (= 10), sedation (= 4), headache (= 4), epigastric pain/nausea (= 4), weakness (= 3), anxiety/irritability (= 3), vertigo (= 3), dysarthria (= 2), and confusion (= 2). Four participants reported “improved mood.” Two dropped out because of AEs: one because of ataxia and one owing to psychotic symptoms (as described previously). Evoked potentials remained unchanged for somatosensory testing. However, some brainstem auditory potential latencies did decrease slightly and P100 latencies of visual evoked potentials that were abnormally prolonged for some patients before vigabatrin therapy approached normalization during therapy.

Mean serum AED concentrations were unchanged, except for a significant decrease in phenytoin concentration (from a mean 13.8 μg/ml before long-term vigabatrin therapy to a mean 9.4 μg/ml during this extension study) for five patients.

Conclusions —  A decrease in seizure frequency >50% was observed with long-term vigabatrin therapy and was durable for up to 4 years. AEs were mostly mild and included many common effects associated with AEDs and vigabatrin. The most common AE was weight gain, which occurred in nearly half of study participants. The increase in weight tended to occur after 3–6 months of treatment and reached a plateau within a few months afterward. Vigabatrin was well-tolerated by most in this study and provided clinical efficacy over a period of months to years for the majority of participants.

Double-blind, placebo-controlled crossover study of vigabatrin for rCPS: Tassinari et al. (19)

This was a randomized, double-blind, placebo-controlled crossover study of 31 participants aged 10–58 years who had rCPS with or without secondary generalization, associated with atonic seizures, or various partial seizures (Table 2) (19). Those in the study had a median duration of epilepsy of 26 years and were experiencing ≥4 seizures per month despite treatment with ≤4 concomitant AEDs (most were receiving three AEDs). On the basis of EEG and other clinical data, study participants were stratified into two groups for efficacy analysis: Group 1 included those with rCPS and temporal EEG abnormalities, and Group 2 included all others who had various seizure types (i.e., atonic, elementary partial, absences, and secondary generalized) and multifocal EEG abnormalities or diffuse slow discharges with focal abnormalities. The 9-month study included an initial run-in period of 8 weeks for treatment with constant dosages of current AEDs. Patients were then treated for 12 weeks with vigabatrin 2 or 3 g/day (given as 1 or 1.5 g twice daily) or matching placebo administered without taper or washout between crossing over (e.g., treatments were switched abruptly). Dosing was based on weight, with those weighing ≤60 kg receiving vigabatrin 2 g/day and those weighing >60 kg receiving 3 g/day. Following the second treatment period, all entered a 4-week single-blind placebo period. Seizure frequency and type were reported by participants, and weights and AEs were recorded at clinic visits. EEGs were recorded at baseline and after each blinded treatment period. Laboratory analysis of concomitant AED serum concentrations during treatments, as well as serum chemistry and hematology laboratory tests, was conducted. At study conclusion, participants’ and investigators’ treatment preferences were recorded.

Results —  Of 31 who were enrolled, 30 completed the study. One was withdrawn because of leucopenia that developed after 6 weeks of treatment with vigabatrin 2 g/day, but which was presumed to be related to concomitant trimethadione therapy. Seizure frequency was reduced, but not significantly, for the overall study sample. Ten participants (33.3%) had a ≥50% decrease in frequency (Fig. 1), six (20.0%) had a ≥25% decrease, and five (16.7%) had a <25% decrease. Analysis of Groups 1 and 2 indicated a differential response to vigabatrin. Those in Group 1 (= 15) had a significant decrease in seizure frequency, with six (40.0%) achieving a ≥50% decrease and 11 (73.3%) attaining a ≥25% decrease. For Group 2 (N = 15), none had a ≥50% decrease, and five (33.3%) had a ≥25% decrease in seizure frequency (not significant).

In the overall sample, 12 of the 16 who had ≥25% decreases in seizure frequency on vigabatrin had an increase in seizure frequency to pretreatment concentrations within 3 weeks of vigabatrin discontinuation. The other four who had seizure frequency decreases also had episodes of quasicontinuous seizures within days after crossing over from vigabatrin to placebo, suggesting a potential withdrawal effect.

The most common AEs were drowsiness and irritability (Table 3). AEs were mostly mild and tended to occur during the first weeks of treatment. One participant discontinued because of an AE (leucopenia), as described previously.

Thirteen participants (43.3%) favored vigabatrin, six (20.0%) preferred placebo, and 11 (36.7%) were unable to choose between the treatments. Investigators preferred vigabatrin therapy for 20 individuals (66.7%), which included 12 from Group 1 and eight from Group 2 (80.0% and 53.3%, respectively). Placebo was preferred for three individuals (10.0%) overall. No preference was indicated for seven participants (23.3%).

There were no significant changes in EEG activity, weight, or laboratory investigations. One exception was a single patient who had decreased hemoglobin, hematocrit, and red blood cells. However, these were present throughout the study, across treatments, and she had reported menorrhagia during vigabatrin treatment.

Antiepileptic drug concentrations were unaltered in the study, with the exception of phenytoin. The mean phenytoin concentration for 19 study participants decreased significantly from baseline: 21.1 μg/ml to 14.8 μg/ml during double-blind vigabatrin therapy, compared with 18.7 μg/ml during double-blind placebo and 18.65 μg/ml during single-blind placebo (< 0.01).

Conclusions —  The overall study group achieved a decrease in seizure frequency by ≥50% on vigabatrin therapy, but the subset of those with rCPS and temporal EEG abnormalities had the greatest response. Overall, vigabatrin 2 or 3 g/day was efficacious for those with rCPS and was well-tolerated by most participants in the study. Investigators’ and participants’ preferences were primarily for vigabatrin, although a substantial percentage of participants were unable to choose between treatments. This is similar to patient preference outcomes in another study (16), in which the authors suggested that memory dysfunctions may have played a role in the inability to select which treatment was favored.

Double-blind, placebo-controlled crossover study of vigabatrin for rCPS: McKee et al. (20)

This randomized, double-blind, placebo-controlled crossover study enrolled 24 participants aged 17–53 years who had rCPS with or without secondary generalization or generalized tonic–clonic seizures (Table 2) (20). Participants had a mean duration of epilepsy of 21 years and were having ≥2 seizures per week despite receiving up to one or two concomitant AEDs. The study was 22 weeks and included a 4-week run-in period to assess seizure frequency. The two subsequent treatment periods were 12 weeks each, separated by a 4-week washout period. For the first 6 weeks of each treatment period, patients received vigabatrin 1.5 g/day or placebo (given as vigabatrin 1.5 g or placebo plus placebo twice daily). Dosages were increased to vigabatrin 3 g/day or matching placebo during the remaining 6 weeks. Seizure frequency and type were recorded in participants’ diaries, and AEs were reported at clinic visits. Compliance was measured by medication counts. Laboratory analysis was performed for vigabatrin and concomitant AED serum concentrations during treatment periods.

Results —  Twenty-one of 24 participants completed the study. Two withdrew consent and one patient discontinued because of a delusional psychotic episode. Two were excluded from analyses for inadequate records of seizures. The overall occurrence of seizures was significantly reduced by vigabatrin therapy compared with placebo (< 0.05). The decrease in seizure frequency with vigabatrin vs placebo was significant for the vigabatrin 2 g/day dosage (median 10 vs. 16 seizure days, < 0.05), but not for the 3 g/day dosage (median 20 vs. 24 seizure days). Subgroup analysis indicated a similar result for partial seizures and no effect on generalized tonic–clonic seizures at either dosage of vigabatrin. Overall, eight of 19 (42.1%) achieved a ≥50% decrease in seizure frequency compared with placebo (Fig. 1). This effect was more prevalent during vigabatrin 2 g/day (= 9) compared with vigabatrin 3 g/day (= 6). The same trend was evident for the subset of patients with rCPS. Three of 12 patients (25.0%) with generalized tonic–clonic seizures had a ≥50% decrease in seizure frequency.

Common AEs were neurologic and occurred within the first 6 weeks (Table 3). There were no significant changes in mean concentrations of AEDs.

Conclusions —  Vigabatrin was efficacious and well-tolerated at dosages of 2 and 3 g/day but was most efficacious at the lower dosage. This did not appear to be a result of poor compliance during treatment with the greater dosage, which was confirmed by medication counts, and long-term efficacy data from other studies suggest that development of tolerance to vigabatrin is not likely (21, 22).

Double-blind, placebo-controlled parallel-group study of vigabatrin for rCPS: Grünewald et al. (21)

This was a randomized, double-blind, placebo-controlled parallel-group study of 45 individuals aged 15–61 years who had simple partial seizures, rCPS with or without secondary generalization, or secondary generalized seizures (Table 2) (21). Study participants had onset of epilepsy at 10–11 years of age on average and were experiencing refractory partial seizures despite treatment with up to three concomitant AEDs (most were receiving one to two AEDs). The double-blind study was 28 weeks, followed by up to 18 months of optional open-label treatment with vigabatrin. An 8-week baseline permitted assessment of seizure frequency and neuropsychologic assessment. Participants were then randomized to the 20-week treatment period to receive vigabatrin or placebo. For the first 2 weeks of treatment, individuals received vigabatrin 1.5 g/day or placebo (given as vigabatrin 1.5 g or placebo, plus placebo, twice daily). Dosages were increased to vigabatrin 3 g/day or matching placebo during the remaining 18 weeks. After the double-blind treatment period, patients were offered open-label treatment with vigabatrin 3 g/day without breaking the blind. Those who had received double-blind placebo received 2 g/day vigabatrin for 2 weeks and then increased to 3 g/day. Seizure frequency and type were recorded in participants’ diaries. Neuropsychologic assessments were repeated at the end of the baseline and treatment periods and during open-label therapy. AEs were recorded, and laboratory assessments (chemistry, hematology) were performed at clinic visits.

Results —  The median seizure frequency for CPS decreased by 66% for those in the vigabatrin group (= 22) between 4 and 12 weeks of treatment and was significantly better than the 50% decrease in the placebo group (= 23, = 0.009). By 12–20 weeks of therapy, the vigabatrin group had achieved a 69% decrease in CPS frequency compared with a 25% increase in the placebo group (= 0.001). Decreases in simple partial and secondary generalized seizures were not significantly different between treatments at both time points. Nine of 22 (40.9%) in the vigabatrin group had a ≥50% decrease in CPS frequency (Fig. 1) between 4 and 12 weeks, and 10 of 22 (45.5%) had a decrease during Weeks 12–20, compared with baseline. In comparison, four of 23 (17.4%) in the placebo group had a ≥50% decrease in CPS frequency during Weeks 4–12 (not significant), and three of 23 (13.0%) had a decrease during Weeks 12–20 (P = 0.016).

Deterioration of seizure control, with a seizure frequency increase of >50%, was less common for patients receiving vigabatrin, occurring for only one patient during Weeks 4–12 and for none during Weeks 12–20. In contrast, in the placebo group, seven of 23 (30.4%) had an increase of >50% in seizure frequency in the analysis periods: Weeks 4–12 (P = 0.023) and Weeks 12–20 (= 0.005).

A total of 43 entered the vigabatrin open-label treatment period of the study, including 20 from the double-blind vigabatrin group and 23 from the placebo group. Ten of those from the vigabatrin group who were responders (with a ≥50% decrease in seizure frequency) continued at 3 g/day. Three withdrew from open-label therapy for other therapy (= 2) or surgery (= 1). At 44 weeks after beginning vigabatrin therapy, six of seven patients continued to improve or maintained improvement. One experienced a slight deterioration in seizure activity.

Common AEs included neurologic events, weight gain, constipation, and mild depression (Table 3). Two in the vigabatrin group experienced severe depressive symptoms and withdrew within 4 weeks of starting treatment. Both patients improved within 4 weeks of discontinuing vigabatrin therapy.

Neuropsychologic testing indicated a significant decrease in motor function, with decreased dominant hand tapping frequency and significant deterioration in the memory design learning task on vigabatrin therapy. There was no significant worsening during open-label therapy. No significant vigabatrin-related changes were observed for measures of mood or behavior for those who tolerated double-blind treatment. During open-label treatment, some mood scores demonstrated significant improvement for vigabatrin responders compared with non-responders: depression, = 0.01; fatigue, = 0.03; and aggression, = 0.04.

There were no significant interactions with AEDs or findings from laboratory investigations. Minor modifications in AED regimens were necessary for two patients in the vigabatrin group and two in the placebo group during long-term treatment.

Conclusions —  In this controlled parallel-group study, vigabatrin 3 g/day significantly decreased seizure frequency vs baseline and vs placebo. Those who were vigabatrin responders (with ≥50% decrease in seizure frequency) maintained treatment effect for up to 44 weeks in the long-term, open-label extension to the study. Treatment was generally well-tolerated, although several study participants developed potentially serious depressive symptoms during double-blind treatment, and one did during open-label treatment. The symptoms subsided after discontinuation of vigabatrin therapy. Neuropsychologic testing results were generally consistent with the effects of vigabatrin and other AEDs.

Summary of efficacy results

In these well-controlled, double-blind studies, vigabatrin add-on therapy reduced the frequency of seizures in those with rCPS receiving ≤4 concomitant AEDs. All studies demonstrated a significant percentage of patients achieving ≥50% decrease in seizure frequency, except for one study (19), which demonstrated significance only within a subgroup of those with CPS (vs other seizure types). Three of the studies also demonstrated a significant percentage with ≥75% decreases in seizure frequency (15–17). Freedom from seizures was not reported in these studies.

Treatment preference was assessed in four studies. In two of these (15, 17), participants’ and physicians’ treatment preferences significantly favored vigabatrin vs placebo. In the other two (16, 19), a significant percentage of physicians favored vigabatrin, but the percentages of those who preferred vigabatrin did not constitute a majority nor were they significant. Personal treatment preferences in both of these studies were potentially confounded by a sizeable percentage of those who were unable to choose between vigabatrin and placebo, instead selecting neither. The inability to choose a treatment was suggested by Gram et al. (16) to be related to memory impairment observed during the course of the study. Participants may have had poor recall of their impressions of the two add-on therapies (vigabatrin and placebo) received during the crossover study.

Neuropsychologic assessment in one study (21) suggested no significant change over time for measures of mood or behavior for those who had tolerated vigabatrin therapy. Vigabatrin therapy was associated with modest, but significant changes in motor function and memory compared with placebo. These effects are generally recognized as associated with vigabatrin and other AEDs.

Summary of safety experience

The percentages of study participants experiencing commonly reported AEs during vigabatrin and placebo therapy in well-controlled studies are shown in Fig. 2. The most commonly reported AEs were mild and CNS-related, including drowsiness, dizziness, fatigue, ataxia, and headache. Psychological AEs were also generally mild, including irritability and depression. Vigabatrin is one of several AEDs associated with the occurrence of depression (23). Vigabatrin-associated depression is typically mild and is responsive to dosage adjustments, tapering, or discontinuation of vigabatrin therapy (24).

Figure 2.

 Percentages of patients experiencing common adverse events during vigabatrin and placebo therapy in well-controlled studies. The total number who experienced the most commonly reported adverse events were pooled across double-blind, placebo-controlled studies (15–21) and are represented as percentages of the pooled population (= 179). For studies with open-label extensions (18, 21), only adverse events reported during the double-blind period of the study are shown. PBO, placebo; VGB, vigabatrin.

Other AEs reported that are common to AEDs in general were weight gain and impaired memory. These may have been a result of add-on vigabatrin therapy or the combination of concomitant AEDs. Weight gain is known to occur more often in those with epilepsy treated with vigabatrin therapy vs placebo. Several AEs of diplopia were reported in two studies (Rimmer & Richens [15]; Gram et al. [16]), and a few isolated cases of blurred vision were reported (data not shown).

The majority of AEs observed in these studies were considered mild. This may be a result, at least in part, of the use of lower dosages (i.e., 1.5–2 g/day) of some participants during these early studies. Currently, 3 g/day is considered therapeutically optimal for most patients, and dosage-dependent AEs may be impacted accordingly.

A more detailed overview of safety findings in studies on vigabatrin is presented in a separate review (25). Importantly, peripheral visual field defects (pVFDs) are now known to be a risk with long-term vigabatrin use. In these well-controlled studies conducted in Europe, vision was not formally monitored. More detailed overviews of vision safety issues (26) and visual field testing methods (27) are provided in separate reviews.

Conclusions

Vigabatrin is an efficacious add-on therapy approved in Europe and the United States for the treatment of rCPS. Several well-controlled studies of rCPS conducted in Europe have demonstrated efficacy of vigabatrin in reducing the frequency of seizures. Vigabatrin was well-tolerated in these studies, with mostly mild AEs reported and few study discontinuations because of AEs. pVFDs were not observed, which is likely a consequence of the natural latency of the defects that have been since realized. Several long-term studies of up to 4 years provided evidence of the durability of vigabatrin efficacy, with long-lasting tolerability and sustained decreases in seizure frequency. Vigabatrin is an appropriate add-on treatment option for those who achieve substantial clinical benefit given the severe consequences of rCPS, although ongoing benefit-risk assessments are critical to ensure that the benefits of treatment outweigh the risk of vision loss.

Conflicts of interest

Elinor Ben-Menachem has received consultancy fees, research grants, or speaker fees from Eisai, GlaxoSmithKline, Janssen-Cilag, Johnson & Johnson, Lundbeck Inc., and UCB. Josemir W. Sander has received consultancy fees or research grants from Eisai, GlaxoSmithKline, and UCB.

Acknowledgments

Medical writing and editorial assistance were provided by Jennifer N. Bodie, PhD, 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.

Ancillary