The gold standard for evaluating the safety and efficacy of an antiepileptic drug (AED) is the randomized, double-blind, placebo-controlled clinical trial. Few drugs—felbamate, lamotrigine, topiramate, and intravenous immunoglobulin (IVIG)—have undergone this rigorous testing to determine their safety and efficacy in patients with Lennox-Gastaut syndrome (LGS). Felbamate, lamotrigine, and topiramate successfully demonstrated efficacy against seizures in patients with LGS. Results for IVIG did not prove effectiveness, reinforcing the need for randomized, double-blind, controlled trials to establish definitively the efficacy of any proposed therapy.
Felbamate was found to be safe and effective in patients with LGS in a randomized, double-blind, placebo-controlled adjunctive therapeutic trial.103 Seventy-three patients with LGS, ages 4–36 years, were enrolled. The felbamate dose in the double-blind portion was 45 mg/kg per day (maximum, 3,600 mg per day). The felbamate treatment group experienced a 34% reduction in atonic seizures, as compared with 9% in the placebo group (p = .01). Total seizure frequency dropped 19% in the felbamate group, as compared with a 4% increase in the placebo group (p = .002). The percentage of patients experiencing at least a 50% reduction in atonic seizures and total seizure frequency was significantly higher for the felbamate group than for the placebo group (p <.001) (See Table: Responder Rates). Felbamate also was significantly better than placebo in improving global evaluation scores.103 During a 12-month follow-up period in patients who completed the controlled part of the study, long-term efficacy was confirmed.104
Despite its efficacy, felbamate is regarded as a good third-line or fourth-line drug for LGS because it is associated with significant risks from idiosyncratic reactions. These involve the formed elements of the blood and the liver. The most common severe felbamate-associated idiosyncratic reaction is aplastic anemia,105 the incidence of which is approximately 127 cases per million treated with felbamate (approximately 1 in 4,000 to 8,000 treated patients), as compared with 2 to 2.5 cases per million persons in the general population.80,105,106 Another report estimates the risk of aplastic anemia in patients receiving felbamate to be 1 in 3,000, with a death rate of 1 in 10,000 felbamate-treated patients.80,107 This estimated risk is roughly up to 20 times greater than that for carbamazepine-associated aplastic anemia.105
The second most common severe felbamate-associated idiosyncratic reaction is hepatotoxicity. Its estimated incidence is between 64 and 164 cases per million (approximately 1 in 18,500 to 25,000 felbamate-treated patients).105 This rate is roughly the same as for valproic acid–associated hepatotoxicity.105
A suggested management strategy is to employ careful clinical monitoring and routine laboratory testing and to discontinue felbamate use if no substantial clinical benefit is observed after 3 to 6 months of therapy.
Lamotrigine’s efficacy against seizures associated with LGS was examined in multiple open-label studies and two controlled trials. In five open-label trials of lamotrigine in patients with LGS, 58% (31 of 53) experienced at least a 50% reduction in seizure frequency.109–113 A double- blind, placebo-controlled, crossover study of lamotrigine as adjunctive therapy in 30 patients with treatment-resistant generalized epilepsy (20 of whom had LGS) was reported in 1998. Of the 20 children with LGS, 7 responded to lamotrigine therapy with at least a 50% reduction in seizure frequency, and two became seizure-free.114
Lamotrigine proved to be safe and effective in patients with LGS in a randomized, double-blind, placebo-controlled, adjunctive therapeutic trial in which 169 patients enrolled and were randomized to either lamotrigine (n = 79) or placebo (n = 90) adjunctive therapy.115 Patients on the lamotrigine treatment arm had a greater median percentage reduction from baseline in weekly seizure counts (for drop attacks, tonic-clonic seizures, and all major seizures [defined as drop attacks plus tonic-clonic seizures]) as compared with patients on the placebo treatment arm. The responder rate (percentage of patients experiencing at least a 50% reduction in seizures) for major seizures (drop attacks and tonic-clonic seizures) was greater in the lamotrigine group than in the placebo group.115 (See Table: Responder Rates)
Despite its efficacy, lamotrigine can be associated with idiosyncratic reactions, predominantly involving the skin. The most common skin manifestation is a rash affecting 10–12% of lamotrigine-treated patients.80,116,117 The rash rapidly resolves when lamotrigine is withdrawn, and sometimes it may even resolve without changing the lamotrigine dosage.118 In some patients, however, it can progress to erythema multiforme, Stevens-Johnson syndrome, or even toxic epidermal necrolysis.80,116,118 The mortality rates for these related severe mucocutaneous disorders are less than 5% for Stevens-Johnson syndrome and 30% for toxic epidermal necrolysis.116 The risk of a potentially life-threatening rash (based on clinical trials and postmarketing reports) is 0.3% in adults and approximately 1% in children 16 years of age and younger.118
Despite the risk of idiosyncratic reactions, lamotrigine is a very valuable medication for patients with LGS and should be considered for use as soon as the diagnosis of LGS is made. Proper attention to concomitant medications, a low starting dose, and a very slow titration rate can minimize the risk of dermatologic idiosyncratic reactions. The prompt evaluation of any rash is prudent.
Topiramate was found to be safe and effective as adjunctive therapy for patients with LGS in a multicenter, double-blind, placebo-controlled trial.119 Ninety-eight patients with LGS (ages 1+ to 30 years) were randomized to either topiramate adjunctive therapy (target dose, 6 mg/kg per day) or placebo adjunctive therapy. The median percentage reduction from baseline in average monthly seizure rate for drop attacks was 14.8% for the topiramate group and –5.1% (an increase) for the placebo group (p = .04). Using parental global evaluations, topiramate-treated patients demonstrated greater improvement in seizure severity than did placebo-treated patients (p = .04). The responder rate for major seizures (drop attacks and tonic-clonic seizures) was greater in the topiramate group (15 of 46, or 33%) than in the control group (4 of 50, or 8%; p = .002). The responder rate for drop attacks in the topiramate group was higher than in the placebo group (28% vs. 14%) but did not reach statistical significance.119 (See Table: Responder Rates)
In the long-term, open-label extension portion of the preceding trial, all but 1 of the patients were followed up, and their topiramate dose was adjusted as clinically indicated.120 The mean topiramate dosage in those patients who had completed 6 months of open-label therapy was 10 mg/kg per day. For those patients who had completed 6 months of open-label topiramate therapy, drop attacks were reduced by at least 50% in 55% of patients (as compared to 28% during the double-blind portion of the trial), and 15% of patients were free of drop attacks for at least 6 months at the last visit. In this open-label extension phase, the median percentage reduction in drop attacks was 56% (as compared to 14.8% during the double-blind portion of the trial). The median percentage reduction in overall seizure frequency was 44%, with 45% of the patients having at least a 50% reduction in all seizure types and 2% being seizure-free for the previous 6 months.
The most common adverse events were somnolence, injury, and anorexia. Only 5% of the patients reported behavioral problems during the last 6 months of topiramate long-term therapy. Therefore, topiramate was well tolerated during long- term therapy and was effective in controlling drop attacks and seizures associated with LGS.120
The double-blind, placebo-controlled adjunctive therapeutic trials for patients with LGS show efficacy and safety for felbamate, lamotrigine, and topiramate. Which is better? Because no comparative trial of AEDs has been performed in patients with LGS, the best method of resolving the issue is a meta-analysis of these three trials.
A clinically useful measure of treatment effect of a study medication is the number needed to treat (NNT).121 This calculated number represents the number of patients that a clinician must treat with a study medication in order to find one with the desired outcome. The NNT is calculated as 1/(A – P), where A is the percentage of responders in the active treatment arm and P is the percentage of responders in the placebo treatment arm. This represents the inverse of the absolute risk reduction.121 In clinical trials of AEDs involving patients with treatment-resistant epilepsy, the desired outcome is usually at least a 50% reduction in seizure frequency. A patient who achieves this reduction is called a responder. The 95% confidence interval (CI) can then be calculated.
To date, only trials involving adults with treatment-resistant partial seizures have been considered in other meta-analyses of AEDs.122 The NNT can be calculated using the felbamate, lamotrigine, and topiramate double-blind placebo-controlled adjunctive therapy trials described earlier. In each study, drop attacks (atonic and tonic seizures) were used as the seizure outcome variable, because they are the most debilitating seizure type. The calculated NNT for each drug is:
|(95% CI, 1.4-3.8)||(95% CI, 3.4-54.8)||(95% CI, 3.3-1,000)|
Although felbamate had a lower NNT than the other two agents, the confidence intervals overlap. The very large confidence intervals shown for lamotrigine and topiramate are due to variability in the results. More studies are needed before it is possible to conclude that there will be an observable clinical difference among these three AEDs in terms of efficacy against drop attacks.
Reviewed and revised May 2004 by Steven C. Schachter, MD, epilepsy.com Editorial Board.
© 2014 Epilepsy.com. All rights reserved.