Antiseizure medications for idiopathic generalized epilepsies: a systematic review and network meta-analysis

Historically, idiopathic generalized epilepsies (IGEs) have included juvenile myoclonic epilepsy (JME), childhood absence epilepsy (CAE), juvenile absence epilepsy (JAE), and generalized tonic–clonic seizures alone (GTCA) [1]. The term “idiopathic” refers to “self-originating” or “spontaneously arising” and implies that the condition is genetic [2]. Characterized by 2.5–6 Hz generalized spike waves, IGEs account for approximately 15–20% of all epilepsies, as well as 55% of newly diagnosed generalized epilepsy in children and adolescents [3, 4]. Because IGEs have a strong underlying genetic basis, the updated classification from the International League Against Epilepsy suggested that IGEs should be a subgroup of “genetic generalized epilepsies,” but reserved the term to describe the aforementioned four overlapping syndromes [1].

The diagnosis of IGEs has important implications for their treatment and prognosis. In patients with IGEs, development, neurological examinations, and radiographic results are typically normal [5]. Additionally, because most cases arise in children and adolescents, IGEs are often emphasized to be of pediatric importance only; however, considerable psychosocial symptoms—such as mood disorders, attention deficits, and learning disabilities—can be observed until adulthood [6, 7]. Long-term follow-up studies have revealed correlations between IGEs and outcomes such as poorer employment/financial conditions, decreased interactions with families, and unplanned pregnancies [8]. Thus, although IGEs may seem easier to manage than symptomatic or partial epilepsies, they should not receive less attention than these other epilepsies.

Antiseizure medications (ASMs) are the cornerstone of treatment for IGE syndromes. A good response rate, at 60–80% of seizure control (i.e., more than 1 year without seizure), can be achieved with appropriate ASM selection [9, 10]. First-line monotherapy controls symptoms in the majority of patients with IGEs. Among the first-line treatments, valproate (rather than lamotrigine or topiramate) monotherapy is the recommended first choice for IGEs in boys and men because it was shown to have better efficacy and tolerability in the SANAD study (Level I evidence) [11]. By contrast, levetiracetam monotherapy is favored in women able to have children; it seldom induces drug–drug interactions. Although levetiracetam was inferior to valproate in the SANAD II study [12], it has shown good efficacy in seizure control in cohort studies [13]. The choice of optimal initial monotherapy in IGEs is very important. Management decisions are different for JME, CAE, JAE, and CTSA, and need to be individualized. However, a limited number of randomized controlled trials (RCTs) have compared various ASMs head-to-head as initial monotherapies for IGEs and related subsyndromes. For example, some ASMs, such as carbamazepine or oxcarbazepine, may exacerbate absence seizures, whereas lamotrigine and gabapentin can exacerbate some myoclonic seizures [14]. A comprehensive integration of the evidence is thus needed so that a tailored plan can be developed for each patient.

Adjunctive therapy should be started when two different monotherapies have been unable to successfully control IGEs. The drug of choice generally depends on the main seizure subtype. Lamotrigine and levetiracetam are recommended as adjunctive therapies to valproate, except in JME. Topiramate, zonisamide, and perampanel have also demonstrated efficacy in RCTs or observational studies and are recommended adjunctive options (e.g., in myoclonic seizures). However, the majority of adjunctive medications do not have proof of efficacy in placebo-controlled RCTs. Hence, a comparison of the efficacy and tolerability of adjunctive ASMs remains lacking; clinically, the choice of adjunctive drug often relies on class III or IV evidence. Furthermore, with numerous established and new medications currently available, physicians face difficult decisions when choosing the most appropriate adjunctive drugs because of the limited high-quality evidence [15].

To the best of our knowledge, no previous review has compared the efficacy and tolerability of ASMs for IGEs (neither as monotherapies nor as adjunctive therapies). This network meta-analysis (NMA) aims to provide comprehensive evidence for the relative efficacy and safety of ASMs for controlling IGEs.

This NMA was conducted following a protocol that was prospectively registered with PROSPERO (No. CRD42022372358) and adhered to the Preferred Reporting Items for Systematic Review and Meta-Analysis (PRISMA) statement for network meta-analysis [16].

Our NMA indicated that all of the included ASMs were more effective than the placebo. The network forest plots compared monotherapies with valproate and adjunctive therapies with placebo. Significant superiority was identified for adjunctive levetiracetam and topiramate, while inferiority was identified for lamotrigine monotherapy. Non-significant differences were also identified. Using SUCRA, rankings of efficacy and tolerability were summarized.

The analyses of the efficacy outcomes of being seizure free for 3–6 or 12 months did not affect the status of valproate as the first-choice monotherapy for overall IGEs without contraindications. Although ethosuximide ranked first according to SUCRA, its optimal efficacy and tolerability are probably only favorable for absence epilepsies rather than for overall IGEs, and especially tonic–clonic seizures [49]. It is supported as the drug of choice for absence seizures without other seizure types, in accordance with other reviews and guidelines (April 2022, NICE guidelines, https://​www.​nice.​org.​uk/​guidance/​ng217).

Lamotrigine monotherapy unexpectedly had the lowest efficacy in both short- and long-term seizure-free outcomes in the SUCRA analysis and showed significant inferiority in the forest plots for overall IGEs. Although a longer duration of titration may partly account for this short-term result (because lamotrigine must be titrated very slowly to avoid unwanted side effects), this finding is in accordance with a SANAD study suggesting that lamotrigine should not be interpreted as a “broad spectrum” antiseizure medication because it provides worse seizure control than valproate or topiramate in generalized epilepsies [11]. However, these findings should be interpreted with caution. Adjunctive lamotrigine is advantageous in controlling unclassified generalized tonic–clonic seizures (i.e., those unable to be classified as IGEs or partial epilepsy) [50].

In recent years, levetiracetam has been increasingly prescribed and recommended. The present NMA supports its efficacy as a second-line monotherapy and as an ideal adjunctive choice in overall IGEs according to its efficacy ranking. Although it was not found to be a non-inferior monotherapy to valproate in a previous study [12], its favorable efficacy, fast action, and good tolerability (leading to less TEAEs than placebo in our ranking) indicate its considerable potential. However, longitudinal studies are needed in the future (for both levetiracetam and brivaracetam).

In the present study, adjunctive topiramate ranked first in adjunctive therapies according to SUCRA. As a second-generation ASM, topiramate is especially effective in JME and GTCA [51]. In a Cochrane review, the efficacy of topiramate monotherapy in JME was not significantly different from that of valproate (the current drug of choice) [51]. Although topiramate is associated with cognitive TEAEs such as dulling or memory problems [52], which are especially unfavorable in neurodevelopmental disorders, its tolerability was the best ranked in terms of TEAEs leading to discontinuation.

The head-to-head comparison of third-generation ASMs used as adjunctive therapies is of great importance because there is a lack of accumulated evidence, especially for perampanel and lacosamide. In the present analysis, SUCRA demonstrated that their efficacies seem to fall between those of levetiracetam and lamotrigine. In addition, lacosamide may be more effective than perampanel for seizure-free outcomes in GTCA. Although there was previously a lack of high-quality long-term evidence, recent unblinded controlled studies have revealed that perampanel reduces generalized seizures with a median of 90% in 52-week follow-up, and has the potential to increase seizure freedom [42, 53, 54]. Moreover, after 24 weeks of lacosamide adjunctive treatment, the freedom rate from generalized seizures was 27.5% (versus placebo 13.2%) in an RCT [44].

In the current study, perampanel was the best-ranked therapy for tolerability as both a monotherapy and an adjunctive therapy when any adverse event was considered. The characteristic TEAEs of perampanel are irritability and aggression [52, 55]. In contrast, when ranking the therapies in terms of serious TEAEs leading to discontinuation, perampanel was inferior to placebo and levetiracetam. Similarly, the PERMIT study indicated a discontinuation rate of 17.6% at 12 months, in which psychiatric TEAEs were the most common reason for discontinuation [56]. We thus predict a future in which new-generation ASMs, like perampanel, are used to control generalized seizures. However, more high-quality research is warranted to draw stronger conclusions.

We must note that, although all included ASMs significantly improved the seizure-free rate compared with placebo, ASMs can neither cure epilepsies nor treat the underlying pathology that causes them; they merely aim to stop the occurrence of seizure symptoms. A better understanding of the molecular mechanisms underlying the pathogenesis, epileptogenesis, and pharmacoresistance of epilepsies is needed to change our clinical approach. For example, genetic therapies and stem cell therapies will likely cure epilepsies in the future [57].

Despite this, the importance of ASMs should not be ignored, although many patients do not achieve or retain complete seizure freedom. Improved seizure reduction may significantly downgrade the risk of injury and unexpected death [58]; however, in the present review, the NMA of the seizure reduction rate had to be stopped because insufficient data were provided in the included studies. In addition, although some well-controlled complex epilepsies might be disrupted by a single poor night’s sleep or missing dose and breakthrough, substantial improvements have been achieved [59]. Compared with invasive options like vagus nerve stimulation or corpus callosotomies, the use of established and new ASMs may provide more tolerable, incremental benefits. Furthermore, the increase in available ASMs makes it possible to devise more individualized plans, thus benefitting patients. Longitudinal comprehensive studies are therefore warranted to evaluate efficacy in particular populations (or genotypes), more in-detail tolerability, effects on quality of life, and cost-utility for ASMs.

The present study had some limitations. Methodologically, a limited number of outcomes restrained us from analyzing other important efficacy outcomes, such as seizure reduction or electroencephalogram improvements. Furthermore, because specific TEAEs were not evaluated, the tolerability outcome analysis lacked details, and only rough results were obtained because of a lack of information. The search strategy mainly focused on the idiopathic generalized seizure type. Thus, some important ASMs (such as cenobamate, brivaracetam, etc.) most frequently used in focal epilepsies, although recently proven adjunctive use in generalized seizures, were not involved. More meaningful future studies are necessary to elucidate the efficacy of these ASMs.

Although low heterogeneity was identified according to the I² test, differences between RCTs existed such as the inclusion criteria, time of treatment, and concomitant drugs. Furthermore, although a statistically suitable and well-known analysis was used, certain overestimations or underestimations may still exist. For example, relatively broad 95% CIs were obtained because relatively few RCTs were included. Further studies are therefore required to further confirm our conclusions.

Among the included ASM monotherapies, valproate ranked best for overall IGEs in efficacy and was the third best in tolerability. For the adjunctive therapies, topiramate ranked best for GTCA and overall IGEs, whereas levetiracetam ranked best for myoclonic seizures. Moreover, perampanel ranked best in tolerability measured by any TEAE when used either as a monotherapy or an adjunctive therapy. Overall, valproate is recommended as the monotherapy of choice for overall IGEs without contraindications. However, our results should be interpreted with caution considering the limited available information and the inherent methodological limitations of the NMA.