Efficacy and safety of adjunctive antiseizure medications for dravet syndrome: A systematic review and network meta-analysis

Abstract

Introduction

Dravet syndrome (DS) is a treatment-resistant, developmental, and genetic form of epileptic encephalopathy (Mei et al., 2019; Lo et al., 2021). The main features of DS include an early onset, various forms of seizures, delayed psychomotor development, and drug resistance (Dravet, 2011). Dravet syndrome progression is divided into three stages: heating, deterioration, and stability (Wheless et al., 2020). The overall incidence of DS is approximately 1/40,900–1/22,000, accounting for approximately 29.5% of various types of myoclonic epilepsy in children (Brunklaus et al., 2012; Bayat et al., 2015; Wu et al., 2015; Bjurulf et al., 2022). Studies have found that DS is related to mutations in the SCN1A gene, which encodes the NaV1.1 channel, and more than 85% of patients have mutations in this gene (Claes et al., 2001). These changes make the channels of inhibitory interneurons less functional, which can overexcite neurons, leading to seizures (Catterall et al., 2010). There are many treatment options for DS (Ziobro et al., 2018). However, antiseizure medication treatment remains the mainstay treatment for DS (Chin et al., 2021). Valproic acid and clobazam are the first-line drugs for treating DS according to the practical guidelines for treating DS with anti-seizure medication (Strzelczyk and Schubert-Bast, 2022). Medications that may exacerbate seizures, including sodium channel inhibitors (e.g., carbamazepine, oxcarbazepine, lamotrigine, and phenytoin) and the γ-aminobutyric acid (GABA) transaminase inhibitor vigabatrin, should be avoided (Wirrell et al., 2017). However, most patients with DS exhibit drug resistance, which leads to deficient treatment effects on epilepsy (Devi et al., 2021). In such cases, multiple drugs must be used simultaneously. Recently, the U.S. Food and Drug Administration (FDA) approved medications for DS, which included stiripentol, fenfluramine, and cannabidiol (Sharawat et al., 2021). Moreover, it was found that soticlestat, a new drug used to treat DS, showed promising efficacy and safety in a recently completed phase 2 clinical trial (Strzelczyk and Schubert-Bast, 2022). These drugs act through different mechanisms. Stiripentol affects the metabolism of other antiseizure medications such as clobazam by enhancing central γ-aminobutyric acid transmission and inhibiting several P450 cytochromes (Devi et al., 2021). Cannabidiol reduces neuronal excitability through inhibition of adenosine transport and modulation of intracellular Ca2⁺(Gray and Whalley, 2020). Fenfluramine enhances the effects of 5-hydroxytryptamine to exerts antiepileptic activity (Sourbron et al., 2017). Fenfluramine is also a positive regulator of the sigma-1 receptor (Martin et al., 2020). Soticlestat is a cholesterol 24-hydroxylase inhibitor that can reduce the level of 24S-hydroxycholesterol (24HC) (Hawkins et al., 2021; Koike et al., 2021). There is no direct head-to-head study on the efficacy and safety of these four drugs, and clinical decision-making largely depends on the availability of drugs in different regions, patient characteristics and needs, and personal preferences, making it difficult for clinicians to choose the best treatment method (Zhang et al., 2022). Therefore, this study aimed to compare and evaluate the efficacy and safety of currently available DS treatment medications by performing a network meta-analysis of randomized controlled trials (RCTs), and to evaluate the long-term efficacy and safety of antiseizure medications by performing a meta-analysis of open-label extension (OLE) studies, to provide a reference for the selection of clinical treatment medications for DS.

Materials and mehthods

The conduct and reporting of this study followed the Preferred Reporting Items for Systematic Reviews (Page et al., 2021) and the PRISMA Extension Statement for Reporting of Systematic Reviews Incorporating Network Meta-analyses (Hutton et al., 2015), using standardized protocols for review and data extraction. Furthermore, they were registered in PROSPERO. The primary outcome measure for efficacy was the number of patients with a ≥50% reduction in seizure frequency from the baseline and a nearly 100% reduction in seizure frequency from the baseline. Seizures included tonic, epileptic spasms, tonic-spasmodic, and atonic seizures (Golub and Reddy, 2021). The primary safety outcome measures were the incidence of adverse events (AEs) and serious adverse events (SAEs). Adverse reactions mainly include somnolence, decreased appetite, and diarrhea (Lagae et al., 2019; Lattanzi et al., 2021). The effect size was defined as the risk ratio (RR) and prevalence of events. A meta-analysis and a network meta-analysis of efficacy and safety indicators were performed according to the frequentist theory and a consensus model.

Result

Discussion

Dravet syndrome is a rare, severe form of hereditary epilepsy. It has attracted much attention in epileptic encephalopathy and is one of the most clinically challenging epilepsy syndromes (Selvarajah et al., 2021). The persistent threat of intractable seizures, multiple comorbidities, and premature death severely affects the quality of life of children and their families (Mudigoudar et al., 2016; Lagae et al., 2018). Epilepsy mortality in DS is 5.1 times higher than that in adults with refractory epilepsy (Harris et al., 2015). The main causes of increased mortality in patients with DS are status epilepticus, seizure-related complications, and sudden epileptic death (Sharawat et al., 2021). Currently, many treatment options are available for DS. The three primary forms of treatment available are antiseizure medications, dietary modifications (usually a ketogenic diet), and surgical intervention (Chin et al., 2021). Nevertheless, antiseizure medications remain the mainstream treatment for epilepsy (Chin et al., 2021). The British National Institute of Clinical Practice guidelines and Chinese epilepsy diagnosis and treatment guidelines recommend valproic acid, topiramate, and/or clobazam as first-line treatment drugs for DS; stiripentol, levetiracetam, and zonisamide can be used as additional therapeutic drugs, while sodium channel blockers are not recommended for the treatment of DS (Nunes et al., 2012). Because DS has no specific symptoms in the early stage, it is often misdiagnosed as other types of epilepsy, and the application of sodium channel inhibitors (carbamazepine, oxcarbazepine, and lamotrigine) and benzodiazepines may promote encephalopathy and worsen the condition (Strzelczyk and Schubert-Bast, 2022). Despite reasonable multidrug therapy and ketogenic diets, approximately 45% of children experience more than four seizures per month due to limited drug options for DS (Wheless et al., 2020). Therefore, there is an urgent need to identify new drugs for DS treatment. In recent years, with the development of large-scale clinical trials, significant breakthroughs have been made in the development of DS treatment drugs. The U.S. FDA approved stiripentol and cannabidiol in 2018, and fenfluramine in 2020 for the treatment of DS (Lattanzi et al., 2018; Yıldız et al., 2019; Sharawat et al., 2021). In 2021, it was found that soticlestat, a new drug for the treatment of DS, showed good safety and effectiveness in the treatment of DS.

A series of recently published randomized controlled trials reveals growing interest in the role of stiripentol, fenfluramine, cannabidiol, and soticlestat in DS treatment. Two double-blind, randomized, placebo-controlled trials of stiripentol as an adjuvant therapy for valproic acid and clobazam in France [STICLO-France (Chiron et al., 2000)] and Italy [STICLO-Italy (Hahn et al., 2022)] showed that DS patients over 3 years of age showed better responses to these drugs than placebo in both the ≥50% reduction in seizure frequency and complete control of seizures. The initial dose of stiripentol was 15–20 mg/(kg d), which was increased to a target dose of 50 mg/(kg d) in two to 4 weeks. The maximum dose for children can be as high as 100 mg/(kg d). The main adverse reactions included drowsiness, fatigue, ataxia, appetite loss, and liver damage (Jullien et al., 2015).

The Nordic Cochrane Centre, T.C.C, 2020, Devinsky et al., 2017 [GWPCARE1 (Devinsky et al., 2017)] and Miller et al., 2020 [GWPCARE2 (Miller et al., 2020)] had investigated the efficacy and safety of cannabidiol for the treatment of DS. The results showed that the effective rate of cannabidiol in patients with refractory DS was significantly higher than that in the placebo group (42.6–49.3% vs. 26.2–27.1%), while in the subgroup that used clobazam at the same time, the effective rate was higher (47.5–62.5% vs. 23.7–36.6%). The initial dose of cannabidiol was 2–5 mg/(kg·d), and the maximum dose was 25 mg/(kg·d), according to attack and tolerance. Adverse reactions included drowsiness, fatigue, diarrhea, and anorexia (Devinsky et al., 2015).

In recent years, two double-blind, randomized controlled trials [Lagae 2019 (Lagae et al., 2019) (without the concurrent use of stiripentol) and Nabbout 2020 (Nabbout et al., 2020) (with the concurrent use of stiripentol)] have confirmed the effectiveness and safety of fenfluramine in the treatment of DS. The percentage of patients with seizures that decreased by ≥50% was significantly higher than that of the placebo group (68–38% vs. 5–12%). The recommended dose of fenfluramine is 0.12–0.90 mg/(kg·d) (Ceulemans et al., 2012). Low-dose fenfluramine is generally well-tolerated in the treatment of DS. The most common adverse reactions were loss of appetite and weight loss. Other common adverse reactions include diarrhea, fatigue, lethargy, fever, and upper respiratory tract infections (Strzelczyk and Schubert-Bast, 2022).

Soticlestat (≤600 mg/day weight-adjusted) is a new drug used for the treatment of DS. It showed better efficacy and safety in a recently completed phase 2 clinical trial, Elektra. Recruitment to phase 3 clinical trials is ongoing. The main adverse reactions were lethargy and constipation (Hahn et al., 2021).

With the advent of these RCTs, different mechanisms of action, different administered doses, and AEs have led to difficulties in the selection of clinically appropriate drugs. Therefore, it is important to determine the efficacy and safety of these four antiseizure medications (stiripentol, fenfluramine, cannabidiol, and soticlestat) in DS patients. Thus, the current study included the seven studies described above and a network meta-analysis of the efficacy and safety of these drugs in the hope of providing a reference for clinicians when choosing treatments.

The drug comparison demonstrated that all four antiseizure medications resulted in a more significant reduction in convulsive-seizure frequency than the placebo among patients with DS. Soticlestat was the most efficacious adjunctive therapy, followed by fenfluramine, and stiripentol. However, since the soticlestat sample size was very small, this result needs further verification. We hope that soticlestat will achieve good results in future clinical trials. Stiripentol was the only medication with a statistically significant difference in the incidence of AEs compared to the placebo. In addition, because stiripentol has the highest probability of AEs and SAEs, fenfluramine is a better choice for DS treatment. This result is consistent with the conclusions of Devi et al. (2021). There were no significant differences in the effects of the four antiseizure medications in achieving complete control of epilepsy.

However, clinical decisions depend on all available information, and the RCTs included in our study were only short-term trials, which would affect the accuracy of the conclusions. To further determine the efficacy and safety of these four antiseizure medications, we included post-commercial studies or open-label RCTs that provided important information on long-term efficacy and safety for a meta-analysis. In this study, soticlestat was included one article, while each of stiripentol, fenfluramine and cannabidiol were included three articles. Due to the small sample size of soticlestat and the lack of corresponding efficacy outcome indicators, the long-term efficacy and safety of soticlestat need further study. In addition to soticlestat, the comprehensive safety and efficacy can be concluded that fenfluramine is the first choice, followed by stiripentol. This is consistent with our conclusion.

In addition, it is also important to consider drug-drug interactions between antiseizure medications. Because it will affect the efficacy and safety of antiseizure medications. When stiripentol is combined with fenfluramine, the maximum plasma concentration and systemic exposure of fenfluramine are significantly increased, so the dose of fenfluramine can be appropriately reduced (Boyd et al., 2019). Valproic acid does not impact on stiripentol pharmacokinetics (PK), and some data suggest that clobazam may increase stiripentol concentrations by around 25% (Peigné et al., 2018). Studies have shown that fenfluramine had no significant effect on the PK of valproic acid, stiripentol, clobazam, or nor-clobazam (the major and active metabolite of clobazam), and thus no dose adjustments are needed for these commonly prescribed anticonvulsant medications when administered with fenfluramine (Balagura et al., 2020). When clobazam is coadministered with cannabidiol, the area under the curve (AUC_0-t) of the active cannabidiol metabolite 7-hydroxy-cannabidiol increases 50% (Franco and Perucca, 2019). It has already been demonstrated that concomitant clobazam affects cannabidiol’s safety profile and raises the likelihood of adverse events, particularly somnolence, drowsiness, and pneumonia (Lattanzi et al., 2020). Therefore, dose reductions of clobazam may be considered based on a benefit-risk assessment (Morrison et al., 2019). Based on PK alone, dose adjustments are not likely to be necessary when cannabidiol is given concomitantly with stiripentol or valproic acid (Morrison et al., 2019). Coadministration of stiripentol with cannabidiol resulted in a slight increase in stiripentol exposure, but this is unlikely to be clinically relevant, so dose adjustment of stiripentol may not be necessary. Cannabidiol had no significant effect on the PK of valproic acid (Morrison et al., 2019). So far, the clinical relevance of these interactions remains largely unknown, and so will have some impact on our findings.

To date, epilepsy control in DS remains a great challenge, and no single drug treatment is highly effective for this syndrome (Wirrell, 2016). Our study showed that soticlestat is likely to be a better adjunctive antiseizure medication in terms of combined efficacy and safety. However, the accuracy of the results needs to be verified further because of the insufficient number of patients evaluated. In addition, owing to the lack of RCTs for a direct comparison, the present analysis only provides indirect comparisons from methodological RCTs, which may differ from reality. Therefore, while our findings provide a reference for clinicians, in practical clinical applications, clinicians need to consider each patient and their situation individually and weigh the advantages and disadvantages of each treatment before administering drugs. In other words, the ranking of treatments in this study is based on the mean SUCRA value but does not imply that a higher-ranked treatment is substantially better than a lower-ranked treatment. Differences in the number of trials per comparison and effect sizes between treatments probably affected ranking probabilities. Furthermore, considering the small number of patients in a single experiment for some drugs and the large confidence intervals for the efficacy and safety outcomes, the results of the final network meta-analysis should be interpreted with caution.

Our study has some limitations that need consideration. First, the sample size was small and some drugs were tested on fewer than 100 participants. Second, because DS is a relatively rare disease with a limited number of related RCTs, some of the included RCTs were performed more than a decade ago. Third, the mean age of the study population was approximately 9 years, which may have further confounded the results due to the effects of age and regression to the mean. Fourth, most of the included studies used only seizures as the main result. However, DS includes multiple types of convulsive seizures, and non-convulsive seizures are associated with DS. The number of included studies was small, and although a systematic literature search was conducted, the bias of the small-study effects persisted. Finally, the shorter treatment periods (up to 2 months) in each study may not be sufficient to assess the true longer-term efficacy of each drug, and more direct clinical trials (head-to-head comparisons) are needed to further validate these results.

Conclusion

The efficacy of all four antiseizure medications was significantly higher than that of the placebo. The SUCRA ranking showed that stiripentol was the most likely to cause AEs and SAEs. Cannabidiol was slightly inferior to the other three antiseizure medications in terms of efficacy and safety. The results of this indirect comparison suggest that soticlestat, fenfluramine, and stiripentol may have little difference in efficacy, but soticlestat and fenfluramine are safer. Regarding comprehensive efficacy and safety, soticlestat may be the best adjuvant therapy for DS, followed by fenfluramine. This conclusion is consistent with the comparison of long-term efficacy and safety. However, soticlestat has only recently passed phase 2 clinical trials, and the results are yet to be verified. We look forward to achieving positive results in subsequent clinical trials. Nonetheless, this still provides a reference for the clinical selection of the four antiseizure medications. We hope there will be more clinical, drug interaction, and pharmacoeconomic studies related to DS in the future, which will provide more scientific references for the selection of therapeutic drugs for patients with DS in clinical practice.

References