Stiripentol: Unpublished results from the first phase 2 clinical trial in Lennox–Gastaut syndrome conducted in the early 1990s

Abstract

Objective

This study was undertaken to present the results of an exploratory phase 2 trial of stiripentol in Lennox–Gastaut syndrome (LGS).

Methods

This exploratory single‐blind, single‐arm, nonrandomized sequential‐period phase 2 study was conducted at four centers in France between January 1989 and August 1993. Eligible patients were aged 2–20 years with LGS and experienced at least one seizure per week despite optimized therapy. After a 1‐month baseline period under a stable treatment regimen, patients received placebo for 1 month followed by stiripentol for 2 months.

Results

Sixteen patients with LGS were enrolled, and efficacy was assessable in 14. The median [Q1–Q3] overall seizure frequency decreased from 31 [16–89] at baseline to 14 [8–21] after the first month of stiripentol (p = .044) and further to 4 [0–18] after the second month (p = .044). The reduction was consistent across seizure types. Eight patients (57%) were responders (≥50% reduction in overall seizure frequency) at the end of stiripentol treatment, including five (36%) who achieved complete seizure freedom and two with a ≥75% reduction in seizure frequency. All patients reported at least one adverse event during stiripentol treatment, most commonly somnolence, decreased appetite, and vomiting. These effects may be related to the high stiripentol doses administered (median = 91 mg/kg/day in the second month). Despite their frequency, no serious adverse events were reported during the stiripentol treatment period, and no clinically meaningful changes in hematological parameters or liver enzyme levels were observed.

Significance

Although this study was conducted before standardized clinical trial designs for LGS were established, a more comprehensive evaluation of stiripentol could have provided further insight into its potential benefits in this severe developmental and epileptic encephalopathy.

Key points.

• Stiripentol demonstrates antiseizure efficacy in LGS, with a 73% median reduction and 36% seizure‐free rate after 2 months.

• The placebo period showed no improvement, whereas efficacy appeared rapidly after stiripentol initiation.

• High doses of stiripentol were used and led to frequent, but generally manageable, adverse events.

1. INTRODUCTION

Lennox–Gastaut syndrome (LGS) is one of the most challenging developmental and epileptic encephalopathies (DEEs) of childhood, accounting for approximately 1%–2% of all epilepsy cases and 3%–4% of pediatric epilepsy populations. ¹ This syndrome is characterized by the presence of (1) multiple types of drug‐resistant seizures with onset before 18 years (one of which must include tonic); (2) cognitive and often behavioral impairments, which may not be present before seizure onset; and (3) diffuse slow spike‐and‐ wave and generalized paroxysmal fast activity on electroencephalography (EEG). ² Although several antiseizure medications (ASMs) are approved for the treatment of LGS‐associated seizures, achieving optimal seizure control remains unlikely for most patients. ³

Stiripentol (STP) is an ASM with multiple mechanisms of action, including direct positive allosteric modulation of γ‐aminobutyric acid subtype A receptors, inhibition of lactate dehydrogenase, and blockade of T‐type calcium channels. ⁴ Its clinical development began in the mid‐1980s with early clinical trials in patients with drug‐resistant epilepsy, ⁵ , ⁶ , ⁷ all of which suggested a beneficial effect of STP. Subsequently, two key clinical trials were initiated at about the same time.

The first one was a prospective, exploratory, phase 2 trial that assessed the efficacy and tolerability of STP as an add‐on therapy in a large series of children with severe drug‐resistant epilepsy, to identify the most suitable indication for STP. ⁸ Among the 227 enrolled patients, 25 had Dravet syndrome (DS), in whom STP demonstrated significant efficacy. These findings prompted the initiation of two double‐blind, placebo‐controlled, phase 3 trials in DS, both of which confirmed the high efficacy of STP compared with placebo. ⁹ , ¹⁰ These trials provided the basis for regulatory approval of STP in DS, representing the first marketing authorization of an ASM for a DEE.

The second study was the present exploratory phase 2 trial conducted in patients with LGS. The primary objective of the study was to evaluate the efficacy of STP administered in combination with standard‐of‐care therapy, as assessed by seizure frequency (total number of seizures reported by the patient or caregiver) and by the maximum duration of interseizure intervals. Although results were positive, because subsequent clinical development of STP focused on DS, the findings of the LGS trial remained undisclosed.

Despite the trial having been conducted several years ago, dissemination of clinical study results remains good scientific practice. We report these data here for the first time, as the publication of all clinical trial results, regardless of outcome, is essential, particularly in rare diseases where evidence remains limited. Given the severity of LGS and the persistent unmet medical need, these findings may be clinically relevant for physicians managing this difficult‐to‐treat condition.

2. MATERIALS AND METHODS

2.1. Ethics and regulation

The study protocol was reviewed and approved by the Marseille Consultative Committee for the Protection of Persons in Biomedical Research and was declared to the Ministry of Health (Directorate for Pharmacy and Medicinal Products), in accordance with the French regulations in force at the time of the study.

2.2. Study design and data collection

This was a single‐blind, single‐arm, nonrandomized sequential‐period phase 2 study, conducted at four centers in France, from January 1989 to August 1993. All male and female patients with LGS were eligible for inclusion, regardless of seizure type. Eligible patients were required to be 2–20 years of age and to experience at least one seizure per week despite optimized treatment with up to three concomitant ASMs. The main exclusion criteria included the presence of severe concomitant disease, inability to comply with treatment follow‐up or inadequate seizure monitoring (i.e., imprecise seizure assessment), and failure to obtain informed consent.

The trial consisted of three distinct phases. Following a 1‐month baseline period during which patients remained on their stable treatment regimen, they received placebo for 1 month, followed by STP for 2 months (Figure 1). The placebo was identical in appearance to STP and was administered either as capsules or sachets. Patients and their caregivers were not informed of placebo administration, and blinding was maintained throughout the study.

FIGURE 1.

Schematic representation of the study design.

At the end of the placebo phase, STP was titrated over a 3‐day period. Given its well‐established pharmacokinetic interactions with other ASMs, doses of concomitant medications were adjusted in parallel to maintain stable plasma concentrations. Therefore, in accordance with the study protocol, phenobarbital and phenytoin doses were reduced by 30%–50% and carbamazepine by one third. Benzodiazepine doses were reduced only in case of adverse events, and valproate doses remained unchanged. The daily dosage of STP ranged from 2000 to 3000 mg, adjusted according to body weight: 2000 mg/day for patients weighing 29 kg or less, 2500 mg/day for those weighing 30–39 kg, and 3000 mg/day for those weighing 40 kg or more. In the case of adverse events, the investigator was permitted to reduce the dosage of STP or concomitant ASMs.

Caregivers or patients documented seizure type (convulsive seizures, absence seizures, drop seizures, or sleep‐related seizures), frequency, and time of occurrence (morning, afternoon, or evening/night) in daily diaries. Study visits were scheduled at Day 1 (enrollment), Day 30 (end of baseline), Day 60 (end of the placebo phase), Day 90, and Day 120 (end of STP treatment). For safety monitoring, caregivers and/or patients were instructed to report any adverse events observed during each treatment period. Blood samples were collected at baseline and at the end of the treatment period to assess hematological parameters and liver function.

Since the original conduct of the study, data management practices and statistical analysis standards have evolved substantially. Therefore, between June 2024 and May 2025, all case report forms and patient diaries were comprehensively reviewed. Individual efficacy data, originally presented as listings in the study report, were re‐entered into updated, standardized databases that served as the basis for the present analyses. In addition, all adverse events recorded in the case report forms were coded using the latest version of the Medical Dictionary for Regulatory Activities, version 28.0. The number of events and the number and percentage of patients reporting at least one event were summarized by primary system organ class and preferred term.

2.3. Population

The study protocol allowed the inclusion of patients with LGS or DEE with diffuse EEG abnormalities. In this article, we present results from patients with a confirmed diagnosis of LGS. Diagnosis of LGS was based on seizure types (atypical absence seizures, tonic–clonic seizures = convulsive seizures, sleep‐related seizures, seizures leading to falls = drop seizures), EEG features (diffuse slow spike‐waves, fast rhythm discharges during sleep), and nonseizure features such as intellectual disability and behavioral disturbances.

2.4. Outcome measures

The primary objective of the study was to evaluate the efficacy of STP at the end of the second month of treatment (Days 91–120) compared with the reference phase (Days 1–30). The primary outcomes were as follows: (1) the change in the total number of seizures; (2) the change in the duration of the interseizure interval, defined as the maximum time in hours between two consecutive seizures; and (3) response rates based on seizure frequency reduction, categorized as an increase (>0%) or a decrease (<50%, ≥50%, ≥75%, or 100%) in seizure frequency.

Secondary outcomes included seizure frequency, interseizure interval, and responder rates during the placebo phase (Days 31–60) compared with the reference phase (Days 1–30), as well as changes at the end of STP treatment compared with the placebo phase. Tolerability was assessed by the proportion of patients experiencing adverse events during the reference phase, the placebo phase, or the STP phase (Days 61–120).

2.5. Statistical analysis

Quantitative variables were summarized by mean, SD, median, quartiles, and range, and qualitative variables by counts and percentages. Seizure counts for each study period were normalized to a 30‐day equivalent to ensure comparability across patients. Absolute and relative changes from baseline (Days 1–30) or placebo phase (Days 31–60) were calculated.

Primary efficacy analysis compared seizure frequency and interseizure interval between Days 1–30 and Days 91–120 using Wilcoxon signed‐rank tests. Responder rates (increase in seizure frequency, reduction in seizure frequency of >0%, ≥50%, ≥75%, and 100%) were compared using McNemar test. Secondary analysis assessed placebo effect (Days 31–60 vs. Days 1–30), STP effect at Day 90 (vs. Days 1–30 and vs. Days 31–60), and STP effect at Day 120 relative to placebo (Days 91–120 vs. Days 31–60) using the same nonparametric methods. All tests were two‐sided with a significance level of 5%. To address the risk of false‐positive findings due to multiple comparisons, p‐values were adjusted for multiplicity using the false discovery rate procedure of Benjamini and Hochberg. ¹¹ Analyses were performed using R (version 4.3.2 or higher).

3. RESULTS

Sixteen patients with LGS were enrolled in the study. One patient discontinued STP treatment shortly after initiation because of adverse events (described in detail below in the section on safety results). As the case report form for this patient was only partially completed, she was excluded from the efficacy analyses, leaving 15 patients for analysis. Among these 15 patients, 11 were male, and the median age at enrollment was 7.3 years (range = 1.6–22.2 years; Table 1). At baseline, the median total number of seizures was 31 (range = 3–203), with the most frequent seizure type being convulsive seizures in seven patients and absence seizures in five patients. Patients were receiving a median of three concomitant ASMs at baseline, most commonly valproate (10 patients) and clobazam (eight patients).

TABLE 1.

Median seizure frequency and median percentage change from baseline over time.

	Baseline, Days 1–30	Placebo, Days 31–60	Stiripentol, Days 61–90	Stiripentol, Days 91–120
All seizures
Frequency [Q1; Q3] (n)	31[16; 89] (14)	30 [12; 71] (14)	14 [8; 21] (15)	4 [0; 18] (15)
% change [Q1; Q3] (n)	–	6 [−2; +88] (13)	−50 [−84; +3] (14)	−73 [−100; −4] (14)
p‐value vs. baseline a	–	.28	.044	.044
p‐value vs. placebo a	–	–	.044	.073
Convulsive seizures
Frequency [Q1; Q3] (n)	7 [1; 27] (14)	7 [2; 21] (14)	4 [0; 15] (15)	1 [0; 10] (14)
% change [Q1; Q3] (n)	–	48 [−61; +343] (11)	−61 [−92; +14] (12)	−84 [−100; −18] (12)
Absence seizures
Frequency [Q1; Q3] (n)	10 [0; 22] (14)	3 [1; 19] (13)	3 [0; 10] (15)	0 [0; 3] (15)
% change [Q1; Q3] (n)	–	28 [−64; +333] (8)	−23 [−100; +54] (10)	−100 [−100; −10] (10)
Drop seizures
Frequency [Q1; Q3] (n)	0 [0; 7] (14)	0 [.0; 2.6] (13)	0 [.0; .0] (15)	0 [.0; .0] (15)
% change [Q1; Q3] (n)	–	−67 [−100; −22] (4)	−100 [−100; −40] (6)	−100 [−100; −100] (6)
Sleep seizures
Frequency [Q1; Q3] (n)	0 [0; 2] (14)	0 [.0; 5.4] (14)	0 [.0; 1.1] (15)	0 [.0; 1.1] (15)
% change [Q1; Q3] (n)	–	−76 [−100; +6] (3)	−100 [−100; −90] (4)	−100 [−100; −64] (4)

Note: n, number of patients; Q1, 1st quartile; Q3, 3rd quartile.

^a Wilcoxon signed‐rank test with p‐values adjusted using the false discovery rate procedure of Benjamini and Hochberg.

STP was administered at a mean (±SD) dosage of 86 ± 39 mg/kg/day during the first month of treatment. This dosage remained stable throughout the study, with a mean of 91 ± 36 mg/kg/day in the second month. In accordance with the study protocol, the dosages of certain concomitant ASMs were reduced following STP initiation to minimize the risk of overdose‐related adverse effects. This was the case for clobazam, whose mean (±SD) dosage was decreased from .7 ± .3 to .4 ± .2 mg/kg/day; carbamazepine was decreased from 22 ± 5 to 11 ± 6 mg/kg/day, and phenobarbital was decreased from 2.3 ± .9 to 1.3 ± .1 mg/kg/day. The valproate dosage remained stable throughout the study, ranging from 33 ± 9 mg/kg/day at Day 30 to 28 ± 6 mg/kg/day at Day 120.

With regard to seizure evaluation, one patient had seizure types documented at baseline, but corresponding seizure counts were unavailable, precluding any assessment of treatment efficacy in this individual. Among the 14 evaluable patients, the median [Q1; Q3] monthly overall seizure frequency at baseline was 31 [16; 89], with no significant change observed after the 1‐month placebo phase (30 [12; 71]; p = .28; Table 1). In contrast, a significant reduction in overall seizure frequency was observed following STP treatment; the median monthly overall seizure frequency decreased to 14 [8; 21] at the end of the first month (p = .044) and to 4 [0; 18] at the end of the second month (p = .044), representing median reductions of 50% [−84; +3] and 73% [−100; −4], respectively (Table 1 and Figure 2). STP efficacy was consistent across seizure types, with median frequency reductions at the end of the 2‐month treatment period reaching −84% for convulsive seizures (n = 12), −100% for absence seizures (n = 10), −100% for drop seizures (n = 6), and −100% for sleep‐related seizures (n = 4) compared with baseline (Table 1). Overall, at the end of the study period, five of 12 patients were free from convulsive seizures, six of 10 from absence seizures, five of six from drop seizures, and three of four from sleep‐related seizures.

FIGURE 2.

Median [Q1–Q3] monthly overall seizure frequency across study time points.

From a responder perspective, defined as a ≥50% reduction in seizure frequency, when considering the overall seizure frequency, one patient met this criterion at the end of the placebo phase (Figure 3). In contrast, after 1 month of STP treatment, seven patients (50%) were classified as responders, of whom one was seizure‐free, and three had a ≥75% reduction in seizure frequency. By the end of the 2‐month treatment period, eight patients (57%) were responders, including five patients (36%) who achieved complete seizure freedom and two patients who had a ≥75% reduction in seizure frequency (Figure 3). From the first to the second month of STP therapy, treatment efficacy was maintained or further improved in all patients, except for two patients who experienced an increase in seizure frequency.

FIGURE 3.

Individual evolution of treatment response over time: relative change in total seizure count from the baseline reference period (Days 1–30).

In line with the reduction in seizure frequency, interseizure intervals increased significantly on STP (Table 2). The median duration of seizure‐free intervals increased nonsignificantly from 60 h at baseline to 80 h during the placebo phase. It then rose to 168 h during the first month of STP treatment (p = .044 vs. baseline) and further extended to 376 h during the second month (p = .044 vs. baseline).

TABLE 2.

Interseizure intervals (in hours) according to time periods.

	Baseline, Days 1–30	Placebo, Days 31–60	Stiripentol, Days 61–90	Stiripentol, Days 91–120
n	14	14	14	15
Mean (SD)	86 (83)	121 (102)	227 (165)	401 (272)
Median [Q1; Q3]	60 [16; 112]	80 [64; 184]	168 [128; 288]	376 [160; 720]
p‐value vs. baseline a	‐	.9	.044	.044
p‐value vs. placebo a			.085	.044

Note: n, number of patients; Q1, 1st quartile; Q3, 3rd quartile.

^a Wilcoxon signed‐rank test with p‐values adjusted using the false discovery rate procedure of Benjamini and Hochberg.

Regarding safety, an 18‐year‐old female patient rapidly developed marked anorexia, nausea, vomiting, balance disturbances, and drowsiness after initiation of STP. These adverse events did not resolve following a reduction in phenytoin dosage, and STP was therefore discontinued. All patients (100%) experienced at least one adverse event during the STP treatment period, with a total of 51 events reported, compared to 33% during the baseline period and 47% during the placebo period (Table 3). The three most reported adverse events during STP treatment were somnolence (60%), decreased appetite (47%), and vomiting (27%).

TABLE 3.

Adverse events reported in more than one patient.

Study phase	Baseline period		Placebo period		Stiripentol period
	Patients, n (%)	Events, n	Patients, n (%)	Events, n	Patients, n (%)	Events, n
Overall	5 (33%)	5	7 (47%)	11	15 (100%)	51
Gastrointestinal disorders	1 (6.7%)	1	2 (13%)	2	6 (40%)	11
Constipation	0 (0%)	0	0 (0%)	0	2 (13%)	2
Nausea	0 (0%)	0	0 (0%)	0	3 (20%)	3
Vomiting	0 (0%)	0	1 (6.7%)	1	4 (27%)	4
General disorders	0 (0%)	0	1 (6.7%)	1	2 (13%)	2
Infections and infestations	1 (6.7%)	1	3 (20%)	3	4 (27%)	5
Investigations	0 (0%)	0	0 (0%)	0	3 (20%)	3
Metabolism and nutritional disorders	2 (13%)	2	0 (0%)	0	7 (47%)	7
Decreased appetite	2 (13%)	2	0 (0%)	0	7 (47%)	7
Nervous system disorders	0 (0%)	0	2 (13%)	2	9 (60%)	13
Balance disorder	0 (0%)	0	1 (6.7%)	1	2 (13%)	2
Somnolence	0 (0%)	0	1 (6.7%)	1	9 (60%)	9
Psychiatric disorders	1 (6.7%)	1	2 (13%)	3	4 (27%)	5
Agitation	0 (0%)	0	2 (13%)	2	1 (6.7%)	2
Skin and subcutaneous tissue disorders	0 (0%)	0	0 (0%)	0	3 (20%)	3

One serious adverse event was reported; during the baseline period, a patient required a 3‐day hospitalization for pneumonia. This event resolved following antibiotic treatment, and the patient subsequently completed the study.

There were no clinically meaningful changes in hematological parameters, including red blood cell count, white blood cell count, and platelet count, between baseline and the end of the STP treatment period. Regarding liver enzymes, alanine aminotransferase and aspartate aminotransferase levels remained stable throughout the study. In contrast, γ‐glutamyltransferase (γ‐GT) levels increased from 17 ± 8 IU/L (mean ± SD) at baseline to 55 ± 19 IU/L at the end of treatment. One case of γ‐GT elevation, from 38 to 73 IU/L, was reported as an adverse event; however, the investigator assessed it as very mild.

4. DISCUSSION

This multicenter, single‐blind, single‐arm, sequential‐period phase 2 study conducted in France between January 1989 and August 1993 recruited 16 patients with LGS. STP significantly reduced seizure frequency, with a median decrease of 73% after 2 months, and 36% of patients became seizure‐free. Adverse events, mainly somnolence, decreased appetite, and vomiting, were frequent but generally manageable.

Since the first randomized controlled trial in LGS in 1993, ¹² several clinical investigations have led to the approval of seven ASMs: cannabidiol, clobazam, felbamate, fenfluramine, lamotrigine, rufinamide, and topiramate. ¹³ Some of these ASMs were initially evaluated in open‐label exploratory or phase 2 trials, for example, cannabidiol in 2016, ¹⁴ fenfluramine in 2018, ¹⁵ and topiramate in 1997. ¹⁶ More recently, similar exploratory studies have led to ongoing randomized controlled trials (soticlestat, bexicaserin). ¹⁷ , ¹⁸

The present study was the first exploratory clinical trial specifically designed to evaluate the efficacy of an ASM in patients with LGS. The results suggest that STP may provide clinical benefit in a significant proportion of LGS patients. Importantly, STP showed superiority over placebo, even though the study was not designed or powered to allow formal comparisons between these two treatment periods. STP efficacy was consistent across seizure types, with median reductions observed at the end of the 2‐month treatment period for convulsive seizures (−88%), absence seizures (−100%), drop seizures (−100%), and sleep‐related seizures (−100%) compared with baseline. It should be noted, however, that certain seizure types (notably sleep‐related seizures and absence seizures) are not considered regulatory countable seizures.

No further clinical trials have been conducted to evaluate STP as add‐on therapy in LGS. However, additional information can be derived from real‐world, retrospective observational studies that included patients with LGS within broader patient cohorts. Given the substantial heterogeneity in epilepsy types across these cohorts, the results are inherently heterogeneous. Habermehl et al. ¹⁹ evaluated the safety and efficacy of STP in 22 adults with pharmacoresistant focal or multifocal epilepsy, including eight with LGS. After 12 months, 54.4% of patients continued STP treatment; the responder rate was 36.4%, and 13.6% achieved seizure freedom. In a large multicenter study, Gil‐Nagel et al. ²⁰ analyzed 82 patients, 55 with DS and 27 with other refractory DEEs, including eight LGS cases. After 12 months, overall retention, responder, and seizure‐free rates were 68%, 65%, and 18%, respectively. More recently, Soto‐Insuga and colleagues ²¹ evaluated a cohort of 18 DS and 17 non‐DS patients, including 13 patients with DEE, of whom four had LGS. After 3 months of STP treatment, seizure characteristics (frequency, duration, and/or intensity) improved in 76.5% of non‐DS patients, with 58.8% classified as responders. In this latter study, STP was shown to be effective in reducing absence, myoclonic, and focal seizures, among others, in both the non‐DS and DS cohorts.

It is noteworthy that in this study, STP was administered at a fixed daily dose ranging from 2000 to 3000 mg/day, increasing with the patient's weight. Given the young age of the study population, this approach resulted in a particularly high median dose of 91 mg/kg/day, with maximum doses reaching up to 182 mg/kg/day, substantially exceeding the recommended dose of 50 mg/kg/day. These elevated doses may have contributed to the observed efficacy but may also explain the 100% incidence of adverse events, which was higher than the frequencies reported in phase 3 clinical trials in LGS with other ASMs, such as clobazam (up to 88.7%), ²² rufinamide (93.1%), ²³ or cannabidiol (86%). ²⁴ This pattern aligns with the findings of Chiron et al., ²⁵ who similarly reported enhanced efficacy of STP at doses exceeding 50 mg/kg/day in children younger than 2 years of age, accompanied by an increased incidence of adverse events. Interestingly, despite the high frequency of adverse events with STP, their severity remained acceptable, with no serious adverse events reported during the treatment period and no clinically meaningful changes in hematological parameters or liver enzyme levels. Although STP doses exceeding 50 mg/kg/day should not be recommended as standard practice, these observations suggest that higher doses may be considered in selected patients, provided that close monitoring and rigorous safety management are ensured.

Several limitations of this study should be acknowledged, primarily related to the long interval between its conduct and the present day. First, international classifications and clinical guidelines have progressively refined and operationalized the diagnostic criteria for LGS, potentially introducing selection bias or limiting the representativeness of the study population. Nevertheless, from the early 1990s to the present, the core diagnostic concept of LGS has remained largely unchanged. LGS continues to be defined as an electroclinical syndrome characterized by drug‐resistant multiple seizure types, a typical slow spike–wave pattern on interictal EEG, and cognitive impairment.

Since the completion of this study, several new ASMs have been introduced, some of which have received marketing authorization specifically for LGS. Consequently, the therapeutic landscape and clinical management of LGS have evolved substantially. Patients enrolled in contemporary clinical trials are typically exposed at baseline to a much broader therapeutic armamentarium, including cannabidiol, felbamate, and fenfluramine, as well as nonpharmacological interventions that were not widely available at the time this study was conducted. As a result, the present findings may not fully reflect current clinical practice, and the magnitude of benefit observed with STP may be overestimated, particularly in light of the relatively high doses of STP used in this study.

In addition, seizure types were classified differently at the time the study was conducted than in current practice, complicating direct comparisons with more recent LGS clinical trials.

An additional limitation is the sequential placebo–treatment design without a parallel control group, which may introduce temporal and expectancy biases. Moreover, the reduction in concomitant ASM dosages at the initiation of STP represents a potential confounding factor, and the relatively small sample size limits the analyses' statistical power. Nevertheless, the within‐patient comparison design, in which each patient serves as their own control, partially mitigates these limitations and strengthens the internal consistency of the findings.

An important takeaway from this study is the need to publish and share all clinical trial results, regardless of outcome, especially in the field of rare diseases, where data remain scarce. When this study was conducted, management of LGS was poorly standardized, and earlier dissemination could have encouraged targeted use of STP and generated valuable real‐world evidence. Considering that more than half of LGS patients experience status epilepticus ²⁶ and that STP potentially reduces both the frequency and duration of these episodes, ²⁷ a broader evaluation of STP in LGS could have provided further insight into its potential benefits in this syndrome. It would also be informative to gather feedback from clinicians treating LGS on whether they have already prescribed STP, and, if so, to document the efficacy and safety outcomes observed in their patients.

5. CONCLUSIONS

In this phase 2 study, STP significantly reduced seizure frequency in patients with LGS, a severe DEE in which seizure control is rarely achieved. As this study was conducted before standardized clinical trial designs for LGS were established, these results warrant confirmation in adequately powered, well‐designed clinical trials.

AUTHOR CONTRIBUTIONS

Stéphane Auvin, Benjamin Serraz, and Laurent Chancharme all contributed to the conceptualization of the article, oversaw data curation, and participated in validating results, interpreting findings, and writing, reviewing, and editing the manuscript. Jérémie Lespinasse conducted the statistics, oversaw data curation, participated in validating results and interpreting findings, and reviewed the manuscript.

CONFLICT OF INTEREST STATEMENT

At the time the study was conducted, it was sponsored and supervised by Biocodex, which was responsible for the study design (with input from investigators and other experts), study management, monitoring, pharmacovigilance, statistical analyses, data analysis, and the supply of investigational medicinal products. The present reanalysis and manuscript preparation were also supported by Biocodex. S.A. is Deputy Editor of Epilepsia. He has received personal fees for lectures or advice from Angelini, Biocodex, Eisai, Encoded, GRIN Therapeutics, Jazz Pharmaceuticals, Longboard, Lundbeck, Neuraxpharm, Nutricia, Orion, Proveca, Servier, Stoke, UCB Pharma, and Xenon. He has been an investigator for Eisai, Lundbeck, Proveca, Servier, Takeda, and UCB Pharma. J.L. has no conflict of interest. B.S. and L.C. are full‐time employees of Biocodex. We confirm that we have read the Journal's position on issues involved in ethical publication and affirm that this report is consistent with those guidelines.

Supporting information

Table S1.

DATA AVAILABILITY STATEMENT

Research data are not shared.