Audit of use of stiripentol in adults with Dravet syndrome

Objectives There are very few data available in the literature on the use of stiripentol in adults with Dravet syndrome (DS). DS cases are increasingly recognized in adulthood, and more children with DS now survive to adulthood. The aim of the study was to document the effectiveness and tolerability of stiripentol in adults with DS. Material and methods We conducted an observational clinical audit in the epilepsy service of the National Hospital for Neurology and Neurosurgery, London (UK). Results We included 13 adult subjects with DS (eight females, five males). The responder (defined as more than 50% reduction in all seizure types) rate was 3/13 (23%) at 36 months. The following other outcomes were reported: seizure exacerbation (3/13, 23%), no change (3/13, 23%), less than 50% reduction in seizures (2/13, 15%), more than 50% reduction in generalized tonic‐clonic seizures but no other seizure types (1/13, 8%), undefined response (1/13, 8%). The retention rate was 62% after 1 year and 31% after 5 years. Adverse effects were reported in 7/13 (54%): the most frequent were anorexia, weight loss, unsteadiness and tiredness. Withdrawal due to adverse effects occurred in 3/13 (23%). Conclusions Compared with previous studies on children with DS, our results show a lower responder rate and a similar tolerability profile. Stiripentol can be effective with a good tolerability profile. Our audit is small, but supports the use of stiripentol in adults with DS when first‐line treatments are ineffective or not tolerated, in keeping with published guidelines.

least a 50% reduction in seizure frequency, compared with placebo.
Very few data are available about the use of stiripentol in adults, with reports for fewer than fifteen DS cases older than 18 years, [18][19][20][21] and twelve healthy adult volunteers. 22 We conducted a clinical audit to document the effectiveness and tolerability of stiripentol in adults with DS.

| MATERIAL AND METHODS
This project was reviewed and registered as an observational clinical audit at the National Hospital for Neurology and Neurosurgery, London (UK). Ethics committee approval is not required for an audit, as clinical audit by definition does not involve intervention beyond usual clinical management. 23 We included all adults (>18 years) with DS attending the epilepsy service with previous or current treatment with stiripentol, from Jan- The following variables were included in the analysis: age, gender, age at clinical diagnosis, age at genetic testing, SCN1A mutation details, seizure types, history of status epilepticus, antiepileptic drugs (AEDs) tried before the introduction of stiripentol, age at introduction of stiripentol, interval between genetic diagnosis and introduction of stiripentol, AEDs initially used with stiripentol, mean monthly seizure frequency before the introduction of stiripentol and while on stiripentol treatment, maximum dose of stiripentol, adverse effects during treatment with stiripentol, and total duration of exposure to stiripentol. Seizure exacerbation was defined as an increase of more than 25% in seizure frequency compared with seizure frequency before the introduction of stiripentol.

| RESULTS
We identified 32 adult cases with DS in our service. Thirteen adults with DS and previous or current treatment with stiripentol were identified and comprised eight females and five males. DS was diagnosed on the basis of clinical findings, at a mean age of 24 (SD 11, range 2-41). Genetic analysis was undertaken at a mean age of 27 (SD 9, range 12-43). SCN1A mutation or rearrangements were found in 12 of the 13 patients included. All cases had multiple seizure types. Eight patients (62%) had a history of status epilepticus prior to the introduction of stiripentol. In five cases (29%), stiripentol was introduced in adolescence and continued into adulthood. The mean age at introduction of stiripentol was 26 years (SD 12, range 12-47).
In four subjects, it was introduced seven to ten years (median 7.5) before the genetic analysis. In the other nine cases, it was introduced a few months to 6 years (median 1 year), after positive genetic testing for SCN1A mutation. At the time of stiripentol introduction, the mean number of AEDs already tried was 9 (range 3-15), and it was used with one to four concurrent AEDs. None of the cases was ever on monotherapy with stiripentol. Details of SCN1A mutations, seizure types, AED history and age at introduction of stiripentol are shown in Table 1.

| Effectiveness
The frequency of seizures (of any type) at the time of stiripentol introduction varied from daily to weekly. The daily dose varied from 250 mg to 3000 mg with a mean maximum daily dose of 1604 mg. Over an average time of exposure to stiripentol of 42 months (range 3-139), more than 50% reduction in frequency of all seizure types was reported in three cases (23%). The other outcomes were: less than 50% reduction in frequency of all seizure types in two cases (15%), no change in seizure frequency in three cases (23%), more than 50% reduction in frequency of generalized tonic-clonic seizures (GTCs) but no change in frequency of other seizure types in one case (8%), seizure exacerbation in three cases (23%) and undefined response in one case (8%). One case had initially more than 50% reduction in frequency of GTCs, from weekly to monthly GTCs with a 45 day period without any GTCs, and became then free of GTCs for about 12 months, followed by GTCs recurrence precipitated by intense physical exercise. Before introduction of stiripentol, eight patients (62%) had had from one to multiple episodes of status epilepticus. Of these cases, four had no further episodes of status after the introduction of stiripentol until the latest followup, over an average time of 43 months (range 3-139) on stiripentol.
Withdrawal due to lack of efficacy occurred in four cases (31%).
In seven cases (54%), treatment was still ongoing at the time of the audit (July 2015), after they had been on stiripentol for an average time of 62 months (range 10-139). The following medication changes were undertaken in these seven cases during the course of treatment with stiripentol: in two clobazam was introduced; in one carbamazepine was withdrawn and clobazam introduced; in one phenytoin was withdrawn; in one sodium valproate was introduced; in one topiramate was withdrawn, sodium valproate, levetiracetam, zonisamide and perampanel were introduced sequentially (zonisamide was soon withdrawn due to lack of effectiveness) and a vagus nerve stimulator was recently inserted; one did not have any changes to concurrent AEDs. Five cases (39%) were on clobazam and 10 cases (77%) were on sodium valproate, as concurrent treatment when stiripentol was initiated. In three cases, (23%) clobazam and in one case (8%) sodium valproate were introduced later on during the course of treatment with stiripentol. Overall, at some point eight cases (62%) had concurrent treatment with clobazam and 11 cases (85%) with sodium valproate. A summary of stiripentol effectiveness in the audited cohort is reported in Table 2.

| Tolerability
Adverse effects were not systemically sought as this was not a prospective study, but were reported in seven cases (54%). These were: anorexia (n=4), weight loss (n=4), unsteadiness (n=3), tiredness/somnolence (n=2), nausea (n=1), abdominal pain (n=1), diarrhoea (n=1), myelodysplasia (thrombocytopenia and neutropenia) (n=1), behavioural disturbance (n=1), increased tremor (n=1). In six cases, reduction or withdrawal of one or two concurrent AEDs was reported. Withdrawal of stiripentol due to adverse effects occurred in three cases (23%): one due to anorexia and weight loss together with lack of efficacy, one due to behavioural disturbance and diarrhoea and one due to anorexia, unsteadiness T A B L E 1 Details of SCN1A mutations, seizure types, AED history, and age at introduction of stiripentol

| DISCUSSION
In a small cohort, we show that stiripentol can be effective in adults with DS, with a reasonable tolerability profile. The responder rate in our cohort was lower than that from studies in children with DS. Not enough data on adult DS cases treated with stiripentol are available to make adequate comparisons with our findings.
We note that some adults have no useful response to stiripentol.
A randomised placebo-controlled trial conducted in children with DS 17 showed 15 of 21 (71%) were responders (seizure-free or experienced at least a 50% reduction in seizure frequency) on stiripentol, whereas there was only one responder of 20 (5%) on placebo (none were seizure-free). T A B L E 2 (Continued) had more than 50% reduction in frequency of clonic seizures or GTCs during the last 4 weeks of the fixed-dose phase vs baseline, including one who became seizure-free. After the 12-week fixed-dose phase, 21 of the initial 24 DS cases entered the 40-week long-term administration phase. 21 None of the older cases remained responders in the long-term (one terminated the study due to poor response without entering the long-term phase). Adverse events were reported in all patients enrolled; the most frequent included somnolence, loss of appetite and ataxia, occurring mostly during the dose-adjustment phase and were of mild-to-moderate severity. Insufficient data were available to establish a relationship with the dose of stiripentol. Due to the retrospective design of the study and the small size of the cohort, effectiveness was assessed on all seizure types and GTCs, but not on the other single seizure types. In terms of effectiveness, the responder rate (defined as more than 50% reduction in all seizure types) was 23% (3/13). More than 50% reduction in GTCs was observed in four cases (31%). It is important to note that all the reported cases had been exposed to sodium channel blockers, known to aggravate seizures and potentially precipitate status epilepticus in DS. 2,6,9 This is likely related to the fact that ours was an adult cohort, with often We report seizure exacerbation in three cases (23%). The reason for this is unknown, and this outcome is not described in the stiripentol literature for adult cases and it is rarely reported in children with DS. [17][18][19][20][21] Dravet syndrome cases are increasingly recognized in adulthood, and many children with DS survive to adulthood. Better treatment strategies are required. Stiripentol is an allosteric modulator of the GABAA receptor, 29 proven to be effective and well-tolerated in children with DS. 21,33 There are no data available in the literature on treatment with stiripentol of DS cases older than 24 years. Although this was a small retrospective study with incomplete information for some cases, it is the largest and oldest sample of DS adult cases treated with stiripentol and with the longest follow-up to our knowledge. In some cases, because there were other drug changes made on clinical grounds, we cannot be certain that the entire response observed after stiripentol introduction was due solely to stiripentol: this was an audit, not a prospective clinical trial.
In view of our findings, the effectiveness demonstrated in previous studies 17,21 and its reasonable tolerability profile, we suggest stiripentol is worth considering for adults with DS and drug-resistant epilepsy when first-line treatments are ineffective or not tolerated, in keeping with recommendations from National Institute for Health and Care Excellence guidelines in the UK. 34 Stiripentol use may lead to improved seizure control, with reduction of hospitalization costs and subsequent better health-care service utilization, as previously shown. 20 Prospective studies, including well-designed trials in larger cohorts, would seem warranted.

ACKNOWLEDGMENTS
We thank the Wellcome Trust for funding Dr Simona Balestrini through a Wellcome Trust Strategic Award (WT104033AIA).