Abstract
Sirolimus for Epilepsy in Children with Tuberous Sclerosis Complex: A Randomized Controlled Trial
Overwater IE, Rietman AB, Bindels-de Heus K, Looman CWN, Rizopoulos D, Sibindi TM, Cherian PJ, Jansen FE, Moll HA, Elgersma Y, de Wit MCY. Neurology 2016;87:1011–1018.
OBJECTIVE: To investigate whether mammalian target of rapamycin complex 1 (mTORC1) inhibitors could reduce seizure frequency in children with tuberous sclerosis complex (TSC). METHODS: Due to slow inclusion rate, target inclusion of 30 children was not reached. Twenty-three children with TSC and intractable epilepsy (age 1.8–10.9 years) were randomly assigned (1:1) to open-label, add-on sirolimus treatment immediately or after 6 months. Sirolimus was titrated to trough levels of 5–10 ng/mL. Primary endpoint was seizure frequency change during the sixth month of sirolimus treatment. RESULTS: Intention-to-treat analysis showed sirolimus treatment resulted in 41% seizure frequency decrease (95% confidence interval [CI] −69% to +14%; p = 0.11) compared to the standard-care period. Per protocol analysis of 14 children who reached sirolimus target trough levels in the sixth sirolimus month showed a seizure frequency decrease of 61% (95% CI −86% to +6%; p = 0.06). Cognitive development did not change. All children had adverse events. Five children discontinued sirolimus prematurely. CONCLUSIONS: We describe a randomized controlled trial for a non-antiepileptic drug that directly targets a presumed causal mechanism of epileptogenesis in a genetic disorder. Although seizure frequency decreased, especially in children reaching target trough levels, we could not show a significant benefit. Larger trials or meta-analyses are needed to investigate if patients with TSC with seizures benefit from mTORC1 inhibition. This trial was registered at trialregister.nl (NTR3178) and supported by the Dutch Epilepsy Foundation. CLASSIFICATION OF EVIDENCE: This study provides Class III evidence that sirolimus does not significantly reduce seizure frequency in children with TSC and intractable epilepsy. The study lacked the precision to exclude a benefit from sirolimus.
Commentary
Childhood onset intractable epilepsy remains the most common and defining morbidity of tuberous sclerosis complex (TSC) (1). Traditional antiepileptic drugs and dietary therapies fall woefully short in most children to induce a long-term seizure remission, and to allow learning and independence. Epilepsy surgery can be offered to few selected children despite the multi-modal presurgical evaluation with newer tests and tools designed to map the “epileptogenic” tuber complex and a safe surgical strategy (2). After discovery that the mTOR pathway upregulation is the pivotal mechanism of TSC pathology in human disease, interest in using already available mTOR inhibitors was only natural. Clinical trials led to the FDA approval of mTOR inhibitor drugs in selected cases of subependymal giant cell astrocytomas (SEGA), renal angiomyolipomas, and pulmonary lymphangioleiomyomatosis (LAM). However, proving the usefulness of mTOR inhibitors in epilepsy turned out to be much more difficult. Preclinical studies and mouse models have shown that mTORC1 inhibition can prevent and reverse epileptogenesis in TSC (3, 4). Treatment with mTORC1 inhibitors has been shown to fully rescue the phenotype, suggesting that mTORC1 inhibitors might be useful for targeted treatment of epileptogenesis in TSC (4). However, such convincing evidence is yet to be established in human epilepsy thus far. This could partly be due to the multi-faceted pathophysiological mechanisms set in motion at the onset of epilepsy, the influence of other environmental and genetic factors, as well the complexity inherent in defining and measuring outcomes in epilepsy associated with a rarer disease especially in cognitively impaired children. Overwater and colleagues should be congratulated on conducting an astutely designed trial of sirolimus in epilepsy for TSC. Unfortunately, the trial was negative and could not confirm the benefit of sirolimus in epilepsy. However, it lacked the precision to exclude a possible benefit from sirolimus, and therefore has set the bar for future trials.
This trial was an add-on, open-label sirolimus treatment in 23 children (age 1.8–10.9 years) with TSC. The one year trial had a cross-over design divided into two phases of 6 months each: the sirolimus phase (SP), when sirolimus was added to the patient's antiepileptic drug regimen; and the standard antiepileptic drugs phase (SADP), without sirolimus. Each patient was randomly assigned to SP or SADP in the first phase and then crossed over to the alternative phase. Primary outcome measures of median seizure frequency were 25 seizures per month (interquartile range [IQR], 7–47) after the SP versus 32 (IQR, 9–62) after the SADP compared to 35 (IQR], 20–65) at baseline. Patients in the SP had 41% fewer seizures compared with the SADP. 14 patients reached the target sirolimus level (>5 ng/mL) during the last month of the SP with a higher mean seizure frequency reduction of 61%, suggesting dose (level) response relationship. Nine children were responders (>50% seizure reduction) and three became seizure free in SD, while in SADP, six were responders and one became seizure free. Considering that mTOR inhibitors target specific pathobiological derangements in the TSC related epilepsy, the reduction in seizure frequency and responder rates were neither universal nor dramatic especially when compared to responses seen in SEGA. The lack of robust seizure response could be due to late intervention after epilepsy onset, inability to deliver adequate doses of sirolimus without intolerable side effects, and/or underlying additional primary and/or secondarily mechanisms set in motion by epilepsy. Another recently concluded large (n=366) phase 3, randomized, double-blind placebo controlled everolimus trial for epilepsy in TSC has shown roughly similar results (5). In this 18-week trial, median reduction in seizure frequency was 14.9% with placebo versus 29.4% with low-exposure and 39.6% with high-exposure everolimus. Responder rates were 15.1% with placebo compared with 28.2% for low-exposure and 40% for high-exposure everolimus (5). Only an open label study of everolimus in 20 TSC epilepsy patients reported more dramatic results of 60% responder rates and 73% median seizure reduction (p<0.0001) (6). Another inherent difficulty in TSC is quantification of seizure burden and seizure response. TSC patients have a high frequency of epileptic encephalopathy with interictalictal continuum of epileptiform spike and waves that is sometimes sleep-state dependent, and clearly demarcating an ‘electro-clinical ictal event’ can be challenging even on a VEEG let alone by patient diary or a seizure log. It is therefore not surprising that in the secondary outcome measures, the trial reported no significant differences in EEG or adaptive behavior measures, cognitive or motor development, behavior problems, or proportions of patients with secondarily generalized seizures.
Unknown optimum duration of treatment and adverse effects of mTOR inhibitors, especially immune suppression, is another issue for use in children. During the SP, serious adverse events occurred in four patients (3 hospitalized with pneumonia, 1 otitis media) and five patients discontinued sirolimus owing to adverse events. Reduction in sirolimus dose owing to adverse events was required in another 12 children.
One of the best aspects of this trial is the relatively longer 6-month duration of sirolimus treatment, more than most epilepsy drug trials and the recently concluded phase 3 randomized trial (18 weeks duration) (5), allowing sufficient period to assess the clinically relevant effect on seizure frequency. Also the cross-over design was quite powerful in showing the effect in a rare disease with a relatively small number of patients. The trial's limitation was lack of a placebo arm and unmasked treatment planned to ease the burden of recruitment based on input from the TSC parent association. Despite foregoing placebo group, the trial fell short of its target of recruiting 30 patients. Because of adverse effects, only 14 patients could reach the target trough of sirolimus during which higher rates of seizure reduction were seen. It is unclear if the sirolimus effect on seizure frequency is dose related, and if adverse effects could be proactively managed more aggressively to reach the target blood levels. Despite best efforts, several changes in the antiepileptic drug regimen were necessary particularly during the SADP. This could have further contributed to the observed seizure reduction during the SADP, and possible underestimation of the sirolimus effect during the SP. The trial provides class III evidence that sirolimus does not significantly reduce seizure frequency in children with TSC and intractable epilepsy. The benefits and risks of mTOR inhibitors in treating epilepsy in TSC are yet to be established. In addition, we do not know if there is a downside of mTOR inhibition in a developing brain but that could be less of a concern as an undertreated epilepsy in TSC is equally if not more devastating and life-threatening.