You Can Teach an Old Drug New Tricks

Commentary

Dravet syndrome (DS), a rare and severe epileptic encephalopathy of childhood, has long had treatment guided by expert consensus, yet recently has enjoyed star status among industry seeking DS indications for their compounds. In the last 2 years, both cannabidiol (CBD as Epidiolex) and stiripentol (Diacomit) have received Food and Drug Administration (FDA) indications for DS with at least 8 additional therapies in various stages of development. Fenfluramine, a drug previously marketed for weight loss in adults and removed from the market due to occurrence of cardiac valvulopathy and pulmonary hypertension, represents the next most likely candidate to receive FDA approval for DS on the heels of impressive results from 2 phase 3 trials.

Using a randomized, double-blinded, placebo-controlled design, fenfluramine was compared at 2 doses (base equivalent 0.2 and 0.7 mg/kg/d) to placebo as treatment for convulsive seizures in children with DS with primary end point reduction in monthly convulsive seizure frequency. ¹ At 0.7 mg/kg/d dosing, fenfluramine demonstrated an impressive 75% median seizure reduction (42% at 0.2 mg/kg/d) compared to 19% in the placebo group. These results compare favorably to phase 3 results of CBD ^2,3 (39%-49% median reduction) and stiripentol ⁴ (84% median reduction) with notable differences in seizure types studied. Although convulsive seizures were categorized as hemiclonic, tonic, clonic, tonic–atonic, tonic–clonic, and focal with clear motor signs in the fenfluramine trials, only clonic and tonic–clonic were included for stiripentol, and CBD did not include focal seizures in their primary end point. More importantly, patients treated with fenfluramine experienced longer mean durations of seizure freedom compared to placebo and 19% of patients experienced 1 or less seizures during the 14-week trial with the mean pretreatment baseline convulsive seizure frequency 40. This response is not only statistically significant but clinically significant for a population at considerable risk of sudden unexpected death in epilepsy (SUDEP). Although the initial phase 3 trial excluded patients on stiripentol due to absence of available pharmacokinetic data to evaluate dosage adjustments, the second trial exclusively enrolled patients on an antiseizure medication regimen including stiripentol. ⁵ The study was similarly designed, other than a maximum fenfluramine dose of 0.4 mg/kg/d, and patients enjoyed nearly identical favorable outcome to the initial phase 3 trial.

Although seizure reduction is immensely important for patients with DS, this diagnosis comes with a myriad of other comorbidities that warrant treatment. The duration of these studies was inadequate to determine the long-term impact meaningful seizure reduction may have on cognition, behavior, SUDEP, and sleep, yet results did provide a glimpse of the possibilities. Using the Behavior Rating Inventory of Executive Function to assess negative impact of fenfluramine on cognition, the authors found significant improvements from baseline in the Behavioral Regulatory Index and Global Executive Composite score for patients treated at 0.7 mg/kg/d, while scores declined for those on placebo. This is promising as the treatment paradigm for DS is shifting from seizure reduction to disease modification.

As encouraging as these results are, to truly understand how fenfluramine will fit into DS treatment several unanswered questions will need to be addressed. Both atypical absence seizures and myoclonic seizures were notably absent as primary seizure types measured in these studies. Myoclonic seizures occur in nearly 70% of patients with DS, while atypical absence occurs in at least 50%, representing a considerable burden. ^6,7 Recorded as “other” seizures, patients treated with high-dose fenfluramine experienced a 76% reduction, just reaching statistical significance (P = .0458) compared to placebo. The relatively high “other” seizure reduction seen in the placebo group (56%) underscores the difficulty in counting these seizures for study purposes, representing a likely source of considerable bias limiting any meaningful conclusions of efficacy for these seizure types. Additionally, CBD is expected to be a component of many DS treatment plans now that is has received FDA approval, yet patients on any type of CBD were excluded from the trials, as FDA approval of CBD was not achieved prior to the enrollment period. However, CBD has been allowed as concomitant treatment in the Expanded Access Program for fenfluramine and evidence describing this therapeutic combination is likely forthcoming. Finally, a number of medications commonly used to treat comorbidities of DS were excluded from these trials due to similar mechanisms of action. Such treatments include drugs for behavior (ie, SSRI, stimulants), sleep (ie, trazadone), and common ailments such as vomiting (ie, ondansetron), limiting our understanding of how fenfluramine will impact other concomitant treatment plans.

Fenfluramine demonstrated a favorable safety profile. The most frequent adverse effects included those commonly encountered in trials of antiseizure medications—namely diarrhea, decreased appetite, fever, lethargy, and nasopharyngitis. Perhaps more important than the adverse effects present were those that were absent. Cardiac valvulopathy received immense attention throughout treatment with standardized echocardiograms conducted intermittently throughout the trial and continuing into the open-label phase. No pulmonary arterial hypertension or significant cardiac valvulopathy was noted during the study or during the first 250 days of open-label follow-up. ⁸ When new trace mitral or aortic regurgitation arose during the study, it frequently resolved or remained unchanged on follow-up testing. Although long-term safety data are still being collected, several open-label studies in Europe have supported long-term cardiac safety when used for epilepsy at low dosing. ^9,10 Although patients with mitral or aortic regurgitation of any grade at baseline were excluded from the study, these findings would suggest that trace regurgitation is not likely a contraindication to treatment provided there is periodic cardiac follow-up. Weight loss is another expected adverse effect of treatment given the drug’s prior indication. Between the 2 trials, 17% of patients experienced weight loss greater than 7% of baseline weight with the majority remaining on therapy. Thus, while appetite suppression and weight loss may occur, it is often tolerable, particularly in the face of considerable efficacy.

Dravet syndrome has reaped the benefits of elevated interest into the epileptic encephalopathies, especially as it pertains to development of new treatments. Fenfluramine, a repurposed old therapy made new again, has benefits that appear to outweigh any potential risks. Even more refreshing is the realization that DS, as devastating an epileptic encephalopathy as it is, is not impenetrable to treatment. Meaningful seizure reduction and seizure freedom are possible, and we should continue working to ensure that outcome for all patients with the disorder.