Fenfluramine in clinical practice: new therapy option for Dravet and Lennox–Gastaut syndromes

Fenfluramine is derived from amphetamine, which modulates serotonergic neurotransmission, and as a result, it produces an overall increase in serotonin levels in the synaptic cleft by increasing central serotonin release and simultaneously reducing presynaptic serotonin reuptake [16, 17]. Fenfluramine acts directly on the serotonin receptors (HTR) including 5‑HT_1D and 5‑HT_2C receptors through its main metabolite, norfenfluramine, and seems to additionally target the sigma‑1 receptor, which belongs to a group of chaperone proteins. Upon activation, the sigma‑1 receptor modulates various voltage-gated ion channels and neurotransmitter receptors, which influence excitatory and inhibitory neurotransmission [16]. The synthesis, release, reuptake, and metabolism of serotonin in the nervous system is illustrated in Fig. 1, along with the connections of the serotonergic system in the central nervous system. Fenfluramine is administered orally, is fat-soluble, and has a bioavailability which is not influenced by nutrition. A steady state is usually reached after 4 days and the plasma-elimination half-life is 20 h, with primarily renal elimination. The FFA solution contains 0.627 mg/mL corn-based glucose.

Antiseizure effects were first noted in a report by Aicardi and Gastaut [18] in which FFA was used as an amphetamine-like derivative to treat psychosis and autism. Three of the patients in that study who also experienced photosensitive epilepsy with self-induced seizures showed significant improvements with the use of FFA. In addition to treating self-induced seizures [18], Gastaut achieved a seizure reduction of more than 50% in approximately half of the patients with severe childhood-onset epilepsy (46%) who were treated. With regard to the reduction of self-induced photogenic epileptic activity [19], Boel and Casaer described the use of FFA in 11 patients with treatment-resistant and self-induced epilepsy [20], of whom eight were photo- or pattern-sensitive. With FFA as an adjunct therapy, 64% of patients (n = 7) were seizure-free and the remaining 36% showed a seizure reduction of ≥ 75%, with five of these patients retrospectively diagnosed with DS. A royal Belgian decree was subsequently granted to Ceulemans et al., which provided permission for the use of FFA in the treatment of DS patients. The average dose for the 12 DS patients was 0.34 mg/kg body weight (0.1–0.9 mg/kg bw/day; [21]). Following a median 11 years of treatment (1–19 years), 58% of patients (7 of 12) had seizure-free intervals of at least 1 year (average 6.6 years). The cohort in that study included the five DS patients identified in the study by Boel and Casaer in 1996 [20]. The follow-up of this extended cohort by Ceulemans et al. revealed that three of the patients remained seizure-free, while four patients showed seizure-free phases greater than 12 years [22]. These positive results were confirmed in a subsequent prospective study with nine DS patients, whereby seven patients showed reductions in convulsive seizures of ≥ 50% [23].

These observations led to double-blind, placebo-controlled randomized phase III studies in DS and LGS patients, which proved the efficacy of FFA treatment for these conditions. The good responder rates (> 50% and > 75% reduction) and in some cases even seizure freedom were confirmed in the following compassionate-use programs (CUPs) before the drug was launched [24‐27]. In addition, FFA has been investigated in an open-label pilot study for Sunflower syndrome [28], in a case series for SCN8A-related epilepsy [29], and a proof-of-principle study assessing the use of FFA in Doose syndrome (myoclonic–astatic epilepsy, MAE) that is currently in progress. Studies with FFA for CDKL5 [30] and DS syndromes for infants younger than 2 years of age are in the planning stage. Furthermore, an FFA registry to capture long-term efficacy and safety effects is underway. Study details are presented in Table 1.

Electrocardiograms (ECGs) were recorded in each clinical appointment of the phase II and III studies and all results were normal; therefore, the ECG was no longer required as part of the routine examination. However, it is advisable to record a baseline ECG before initiating FFA therapy, as this is a common practice prior to starting any new ASM treatment [37].

When FFA was used as an appetite suppressant, the main valvular changes involved valve insufficiencies on the left side of the heart, which can be explained by the higher density of 5‑HT_2B serotonin receptors on this side [38]. Thickening of the mitral and/or aortic valve led to valvular heart disease that required surgery in some cases. The development of severe pulmonary arterial hypertension (PAH) with possible lethal outcomes was also observed [12, 13]. Therefore, the FFA-approval trial had strict exclusion criteria. For example, patients with evidence of a “physiological” MI during screening were excluded.

During the phase III trial (average treatment duration, 8.4 months), valvular heart disease and PAH were absent in all patients. However, inconsistent “trivial” MIs were observed during echocardiography monitoring, which were classified as physiological and thus normal. The frequency of these valve insufficiencies in DS patients was 17.9% in the low-dose (0.2 mg/kg bw/day) FFA group, 22.5% in the high-dose (0.7 mg/kg bw/day) group, and 12.5% in the placebo group [39]. These insufficiencies were of no clinical relevance and occurred intermittently. A progression to valve insufficiency, which was a criterion for discontinuation, did not occur in any patients.

No relevant cardiopulmonary side effects were observed in the original Belgian cohort (DS patients) over the course of up to 30 years. In these two studies, a daily low dose of FFA of 7.5–20 mg was administered and patients were carefully monitored and examined [22, 23].

In the European CUP, any normal findings such as “trivial” or “physiological” MIs or low-grade MIs that were not significantly pathological were not considered contraindications if there was no pre-existing anatomical pathology of the valves. However, it was important to carefully follow up patients with low-grade MIs to avoid overlooking a progressive worsening. Even greater caution is required in cases of low-grade AI, which is less frequent in children than in adults. With increasing age, low-grade valvular insufficiencies are common in adults.

A standardized complete echocardiographic examination should be performed according to the European Association of Cardiovascular Imaging (EACVI) and the American Society of Echocardiography (ASE) guidelines, which include ECG recording and blood pressure measurement.

The optimal configuration of the device is essential for comparability of follow-up examinations, and this includes appropriate penetration depth and color Doppler settings. Insufficiencies greater than a low grade (grade I) in the mitral or aortic valve and right ventricular pressure of > 27 mm Hg as estimated by tricuspid regurgitation are contraindications for treatment with FFA. In addition, new-onset AI and MI detected during follow-up examinations must be monitored at shorter intervals than those of patients without these conditions. Follow-up examinations should occur at 6 months after therapy discontinuation for the safety of the patient.

Before introducing FFA as an add-on therapy, a review should be performed of the prior ASMs administered in terms of their efficacy in previous courses of treatment. A polytherapy that includes more than three to five medications should be avoided; therefore, ASMs with lower efficacies should be gradually discontinued before the introduction of a new medication. Special attention should be given to combinations with stiripentol because FFA is metabolized via multiple CYP enzymes (in particular CYP1A2, CYP2B6, and CYP2D6) and drugs that inhibit any of these enzymes may increase serum concentrations of FFA. This interaction can enhance its therapeutic effect but also its side effects, as is the case with stiripentol. Therefore, the FFA dose must be reduced when used concurrently with stiripentol (Table 3). Moreover, it should be considered that drugs that induce corresponding enzymes (e.g., phenobarbital) can reduce the plasma level of FFA and consequently its efficacy.