The fact that pinealectomy induced violent convulsions in parathyroidectomiced rats as well as observations on the epileptogenic effects of melatonin antibodies intraventricularly injected into rats [
75,
76] was the first data suggesting a possible relationship between melatonin and epilepsy. In this context, experimental studies demonstrate that suppression of melatonin, by pinealectomy, increased the brain damage after kainic acid-induced seizures in rats, thus suggesting a neuroprotective role of melatonin [
77] on neural tissue. Melatonin was reduced in patients with epilepsy at baseline compared with controls and increased threefold following seizures [
78]. Other authors have also reported that patients with seizures of diverse origins show an alteration of the melatonin rhythm [
79,
80]. Recent studies, carried out in children with refractory epilepsy or febrile seizures [
81,
82], revealed a lowered level of melatonin in these children in comparison with those without seizures. Melatonin could control convulsive crises by acting on both
γ-aminobutyric acid (GABA) and glutamate receptors [
79,
80].
The first trials on the possible usefulness of melatonin in epilepsy were carried out on patients refractory to conventional therapies. Although melatonin alone, at a single evening dose of 5–10

mg, has been reported to reduce the frequency of epileptic attacks in children [
83], most studies have been focused on the use of melatonin associated with other conventional antiepileptic drugs (vigabatrin, valproate, phenobarbital, etc.). One of the first case reports was concerned with a young girl diagnosed of severe myoclonic epilepsy, with convulsive seizures since her first month of age; melatonin was added to the conventional treatment (Phenobarbital) when the patient was in a precomatose stage, at the age of 29 months. After one month of this therapy and for a year thereafter, the child's seizures were under control [
84]. In another study [
85,
86] on children with severe intractable seizures, a 3-month treatment with oral melatonin (3

mg/day, 30

min before bedtime) associated with a conventional antiepileptic drug achieved a significant clinical improvement in seizure activity during treatment, particularly during the night.
The only negative results were reported in a trial on a sample of six children treated with melatonin (5

mg at bedtime). Although melatonin had a positive effect on the patient's sleep disorder, four of six children had elevated seizure activity after treatment [
87]. This study had, however, serious weaknesses. One major factor is the small number and heterogeneity of patients: 6 cases with ages ranging from 9 months to 18 years; the comparison between neonatal, infant, or juvenile response to the same doses of melatonin seems inappropriate. Furthermore, the neurological lesions of the six patients are also markedly different. Moreover, the hypothesis of the proconvulsant effects of melatonin has never been confirmed in subsequent studies. Other authors [
88] also found that melatonin can be helpful for sleep disturbance in young people with significant neurological impairment although did not find a demonstrable influence of the indoleamine on seizure control.
Since melatonin plays a protective role against the oxidative stress and prevents neuronal damage associated with epilepsy, several clinical trials were focused on the evaluation of the changes in the oxidative status of the patients treated with melatonin alone or associated with other anticonvulsive drugs. In relation to this, 31 children with epilepsy receiving carbamazepine monotherapy [
89] and 31 other children treated with valproate [
90,
91], who were seizure-free at least for the last 6 months, were involved in a double-blind, randomized, parallel-group, placebo-controlled trial. The effect of add-on melatonin (6–9

mg/day for 14 days) on the antioxidant enzymes glutathione peroxidase and glutathione reductase demonstrated that melatonin is a putative neuroprotector in conditions involving oxidative stress such as that occurs in epilepsies.
Children with epilepsy have high rates of sleep problems due to seizures or anxiety, frequently associated to mental retardation. Twenty-five patients (16 males and 9 females; mean 10.5 years) with these characteristics were randomized to oral fast-release melatonin at bedtime (3

mg/day extended, in case of inefficacy, up to 9

mg/day at 3

mg/week steps) or placebo. Patients treated with melatonin improved their wake-sleep disorders although the seizure frequency was poorly influenced by the treatment [
92]. A more recent study, conducted on 23 children with intractable epilepsy treated with oral melatonin before bedtime for 3 months, noticed a significant improvement of both sleep-related phenomena and the severity of seizures in these patients [
93].
Although the antiepileptogenic properties of ramelteon, a selective melatonin receptor agonist recently patented, have been described in a rat model of chronic epilepsy [
94], clinical trial on its efficiency for human treatment has not been studied.