Juvenile-onset Alpers' syndrome is characterized by refractory epilepsy, predominating in the occipital lobe, with frequent episodes of spreading and status epilepticus, for which a combination of several anti-epileptic drugs (AEDs) is often required [
10]. Benzodiazepines are the main-stay in the treatment of status epilepticus, but often necessitate ventilatory support. Importantly, valproic acid should be avoided, since it carries a high risk of inducing hepatic failure. In our own experience, propofol and thiopental may also lead to hepatic failure and rapid neurological deterioration in children with Alpers' syndrome. In juvenile-onset patients with POLG1 mutations, these drugs should perhaps also be avoided. When the combination of several high-dose AEDs could not stop focal or generalized status epilepticus in our two patients, we chose to add magnesium according to the use in (pre-)eclampsia [
11]. The anticonvulsive effect of magnesium has been well established in (pre-)eclampsia, but it is rarely used in the treatment of epilepsy of other origin. In (pre-)eclampsia, an immediate effect is pursued by a loading dose of magnesium sulphate followed by continuous infusion for 24 h, resulting in serum magnesium levels of 1.7–3.5 mmol/l [
15], but even at these high doses, side-effects are mild and sparse [
11,
16]. In particular, renal or hepatic failure due to high-dose magnesium therapy have not been described. Although magnesium is effective in preventing and treating eclampsia, its mechanism of action remains unclear. One of the suggested anticonvulsant mechanisms is its action as a
N-methyl-
d-aspartic acid (NMDA-) receptor antagonist. The NMDA-receptor is a glutamate receptor which, when activated, leads to a massive depolarization of neuronal networks and bursts of action potentials. Magnesium may act to increase the seizure threshold by blocking the NMDA-receptors [
15]. Animal models suggest that AEDs enhancing GABA-receptor activation (e.g., barbiturates and benzodiazepines) are effective in the treatment of status epilepticus but resistance often develops over prolonged status epilepticus. Ketamine, which is also an NMDA-receptor antagonist, however, showed no effect when administered within the hour after onset of status epilepticus, but was effective in treating prolonged status epilepticus, suggesting a shift from inadequate GABA-ergic inhibition to excessive glutamergic excitation in refractory status epilepticus [
17,
18]. Magnesium also acts as a voltage-dependent calcium channel antagonist and prevents membrane depolarizations, which could be an additional mechanism of anti-epileptic action [
19].
The potential use of magnesium in epilepsy is adapted from practice in eclampsia [
11]. While magnesium as an anticonvulsant agent has been tested to be safe in (pregnant) women, the safety of using magnesium for treating epilepsy in adolescents with juvenile-onset Alpers' syndrome has not been studied. We cannot of course be certain that the magnesium infusion was the key factor in terminating otherwise refractory status in our two patients, particularly in the second case who resolved some hours later. However, the timing in the first, together with biological plausibility, is persuasive. Given the difficulties encountered while treating refractory status epilepticus in patients with juvenile-onset Alpers' syndrome, and the potential danger of some commonly used antiepileptic drugs in this syndrome, additional magnesium therapy should be considered.