The blood–brain barrier hypothesis in drug resistant epilepsy

Sir, We read with great interest this succinct, informative and focused update on the therapeutic approach to super-refractory status epilepticus (SRSE) by Shorvon and Ferlisi (2011). In particular, we were pleased to see the ‘blood–brain barrier failure’ displayed as a potential seizurogenic mechanism. The authors provide a comprehensive update on the treatment options, and a thoughtful rationale for the timing of interventions. We noted that the authors present in Fig. 1 and Table 1 a number of anti-SRSE choices spanning from drugs acting on γ-aminobutyric acid (GABA) receptors (benzodiazepines, anaesthetics and barbiturates) to anti-inflammatory drugs (corticosteroids) or to other treatments with direct or indirect, documented or unknown, mechanisms of action on neurons.

It is reasonable to assume that if and when GABA-ergic drugs or anaesthetics quickly (<1 h) achieve the desired anti-status epilepticus effect, then their action is consistent with the modulation of inhibitory synapses. However, anaesthetic drugs also have immunomodulatory effects partially overlapping with those of corticosteroids. These effects may become of therapeutic value only when an immediate short-term, GABA-mediated effect is unachievable (e.g. Stage IV in Fig. 1; Shorvon and Ferlisi, 2011). Propofol or thiopental exert potent anti-inflammatory effects mediated by decreased NF-κB expression (Roesslein et al., 2008; Sanchez-Conde et al., 2008). Interestingly, sevoflurane, which has similar anaesthetic potency but lacks anti-inflammatory action, is not the first choice among halogenated anaesthetics for status epilepticus. Furthermore, sevoflurane may actually have an epileptogenic effect (Jaaskelainen et al., 2003). These anti-inflammatory mechanisms predict prevention of blood–brain barrier disruption or recovery of blood–brain barrier integrity; therefore, these effects may be comparable to what was observed with corticosteroids in status epilepticus or other forms of seizures (Verhelst et al., 2005; Marchi et al., 2009, 2011). Moreover, the authors point out that SRSE elicits as a consequence of stroke, trauma or infection, conditions where blood–brain barrier dysfunction and altered brain homeostasis were demonstrated to play a role.

Another therapeutic approach to refractory status epilepticus reviewed by Shorvon and Ferlisi (2011) is the use of intravenous infusion of magnesium sulphate. Magnesium has negligible blood–brain barrier permeability and in healthy individuals, brain levels exceed serum concentrations (Amtorp and Sorensen, 1974; Heath and Vink, 1998). Under conditions of disrupted blood–brain barrier (e.g. during SRSE), brain Mg²⁺ concentration may decrease (e.g. as a result of brain-to-blood leakage), leading to N-methyl-d-apartic acid (NMDA) receptor disinhibition. Systemic administration of high magnesium may restore brain levels of this ion, with a pronounced effect on excitatory synapses and thus on status epilepticus.

The most heterogeneous group of treatments for SRSE includes anti-inflammatory corticosteroids, vagal nerve stimulators, ketogenic diet and hypothermia. All these treatments have beneficial (protective or restorative) effects on the blood–brain barrier, and as predicted by the proposed aetiological role for blood–brain barrier disruption in status epilepticus, may terminate seizures by promoting cerebrovascular repair (Table 1).

Table 1.

Drugs and interventions used to control status epilepticus have anti-inflammatory effects that may result in blood-brain barrier protection/repairing

Therapeutic intervention	Classification/Drug class	Accepted mechanism of action	Anti-inflammatory potency	Predicted or demonstrated effects on BBB integrity	References
Propofol	Anaesthetic
	Short-acting hypnotic agent	GABA	Inhibits NF-κB	Protection/repair	Jaaskelainen et al., 2003; Sanchez-Conde et al., 2008; Schneemilch et al., 2005
Thiopental	Anaesthetic
	Short-acting hypnotic agent	GABA	Inhibits NF-κB	Protection/repair	Roesslein et al., 2008; Schneemilch et al., 2005
Ketamine	Anaesthetic
	Dissociating anaesthetic	NMDA antagonist	Inhibits NF-κB and IL-1β, TNF-α surge	Protection/repair	Beilin et al., 2007; Welters et al., 2010, 2011;
Magnesium	Electrolyte
		NMDA blocker	NA	Restores NMDA receptor blockade after BBB disruption	Amtorp and Sorensen, 1974; Heath and Vink, 1998
Vagal nerve stimulator	Device
		Unknown	Nicotinic receptors Ghrelin	Protection/repair	Cheyuo et al., 2011; Rosas-Ballina and Tracey, 2009; Rosas-Ballina et al., 2011
Ketogenic diet	Dietary regimen
		Unknown	NA	Protection/repair	Janigro, 1999; Nabbout et al., 2011
Hypothermia	Medical management
		Unknown	Inhibits NF-κB	Protection/repair	Oztas and Kaya, 1994; Polderman, 2009; Webster et al., 2009
Cortico-steroids	Anti-inflammatory agents
		Immunodepression	Similar to NFκB inhibition	Protection/repair	Marchi et al., 2009, 2011

BBB = blood–brain barrier; NMDA = N-methyl-d-aspartic acid; Nf-κb = nuclear factor kappa B; IL = interleukin; TNF = tumor necrosis factor; NA = not available.

Last but not least, even traditional antiepileptic drugs exert immunological effects. While an ample array of contradictory data exists, the fact that antiepileptic drugs interfere with the immune system hints at unconventional pathways leading to, or sustaining, seizure activity (Beghi and Shorvon, 2011).

In conclusion, we suggest that even ‘traditional’ or device-based interventions routinely used to treat refractory status epilepticus may act by improving brain homeostasis. This may be achieved by anti-inflammatory effects leading to blood–brain barrier repair. Whether similar mechanisms may be involved in the treatment of non-status epilepticus seizures remains to be investigated.

Funding

Supported by Epilepsy Foundation Research Grant awarded to Nicola Marchi and NIH R41MH093302 and R21HD057256 awarded to Damir Janigro.