Rational polytherapy in the treatment of cholinergic seizures

The initiation and maintenance phases of cholinergic status epilepticus (SE) are associated with maladaptive trafficking of synaptic GABA_A and glutamate receptors. The resulting pharmacoresistance reflects a decrease in synaptic GABA_A receptors and increase in NMDA and AMPA receptors, which tilt the balance between inhibition and excitation in favor of the latter. If these changes are important to the pathophysiology of SE, both should be treated, and blocking their consequences should have therapeutic potential. We used a model of benzodiazepine-refractory SE (RSE) (Tetz et al., 2006) and a model of soman-induced SE to test this hypothesis. Treatment of RSE with combinations of the GABA_AR agonists midazolam or diazepam and the NMDAR antagonists MK-801 or ketamine terminated RSE unresponsive to high-dose monotherapy with benzodiazepines, ketamine or other antiepileptic drugs (AEDs). It also reduced RSE-associated neuronal injury, spatial memory deficits and the occurrence of spontaneous recurrent seizures (SRS), tested several weeks after SE. Treatment of sc soman-induced SE similarly showed much greater reduction of EEG power by a combination of midazolam with ketamine, compared to midazolam monotherapy. When treating late (40 min after seizure onset), there may not be enough synaptic GABA_AR left to be able to restore inhibition with maximal GABA_AR stimulation, and further benefit is derived from the addition of an AED which increases inhibition or reduces excitation by a non-GABAergic mechanism. The midazolam-ketamine-valproate combination is effective in terminating RSE. 3-D isobolograms demonstrate positive cooperativity between midazolam, ketamine and valproate, without any interaction between the toxicity of these drugs, so that the therapeutic index is increased by combination therapy between GABA_AR agonist, NMDAR antagonist and selective AEDs. We compared this drug combination based on the receptor trafficking hypothesis to treatments based on clinical practice. The midazolam-ketamine-valproate combination is far more effective in stopping RSE than the midazolam-fosphenytoin-valproate combination inspired from clinical guidelines. Furthermore, sequential administration of midazolam, ketamine and valproate is far less effective than simultaneous treatment with the same drugs at the same dose. These data suggest that we should re-evaluate our traditional treatment of RSE, and that treatment should be based on pathophysiology. The search for a better drug has to deal with the fact that most monotherapy leaves half the problem untreated. The search for a better benzodiazepine should acknowledge the main cause of pharmacoresistance, which is loss of synaptic GABA_AR. Future clinical trials should consider treating both the failure of inhibition and the runaway excitation which characterize RSE, and should include an early polytherapy arm.