Should phenytoin or barbiturates be used as second-line anticonvulsant therapy for toxicological seizures?

Abstract

Introduction. Seizures are a common sequela of self-poisoning. However, their mechanism differs from seizures of other etiologies. Toxicological seizures result from alterations in the excitatory and inhibitory balance of otherwise normal neurons. In contrast, idiopathic or trauma related seizures usually start with a focus of abnormal neurons. For both forms of seizures, benzodiazepines are recommended as first-line therapy; however, there is debate about the use of phenytoin or barbiturates for second-line therapy. Methods and Results. In this article, we systematically review the evidence for the use of these drugs as second-line therapy for toxicological seizures. Barbiturates complement the anticonvulsant effect of benzodiazepines at the GABAA receptor by increasing the duration of chloride channel opening; phenytoin blocks voltage-dependent sodium channels to inhibit propagation from active electrical foci, an effect more useful for nontoxicological seizures. We found no randomized controlled trial comparing phenytoin and barbiturates in toxicological seizures refractory to benzodiazepines; similarly no trial was found comparing the use of these drugs in nonpoisoned patients. Animal studies indicate that phenobarbital has greater effectiveness than phenytoin for many poisons; a few case reports suggest a better response in patients. Conclusion. Despite the lack of high-quality clinical trial data, pharmacological knowledge and animal studies suggest that phenobarbital or thiopentone should be second-line agents for controlling toxicological seizures. The role of newer agents such as propofol and levetiracetam in toxicological seizures is currently unclear because of a lack of clinical or animal studies.

• Drug-induced seizures
• Barbiturates
• Toxin-induced seizures
• Phenytoin
• Phenobarbital

Acknowledgments

We thank Nick Buckley for critical review of the manuscript. ME is a Scottish Senior Clinical Research Fellow, funded by the Scottish Higher Education Funding Council and the Chief Scientist Office.

Declaration of interest

The authors report no conflicts of interest. The authors alone are responsible for the content and writing of this paper.