Reviewed by Shannon Hextrum, MD
Early coma induction for status epilepticus after benzodiazepine therapy shortens duration of status and ICU length of stay.
De Stefano P, Baumann SM, Semmlack S, et al. Safety and Efficacy of Coma Induction Following First-Line Treatment in Status Epilepticus: A Two-Center Study [published online ahead of print, 2021 May 27]. Neurology. 2021;10.1212/WNL.0000000000012292. doi:10.1212/WNL.00000000000122
Read the full article.
Background: The treatment approach for status epilepticus (SE) endorsed by international guidelines recommends benzodiazepines as first-line therapy, followed by additional non-anesthetic antiepileptic medications if seizures continue. Anesthetic agents and induction of coma is recommended after failure of these approaches. There are clinical scenarios that may warrant direct initiation of anesthetic agents and artificial coma after first-line agents have failed in status epilepticus. The risks of deep sedation and artificial coma must be considered as well. This study examines the risks and benefits of artificial coma after first-line antiepileptic therapy in the setting of SE.
Methods: This retrospective design includes data from patients at two Swiss medical centers between January 2017 and December 2018. The definition used for SE was clinical and electrographic seizures longer than 5 minutes duration or multiple seizures without recovery to clinical baseline. Various clinical and demographic data was collected, including etiology, severity and duration of SE, as well as illness severity scores. Patients were excluded from analysis if the primary reason for artificial coma induction/mechanical ventilation was not for SE management. Also excluded were those patients with SE from hypoxic-ischemic brain injury after cardiac arrest. The primary outcomes assessed were return to functional baseline at discharge and inpatient mortality. Secondary outcomes included ICU and hospital length of stay, SE duration and complications arising during SE.
Results: 230 patients included in the analysis met criteria for SE, and 205 of those were treated with escalation of therapy after first-line treatment. 56 received artificial coma as the next step, while 149 received non-anesthetic antiepileptic medications (guideline treatment), and 34 of these 149 patients later required artificial coma. Those patients with early escalation after first-line treatment had a younger median age (64 vs. 69), presented more frequently with convulsive seizures, and had a greater severity of SE, as determined by a validated severity score. The primary outcome of inpatient mortality and return to neurologic baseline at discharge was not significantly different between groups. However, early coma induction was correlated with shorter duration of SE and shorter length of stay in ICU and hospital. There was no significant difference in SE complications between treatment groups. Patients in the standard treatment group who went on to require artificial coma were less likely to return to neurologic baseline when compared with the early coma induction group.
Commentary: This study suggests that early and aggressive treatment of SE with coma after first-line antiepileptics is associated with shorter duration of SE and even shorter ICU length of stay. Complications including rates of hypotension requiring vasopressors were not significantly different between treatment groups. However, when comparing all patients receiving induced coma (early coma vs. guideline directed), significantly more patients had hypotension requiring vasopressors when they had been escalated from the guideline treatment group versus those patients with early induced coma. This suggests that if SE is severe enough to ultimately require artificial coma, early treatment with anesthetics is preferable.
Patients in the early coma induction group were more likely to suffer from convulsive SE, and perhaps this type of status is the appropriate condition for such aggressive initial treatment. This result also suggests confounding in the group selected for artificial coma, which may be expected in this observational, retrospective study design. In fact, only one patient in the early coma induction group had SE classified as focal NCSE without coma, while 52 patients in the standard group were classified as such. Another important point about study design is the method by which SE was monitored. Options for EEG included continuous or routine EEGs of at least 30 minutes every 12 hours. The true duration of SE may not be effectively captured with such practice of routine EEGs. Nonetheless, this is a compelling study suggesting that early coma induction after benzodiazepine treatment may be safe and effective in severe, convulsive SE.