Studies have compared lorazepam, phenobarbital, diazepam, and phenytoin to phenytoin alone and determined that lorazepam was most likely to terminate convulsive SE. It was statistically significantly more likely to terminate SE than phenytoin although not statistically superior to the other agents [15]. It has emerged as the drug of choice for the initial treatment of SE because it is also faster and more convenient to administer than the other drugs tested. The dose of lorazepam studied was 0.1 mg/kg; lower doses may be efficacious but have not been systematically studied. At all time points studied, patients randomized to receive lorazepam were more likely to have stopped clinical seizure activity than those who received diazepam (or placebo) [15]. Confounding these studies, however, is the use of EEG. Further investigations in these studies went on to show that many of the patients who were no longer clinically seizing after “successful treatment” of SE were in fact still in NCSE. Thus, it is likely that studies only looking at clinical symptoms overestimate the efficacy of treatment. Recent studies have compared prehospital administration of IM midazolam with IV lorazepam, testing the theory that a rapidly absorbed IM drug would yield the same success rates as an IV drug because the IM drug could be administered more rapidly. The dose of midazolam was 10 mg and for lorazepam was 4 mg, and if patients weighed between 13 and 40 kg, the doses were halved. The study showed that the more rapidly administered midazolam was superior at terminating SE than lorazepam, likely because of the ease with which it could be given [16]. Most clinicians still agree, however, that if an IV line is already in place (such as in the inpatient setting), then IV lorazepam should be administered.

One of the most important confounding factors in assessing the efficacy of treatment of SE is the latency from seizure onset to treatment. Studies have investigated this based on the EEG pattern at the time treatment was initiated [11]. Table 1 shows the likelihood of terminating SE based on the initial EEG pattern, while Figs. 8, 9, and 10 demonstrate different SE patterns. While it is difficult to know the time of seizure onset, the progression from one EEG pattern to the next is well established and provides an estimate of the latency to treatment. For example, patients with a burst suppression pattern likely have been seizing for longer than those with a waxing-and-waning pattern. Patients classified as having subtle SE most likely have been experiencing seizures for a long period of time or have a catastrophic underlying condition, which helps to explain their poor response rates to any of the conventional antiepileptic drugs [11].

Available data shows that only the first anticonvulsant has a reasonable chance of terminating SE (Table 2) and does not have any implications on preventing recurrence [15]. While fosphenytoin is the only AED to be approved by the US Food and Drug Administration (USFDA) for the termination of SE, it is not superior to phenytoin as a first- or second-line agent. It is used because it carries a lower risk of complications related to intravenous infusion, although complications like bradycardia and hypotension are still possible. Other second-line agents with published data include valproate, levetiracetam, lacosamide, and topiramate (given enterally as no parenteral form is currently available). There is insufficient data to recommend one over the other at this time. However, the clinician should be familiar with the reasons to choose one agent over another, e.g., side effects or interactions with other medications.

Regarding second-line agents, many believe that if SE has lasted through the time that it took to administer and assess the efficacy of the first-line AED, a more definitive treatment should be employed for the second-line treatment – usually an anesthetic agent (summarized in Table 3). Endotracheal intubation and mechanical ventilation should be started prior to the anesthetic agent. At this point, cEEG is necessary because these agents will usually terminate all movements before abolishing GCSE, and the only way for the intensivist to titrate the anesthetic agent to resolution of status epilepticus is via cEEG monitoring. Vasopressor and inotropic support is almost always necessary, as is support of core body temperature since anesthetic agents will often lead to poikilothermia. Enteral feeding may be possible, but parenteral nutrition may be required if an ileus develops. Infections are common, at least in part due to immunosuppression.