A retrospective observational study of current treatment for generalized convulsive status epilepticus

Highlights

• An observational study of treatment for generalized convulsive status epilepticus (GCSE) is performed.

• Majority of patients with GCSE presenting for emergency treatment receive benzodiazepine.

• Benzodiazepine, as first-line treatment, was effective in 56% of the patients.

• A second-line treatment was effective in an additional 28% of the patients, and a third-line treatment was effective in 12% of the patients.

• Phenytoin was the most prescribed but least effective second-line treatment.

Abstract

Objectives

This study aimed at determining the current state of practice of treatment for acute generalized convulsive status epilepticus (GCSE) and responsiveness to therapy.

Methods

This observational study was performed by retrospectively identifying patients with GCSE presenting to an emergency room setting. The primary outcome was seizure cessation following medication administration. Secondary outcomes were rates of intubation and mortality.

Results

One hundred seventy-seven episodes of GCSE were identified. All patients, except 1, received a benzodiazepine for first-line treatment. Only 11% of these patients, all children, were treated with at least 0.1 mg/kg of lorazepam or an equivalent dose of an alternative benzodiazepine. A first-line treatment was effective in 56% of the patients, a second-line treatment in an additional 28%, and a third-line treatment in 12%. Phenytoin was the most prescribed second-line treatment (41%) but statistically significantly least effective (22% versus 86% seizure cessation, p < 0.0001) compared with all other second-line agents together. Propofol was the most prescribed third-line treatment.

Conclusions

Results emphasize that, in clinical practice, approximately half of GCSE patients respond to first-line therapy and, among nonresponders, approximately two-thirds respond to second-line and approximately three-quarters respond to third-line therapies. The variations in treatment selection reflect that there are no randomized controlled trials to guide treatment beyond use of benzodiazepines for first-line treatment. The observation that phenytoin is statistically substantially worse than other second-line treatments raises the possibility that the most commonly selected second-line treatment is the least effective and provides equipoise for a large randomized controlled trial of second-line therapies.

Introduction

Status epilepticus (SE) is a common medical and neurological emergency. The Richmond-based population study projected an estimated 126,000–195,000 cases of SE per year in the United States with annual mortality of 22% despite aggressive medical treatment [1]. Along with high mortality, SE is associated with significant morbidity, including cognitive dysfunction and risk of subsequent epilepsy [2].

Benzodiazepines (BZDs) are considered first-line treatment based on randomized controlled trials. The VA Cooperative Study, a large multicenter randomized controlled trial, compared 4 treatments for SE and found that, in patients with overt SE, 0.1 mg/kg lorazepam was the most successful, terminating episodes in 65% of the patients which was statistically significantly greater compared with phenytoin (PHT) and numerically greater compared with the other treatments [3]. Two other randomized controlled trials also suggest superiority of lorazepam [4], [5]. Although a recent double-blind placebo controlled trial found superiority of intramuscular midazolam over intravenous lorazepam for out-of-hospital initial treatment for status epilepticus, it is likely that many emergency physicians still use lorazepam [6]. It is probable that BZDs are used almost uniformly for patients with SE, but it is unclear what types or doses are being used in current clinical practice.

While there are excellent data to support the use of BZDs for first-line treatment, there are no large prospective randomized controlled trials to guide second-line treatment. The majority of data come from case reports or case series of individual drugs. There have been two single center randomized controlled trials that found equivalent efficacy of VPA and PHT [7], [8] and one which found VPA to be statistically significantly more effective than PHT (66% vs 42%, p = 0.046) [9]. Alvarez et al. recently reported a retrospective analysis of patients in a clinical protocol who received a standardized dose of benzodiazepine followed by a choice of PHT = 20 mg/kg, VPA = 20 mg/kg, or LEV = 20 mg/kg and found that VPA failed to control SE in 25.4% of the patients, PHT in 41.4%, and LEV in 48.3% [10].

There have been no observational studies to determine what the current state of clinical practice is for the treatment for GCSE. A 2003 survey of members of the critical care or epilepsy sections of the American Academy of Neurology reported that 95% of participants would use fosphenytoin or PHT as a second-line agent [11]; however, this study was based on a response to a clinical vignette and may not reflect true clinical practice. More recently, a 2012 survey of the American Critical Care Society also came to a similar conclusion [12].

The lack of rigorous systematic data to support decision-making beyond BZDs, especially the lack of data about second-line therapy, prompted us to study how GCSE is actually treated in routine emergency room practice. We hypothesize that treatment is likely to be highly variable, especially benzodiazepine use, and that some routinely used treatments may be more effective than others.