Introduction

Status epilepticus is a potentially devastating medical emergency that affects 60,000–120,000 Americans per year [1]. It is classically defined as a seizure that lasts more than 30 min, or repeated seizures that prevent the patient from regaining full consciousness [2]. Thus defined, its mortality is approximately 25%, compared to less than 5% from seizures shorter than 30 min. Its treatment is particularly relevant to the neurocritical care community, because it often requires prolonged use of anesthetic agents in an intensive care unit (ICU) [3].

Most patients with status epilepticus are successfully treated with benzodiazepines and conventional anti-epileptic drugs such as phenytoin, but some require anesthetic agents such as propofol, midazolam, or pentobarbital to suppress seizures [3, 4]. This requires endotracheal intubation, admission to an ICU, and continuous electroencephalographic (EEG) monitoring. Anesthetic agents are important in preventing cortical damage from sustained seizures [5], but eventually other means of seizure control are needed, so high doses of multiple conventional anti-epileptic drugs are given and anesthetics are weaned. In some patients this fails, leading to so-called refractory status epilepticus, defined as ongoing seizures in the face of multiple anti-epileptic medications [6, 7].

Refractory status epilepticus carries a grim prognosis [8]. Patients with uncontrolled seizures often succumb to complications of prolonged anesthesia and critical illness or the toxicity of anti-epileptic drugs [6]. However, remarkable recovery may occur if seizures are stopped, especially in cases without significant underlying brain damage [8]. Therefore, alternative methods of stopping seizures are crucial when conventional anti-epileptic drugs fail. One such method is electroconvulsive therapy (ECT), which case reports suggest may effectively treat refractory status epilepticus [9–13], possibly by inducing endogenous anti-convulsant pathways [14, 15]. We report here the effect of ECT on three consecutive patients with refractory status epilepticus treated at our institution. The patients’ surrogates provided informed consent for ECT, which was administered in accordance with federal and state laws. This case series was certified as exempt from review by our institutional ethics committee.

Case One

A previously healthy 32-year-old woman was admitted to another hospital with fever, headache, and seizures. Intravenous acyclovir was given for possible viral encephalitis, but she developed refractory status epilepticus. One week later, computed tomography (CT) of her brain revealed diffuse cerebral edema, and she was transferred to our Neuroscience ICU for further management. Extensive diagnostic evaluation did not reveal an etiology, although infectious encephalitis was considered most likely (Table 1). Despite high doses of multiple anti-epileptic drugs (Table 2), EEG monitoring revealed refractory status epilepticus whenever propofol and midazolam were weaned (Fig. 1). High doses of methylprednisolone were given for possible autoimmune encephalitis, with no abatement of seizures. Three successive trials of pentobarbital to achieve complete suppression of the EEG for 72 h resulted in no improvement. A trial of ketamine also failed to control her seizures. Six weeks into her illness, we weaned all anti-epileptic drugs and performed 4 cycles of ECT over a period of 5 days (Table 3). Propofol and midazolam were given between treatments to maintain burst-suppression on the EEG. ECT was administered via bifrontotemporal leads with a Somatics Thymatron System IV device (Somatics, LLC, Lake Bluff, IL, USA) programmed to deliver a pulse width of 0.5 ms, at the longest stimulus duration available to achieve the standard maximum output of 504 mC. Continuous EEG monitoring was used to ensure that the ECT was provoking generalized electrographic seizures (Fig. 1). Rocuronium was used for paralysis to avoid succinylcholine in a debilitated patient. On each day of therapy, at least three electrical stimuli were administered in succession. After the last round of ECT, the EEG showed that the patient’s seizures had resolved (Fig. 1), and she began to regain consciousness. Over several weeks, she became more conversant and interactive, and was discharged to an acute rehabilitation facility. Two months after discharge, she was seen in our clinic; she complained of retrograde amnesia for the months prior to her illness and anterograde amnesia for the duration of her stay in the ICU, but otherwise felt well. She scored 30 out of 30 on her Mini-Mental State Examination and had a normal neurological examination.

Table 1 Pertinent results of diagnostic investigations
Full size table
Table 2 Therapies used in an attempt to stop seizures
Full size table
Fig. 1
figure 1

Samples of electroencephalograms from case 1. Displayed are four electroencephalograms in longitudinal bipolar montage showing (a) bilateral periodic lateralized epileptiform discharges (PLEDs) prior to electroconvulsive therapy (ECT), b a focal left posterotemporal seizure (arrow) with concurrent right posterotemporal PLEDs (arrowhead) prior to ECT, c a generalized seizure associated with ECT, and d the background roughly 1 week following ECT demonstrating the return of some anteroposterior organization with resolution of PLEDs, new blink artifact (arrow), and persistent slowing in the region of prior epileptiform activity (arrowheads)

Full size image
Table 3 Details of electroconvulsive therapy used in case 1
Full size table

Case Two

A previously healthy 41-year-old woman developed fever, headache, and changes in her personality. Cerebrospinal fluid (CSF) analysis at an outside hospital revealed a moderate pleocytosis (Table 1), and she was treated with intravenous acyclovir for viral encephalitis. Fosphenytoin was started after a seizure, but she continued to have seizures and a decline in her mental status. Two weeks after her symptoms began, she was transferred to our Neuroscience ICU for management of refractory status epilepticus. Laboratory studies did not reveal the cause of her encephalitis, and an MRI study of the brain was normal (Table 1). Despite receiving high doses of multiple anti-epileptic drugs (Table 2), she continued to have electrographic seizures whenever anesthetic agents were weaned. A trial of ketamine was unsuccessful. She underwent four daily cycles of ECT using a protocol similar to that described in case 1, but her seizures continued, so she was given pentobarbital to achieve complete suppression of the EEG. After several days of pentobarbital infusion, she developed multiple complications, including multi-drug resistant nosocomial pneumonia and acute renal failure. At the request of her family, life support was withdrawn, and the patient died. Pathologic examination of the brain revealed an active inflammatory infiltrate consistent with ongoing meningoencephalitis; no organism was identified.

Case Three

A previously healthy 26-year-old woman experienced seizures after several days of flu-like symptoms. Analysis of her CSF at another hospital showed a mild pleocytosis (Table 1). Intravenous acyclovir and phenytoin were administered, but she continued to have seizures. One week into her course, she was transferred to our Neuroscience ICU for further management. An MRI revealed abnormal T2 prolongation and reduced diffusion in the left claustrum and superior frontal gyrus, suggesting infectious cerebritis. Laboratory testing did not reveal the etiology of her imaging findings, and she continued to receive broad-spectrum antibiotics and anti-viral drugs. Despite high doses of conventional anti-epileptic drugs, she experienced both focal and generalized electrographic seizures whenever anesthetic agents were weaned. This continued despite three trials of pentobarbital to achieve suppression of the EEG. Trials of isoflurane and ketamine were unsuccessful. High doses of methylprednisolone were given without any effect. Two months into her course, we administered ECT using a protocol similar to that in case 1. The first two cycles failed to provoke a generalized electrographic seizure, but the 3rd round succeeded after midazolam was pharmacologically reversed. After the 4th cycle of ECT, the patient’s seizures decreased in frequency but continued, so another four rounds of ECT were administered, for a total of eight cycles. During this time, she began to slowly awaken, and she was discharged to an acute rehabilitation facility 3 months after her symptoms began. She has been seen regularly in our clinic for several years. She has frequent seizures and mild difficulty with short-term memory and concentration, but she lives independently and has had no further episodes of status epilepticus.

Discussion

Status epilepticus can have multiple underlying causes. Some, such as hypoxic-ischemic brain injury, are associated with a mortality of 80% [1]. However, many etiologies of status epilepticus are much more benign, especially if they do not involve direct insults such as hypoxia or infarction [1]. If seizures can be stopped, patients in this second category can make a remarkable recovery, as demonstrated in cases 1 and 3.

Most cases of status epilepticus are not refractory to conventional treatment, but sometimes alternative methods are needed to stop seizures. In the cases described earlier, we turned to ECT after we had exhausted potent combinations of conventional anti-epileptic drugs, multiple trials of complete EEG suppression with anesthetic agents, and more infrequently used therapies such as inhaled anesthetics and ketamine. We chose ECT because experimental evidence indicates that it augments endogenous anti-convulsant neurotransmitters such as gamma-aminobutyric acid (GABA) and reduces neuronal metabolic activity [14, 15], producing a refractory period that can interrupt the continuous seizure activity seen in status epilepticus, a condition in which endogenous anti-convulsant mechanisms have failed [12]. Several case reports [9–13] (Table 4) and our experience as described earlier suggest that these theoretical anti-epileptic properties of ECT are in fact clinically effective. Furthermore, compared to other treatments for refractory status epilepticus, such as prolonged barbiturate coma, ketamine, inhaled anesthetics, or surgery to excise seizure foci [8], the safety of ECT has been proven during widespread use in the treatment of psychiatric disorders [16].

Table 4 Summary of published case reports describing the use of electroconvulsive therapy (ECT) for status epilepticus or frequent seizures
Full size table

In the absence of other scientifically proven therapies for refractory status epilepticus, we feel that our case series and the case reports cited above justify further use and study of ECT for this indication (Fig. 2). Based on this combined experience, a reasonable strategy would be to administer daily sessions of ECT for 3–8 days, with adjustment based on the patient’s response. This recommendation should be considered in light of the limitations of our study, including its small size and lack of controls. The efficacy of ECT should be demonstrated in a larger group of patients, preferably with a control group. To be feasible, such a study may require a crossover design in which patients are treated with either standard therapy (e.g., pentobarbital infusion) or ECT, and then receive the alternative therapy if the first treatment fails. We hope that our small success in using ECT to treat refractory status epilepticus will lead to such studies, and eventually to better care for this common and potentially devastating but treatable illness.

Fig. 2
figure 2

Suggested algorithm for treatment of refractory status epilepticus: ongoing seizures despite high doses of at least two anti-epileptic drugs (AEDs)

Full size image