I’ve often wondered if the devastating outcomes observed in patients presenting in status epilepticus (SE), are far more influenced by the underlying etiology inciting this seizure activity than the time it takes us to control this neurological outburst. And while the HYBERNATUS study does not directly address this question, it does allow for some interesting musings.
Published in the NEJM, Legriel et al examined the effectiveness of therapeutic hypothermia to control seizure activity and improve clinical outcomes in patients presenting in SE (1). The authors randomized 278 patients admitted to 11 ICUs throughout France, to receive the addition of therapeutic hypothermia to standard care or standard care alone. The authors defined SE as continual seizure activity lasting greater than five minutes or two or more seizures without a return to baseline over the same time interval. Patients randomized to the hypothermia arm were rapidly cooled to 32-34 degrees Celsius using ice packs, cold saline, and surface cooling techniques. Continuous EEGs were deployed in both groups, and propofol boluses followed by infusions were titrated to control EEG evidence of seizure activity.
The median age of the patients enrolled was 57, 65% were male and 47% had a previous history of epilepsy. In the majority of patients (65%), the epileptic episode began prior to arrival to the hospital, and most patients required at least two medications to control their seizure activity. The median time to electrical seizure control on EEG was 80 minutes.
Patients in the hypothermia arm reached the targeted temperature 3.5-7.1 hours after randomization. This cooling process effectively decreased the frequency and duration of seizure activity. The median seizure duration was 75 and 90 minutes in the hypothermia and control arms respectively. 11% vs 22% of patients had EEG confirmed status, 6-12 hours after randomization. 31% vs 38% had refractory SE (defined as continuous or intermittent clinical seizures, EEG-confirmed seizures, or both despite two types of antiepileptic drugs up to 24 hours after randomization). And 17% vs 23% had super refractory SE (defined as ongoing or recurrent status epilepticus between 24 and 48 hours after the initiation of anesthetic treatment).
Despite these differences in the rates of EEG-confirmed status, improvements in patient outcomes varied little between the two groups. The authors found no difference in their primary endpoint, the rate of neurologically intact patients at 90-days (GOSE score of 5), which was seen in 49% and 43% respectively (p of 0.43). Nor did the authors observe a difference in the ICU mortality (9% vs 12%), in-hospital mortality (12% vs 15%), or 90-day mortality (13% vs 15%) between the two groups.
This is a somewhat difficult trial to interpret as hypothermia was utilized fairly early in the course of treatment. Only 35% had refractory status in the first 24-hours and only 19.8% had super refractory SE. And while it is safe to say that empiric hypothermia in all patients admitted to the ICU for SE requiring sedation and mechanical intubation is unlikely to lead to improved outcomes, its utility as a salvage therapy in the subset of patients who do have refractory SE is still unclear.
This study brings up an interesting side note. Is time to termination of seizure activity a relevant surrogate marker for improvements in true clinical outcomes? Trials that have demonstrated time to cessation of seizure activity as a poor prognostic marker have all been observational and thus can only demonstrate an association between time and poor outcomes (2). And so it is difficult to decipher whether patients had poor neurological outcomes because they had prolonged seizure activity, or they had prolonged seizure activity because of the severity of their underlying illness. In fact, no randomized trial has found that improvements in time to termination of the seizure activity has ever translated into improvements in clinical outcomes.
When Alldredge et al published the findings of their pre-hospital study examining the efficacy of IV lorazepam vs IV diazepam vs placebo for the treatment of SE in the NEJM in 2001, it was universally thought of to be in favor of IV lorazepam (3). The authors reported in patients randomized to receive IV lorazepam there was an 16.5% increase in the rate of patients who experienced termination of seizure activity upon arrival to the ED when compared to the IV diazepam group, and a 38% difference when compared to placebo. And yet despite this improvement in early termination of seizure activity the authors reported no difference in the rates of neurologic deficits between the 3 groups (16.9%, 17.9%, 14.3% in the lorazepam, diazepam and placebo group respectively).
In 1998, when The Veterans Affairs Status Epilepticus Cooperative Study Group published the findings of their four arm trial examining the efficacy of four treatment strategies for patients presenting to the ED in SE, Treiman et al again found IV lorazepam to be superior to the various other treatment strategies (4). The authors reported a statistically significant improvement in the number of patients controlled with IV lorazepam when compared to either IV phenobarbital, IV diazepam and phenytoin, or IV phenytoin alone. Despite this significant improvement in the rate of termination of seizure activity, the authors reported no difference in clinical outcomes at 30-days. The only factors that determined 30-day outcomes were response to the initial agent and presence of subtle non-convulsive SE.
The HYBERNATUS study was not dissimilar. The use of hypothermia led to far earlier termination of seizure activity, fewer episodes of refractory and super refractory SE, and yet the authors found no difference in clinically important outcomes. This very well may mean that time to termination of seizure is not a clinically meaningful variable or that the effect size is too small to be detected in these diminutive study populations.
I am sure these musings will be mistaken by some as therapeutic nihilism. I am in no way advocating we should discontinue our current aggressive management of SE. Rather offering this post to serve as a reminder that seemingly important surrogate measures do not necessarily translate into clinically important patient outcomes. The violent spasms that accompany the neurological storms are often a symptom, representative of an underlying etiology that is likely a far more important determinant of patient outcomes.
Sources Cited:
- Legriel S, Lemiale V, Schenck M, et al. Hypothermia for Neuroprotection in Convulsive Status Epilepticus. N Engl J Med. 2016;375(25):2457-2467.
- Legriel S, Azoulay E, Resche-Rigon M, et al. Functional outcome after convulsive status epilepticus. Crit Care Med 2010;38: 2295-303.
- Alldredge BK, Gelb AM, Isaacs SM, et al. A comparison of lorazepam, diazepam, and placebo for the treatment of out-of-hospital status epilepticus. N Engl J Med 2001;345:631-7
- Treiman DM, Meyers PD, Walton NY, Collins JF, Colling C, Rowan AJ, Handforth A, Faught E, Calabrese VP, Uthman BM, Ramsay RE, Mamdani MB.. DVA Status Epilepticus Cooperative Study Group: A comparison of four treatments for generalized convulsive status epilepticus. N Engl J Med 1998;339: 792–798
- EM Nerd-The Case of the Partial Cohort - May 24, 2020
- EM Nerd: The Case of the Sour Remedy Continues - January 20, 2020
- EM Nerd-The Case of the Adjacent Contradictions - December 23, 2019
Great post, raises some very interesting questions. This is admittedly a big stretch, but if you re-analyze the Alldredge study by combining the lorazepam and diazepam groups (benzo-positive) vs. placebo (benzo-negative) some interesting findings emerge. The mortality rate in patients who received benzodiazepines was 6% (8/134) vs. 15% (11/71) among patients who didn’t receive benzodiazepines (p=0.04). Signal or noise? Who knows. Overall I find myself on both sides of the seizure argument. I continue to believe in the value of rapid seizure control initially (based largely on animal data showing neuronal cell death from prolonged status). However, I’m skeptical about… Read more »
Thanks Josh! I completely agree. I have no doubt that there is some benefit to early and aggressive control of SE. I think the question becomes more and more complex the more aggressive and burdensome our interventions become. Like you said burst-suppression comas come with much more baggage than a few doses of benzodiazepines at the onset of seizure activity. Most importantly I think we have to focus on patient centered outcomes as we try to answer these questions out. I have no doubt we can suppress the seizure activity if we try hard enough. But at what cost?