Finding the Lesser of Two Evils: Treating Refractory Status Epilepticus

Commentary

Anesthetic Drugs in Status Epilepticus: Risk or Rescue? A 6-Year Cohort Study.

Sutter R, Marsch S, Fuhr P, Kaplan PW, Rüegg S. Neurology 2014. Feb 25;82:656–664.

OBJECTIVE: To evaluate the risks of continuously administered IV anesthetic drugs (IVADs) on the outcome of adult patients with status epilepticus (SE). METHODS: All intensive care unit patients with SE from 2005 to 2011 at a tertiary academic medical care center were included. Relative risks were calculated for the primary outcome measures of seizure control, Glasgow Outcome Scale score at discharge, and death. Poisson regression models were used to control for possible confounders and to assess effect modification. RESULTS: Of 171 patients, 37% were treated with IVADs. Mortality was 18%. Patients with anesthetic drugs had more infections during SE (43% vs 11%; p < 0.0001) and a 2.9-fold relative risk for death (2.88; 95% confidence interval 1.45–5.73), independent of possible confounders (i.e., duration and severity of SE, nonanesthetic third-line antiepileptic drugs, and critical medical conditions) and without significant effect modification by different grades of SE severity and etiologies. As IVADs were used after first- and second-line drugs failed, there was a correlation between treatment-refractory SE and the use of IVADs, leading to insignificant results regarding the risk of IVADs and outcome after additional adjustment for refractory SE. CONCLUSION: Our findings heighten awareness regarding adverse effects of IVADs. Randomized controlled trials are needed to further clarify the association of IVADs with outcome in patients with SE. CLASSIFICATION OF EVIDENCE: This study provides Class III evidence that patients with SE receiving IVADs have a higher proportion of infection and an increased risk of death as compared to patients not receiving IVADs.

Is Pentobarbital Safe and Efficacious in the Treatment of Super-Refractory Status Epilepticus: A Cohort Study.

Pugin D, Foreman B, De Marchis GM, Fernandez A, Schmidt JM, Czeisler BM, Mayer SA, Agarwal S, Lesch C, Lantigua H, Claassen J. Crit Care 2014;18:R103.

INTRODUCTION: Seizures refractory to third-line therapy are also labeled super-refractory status epilepticus (SRSE). These seizures are extremely difficult to control and associated with poor outcome. We aimed to characterize efficacy and side-effects of continuous infusions of pentobarbital (cIV-PTB) treating SRSE. METHODS: We retrospectively reviewed continuous electroencephalography (cEEG) reports for all adults with RSE treated with cIV-PTB between May 1997 and April 2010 at our institution. Patients with post-anoxic SE and those receiving cIV-PTB for reasons other than RSE were excluded. We collected baseline information, cEEG findings, side-effects and functional outcome at discharge and one year. RESULTS: Thirty-one SRSE patients treated with cIV-PTB for RSE were identified. Mean age was 48 years old (interquartile range (IQR) 28,63), 26% (N = 8) had a history of epilepsy. Median SE duration was 6.5 days (IQR 4,11) and the mean duration of cIV-PTB was 6 days (IQR 3,14). 74% (N = 23) presented with convulsive SE. Underlying etiology was acute symptomatic seizures in 52% (N = 16; 12/16 with encephalitis), remote 30% (N = 10), and unknown 16% (N = 5). cIV-PTB controlled seizures in 90% (N = 28) of patients but seizures recurred in 48% (N = 15) while weaning cIV-PTB, despite the fact that suppression-burst was attained in 90% (N = 28) of patients and persisted >72 hours in 56% (N = 17). Weaning was successful after adding phenobarbital in 80% (12/15 of the patients with withdrawal seizures). Complications during or after cIV-PTB included pneumonia (32%, N = 10), hypotension requiring pressors (29%, N = 9), urinary tract infection (13%, N = 4), and one patient each with propylene glycol toxicity and cardiac arrest. One-third (35%, N = 11) had no identified new complication after starting cIV-PTB. At one year after discharge, 74% (N = 23) were dead or in a state of unresponsive wakefulness, 16% (N = 5) severely disabled, and 10% (N = 3) had no or minimal disability. Death or unresponsive wakefulness was associated with catastrophic etiology (p = 0.03), but none of the other collected variables. CONCLUSIONS: cIV-PTB effectively aborts SRSE and complications are infrequent; outcome in this highly refractory cohort of patients with devastating underlying etiologies remains poor. Phenobarbital may be particularly helpful when weaning cIV-PTB.

Status Epilepticus: Impact of Therapeutic Coma on Outcome.

Marchi NA, Novy J, Faouzi M, Stähli C, Burnand B, Rossetti AO. Crit Care Med 2015;43:1003–1009.

OBJECTIVES: Therapeutic coma is advocated in guidelines for management of refractory status epilepticus; this is, however, based on weak evidence. We here address the specific impact of therapeutic coma on status epilepticus outcome. DESIGN: Retrospective assessment of a prospectively collected cohort. SETTING: Academic hospital. PATIENTS: Consecutive adults with incident status epilepticus lasting greater than or equal to 30 minutes, admitted between 2006 and 2013. MEASUREMENTS AND MAIN RESULTS: We recorded prospectively demographics, clinical status epilepticus features, treatment, and outcome at discharge and retrospectively medical comorbidities, hospital stay, and infectious complications. Associations between potential predictors and clinical outcome were analyzed using multinomial logistic regressions. Of 467 patients with incident status epilepticus, 238 returned to baseline (51.1%), 162 had new disability (34.6%), and 67 died (14.3%); 50 subjects (10.7%) were managed with therapeutic coma. Therapeutic coma was associated with poorer outcome in the whole cohort (relative risk ratio for new disability, 6.86; 95% CI, 2.84–16.56; for mortality, 9.10; 95% CI, 3.17–26.16); the effect was more important in patients with complex partial compared with generalized convulsive or nonconvulsive status epilepticus in coma. Prevalence of infections was higher (odds ratio, 3.81; 95% CI, 1.66–8.75), and median hospital stay in patients discharged alive was longer (16 d [range, 2–240 d] vs 9 d [range, 1–57 d]; p < 0.001) in subjects managed with therapeutic coma. CONCLUSIONS: This study provides class III evidence that therapeutic coma is associated with poorer outcome after status epilepticus; furthermore, it portends higher infection rates and longer hospitalizations. These data suggest caution in the straightforward use of this approach, especially in patients with complex partial status epilepticus.

“When you choose the lesser of two evils, always remember that it is still an evil.”

–Max Lerner

A lack of data always leads to differences of opinion. This is clearly the case when it comes to treatment of refractory status epilepticus. On the one hand, the harms of prolonged seizures on the brain are well accepted. Although the underlying etiology is the greatest predictor of outcome, there is extensive evidence that within a given diagnosis or severity, nonconvulsive seizures are associated with adverse acute physiologic effects and independently associated with worse outcomes. This has been demonstrated in some form in all age groups, via multiple different outcome measures, and in both the short and long term (1, 2). Recent studies have even shown the equivalent of a dose-response curve, in which the duration of nonconvulsive seizure activity was directly proportional to decline in neurological function (3). Thus, nonconvulsive seizure activity, especially if prolonged, is Evil #1.

On the other hand, anesthetic doses of continuous IV anti-seizure drugs (cIVADs; typically midazolam, propofol, pentobarbital, or thiopental), though highly efficacious for seizure cessation, are associated with significant risks of their own, mostly non-neurological. Common adverse events associated with cIVADs include prolonged sedation, infections (especially pneumonia), hypotension, and need for intubation. Thus, iatrogenic adverse effects of cIVADs are Evil #2. The three studies reviewed here—all from reputable centers with extensive experience in this area—attempt to shed light on this issue.

Sutter et al. reviewed 171 consecutive patients with status epilepticus (SE) treated in an intensive care unit, 63 (37%) of whom received cIVADs, and excluded patients post-cardiac arrest. They accounted for the major known predictors of outcome—about as well as can be done retrospectively—and found that the use of cIVADs was associated with increased mortality, as well as more frequent intubation (90% vs 25%), severe hypotension (16% vs 2%), and poor functional outcome on the Glasgow Outcome Score with long-term follow-up (78% vs 63%). The difference in outcome was most pronounced in those with milder seizure types (simple partial, complex partial, or absence SE) rather than convulsive SE or nonconvulsive SE in coma. Not surprisingly, the group receiving cIVADs was much sicker, with greater prevalence of stupor or coma, a higher rate of acute symptomatic etiology, and longer duration of SE. When adjusting for the refractoriness of the seizures, there were no longer any significant findings regarding the risks of cIVADs. It also important to note that at this center, barbiturates “were used to induce an isoelectric EEG,” a more aggressive approach than is typically used, although only 12 patients received barbiturates. There was no attempt to determine cognitive impairment or presence or absence of later epilepsy; these are the things one is trying to prevent when treating SE more aggressively.

My personal take-home message from this study (and reflective of the opinions of many others expressed via letters to the editor and comments at meetings where this study has been presented) is simply that those who needed cIVADs had more severe illness and more refractory SE than those who did not need them. It remains unclear whether the brain function that was potentially saved by controlling seizures was worth the risk of the cIVADs. The use of barbiturates to induce an isoelectric EEG may have contributed to the adverse effects of cIVADs seen in this study but probably only modestly.

Pugin et al. reviewed 31 patients treated with pentobarbital for refractory SE at their center. Their goal of treatment was seizure suppression, although 90% ended up in suppression-burst. Virtually all the patients had failed midazolam prior to receiving pentobarbital. Pentobarbital stopped seizures in 28 of 31 patients, but 48% had withdrawal seizures. Adding phenobarbital allowed for successful withdrawal of pentobarbital in most cases. With regard to safety: The need for vasopressors increased from 13 to 32 percent when pentobarbital was utilized; one-third developed pneumonia; 10% had a deep venous thrombosis; and 10% developed an ileus. One-third of patients had no identifiable complications. The median ICU stay was 30 days, with a median hospital stay of 48 days. At one year of follow-up, 74% had died or were unresponsive, 16% had moderate disability, and 10% had no or minimal disability.

The authors concluded that “complications are infrequent” (though seen in two-thirds of patients), that “treatment-associated morbidity was manageable in the ICU setting,” and that none of the above complications influenced the outcomes. It is also quite clear from this study that patients requiring pentobarbital were very sick, had very long hospital stays, and the majority—but certainly not all—had poor outcome. The authors stress that 10% of these cases of severe refractory SE (defined in this study as ongoing or recurrent SE 24h or more after starting cIVADs) had excellent outcome. This is similar to the conclusion of a recent multicenter study of patients with prolonged refractory SE (ongoing or recurrent SE after at least 1 week of cIVADs), where 14 of 63 patients had good outcome (moderate disability or better) and 6 (10%) had no disability (4).

My personal take-home message from this study: Intensivists are comfortable managing the complications of cIVADs and do not feel that these contribute significantly to worse outcomes. However, complications are common and outcomes are often poor, as already known.

Marchi et al. reviewed 467 consecutive adult cases of SE lasting 30 minutes or longer at their center, including those not in an ICU setting. Overall, 35% developed new disability and 14% died. They accounted for etiology, severity and comorbidities in their outcome analyses, variables that they report account for more than 90% of the mortality prediction in SE. They found that therapeutic coma (equivalent to use of cIVADs as used in this commentary) was highly associated with worse outcome at hospital discharge, including new disability (relative risk of 4.6) and mortality (relative risk 5.5). This effect was most marked for patients with milder seizure types (complex partial SE rather than generalized convulsive SE or nonconvulsive SE in coma), similar to the article by Sutter et al. Use of cIVADs was also associated with higher rates of infection and longer hospital stays. At this center, therapeutic coma was used with a goal of either seizure suppression or suppression-burst. Again, those treated with cIVADs were more likely to have more severe forms of SE (generalized convulsive and nonconvulsive SE in coma). Outcome after hospital discharge was not assessed.

The authors conclude that therapeutic coma should be used cautiously, especially in those with complex partial SE, but that it appears “fully justified” in those with generalized convulsive SE or nonconvulsive SE in coma. The second half of that statement did not make it into the abstract but is in their concluding paragraph. My personal take-home message is similar to that of the article by Sutter et al.

Concluding Comments

So, where does this leave us? Unfortunately, the same place we started. In studies by Sutter et al. and Marchi et al., the potential benefits of more aggressive treatment (including cIVADs) for long-term neurological function and prevention of later epilepsy—the main reasons more aggressive treatment is suggested—were not assessed (other than very crudely in Sutter et al. via the Glasgow Outcome Scale). Thus, both these papers address the potential harms of aggressive care in detail but not the main potential long-term benefits.

It is clear that patients requiring cIVADs—especially multiple cIVADs for prolonged periods—are critically ill and outcomes are often poor. There is at least one article (of which I was a co-author) that argues that more aggressive and higher doses of cIVADs are superior: Switching from a low-dose continuous IV midazolam for refractory SE to a more aggressive protocol including higher-dose continuous IV midazolam at one center was associated with fewer withdrawal seizures and significantly lower mortality (5). This paper has methodological limitations as well, especially the historical comparison group.

Prolonged or repetitive seizures, including nonconvulsive ones, should be identified and treated as quickly as possible with anti-seizure medication; that part is not controversial. My personal recommendation beyond that is as follows: When seizures are intermittent or focal, or if the patient is interactive, trying multiple nonsedating antiseizure medications before resorting to cIVADs is reasonable and perhaps wisest, but this should be done with loading doses and quickly (minutes to hours, not days). If this fails and there is a significant seizure burden, cIVADs are needed to stop seizures to minimize the chance of additional neuronal injury and its permanent effects on cognition and possible new development of epilepsy. When cIVADs are used, they should be used with a goal of seizure suppression (with continuous EEG monitoring to confirm this) and in an ICU setting with careful attention to the common but manageable adverse effects. When we have reliable biomarkers of seizure-induced neuronal injury, this can be done more rationally. If a patient has a progressive and untreatable disorder, a less aggressive approach may be more reasonable as there is no long-term function to improve.

Retrospective studies can never fully account for severity of illness or refractoriness of SE; when these studies are attempted, outcome should be long-term (at least a few months) and include cognition and newly developed epilepsy. It is not a fair comparison to look only at the adverse effects of a treatment and not the potential benefits (on long-term brain function, in this case). It remains possible that more aggressive very early treatment could be the most effective approach; this may minimize the early changes in the brain that lead to refractoriness. Only prospective, randomized trials might resolve this debate, and these will not be easy or cheap. Hopefully, they will be feasible.