Abstract
Prehospital Treatment With Levetiracetam Plus Clonazepam or Placebo Plus Clonazepam in Status Epilepticus (SAMUKeppra): A Randomised, Double-Blind, Phase 3 Trial.
Navarro V, Dagron C, Elie C, Lamhaut L, Demeret S, Urien S, An K, Bolgert F, Tréluyer JM, Baulac M, Carli P; SAMUKeppra investigators. Lancet Neurol 2016;15:47–55.
BACKGROUND: Generalised convulsive status epilepticus (GCSE) should be treated quickly. Benzodiazepines are the only drug treatment available so far that is effective before admission to hospital. We assessed whether addition of the antiepileptic drug levetiracetam to the benzodiazepine clonazepam would improve prehospital treatment of GCSE. METHODS: We did a prehospital, randomised, double-blind, phase 3, placebo-controlled, superiority trial to determine the efficacy of adding intravenous levetiracetam (2·5 g) to clonazepam (1 mg) in treatment of GCSE in 13 emergency medical service centres and 26 hospital departments in France. Randomisation was done at the Paris Descartes Clinical Research Unit with a list of random numbers generated by computer. Adults with convulsions lasting longer than 5 min were randomly assigned (1:1) by prehospital physicians to receive levetiracetam or placebo in combination with clonazepam. All physicians and paramedics were masked to group assignments. If the status epilepticus lasted beyond 5 min after drug injection, a second dose of 1 mg clonazepam was given. The primary outcome was cessation of convulsions within 15 min of drug injection. We analysed the modified intention-to-treat population that had received at least one injection of clonazepam and levetiracetam or placebo, excluding patients without valid consent and those randomised more than once. The trial is registered at EudraCT, number 2007-005782-35. FINDINGS: Between July 20, 2009, and Dec 15, 2012, 107 patients were randomly assigned to receive placebo and 96 were assigned to receive levetiracetam. The trial was discontinued on Dec 15, 2012 when interim analysis showed no evidence of a treatment difference, and 68 patients in each group were included in the modified intention-to-treat analysis. Convulsions stopped at 15 min of drug injection in 57 of 68 patients (84%) receiving clonazepam and placebo and in 50 of 68 patients (74%) receiving clonazepam and levetiracetam (percentage difference −10·3%, 95% CI–24·0 to 3·4). Three deaths, 19 of 47 (40%) serious adverse events, and 90 of 197 (46%) non-serious events were reported in the levetiracetam group, and four deaths, 28 of 47 (60%) serious events, and 107 of 197 (54%) non-serious events were reported in the placebo group. INTERPRETATION: The addition of levetiracetam to clonazepam treatment presented no advantage over clonazepam treatment alone in the control of GCSE before admission to hospital. Future prehospital trials could assess the efficacy of clonazepam alone as a first-line treatment in status epilepticus and the efficacy of a second injection of clonazepam with another antiepileptic drug as second-line treatment.
Commentary
One of the incontrovertible basic tenets in the management of status epilepticus (SE) is that early treatment leads to a more favorable outcome. This was the principle behind the Prehospital Treatment of Status Epilepticus (PHTSE) study in 2001, which compared the out-of-hospital treatment of SE with IV lorazepam (2 mg), diazepam (5 mg), or placebo (1). The study revealed superiority of lorazepam (59.1%) and diazepam (42.6%) over placebo (21.1%) for cessation of SE by arrival at the emergency department. The subsequent Rapid Anticonvulsant Medication Prior to Arrival Trial (RAMPART) study comparing the more easily administered midazolam (10 mg) vs lorazepam (4 mg) solidified these results, with 73.4% of the midazolam and 63.4% of lorazepam patients reaching cessation of SE at arrival to the emergency department (2).
Navarro et al., in their Service d'Aide Médicale Urgente (SAMUKeppra) study, addressed the natural next question as to whether AED polytherapy of out-of-hospital status would result in improved outcome (3). The investigators chose to study IV clonazepam (1 mg) vs simultaneous IV clonazepam in combination with IV levetiracetam (2500 mg). There is sound theoretical basis that such a treatment may be effective (4). Evidence that this might potentially be successful was present in case series, observational, and open-label studies. Levetiracetam certainly appeared safe—likely the safest of any AEDs they could have chosen. It certainly appeared to potentially act fast enough. The clinical trial design was modern, and the endpoints were of clinical interest and well-thought-out.
Nevertheless, the trial was terminated early due to futility. The co-administration of levetiracetam, albeit safe, conferred no benefit in the primary outcome measure of seizure termination at 15 minutes and no benefit of clinical significance to meaningful secondary outcome measures, such as avoidance of intubation or length of time in either the intensive care unit or in the hospital. In retrospect, a few unfavorable conditions can be identified. Early polytherapy has a long and thus far undistinguished history in the treatment of epilepsy in the outpatient setting (5). In the now-classic Veteran Affairs Cooperative Study, the one polytherapy arm of diazepam administered with phenytoin did not fare better than monotherapy lorazepam (6). Furthermore, in their power calculation, the authors assumed a baseline response of 50%. It is unclear why it was thought that the study would underperform RAMPART; the clinical trial pessimist/realist would have assumed that it would have at least comparable efficacy, and if aiming for a clinically relevant 15% response, the expected success rate would have been considerably more difficult to reach.
In some respects, they did reach that response rate—but for an unexpected reason. At 5 min, ~43% of patients had not reached seizure control in both arms, which is somewhat comparable to the RAMPART study. The second 1 mg of clonazepam that was given in those patients appears to have been critical, which boosted the response in both arms to much better than expected rates of 84% in the control arm and 74% in the levetiracetam arm. Although it would have been easier to compare this study to previous studies had they used lorazepam or avoided the second dose, this would have been difficult due to the standard of care at the study site.
What other conclusions can we draw from this trial? It vividly demonstrates limits of the conclusions that can be drawn from retrospective and even prospective observational studies. Randomized controlled trials still provide the searing honesty that helps change management practice. There may be concern regarding the inclusion of levetiracetam into the newly published status epilepticus treatment guidelines of the American Epilepsy Society (7) as well as the Neurocritical Care Society guideline (8) as a second-line medication without the benefit of such a trial. Many of the patients described in this study would have received levetiracetam once they reached the hospital. On the other hand, the results of this trial should not be extrapolated to its effectiveness (or lack thereof) as a second-line treatment in the in-hospital scenario; we can only conclude that levetiracetam in conjunction with IV clonazepam has no role in out-of-hospital treatment for early termination of status epilepticus. The upcoming Established Status Epilepticus Treatment Trial (ESETT) may provide the definitive answer to the question of second-line treatment.
This study also provides evidence that there may be efficacy to be gained simply by optimizing the dose of benzodiazepines given. Several observational studies (9, 10) suggest that adherence to protocol, particularly in the dosing of benzodiazepines, may lead to superior outcome in terms of seizure termination, fewer number of AEDs required, fewer patients transitioning into refractory status epilepticus, and decreasing intensive care unit stay. It should be noted that 40 to 50 percent of the patients in the PHTSE study received 2 injections of lorazepam or diazepam for recurrent or prolonged seizures. In addition, the study employed physician-led mobile medical emergency teams, a resource not widely available. The better-than-expected response rate leaves one to wonder whether these teams have unmeasured benefits that have led to superior outcomes. The study does provide impetus for all practitioners in the field to examine their out-of-hospital protocol to try to achieve the outcomes described in the study.
The authors suggest that another AED may have been more efficacious. Is another trial a reasonable undertaking? In its current design form and endpoint, it would be difficult to imagine that this would be performed if optimization of benzodiazepines truly does lead to an 84% control rate, and a 15% difference is considered clinically significant. Would this treatment be effective in complex partial status epilepticus, or another condition that is not readily remediable by benzodiazepines? Perhaps, though this type of trial may potentially necessitate the use of an out-of-hospital EEG, a technology that we can hope to see implemented in the future. A consequence of this trial design is that due to early termination, potentially interesting secondary endpoints remain incompletely explored. Can patients be “rescued” from refractory status epilepticus and the need for anesthetic medications by early co-administration of an AED? Seizure recurrence during the hospital stay was half as likely in patients receiving levetiracetam in this study, an endpoint that could well have reached statistical significance had the population been larger. More tantalizingly, there was a statistically significant decrease in patients with new neurological deficits in the levetiracetam cohort. Is co-administration of an AED neuroprotective? These are some questions that remain to be determined in future studies.
The investigators should be congratulated for performing this beautifully designed trial. Its adaptive design of a trial of this magnitude in the field of epilepsy is pioneering. It is hoped that such AED clinical trial designs can be utilized to establish efficacy and obtain FDA approval. Although the study failed to reach its planned endpoint, it is substantive for the clinical and scientific questions it has raised, for establishing a high target for out of hospital treatment of status epilepticus, and for its daring implementation of an innovative clinical trial design in a field where data is still scarce.