Which anticonvulsant best terminates benzodiazepine-refractory status epilepticus?

Takeaway

• Levetiracetam, fosphenytoin, and valproate were similarly efficacious and safe for stopping seizures and improving level of consciousness in patients with benzodiazepine-refractory convulsive status epilepticus.

Why this matters

• Early seizure termination reduces complications, mortality.

Key results

• Trial stopped early for futility.
• Main efficacy outcome:
  • 47% with levetiracetam (95% credible interval [CrI], 39%-55%),
  • 45% with fosphenytoin (95% CrI, 36%-54%), and
  • 46% with valproate (95% CrI, 38%-55%).
• Posterior probability that drug was the most effective:
  • 0.41 for levetiracetam,
  • 0.24 for fosphenytoin, and
  • 0.35 for valproate.
• Groups also similar on:
  • ICU admission;
  • Median lengths of ICU, hospital stay; and
  • Median time to seizure termination when treatment was successful.
• Numeric but nonsignificant differences:
  • More episodes of hypotension, intubation with fosphenytoin.
  • More deaths with levetiracetam.

Expert comment

• In an editorial, Phil E. Smith, MD, writes, “Having three equally effective second-line intravenous medications means that the clinician may choose a drug that takes into account individual situations that may be modified by factors such as the presumed underlying cause of status epilepticus; coexisting conditions, including allergy, liver and renal disease, hypotension, propensity to cardiac arrhythmia, and alcohol and drug dependence; the currently prescribed antiepileptic treatment; the cost of the medication; and governmental agency drug approval.”

Study design

• Randomized adaptive trial comparing 3 drugs in 384 children and adults with convulsive status epilepticus unresponsive to treatment with benzodiazepines (ESETT trial).
• Main outcome: absence of clinically evident seizures and improvement in level of consciousness 60 minutes after start of drug infusion, without additional anticonvulsant medication.
• Funding: National Institute of Neurological Disorders and Stroke.

Limitations

• Need for unblinding in some cases.
• Reliance on clinical vs electroencephalographic criteria for main outcome.
• Only single dose of each drug tested.
• Late nonserious adverse events not collected.