Managing refractory status epilepticus after third-line agent failure

32F with no prior seizure history, now in refractory status epilepticus for 4 hours. Failed lorazepam 0.1 mg/kg, fosphenytoin 20 mg PE/kg, and levetiracetam 60 mg/kg. EEG showing continuous generalized seizure activity.

Currently intubated for airway protection. Hemodynamically stable. CT head negative. CSF analysis pending.

My questions:

Propofol vs midazolam infusion as first-line continuous infusion?
Target on cEEG: burst-suppression vs seizure suppression?
Duration of continuous infusion before attempting wean?
Role of ketamine in refractory SE?
When do you consider immunologic workup (autoimmune encephalitis)?

MRI brain is scheduled for tomorrow but I need to stabilize seizures first.

For the immediate seizure management:

1. Continuous infusion choice: I start with midazolam (0.2 mg/kg bolus, then 0.1 to 2 mg/kg/hr infusion) rather than propofol as first-line. Reason: propofol infusion syndrome risk, especially in younger patients on prolonged infusions. The ESETT trial established benzodiazepine equivalence for initial treatment, but for continuous infusion, midazolam has a more predictable titration profile.

2. cEEG target: I target seizure suppression, not burst-suppression, as first-line. Burst-suppression requires deeper sedation with more hemodynamic compromise. If seizures recur despite adequate midazolam levels, then I escalate to burst-suppression.

3. Duration: I maintain the infusion for 24 to 48 hours after seizure cessation on cEEG, then wean slowly (reduce by 1 mg/hr every 2 to 4 hours). If seizures recur during wean, I restart at the effective dose and add a second maintenance AED before reattempting.

4. Ketamine: Excellent question. Ketamine (1 to 5 mg/kg/hr) is my second-line continuous infusion if midazolam fails. The NMDA receptor antagonism provides a mechanistically distinct approach. The KETASER-01 trial showed safety but was underpowered for efficacy.

5. Autoimmune workup: In any new-onset SE without clear structural or metabolic cause, I send NMDA receptor antibodies, LGI1, CASPR2, and a paraneoplastic panel on day 1. In a 32F, anti-NMDA receptor encephalitis should be at the top of the differential.

Update from the case: CSF came back with lymphocytic pleocytosis (WBC 45, 85% lymphocytes, protein 68). MRI showed bilateral mesial temporal FLAIR hyperintensity. NMDA receptor antibodies are pending but clinical picture is strongly suggestive of autoimmune encephalitis.

Started on IV methylprednisolone 1g daily and IVIG 0.4 g/kg/day. Seizure burden has reduced significantly on cEEG. Currently on midazolam infusion at 0.2 mg/kg/hr.

This case underscores the importance of early immunologic workup in new-onset refractory SE, especially in young women.

From the ICU perspective, I want to emphasize the hemodynamic management alongside seizure control. Patients on midazolam infusions at the doses needed for refractory SE (sometimes 1 to 2 mg/kg/hr) frequently develop hypotension requiring vasopressor support.

My practice:

Arterial line for continuous BP monitoring before starting the infusion
Norepinephrine ready to go
Target MAP > 65, but I aim higher (> 75) during active seizures to maintain cerebral perfusion pressure
Electrolyte panel q6h (magnesium, calcium, sodium) since hypomagnesemia is both a cause and consequence of prolonged seizures

Also, do not forget to check for nonconvulsive SE after the clinical seizures stop. I have seen multiple cases where clinical seizure activity resolved but cEEG showed ongoing electrographic seizures.

3 Answers