Anti-NMDA-receptor encephalitis
ICD-10 G04.8 · ICD-11 8E4A.0.3

What to Do When Second-Line Immunotherapy Fails in Children Under 18 with Severe Anti-NMDA-receptor Encephalitis

This protocol applies to children (age <18 years) with NMDA receptor antibody encephalitis classified as severe disease who have not achieved adequate improvement following a first course of second-line immunotherapy given approximately 1–3 months earlier.

Clinical Scenario — Severe Disease in Children Under 18

Severe disease is defined by issues affecting safety and function. Classification requires at least one severity marker to be present: intensive care need, airway support, dysautonomia threatening safety, being bed-bound, suicidal thoughts, or an adapted mRS score of 4–5.

Any single qualifying marker is sufficient to place the patient in the severe category, which determines both initial treatment intensity and the threshold for escalation when response is inadequate.

Previous Treatment — and Why Escalation is Needed

For severe disease, second-line immunotherapy — rituximab or cyclophosphamide — is offered approximately 2 weeks after first-line initiation. Adequate clinical improvement is reassessed at approximately 1–3 months following that start.

This protocol is indicated when that 1–3 month reassessment shows the first second-line agent did not produce adequate improvement. The failure to meet that improvement target is what triggers this escalation step.

Next Treatment Step (Overview Only)

When the first second-line agent has not produced an adequate response, the approach involves switching to an alternative second-line immunotherapy. In the most refractory cases — those who still fail to improve after this switch — further escalation to an additional targeted therapy may be considered. The full protocol details which agents apply in which order and under what conditions.

Goal: adequate improvement assessed approximately 1–3 months after the next treatment step.
Instant Access to Structured Evidence-Based Regimens

References

DOI: 10.1212/NXI.0000000000001052

View source ↗