Symptomatic treatment of children with anti-NMDAR encephalitis

Aim We performed the first study on the perceived benefit and adverse effects of symptomatic management in children with anti-N-methyl-d-aspartate receptor (NMDAR) encephalitis. Method A retrospective chart review was undertaken at two tertiary paediatric hospitals in Australia and New Zealand. We included 27 children (12 males, 15 females; mean age at admission 7y 1mo) with anti-NMDAR antibodies in serum or cerebrospinal fluid with a typical clinical syndrome. Results Only two out of 27 patients were white, whereas 16 out of 27 patients were from the Pacific Islands/New Zealand Maori. The mean duration of admission was 69 days (10–224d) and 48% of patients (13/27) needed treatment in an intensive care setting. A mean of eight medications per patient was used for symptomatic management. Symptoms treated were agitation (n=25), seizures (n=24), movement disorders (n=23), sleep disruption (n=17), psychiatric symptoms (n=10), and dysautonomia (n=four). The medications used included five different benzodiazepines (n=25), seven anticonvulsants (n=25), eight sedatives and sleep medications (n=23), five antipsychotics (n=12), and five medications for movement disorders (n=10). Sedative and sleep medications other than benzodiazepines were the most effective, with a mean benefit of 67.4% per medication and a mean adverse effect-benefit ratio of 0.04 per medication. Antipsychotic drugs were used for a short duration (median 9d), and had the poorest mean benefit per medication of 35.4% and an adverse effect-benefit ratio of 2.0 per medication. Interpretation Long-acting benzodiazepines, anticonvulsants, and clonidine can treat multiple symptoms. Patients with anti-NMDAR encephalitis appear vulnerable to antipsychotic-related adverse effects. Pacific Islanders appear to have a vulnerability to anti-NMDAR encephalitis in our region.