Patients with AE admitted at two tertiary care centers within 3 months from symptomatic onset were retrospectively identified using International Disease Classification (ICD) codes for encephalitis. Clinical and immunologic factors were assessed for the development of seizures (either clinical or electrographic) using Wilcoxon Rank Sum test and Fisher’s exact test. Variables that were found to be statistically significant were incorporated into a logistic regression predictive model for the development seizures. A survival model with Cox hazard ratios, using delay to achieving seizure-freedom from treatment initiation as the primary outcome, to assess for the effects of immunotherapy and antiseizure medication use. 

Of 131 patients with AE, 103 were admitted within 3 months from symptom onset and included. Seventy (70%) were antibody positive. Eighty patients (78%) had an inpatient EEG and were collectively recorded for 854 days (median=7 days; range=1-111). Sixty-two patients (60%) had a seizure during the study period: 22 had clinical-only seizures (35%), one electrographic-only seizures (2%), and 38 had both clinical and electrographic seizures (61%). Patients with negative or cell-surface antibodies were more likely to have seizures (odds ratio[OR]=1.61; p=0.02), as were patients with signal change in the temporal lobes on MRI (OR=2.58; p=0.04). Early antiseizure medication and immunotherapy treatment was associated with decreased duration of seizures (OR=1.05; p<0.01). 

Patients with negative or cell-surface antibodies and those with signal change in the temporal lobes on MRI were more likely to develop seizures. Combination treatment of immunotherapy with antiseizure medication achieved shorter delays to seizure freedom.