Objective. Electroencephalogram (EEG) is a sensitive method for evaluation of anti-N-methyl-d-aspartate receptor (anti-NMDAR) encephalitis. The aim was to investigate the EEG and clinical features and correlations in patients with anti-NMDAR encephalitis. Methods. Anti-NMDAR encephalitis patients (n = 98) with at least 1 EEG recording in our hospital were recruited between January 2012 and October 2017. We reviewed and analyzed the EEG and clinical data. Results. In our cohorts, 79 patients underwent EEG in the disease acute stage, and 39 in the recovery stage. Of the 79 patients, 70 (88.6%) EEG recordings in acute period were abnormal. Symptoms, including consciousness, movement disorder, coma, were correlated to the degree of EEG abnormalities (P < .05). The patients with more severe EEG abnormalities also had longer hospitalized and intensive care unit stay time (P < .05). We found that the EEG pattern of abnormal occipital alpha rhythm had a correlation with the clinical severity, and the Spearman coefficient was 0.448 (P = .000). Neither delta activities distribution nor prevalence showed correlations with clinical severity in acute stage. However, delta activities significantly decreased in the disease recovery stage. The other findings of EEG records were extreme delta brush (7cases, 8.9%), excess delta activities (diffuse slowing 30 cases, 38.0%), and epileptiform discharge (10 cases, 14.3%). Conclusion. This is the largest study of EEG recording in patients with anti-NMDAR encephalitis patients in China. EEG abnormalities, particularly occipital alpha rhythm, are correlated with clinical severity. EEG is useful for diagnosis and monitoring of treatment response in conjunction with clinical improvement.
Keywords: EEG; anti-NMDAR encephalitis; delta activities; extreme delta brush; occipital alpha rhythm.