Background& Objective: Investigation modalities, such as MRI and CSF examination, are neither sensitive nor specific in the early phase of anti-NMDAR encephalitis. Nuclear imaging may be useful to monitor the response to treatment but limited by the availability.We aimed to determine the role of EEG as a tool for early diagnosis as well as a tool to assess disease progression and response to treatment. Methods: A total of 99 EEGsdone in 16 patients diagnosed with anti-NMDAR encephalitis throughout the course of illness, were reviewed retrospectively. The EEG changes were correlated with the clinical presentations and response to treatment. Sixteen EEGs of patients with schizophrenia and mood disorder, and 10 EEGs of patients with infective encephalitis were included as control. Results: EEGs performed during the psychiatric and cognitive dysfunctionphase in patient with anti-NMDAR encephalitis, showed diffuse background slowing in the delta-theta range in all the patients. Serial EEGs showed that the dominant background frequency improved with improvement in cognitive status. Nine patients had complete recovery with normalisation of the EEG abnormalities. Eight patients had their typical clinical seizure recorded during EEG monitoring, but only 2 (25.0%) with EEG correlation. Ten patients had status epilepticus (62.5%), 5 had EEG recorded during their status epilepticus, of which only one with EEG correlation (20.0%). Eleven patients had asymmetric background (68.8%), but only 1 has correlation with focal changes in the MRI brain (9.1%). Even though the EEGs of patients with infective encephalitis also showed background slowing, their CSF analysis was supportive of an infective cause. EEGs of patients with established psychiatric disorder were within normal limits.
Conclusion: EEG abnormality has a good correlation with the degree of psychiatric and cognitive dysfunction in patient with anti-NMDAR encephalitis, and is useful in early diagnosis, monitoring the progress and the response to treatment. However, it has poor correlation with clinical seizures.