Predictors of Cognitive Recovery in Paediatric Autoimmune Encephalitis

Abstract

Paediatric autoimmune encephalitis (AE) is an inflammatory brain disease associated with acute cognitive and behavioural difficulties, seizures, magnetic resonance imaging (MRI) and electroencephalography abnormalities. Some children with AE experience difficulties years after acute illness, but identifying those at high risk is difficult; factors able to predict them are needed. Advanced neuroimaging methods including magnetoencephalography (MEG) and quantitative measures of structural magnetic resonance imaging (MRI; for example cortical thickness) are promising tools with which to predict neurobehavioural outcome. This thesis developed and investigated the ability of these techniques to predict cognitive and behavioural outcome in children with AE. Children with autoimmune encephalitis (including anti-NMDA and ADEM) were recruited at least 18 months after initial presentation; along with typically developing children. Participants underwent MRI scans; MEG recordings at rest and during an auditory oddball task; and completed neuropsychological assessments. Through a series of experiments, long-term psychological outcomes were examined, and brain structure and function interrogated. Overall, the thesis found that behavioural assessments highlighted long-term difficulties in children with AE. MRI analyses showed brain cortical thinning in the left superior occipital and parietal gyri, and orbitofrontal cortex. Functional network analyses highlighted alterations within the delta frequency, with lower efficiency in information transmission within local connections. However, network measures, cortical thickness and auditory evoked responses did not predict neurobehavioural outcomes. The present findings show that these neuroimaging approaches are applicable in paediatric AE and enhance our understanding of long-term alterations, however, it is not established whether they can predict long-term difficulties. The findings highlight opportunities to further develop neuroimaging approaches that can potentially uncover clinically useful findings, and will be relevant for approaches that will better identify children with AE who warrant ongoing surveillance and early intervention to ameliorate long-term consequences of early life brain inflammation.

Date of Award	Mar 2023
Original language	English
Supervisor	Sukhvir Wright (Supervisor), Gavin L Woodhall (Supervisor), Elaine Foley (Supervisor), Evangeline Wassmer (Supervisor), Amanda Wood (Supervisor) & Klaus Kessler (Supervisor)