Metabolic Signatures and Diagnostic Strategies for Different Types of Encephalitis

To characterize metabolic signatures associated with different encephalitis types and to develop differential diagnostic models based on metabolic dysregulation profiling.

Untargeted metabolomics was performed on cerebrospinal fluid (CSF) from a prospective cohort of 232 patients classified into viral encephalitis (VE, n = 82), autoimmune encephalitis (AE, n = 63), other infectious encephalitis (OIE, n = 25), and non-encephalitis control (NE, n = 62) groups, with follow-up samples from 10 participants.

Compared with controls, taurine-associated metabolites were consistently downregulated across the encephalitis groups. Arginine and proline metabolism were identified as commonly dysregulated pathways in the encephalitis groups. Highly similar metabolic network topologies were observed between VE and AE, despite different key node metabolites (lysophosphatidylcholine vs. acylcarnitine, respectively). Three diagnostic models developed to distinguish VE from other groups achieved an area under the receiver operating characteristic curve > 0.8 in the training and validation sets. Glial fibrillary acidic protein antibody-positive AE exhibited a unique carnitine metabolism signature. Longitudinal follow-up identified seven metabolites significantly associated with clinical remission in patients with AE.

This study systematically maps shared and distinct metabolic perturbations in different types of encephalitis and suggests potential metabolic features associated with AE subtype-specific pathophysiology, providing a preliminary metabolomic basis for precision diagnosis.