The rapid differentiation between autoimmune and infectious encephalitis in children is a critical clinical decision that dramatically impacts treatment and outcome. Metagenomic next-generation sequencing (mNGS) of cerebrospinal fluid (CSF) is a powerful but often underutilized tool, as its host-derived RNA component is typically discarded. We hypothesized that this host response data could be translated into a diagnostic tool for autoimmune encephalitis (AE). We enrolled 180 pediatric patients with suspected encephalitis to evaluate the clinical performance of CSF mNGS against conventional methods. Host transcriptomic analysis was performed on CSF cells from 88 patients (autoimmune, bacterial, and viral encephalitis). A novel biomarker was validated using RT-qPCR in an independent cohort, and its functional role was investigated in neuronal cultures challenged with NMDAR1 antibodies. A diagnostic model was developed and validated. mNGS demonstrated a significantly higher pathogen detection rate than conventional methods (29.4% vs. 16.7%). Host transcriptomic profiling revealed that AE shared a hyperinflammatory signature with viral encephalitis but was uniquely associated with dysregulation of receptor tyrosine kinase and heme signaling pathways. Furthermore, memory B cells and activated mast cells were specifically elevated in AE. We identified and validated RAD54B as a novel biomarker specifically upregulated in AE. Functionally, RAD54B upregulation protected neurons from DNA damage stress induced by NMDAR1 antibodies. A multi-gene diagnostic model based on host-response genes robustly differentiated AE from infectious encephalitis (AUC > 0.923) in a validation set. We present a validated translational pipeline that repurposes routine CSF mNGS data into a dual-purpose diagnostic tool. By leveraging the host RNA data inherent in CSF mNGS, clinicians can now simultaneously investigate infectious and autoimmune etiologies in a single, rapid test. This strategy has the immediate potential to reduce diagnostic delay, guide timely therapy, and improve outcomes in children with encephalitis.
Han et al. (Wed,) studied this question.