Serostatus‐Specific Prognostic Profiles and Phenotype‐Dependent Therapeutic Windows in Autoimmune Encephalitis: A Multicenter Cohort Study

ABSTRACT Background While autoimmune encephalitis (AE) outcomes vary, serostatus‐specific prognostic determinants and institutional clustering effects remain poorly characterized. We aimed to delineate the distinct prognostic architecture of AE, identifying serostatus‐specific risk factors and potential therapeutic windows. Methods This multicenter, retrospective cohort study (2016–2024) included 498 patients (382 antibody‐positive [Ab+], 116 antibody‐negative [Ab−]) across 11 tertiary centers. The primary endpoint was 1‐year poor functional outcome (mRS > 2). We utilized Least Absolute Shrinkage and Selection Operator (LASSO) regression and generalized linear mixed models (GLMM) to adjust for institutional clustering, alongside phenotype‐treatment interaction analyses. Results Poor outcomes occurred in 12.9% of patients. GLMM analysis demonstrated negligible institutional heterogeneity. Prognostic architectures diverged by serostatus: Ab+ outcomes were dictated by failure to achieve early functional remission (discharge mRS: OR = 3.74; CASE: OR = 1.56; both p < 0.01) and age ≥ 65 years (OR = 5.78, p = 0.040). Conversely, Ab− outcomes were independent of discharge severity metrics, driven by baseline demographic and phenotypic features: age ≥ 65 and atypical presentation. Crucially, clinical phenotype modified the therapeutic response ( p interaction = 0.025). Delayed immunotherapy (≥ 30 days) significantly increased poor outcome risk in atypical presentations (OR = 9.16, p = 0.025) but not in canonical phenotypes. A derived prognostic nomogram yielded excellent discrimination (AUC = 0.909). Discussion Long‐term AE outcomes are dictated by patient‐level clinical severity and phenotype, with negligible influence from institutional variations. These findings advocate for a serostatus‐ and phenotype‐stratified management strategy. Crucially, atypical presentations exhibit a highly time‐dependent therapeutic window requiring rapid immunotherapy, while elderly patients warrant prioritized intervention to mitigate inherent baseline risks.