Clinical features, prognosis, and influencing factors of relapse in autoimmune encephalitis associated with antibodies against cell surface antigens: a single-center retrospective study.

Abstract

Objective: This study aimed to systematically analyze the clinical features, prognosis, and relapse of patients with autoimmune encephalitis associated with antibodies against cell surface antigens (AEASA), and to explore the independent risk factors for poor prognosis.

Methods: A retrospective study was conducted on 91 patients diagnosed with AEASA at our hospital between 2015 and 2023. Demographic data, clinical characteristics, and follow-up information were collected and analyzed. Patients were grouped based on antibody type, modified Rankin Scale (mRS) score, relapse status, and Clinical Assessment of Autoimmune Encephalitis (CASE) score. Statistical analyses included t-tests, analysis of variance (ANOVA), chi-square tests, Spearman correlation analysis, univariate and multivariate binary logistic regression, and receiver operating characteristic (ROC) curve analysis to evaluate the predictive efficacy of risk factors.

Results: Among the 91 patients, 58.3% were male, with a mean age of 47.9 ± 18.8 years. Clinical manifestations varied by antibody subtype: Anti-NMDAR encephalitis was more common in young females, with prominent neuropsychiatric symptoms, and significantly elevated CSF inflammatory markers (pressure, lymphocyte count) and systemic inflammatory indicators (CRP, CAR). Anti-LGI1 encephalitis was more prevalent in middle-aged and elderly males, often accompanied by hyponatremia, hypochloremia, and hippocampal MRI abnormalities. Anti-GABABR encephalitis had high rates of consciousness disturbance and cognitive impairment, with a significantly higher tumor comorbidity rate (37.5%) than other groups. Follow-up showed 71.4% of patients had a good prognosis, and 19.8% experienced relapse. Multivariate logistic regression analysis revealed that prolonged hospital stay (OR = 1.029, 95% CI: 1.008-1.066), elevated C-reactive protein (CRP; OR = 1.040, 95% CI: 1.032-1.058), and increased CRP/albumin ratio (CAR; OR = 1.032, 95% CI: 1.005-3.043) were independent risk factors for poor prognosis. ROC curve analysis confirmed these three factors had good predictive value for poor prognosis (AUC: 0.706, 0.697, 0.714, respectively).

Conclusions: AEASA subtypes differ significantly in demographic characteristics, clinical manifestations, laboratory results, and imaging findings, forming distinct disease spectra. Hospital stay duration, CRP, and CAR are robust, independent biomarkers for predicting poor prognosis in AEASA patients, providing important guidance for clinical risk stratification and treatment decisions. Early identification and intervention for high-risk patients with relapse potential are crucial for improving long-term prognosis.