Systematic Review and Meta-Analysis of the Clinical Features Associated With Seronegative Autoimmune Encephalitis

Background and objectives: Seronegative autoimmune encephalitis (AE), defined by an appropriate clinical phenotype in the absence of known neuronal autoantibodies, poses diagnostic and therapeutic challenges due to clinical heterogeneity and lack of definitive biomarkers. We conducted a systematic review and meta-analysis of individual patient data to characterize the phenotypes, treatment responses, and prognostic factors in seronegative AE. Methods: We included 213 cases from 30 studies published between 2014 and 2024 and 11 from a local cohort meeting Graus criteria for seronegative AE. We extracted details on clinical and paraclinical features, immunotherapy, and outcomes measured via the modified Rankin Scale (mRS) and Clinical Assessment Scale for Autoimmune Encephalitis (CASE). Multivariate regression and dimensionality reduction analyses identified prognostic markers. Results: Of 224 patients (median age 49 years, 50.9% male), 72 (32.1%) had limbic encephalitis (LE) and 152 (67.9%) had antibody-negative probable AE (ANPRA). Good outcome (mRS score ≤2) was more common in LE (49/72, 68.1%) than in ANPRA (76/154, 50.0%) (p < 0.05). Delayed immunotherapy was associated with an increased risk of poor outcome. Additional predictors of poor prognosis included age older than 60 years, the ANPRA subtype, an underlying tumor, striatocapsular or thalamic involvement on MRI, and presentation with refractory status epilepticus. Dimensionality and clustering analysis identified heterogeneity among seronegative AE, with 3 distinct subtypes. Discussion: Seronegative AE comprises clinically and prognostically distinct subtypes. Early immunotherapy is the key modifiable factor influencing outcome. We advocate for biomarker discovery and prospective, systematically reported cohort studies to improve stratification and treatment strategies in this diagnostically challenging population.