Long-Term Outcomes in Antibody-Negative Autoimmune Encephalitis: A Systematic Review and Meta-Analysis

Seronegative autoimmune encephalitis (AE), characterized by the absence of known neuronal autoantibodies despite meeting clinical criteria, presents significant diagnostic and therapeutic challenges. Although immunotherapy is frequently used, long-term outcomes in this population remain poorly characterized. This systematic review and meta-analysis aims to synthesize evidence on long-term outcomes in seronegative AE, focusing on functional status, seizure burden, cognitive impairment, relapses, immunotherapy practices, and prognostic indicators. We included studies involving patients 12 years or older diagnosed with antibody-negative probable AE (ANPRA), possible AE, or definite limbic encephalitis (LE) with negative neuronal antibodies according to Graus et al. (2016), with a minimum median follow-up of 6 months.

Nineteen studies, predominantly retrospective cohorts with moderate risk of bias, met the inclusion criteria. The pooled proportion of patients achieving favorable functional outcomes (modified Rankin Scale 0-2) was 65% (95% CI 51% - 78%; I² = 88.1%; 12 studies, n = 782). Outcomes differed by seronegative subtype, with better functional recovery in seronegative LE (73%; 95% CI 45%-90%; 6 studies, n = 95) compared with ANPRA (47%; 95% CI 38%-55%; 5 studies, n = 277; p = 0.005). When compared with seropositive AE (n = 466), seronegative patients (n = 373) demonstrated comparable odds of favorable functional outcomes (odds ratio 0.50; 95% CI 0.19-1.31; p = 0.16) but had a significantly higher risk of persistent seizures (risk ratio 1.66; 95% CI 1.01-2.72; p = 0.04). The pooled relapse proportion among seronegative patients was 17% (95% CI 11%-26%; 7 studies, n = 350), with no significant difference by serostatus. Residual disability assessed using the Clinical Assessment Scale in Autoimmune Encephalitis suggested mild persistent impairment (mean score 3.11; 95% CI 1.79-4.43; 5 studies, n = 409). Owing to heterogeneity and inconsistent reporting, meta-analysis of cognitive outcomes, immunotherapy strategies, and RAPID (Refractory status epilepticus, Age ≥ 60 years, Probable AE, Infratentorial involvement, Delayed immunotherapy ≥ 1 month) scores was not feasible; narrative synthesis suggested persistent cognitive deficits and lower utilization of second-line immunotherapy in seronegative cohorts.

Seronegative AE is associated with heterogeneous outcomes, with antibody-negative LE demonstrating better recovery than ANPRA. Subtype-specific variability and elevated seizure burden highlight the need for individualized management, timely diagnosis, structured follow-up, and well-designed prospective research.