Objective:

To conduct an umbrella review synthesizing existing systematic reviews and meta-analyses characterizing the incidence, clinical phenotyping, time-to-onset, and mortality of neuro-irAEs across all approved ICI classes and solid tumor indications.

Background:

Immune checkpoint inhibitors (ICIs) including anti-PD-1, anti-PD-L1, and anti-CTLA-4 agents have revolutionized oncologic care but carry an emerging burden of neurological immune-related adverse events (neuro-irAEs). These range from mild peripheral neuropathies to fulminant autoimmune encephalitis and myasthenic crisis. Despite escalating ICI utilization across tumor types, the true epidemiological landscape of neuro-irAEs remains fragmented across heterogeneous case series and single-agent trials.

Design/Methods:

PubMed, Embase, Cochrane CENTRAL, and SCOPUS were searched through January 2025. Eligible studies included systematic reviews or meta-analyses reporting neurological adverse events in adult cancer patients receiving ICI monotherapy or combination regimens. Methodological quality was assessed using AMSTAR-2. Pooled incidence rates were extracted across phenotypic categories: encephalitis, Guillain-Barré syndrome (GBS), myasthenia gravis (MG), peripheral neuropathy, aseptic meningitis, and transverse myelitis. Overlap across included reviews was quantified using the corrected covered area (CCA) index.

Results:

Twenty-six systematic reviews encompassing data from 612 clinical trials and 94,700 ICI-treated patients were included. Overall pooled incidence of any-grade neuro-irAE was 3.8% (95% CI 2.9–4.9%), rising to 6.1% with combination PD-1/CTLA-4 blockade. MG and GBS carried the highest case-fatality rates at 11.4% and 8.7%, respectively. Median time-to-onset was significantly shorter with combination therapy (28 vs. 49 days, p<0.001). Anti-PD-1 agents demonstrated disproportionate encephalitis risk versus anti-PD-L1 (OR 1.74, 95% CI 1.21–2.50). Neurological irAEs preceded cancer progression detection in 18% of cases, suggesting potential immunological cross-reactivity mechanisms.

Conclusions:

 Neuro-irAEs represent a clinically underrecognized, heterogeneous, and potentially fatal complication of modern immunotherapy. The disproportionate risk with combination ICI regimens and the early temporal onset profile demand heightened neurological surveillance protocols integrated into oncology care pathways. Collaborative neuro-oncology frameworks are urgently needed to standardize grading, treatment escalation, and long-term follow-up.