To compare the treatment response and overall prognosis of pure iatrogenic versus paraneoplastic immune checkpoint inhibitor (ICI)-induced encephalitis.

ICIs offer favorable outcomes over conventional cytotoxic therapies but can also cause neurologic immune-related adverse events (irAE-Ns), which either arise through de novo induction of an anti-neural immune response or unmasking of a latent paraneoplastic process. As existing literature fails to differentiate these two etiologies despite their distinct prognoses, we compared neurologic response to treatment between pure iatrogenic and iatrogenic/paraneoplastic ICI-induced encephalitis.

Inclusion criteria included ICI encephalitis-related conditions. Cases were categorized using the 2021 Paraneoplastic Neurologic Syndrome (PNS) Care Score by clinical phenotype, antibody type, high-risk cancer presence, and follow-up time. Definite/probable PNS scores indicated iatrogenic/paraneoplastic cases; possible/non-PNS scores indicated iatrogenic cases. Neurologic response was defined as complete, partial, and no response/worsening. These were compared using a chi-squared test to determine significance and dichotomized into favorable (complete or partial) and unfavorable neurologic responses (no response or worsening).

We conducted a meta-analysis of 185 articles documenting ICI-encephalitis and encephalitis-related cases (2011–May 2025; PubMed/Web of Science). The cohort included 309 patients (168 males, 127 females; median age 65.0 years). 122 patients had neuronal autoantibodies (most commonly anti-Hu and anti-Ma2). Applying the PNS score revealed 169 patients (55%) had probable or definite paraneoplastic origin triggered by ICIs (iatrogenic/paraneoplastic) while 140 patients (45%) were not paraneoplastic (pure iatrogenic encephalitis induced by ICIs). Patients primarily received anti-PD1 (47%), anti-PDL1 (29%), combination therapy (20%), or anti-CTLA4 (4%). Pure iatrogenic cases showed significantly more favorable neurologic responses than iatrogenic/paraneoplastic cases (86% vs 62%, p=0.00003).

Accurate diagnosis of PNS is essential—ICIs induce two distinctive patterns of encephalitis: pure iatrogenic with a favorable prognosis or paraneoplastic with a guarded prognosis. Applying the PNS score to patients with ICI-induced irAE-Ns is crucial for proper prognostication, informing clinical decision-making, and optimizing patient management.