To the Editor: Ependymomas account for approximately 8%–10% of pediatric central nervous system tumors, and anaplastic variants are characterized by aggressive behavior and high recurrence rates. Despite multimodal treatment strategies, outcomes for relapsed or refractory disease remain poor, prompting investigation of novel therapeutic approaches, including immune checkpoint inhibitors 1, 2. Nivolumab, a programmed cell death-1 (PD-1) inhibitor, enhances T-cell-mediated antitumor immune responses and has shown potential activity in selected pediatric CNS tumors, particularly those expressing programmed death-ligand 1 (PD-L1) 1, 3. However, immune checkpoint inhibition may disrupt immune tolerance and lead to immune-related adverse events (irAEs). While neurological irAEs are rare, they can be severe and potentially life-threatening, with limited data available in pediatric patients 4-7. We report a 15-year-old boy with multiply relapsed anaplastic ependymoma who developed anti–glutamic acid decarboxylase (GAD)–positive autoimmune encephalopathy following nivolumab therapy. The patient was initially diagnosed 7 years earlier and received surgery, radiotherapy, and multi-agent chemotherapy. After multiple relapses treated with craniospinal irradiation, combination chemotherapy, bevacizumab-based therapy, and sirolimus, he underwent reoperation for progressive disease in January 2025. Tumor tissue demonstrated PD-L1 expression, and nivolumab was initiated following off-label consent. After three doses of nivolumab, the patient developed hallucinations, dysarthria, and gait instability. Cerebrospinal fluid analysis obtained via thoracic puncture revealed negative infectious studies, while autoimmune encephalitis panel testing demonstrated anti-GAD positivity. Electroencephalography and magnetic resonance imaging findings were consistent with encephalopathy. The clinical presentation was interpreted as nivolumab-associated autoimmune encephalopathy. High-dose corticosteroids and intravenous immunoglobulin failed to achieve sufficient clinical improvement. Subsequent plasmapheresis resulted in marked neurological recovery. Steroid therapy was continued with gradual tapering. At the last follow-up, the patient remained neurologically intact and was receiving oral etoposide with stable disease. Immune checkpoint inhibitor–associated encephalitis is a rare but serious complication with heterogeneous clinical features. Autoantibody positivity, including anti-GAD antibodies, has been described in a subset of cases, although cell-mediated immune mechanisms are also thought to play a significant role 4-7. Current management recommendations are largely extrapolated from adult experience, with corticosteroids as first-line therapy and intravenous immunoglobulin or plasmapheresis reserved for refractory cases. In our patient, neurological improvement was achieved only after plasmapheresis, highlighting its potential role in steroid-refractory pediatric cases. In conclusion, this case underscores the need for heightened awareness of immune-related neurological toxicity in children receiving immune checkpoint inhibitors. While immunotherapy may represent a promising option for refractory pediatric ependymoma, careful patient selection, close neurological monitoring, and prompt immunosuppressive intervention are essential to optimize outcomes. The authors declare no conflicts of interest.
Bulut et al. (Tue,) studied this question.