236P Development and external validation of a predictive model for early severe immune-related adverse events in immune checkpoint inhibitor recipients, with deployment as a web-based calculator

Severe (grade ≥3) immune-related adverse events (irAEs) occur in ∼10–20% of patients receiving immune checkpoint inhibitors (ICIs), often within 90 days of treatment. These toxicities can cause treatment interruption, hospitalization, and long-term morbidity. Early identification of high-risk patients could enable tailored monitoring and proactive management. No broadly validated, generalizable prediction tool currently integrates clinical, laboratory, and immune data into a user-friendly format for oncology practice.