Machine learning models improve prediction of immunotherapy outcomes across multiple cancer types

Dr Alyssa Pybus speaks to ecancer about machine learning based survival models to predict outcomes following immune checkpoint blockade therapy in patients with advanced melanoma, non-small cell lung cancer, and renal cell carcinoma.

Using over 2,000 patients and routinely collected clinical and laboratory features, the models identified key predictors of survival and disease progression, including blood-based biomarkers and performance status indicators.

The resulting algorithms demonstrated strong predictive accuracy across cancer types, outperforming established biomarkers such as PD-L1 expression and tumour mutational burden.

These findings highlight the potential of data-driven approaches using standard clinical information to improve patient stratification, support treatment decision-making, and better identify individuals most likely to benefit from immunotherapy.