Immune Oncology Research Institute shared on LinkedIn:
“What if we could know who is most likely to benefit from immunotherapy – before treatment even begins?
In a new Nature Medicine study, Wanxiang Shen and colleagues introduce COMPASS, a pan-cancer AI foundation model designed to predict response to immune checkpoint inhibitors using pretreatment tumor transcriptomes.
Trained on 10,184 tumors across 33 cancer types and validated across 16 independent clinical cohorts, COMPASS outperformed 22 existing prediction methods.
But what makes this work particularly compelling is that the model does more than predict outcomes – it also helps explain the biology behind them, connecting response and resistance to mechanisms such as IFN-γ signaling, cytotoxic T-cell activity, TGF-β signaling and immune exclusion.
This points toward a potentially important shift in precision immuno-oncology: moving beyond isolated biomarkers toward AI systems that can interpret the complex interaction between tumors and the immune system.
From better patient selection to understanding resistance and identifying new therapeutic strategies, the implications are substantial. If you work in immuno-oncology, precision medicine or AI-driven cancer research, this is a study worth reading in full.”
Title: Generalizable AI predicts immunotherapy outcomes across cancers and treatments
Authors: Wanxiang Shen, Intae Moon, Thinh H. Nguyen, Michelle M. Li, Yepeng Huang, Nitya Nair, Daniel Marbach, Marinka Zitnik
Other articles about immuno-oncology on OncoDaily.