ASCO 2025 | Novel approach to biomarker development in kidney cancer

So I think there’s a couple of key points as background. One is that kidney cancer really is distinct biologically from other tumor types. So we can’t just establish a biomarker in lung cancer or melanoma and assume it’s going to work in kidney. We need something that’s really kidney cancer specific. The other is the way that biomarker development has typically been done, including by our own group, has been to look at one thing at a time. You look at a mutation, or that’s it, or you look at the T-cell phenotype and that’s it. But we know these are really complex systems. There’s lots of things within the tumor that matter. And there’s a lot of things within the peripheral blood, circulating factors that matter as well. And that’s really what we set out to do in this study. We use both tumor material, but also peripheral blood. We use both conventional ways of analyzing it and some pretty novel ways as well. But then critically, we use the machine learning-based approach to try to really integrate that large-scale data and come up with an integrated biomarker model to predict response or resistance.