Presented at AACR Annual Meeting 2025
Jeannette Fuchs1, Luke Piggott1, Kristian Urh2, Matjaž Žganec2, Eva Lavrencic2, Mark Uhlik2, Carolina Haefliger1
1 Debiopharm International S.A., Lausanne, Switzerland
2 Genialis Inc., Boston MA, United States
Summary
Debio 0123 is an investigational, orally available, highly selective, and brain-penetrant adenosine triphosphate (ATP)-competitive inhibitor of the WEE1 tyrosine kinase, currently in phase one clinical trials. Inhibition of WEE1 presents an opportunity as a therapeutic target in cancer therapy, either in cells relying on cell cycle checkpoints regulated by WEE1 or to potentiate DNA damaging agents.
In this study, we present a first-generation digital biomarker leveraging the ResponderID™ framework and Genialis™ Supermodel to predict response to Debio 0123. Using an Extra-Trees machine learning algorithm trained on DNA damage response (DDR)-related pathways and validated on patient-derived organoid RNA-seq datasets, our predictor initially achieved an accuracy of 0.76, later optimized to 0.82. Notably, the optimized model successfully predicted independent patient-derived xenograft responses, distinguishing responders from non-responders, even in cases where ex vivo and in vivo outcomes differed. The biomarker model comprises biomodules previously connected to WEE1 inhibition as well as less characterized pathways.
These findings highlight the potential of a machine learning-driven approach to refine patient selection for WEE1 inhibitor therapies, providing a strong foundation for further clinical validation of Debio 0123.