Human-relevant bone marrow-on-chip and AI-guided NAMs to predict and prevent resistance to targeted therapies in hematologic malignancies

Many hematologic malignancies are defined by specific driver mutations, and targeted therapies have been developed to inhibit these molecular abnormalities. However, over time, malignant cells often develop resistance to targeted agents. Due to their high proliferative rate and genomic instability, these malignancies acquire additional mutations that contribute to therapeutic resistance. An important question is which molecular pathways could be targeted to reduce the emergence of new mutations and delay or prevent resistance. This project will develop human-relevant bone marrow-on-chip and 3D organoid models of hematological malignancies, integrating patient-derived cells and AI-guided high-content analysis to predict targeted therapy response and resistance, enabling translational preclinical testing and regulatory adoption while reducing reliance on animal models. Therapy-induced mutagenesis and resistance in hematologic malignancies are driven by replication stress, DNA repair dysregulation, and microenvironment-mediated survival signaling. Targeting these pathways—identified through human NAM platforms—can reduce mutation emergence and delay resistance.