Face/off: phase-specific modeling of lineage plasticity using near-patient models in genitourinary cancers

Lineage plasticity has emerged as a fundamental mechanism of adaptive therapy resistance in prostate and bladder cancers, enabling malignant cells to bypass lineage-restricted dependencies through non-genetic reprogramming. However, plasticity is often studied as a terminal phenotype rather than as a dynamic process that unfolds over time. The endpoint-centered view can obscure the early and potentially reversible events that initiate identity loss, as well as the intermediate states through which resistant phenotypes emerge. In this review, we propose a phse-based framework for modeling lineage plasticity in genitourinary cancers. We conceptualize plasticity as a temporally ordered trajectory comprising three phases: priming, transition, and stabilization. We examine how commonly used near-patient experimental and computational platforms, including patient-derived organoids (PDOs), patient-derived xenografts (PDXs), genetically engineered mouse models (GEMMs), and artificial intelligence-based approaches, tend to sample different portions of this trajectory. By mapping these model systems to the phases, they are best equipped to simulate, we demonstrate how a phase-aware approach can resolve long-standing discrepancies in the literature and clarify the gene-regulatory logic underlying tumor identity shifts. Ultimately, this framework provides a roadmap for integrating cross-platform insights to identify critical "windows of vulnerability," informing novel strategies for the early therapeutic interception of lineage-driven resistance before it reaches a terminal, irreversible state.