The right patient, the right moment, the right therapy: reimagining precision oncology from the bedside.

Precision oncology has delivered an unprecedented expansion of molecularly targeted therapies, yet patient selection remains the defining unmet need of the field. Too many patients receive treatments that will not benefit them, while those most likely to respond go unidentified. Drawing on two decades of translational work in treatment-refractory triple-negative breast cancer (TNBC) and inflammatory breast cancer (IBC), this Perspective argues that the central question of oncology's next decade is not what new drug to develop, but which patient should receive which therapy, at which moment, and why. We articulate a biology-first framework for precision patient selection built on multi-omic tumor profiling, longitudinal liquid biopsy with circulating tumor DNA (ctDNA) monitoring, tumor microenvironment (TME) characterization, and disease-specific criteria for rare subtypes that do not follow conventional patterns of response and progression. The poor response of metastatic hormone receptor-positive IBC to CDK4/6 inhibition illustrates why selection criteria validated in common populations can mislead when applied to biologically distinct diseases. From these lessons, we propose five principles for next-generation trial design: multi-omic entry criteria, longitudinal biomarker integration as adaptive decision points, microenvironment-aware stratification, disease-specific frameworks for rare subtypes, and preemptive resistance monitoring. Equitable implementation is framed not as an aspiration but as a structural requirement, since a selection framework that cannot reach underrepresented patients has failed before deployment. Realizing true precision oncology requires treating eligibility as a dynamic biological state, reassessed throughout treatment, rather than a snapshot fixed at enrollment.