Current clinical trials are based on rigid designs and drug-centric approaches that can stifle flexibility and innovation. With advances in molecular biology and technology, there is an urgent call to revitalize trial designs to meet these evolving demands. We propose a reshaped, prismatic vision of clinical trials combining different knowledge layers, synergized with modern computational approaches. This paradigm based on iterative learning will enable a more adaptive and precise framework for oncology drug development.