Accelerating discovery: Transformative clinical trial models in neuro-oncology

Abstract

Progress in neuro-oncology drug development has been slower than the pace of biological discovery, and trial design is increasingly recognized as a central contributor to this gap. Traditional trial paradigms, largely developed for systemic malignancies, face distinct challenges when applied to central nervous system (CNS) tumors - among them intratumoral heterogeneity, rapid clinical progression, limited tissue access, and the evolving molecular classification of gliomas. These biological and logistical realities have strained the assumptions underlying conventional phase II and III designs, and there is growing consensus that new approaches are needed. This review examines a framework of emerging trial models that better align clinical testing with the realities of CNS tumors. Master protocol designs enable simultaneous evaluation of multiple therapies within shared, molecularly informed infrastructures, supporting precision medicine approaches and improving efficiency. Bayesian adaptive frameworks allow trials to learn during conduct, reallocating patients toward promising therapies and incorporating emerging biological data in real time. Beyond efficacy assessment, trials can serve as active discovery platforms embedding longitudinal tissue sampling, window-of-opportunity designs, and multi-omic profiling to interrogate CNS drug penetrance, pharmacodynamic target engagement, and treatment-induced tumor evolution directly in human disease. Decentralized trial models and artificial intelligence offer additional tools to broaden access, improve enrollment, and reduce operational burden. Reimagining clinical trials as dynamic, biologically integrated, learning-based systems rather than static tests of individual agents is essential to accelerate therapeutic progress in neuro-oncology.