This 2026 white paper from Friends of Cancer Research, developed by a multidisciplinary working group including experts from Quibim, explores how AI-enabled tumor assessment tools could modernize treatment-response evaluation and clinical trial endpoints in solid tumors.

RECIST has standardized tumor response assessment for decades, but its reliance on manual, one-dimensional measurements of selected lesions can introduce variability and incompletely capture tumor burden and biological complexity. AI offers the possibility of detecting, segmenting, tracking and quantifying tumors across the body more consistently and comprehensively.

The white paper examines emerging approaches including AI-enhanced RECIST, radiomics, volumetric tumor burden and tumor growth kinetics, highlighting their potential to provide earlier and more sensitive measures of therapeutic response.

Crucially, it moves beyond technological potential to propose a stepwise roadmap for integrating and validating AI-based imaging endpoints in clinical trials, addressing context of use, analytical and clinical validation, endpoint definition, cross-tool reproducibility and regulatory requirements.

The ultimate goal is not simply to automate RECIST, but to establish more quantitative, reproducible and biologically informative imaging endpoints that could improve trial efficiency, support regulatory decision-making and accelerate oncology drug development.