Tumor-Specific Success Probabilities and Factors Associated With Phase III Trials in Oncology Drug Development

Despite advances in novel therapeutic modalities, such as molecularly targeted agents and immunotherapies, the probability of success in Phase III oncology trials remains low. This study quantitatively characterized success probabilities of Phase III trials in oncology drug development and evaluated the factors associated with trial success. Phase III interventional oncology drug trials registered at ClinicalTrials.gov between 2007 and 2023 with publicly available primary endpoint results were included. Trial success was defined as the achievement of at least one primary endpoint, and tumor-specific success probabilities were calculated. Multivariable logistic regression analyses were conducted to assess the association between trial success and key trial characteristics. We analyzed 824 trials (358 successful and 466 unsuccessful). Overall success probabilities were comparable between solid tumors and hematologic malignancies, although substantial heterogeneity was observed across solid tumors, with particularly low success probabilities for central nervous system tumors and pancreatic cancer. Multivariable analyses showed that biomarker-based patient selection, more recently initiated trials, line of therapy, and the number of primary endpoints were associated with trial success. Trials evaluating molecularly targeted therapies and those with short-term evaluable endpoints showed higher success probabilities, whereas trials evaluating chemotherapy or assessing overall or event-free survival endpoints showed lower success probabilities. Phase III trial success in oncology is associated with tumor-specific characteristics and development-stage factors, including biomarker-based patient selection and trial design. These results provide quantitative evidence to inform decision-making and trial design in oncology drug development.