Clinical trials are essential in the process of new drug development. As clinical trials involve significant investments of time and money, it is crucial for trial designers to carefully investigate trial settings prior to designing the trial. Utilizing trial documents from ClinicalTrials.gov, we aim at understanding common characteristics of successful and unsuccessful cancer drug trials to provide insights about what to learn and what to avoid. In this research, we first computationally classified cancer drug trials into successful and unsuccessful cases and then utilized natural language processing to extract information of eligibility criteria from the trial documents. Contrast mining was applied to discover highly contrasted patterns with significant difference in prevalence between successful and unsuccessful groups. Our method identified contrast patterns consisting of combinations of drug categories, eligibility criteria, study organization, and study design for nine major cancers. In addition to literature review as the qualitative validation of mined contrast patterns, we also found classifiers based on quality contrast patterns achieved 80% of average F1 score for most cancer types. In summary, our study demonstrates that contrast mining is a useful approach to inform the decision-making process for trial investigators and therefore facilitate better design of cancer drug trials.
For Zoom information, please contact Robert Sanders (email@example.com).