The dice of clinical trials in oncology: Strategies from mathematical permutation, combination, and theoretical oncology.

Abstract

e23004 Background: With the integration of artificial intelligence and machine learning into clinical trials in oncology, the development of the oncology field has entered a transition period from precision medicine to artificial intelligence medicine. To better apply artificial intelligence and big data to clinical trials in oncology, we developed a clinical research support system called the Dice of Clinical Trials. Combining the principles of mathematical permutation and combination in the selection of drugs and treatment models via theoretical oncology, the Dice of Clinical Trials provided the basis for alternative intervention plans in clinical trials in oncology. Methods: On the basis of the current anticancer treatment options, modalities, types, prognostic indicators, tumor types and principles of theoretical oncology, permutations and combinations were conducted via the Dice of Clinical Trials. According to the results of published research and clinical trials, information such as the mutual influence and incompatibility of drugs in absorption, distribution, metabolism and excretion can be entered into the system. Through this system, clinical trial investigators can prescreen the intervention protocol, predict its efficacy and adverse reactions, and select the most appropriate control and experimental groups according to the research purpose. Additionally, the system can evaluate and analyze schemes according to information from ClinicalTrials.gov and ChiCTR to provide real-time information concerning the registration of new clinical trials in oncology. Results: According to the permutations and combinations of treatment types and modes using theoretical oncology rules, the Dice of Clinical Trials can provide 4374 different combinations of anticancer therapy regimens in the database. After preliminary screening and exclusion of drug combinations on the basis of published literature and pharmacological interactions between drugs, 1712 treatment regimens can be preliminarily studied in clinical trials. After information matching with ClinicalTrials.gov and ChiCTR, 239 protocols have currently completed clinical trials, 546 protocols are being carried out, and 927 protocols can be registered for new clinical trials. Conclusions: The Dice of Clinical Trials can enable preliminary screening of cancer clinical trial protocols for clinical trial investigators. Theoretical oncology involves the pre-research assessment, quantification and prediction of clinical oncology. Combining the Dice of Clinical Trials and theoretical oncology will accelerate the transformation of clinical cancer research from the era of precision medicine to artificial intelligence medicine.