A novel model of the continual reassessment method in Phase I trial

• Published in: Scientific reports Vol. 13; no. 1; p. 5047
• Main Authors: Zhang, Weijia, Lei, Wanni, Zhu, Xiaojun
• Format: Journal Article
• Language: English
• Published: London Nature Publishing Group UK 28.03.2023; Nature Publishing Group; Nature Portfolio
• Subjects: 639/705/531; 692/308/2779/109/1940; Algorithms; Bayes Theorem; Clinical trials; Computer Simulation; Design; Dose-Response Relationship, Drug; Drug dosages; Humanities and Social Sciences; Likelihood Functions; multidisciplinary; Patients; Performance evaluation; Probability; Reproducibility of Results; Research Design; Science; Science (multidisciplinary); Survival analysis; Toxicity
• ISSN: 2045-2322; 2045-2322
• DOI: 10.1038/s41598-023-28148-4

For the model-based designs, the continual reassessment method (CRM) is widely used to identify the maximum tolerated dose (MTD) in phase I clinical trials. To improve the performance of classic CRM models, we propose a new CRM and its dose-toxicity probability function based on the Cox model whatever the treatment response is immediately observed or delayed. In the process of dose-finding trial, we can use our model in situations when either the response is delayed or not and can derive the likelihood function and posterior mean toxicity probabilities to find the MTD. Simulation is carried out to evaluate the performance of the proposed model with the classic CRM models. We also evaluate the operating characteristics of the proposed model by the Efficiency, Accuracy, Reliability, and Safety (EARS) criteria.