Keywords: dose escalation, Oncology, CRM
Typically, most dose escalation methods in Phase I clinical trials have used a maximum tolerated dose (MTD)based on toxicity alone, while ignoring other variables such as efficacy or biomarker data. However, a dose escalation method using two variables such as both efficacy and toxicity, will identify an optimal dosemore accurately with better efficiency for further drug development. Specially, for drugs with a low toxicity rate such as most biologic drugs, a biologically effective dose will be reached much lower than a MTD. Therefore, a method using another endpoint other than toxicity is needed in a dose escalation study specially in immuno-oncology drug development. Currently, the two well-known methods, bCRM by Braun(2002) and EffTox by Thall(2004), for using two variables employ a bivariate distribution which requires an association parameter between the two variables. It is hard to define and the results tend to be sensitive to the association parameters.