报告人： Theis Lange（University of Copenhagen)
时间：2021-11-04 14:00 - 15:30
地点：Tecent Meeting(979 3153 9451)
Not least within oncology are combination treatments [REF AAA paper], which combines two or more drugs to treat the same patient, becoming more common. One necessary step in a drug development process is to determine the safe doses. In combination therapy this problem is an order of magnitude harder since there is no single “dose with acceptable safety”, but instead a whole frontier of dose combinations which are all at the acceptable level of safety risk. So how to do dose escalation?
We propose a two-stage design, where an initial rule based escalation (ie. 3+3 like along the diagonal of doses) is followed by an adaptive stage where a continuously updated Bayesian logistic regression is used to suggest the dose combination which is likely to reduce maximal uncertainty (or other sponsored chosen metric) on the location of the acceptable-risk-frontier. The latter is a unique featured among proposed designs for combo dose escalation. In addition our proposed method can incorporate a) concurrently running mono-trials, b) insert new doses if deemed required, c) incorporate historical information in case on of the combo drugs is well-known and d) initiate several drug-dose cohorts in parallel to speed up trial completion. The proposed method is compared to other proposed methods in terms of precision and trial speed.
About the Speaker:
Theis Lange is professor at the Department of Biostatistics, University of Copenhagen. He obtained his PhD of Mathematical Statistics from university of Copenhagen. He is interested in causal inference and mediation, statistical analysis of clinical trials, non-linear dynamic models. His current research includes analyzing causal pathways involved in creating social inequality in cancer, comparing different measures of mediation in a survival analysis context and gene-disease association studies, etc...
Prof. Lange is editor for a few top journals. Such as Biometrics, Journal of Business and Economic Statistics, Journal of Econometrics, Oxford Bulletin of Economics and Statistics, Scandinavian Journal of Statistics, Statistica Sinica, Statistics in Medicine, Stochastic Processes and their Applications, The Manchester School, The R Journal.
Tencent Meeting ：https://meeting.tencent.com/dm/WStyUVrbn5bJ
Meeting ID：979 3153 9451
Your participation is warmly welcomed!