Theoretical And Application Research Of Selective Crossover Design Based On The Principle Of RPW

In parallel-designed new antitumor medicine clinical trial,the benefit of progressed patients in control group may be missed due to group restriction.Consequently,t he related ethical issues as well as the lowered protocol compliance in patients and investigators restricted the development of new medicines seriously.For the problem,the theory of Selective Crossover design in oncology clinical trial was developed,which allows the progressed patients in control group to switch to the treatment group with higher efficacy.As a result,the patients will be most benefited.However,the Selective Crossover design required at least once interim analysis to identify the group with higher efficacy,so there is an ethical issue that the progressed patients before the interim analysis cannot be treated effectively.To solve this issue,we studied the RPW-SC(Randomized Play-the-Winner-Selective Crossover)design based on the combination of the theory on the Selective Crossover design and the principle of Randomized Play-the-Winner in adaptive design.As a result,we developed the AUM-SC design,which is a much more scientific and beneficial method in new developed antitumor medic ine clinical trial for terminal cancer patients.Through the AUM-SC design,the first progressed patients in the control group will be able to switch to the treatment group with a higher ratio when the test drug has better efficacy so as to satisfy the ethical consideration,and receive more efficatious treatment even earlier before the interim analysis as in the Selective Crossover design.Our study included the next three parts:The first part we developed AUM-SC design based on the RPW-SC design and performed Monte Carlo simulation for exploring the optimism range of the control factor function parameter k in AUM-SC.We studied 4 scenarios,hazard ratio HR(control group vs treatment group)selected 1,1.2,1.5 and 3 respectively.The boundary value H and sample size was all seleted as 1.2 and 100 per group.For view the whole shape,the whole process was divided into 10 stages.(1)HR=1,the effect of treatment group is similar to the control,as selected HR is smaller than H,so whatever the parameter k selected,the control group switched possibility always decreased from 20% at the beginning stage to 10% at the end.(2)HR=1.2,as k=1.5,3 or 5,the control group switched possibility increased from 40% to 50%,and remained stable and no more than 50% during all process.While as k=0.5 or 1,the possibility fluctuated irregularly and no more than 30% till the end.(3)HR=1.5,as k=1.5,3 or 5,the control group switched possibility increased from 40% to 80% slowly.While as k=0.5 or 1,the possibility still fluctuated irregularly.(4)HR=3,as k=1.5,3 or 5,the control group switched possibility increased from 60% to 90% quickly,and reached 100% finally.The possibility shape of k=3 and k=5 were almost same.While as k=0.5 or 1,the possibility still fluctuated irregularly.In the second part,we compared the type ? error and the power of three method based on the AUM-SC design,such as ITT,RPSFTM and IPE in oncology clinical trials.The simulation results suggested that on the basis of AUM-SC design,although the type ? error of RPSFTM and IPE was slightly higher than that of ITT,the power of RPSFTM and IPE was higher than that of ITT obviously when the treatment group has better efficacy.The power of RPSFTM was higher.In the practical application of new antitumor medicine clinical trial,the higher statistical power is more important for the patients with terminal cancer.In the third part,we developed Randomization Switch group System,which is based on the existing EDC system(http://www.epiedc.com)and basic RPW-SC principle.The system includes modules such as AUM-SC operation,medical imaging evaluation and the dispensing system etc.The functions of the system including real-time data entry and acquisition,IDMC network review,website back end unblinding and selective group switch therapy on the principle of RPW.Using the new developed system,the patients in double-blinded clinical trails can still be remained blind even after be switched.The simulation results showed that the subject in the placebo group was able to switch into the treatment group with a higher ratio when the test drug has better efficacy,so the ethical consideration can be satisfied well.The application of the Randomization System of Switch group can promote the application of AUM-SC design in oncology clinical trials.The application of this system will improve the quality of clinical trials,reduce the cost of new drug research,promote to the development of antitumor drugs in clinical trials.The innovations of this study are as follows:(1)According to different control mode,based on the principle of RPW a nd SC design,we put forward the design of RPW-SC firstly,and performed Monte Carlo simulation for exploring the optimism range of the control factor function parameter k in AUM-SC.(2)We developed Randomization Switch group System firstly,which is based on the existing EDC system(http://www.epiedc.com)and basic RPW-SC principle.