基于阿基米德Copula回归模型双计数数据新药安全性有效性评价

Dose finding based on count efficacy-toxicity response using Archimedean Copula

目的:研究新药安全性指标和有效性指标为计数数据以及相关条件下的联合评价及剂量选择问题。方法:基于阿基米德Copula性质建立了双计数数据联合评价模型,并提出了联合概率法进行剂量选择。结果与结论:阿基米德Copula具有良好的效果。首先,可以方便地刻画变量间相关性。另外,安全性和有效性回归模型参数具有明显生理学意义。同时,阿基米德Copula对概率分布没有任何限制。模拟结果表明,阿基米德Copula联合模型可以很好地刻画具有相关性安全性有效性指标,同时选择最优给药剂量。

Objective：To propose a new strategy for dose-finding in clinical trials when both efficacy and toxicity responses are count data with dependence. Methods：A bivariate count responses model was established using Archimedean Copula,and a dose decision rule was developed based on the joint probability of both efficacy and toxicity. Results and Conclusion：Significant gains can be achieved when using Archimedean Copula joint model. Firstly,the correlation between efficacy and toxicity can be estimated. Secondly,the regression parameters in Archimedean Copula for both efficacy and toxicity are meaningful. Thirdly,Archimedean Copula has no restrictions on the probability distribution of response. The simulation study showed that the new joint model based on Archimedean Copula can enable physicians to evaluate the efficacy and toxicity simultaneously and select the optimal dose successfully.