強化設計標的臨床試驗下設限資料統計推論之研究

Abstract

在傳統的臨床試驗中，納入和排除標準通常是基於一些臨床指標而未考量受試者的基因或基因的變異。在完成人類基因體計畫後，因可鑑別疾病的分子標的，進而發展出分子標的治療方法。但是分子標的鑑定的診斷試劑通常並非百分之百準確，所以納入標的臨床試驗的陽性診斷病人實際上有些可能並沒有此分子標的。因此，標的臨床試驗下之標的療法對於真正擁有分子標的之病人族群而言的療效估計值會有偏差。因此，我們提出對於真正擁有分子標的之病人配合標的療法之不偏推論的統計方法。在強化設計的臨床試驗及指數分佈及比例化風險迴歸模式下，我們提出利用EM演算法配合拔靴技術並考慮鑑定分子標的之診斷試劑的準確度，針對設限資料來進行處理效應之推論。並運用模擬研究加以評估所提出估計式與檢定方式的表現，及提出實例數據以說明方法的應用。

For the traditional clinical trials, inclusion and exclusion criteria are usually based on some clinical endpoints, the genetic or genomic variability of the trial participants are not totally utilized in the criteria. After completion of the human genome project, the disease targets at the molecular level can be identified and can be utilized for the treatment of diseases. However, the accuracy of diagnostic devices for identification of such molecular targets is usually not perfect. Some of the patients enrolled in targeted clinical trials with a positive result for molecular target might not have the specific molecular targets. As a result, the treatment effect may be underestimated in the patient population truly with the molecular target. To resolve this issue, under the exponential distribution and the Cox-Proportional hazard model, we develop inferential procedures for the treatment effects of the targeted drug based on the censored endpoints in the patients truly with the molecular targets. Under an enrichment design, we propose using the EM algorithm in conjunction with the bootstrap technique to incorporate the inaccuracy of the diagnostic device for detection of the molecular targets on the inference of the treatment effects. A simulation study was conducted to empirically investigate the performance of the proposed methods. The impact of the simulation of the assumption for the proportional hazard model was also examined in the simulation study. Numerical examples illustrate the proposed procedures.