藥理遺傳研究樣本數計算

Sample Size Requirements for Pharmacogenetic Studies

林滋堉(Tzu-Yu Lin)

指導教授 ： 戴政

國立臺灣大學/公共衛生學院/流行病學研究所/碩士(2007年)

https://doi.org/10.6342/NTU.2007.10349

摘要

藥理遺傳研究將人體體質的差異納入藥理研究中，深入探討藥物與遺傳基因對藥效的影響性。藥理遺傳研究可依照是否採取隨機試驗步驟區分為觀察性藥理遺傳研究及試驗性藥理遺傳研究，本論文循序漸進介紹採用單一藥物只關心藥物反應與遺傳基因相關的病體－對照觀察性藥理遺傳研究，及延伸至使用多種藥物的試驗性藥理遺傳研究，萃取藥物反應與藥物試驗、遺傳基因與試驗－基因交互作用的相關訊息，運用由簡而繁的藥理遺傳研究設計勾勒出藥理遺傳研究個人化投藥目的藍圖；本論文針對每一種研究設計提出偵測相關訊息的分析方法並計算研究所需樣本數；此外，本研究不但採用群體為基礎的資料結構作為分析依據，更探討家庭為基礎的同胞對研究設計，給定相似的背景之下進行藥物、遺傳與藥效間的資料探勘步驟，以供研究人員處理族群混合的藥理遺傳研究方式。

關鍵字

藥理遺傳研究 ； 病體－對照研究 ； 藥理遺傳試驗 ； 同胞對 ； 樣本數

並列摘要

Pharmacogenetic studies investigate the inter-individual variability due to the genetic effect on drug response. Pharmacogenetic studies can be classified to experimental studies and observational studies with or without randomization. In this thesis, we aim to investigate the useful designs, from simple to delicate, in pharmacogenetic studies. In our framework, we use case-control study design to dissect the association between genetic effect and drug response when there is one drug treatment. Furthermore, when there are two or more treatment groups, we detect the effects of treatments, genotypes, and gene-treatment interaction on drug response under trial design. For each design we present the analysis methods for detecting association and the methods for calculation of required sample size. The family-based (sib pairs) design is especially discussed for the intention of setting up a robust design against population stratification.

並列關鍵字

pharmacogenetic study ； case-control study ； pharmacogenetic trials ； sib pairs ； sample size

參考文獻

CIOMS (2005) Pharmacogenetics: towards improving treatment with medicines, Geneva.

Duncan C. Thomas (2004) Statistical methods in genetic epidemiology, Oxford University Press, New York.

Elston RC, Idury RM, Cardon LR, Lichter JB. (1999) The study of candidate genes in drug trials: sample size considerations. Statistics in Medicine. 18:741-751.

Garrod A.E. (1931) Inborn Factors in Disease: An Essay, Oxford University Press, (Abstract).

Greenland S. (1985) Power, sample size and smallest detectable effect determination for multivariate studies. Statistics in Medicine. 4:117-127.