基因改造產品統計評估方法的初步研究

Abstract

基因改造 (Genetically modified, GM) 產品泛指基因改造植物(作物)以及延伸出之飼料及食品。基因改造作物的種植面積由1996年的1.7百萬公頃增加到2006年的100百萬公頃。目前市面上已經有11種基因改造作物包括玉米、大豆、棉花、油菜籽、水稻、南瓜、木瓜、甜菜、蕃茄和馬鈴薯。2006年，經濟合作發展組織 (Organization of Economic and Cooperation Development, OECD) 的會員國已有96種商業化GM產品。目前全球基因改造產品市場大約價值62億美金並且以2位數字成長，基因改造產品也佔了商業種子市場的25%。因此，基因改造產品是勢不可擋的。有鑑於此，當前首要之急是有效的管理以及適當的法規來確保基因改造產品的種植與食用安全。 在本初步研究中，我們先回顧基因改造產品在美國、歐盟、日本和台灣的相關法規。然後我們將介紹實質等同 (Substantial equivalence) 的概念，此概念用來說明基因改造產品與非基因改造產品沒有明顯的差異。這表示基因改造產品與非基因改造產品在生理表現、農藝性狀的相等，也就是在生化特徵上為等同。 最近，平均族群相等性 (Average bioequivalence, ABE) 的概念被應用在基因改造產品上(原本應用在學名藥上)。然而，學名藥是應用在很小的病人族群上，需要觀測的成分只有1~2種。反觀基因改造產品是應用在廣大的健康人口上，而且高達100多種需要觀測的營養、毒性成分。因此在本研究中，我們進行一個初步的研究關於應用容忍區間 (Tolerance interval) 來評估基因改造產品的實質等同。這個議題考慮到的層面包括了：容忍區間概念的適用性、決定相等性的界限、使用何種設計、樣本大小、季節與地區變異大小。換句話說實質等同的意義在於基因改造產品的生化特性，是否落入由非基因改造產品生化特性所建構出的正常範圍之內。這代表了基因改造生化特性的族群與非基因改造產品生化特性的族群相等。這也是族群相等性 (Population bioequivalence, PBE) 的概念。當分布近似於常態，可藉由觀察平均和變異是否相等來建構族群相等性。在本研究中，我們觀察了為了達到族群相等，平均和變異所需控制的範圍。提供的資料用於示範所提供的方法。

Genetic modified (GM) products are defined as the genetic modified plants (crops) and their derived feed or food. The cultivated area of GM crops has increased from 1.7 million hectares in 1996 to 100 million hectares in 2006. There are already 11 GM crops on the market including corn, soybean, cotton, canola, rice, squash, papaya, sugar beet, potato and tomato. In addition, in 2006, there are 96 commercial GM products for member states of Organization of Economic and Cooperation Development (OECD). The current worldwide market for GM products is US $6.2 billions with a double digit growth rate and the GM products also account about 25% of commercial seed market. Therefore, trend of GM products seems irreversible. As a result, the current focus is the effective management and adequate regulations to ensure that GM products are safe to grow and to consume. In this preliminary study, we will first review the regulations of GM products of the U.S.A., European Union (EU), Japan and Taiwan. Then we introduce the concept of substantial equivalence for approval of GM products which is defined as the no meaningful difference between the GM products and their conventional counterpart. This implies that the GM products are equivalent to the conventional counterpart with respect to phenotypic, agronomic, and compositional parameters, in other word, the entire biochemical profiles. Not until recently, the principle of average bioequivalence (ABE) for approval of generic drugs is applied to evaluation of substantial equivalence for approval of GM products. However, generic drugs are intended to be used in a small population of patients with one or two active ingredients to compare. On the other hand, GM products are to be consumed by a much larger population of healthy people with up to 100 nutritional, safeties, and other characteristics under investigation. Therefore, in this research, we conduct a preliminary study on feasibility and issues concerning about the application of tolerance intervals to evaluation of substantial equivalence for GM products. The issues considered are the probability content of the tolerance interval, determination of equivalence limits, design of the study, the sample size, and variability due to season and location. On the other hand, the concept of substantial equivalence also suggests that the biochemical profiles of the GM products fall within the range of natural variation established by the biochemical profiles of the conventional counterpart. This implies that the distribution of the biochemical profiles of the GM products is equivalent to that of the conventional products. This concept is referred to as population bioequivalence (PBE). When the distributions approximately follow a normal distribution, the PBE can be established as showing equivalence in both mean and variance. In our research, we also investigate the possible equivalence limits for both mean and variance. Published data are employed to illustrate the proposed procedures.