基于遗传算法的非线性建模法研究中药口服固体制剂多成分群的体内外相关性

项目名称： 基于遗传算法的非线性建模法研究中药口服固体制剂多成分群的体内外相关性

项目编号： No.81202926

项目类型： 青年科学基金项目

立项/批准年度： 2013

项目学科： 医学八处

项目作者： 熊阳

作者单位： 浙江中医药大学

项目金额： 23万元

中文摘要： 口服中药固体制剂的药效是药动学行为不同的多成分群在体内共同作用的结果，与各成分群性质和体外溶出等有一定相关性。但目前中药制剂的IVIVC常集中在某一个或少数几个成分，而中药成分复杂，且同一制剂针对不同疾病的有效物质群可能不同。因此其IVIVC应针对具体疾病，构建多成分群体外释放与病理状态下效应部位药动/药效（PK/PD）的IVIVC模型。本项目基于中医药整体观念，拟以银杏叶提取物（EGB）为研究对象，制备体外释药速度不同的微丸，采用微透析技术研究多成分群在脑缺血再灌注大鼠血液的PK及脑部的同步PK/PD，通过遗传算法对影响EGB体内行为的体外参数进行筛选，借助支撑向量机等非线性建模法构建具有一定信度和效度的IVIVC数学模型。该模型将中药多成分群IVIVC研究中的复杂问题转化成数学模型，以望解决其IVIVC面临的建模难点，并可指导中药制剂组分机理、剂型设计、处方优化、制备工艺、质量评价等。

中文关键词： 银杏酮酯；中药固体制剂；体内外相关性；多成分群；非线性建模

英文摘要： The therapeutic effect of oral solid preparations of traditional Chinese medicine (TCM) is the combined results of multi-component groups at work in vivo with different pharmacokinetic behaviors, and also correlated at a certain degree with the characters and in vitro dissolution of the multi-component groups. At present, in vitro and in vivo correlation (IVIVC) of multi-components of TCM often paid attention to one or several ingredients, while the components in TCM were varied and the active ingredient groups in the same preparations would be different depending on diseases. Therefore, the IVIVC model should be built upon the in vitro release and in vivo pharmacokinetics/pharmacodynamics (PK/PD) of effect side for multi-component groups and also based on a certain disease. In this project, with the guidance of holistic concept of TCM, Ginkgo biloba leaves extraction was selected as a model drug, micro pills with different release profiles in vitro would be prepared and micro-dialysis method would be used to get PK/PD parameters of multi-component groups in blood and brain in rats with cerebral ischemia reperfusion. Genetic algorithm would be used to screen parameters in vitro, with support vector machine method et al utilized to build IVIVC models with certain reliability and validity. These models would tran

英文关键词： Ginkgo biloba extract；oral solid preparation of Chinese medicine；IVIVC；multicomponent groups；nonlinear mathmatical modeling