DOE药学应用实践与指南-响应曲面之CCF设计与minitab实践指南.pdf

DOE 药学应用实践与指南 编著：刘春平 响应曲面之CCF 设计与minitab 实践指南 CCF 案例：ER Polymer Coating Process Optimization Study at Lab Scale (4 kg) 实验室规模(4 kg) 下的ER 聚合物包衣工艺优化研究 An optimization study was performed to optimize the three CPPs (product temperature, spray rate and air volume) identified in the previous screening and feasibility studies. The same equipment (Glatt GPCG-5 with a 7 ″Wurster insert) was used as in the process screening study. Since the impact of atomization air pressure was not significant (α = 0.05) in the screening study, it was kept constant at 1.6 bar for the optimization study. 进行了优化研究以优化先前的筛查和可行性研究中确定的3 个CPPs (产品温度， 喷速和风量) 。使用与工艺筛查研究相同的设备(制备 7 ″ Wurster 插入腔体的 Glatt GPCG-5) 。因为筛查研究中，雾化气压的影响不显著(α = 0.05) ，所以优化研究中保 持不变为1.6 bar。 In a full factorial design for three factors at three levels including six center point replicates, the total number of experiments needed would be 33. Though a full factorial design offers maximum resolution, a Central Composite Face-centered (CCF) design was chosen to reduce the number of runs and allow a quadratic model fit while evaluating three levels for each factor. The design had one center point, no replicates, α = 1.0 and 15 runs total. Table 72 shows the study design and acceptance criteria. The measured responses were the drug release profiles (T20%, T50% and T80%), the amount of fines and agglomerates generated and process efficiency. Table 73 summarizes experimental results (Y1 and Y4, other responses not shown). 3 因素 3 水平的中心复合设计包括了 6 个中心点复制数，实验的总数量将为 33 。虽然中心复合设计提供了最大分辨率，但是选择了中心复合表面(CCF)设计以降 低批数量并允许二次模型拟合当评估每个因素 3 水平时。该设计有一个中心点，无 复制数，α = 1.0 ，共15 批。表 72 显示了研究设计和可接受标准。测量的响应 为药物释放曲线(T20%, T5