Monte Carlo法确立IgG4抗体的复合离子纯化工艺设计空间

Monte Carlo Method Establishes the Design Space of Composite Ion Purification Process for IgG 4 Antibody

关键词:  Monte Carlo  单克隆抗体  纯化  复合阴离子层析  设计空间

Key words:  Monte Carlo  Monoclonal antibody  purification  composite anionic chromatography  design space

摘要:

摘要： 目的 用Monte Carlo方法确立IgG4抗体的复合型阴离子交换层析填料NM 90-AgaroseTM HAM纯化工艺的设计空间，并进行验证。方法 利用DOE全因子实验设计对影响复合阴离子交换层析的3个因素（上样 pH值、上样电导、上样载量）进行筛选，建立模型。用Monte Carlo方法确定工艺的设计空间，并模拟设计空间内工艺稳健性，将确定的 NM 90-AgaroseTM HAM纯化工艺与亲和层析相结合，纯化CHO细胞表达的IgG4单抗样品，验证工艺参数的稳健性。结果 Monte Carlo方法确定了复合阴离子层析纯化的3个关键工艺参数的设计空间，其中：上样液 pH 值设计空间为4.6±0.1，上样液电导值设计空间为（50±0.03）ms·cm-1，上样量不大于60 mg·ml-1填料；设计空间内参数的运行能力CPK最低在1.22以上，且预测值与实际值的偏差小于5%。Monte Carlo所建模型有较好的预测性能。设计空间内试验点达标，可以保证复合阴离子层析步骤单元CQA指标的动态稳定。而设计空间范围外，运行1批次，部分指标不达标，表明此法建模可满足一定的指标预测，对工艺过程控制起到预警作用。同时与传统叠加法确定设计空间相比，在处理多个变量时，更为有效，简便，且提高了设计空间边缘的可靠程度。结论 用Monte Carlo模型评价IgG4抗体的复合型阴离子层析纯化工艺，是有效、简便、可行的。

Abstract:

ABSTRACT: OBJECTIVE The design space of the NM 90-AgaroseTM HAM purification process of IgG 4 antibody was established by Monte Carlo method and verified. METHODS Using the Design of Experiment(DOE)all-factor experimental design, the three factors ( pH,conductivity,loading capacity were screened, and the model was established. The design space of the process was determined by Monte Carlo and simulated process robustness in the design space, combining the defined NM90-AgaroseTMHAM purification process with affinity chromatography to purify IgG 4 mAb expressed by Chinese Hamster Ovary cells(CHO)cells to verify the robustness of the process parameters. RESULTS Monte Carlo determines the design space of three key process parameters purified by composite anion chromatography, in which the pH is 4.6±0.1, the conductivity is （50±0.03）ms·cm-1, the loading capacity is no more than 60mg ·ml-1 ; the Complex Process Capability index (CPK) is no less than 1.22,and the deviation is less than 5%. Monte Carlo has a relatively good prediction performance. The design space can ensure the dynamic stability of the Critical Quality Attributes(CQA)of the process.However,outside the design space,one batch was run,and some of the indicators were unqualified,indicating that this method can meet certain index prediction and play an early warning role in process control. Compared with the traditional superposition method,it is more effective and simple to deal with multiple variables, and improves the reliability of the design space . CONCLUSION It is effective, simple and feasible to evaluate the purification process of composite anion chromatography by Monte Carlo model.