Division of Product Quality Research, Office of Testing and Research, OPS, CDER, FDA, United States.
Int J Pharm. 2012 Nov 15;438(1-2):167-75. doi: 10.1016/j.ijpharm.2012.08.033. Epub 2012 Aug 25.
The purpose of this study was to use QbD approaches to evaluate the effect of several variables and their interactions on quality of a challenging model murine IgG3κ monoclonal antibody (mAb), and then to obtain an optimized formulation with predefined quality target product profile. This antibody was chosen because it has a propensity to precipitate and thus represents a challenge condition for formulation development. Preliminary experiments were conducted to rule out incompatible buffer systems for the mAb product quality. A fractional factorial experimental design was then applied to screen the effects of buffer type, pH and excipients such as sucrose, sodium chloride (NaCl), lactic acid and Polysorbate 20 on glass transition temperature ( [Formula: see text] ), monoclonal antibody concentration (A(280)), presence of aggregation, unfolding transition temperature (T(m)) of the lyophilized product, and particle size of the reconstituted product. A Box-Behnken experimental design was subsequently applied to study the main, interaction, and quadratic effects of these variables on the responses. Pareto ranking analyses showed that the three most important factors affecting the selected responses for this particular antibody were pH, NaCl, and Polysorbate 20. The presence of curvature in the variables' effects on responses indicated interactions. Based on the constraints set on the responses, a design space was identified for this mAb and confirmed with experiments at three different levels of the variables within the design space. The model indicated a combination of high pH (8) and NaCl (50mM) levels, and a low Polysorbate 20 (0.008 mM) level at which an optimal formulation of the mAb could be achieved. Moisture contents and other analytical procedures such as size exclusion chromatography, protein A analysis and SDS-PAGE of the pre-lyophilized and final reconstituted lyophilized products indicated an intact protein structure with minimal aggregation after formulation and lyophilization. In conclusion, experimental design approach was effective in identifying optimal concentrations of excipients and pH for this challenging monoclonal antibody formulation.
本研究旨在采用 QbD 方法来评估多个变量及其相互作用对一种具有挑战性的模型鼠源性 IgG3κ 单克隆抗体（mAb）质量的影响，然后获得具有预定义质量目标产品特性的优化配方。选择该抗体是因为它有沉淀的倾向，因此代表了制剂开发的挑战条件。初步实验排除了不兼容的 mAb 产品质量缓冲体系。然后应用部分因子实验设计筛选缓冲类型、pH 值以及蔗糖、氯化钠（NaCl）、乳酸和聚山梨酯 20 等赋形剂对玻璃化转变温度（[Formula: see text]）、单克隆抗体浓度（A（280））、聚集物的存在、冻干产品的解折叠转变温度（T（m））以及重溶产品的粒径的影响。随后应用 Box-Behnken 实验设计研究这些变量对响应的主要、相互作用和二次影响。Pareto 排序分析表明，对该特定抗体的所选响应影响最大的三个因素是 pH 值、NaCl 和聚山梨酯 20。变量对响应的影响存在曲率表明存在相互作用。根据对响应设定的约束，确定了该 mAb 的设计空间，并通过在设计空间内的变量三个不同水平进行实验进行了验证。该模型表明，在高 pH 值（8）和高 NaCl 水平（50mM），以及低聚山梨酯 20 水平（0.008mM）的组合下，可以获得 mAb 的最佳配方。预冻干和最终重溶冻干产品的水分含量和其他分析程序，如尺寸排阻色谱法、蛋白 A 分析和 SDS-PAGE 分析表明，在制剂和冻干后，蛋白质结构完整，聚集最小。总之，实验设计方法有效地确定了这种具有挑战性的单克隆抗体配方赋形剂和 pH 的最佳浓度。
