Departments of BioPharmaceutical Product Sciences and.
Statistics and Programming, GlaxoSmithKline Inc., 1250 S Collegeville Road, Collegeville, PA 19426.
PDA J Pharm Sci Technol. 2020 Jan-Feb;74(1):15-26. doi: 10.5731/pdajpst.2018.009324. Epub 2019 Sep 13.
Understanding the contribution of relevant factors to the analytical variability of the micro-flow imaging (MFI) technique is of prime importance because of the significance of the subvisible particulate data in biopharmaceutical product development. The current study was performed to determine the contribution of several key variables to the variability of the subvisible particle counts (e.g., day-to-day, vial-to-vial, sample-to-sample, and measurement-to-measurement variabilities) using a nested statistical analysis. The variability was measured in the <10 μm, ≥10 μm, ≥25 μm, and ≥50 μm size ranges along with the total particle count and the maximum and the mean particle size. The contribution of the vial to the variability of the subvisible particle counts was found to be greater than those of the other factors evaluated in the current study. The analytical method variability in terms of percent relative standard deviation with respect to the particle count in the <10 μm, ≥10 μm, and ≥25 μm size ranges was found to be 16%, 40%, and 44%, respectively. A thorough understanding of the contribution of key factors to the analytical variability revealed how the corresponding contribution can be minimized, that is, by increasing the number of vials, samples, and measurements. The results of the current study may be leveraged for the optimization of the analytical method or for minimization of the analytical variability with the MFI technique.
了解相关因素对微流成像(MFI)技术分析变异性的贡献至关重要,因为亚可见颗粒数据在生物制药产品开发中具有重要意义。本研究采用嵌套式统计分析,旨在确定几个关键变量对亚可见颗粒计数(例如,日-日、瓶-瓶、样品-样品和测量-测量变异性)的变异性的贡献。在<10μm、≥10μm、≥25μm 和≥50μm 粒径范围内以及总颗粒计数、最大和平均粒径来测量变异性。结果表明,与当前研究中评估的其他因素相比,瓶对亚可见颗粒计数变异性的贡献更大。在<10μm、≥10μm 和≥25μm 粒径范围内,分析方法变异性(以相对于<10μm 粒径范围内颗粒计数的百分比相对标准偏差表示)分别为 16%、40%和 44%。深入了解关键因素对分析变异性的贡献,可以揭示如何通过增加瓶数、样品数和测量数来最小化相应的贡献。本研究的结果可用于优化分析方法或最小化 MFI 技术的分析变异性。