Storme-Paris I, Clarot I, Esposito S, Chaumeil J C, Nicolas A, Brion F, Rieutord A, Chaminade P
Groupe de Chimie Analytique de Paris-Sud, EA 4041, IFR 141, School of Pharmacy, Univ Paris-Sud, Châtenay-Malabry, France.
Eur J Pharm Biopharm. 2009 May;72(1):189-98. doi: 10.1016/j.ejpb.2008.11.002. Epub 2008 Nov 25.
Near InfraRed Spectroscopy (NIRS) is a potentially powerful tool for assessing the homogeneity of industrial powder blends. In the particular context of hospital manufacturing, we considered the introduction of the technique at a small pharmaceutical process scale, with the objective of following blend homogeneity in mixtures of seven components. This article investigates the performance of various NIRS-based methodologies to assess powder blending. The formulation studied is prescribed in haematology unit, as part of the treatment for digestive decontamination in children receiving stem-cell transplantation. It is composed of the active pharmaceutical ingredients (APIs) colimycin and tobramycin and five excipients. We evaluated 39 different blends composing 14 different formulations, with uncorrelated proportions of constituents between these 14 formulations. The reference methods used to establish the NIRS models were gravimetry and a High Performance Liquid Chromatography method coupled to an Evaporative Light Scattering Detection. Unsupervised and supervised qualitative and quantitative chemometric methods were performed to assess powder blend homogeneity using a bench top instrument equipped with an optical fibre. For qualitative evaluations, unsupervised Moving Block Standard Deviation, autocorrelation functions and Partial Least Square Discriminant Analysis (PLS-DA) were used. For quantitative evaluations, Partial Least Square Cross-Validated models were chosen. Results are expressed as API, and major excipient percentages of theoretical values as a function of blending time. The 14 different formulations were only satisfactorily discriminated by supervised algorithms, such as an optimised PLS-DA model. The homogeneity state was demonstrated after 16 min of blending, quantifying three components with a precision between 1.2% and 1.4% w/w. This study demonstrates, for the first time, the effective implementation of NIRS for blend homogeneity evaluation, as early as the preformulation step in a small hospital manufacturing unit. It shows how NIRS involving sampling with an optic fibre can be useful to characterise, optimise and control a small-scale mixing processes on the basis of the distribution of APIs and excipients during blending.
近红外光谱(NIRS)是一种评估工业粉末混合物均匀性的潜在强大工具。在医院制剂生产的特定背景下,我们考虑在小型制药工艺规模引入该技术,目的是跟踪七种成分混合物的混合均匀性。本文研究了各种基于近红外光谱的方法评估粉末混合的性能。所研究的配方是血液科规定的,作为接受干细胞移植儿童消化去污治疗的一部分。它由活性药物成分(API)黏菌素和妥布霉素以及五种辅料组成。我们评估了由14种不同配方组成的39种不同混合物,这14种配方中成分比例不相关。用于建立近红外光谱模型的参考方法是重量法和与蒸发光散射检测联用的高效液相色谱法。使用配备光纤的台式仪器,采用无监督和有监督的定性和定量化学计量学方法来评估粉末混合均匀性。对于定性评估,使用了无监督移动块标准差、自相关函数和偏最小二乘判别分析(PLS-DA)。对于定量评估,选择了偏最小二乘交叉验证模型。结果以API以及理论值的主要辅料百分比表示为混合时间的函数。只有通过有监督算法,如优化的PLS-DA模型,才能令人满意地区分这14种不同配方。混合16分钟后证明了均匀性状态,对三种成分的定量精度在1.2%至1.4%w/w之间。本研究首次证明了近红外光谱在小型医院制剂生产单位的预制剂步骤中就可有效用于混合均匀性评估。它展示了涉及光纤采样的近红外光谱如何有助于根据混合过程中API和辅料的分布来表征、优化和控制小规模混合过程。