基于物质流分析的丹红注射液制药过程质量控制研究

Process quality control for digital pharmaceutical manufacturing of Danhong injection based upon material flow analysis

[目的]通过研究丹红注射液生产过程的中间工序物料化学成分分析数据和紫外光谱分析数据,阐明生产过程中各成分的含量变化与转移规律,辨析关键生产工序和关键质控指标,建立各个工序的质量监测模型以保障丹红注射液质量一致性。[方法]收集丹红工业生产数据32个批次7个工序中间产品物料,以高效液相色谱法(HPLC)测定9个主要化学成分含量,并同时采集紫外光谱数据。通过分析各化学成分的相对标准偏差,确定各个成分含量的批次间一致性。通过各个化学成分在每个工序中的收率与物料衡算分析,确定各工序对成分变化的影响及相应的变化原因。以主成分得分、Hotelling-T2、DMod X等指标,建立并比较了丹红注射液各工序的基于HPLC数据和紫外光谱数据的质量监测模型。[结果]丹红注射液主要化学成分在各个工序中均具有较好的批次间质量一致性,RSD均在10%以内。物料衡算结果显示丹酚酸A存在化学转化,其他成分以物理转移为主。以HPLC和UV光谱数据建立了丹红注射液各工序的生产监测模型。[结论]丹红注射液制药过程物质流分析可以明确药效物质的含量变化与转移规律,辨析关键生产工序和关键质控指标,并以此建立质量监测模型。

[Objective] It is the objective of this study to analyze the material in process using HPLC and UV spectroscopyto elucidate the material change and transfer mechanism. We aim to developing a methodology to identify critical process and parameters for statistical quality control（SPC） models to enhance the batch-to-bath quality consistency of Danhong injection. [Methods] The concentrations of 9 main chemical components and UV spectroscopy were collected for 32 batches and 7 stages of the manufacturing process of Danhong injection. The batch-to-batch quality consistency for each stage was evaluated based on the relative standard deviation（RSD） of each chemical component. The material change phenomenon was studied by evaluating the yield ratio of each chemical component and the mass balance to identify the influence of each stage on the chemical components. The SPC models were established with PCA analysis of the HPLC and UV spectroscopy data respectively, with PC1 score, Hotelling-T2 and DMod Xused as index for quality monitoring. [Results] The major chemical components demonstrated good consistency in the multistage process of Danhong injection manufacturing with RSD less than10%. Mass balance analysis indicates chemical reactions of Salvianolic acid A and physical transfer for other components. [Conclusion] Mass flow analysis of Danhong injection process elucidated the material change and transfer mechanism. We identified critical process for Danhong injection and established quality monitoring models for each stage.