Studies On The Methodology For Quality Analysis Of Traditional Chinese Medicine Using Qualify By Design Principle

Since quality by design (QbD) has been introduced as a fundamental pharmaceutical quality model in the development of pharmaceutical products and processes, the opportunities for analytical method developments to follow a similar approach have been discussed. A key component of QbD compliant analytical method is the definition of analytical design space where assurance of quality is provided. Analytical design space is a measure of method robustness, and allows the inclusion of the measurement uncertainty to provide assurance of analytical quality. The aim of this dissertation is thus to study the methodology for quality analysis of traditional Chinese medicine (TCM) using QbD principle. The main innovation and academic contributions of this study are summarized as follows:1 The feasibility of QbD in the development and optimization of a quantitative method was evaluated for a botanical injection. Analytical design space of a solid phase extraction-high performance liquid chromatography-ultraviolet tandem evaportive light scattering detector (SPE-HPLC-UV/ELSD) method to assay nine chemical compounds in Shenqi Fuzheng Injection was defined. The impact of the operating parameters, including sample preparation, chromatographic separation, and compound detection, on the method performance were investigated simultaneously. And the relationships between critical parameters and the analytical quality of SPE-HPLC-UV/ELSD were further studied. Compared with analytical method developed using univariate approach traditionally, the method developed following QbD displayed improved robustness. It is revealed that the analytical reliability and method robustness were ensured by analytical design space. This presented approach provided a tutorial to the development of a robust quantitative method for samples in complex matrics.2 The methodology of consistency evaluation of botanical analysis was presented. SPE-HPLC-UV/ELSD for simultaneous determination of nine chemical compounds in Shenqi Fuzheng Injection was taken as an example. A total of 12 runs within analytical design space were used to develop a principal component analysis (PCA) model. Two kinds of multivariate control charts, including Hotelling T2 and DModX charts, were applied to examine the analytical consistency of the tested runs. The results showed that the presented strategy distinguished the abnormal operation during the analytical procedures successfully. Futher understanding of the analytical processes was achieved by investigating the contribution plots. It is demonstrated that consistency evaluation of analytical quality provided a tutorial to the quality control of analytical procedures for botanical drugs.3 The feasibility and methodology of QbD in the optimization of mass spectrometry response was discussed for a botanical injection. Direct analysis in real time mass spectrometry (DART-MS) was adopted to analyze Compound Kushen Injection. The impact of a number of factors on DART-MS method performance were screened, and the relationships between critical factors and DART-MS response were investigated. The robustness of analytical design space were evaluated through the critical point verification. It is revealed that the analytical reproducibility and robustness of DART-MS were significantly improved by analytical design space. This presented study provided a tutorial to the improvement of DART-MS response and a guidance to the development of a robust QbD compliant MS method for samples in complex matrics.4 The application of DART-MS method, as a promising process analytical technology (PAT) tool, to achieve real-time process monitoring in botanical drug manufacturing was presented. A percolation process with multi-constituent substances for Compound Kushen Injection was taken as an example. Multi-way partial least squares (MPLS) model was established by the obtained multivariate data. A total of eight batches with normal operations were utilized to derive control trajectories, which detected and diagnosed the process faults successfully. It is demonstrated that DART-MS is efficiency for real-time process monitoring in botanical drug manufacturing, and can be potentially used to improve batch quality and process consistency of samples in complex matrices.