Institutes outside Greece
Ειδικά προβλήματα ποιότητας φαρμάκων χορηγούμενων από το στόμα: πεδίο σχεδιασμού, βιοφαρμακευτικό σύστημα ταξινόμησης και στατιστικός έλεγχος διεργασίας με ανατροφοδότηση
Abstract
dc:descriptionQuality by Design is a modern scientific approach in pharmaceutical development, and in this framework, methodologies to determine the Design Space (DS) of a pharmaceutical product are presented. According to the literature review performed, the determination of the DS was approached using various methods with little or no regard of the type of experimental data obtained to this end. This thesis shows the importance of the type of experimental data on deciding the type of the method used; thus the type of methodology applied depends on the type of experimental data. However, if there is doubt about the data collected, their quality can be determined from successive application of the various methods and evaluation of the “composite desirability” measure for each method. Especially in the case of generic drugs, bioequivalence is one of the primary drug quality requirements in tablet form and its in-vitro evaluation can be performed with the dissolution profiles comparison. Though most of the response variables of the DS are single-point-value attributes, a dissolution profile is a multi-point-value attribute and is assessed using integral measures from pair-wise model independent methods. This way, bioequivalence is introduced in the DS determination and process optimization as a Critical Quality Attribute (CQA).The DS is determined by the following multivariate methods: Response Surface Method (RSM) and Mixture Design Analysis, Broyden method, Artificial Neural Network (ANN) and Bayesian Posterior Predictive Approach (BPPA). Each process is evaluated as a single step, enabling further scale-up procedures. If the data are complete and there is little uncertainty, the proper method of determining the DS is the RSM and Mixture Design Analysis. It is recommended that the regression models and the results of this method to be verified by the Broyden method, which is their inverse approximation. If the data involve high uncertainty and correlation structures of the variables, the proper method is the BPPA. Finally, if data are missing, fuzzy or with unknown correlations, the suggested method of determining DS is ANN, where the process is considered a “black box”. Apart from the applied method, two main ways of determining the DS are identified: multi-response optimization and “overlapping” responses. The first approach determines targets, maxima and minima of the responses and leads to DS with bounds in the neighborhood of optimum conditions. The second approach leads to DS with bounds determined from the global lower and upper specification limits of the response variables. Obviously the above analysis requires a well-defined framework of the final specifications, especially in the case of generics development. The method developed here, evaluates the measurement system of samples from different batches of the reference product, identifies the data distribution and applies the “six sigma” method for Normal data, suitably modified or not.Two of the most important CQAs are bioavailability and bioequivalence of pharmaceuticals. Their classification in four classes according to BCS (Biopharmaceutical Classification System) and BDDCS (Biopharmaceutical Drug Disposition Classification System) is widely adopted by companies and regulatory authorities worldwide, in order to set standards in immediate release orally administered drugs, offering many advantages in the development and evaluation of innovative and generic drugs. However, some drawbacks include difficulty or variety of determining methods of the solubility and permeability values, the heterogeneous nature of these classification criteria, which can also vary between the different Regulatory Authorities worldwide, including fuzziness in the boundaries between classes. Thus, two new frames of reference are proposed for mapping these four classes: one is three-dimensional with orthogonal coordinate axes the basic rate constants of the system of mass balance equations for the drug in the lumen and the epithelium of the gastrointestinal tract and in the adjacent plasma, specifically representing drug dissolution, adsorption, degradation, membrane permeation and systemic elimination and the second frame of reference is two-dimensional with orthogonal coordinate axes the ratios of dissolution and systemic elimination to effective membrane permeation rate constants. In both frames, the boundaries of the regions into which the four classes are mapped, are clear and distinct, while BDDCS seems to be rather determining on this classification, confirming the important role of metabolism categorization of drugs. Finally, after the design and development, follows the production stage, an effective control of which is very important to ensure a flexible and robust quality process and final product. The variability of raw materials or process parameters is a common phenomenon, with significant impact on quality, lot losses and high costs of delayed corrections. Feed Forward control is an effective tool in this direction, including Latent Variable methods (PLS, PCA), aiming to reduce the high-dimensional space of numerous variables of the process in a sub-space, more easily depicted and controlled. PLS models are developed, where non-linear models show better forecasting ability, while a new approach is proposed, that effectively combines PCA sub-space design variables with regression models from RSM analysis. Both methods involve quadratic terms, while the second one also presents interaction effects and regression models directly related to the responses. Finally, this approach displays graphically the Control Space, offering a more direct evaluation of new points of the process, without the need for further experimentation, as well as a better optimization of the process in case of variability of material attributes or process parameters.
Degree
thesis:*- Grantor dc:publisher
- Institutes outside Greece
- Year dc:date
- 2016
Author and committee
dc:creator, dc:contributor.*- Authors dc:creator
-
- Chatzizacharia, Kalliopi
- Χατζηζαχαριά, Καλλιόπη
Subjects
dc:subject × 12Rights
dc:rights- Statement dc:rights
-
- BY_NC_ND
- Language dc:language
- gre
Identifiers
dc:identifier.*- Identifier
- 10.12681/eadd/37301