MANUFACTURING
KNOWLEDGE SPACE DESIGN SPACE
CONTROL SPACE
Figure 1: A graphical representation of the relationship between knowledge space, design space and control space.
Key Factors in the Design of Experiments
It is essential for topical formulations that during the pre-screen and full DoE experiments are conducted using equipment that is representative of larger-scale equipment in order to derive meaningful qualitative CPP data. At MedPharm, IKA LR1000 lab reactors are used, which allow for the control of all typical processing parameters. This approach crucially avoids ‘noise’ generation and ensures the quality of the output and therefore the robustness of the resultant control strategy. Understanding the influence of scale on the CPP from high-quality experimental work conducted on a small scale forms the basis of any future scale-up work and technical-transfer activities. For a complex cream, typically 12 experiments covering two to four CPPs are conducted in a pre-screening study in preparation for the manufacture of toxicity or clinical batches. The need to expand the actual number of experiments will depend on the outputs from this early screen and the associated risks. Experimental design in both the pre-screen and full study should ideally attempt to push the product to failure to allow clear understanding of the design and control space boundaries. Problems can arise if the developer is conservative in experimentation due to a misunderstanding of the boundaries between success and failure (see Figure 1). A third important factor is the identification of what six-sigma calls the Key 60
P H A RM A F O C U S A S I A
ISSUE 36 - 2019
Process Output Variable(s) (KPOV), those that determine success. It is critical to have evaluation criteria to establish whether the method employed to measure the KPOV will detect critical failure. If not, a mitigation plan needs to be in place as the output of any experimental work will only be as good as the analytical method allows. An array of analytical techniques is required to evaluate the quality of topical pharmaceutical products. These range from commonplace methods, such as High-Performance Liquid Chromatography (HPLC) and viscosity testing, to more sophisticated methodology, such as rheological evaluation, accelerated stressing to show the potential for separation, and in-vitro release testing to check that there is no change in the release/thermodynamic activity of the drug from the formulation. The design space is defined as the multidimensional combination and interaction of input variables and process parameters that have been demonstrated to provide assurance of quality 3. It satisfies the QTPP and CQAs for the product and provides the boundary of any process parameters to which a product can be made. It is important to stress that the knowledge space is not a hard boundary and that defined process inputs are just as important as the measured output. Interpretation of the DoE using statistical techniques such as Analysis of Variants (ANOVA) plots, which show the mean and distribution around the mean, should clearly show the significance of any parameter and any key interactions between parameters and any associated variability. With an increase in sophistication of many experimental design software packages (e.g., SPSS, Minitab, JMP, and Design Expert) it is easier than ever to highlight statistically significant affects across a range of parameters and present them in a concise graphic that decision makers find easy to understand. It is important to note that any process change made within the design space is not considered a regulatory 3 https://www.camo.com/products/pat-qbd-overview.html
change, and hence, directly allows for the flexibility of any future manufacturing process. Having a strong understanding about the boundaries of success and failure from the outset, creates a foundation for a good control strategy. Furthermore, it reduces the impact any changes in excipient suppliers and specifications or site of production may have upon a product over its lifetime. Ensuring a Robust Process
The control space represents a range of critical parameters within which the process will yield an assured output to meet the CQAs and target specification at all times. The further use of experimental design is essential when navigating from the design space to the control space. This will typically involve, two to five factors in full, or fractional factorial, or some other surface response design informed by the data from previous experiment design work. A useful guide to DoE in this scenario can be found online in the form of the engineering statistics handbook . Undoubtedly, the control space for the CPP must sit well within the boundaries of the design space and sufficiently away from the edge of failure to ensure robustness. The outcomes of an optimisation design will provide interpretation and conclusions to show the most desirable settings to achieve a topical product that meets the QTPP and CQAs. A confirmation batch using these settings aims to demonstrate that the response values from the DoE are close to their predicted values. Conclusion
Using a step-by-step and disciplined QbD approach during the development and late-stage formulation of topical products provides a sound and robust platform in establishing the design space for process development. This methodology will ultimately enable the developer to provide a reliable control strategy for manufacturing. The often-complex liquid and semi-solid processing for topical products cannot be underestimated