Mathematical modelling in the pharmaceutical industry in Ireland

By Dr Kevin Moroney

The pharmaceutical industry like many others is undergoing a technological revolution as we enter the much vaunted era of Industry 4.0. This is of particular strategic importance to Ireland, home to all of the world’s top 10 pharmaceutical companies and one of the largest exporters of pharmaceutical products in the world. One key area that this is manifesting itself is in the introduction of continuous manufacturing lines for drug products, such as tablets, instead of traditional batch processes. This transition poses a number of technical challenges in terms of process understanding, modelling and control as well as regulatory challenges. Unsurprisingly, the pharmaceutical industry is highly regulated to ensure safe and consistent products. Replacing tried and trusted processes with new ones requires modernisation of regulations and proof that previous standards can be matched or improved upon. Researchers from MACSI in the University of Limerick have been working on some of these issues with collaborators from the Synthesis and Solid State Pharmaceutical Centre (SSPC) and industry partners in a project called MOMEnTUM (Modelling of Multi-Phase Transport Processes to Enable Automation in Manufacturing).

Tablets consist of a mixture of powders or granules of (i) active pharmaceutical ingredient(s) and (ii) excipient or filler materials, which can be chosen to improve the flow of the powder mixture or the mechanical strength and drug release behaviour of the final tablets. Traditionally drug products have been manufactured in batches. The ingredients of the formulations are initially weighed to make up a target batch size. This is followed by blending and compaction into tablets in successive unit operations, although other processing steps may be required. The tablets then have to be carefully sampled and tested to the specified requirements. Any defects detected in the final tablets can lead to the rejection of the entire batch. In contrast, in continuous processing the raw ingredients are continuously fed into the system, conveyed through the various processing stages and removed as a finished product at the end of the processing train. The continuous nature of the process facilitates the incorporation of in-process monitoring tools allowing for the implementation of control strategies so the process does not deviate from established quality requirements. In an ideal scenario, this allows for the real-time release of tablets, as the company has the data to show a robust process has been implemented. This compares to batch-processing where quality has to be assured by off-line testing of tablets from each batch, which can delay product release for a number of weeks.

Companies have recently attempted to get regulatory approval from the Food and Drug Administration (FDA) to switch manufacture of an oral solid dosage form (tablet) from a batch to a continuous process for a small number of drug products. Evidence of the resulting benefits, including the potential for a faster, more cost-effective and tightly controlled process, mean that our industry partners and other companies are targeting the transfer of many more of their products from batch to continuous processes. Nevertheless, this transfer is highly challenging and each drug product presents its own challenges.  Pharmaceutical companies know how to build and run batch processes. In order to be an economically viable alternative, the development time for continuous processes needs to be cut or the benefits of a more efficient process are lost in the development cost and the risk of failure.

Companies are increasingly turning to mathematical modelling as the engine through which they can capture fundamental process understanding. This involves creating physical models which relate key quality attributes to raw material properties and the design and operating parameters of the process. These models can be incorporated into existing process modelling tools to allow digital design of processes based on the minimum required experimental data, by creating a so-called digital twin of the process. This digital twin can also be used in conjunction with processing monitoring or PAT (Process Analytical Technology) tools to control the process once in place.

There has been significant research into this area in recent years, but there remains a need for more accurate models of various fundamental physio-chemical processes including powder flow and compaction, bioavailability and drug release. Many current approaches still rely on statistical and semi-empirical approaches or the knowledge of experienced practitioners. The MOMEnTUM project has focused on novel methods to optimise formulation design for continuous manufacturing, process scaling and on modelling of specific unit operations.

Following on from this work, MACSI is starting to work on new projects with the SSPC and industry to develop simplified models of fundamental processes which are of key importance to further enable in-silico process design and control and drug release testing. Collaborating with process specialists from academia and industry, the projects aim to exploit MACSI’s core expertise in asymptotic analysis, model reduction and data analysis to isolate the key underlying mechanisms in each process and select model parameters through statistical data analysis techniques. Traditionally Ireland has largely been a manufacturing base for pharmaceutical companies, with limited R&D presence. However, as the major players in the pharmaceutical industry look to establish an increased R&D presence in Ireland and exploit the potential benefits of continuous manufacturing, there are exciting opportunities for young researchers with expertise in mathematical modelling and data analytics.

Dr Kevin Moroney is a postdoctoral researcher at MACSI and at the SFI Pharmaceutical Research Centre, Synthesis and Solid State Pharmaceutical Centre (SSPC)