Digitalizing pharmaceutical development and manufacturing: advanced mathematical modeling for operation design, process monitoring and process control

For pharmaceutical companies, the economic return on investments on research and development has recently been decreasing, mainly due to the large cost (~$2 billion) and time (~10 years) for bringing a new product to the market in the latest years. At the same time, an alarming number of drug shortages and recalls for quality concerns has been registered by regulators. These events affect companies, from the financial side, but also patients, who might experience increasingly large costs for drugs, or unavailability of essential medicines. The lack of adoption of modern technology and approaches for pharmaceutical development and manufacturing is acknowledged as a main actor for these events. A recent example is the sluggish rollout of COVID-19 vaccines, which has been significantly affected by technological limitations in pharmaceutical development and manufacturing, especially regarding process scale-up. In the early 2000s, a modernization momentum of the sector was initiated, culminated into the Quality-by-Design (QbD) initiative. Within the QbD initiative, regulators defined a novel pharmaceutical development and manufacturing paradigm, rooted in product and process understanding and based on sound science and quality risk management. However, much effort is still needed by the pharmaceutical sector to catch up with other industries on the adoption of modern development and manufacturing technologies. Recently, QbD is evolving towards a new phase, that features the adoption of novel emerging technology, the most important ones being continuous processing, active (i.e., closed-loop) quality control and increased use of mathematics. Mathematical modeling can be used for developing digital tools pivotal to the efficient and rapid implementation of QbD, and its adoption has also been recommended by regulators with dedicated guidelines. Mathematical methodologies can support all stages of the pharmaceutical life cycle, and enable the implementation of continuous processing and active quality control. Within this context, the role and expertise of chemical engineers, especially of the process systems engineering field, are of utmost importance. The objective of this Dissertation is to promote the use of advanced mathematical modeling techniques within pharmaceutical development and manufacturing environments to: i) reduce pharmaceutical development time and cost; ii) increase the efficiency and the robustness of pharmaceutical manufacturing. These objectives are achieved by developing and implementing mathematical methodologies in key areas of pharmaceutical development and manufacturing: operation design, process monitoring and process control. The case studies span across the whole pharmaceutical flowsheet, but are particularly focused on continuous manufacturing processes. Applications of mathematical modeling are specifically addressed to tackle current bottlenecks towards the transition to end-to-end continuous pharmaceutical processing. The results presented and discussed in this Dissertation make several steps forward in the journey to adopt model-based methodologies for modernizing pharmaceutical development and manufacturing. Enabling technologies for the novel Quality-by-Control paradigm and for the transition to end-to-end continuous manufacturing have been developed. In particular, the presented results are expected to promote the adoption of advanced fault detection and diagnosis, digital operation design and closed-loop quality control routines in the pharmaceutical industry.