(377c) Integrated Use of Digital Twin-Based Framework for the Development of a Lomustine Manufacturing Pathway

Traditionally, batch-wise operation has accounted for a large portion of small- and large-scale industrial pharmaceutical processes. However, the development of newer technologies through intensified unit operations for pharmaceutical manufacturing, as well as more precise modeling and online control, has accelerated over the past 10 – 20 years. In addition to these advancements, interest in implementing more continuous pharmaceutical manufacturing techniques in industry has also increased. With the adoption of new unit operations and operational modes, quality guarantees through initiatives such as Quality by Design [1] (QbD) and Quality by Control (QbC) must be maintained to ensure both consumer and operational safety through good manufacturing and modeling practices. In this work, we present a case study utilizing a start-to-finish framework for the full analysis of a high value, low volume pharmaceutical drug pathway with simulation-based alternative flowsheet comparisons. Our group is developing an end-to-end simulation and optimization framework for generating and comparing flowsheet alternatives for pharmaceutical drug manufacturing. The framework includes a custom model library and supports estimation of model parameters from experimental data.

In this presentation, the pharmaceutical drug is analyzed under various operational modes, i.e. batch and continuous for various unit operations. Process development was performed starting with parameter estimation from experiments. For example, crystallization kinetic parameters were estimated with metastable width zone determination at the lab scale in batch crystallizers [2] and nucleation monitoring in microfluidic nitrogen-segmented droplets [3]. We simultaneously developed a digital twin of the process using gPROMS FormulatedProducts to analyze the response of critical quality attributes to process disturbances and better characterize process dynamics under uncertainty. Subsequently, flowsheet simulation and dynamic process analysis via the digital twin was performed. The framework utilizes many computational platforms for analysis by exploiting strengths of different software environments. We utilize Python for parameter estimation and flowsheet simulation, Pyomo for unit operation and flowsheet optimization, and gPROMS FormulateProducts for the digital twin of the process. Thus, the fidelity between gPROMS FormulatedProducts and the custom model library was analyzed as well. The start-to- finish framework is shown to be robust in analyzing different operational modes and conditions in pharmaceutical processing. To demonstrate this capability, we analyze in detail one batch flowsheet, one continuous flowsheet, and one hybrid flowsheet. Here, we compare these flowsheet alternatives manually for proof-of-concept; however, this framework is ultimately to become an end-to-end optimization framework which will algorithmically generate and compare flowsheet alternatives, specifically for the comparison of batch, continuous, and hybrid operational modes.

1. Food and Drug Administration. Pharmaceutical cGMPs for the 21st Century—A Risk-Based Approach; Technical Report; U.S. Department of Health and Human Services, Food and Drug Administration, Center for Drug Evaluation and Research (CDER): Rockville, MD, USA, 2004.
2. Nagy, ZK; Fujiwara, M; Woo, XY; Braatz, RD; Determination of the kinetic parameters for the crystallization of paracetamol from water using metastable zone width experiments. Industrial and Engineering Chemistry Research. 2008, 47 (4), 1245-1252.
3. Lu, J; Litster, JD; Nagy, ZK. Nucleation studies of active pharmaceutical ingredients in an air-segmented microfluidic drop-based crystallizer. Crystal Growth & Design. 2015, 15 (8). 3645-3651