Currently, drug substance manufacturing industries are going through a paradigm shift from traditional batch to digitalized continuous manufacturing integrated with real time quality control system [1]. Such an advanced manufacturing process will significantly reduce the footprint, time, and resources needed to manufacture the drug substance. Knowledge of process dynamics is important to maintain a state of control in CM. Specifically, understanding the impact of transient events helps to identify risks to material quality and to develop an appropriate control strategy
1. A digital toolbox is therefore needed for modelling, optimization, and control of continuous drug substance manufacturing process.
In this work, a digital twin for continuous APIs (active pharmaceutical ingredients) manufacturing process using modular components has been developed. The process flowsheet model is the heart of 'digital twin' and consists of mathematical representation of three modules with the options of adding more as needed. The first module is for feeding and dispensing. It consists of a refill unit, feed tank, pump, mixing, and preheater. The second module is for performing the chemical reactions required to produce the target APIs. This module consists of tubular reactors placed inside a heating and cooling jacket. The third module is focused on separation of API from impurities. A continuous chromatographic model is currently used for separation purposes that can be easily replaced with any other type of separation techniques such as continuous crystallization. Currently, the developed model library consists of more than 32 units across 10 types that has been used to generate the integrated flowsheet model. The model has been validated using experimental data.
The applications of the digital twin have been demonstrated for virtual experimentation, scenario analysis, digital dynamic optimization as well as design, and development of suitable control architecture for a continuous APIs manufacturing process. The dynamic optimization provides the optimum operating conditions for the manufacturing process that improves the critical quality attributes (CQAs). A suitable control architecture then has been developed using digital twin to track the optimum set points for the key control variables. The developed control architecture is also modular in nature and can be easily adapted for different manufacturing processes of APIs that may have different material and information flows, dead time variations, and tuning parameters. The proposed digital twin can save the time and resources needed for continuous APIs manufacturing and improve the product quality significantly.
The objective of this presentation is two-fold; first to highlight the developed unit operation model library and integrated flowsheet model and then demonstrate its application for dynamic optimization and design of control architecture of continuous API manufacturing processes.
Reference
[1]. FDA (2023). ICH Q13, Continuous Manufacturing of Drug Substances and Drug Products: Guidance for Industry. U. S. Food and Drug Administration. https://www.fda.gov/media/165775/download