The CMAC Future Manufacturing Research Hub

Digitally Enabled MicroFactory-based Medicines Manufacture and Supply

The Grand Challenge is driven by a pressing need to exploit new technologies to address the cost of medicines. The aims are to achieve agility (reduced development time), sustainability (reduced material consumption) and security of supply (ability to reconfigure and/or deploy new capacity). These are highly relevant manufacturing research aims that are aligned with industry interests. They continue to frame the development of the Hub’s research plan.

The Grand Challenge “Digitally Enabled MicroFactory-based Medicines Manufacture and Supply” presents significant research questions with considerable potential to deliver impact from:

• Accelerating the development of product and optimised manufacturing processes

• Closer integration of Active Pharmaceutical Ingredient and Drug Product manufacture

• Use of Digital Twins to drive MicroFactory design, operation and control

• Advanced characterisation capability advances across length scales from the molecule to the particle

• Enhancement of Drug Product models

Grand Challenge Themes

We have two main themes in the Hub research project in phase II (2021-2024). In Integrated Supply System Design we maintain an emphasis on supply chain network design and identifying supply demands. In Quality by Digital Design Approach to End-to-end Process Design and Operation we incorporate a Quality by Digital Design Approach with aligned Advanced Characterisation and development of Drug Product Performance Models.

The Quality by Digital Design Approach aims to formalise the exploitation of integrated data through modelling and simulation for predictive design, whilst aligning with regulatory requirements. This approach demands a greater focus on understanding variability and uncertainty in processes and their accompanying data to establish digitally-enabled routes to define robust design spaces and effective control strategies.

Integrated Supply System Design

The evaluation of both single molecule selection and product-process supply system workflows to help examine future Active Pharmaceutical Ingredient selection will be extended from methodologies developed previously in the Hub project.

This is part of the integrated supply-system design analysis for multi-New Chemical Entity /Active Pharmaceutical Ingredient processes across the crystallisation, isolation and drug product objectives in the Quality by Digital Design Approach to End-to-end Process Design and Operation theme.

Quality by Digital Design Approach to End-to-end Process Design and Operation

The CMAC Hub aims to digitalise Quality by Design (QbD), exploiting our extensive modelling and data driven decision support tools within an overarching Quality by Digital Design (QbDD) framework. We are developing an integrated, digitally-enabled QbDD approach that will address uncertainty and risk associated with the development of robust, capable continuous processes and extend to develop and implement strategies for advanced process control.

This research sub-theme targets the development of the QbDD workflow, Digital Twins for design and operation of continuous processes, and the use of models to identify robust design spaces and inform control strategies that will be used to implement integrated continuous processes using CMAC Hub MicroFactory platforms. The technical focus, driven by our partners needs, spans Active Pharmaceutical Ingredient particle formation (i.e. crystallisation, filtration, washing, drying) and Drug Product secondary processing (i.e. polymer extrusion/ printing and powder compaction to form tablets).

QbDD Workflow

Our QbDD Workflow provides the framework for using the DataFactory and Digital Twin to steer process development and MicroFactory operation.

Building on the development of various workflows carried out to date in CMAC, we are establishing the activities required to model and predict process outcomes. Optimised experiment and measurement plans are driven by the model requirements. The work includes the development of models for solubility and solvent selection, optimal design of experiments for parameter estimation and validation, with model-based global sensitivity and uncertainty analysis used to inform selection of suitable process models. The QbDD Workflow drives experimental efforts, augmented through the development of an autonomous crystallisation DataFactory that helps populate the Digital Twin of a process, ultimately enabling transfer of process design and control strategy into operation in the MicroFactories.

Model Driven Autonomous Crystallisation DataFactory

The Autonomous Crystallisation Classification DataFactory will automate experiments done on the Technobis Crystalline platform to deliver large structured data sets for interrogation by image analysis and machine learning.

The crystallisation experiments will sweep through physicochemical phase space from molecule to solubility, kinetics, growth, agglomeration and fouling. The data collected will inform and optimise models of crystallisation via smart experiments. The data collected will be used for research on image analysis, model development and machine learning. This will feed into work to establish a Crystallisation Parameter Database.

Digital Twins

The Digital Twin work in CMAC includes designing the integrated digital framework. It collates, analyses, visualises and applies data, models and our knowledge of the rapid design, control, operation and testing of continuous processes. This helps design MicroFactories for Active Pharmaceutical Ingredient crystallisation and Drug Product production. The Digital Platform supports the Digital Twins.

The QbDD workflow develops a specific Digital Twin for a particular process and product. The various Digital Twins we develop will combine the overarching digital definition of both the process and product in each case. For example, we use the The QbDD workflow to gather data and inform model development, optimisation and implementation. These data, models and knowledge are then captured, stored and interrogated in the Digital Twin framework. The models predict the process operating conditions for the MicroFactory being developed.

Model Driven MicroFactories

As we are using the QbDD Workflow to drive our project, process designs will be established from data collected as a result of following the QbDD Workflows, and then modelled, analysed and optimised to establish a Digital Twin for an Active Pharmaceutical Ingredient or Drug Product manufacturing process, that will inform MicroFactory optimal operating ranges and control strategy. Also, data will be generated to inform the Integrated Supply System Design theme.

For Hub Active Pharmaceutical Ingredient MicroFactories, the default equipment we will use will be MSMPR with integrated PAT-enabled closed-loop control of Critical Quality Attributes linked to continuous or semi-continuous filtration and washing stages and batch drying.

The Drug Product MicroFactory manufacture will be done via polymer processing using extrusion-printing technologies to exploit existing infrastructure across the Hub partners, and alternatively by making capsules.

Advanced Characterisation

The objectives for Advanced Characterisation in Phase II are two-fold:

• To integrate the deeper multi-scale understanding from the multi-technique characterisation paradigm into the QbDD MicroFactory programme

• To strategically develop joint activities started in Phase I

The approach for integration of deeper multi-scale understanding is to apply advanced characterisation techniques where locally-available analytical tools provide ambiguous or no answers. The aim is to provide critical missing information required for QbDD and modelling, identification of mechanistic pathways in API formation, explore the molecular basis for solvent selection, and overcome inefficiencies and manufacturability bottlenecks that arise from incomplete understanding of the molecular and mesoscopic structure and their impact on CQAs, such as flowability, compressibility, cohesion, particle morphology.

The strategic development of joint activities with national (Royce, Diamond, Harwell, NPL) and international (ESRF, Argonne, Brookhaven) central research facilities is being done with a view to establishing a Centre of Excellence in Advanced Analytical Science for Medicine Manufacturing, acting as a hub and relay for seamless knowledge transfer between fundamental and industrial research.

Drug Product Performance Models

The aims of developing DP Performance models are to:

• Close the vast knowledge gap for predictive performance of Oral Solid Dosage forms, by developing mechanistic models to describe and predict tablet dissolution and disintegration

• Validate these models using cutting edge techniques

• Integrate these models with existing and in-development model platforms for Oral Solid Dosage manufacture, including traditional tableting and novel polymer process methods developed in the QbDD Workflows approach