Digital: Process Analytical Technology

Cloud-based tools like process analytical technology can be used to improve pharma manufacturing processes, speeding up quality assurance and time-to-market in a cost-effective manner

Jonique Samuels at Optimal Industrial Technologies by SciY

The process of bringing new drugs to market is lengthy, complex and costly. With long development timelines causing the average cost of developing a new drug to increase by $298m to $2.3bn in 2022, finding ways to reduce drug delivery timelines is a priority for the pharmaceutical industry 1.Removing time intensive steps from the drug manufacturing process supports this goal.

Process analytical technology (PAT) is a data-driven framework that aims to improve and accelerate the pharmaceutical manufacturing process by expediting the real-time measurement of the critical process parameters (CPPs) that affect a drug’s critical quality attributes (CQAs) through the automation of process monitoring. As described in detailed guidelines by the US Food and Drug Administration (FDA), the ultimate goal of PAT is to drive innovation and efficiency in pharmaceutical development, manufacturing and quality assurance.2

Huge strides are being made in this field, and these advances are only set to continue as cloud computing paves the way to new opportunities in data mining and predictive analytics. With advanced PAT frameworks, pharmaceutical manufacturers can improve process flexibility, responsiveness, efficiency, quality and regulatory compliance, building in the principles of quality by design (QbD) from the outset.

By using data engineering to support models for data capture, sharing and storage, and data mining to provide unique methods to generate robust prediction models that link CPPs and CQAs, drug manufacturers can create an in-depth, holistic understanding of processes. The integration of these elements supports process optimisation, improves cycle times and reduces raw material use, energy consumption, production costs, as well as rejects and waste.

With the latest generation of analytical spectrometers generating previously unimaginably high volumes of data, manufacturers can look to high capacity computing infrastructures to manage big data analytics, as they may struggle to store them in on-premise servers. Near-infrared (NIR) process spectrometers, for example, which generate multiple spectra per second, produce a volume of data that exceeds the gigabyte range. Cloud computing gives near unlimited space for data storage and processing, enabling the creation of data-driven process intelligence in line with PAT principles.

As well as streamlining analytics, cloud-based centralised platforms for laboratory data can enhance comprehensive process knowledge platforms for drug development, as well as centralised real-time release testing (RTRT). It is also possible to manage PAT methods to create uniform procedures across global sites using a knowledge management platform, enabling shared best practices and supporting multi-site collaboration to allow for operational optimisation and increase consistency. Manufacturers are now able to move away from conventional, centralised high-performance computing strategies that focus on supercomputers. The new, highly interconnected global infrastructure supports the shift towards high throughput computing, which relies on distributed data sharing and content delivery applications that gather data in greater volume, integrating readings across different instruments to create more accurate and precise predictive models.

Cloud computing solutions are fully scalable, giving the flexibility for agile expansions and upgrades, enabling manufacturers to rapidly scale up data processing capabilities and switch back down as needed. Data transparency and availability are also enhanced by a cloud solution, with multi-user access and remote monitoring supported. Instant reporting capabilities also mean that a cloud-based PAT solution can facilitate document creation for quality auditing and control. Cloud computing can support an advanced PAT knowledge management system with computing algorithms to perform complex functions, while time-critical control can be delegated to real-time control systems, such as programmable logic controllers (PLCs) and distributed control systems (DCSs) acting on quality driven setpoint changes from the PAT system.

As well as optimising quality, this cloud-based integration drives the optimisation of existing assets, reducing the need for investment to increase capacity. PAT also enables manufacturers to adopt continuous processing strategies, to help streamline production and processing operations, reducing running costs, cycle times, energy and resource utilisation. Operational costs, for example, have been reduced by up to 40% through the use of continuous flow processes, rather than batch operations, in manufacturing plants.3

Implementing a PAT solution that funnels data to the cloud can help manufacturers to set up advanced automated operations that increase production flexibility and responsiveness, using either push or pull data transfer methods. A push data strategy sends data to a central data lake located in the cloud, while a pull data architecture relies on a central collector to periodically request metrics, and each has its own merits for different applications. For example, if cyber security is the top priority, as is the case in the highly regulated pharmaceutical sector, a PAT ‘data pump’ system is typically based on push methods to support access control.

As pharmaceutical manufacturers seek to cut drug development time and costs, a PAT framework that supports cloud computing for data analysis, storage and reporting can greatly improve flexibility, responsiveness and quality.

With Deloitte reporting that the average expected return on investment for R&D for new drugs fell from 6.8% in 2021 to just 1.2% in 2022 – the lowest percentage recorded in the 13 years of reporting – finding the right data solution that combines global scalability and accessibility with regulatory compliance to streamline process efficiencies is the key to future success.1

References

1. Visit: deloitte.com/ch/en/pages/press-releases/articles/deloittepharma-study-drop-off-in-returns-on-r-and-d-investmentssharp-decline-in-peak-sales-per-asset.html
2. Visit: fda.gov/media/71012/download
3. Calabrese G S et al (2011), ‘From batch to continuous flow processing in chemicals manufacturing’, AIChE Journal, 57(4), pp828-834