Key points
The Importance of PAT in Process Development
Process Analytical Technologies (PAT) is the initiative published by the U.S. Food and Drug Administration (FDA) which aimed to address steps that bio-production facilities could adopt to optimise their processes through the adoption of process-control strategies that minimised deviances by improving the reliability, reproducibility, and cost-efficiency of processes during production.
PAT are the building blocks for ensuring the reliable development, production, and quality assurance of Biomolecules produced in the Biopharmaceutical Industry. Furthermore, PAT empowers innovation in these industries by reducing process costs, increasing yield and product time-to-market.
To achieve this, utilisers of PAT must determine a Quality-by-Design (QbD) approach, which ensures that product quality is checked and maintained throughout its production.
To achieve this approach, the manufacturer must (1) acquire/have Product Knowledge (know the product itself, define product quality attributes (PQAs, physical, chemical, biological, or microbiological property or characteristic that should be within an appropriate limit, range, or distribution to ensure the desired product quality)), and (2) Process Understanding (understand how every component and step of the production process can influence the quality of the final product, define process performance indicators (KPIs, a metric for the status of each production step e.g., cell density) and critical process parameters (CPPs, a process parameter whose variability has an impact on a critical quality attribute e.g., pH, dissolved CO2, dissolved O2, temperature)).
PAT can be measured in different methods, as shown in Figure 1. In-line (and, to an extent, on-line) measurements are the preferred choice for PAT initiatives as they offer direct monitoring of processes in real time for data-driven adjustment of critical process parameters.
Although they may incur higher costs initially (compared to at-line and off-line measurements), this is quickly off-set by the reduced running costs when we consider fewer down-times (and associated loss of expensive resources and products) due to continuous, direct monitoring of processes.
And when we think about the implementation of automation (only possible with in-line and modified on-line measurements), further cost reductions are possible due to the redundancy of personnel and manual handling and analysis, in addition to increasing efficiency of processes by reducing the frequency of process delays and downtimes.
Connectivity innovations for sensor solutions
Since the glass pH formulations of pH sensors were marketed in 1989 at Hamilton, our traditional (analogue) measurement loops at the bioreactor usually consist of wired connectivity between sensors and transmitters relaying sensitive raw measurement values (nA in the case of dissolved oxygen (DO), and mV in the case of pH), and further physical wired connections between transmitter and the process control system which relay robust 4-20 mA signals.
In traditional analogue loops, each sensor must be connected and calibrated to a dedicated transmitter measuring a specific parameter (e.g., pH sensors are connected and calibrated to a transmitter that is separate to sensors measuring DO), therefore at bioreactors measuring multiple parameters this requires multiple transmitters (Figure 2).
For most production areas and labs, this is an inconvenience due to the physical space of the transmitters and cables, in addition to the maintenance and calibration of multiple discrete loops.
In addition to innovations in connectivity, Hamilton’s focus is innovations in measurement principles for CPPs. The release of the Optical oxygen sensors revolutionised DO, and the advent of “Intelligent Sensor” technologies in 2010 through integrating microtransmitters and optional wireless “Arc” adapters revolutionised in-line data acquisition, transmission, and analysis for these CPPs.
In Arc systems, different parameters can be measured, converted, compensated, and transmitted via robust digital signal to a single device (Figure 2). Furthermore, additional data, including temperature, data quality, sensor health status, and sensor life history can be recorded, offering the opportunity for additional quality control traceability of measurements, which is important for regulatory compliance and documentation (Figure 2).
These innovations in sensor technologies paved the way for simplifying handling workflows. Consider an essential parameter for both upstream and downstream bioprocessing: pH. Multiple maintenance, calibration and diagnostic steps are required throughout the lifetime of an analogue sensor compared to intelligent sensors (Figure 3).
Additionally, automated running of these processes and self-diagnostics for intelligent sensors minimise personnel input and associated delays and costs while simultaneously increasing process performance and efficiency.
Measurement of key performance indicators such as cell density are useful for determining process performance. Capacitance sensors selectively measure only viable (living) cells by applying an alternating electric field which polarises intact cell membranes.
Each cell will be polarised at a specific frequency, that is dependent on the cells size and morphology (dependent on cell type and stage of growth and development), therefore analysing cells at different frequencies can allow deeper insights into the state of the cells (Figure 4).
The Incyte capacitance probe from Hamilton uses 17 frequencies between 300 kHz and 10 MHz (from low to high frequency), which provides additional insights into processes. Off-line devices give similar information as well like, circularity, cell size, cell size distribution, even viability.
The path towards Smart Factories:
With the wealth of process and sensor data available using modern digital intelligent sensors, determining what and how data is analysed and processed is the core to understanding the true value of the data collected, and how they can assist the optimisation of manufacturers processes.
Application of data science approaches will enable data-driven interventions of processes to improve process efficiency and productivity for Smart Factories of the future (Figure 5).
References
1. FDA. Guidance for Industry PAT – A Framework for Innovative Pharmaceutical Development, manufacturing, and Quality Assurance. (2004).
2. Metze, S. et al. Multivariate data analysis of capacitance frequency scanning for online monitoring of viable cell concentrations in small-scale bioreactors. Anal Bioanal Chem 412, 2089–2102 (2020).
