September 1, 2018 (Vol. 38, No. 15)
Gareth John Macdonald Freelance Business and Science Journalist
Advancing Real-Time Monitoring, Regulatory Compliance, and the Production of Novel Therapeutics
The ability to monitor culture conditions within a bioreactor during the production of a biopharmaceutical is critical.
The only way to ensure the product is of the desired quality is to monitor key parameters throughout the process, from upstream growth to downstream separation. It is vital that processes function as designed and, in the event that a deviation does occur, that it is recorded and flagged.
Accurate monitoring is also critical from a cost standpoint. Understanding whether a process is running as intended is key to ensuring that sufficient reagents are available at the right time. Likewise, accurate inline process data can help biopharmaceutical manufacturers identify inefficiencies and minimize waste.
Effective biosensors are therefore vital to biopharmaceutical manufacturing. It is important that these technologies continue to keep pace with the changing industry and regulatory demands, as well as emerging types of therapy.
Growing Demand for Customizable Sensors
Early bioreactors typically featured built-in sensors designed to monitor fundamental parameters such as oxygen concentration, pH, and temperature. However, as the biopharmaceutical industry developed, and processes became more complex, the demand for sensors that are capable of tracking a wider range of metrics increased. This in turn has driven the development of bespoke biosensors that are customizable to a specific process.
The evolution of biomanufacturing technologies has also driven biosensor innovation. For example, microbioreactors that have become increasingly common in process development rely on the ability to accurately monitor multiple instances of key reaction parameters such as glucose, lactate, pyruvate, and galactose concentration in real time.1
Furthermore, regulatory demands for accurate information about processes and advocacy for quality-by-design (QbD) principles have further advanced innovation in the field of biosensor development and production.2,3 Biopharmaceutical companies are attracted to technologies that make compliance with regulations more straightforward.
The growing diversity of biopharmaceutical products is also driving the development of biosensors. For example, the biosensors needed to monitor processes used to manufacture protein therapeutics are different from those used to track production of monoclonal antibodies.4
Likewise the technologies needed to monitor the production of antibody-drug conjugates differ from those used to assess the manufacture of cell therapies, gene therapies, and vaccines.5,6
The underlying principle is that the growing complexity of biopharmaceuticals, combined with stricter quality requirements, is increasing the demand for more effective monitoring technologies. This, in turn, is prompting innovation in the field of biosensor development.
The iLine Fast, developed by Applikon Biotechnology, is a case in point. The technology is an inline optical cell image analyzer designed to let manufacturers monitor mammalian cell cultures more accurately, explains Erik Kakes, commercial director, Applikon.
“The iLine Fast takes holograms of the cells and, through advanced image analysis software, generates the cell count, size, distribution, and viability data, as well as the viral load information in real time and other parameters automatically during the cultivation,” he explains.
The advantage of this approach, Kakes says, is that it reduces the risk of human error and produces less hazardous waste than traditional biosensor technologies and techniques.
“Estimates of cell count and cell viability are traditionally performed once a day using Trypan-Blue cell exclusion as a method of choice. Stained samples are destroyed afterward, creating toxic waste. Sampling a bioreactor and counting cells involve manual operations, and weekend work is regularly needed.
“One of the challenges of microscope image analysis performed by humans is the reproducibility between samples and the reproducibility between operators. Readings will differ between each person as well as between different samples by the same operators. This makes optical analysis of cells by operators unreliable from a process control point of view,” Kakes says.
The iLine Fast technology is designed to eliminate such variation by replacing human analysis of optical images with analysis software which, Applikon claims, makes for more reproducible and more reliable information.
“The software can analyze more images and more details of each image—resulting in more and better information on a continuous basis. Human observation differs from person to person and from sample to sample, making it a subjective measurement on a discontinuous basis,” Kakes says.
Core to the iLine Fast technology is the application of a technique known as “differential digital holographic microscopy.” In digital holography, a scattered light beam reflected by an illuminated object interferes with a reference beam on a CCD camera allowing for a 3D numerical reconstruction of the object.
“Differential digital holography is an evolution of this base technology that brings increased stability and an important size reduction of instruments. Digital holography can construct intensity images, quantitative phase images, and 3D images covering the shape and density of an object,” Kakes says.
The technology can optimize processes from a product quality standpoint, according to Kakes. “Using it to determine the infection status in virus cultivation, one can detect the optimal harvest point of these cultures in minutes. This immediately affects the quality of the product, increases shelf-life, and increases productivity of the culture,” he adds.
Handling Big Data
Monitoring biopharmaceutical manufacturing generates data. A lot of data. Ensuring that this information is stored, transferred, and analyzed effectively is another factor driving the biosensor sector, particularly the field of biosensor development.
It is vital that biosensors are able to communicate with one another to ensure biomanufacturers have the most accurate overview of production processes, according to Tony Allman, Ph.D., product manager, fermentation at Infors.
“We have a range of soft sensors, including biomass determination and real-time calculations for scale-up based on oxygen transfer,” says Dr. Allmen. Soft sensors, as the name suggests, are software-based measurement technologies that monitor bioprocess parameters using models created from data generated by hard sensors.
“However, one of the most important aspects of soft-sensor development with Infors’ eve® bioprocess software is that the creation is fully integrated—allowing editing and expansion over time, according to specific challenges,” he continues. “Soft sensors can evolve based on increasing amounts of process data, user experience, and the needs of different groups.”
Integrating biosensors is not straightforward according to Dr. Allman, who explains, “The main challenge for modern bioprocess software is integration of ever-increasing quantities of data from different sources.
The main issue is that older supervisory control and data acquisition (SCADA) software systems concentrated mainly on controlling the bioreactor with limited access to metadata, events, and structured planning across the many experiments that make up a project. This is where more modern technologies, like Infors’ eve soft-sensor system, can help, suggests Dr. Allman.
“Soft sensors represent a move from simple calculations to replacing sensors that are not available as probes or analyzers to a more integrated model of the whole bioprocess as the basis for control strategies,” he adds.
In addition, the ability to establish an integrated monitoring system has advantages from a compliance standpoint.
“Soft sensors are only as good as their proven accuracy, reliability, and capability—just like any physical sensor. Having an integrated approach means all the relevant data on soft-sensor content, creation, and use can be easily accessed by regulatory authorities from within the eve system without accessing any live processes,” asserts Dr. Allman.
Product quality can also be improved by employing soft sensors. “As a general point: The more information available, the better the end result. Note that the key word is information and that this must be created from the raw data by analysis and interpretation,” explains Dr. Allman.
Soft sensors can also help biomanufacturers take a broader range of information into account, which was also a driver for Infors’ development of its technology, according to Dr. Allman.
Soft sensors address other critical factors. These include the need to integrate more third-party and historical data from different document formats, and the increasing use of common protocols for data transfer from bioreactors implemented by manufacturers of whole systems and peripherals.
“When everything is in one place and in a common format, reanalysis of historical data from several sources can lead to new insights for improving a bioprocess for virtually no extra investment,” concludes Dr. Allman.
1. Innovative Microbioreactors and Microfluidic Integrated Biosensors for Biopharmaceutical Process Control
2. Report from the EMA-FDA QbD pilot program
3. (Glycosylphosphatidylinositol-Based) Protein Chips and Biosensors for Biopharmaceutical Process Analytics
4. Quantitative measurements in single-cell analysis: towards scalability in microbial bioprocess development
5. Bioengineering Solutions for Manufacturing Challenges in CAR T Cells
6. Enhancing cell and gene therapy manufacture through the application of advanced fluorescent optical sensors