It’s no secret: the biopharma industry is changing, and some say it’s about time. Compared with other manufacturing industries such as aviation, food, automotive and consumer goods, biopharma has been notoriously slow to adopt automation and other technologies that could drastically improve production and accelerate drug development. In fairness, the manufacture of biopharmaceuticals is incredibly complex. For some, however, that high level of complexity is a good reason to reconsider how things could be done.
Many of these subsequent changes fall under the umbrella of “Biopharma 4.0”, a term used to describe a convergence of informational, operational and processing technologies. It is an extension of the Industry 4.0 operating model, a strategic initiative of the German government designed to drive digital manufacturing forward. The new approach – thought of as the fourth industrial revolution – is expected to enable a significant leap in productivity, by harnessing interconnected instruments, automation, greater transparency and decentralized decision-making. In practice, this may include the use of machine learning, robotics, cloud-based data storage and artificial intelligence – all of which can support real-time monitoring and faster decision-making. Currently, many companies, vendors and research institutes are navigating exactly how these technologies and approaches could benefit biopharmaceutical development and biomanufacturing.
What is driving this digital transformation?
Biopharma 4.0 was borne out of both necessity and opportunity. Developing new therapeutics is highly inefficient, slow and bears a high cost and failure rate. At the same time, analytical technology has evolved significantly, making it easier to screen for impurities that might compromise the safety or efficacy of a compound. Regulatory bodies have also become more process-oriented and expect manufacturers to place a strong focus on quality and data integrity. New approaches are therefore needed to help manufacturers navigate regulatory challenges, remain competitive and maintain global supply chains of highly complex biotherapeutics.
Michael Sokolov, lecturer at the Institute of Chemical and Bioengineering at ETH in Zurich, paints a picture of bioprocess development using a unique analogy: “Imagine process optimization is like baking a cake, and you need to figure out the perfect recipe. Instead of having a dozen parameters to control, there are thousands to consider. But the time you have is not much more than what is needed to actually bake the cake – you might have only three iterations. Therefore, we see as ourselves as engineers under huge pressure; we must make many decisions without having time to gather the evidence that would lead us down one path or another.”
In addition to the race against time – continuing with the cake analogy – manufacturers are trying to bake as cheaply as possible and are therefore seeking ways to improve efficiency. One such opportunity can be found in continuous manufacturing, a style of production where quality control is built into the bioprocess. As continuous manufacturing is dependent on the rapid assessment of critical manufacturing parameters, this trend has only intensified the demand for advanced, automated analytical tools.
A Pan Industry Revolution: Digital Transformation and Lab 4.0
In order to remain competitive a central digital strategy is imperative. The application of smart and digital technologies, such as artificial intelligence (AI), cloud and platform technologies, informatics solutions and the Internet of Things (IoT), are enabling organizations to create connected digital ecosystems and work toward a Lab 4.0 strategy. Machine learning – a subset of AI – is an analytical approach that relies on input data, environment and feedback to improve through experience, produce intelligence and can be deployed to solve complex problems. Download this whitepaper to find out how digital technologies are helping across numerous industries.Download Whitepaper
Perks of greater connectivity: enhanced compliance, process control and efficiency
By embracing technology encompassed by the 4.0 revolution, companies are overcoming a huge number of modern challenges. Centralized, cloud-based data storage capabilities protect businesses against events like power outages, floods and fires – and they are increasingly designed for usability. Alongside improved searchability, simpler version control and streamlined data management, ease of compliance is a welcome feature. A strong example of this can be found by considering data integrity and the tightening regulatory requirements that have emerged to reduce cGMP violations and protect public health. In many cases, existing paper-based and electronic systems cannot support the volume of record-keeping needed to collect metadata and satisfy an audit trail review. Therefore, digital solutions have emerged to enable full traceability and compliance, enabling automated record-keeping, unique user identities and more efficient audit trail reviews. Professor Pauline Rudd, emeritus fellow at University College, Dublin and visiting investigator at the Bioprocessing Technology Institute, A*STAR, Singapore, highlights the benefits of modern solutions: “In bioprocessing, time is really important. And with this focus on process quality and data integrity, it is better to buy instruments with built-in, validated programs where everything is already compliant. And then if a vendor makes a change, they clear that with the regulators – not you.”
Failure to meet regulatory guidelines may have many potential negative implications for biopharma laboratories, such as product recalls or compromised public health – therefore having confidence in bioprocesses is critical. Inefficiencies and downtime can further reduce a laboratory’s ability to meet their own business goals and client requirements, creating additional challenges. In recognition of modern challenges, and the need for greater confidence and efficiency, a range of technologies and workflows are being explored, for example:
- Automated annotation of N-glycans in ultra-performance liquid chromatography/mass spectrometry (UPLC-MS) chromatograms
- Programmable liquid handling robots for automated stem cell culture and production
- Digital Twins; virtual counterparts of physical systems or processes that enable predictive manufacturing, e.g., for cell culture process optimization
- Laboratory Information Management Systems (LIMS) enabling centralized data capture and storage, and a standardized data source for analysis
- Automated headspace sampling for the analysis of volatile organic impurities and residual solvents
- Automated data analysis for metabolomics based on gas chromatography-mass spectrometry
The benefits of embracing automation extend beyond streamlining compliance behind the scenes; automation in the laboratory improves reproducibility through the elimination of manual errors. Using technology like miniaturized purification and simulation technology, manufacturers can plan experiments in greater detail, and test them carefully using automation processes on a small scale. Greater precision, achieved with automation, complements miniaturization to accelerate process optimization and reduce the use of expensive reagents. Modern instruments and software can also lessen the burden of instrument maintenance by enabling users to anticipate any issues before they arise, and schedule maintenance based on usage.
An Interview with Yun Zhao, Admera Health: Using Connected Tools in the Lab
A new technology used for analysis of pipetting was recently discussed with Yun Zhao, Director of Biopharma Services at Admera Health, about how he’s using it in his complex research. This technology, TRACKMAN® Connected, consists of a tablet with accessories and apps that makes pipetting on microplates more traceable and reproducible. The tablet is connected to a pipette, where it can monitor and record administration of liquid samples on microplates, and stores digital data readily available to review. View this article to learn more about TRACKMAN® Connected.View Article
Maximizing data insights for drug discovery and development
For Rudd, harnessing automation in bioprocesses has been a high priority for a good part of her career, and she has seen how the emergence of big data has created both significant opportunities and challenges for drug development. In particular, Rudd has been developing automated technologies for glycoanalytics – the study of proteins with covalently-attached oligosaccharide chains (glycans). “The sugars are really important from a control point of view,” says Rudd, “because if a process is well managed, then those sugars will be consistent. Also, sometimes a sugar may have a really important biological function.” To improve glycosylation analysis for process quality control and biological research, Rudd and colleagues have been developing computer software programs to determine features of glycosylation. Unlike previous, laborious approaches, Rudd says that modern glycoanalysis tools can generate hundreds of structures at the press of a button – and much more thoroughly.
Reducing and predicting drug toxicity is a high priority in the field of drug discovery. Therefore, artificial intelligence and machine learning are being explored for their ability to identify factors affecting a biotherapeutic’s safety and efficacy. When it comes to bioprocess development, the large number of parameters can be overwhelming. “You could have 200 parameters that you could change, and you end up with these huge matrices with no information to fill them in,” says Rudd. “If you’re going to build up a body of information, you can’t do it manually. You’ve got to have some sort of monitoring process.” Like many others, Rudd is setting her sights on multiomics approaches. “When you’re working with different branches of “omics”, and all your data are in different units, it’s difficult to inter-relate them. That’s why we’re using machine learning.”
Having tools to quickly identify patterns in large, complex data sets would be highly beneficial for drug target identification and drug repurposing, and research is underway in these fields to see how machine learning could help find solutions for the COVID-19 pandemic. Recently, significant progress has been made in artificial intelligence, where a network was successful in determining a protein’s 3D shape from its amino acid sequence. “Although we have physics, chemistry and all these disciplines, data analytics is super mighty – and we will see more and more of these tools being enabled,” says Sokolov.
Navigating the Digital Transformation Journey
The idea of fulfilling a digital transformation is appealing, but how is it possible to transition from manual or isolated lab informatics processes to a digital ecosystem? So much of the information available is driven by hype and based on speculation and inflated claims. Download this whitepaper to learn more about how you could successfully transition.Download Whitepaper
Digital transformation and big ideas are gaining traction
At present, there appears to be a positive feedback loop in play: with new, technological advances, more complex challenges are being tackled, and regulatory bodies are placing a greater emphasis on process control. In turn, biopharma companies are adopting 4.0 technologies to ease compliance and accelerate drug discovery and development – and they are becoming more competitive as a result. Now, transitioning towards biopharma 4.0 appears to be an intuitive, necessary move in order to remain competitive and make informed decisions in what looks set to be a data-driven future. Like many others, Rudd is dreaming big and backing the capabilities of her IT team to deliver: “I don't like things to be too complicated, because then they just get uninterpretable. So, I go to my IT person with some very simple ideas. And then I say, if I can think about it, you can do it. And I walk away. And in six months, he comes back with the answers, and I love it!”