Modernizing Pharma Manufacturing

RaZZeRs/Shutterstock.com

The pharmaceutical industry has the reputation of being cautious about implementing new technology and slow to make changes. This conservatism is due, at least in part, to regulatory requirement to prove that any process modifications will not have a detrimental effect on product quality. Regulatory agencies and the industry have made a concerted effort over the past two decades, however, to adopt methods and technologies that will improve quality and manufacturing efficiency. There is also an awareness that more changes will be needed to further adapt to new manufacturing approaches, such as continuous manufacturing, Industry 4.0, and personalized medicine, as well as supporting technologies, such as process analytics and advanced process control.

Emerging technologies in a regulated environment

FDA’s Emerging Technology Program (ETP), formed in 2013 to facilitate adoption of new manufacturing technologies, gives industry a forum to meet with FDA’s Emerging Technology Team (ETT) members early in the development stage, prior to filing regulatory submissions.

“Early engagement through the ETP helps the FDA proactively identify and address potential challenges applicants may face and helps eliminate potential delays in the adoption of promising technologies,” says Sau (Larry) Lee, acting director of the Office of Testing and Research in the Office of Pharmaceutical Quality, Center for Drug Evaluation and Research, FDA. “In addition, when regulatory submissions are received, the ETP will work collaboratively with the pharmaceutical quality assessment team to ensure timely assessment of the submission. The continued involvement of ETP from early technology development through application assessment helps ensure the consistency, continuity, and predictability of the assessment-including the facility inspection.”

The ETP is working well, with positive reception by industry and several drug approvals facilitated by ETT discussions, reports FDA. So far FDA has approved four solid-dosage drugs produced using continuous manufacturing: Vertex’s Orkambi (lumacaftor/ivacaftor), Janssen’s Prezista (darunavir), Eli Lilly’s Verzenio (abemaciclib), and Vertex’s Symdeko (tezacaftor/ivacaftor and ivacaftor), which was approved in February 2018. In addition, tablets made using a 3D-printing manufacturing process-Aprecia’s Spritam (levetiracetam) epilepsy treatment-were approved in 2015.

Although personalized medicine is still futuristic, the flexible manufacturing platforms needed to support clinical development for personalized medicine are in development. “One might say that emerging technologies, including continuous manufacturing, advanced process analytics and control, and new manufacturing paradigms-including 3D printing and portable/modular systems-will be essential to the advancement of personalized medicine,” suggests Lee. “Of course, there will be regulatory challenges in the change from bulk manufacturing a one-size-fits-all medicine to a system of more localized manufacturing.”

Industry 4.0

Another change on the horizon for pharmaceutical manufacturing is the “fourth industrial revolution” or “Industry 4.0.” The first industrial revolution introduced mechanical production; the second, mass production; and the third, computers and automation of production. Industry 4.0 is characterized by cyber-physical systems and direct, machine-to-machine communication via the industrial Internet of things (IIoT). In some ways, the pharmaceutical industry is still transitioning to automation (i.e., “Industry 3.0”), but it is moving (albeit more slowly than other manufacturing segments) toward an Industry 4.0 paradigm.

The pharmaceutical industry should recognize, however, that the rate of change in technology is on an exponential growth trajectory, noted Gary Rathwell, president of Enterprise Consultants, in a presentation at IFPAC 2018 (1). Technology is changing so quickly that it is almost impossible to predict what will happen further than five years in the future, he suggested, but inaction would be a fatal business mistake. A key action to take, according to Rathwell, is to adopt new standards for “Intelligent Industrial Devices,” such as instruments and controllers. Communication standards, such as those from the OPC Foundation and the FieldComm Group, and data exchange standards from the International Organization for Standardization and the International Electrotechnical Commission are crucial for making sure devices and systems are designed to communicate with each other. In addition, standards from the International Society of Automation (ISA) specify engineering design data (e.g., ISA S108 for configuring intelligent devices), and the Capital Facilities Information Handover standard (CFIHOS) standardizes product data to facilitate equipment and device acceptance testing and handover. These standards will be increasingly important as machines communicate directly with each other, with less human involvement.

AI and digital twins

Artificial intelligence (AI) or machine learning is an example of a potentially disruptive Industry 4.0 technology. In a type of AI called “deep learning,” computers train themselves at high speeds by running scenarios and learning from the outcomes, using the massive data available at low cost in cloud-based storage, said Chris Larkin, vice-president of Advanced Analytics at GE Digital, in a presentation at IFPAC 2018 (2).  “AI poses difficult questions from a regulatory perspective: if the algorithm that defines a product is constantly changing as learning increases, how do you define and validate the product? The pharma industry needs to think seriously about how AI will affect the industry and how this change can be managed,” noted Larkin.

Using simulation to model a process on the computer and running scenarios in this virtual system rather than running experiments on real equipment is already used in pharmaceutical process development. The concept of the “digital twin,” however, goes beyond traditional modeling simulation, says Dirk Voelkel, chief technology officer of Innovation and Analytics at GE Healthcare Life Sciences. He explains that GE’s digital twin concept creates a generic digital representation of an asset type. The twin captures multiple characteristics of the asset from sensor data, and these data can be used for deviation or anomaly detection, prediction, and simulation. The digital twin models can also learn continuously as they adapt to new information.

“Full digital twins are still at an early stage and currently are used in research settings, but not in manufacturing or in any regulated space. Full GxP-compliance can be a barrier for learning systems, so taking the technology to a cGMP environment will take time,” adds Voelkel. He predicts that the first GMP implementations of digital twins will most likely be for equipment assets. Benefits that will drive implementation include higher asset availability, process robustness, and cost reduction through faster root-cause analysis. “As the technology matures and systems become better connected, there will be more information available to support decision-making, so that risks can be mitigated in a proactive manner,” he concludes.

Digital Transformation

Several pharma companies have announced corporate initiatives to embrace Industry 4.0. 

At Merck, for example, the vision for “Manufacturing 4.0” is a digital transformation to “smart factories” that are connected across the entire value chain to allow responsive and adaptive manufacturing, explains Francisco Leira, executive director, Internal & External Manufacturing, Biologics, Global Technical Operations, Merck & Co. In this paradigm, data are seen as a core asset to be turned into intelligence to maximize the effectiveness of operations. “Although biopharma as an industry has lagged behind industries such as aerospace and electronics, we are learning from others and taking initial steps,” says Leira. Cybersecurity is a key element that must be fully proven, he notes. The need to better connect information technology and operational technology systems, which may be different at various facilities, is a challenge. Another hurdle to be overcome is developing the capabilities of the workforce to use new tools and systems.

Pfizer is also beginning a digital transformation in a move toward connected manufacturing plants. Such connected plants will make data visible and available on demand, empowering right-first-time execution, the company noted in a presentation at IFPAC 2018 (3). The goal is a connected digital operations center for each worker, with access to tools and data for data-driven performance management and real-time, shop-floor decision making. 

Sanofi is working on the digital transformation of two of its biopharmaceutical manufacturing sites, in Geel, Belgium, and in Framingham, Massachusetts, in the United States, the company said in an Oct. 24, 2017 press release (4). New technologies will include collaborative robots, autonomous mobile robots to “transport raw materials, single-use equipment, and finished products to different points in the facility,” augmented reality for technicians, and paperless operations. The company said, “advanced analytics on data from all these systems will help identify and correct potential issues on the factory floor and continuously optimize flow and orchestration of activities.” Sanofi will use “digital twins” to simulate the process. “I see great scope for using digital models and simulations to transform how we plan, design, and operate our facilities from the concept through to delivering products to patients,” said Philippe Luscan, executive vice-president, Global Industrial Affairs, Sanofi, in the press release. The company added that such simulation “provides the level of manufacturing modularity and future flexibility required to support personalized medicine.”

In the new digital plant, the role of the workforce will change significantly. Robotics may be used to eliminate repetitive and manual activities, with their potential for human error, improving safety and control, noted Sanofi. Operations staff will have new tools, including digital collaboration, data analytics, and augmented reality solutions. Sanofi says it has trained more than 1100 employees in the past three years and is “developing a digital mindset and capability at all levels of industrial affairs.” 

Continuous manufacturing

Modern pharmaceutical manufacturing processes are evolving alongside of digitalization, using the tools of process analytics and advanced process control to improve quality and efficiency.  Automated and integrated manufacturing control using real-time data can be applied to either batch or continuous systems, but it is requisite for operating a continuous manufacturing system.

Continuous manufacturing for solid-dosage drugs has come a long way, with robust process analytical technology (PAT) sending real-time data to the control system and four approved drugs being produced commercially, although it can still be considered an emerging technology. Much work has also been done to develop models of pharmaceutical processes. The concept of using advanced model predictive control of manufacturing is gaining acceptance, but more use cases demonstrating the value in commercial applications are needed for wider adoption, says Doug Hausner, associate director at the Engineering Research Center for Structured Organic Particulate Systems (CSOPS) at Rutgers, the State University of New Jersey.

Integrated, continuous biopharmaceutical manufacturing is also making progress, driven by the need for more flexible production with faster commercialization. Using perfusion technology instead of fed-batch reactors allows intensification to smaller reactors that are more flexible, noted Rohini Deshpande, vice-president of Attribute Sciences Process Development at Amgen, in a presentation at IFPAC 2018 (5). The smaller scale accelerates commercialization by allowing scale up from pilot to commercial-size more seamlessly. Amgen’s advanced manufacturing facility in Singapore using perfusion and single-use systems is the first generation of a flexible and intensified platform, and the next step is to further intensify by integrating unit operations to fully achieve continuous manufacturing, said Deshpande.

One of the barriers to fully realize continuous biopharmaceutical production is the need for PAT to measure product attributes in real time. “Currently, there are more than 30 assays with manual endpoint testing. An alternate for some of these tests in the future will be on line sensor technologies. In the near term, we are shrinking the analytical footprint to enable on the floor testing and investing in multi-attribute methods (MAM),” said Deshpande. “We are working with the ETT to engage proactively and eventually use MAM for release testing. The integration of attribute testing and control will be crucial for moving to integrated continuous manufacturing.”

Investing in modernization

Modern manufacturing technologies offer the potential for robust, efficient, and agile production, and pharma companies are considering the business cases for making these investments. “Hopefully, companies-especially early adopters-will continue to publicize their technology advancements and report the associated gains to encourage wider adoption throughout the industry. FDA and industry may need to take additional steps to address perceived regulatory uncertainty and facilitate further adoption,” concludes Lee.

References

1. G. Rathwell, “Industry 4.0 and the IIoT,” presentation at IFPAC 2018 (North Bethesda, MD, 2018).
2. C. Larkin, “AI, Digital Twins, and Advanced Manufacturing,” presentation at IFPAC 2018 (North Bethesda, MD, 2018).
3. M. Tomasco, “Digital Transformation and IIoT,” presentation at IFPAC 2018 (North Bethesda, MD, 2018).
4. Sanofi, “The Digital Plant: from Collaborative Robots to Virtual Reality,” Press Release, Oct. 24, 2017.
5. R. Deshpande, “Progressive Technologies to Realize the Vision of Advanced Biomanufacturing,” presentation at IFPAC 2018 (North Bethesda, MD, 2018).

Article Details

Pharmaceutical Technology
Vol. 42, No. 4
April 2018
Pages: 20–25

Citation

When referring to this article, please cite it as J. Markarian, "Modernizing Pharma Manufacturing," Pharmaceutical Technology 42 (4) 2018.

About the author

Jennifer Markarian is manufacturing editor for Pharmaceutical Technology.