Elements of Life – Pharmaceuticals
Medications with Built-In Quality
In the past, stringent regulations left little room for innovation in the pharmaceuticals industry. As a result, processes often remained unchanged from licensing onward. Soon, however, advances in process technology will ensure quality to such an extent that final testing will be obviated.
New solutions from Siemens continuously monitor the quality of medications throughout production while automatically controlling production processes themselves. The result: accelerated output and reduced costs
Few industries require a level of purity as high as that demanded by pharmaceutical production. The effectiveness of medications and the health of millions of patients depends wholly on quality. That said, it is not unknown for medications to fail quality test.
Last fall, for example, routine tests at one of the world’s leading producers of flu vaccine revealed that a number of batches were contaminated with bacteria. The consequences were dramatic, since the U.S. market suddenly faced a shortfall of almost half the quantities of vaccine it required. Given that production takes half a year, there was no chance of rapidly sourcing replacement supplies. The website of the U.S. Food and Drug Administration (FDA) permanently lists poor-quality drugs and medicines that are either contaminated or contain too many, too few or deficient active substances. "Product quality varies more in the pharmaceuticals industry than in other sectors," says Ingrid Maes from the Center of Competence for Pharmaceuticals at Siemens Automation and Drives (A&D) in Antwerp, Belgium. A rejection rate of five to 10 % is not uncommon in the pharmaceutical industry.
Exceptional Purity. "The problem is that manufacturing processes are often outdated," says Hans Bijl, marketing manager at A&D in Antwerp. Medications have always been produced in stages. After each process, samples are removed—often by hand—and tested, after which the results are documented and filed. Only then can production continue. In other words, quality is not monitored and controlled during manufacturing itself. And production processes often remain unchanged from the date of licensing onward, which can be decades, since any modification has to be approved by licensing authorities. In its new guidelines on Process Analytical Technology (PAT), however, the FDA has now pledged to allow ongoing enhancement of production processes in the industry. Here, drug manufacturers must be able to show that they understand the production process and know what impact a modification will have on a product. Siemens recognized this trend early and has already put together a PAT concept for the pharmaceuticals industry. Furthermore, the concept is being developed in close cooperation with the FDA, now that the latter has decided to put its faith in PAT.
PAT generates the information required to understand production processes and thereby improve product quality. It enables not only inline quality assurance but also automatic process control. Likewise, it largely obviates the need for elaborate lab tests and long waiting times and thus accelerates the manufacturing process. Lisbeth Thierry-Carstensen, who is responsible for quality enhancement at Danish pharmaceuticals manufacturer Lundbeck, identifies another benefit. "Once a patent has expired, it’s not so easy for rival manufacturers to attain the quality standards of our products." This is because the process know-how required to produce a high-quality drug cannot be teased out of a pill in the same way that its constituents can.
Real-time Release. Implementing a PAT concept for a production process involves more than just making a few inline measurements. It implies an integrated approach that comprises risk, data and process analysis, automation, data processing, documentation and process control. First, data analysis determines which process parameters have a decisive impact on product quality. A selection of appropriate inline analytical tools is then made. These generate a picture of how the process works. Following a short learning period, a "fingerprint" emerges of how the process should look when managed optimally and what degree of variation is acceptable. Finally, inline analytical tools are hooked up to an intelligent process-control system. This is equipped with a database on the various production operations and will automatically intervene whenever the process deviates from a predetermined path.
The objective of PAT technology is to achieve real-time release. If it can be shown that the production process has functioned optimally throughout, the drug can be sold without the need for an elaborate final inspection. Thus, quality is automatically "built in" to the product by the process itself. Although individual automation solutions have already been implemented, fully integrated PAT concepts are still at the developmental stage.
"Siemens is one of the leading players here," says Bijl. For example, since 2003, Siemens has been involved in the PaRel (Parametric Release) research project—together with the Netherlands Vaccine Institute (NVI) and Applikon, a producer of fermentor systems—to develop the first PAT solution for biotech processes. NVI is contributing its experience with vaccines, Applikon its know-how with bioreactors, and Siemens its expertise in process analysis and automation.
In the biotech industry, bioreactors are used to produce an active substance from living bacteria or animal cells. Whereas chemical molecules under identical conditions always react in exactly the same way, living cells vary in their behavior. It’s thus important to monitor the processes inside a fermentor continuously. At present, however, this can only be done indirectly by measuring pressure, temperature and oxygen saturation. "It’s like measuring a room’s humidity and temperature and then trying to say whether people feel happy in it," says Leo Hammendorp, Director of Sales and Marketing at the Center of Competence for Pharmaceuticals.
The PaRel team is initially focusing on the production of a whooping cough vaccine. The project involves the use of near-infrared (NIR) spectrometers integrated in the fermentor using fiber-optic sensors. NIR spectrometers measure how NIR light changes when it passes through a liquid. The light causes the molecules in the liquid to oscillate. A specific molecular composition leaves a characteristic pattern in the absorption spectrum, meaning that cellmetabolic processes can be observed as they happen.
According to Thijs Veerman, Director General of NVI, the project is unique. "It’s the first time that an end customer, an equipment manufacturer and an automation specialist have worked together so closely." Before long, NVI should be able to release the whooping cough vaccine in real time.
"PAT means faster and cheaper production at higher quality," says Maes. "A lot of pharmaceutical companies will take this route in the near future." A crucial aspect of all this is the flexibility that PAT provides. After all, patients will one day demand tailor-made medications. This will require precise control of manufacturing processes.
In the meantime, PAT will provide a major step forward by enabling early identification of contamination and the rapid production of replacement supplies. Then the days of drawing lots for flu vaccine—as happened last fall in the U.S.—should be a thing of the past.
Carola Hanisch