Continuous manufacturing is a relatively new concept for the pharmaceutical industry, although other industries have been pursuing it for years. The strategy entails feeding a constant stream of raw materials into a process that runs continuously at a steady state. Material flows from one step to the next without interruption until a final dosage form is created. A continuous process is not meant to be stopped and restarted, but kept running for an extended period.
This manufacturing strategy could bring the pharmaceutical industry advantages such as consistent and improved product quality, reduced waste, lower inventory and equipment costs, easy scale-up, flexibility, and short cycle times. Continuous manufacturing also could enable chemical reactions that are difficult in batch operations (e.g., those with exceptionally high temperatures or pressures) and, ultimately, real-time release of finished products.
The technology required for continuous manufacturing comes mainly from the oil, gas, and food industries, says Malcolm Berry, head of continuous-processing projects at GlaxoSmithKline Research and Development (London). Plug-flow reactors, heat exchangers, continuously stirred tanks, and distillation columns have long been available for large-scale operations. At first, it was difficult for drugmakers to find models of these machines that were appropriate for small-scale processes, particularly for laboratory processes. Recently, more walk-up laboratory-scale continuous equipment has been introduced.
New developments in reactor technology are encouraging manufacturers to pursue continuous manufacturing. Typically, companies adapt a new process to an existing batch manufacturing plant and often accept suboptimal performance as a result, says Peter McDonnell, senior technical director of Genzyme UK (Haverhill). More pharmaceutical companies began adopting continuous reactors when commercially available packages emerged in recent years such as oscillating baffled reactors from Nitech, mesoscale plate reactors by Corning, and Alfa-Laval’s reactors, which are based on their heat exchangers.
Novartis (Basel) is collaborating with scientists from the Massachusetts Institute of Technology (MIT) to develop new continuous-processing equipment. MIT scientists have developed a new alternative to current granulation processing that combines microscale quantities of active ingredients and excipients to produce a thin film, says Tom Van Laar, head of global technical operations at Novartis. Operators can convert the film into nearly any dosage form. For example, layers of thin film could be combined to form a tablet.
Many pharmaceutical manufacturing sites are reluctant to switch to continuous processing because they have large quantities of batch manufacturing equipment that is underused. But some batch equipment could still be needed to serve as feed tanks, buffer pots, makeup vessels, and waste-handling units. In addition, buffer capacity in a continuous process allows a manufacturer to keep a small inventory of intermediates to allow for on-line maintenance and troubleshooting. Batch equipment also would be needed for long continuous processes such as crystallizations based in continually stirred tank reactors.
Some pharmaceutical companies are intrigued by the potential benefits of continuous manufacturing, but are uncertain about how to convert to it from their existing batch operations. Making the switch requires great understanding of the process, the critical process parameters, and associated critical quality attributes, say Rick McCabe, senior technical project manager of global manufacturing services, and Alex Chueh, director of technology, science, and operations of global manufacturing services, both of Pfizer (New York). A key goal of the transition would be to develop a fully integrated system with robust, high-level control and automation capabilities.
The strategy for converting to a continuous process depends on the number of processing stages a manufacturer requires and the problem being addressed. The more stages a company puts together, the greater the level of process understanding required to ensure that the quality of the end product is fit for its purpose. “For example, many batch processes have an isolation, intermediate purification, or change of solvent in them. It’s very difficult to do that continuously, by definition,” says Berry. Such steps would need to be eliminated from the process if possible by redesigning the entire process at the outset.
The journey toward fully continuous manufacturing typically starts with a single unit operation. A company might introduce continuous reactors or separators, for example, into an existing plant. This strategy is wise until the manufacturer gains confidence in the reliability of the equipment and the process, says McDonnell.
It is also sensible to identify individual operations that could halt the whole production process if they failed. Personnel could either engineer these operations out of the process or provide redundancy by, for example, having one piece of equipment in service and another on standby. Manufacturers also must design the continuous operation to accommodate routine servicing such as water sanitization. Introducing buffer steps could allow parts of the plant to keep running while other parts are closed for maintenance.
One approach to the transition would be to develop a manufacturing-scale continuous process during research and development, then ship the equipment directly to the factory to avoid scale-up and technology transfer. Although the idea may seem cost-effective, a manufacturer would need to pay close attention to installation costs during equipment design. Installing equipment into a new building can be expensive, sometimes more costly than the equipment itself, says Berry. GSK is considering this strategy but has not yet decided to adopt it, he adds.
The industry is going to adopt continuous processing eventually because of its great advantages, and Novartis hopes to be part of the vanguard, says VanLaar. The company began developing continuous technology with MIT in 2007 and plans the first semipilot scale-up of part of the manufacturing process later in 2010. The next milestone will be in 2013, when a segment of the process will be ready for commercialization at close to commercial scale. Novartis intends to market a broad range of products created through continuous manufacturing by 2015. After that, it will take five to 10 more years for the company to switch from its current asset base and product portfolio to the new ones, says Van Laar.
It’s worth the time and effort to convert to continuous manufacturing because the approach is “transformational” and its advantages “huge,” says Van Laar. Patent expirations, generic competition, and economic considerations might lead the industry to reconsider its traditional aversion to risk and new processes. If drugmakers become convinced that continuous manufacturing can help them create products of higher quality than before, in less time and at lower cost than before, it could well become the new industry standard.