“The first rule of any technology used in a business is that automation applied to an efficient operation will magnify the efficiency. The second is that automation applied to an inefficient operation will magnify the inefficiency.” — Bill Gates
Continuous processing (also known as continuous manufacture or flow processing) and semicontinuous processing have been around for quite a while, and are the norm in the majority of other manufacturing industries. For example, continuous processing is prevalent in paper production, automobile manufacture, petroleum and gas production, food processing, and electronic components industries. In the pharmaceutical industry, it is primarily used to produce over-the-counter products considered nonpharmaceutical by regulators, such as toothpaste.
The opportunity to change the manufacturing paradigm from predominantly batch manufacturing to a fully integrated continuous manufacture-based supply chain model is—with some governmental pump priming and some manufacturing vision—becoming possible. Batch processing methods allow for the capacity to manufacture without stop for 24 hours a day, seven days a week, 365 days a year as well as the flexibility to respond to market needs. But continuous processing offers the opportunity to achieve improved, leaner results in less time, at reduced cost, with minimum waste and a flexible delivery system. Still, it is neither common nor currently available other than for partial systems or prototype plants.
What is Continuous Processing?
The term “continuous” is applied to all production or manufacturing processes that run with a continuous flow, or, alternatively, for all stages or interlinked stages that run based on a continuous feed and interfeed of material—generally powder or fluid—for that or any stage of manufacture. So, it is important when discussing continuous processing to understand what element of the process, or even of the overall supply chain, is continuous or, otherwise, has those qualities within a specific industry context that may reasonably be considered continuous.
In regulatory terms, the U.S. FDA definitions of both lot and batch in 21 CFR 210.3 are both applicable to continuous processes. The EMA has adopted the ICH Q7 definition: “In the case of continuous or semicontinuous production, a batch may correspond to a defined fraction of the production. The batch size can be defined either by a fixed quantity or by the amount produced in a fixed time interval.” Thus, from a regulatory standpoint, there is no barrier to development of continuous processing, at least in principle, in the case of manufacturers producing for markets that accept U.S. and/or European regulatory findings.
Batch Processing: The Industry Norm
Batch processing has been the pharmaceutical industry norm since the modern industry began in the 19th century, when manufacturers started using new materials and processes to concoct pills and other medicines. Previously, these were developed using mortar and pestle or mixing spoon stirred liquids from mug to mug or glass to glass (1). Large scale manufacturers generally still follow a simple volume controllable recipe process that is staged and can be readily stage tested and quality checked with reasonable surety. Even now, the tools used are scaled-up items similar in most instances to those of a domestic kitchen—bottle and bag, sieve, mixing bowl and whisk, oven, kettle, etc. This is also the way we teach the processes we have designed, developing the next steps of our curative journey of kitchen experimentation from school laboratory through university laboratory to commercial laboratory research and development and onward, once proven, to developmental research and manufacture.
These readily controllable methods have served our industry well and are sufficient to keep increasing volume through increasing batch sizes, through increase of equipment size, speed or efficiency, or both, of the process. If there is a need for a settlement or resting time between process stages, it is easy to plan to hold over (work in process or WIP). If product forms differ or involve a mix and match of ingredients, it is easy to split at a batch stage to do so. If a batch is found to be out of specification, it is readily discovered by holding in quarantine and testing at each stage prior to retention or rejection. It is a process to which it is easy to apply protocols—for materials, for loading, for cleaning, for holding, for quality checking, etc. The individual automation of stages allows the overall process be maintained without rocking any throughput or quality parameters. If this is all so simple and easy, why consider something as radical as changing the well tried and tested method? After all, consider the common saying, “If it ain’t broke, don’t fix it.”
Shortages Illustrate Need for Change
According to the wisdom of Henry Ford:
Time waste differs from material waste in that there can be no salvage. The easiest of all wastes and the hardest to correct is the waste of time, because wasted time does not litter the floor like wasted material.
shortages are highlighting inefficiencies in the pharmaceutical industry’s traditional process. This is borne out by both aging facilities and inflexibility in the production methodologies used. A recognized key issue with the prevalent production method of batch processing is that—certainly at larger throughputs—it is a lengthy process that retains large volumes of partially developed product within the process and can be significantly wasteful if an ingredient or process step is found to be wanting. The training of staff in lean techniques, while locally useful within product stages, does not significantly address the fundamental issue of maintaining product quality with assured and timely output volume, nor overcome any of the other failings of large scale batch manufacture—large cumbersome equipment requiring primarily manual cleaning, thus making changeovers slow and costly and large inflexible, costly facilities reflecting the process method, etc. With small-scale batch processing or, depending on the process, intermediate-scale manufacture, this is not such an issue, but with the larger scale processing of oral solid dose products, for example, this can be both wasteful and highly inefficient.
Due to the prevalence of significant drug shortages, the use of fundamentally inefficient and inflexible systems illustrates the need to move to less wasteful and more flexible manufacturing methods.
Inevitably, making this change requires considerable confidence for an industry founded on product quality and assurance, where mistakes are both costly and, more significantly, affect patient health. Inevitably, this also means that the industry is reluctant to change from long-proven processes and entrenched cultures and strategies. Inevitably, during a period of recession, companies only invest in new processes if risks and issues can be resolved satisfactorily. Inevitably, research and development is, therefore, focused on development of new products, fixing the existing system or transferring old products to new markets using tried and tested methods without the need to change the vocabulary.
The lack of earlier investment and take up of continuous processing has meant that at larger scales, and for multiple products, the plant layouts for batch manufactured product have become highly cellular, reflecting and at least adopting the techniques of Six Sigma and lean-focused, automotive industry-based lower volume production cells. Newer plants have at least started to recognize the use of smaller scale multiples of quasi-modular equipment supplied by automated retrieval and supply systems (ASRSs) linked with sealed dispensing systems, enclosed automated feeds and semirobotic operation and transfer provides for flexibility in use. This still results in plants and operations physically large and inflexible, costly to build and significantly reliant on many operators following multiple and complex SOPs —both, arguably, at risk to themselves and the product. This issue is being exacerbated by high volume multiproduct facilities currently planned and under construction that are not using continuous processes—missing the potential benefits of continuous processing realized in other industries.
Moving to Continuous Requires Understanding, Planning
Continuous processing is not necessarily a panacea that will cure all the ills of the industry in one fell swoop through pouring raw materials in at one end and retrieving desired product in its appropriate packaging form at the other with real-time release documentation in hand. Neither can it be implemented for all manufacturing stages for every product without investment in research. Issues around actual required volume, quality of raw product, homogeneity of mix, residence time distribution (RTD), surety and characterization of intermediate, continuity of throughput relative to chemical or biological interaction and resting times are all very specific to individual product. That product must also be assured through proof of quality and validation in a way that is acceptable to the regulators and, more importantly, without risk to the consumer, i.e., the patient—this last respect more so than in any other industry. This, in many ways, underlines reluctance for change.
For this industry, unlike others, product verification has to be controlled and risks managed to an extremely high degree. Each product has variances in material and mass balance that affects throughput parameters along with differing expectations of residence times to allow chemical interactions between stages, and so there is a criticality of equipment response that is needed in continuous processing not present in batch production. The need to understand the process stages and interactions relative to volume throughput is significantly more acute as each element of the process needs to be matched to the volume demand of the output—not a prerequisite for batch operation where stages and WIP act as protectors to what otherwise may be seen as a wasteful process.
Quality by design (QbD) addressing a drug’s critical quality attributes (CQAs) from the outset, improving understanding of flow chemistry with recognition of the need for developments in conversion chemistry, modelling and simulation and the acceptance of risk assessment methodologies, provides routes to the development of fitness for manufacture from the earliest stage of product development. This suits a path to the adoption of continuous processing. Linked with current developments of in-process characterization available through process analytical technology (PAT) and opportunities for feedback/feedforward, as opposed to steady state-based process control systems, the level of surety required for appeasing regulatory expectations does not exist for continuous manufacture’s introduction. The adoption of these checks and system/process balances, however, is still in an early stage of development.
Commercially, continuous processing assuredly places demands on a manufacturer’s procurement options as well as on equipment manufacturers. No pharmaceutical company likes to be seen tied to one manufacturer, and the purchase of a fully integrated line from one source—unless the IP is held by the purchaser—tends to mean that there is little appetite for purchasing something on which warranties and service agreements tie to one option. Likewise, recession-hit equipment manufacturers will tend to avoid putting money into researching something for which there is a dubious market.
Left to its own devices, continuous processing will unavoidably be a slow build to anything like a totally integrated line set. That is, unless companies follow the example of Novartis—which partnered with MIT on a continuous processing research initiative—and invest in technologies open marketed sufficiently for manufacturers to take the risk, allowing them to see the value in investment.
Operational Challenges, Threats and Concerns
Culture: Changing processes often means changing systems; this is often considered a threat rather than an opportunity. The challenge with changing to continuous processing—and in many ways the biggest obstacle—is that for existing products it means reregistration, both a costly and time consuming business not considered acceptable without the availability of proven systems when the time to market is restrictive. This means that refiling is avoided and a facsimile of the previous process is adopted, further restricting opportunities for new options for a significant time period thanks to ROI/amortization considerations.
Supply Chain: Changes based on an end-to-end paradigm will challenge current material procurement methodologies often founded on acceptance of the cheapest rather than the best value. A focus on raw material suppliers to the pharmaceutical industry is long overdue.
The introduction of fully integrated through flow continuous processing systems will challenge current facility feed systems founded, in most cases, on pallet-based nonautomated warehousing that dispenses with the retention of large stocks of, again, pallet-based, WIP and quarantined product. Opportunities for improved supplier control upstream, facility size and flexibility will lead to the potential for more diverse, decentralized operational models that include location at point of need and, downstream, greater response to the local regulatory and patient environment with more personalized patient-focused delivery models.
The Path Forward
Size reduction in manufacturing equipment based on continuous processing, with a move to matrix- and cassette-based rather than pure modular design solutions, will provide new opportunities for plant packaging. The disintegration of the static factory model with resources centered on one or two locations feeding a global market has, to an extent, already begun but is still based on a few API facilities feeding secondary factories with local packaging operations. It is increasingly recognized that, with global energy costs increasing, the unsustainable transport of lightly packaged drugs is often the equivalent of transporting high quantities of air.
The availability of small-scale continuous manufacturing facilities capable of being colocated with the area of drug concern or shortage will change the dynamic of material deliveries based on prepackaged, preapproved, and precontained raw materials to location.
Drug designs focused on personalization on one hand and controllable flow for large-scale operation on the other will inevitably change the pharmaceutical landscape.
The benefits of continuous processing are without dispute. The path to introduction, however, is less stable and dependent on conviction. Does the industry need continuous processing? Continuous manufacture is a natural progression in the technology of production and, as techniques develop and systems improve, take up is inevitable.
[Author’s Note: This is the first of two papers on continuous processing. This paper sets the scene and considers philosophical, social and supply chain issues; the second reports on the perceived current status of the developing art of continuous processing related to drug product with some thoughts on the way forward.]
- Daemmrich, A. and Bowden, M.E. Emergence of Pharmaceutical Science and Industry: 1830-1930. Chemical Engineering News 83: tinyurl.com/oet66cb.
About the Author
Robert Bowen is director of Facilities Integration, a consultancy specializing in master planning, concept design and design development. He is a practicing architect with considerable experience in the design of complex specialist facilities from research and development, CT and API through biopharmaceutical manufacture to oral solid dose, fill/finish and warehousing.