Development of Continuous Pharmaceutical Production Processes Supported by Process Systems Engineering Methods And Tools

Keywords::

• continuous manufacturing
• microreactors
• pharmaceutical production
• process developement
• process understanding

The pharmaceutical industry is undergoing a radical transition towards continuous production processes. Systematic use of process systems engineering (PSE) methods and tools form the key to achieve this transition in a structured and efficient way.

Why continuous production?

Several drivers explain the increased focus on continuous production of small-molecule drug substances. Publication of the Process Analytical Technology guidance by the US FDA in 2004 is certainly important [101]; by publishing that document regulatory authorities have clearly encouraged introduction of new production technologies, for example, more online measurements and automatic control, as a means to reduce production costs and improve safety and product quality. Compared with a batch process, where product quality is, in general, measured offline causing long time delays between sampling and corrective actions, a continuous process is more suitable to apply automatic control, and to implement Process Analytical Technology concepts such as ‘real-time release’.

The core problem for traditional pharmaceutical companies, however, is the obvious lack of sufficient R&D productivity: R&D expenditure is increasing all the time, while generating significantly fewer new molecules than 20 years ago [102]. In addition, active pharmaceutical ingredients for the most important diseases were discovered a long time ago and medical treatments have become more individualized. As a consequence, there is a general lack of new ‘blockbuster’ drugs, and traditional research-based pharmaceutical companies now also produce generic drugs, meaning that they have a new focus on decreasing manufacturing costs and establishing flexible production plants – continuous production also comes into the picture here.

Moving from batch towards continuous-pharmaceutical manufacturing is one of the key strategies for improving safety and product quality, while decreasing waste generation and manufacturing costs [1,2]. Batch production, despite its flexibility and multipurpose nature, is highly inefficient: for example, large amounts of solvents are used and solvent recovery is not always carried out [3]. In conclusion, by adopting continuous production (combined with batch processes whenever relevant), several opportunities to improve the efficiency of pharmaceutical processes arise.

What are the challenges?

Continuous processes are increasingly important in the production of small-molecule drug substances, which are typically obtained via organic synthesis. A similar trend towards more continuous production can be seen in formulation operations [4].

Continuous pharmaceutical plants have been envisaged as mini-plants whose constituents are a limited number of standardized, well-characterized and easily connected modules, each one carrying out one or more basic unit operations [5]. Interestingly, these plants need to accomplish two objectives, which, traditionally, have been regarded as contradictory: serving a dedicated equipment function, but also having flexible capacity and operation, as well as adaptability to new chemistries. Microreactors have been intensively studied in recent years for continuous production, featuring excellent mass and heat transfer but still showing some limitation with respect to handling solids [6,7]. Microseparations have also been developed, enabling solvent exchange and multistep chemical synthesis at microscale [8]. Responding to changes in demand can be achieved by replicating the number of units (i.e., numbering up [scaling out] as opposed to scaling up), which should, theoretically, eliminate the uncertainty involved in the scale-up task typical for batch processes. However, this strategy may lead to prohibitive capital costs, while the advantages of operating at a small scale may not be obvious in every situation. A trade-off may be needed, scaling up from the micro- and meso-scale to a scale offering sufficient throughput while maintaining the performance with respect to heat and mass transfer, and, subsequently, replicating those units in order to meet the required throughput. Clearly, a good process understanding is required to select the most relevant operating scale for a given chemistry and physical system. Furthermore, a scale-up/-out strategy must be found such that the advantages of operating at small scales are not lost at high throughputs.

Continuous pharmaceutical production is still in its infancy and much of the research work has focused on the demonstration of mini- and micro-scale operation of individual unit operations [7], some of which have been interconnected in relatively simple multistep processes. Higher reaction yields and separation efficiencies are typically demonstrated at small scales. In many cases, however, there is still a gap in transferring those new technologies to industry. In addition, several important issues remain unresolved:

• Where are the highest benefits found?

• Should we make an effort to enable continuous processes with heterogeneous systems?

• How can a continuous process be designed and implemented?

• Are continuous processes compatible with slow reactions and are they advantageous in this case?

• What is the cost needed to implement a continuous process?

• What is the profit?

• And, most importantly, when should continuous processes be selected?

It is generally accepted that not all reactions are suited for continuous production: it was estimated that only 50% of the reactions carried out at Lonza (Basel, Switzerland) would benefit from continuous operation due to slow kinetics, and, from these, 63% cannot technically be realized in a microreactor due to presence of a solid phase [9]. However, it is important to realize as well that continuous processes are not limited to microreactors only. Examples of other, often mesoscale, continuous reactor types are the continuous oscillatory baffled reactor [10], the agitated cell reactor [11] and the filter reactor [12].

Process systems engineering

Addressing the aforementioned challenges requires a systematic approach towards design. We are convinced that a PSE approach towards pharmaceutical process development and innovation should be stimulated [13]. A major strength of PSE methods and tools is their potential for structuring, managing and exploiting process knowledge, as well as for directing experimental efforts to generate new knowledge [14]. Furthermore, PSE offers a structured approach to solving problems, and the potential to develop generic solutions that apply across several processes [15,16]. The pharmaceutical industry – multipurpose in nature – indeed requires generic approaches in order to solve multiple problems sharing common features, with the purpose of keeping development costs down. An additional benefit of generic PSE-based approaches is that it should be easier to get approval from regulatory bodies such as the FDA. Regulatory bodies should of course be involved early on in the development phase, whenever new methods or tools need to be tested or implemented.

PSE methods and tools – for example, mechanistic models for representing process knowledge, optimization algorithms and advanced process control methods – have been widely used in the chemical industry for efficient design of new processes and products. Despite the obvious potential of such methods, they have only been sporadically used by the pharmaceutical industry in the past. High pressure to reach the market in due time and the high complexity of pharmaceutical product mixtures have been the main reasons for relying almost exclusively on experimental data to enclose the design space. A major hurdle to be taken, according to us, is to adapt existing PSE methods and tools – often developed in an academic environment – to the inherently complex nature of pharmaceutical products and processes, with very strict quality and regulatory requirements, developed against aggressive timelines.

Of course, PSE has its limitations as well. Specifically for the pharmaceutical industry, major limitations are:

• Considerable time and effort is needed to understand PSE tools;

• Limitation of thermodynamic models (e.g., basis for solvent selection, simulation and optimization) to very simple and small molecules;

• Lack of commercial software with specific unit operations for the pharmaceutical industry;

• Lack of case studies demonstrating that the use of these tools actually brings an economical/environmental advantage.

In this respect, we are convinced that much can be gained to address some of those limitations by establishing an even closer collaboration between academia and industry. Academia, on the one hand, needs this to understand the industry’s limitations, which is necessary in order to develop the scientific basis for providing the right solutions, or to come up with PSE tools that fit the needs of the pharmaceutical industry. The pharmaceutical industry, on the other hand, needs this for knowledge transfer, that is, to learn and understand the strengths and limitations of existing PSE methods and tools (e.g., to allow application of the right tool to the right problem). Undoubtedly, it would also be advantageous with increased collaboration between the major companies, since most of their process-development teams struggle with similar problems. Here, we think the industry should present a few case studies, for example, of products that have already gone off patent years ago, which can be used to develop, test and compare new production technology and new PSE tools. Those case studies should be relatively simple (e.g., only a couple of steps each), such that academia can handle such a case study at relatively small scale against a reasonable cost.

A key question when considering the use of PSE methods and tools during the development of a novel continuous production process for an active pharmaceutical ingredient is when to start using these methods. We are convinced that PSE methods and tools should be used throughout the product lifecycle, to facilitate continuous learning and collection of process/product knowledge. As soon as the first information (i.e., properties and kinetic data) is available from the laboratory on a promising compound with potential to make it to clinical testing, one could, in principle, set up a simple kinetic model and try to find out what the flow-sheet of a continuous production process could look like.

In conclusion, PSE methods and tools can potentially help the pharmaceutical industry to evolve from a situation with ad hoc empirical solutions to a systems approach and generic solutions. Despite the obvious potential, a major effort is still needed in order to convincingly demonstrate the potential of applying PSE in pharmaceutical process development. Academia cannot expect that the pharmaceutical industry will use PSE and its tools until their full potential is demonstrated in challenging and real scenarios, with the benefit quantified on an economic basis. Economic support from industry, close collaboration between academia, industry and regulators, and also good communication between all, must be established in order to understand the limitations of each party and to evolve innovative PSE-based development of lean production processes.

Financial & competing interests disclosure

The authors have no relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript. This includes employment, consultancies, honoraria, stock ownership or options, expert testimony, grants or patents received or pending, or royalties.

No writing assistance was utilized in the production of this manuscript.

Additional information

Funding

The authors have no relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript. This includes employment, consultancies, honoraria, stock ownership or options, expert testimony, grants or patents received or pending, or royalties.