Concept Article pubs.acs.org/OPRD
Pharmaceutical Roundtable Study Demonstrates the Value of Continuous Manufacturing in the Design of Greener Processes Peter Poechlauer,*,† Juan Colberg,‡ Elizabeth Fisher,§ Michael Jansen,∥ Martin D. Johnson,⊥ Stefan G. Koenig,# Michael Lawler,∇ Thomas Laporte,∇ Julie Manley,○ Benjamin Martin,◆ and Anne O’Kearney-McMullan¶ †
DSM Chemtech Center, Urmonderbaan 22, 6160 MD Geleen, The Netherlands Pfizer Worldwide Research and Development, 445 Eastern Point Road, Groton, Connecticut 06340, United States § Merck and Co, Inc., Rahway, New Jersey 07065, United States ∥ Hoffmann-La Roche AG, Grenzacherstrasse 124, 4070 Basel, Switzerland ⊥ Eli Lilly and Company, Indianapolis, Indiana 46285, United States # Genentech, Inc., A Member of the Roche Group, 1 DNA Way, South San Francisco, California 94080, United States ∇ Bristol-Myers Squibb, 1 Squibb Drive, New Brunswick, New Jersey 08901-1588, United States ○ ACS Green Chemistry Institute, 1155 Sixteenth Street NW, Washington, D.C. 20036, United States ◆ Novartis Pharma AG, Werk St. Johann, CH-4002 Basel, Switzerland ¶ AstraZeneca, Silk Road Business Park, Macclesfield SK10 2NA, U.K. ‡
ABSTRACT: The American Chemical Society (ACS) Green Chemistry Institute (GCI) Pharmaceutical Roundtable conducted a study to elucidate the value of continuous processing, which had been defined as a key research area for green engineering. In the course of defining the business case for continuous processing, individual cases were collected and evaluated to determine specific drivers to implement continuous processing and to find key success factors. The magnitude and timing of effects and the relation to the principles of green chemistry were investigated. drugs. The team’s first task was to agree on a procedure for developing a “business case for continuous manufacturing”. Ideally, this procedure would be broadly applicable to describe the value of other technologies in a similar way.
1. INTRODUCTION In 2005, the American Chemical Society (ACS) Green Chemistry Institute (GCI) and several global pharmaceutical corporations founded the ACS GCI Pharmaceutical Roundtable (hereafter called “the Roundtable”). Currently, the Roundtable consists of 15 corporations. The activities of this consortium reflect the joint belief that the pursuit of green chemistry and engineering is imperative for a sustainable business and environment. After defining “continuous processing” and “process intensification” as “Key Green Engineering Research Areas for Sustainable Manufacturing”,1 the member companies decided to perform a study of the experiences of manufacturers of active pharmaceutical ingredients (APIs) and finished dosage forms in continuous processing.2 As a next step, the Roundtable decided to elucidate the value that member companies had drawn from continuous processing in their efforts to develop greener routes. A subgroup was formed with the following tasks: • develop a business case for continuous manufacturing; • compile examples from multiple member companies to demonstrate the business case for green engineering. The group consisted of 10 representatives from eight companies (AstraZeneca; Bristol-Myers Squibb; Eli Lilly; Hoffmann-La Roche and its U.S. subsidiary, Genentech; Merck; Novartis; Pfizer, and DSM). The pharmaceutical houses represented in this group generate roughly 50% of turnover of the top 20 pharmaceutical companies in originator © XXXX American Chemical Society
2. DEVELOPING THE BUSINESS CASE Scoping and Shaping of the Business Case. A “business case” analyzes a particular scenario concerning the profitability of an investment opportunity. It is used to present and compare the projected financial and strategic impacts of different options for action that, from a business perspective, will always be understood as “investment”. The first task of the group was to define where to look for individual cases and how to express their value as business cases. The group started by defining fields of activities for developers and manufacturers of pharmaceuticals in which techniques of continuous processing are utilized successfully. In these fields, continuous processing had presumably provided some economic or efficiency-related benefit that could be estimated. These activities ranged from the generation of lead structures and their optimization via design of a process used for launch of a new pharmaceutical to the design of secondgeneration manufacturing processes. Benefits in different fields of activity varied in size and timing. Therefore, the group Received: September 4, 2013
A
dx.doi.org/10.1021/op400245s | Org. Process Res. Dev. XXXX, XXX, XXX−XXX
Organic Process Research & Development
Concept Article
Figure 1. Distribution of cases over activities (horizontal) and motives (vertical).
defined categories of typical reasons (motives) why a continuous process was used in a certain activity. With a list of motives and a list of activities at hand, a matrix was set up to categorize potential cases along these two axes. Figure 1 shows the matrix with fields of activities along the horizontal axis and motives along the vertical axis. In the next step, group members committed to finding and describing actual cases in fields of activities of their choice. These were not necessarily representative of all activities in which the group members’ companies were engaged, since not all of those could be shared publicly, but they were intended to be examples of common activities. Figure 1 shows the distribution of the single cases in this matrix. The focus of these examples was mainly on saving materials and waste and on reducing the footprint of operations between late clinical development and launch. Note that some examples fit more than one primary motive, e.g., waste was saved and company goals of emissions reductions, for example, were also met. Describing the Business Case. The next task was to describe individual cases at a suitable level of detail. This task was critical for two reasons: first, the Roundtable is a strictly
precompetitive group, so defining, disclosing, and comparing cases by their financial value was clearly neither intended nor feasible. Second, it was not the monetary value that was of interest in this context but rather the value seen in greening a process of developing or manufacturing a pharmaceutical. Thus, while the term “business case” insinuates financial motives, the group’s mission was to show that continuous processing offers various categories of value that ultimately have an effect on their business figures or that have undisclosed monetary value. The team agreed to describe each case by providing the following information: • title, author, and company; • technical description (e.g., “development of continuous on-demand synthesis of an unstable intermediate”); • motivation to implement (e.g., consistent product quality, avoiding waste); • effort to implement (time, work, hardware); • value and category of value (e.g., gains in safety); • timing to realize value/effect [immediate/midterm/longterm (>3 years)]; • size of effect [small/medium/large (>10%)]. B
dx.doi.org/10.1021/op400245s | Org. Process Res. Dev. XXXX, XXX, XXX−XXX
Organic Process Research & Development
Concept Article
Several companies decided to use already published material3 and reanalyze it using the above criteria. In total, the team described 17 cases from seven different companies. As noted above, these were not a comprehensive list of examples but were intended to be representative of company activities. These documents were made accessible to all team members and subjected to further analysis. In the following, an example of the procedure is given: Eli Lilly and Company reports an example of a business case for continuous processing: The preferred synthetic route for an active pharmaceutical ingredient involved a high-pressure asymmetric hydrogenation at a pressure up to 70 bar to be economical on catalyst. The decision to invest in specialized capacity to perform this chemistry comprised a comparison of a batch plant and a flow plant: • Investing without knowing whether the pharmaceutical would reach the market would represent a significant financial risk; its uncertain peak volume would add uncertainty to the size of the investment. The capital cost for the flow plant was about 10 times less than that for the batch installation. • Building a suitable batch installation (a new hydrogenation bunker containing a 1000 L high-pressure autoclave) would have taken significantly longer than building an equivalent flow plant. • The envisaged batch process would have been rated as a high-risk operation (volume and pressure of hydrogen, rate of hydrogen use and venting), requiring extensive safety considerations; the respective continuous process could be called a low-risk operation. Therefore the company developed the continuous hydrogenation as a practical, safer, quicker, and flexible alternative. In this study, the case considered the following features: • investment savings; • gain in development speed; • lower technical hazard.
Implementing a new process on the pilot or plant scale may be time-consuming if equipment with a long lead time has to be purchased. Many flow plants can be built without such items. Their construction and temporary installation and use can be seen as operational expenditures rather than as capital expenditures. Features of pilot-scale or plant-scale flow devices are easier to predict from experimental laboratory-scale plants, saving time to scale-up.6 Use of the same equipment to produce and deliver both clinical material and commercial products can eliminate the need for scale-up to larger equipment and can reduce the overall development time. 3. Minimisation of Inventory and Establishment/Validation Stocks. Continuous processing enables production that meets precise demand (in contrast, batch processing is determined by the validated batch size/scale). It also enables just-in-time manufacturing. The reduced operating footprint should also enable manufacturing-at-destination (i.e., the drug substance could potentially be manufactured at the drug production facilities) and hence reduce overproduction and long-term storage of active pharmaceutical ingredients prior to formulation. Overproduction generates large stocks of material before the product is even approved (i.e., high working capital costs), and the material has to be discarded if it has a short shelf life or slow sales uptake. In theory (but not well-proven!), a continuous process needs to be validated only when it attains steady-state; hence, this could dramatically reduce the material/ manufacturing burden. Chemistry/Process Drivers. 1. Desire or Need To Operate a Reaction or Unit Operation at a Temperature Either Below −40 °C or Above +200 °C. Changing the temperature of a whole installation is slower and less energyefficient than changing only the temperature of chemicals as they react by taking them into parts of the installation kept at the required temperature. Countercurrent techniques (preheating or precooling) generate further energy savings. Furthermore, below −40 °C or above +200 °C is outside the operating limits of most standard batch reactors in pharmaceutical manufacturing facilities.7,8 2. A Reaction Yields an Intermediate or Product That Is Unstable under Conditions of Reaction or Workup. Unstable intermediates need to be processed within a short time under well-defined conditions to avoid losses and byproduct formation and make the process more robust.9,10 Continuous processing allows: • fast mixing to minimize degradation of labile substrates when exposed to reaction conditions; • strict control of timing and temperature, allowing fast reactions to be performed in series with unstable intermediates such as lithiation and coupling; • quicker operation, smaller temperature gradients, and better control of the residence time of a product under continuous-workup conditions. 3. Demanding Separations. Continuous-flow equipment allows for time-efficient multiple (countercurrent) extractions on a large scale (in case of unfavorable extraction coefficients). This assists (or even enables) recovery of highly water-soluble products. Safety. 1. Reaction is Very Exothermic. Exothermic reactions are frequently fast and consequently release their high reaction energy within a short period of time. Efficient control requires proficient heat removal over short distances through large areas; both can be provided by continuous-flow
3. ANALYZING CASES: DRIVERS TO IMPLEMENT CONTINUOUS MANUFACTURING The case studies were analyzed for elements they had in common. Many of these elements turned out to be specific features of chemical reactions, unit operations, or chemicals themselves that made continuous processing especially attractive. Such features are termed “drivers” to implement a continuous process option. Most examples contained more than one driver. In total, 12 drivers were identified. These belong to one of three groups: • logistics/quality; • chemistry/process drivers; • safety. In the following sections, a short description of each driver is given. Logistics/Quality. 1. Insufficient Throughput. A reaction operated in batch mode may suffer from insufficient throughput for various reasons: slow dosing, low temperature, high dilution, etc. Continuous processing can speed up the most time-consuming step.4,5 This may also speed up early material deliveries, enabling the route of first synthesis (the “medchem” route) to be scaled up safely. 2. Speed of Implementation and “Right First Time” Performance through Reduction of Lot-to-Lot Variation. C
dx.doi.org/10.1021/op400245s | Org. Process Res. Dev. XXXX, XXX, XXX−XXX
Organic Process Research & Development
Concept Article
Figure 2. Distribution of drivers over the reported individual cases.
Figure 3. Distribution of crucial elements of implementation.
equipment.11,12 Also, the heat to be removed is generated at a constant steady-state rate 24 h per day, 7 days per week, and therefore, the maximum power output is less than in an equally productive batch operation. The volume of chemicals engaged in a failure scenario is always a tiny fraction of a whole batch, while in a batch procedure it is frequently the whole batch. 2. Desire/Need To Run Reactions at High Pressure or with Hazardous Gases (CO, H2, O2). Large-volume pressure reactors
are massive, react slowly, and while in operation contain large quantities of hazardous reagents, causing severe problems in case of leakage. Flow equipment is smaller and has a smaller hold-up, and high pressure is less of a problem because it exerts smaller forces in smaller equipment (for an example, see refs 13−15). 3. Reaction Benefits from Operation with No Vapour Space (100% Liquid-Filled). At high pressure, the fraction of D
dx.doi.org/10.1021/op400245s | Org. Process Res. Dev. XXXX, XXX, XXX−XXX
Organic Process Research & Development
Concept Article
gas volume in a completely liquid-filled system like a flow reactor is very small, so a dangerous “head space” is avoided. This is crucial to avoid evaporation and condensation of lowboiling hazardous materials (e.g., hydrazoic acid16). 4. Reagents or Products Are Highly Potent or Cytotoxic. Demanding cleaning requirements caused by cytotoxic reaction components may motivate engaging disposable equipment. Many components of simple flow plants may be available as single-use or essentially disposable supplies or large-scale laboratory glassware that is portable and inexpensive. Many of these highly potent products will be needed in small amounts only. Therefore, manufacturing sites will need the ability to manufacture small batches and the flexibility to run small campaigns (for an example, see ref 17). There will be increasing cost-of-goods pressures driven by more complex molecules and market price pressures. This will increase the pressure to deliver processes with innovative and cost-efficient chemistry and technology (for an example, see ref 18). The next generation of API processes, equipment, and manufacturing facilities must be smart, effective, and value-oriented. The matrix in Figure 2 shows the distribution of drivers over the cases. Some drivers appear frequently and others more rarely, but all are crucially important for the respective case. This is certainly a nonrandom set of cases; the most-cited drivers in the collected examples could be linked to logistics rather than safety, chemistry, or process technology − operating processes in continuous-flow mode appears to fit well with the logistical requirements for development and manufacturing of active pharmaceutical ingredients. These requirements are, most obviously, access to a quick supply of high-quality material “on demand” while keeping inventories low.
term (>3 years). Each of the 17 cases delivered three estimates. In seven situations, effects were categorized as “unpredictable in size” and were not counted. Accordingly, 44 data points were considered in total. Figure 4 depicts the size and timing of effects of the reported cases.
Figure 4. Size of effects and timing of results.
Remarkably, most of the reported cases (which, as we repeat, are not random samples) show large effects with either immediate or midterm results. This appears to reflect the trend to develop this technology within or in close collaboration with business projects (1−3 years overall project time), and continuous processing obviously keeps its promises.
6. LINKING THE BUSINESS CASE TO THE PRINCIPLES OF GREEN CHEMISTRY Finally, the group members turned their attention to the correlation of continuous operation of a process to its greenness as defined by the 12 principles of green chemistry (Table 1).19 Many case descriptions make explicit mention of
4. CRUCIAL ELEMENTS OF IMPLEMENTATION The reported cases are the result of successful development and application of certain competencies, skills, and know-how. In acquiring these, the single project groups had to learn “what works and what does not” and “what to keep an eye on”. Many groups experienced similar pitfalls related to equipment failure or misjudgment of the effects of deviations. Consequently, they developed routines and “best practices” to accomplish certain tasks. While these practices and solutions differ among the project groups, certain topics appear systematically in the reported cases of successful implementations. Conversely, we may assume that disregard of these topics may cause problems. Figure 3 depicts the distribution of some crucial elements of implementation over the reported cases. Clearly “mixing” and “keeping process flows at defined conditions” appear as critical elements in most cases. This is not surprising, since nearly all of the drivers to implement continuous processing listed in the previous section depend on having excellent control of mixing and stoichiometry/residence time, and batch processing frequently can have scale-up issues in these areas. Consideration of these elements is key to realizing the desired benefits of continuous processing.
Table 1. The 12 Principles of Green Chemistry 1 2 3 4 5 6 7 8 9 10 11 12
prevent waste design safer chemicals and products design less hazardous chemical syntheses use renewable feedstocks use catalysts, not stoichiometric reagents avoid chemical derivatives maximize atom economy use safer solvents and reaction conditions increase energy efficiency design chemicals and products to degrade after use analyze in real time to prevent pollution minimize the potential for accidents
advantages related to at least one green principle (e.g., less hazardous chemical synthesis or use of a catalytic process), but a link of these principles with continuous processing is not a priori obvious. In several cases, continuous processing provides an opportunity for a trade-off: for example, hazardous materials are produced and consumed in a safe way to shorten a route and improve its atom efficiency. Thus, the group investigated the relationships between a “green process” as defined by the 12 principles of green chemistry and the major benefits
5. ANALYSIS OF THE CASES: SIZE AND TIMING OF EFFECTS The group then analysed the semiquantitative data on the size and timing of the effects: for each case, the magnitude of value generated was categorized as small/medium/large (>10%) and the timing of effects as immediate/midterm (1−3 years)/longE
dx.doi.org/10.1021/op400245s | Org. Process Res. Dev. XXXX, XXX, XXX−XXX
Organic Process Research & Development
Concept Article
runs successfully as designed and is scalable. This might also involve investment in laboratory equipment to prove that the proposed process works and retains its advantages over the batch equivalent at all scales. There will also be an initiative to search for a new route that runs best in a traditional batch equipment. “Special equipment” is viewed as a risk because inside a company comparatively few people are experienced in operating it and even fewer are experienced in troubleshooting. Next, unanimous agreement among high-level leadership within a corporation, which includes Development, Manufacturing, Quality, Health/Safety/Environmental, Internal Regulatory, and other business units, is needed to implement the continuous process. Even though the business case exists, funding must be available for manufacturing at the right time in case investment is required. Furthermore, if the continuous process is planned for internal commercial manufacture after development and validation by an external partner, then such a partner must exist who can run the earlier GMP campaigns for earlier clinical trials to produce representative material using the continuous process technology. Finally, attrition rates are high for new molecules in pharmaceutical development: most drug developments have to be terminated before launch. This collection of nontechnical factors contributes to every successful introduction of a new technology. Thus, they are a clear sign of a company’s commitment to improve on its operations to supply new medicines by greener processes.
associated with continuous processing. These major benefits were seen as: • investment savings; • yield/quality improvement; • safety; • speed. Cases were analysed for these categories of benefits as exemplified in section 2. An Euler diagram was used to visualize these relationships as a final analysis. It consists of a number of partially overlapping circles. According to the theory of sets, elements within a circle meet certain criteria. Regions of overlaps of two or more circles contain elements meeting all criteria of the overlapping circles. The discussed example (depicted as example 9) meets the criteria “investment savings”, “speed”, and “safety”. It describes a catalytic process with high atom efficiency and is therefore “green”. An Euler diagram was set up, consisting of a circle with “green” as a criterion at the centre and four partially overlapping circles with investment savings, yield/quality, safety, and speed as criteria. All of the business cases were distributed over this diagram (Figure 5). We found only few cases outside the green circle.
■
AUTHOR INFORMATION
Corresponding Author
*E-mail:
[email protected]. Notes
The authors declare no competing financial interest.
■
REFERENCES
(1) Jiménez-González, C.; Poechlauer, P.; Broxterman, Q. B.; Yang, B.-S.; am Ende, D.; Baird, J.; Bertsch, C.; Hannah, R. E.; Dell’Orco, P.; Noorman, H.; Yee, S.; Reintjens, R.; Wells, A.; Massonneau, V.; Manley, J. Org. Process Res. Dev. 2011, 15, 900. (2) Poechlauer, P.; Manley, J.; Broxterman, Q. B.; Gregertsen, B.; Ridemark, M. Org. Process Res. Dev. 2012, 16, 1586. (3) Johnson, M. D.; May, S. A.; Calvin, J. R.; Remacle, J.; Stout, J. R.; Diseroad, W. D.; Zaborenko, N.; Haeberle, B. D.; Sun, W.-M.; Miller, M. T.; Brennan, J. Org. Process Res. Dev. 2012, 16, 1017. (4) Kappe, O. C.; Pieber, B. Green Chem. 2013, 15, 320. (5) Seeberger, P. H.; Levesque, F. Org. Lett. 2011, 13, 5008. (6) Lawton, S.; Steele, G.; Shering, P.; Zhao, L.; Laird, I.; Ni, X.-W. Org. Process Res. Dev. 2009, 13, 1357. (7) Kappe, O. C.; Glasnov, T. N.; Razzaq, T. Eur. J. Org. Chem. 2009, 1321. (8) Knell, A.; Monti, D.; Baiker, A. Catal. Lett. 1995, 31, 197. (9) Liu, B.; Fan, Y.; Lv, X.; Liu, X.; Yang, Y.; Jia, L. Org. Process Res. Dev. 2013, 17, 133. (10) Roberge, D. M.; Ducry, L. Org. Process Res. Dev. 2008, 12, 163. (11) Dyer, U. C.; Henderson, D. A.; Mitchell, M. B.; Tiffin, P. D. Org. Process Res. Dev. 2002, 6, 311. (12) Nieuwland, P. J.; Koch, K.; van Harskamp, N.; Wehrens, R.; van Hest, J. C. M.; Rutjes, F. P. J. T. Chem.Asian J. 2010, 5, 799. (13) May, S. A.; Johnson, M. D.; Braden, T. M.; Calvin, J. R.; Haeberle, B. D.; Jines, A. R.; Miller, R. D.; Plocharczyk, E. F.; Rener, G. A.; Richey, R. N.; Schmid, C. R.; Vaid, R. K.; Yu, H. Org. Process Res. Dev. 2012, 16, 982. (14) Murphy, E. R.; Martinelli, J. R.; Zaborenko, N.; Buchwald, S. L.; Jensen, K. F. Angew. Chem., Int. Ed. 2007, 46, 1734. (15) Stevens, J. G.; Bourne, R. A.; Twigg, M. V.; Poliakoff, M. Angew. Chem., Int. Ed. 2010, 49, 8856.
Figure 5. Euler diagram depicting the distribution of cases.
The final distribution shows that even in cases where “greening” a process might not have been the primary goal, operation in continuous flow mode, in most cases, allows a process to become “greener” according to the 12 principles of green chemistry. Thus, a variety of different goals help to support the business case for continuous processing, and conversely, operating processes in flow mode helps achieve the objectives of management to realise improvements across a span of dimensions.
7. SHOWING THE BENEFITS TO BUSINESS IS JUST THE BEGINNING After a case for a continuous process has been established, many additional hurdles must be overcome in the early stages of development of active pharmaceutical ingredients. Most identified opportunities do not make it to commercial manufacturing, and there are many reasons: After a process has been selected for continuous operation instead of batch operation, the next goal is to prove that the continuous process F
dx.doi.org/10.1021/op400245s | Org. Process Res. Dev. XXXX, XXX, XXX−XXX
Organic Process Research & Development
Concept Article
(16) Kopach, M. E.; Murray, M. M.; Braden, T. M.; Kobierski, M. E.; Williams, O. L. Org. Process Res. Dev. 2009, 13, 152. (17) Bedore, M. W.; Zaborenko, N.; Jensen, K. F.; Jamison, T. F. Org. Process Res. Dev. 2010, 14, 432. (18) White, T. D.; Berglund, K. D.; McClary Groh, J.; Johnson, M. D.; Miller, R. D.; Yates, M. H. Org. Process Res. Dev. 2012, 16, 939. (19) Anastas, P. T.; Warner, J. C. Green Chemistry: Theory and Practice; Oxford University Press: New York, 1998.
G
dx.doi.org/10.1021/op400245s | Org. Process Res. Dev. XXXX, XXX, XXX−XXX
