www.pharmpro.com The Path Towards Continuous Processing Featured In: Manufacturing & Capabilities | Pharm Pro Magazine | Newsletter | FDA By Girish Malhotra, PE, President, EPCOT International Monday, August 2, 2010 In recent years, there has been discussion of value manufacturing technology innovations (batch vs. continuous manufacture, PAT, QBA, QBD) for active pharmaceutical ingredients (API) and formulations. In addition, the branded pharmaceutical companies are changing the playing field in order to bolster their revenues and profits. This combination will impact the global pharmaceutical business. I am presenting my observations of the circumstances that can improve manufacturing technologies, revenues and profits. They are different for ethical (branded) and generic drugs. Market Conditions Generics are continuously challenging brand company revenues through patent challenges. In addition, due to patent expiration in the next few years, branded pharmaceuticals are projected to loose about $80 billion in revenue. With the loss of revenue due to dwindling patent protected products, drying pipelines and uncertainties of new blockbuster drugs (bio-tech/small molecule), the brand companies have started to reduce drug prices in high growth markets (the developing countries) (1,2,3). They are also selectively establishing relationships (4,5,6,7,8,9) with companies in developing countries to have market presence. It is their hope that the larger market size will replenish some of their lost revenue and profits. The current practice of batch manufacturing will assist, but it will take longer time compared to using continuous processes (10). With continuous processes, in addition to regaining revenue quicker, they will generate higher profits, produce higher quality products and simplify the total business process. API suppliers, formulators, pharmaceutical companies and customers will benefit from better manufacturing technologies. Brand companies, in order to regain their lost revenue, have also started to increase drug prices in the countries where they can (11). This can have negative connotations and can also create opportunities for generics in the brand company’s most profitable markets. Manufacturing Technologies Ethical (brand) companies resort to batch processes due to the following boundary conditions. 1. Batch processes are easier and simpler to develop, scale-up and commercialize. 2. Pressure of first to market lowers the possibility of using a continuous process. 3. Since the success of a molecule is not known, batch process is the logical choice. 4. Patent protection negates the need for technology innovation. Due to the factors listed above and limited patent life after approval; the effort needed to comply with the necessary regulations and convert a batch to a continuous process has not been a branded drug company’s priority. However, the market conditions might change this in the near future. In addition, my conjecture is that due to lack of competition and the ability of API manufacturers and drug formulators to achieve their profit margins, there is no push to incorporate "quality by design (QBD)" methods in the manufacturing methods. Chemists and chemical engineers are taught to develop, design and implement QBD processes
but since the pharmaceutical manufacturing culture is "quality by analysis (QBA)", it prevails and they have to conform. After the patents have expired, any company can produce API molecules after appropriate approval. They can use a continuous process, if the chemistry/volume warrants it. Under the current business conditions generic companies have higher possibilities of designing processes that follow the QBD regimen and using continuous processes for their manufacturing versus brand companies. We need to understand why the pharmaceutical industry or the technology innovators have not adopted continuous processes or have had difficulty in improving their current batch processes. One reason could be due to significant price differential between the average wholesale price and the cost of APIs and formulated drugs (12). This difference gives them sufficient monies for the development of new molecules and shareholder profits. Compared to the gains due to price differential, financial benefits due to manufacturing technology innovation are very low, making such investments unacceptable. In addition, since the generics take over the business after patent expiration, the brand companies have no incentive to invest in the products they will not produce. Contract API manufacturers and formulators have sufficient financial incentive for technology innovation but have not made any progress. The industry’s hesitation to introduce innovation and the reasons are enumerated in the U.S. FDA’s PAT Guidance for Industry (13). Since the brand companies are beginning to encroach on the "pharmerging" markets, we could see generic companies, who also produce API’s, in order to keep their markets, incorporating continuous processes, as they would not want to loose their market share on their home turf. Continuous processing through lower costs would allow them to retain/increase their market share and margins. This is a challenge they have to take. I foresee that we will see generic companies incorporating continuous processing as part of their manufacturing practices. Adoption of continuous processes will demand implementation of QBD methods for the manufacturing technologies. Review of APIs Market size and drug dosage determine the volume of the API (12). Process chemistry will determine the feasibility of a continuous process. I am using four APIs that have varying yearly needs to illustrate the viability of continuous processing. Some of these products can be produced using standard commercially available equipment that is used for the production of various fine and specialty chemicals. For the lower volume products, modular plants could be used to have a continuous process. Such methodologies have not been considered. Innovative use of equipment would be needed. Creativity and imagination would be valuable. In each of the processes, it is expected that we have the necessary knowledge and understanding of the chemistry, kinetics, physical properties and their interaction and unit operations to design an economic commercial plant that will produce the desired product. Metformin hydrochloride (Glucophage) Metformin hydrochloride is used for Type II diabetes. Table 1 is an estimate of the global metformin market. It is based on every diabetic patient taking 1000 mg of Metformin every day of the year.
Table 1: Yearly Global Metformin Need
If only two thirds of the global diabetic population (285 million) could afford the medicine and Metformin was the drug of choice, then about 69,000 metric tons per year (about 152 million pounds per year) of metformin hydrochloride (API) would be needed. This is a significantly large quantity of a specialty chemical that has a disease curing value. At this volume, it should be produced by a continuous process. At 2% growth in diabetic population, the global demand for metformin in the next 10 years would increase to about 85,000 MT per year. The chemistry of metformin hydrochloride (biguanide class), Figure 1, is simple and straightforward. U.S. Patents 3,174,901 and 4,017,539 and other patents suggest different methods of manufacture of biguanide compounds. Any chemist/chemical engineer familiar with such chemistries can easily produce this product. Currently the chemistry is practiced using solvent in more than a one step process. Reported yield of such processes is about 85-90%. It is possible to improve these processes and their yield.
Figure 1: Metformin Chemistry Dimethyl biguanide hydrochloride chemistry is simple and can be produced at 95%+ yield in a single step followed by continuous crystallization and drying. The stoichiometry will have to be controlled to minimize product impurities. Subsequent crystallization will result in the desired quality product. Properly designed unit operations with well-controlled operating parameters will deliver high yield and quality. A continuous process will have much smaller carbon footprint and can increase profits by more than 10% relative to a batch process. Similar single step chemistries are being practiced to produce fine chemicals of 99+% purity. Process design and operation has to be flawless because any deviation outside the permissible operating parameters will result in significant financial losses. Process control technologies exist to deliver such performance. According to the U.S. FDA’s DMF filings there are 18 active global producers. The average plant capacity of each producer would be about 8.5 million pounds per year. For this throughput, based on my experiences with batch plants, multiple reactors and dryers would be required. Meeting cGMP requirements could be a challenge. Logistics and paperwork would be complex and prone to errors. I am surprised that metformin, being a high volume product like ibuprofen and acetaminophen, is not being produced by any big specialty chemical company. Some of them practice similar technologies. Perhaps they are afraid of the regulatory hurdles. A continuous plant operating at 3000 pounds per hour can produce about 21 million pounds of metformin hydrochloride per year. At this capacity about eight plants would be needed to meet the global demand. Such a plant can be easily debottlenecked to increase the capacity. These plants will have higher yield and lower cost compared to the existing plants i.e. higher profits. Since they will be operating continuously, cGMP requirements would be easier to implement. Based on cost estimates using methods that are taught in any chemical engineering curriculum, the factory-selling price of a 500 milligram tablet after 50% profit would be approximately one cent per tablet. Sixty tablets can be bought at Wal-Mart for $4.00 for a month supply. There is sufficient profit for all involved. Bupropion hydrochloride (Wellbutrin) This antidepressant has global sales of about $1.5 billion per year. Based on 150 mg per day dose (assumption) and using an approximate global wholesale price ($0.40 per tablet), the total global API need for the market would be about 1.2 million pounds per year. Coincidentally, there are 14 plants that supply the global needs, i.e., each plant is producing about 86,000 pounds per year. These would be batch operations. In operating terms, each plant has to go through rigorous cleaning before and after each batch and between different products. Basically, companies have created their own scheduling and operating
challenges that can be alleviated by having a continuous manufacturing process. At 200 pounds per hour operating rate, a single plant operating 24/7 can produce sufficient API to meet global demand. Synthesis of this API is fairly simple and is a routine organic chemistry involving bromination of mchloropropiophenone, subsequent amination with t-butyl amine, reaction with hydrochloric acid and crystallization. Chemistry (Figure 2) and processes mentioned in US patents (15,16,17) can be optimized as one scales-up and commercializes.
Figure 2: Bupropion Hydrochloride Chemistry The process can be an all-liquid process that will have complete control of the process and subsequent crystallization. Since the API volume is moderate for conventional equipment, this can be an excellent platform for a modular plant. We do not think about or are taught about modular plants in our engineering curriculum, as there has been no need. Maybe their time has come! The next two APIs, Omeprazole and Modafinil, are actives to treat two different ailments. Omeprazole is a proton pump inhibitor used in the treatment of dyspepsia, peptic ulcer disease and gastroesophageal reflux disease. Modafinil is used for the treatment of narcolepsy, shift work sleep disorder, and excessive daytime sleepiness associated with obstructive sleep apnea. Knowledge of processes should be shared and used for process simplification. In the chemistries of these two products the oxidation step is same. Chemistries illustrated in figures 3 and 4 show that even though two different oxidation agents are used, any other agents that will result in an economic process could be used. Their selection will depend on creativity of the chemist and chemical engineer involved in scale-up and commercialization of the process (18,19,20). Omeprazole The global market for omeprazole is about $3 billion. At a 20 milligram dosage, and an average global wholesale price of about $0.08 per tablet, the annual requirement for omeprazole API is estimated at about 1.65 million pounds per year. Based on about 20 plants worldwide, it is safe to assume that each plant is producing about 82,500 pounds per year. Since the chemistry is simple enough, it is feasible to have a continuous process for Omeprazole. One plant operating 24/7 at about 250 pounds per hour rate can more than meet the global demand. A single plant operating five days per week at about 350 pounds per hour rate could also meet the global demand. Nexium, which is the S enantiomer of omeprazole, could be part of the same plant to meet the global demand.
Figure 3: Omeprazole Chemistry Omeprazole and Nexium chemistry is illustrated in Figure 3. A properly designed chemical process will deliver the desired product. We need to consider the cost and method that will result in an economic and simple process. Modafinil Modafinil synthesis is typical of many fine/specialty chemicals. With its billion-dollar annual sales, it is considered a "blockbuster" drug. Based on its dosage, average wholesale price (12) and sales one can calculate annual volume of API needed to fulfill the market demand. Global demand for Modafinil (API) would be about 25,000 Kg per year. Based on FDA’s DMF files, 11 plants are producing modafinil globally, i.e., on average each plant is producing about 5,000 pounds year. This is definitely being produced by a batch process at these plants. Like omeprazole, modafinil chemistry is simple and straightforward. Physical properties and unit processes can be finessed and simplified to increase productivity of a batch process. Knowledge gained from these improvements can be used to have a single continuous processing plant operating at about 15 pounds per rate operating 24/5 a week for fifty weeks of the year. The production rate is low compared to the rates many of us would consider for a continuous chemical operation. However, a single modular plant can deliver the annual global demand for many APIs. Since we are not tuned to such plants, we never think about it.
Figure 4: Modafinil Chemistry Different products could be campaigned at such a site. If modafinil was to be campaigned, it could be produced continuously at 200 pounds per hour rate operating 24/7 for about two weeks to deliver the annual API needed. An API manufacturer, who has modular plant capability, could be an ideal location to produce various active ingredients. Modular plants could be considered a novelty but the technology and equipment exists to build one. Many have not considered such methods and would be apprehensive because we are not taught to think small. Basic principles of chemistry and chemical engineering would still apply. However, if a batch process is the methodology of choice for modafinil, a single plant can produce the global demand in one campaign of a few months. As I have stated earlier, when moving from a batch to continuous process fundamental chemistry knowledge, ability to manipulate and capitalize on physical properties, reaction mechanism, process operating variables and unit operations would be needed. Breaking down each reaction step and its understanding will assist in simplification of the total process. Again, human ingenuity and creativity would be helpful. Our command of the process will result in simplifying the information that would be needed for regulatory approval and will also be conducive to process innovation that has been lacking in the pharmaceutical industry. The process would be considered a "quality by design" process. Opportunities A latent reason for lack of pharmaceutical manufacturing innovation could be due to the fact no one wants to disturb the demand/supply apple cart. Another reasons could be the daunting task of regulatory reapproval and associated costs or lack of financial justification. Since 2005, generics have caused a perturbation in the global pharmaceutical business. A new business model might be needed as the demand for proven drugs is ever increasing in developing countries. Since we humans want to live forever, there is high demand for the drugs that will extend or assist in extending life. Demand (21) exists. It is possible that the population that cannot afford the drugs might not be included in the demand projections. If it is so, then global drug needs are underestimated. Management guru Dr. C. K. Prahlad said it right "If you build for the poor, the rich can come. If you build it for the rich, the poor can’t come (22)."
If the "medicines of mass need" can be made available at affordable prices through manufacturing innovation, be it a batch or a continuous process, we could see growth of 10+% per year drug sales, i.e., the market could reach about $2.0 trillion dollars in the next ten years or sooner. With the potential for such a high revenue market, it is very possible that companies from the developing countries could capitalize on and satisfy the unmet market need through better technologies. This scenario would have been unthinkable 5-8 years ago, but with the changing landscape anything is possible. The fastest way to innovate would be to move from "chemistry centricity" methods to "process centricity" methods (23). Manufacturing innovation will benefit every pharmaceutical company especially companies who are API producers, formulators, and sell to mass merchandisers, as every one will generate higher profits through larger markets and lower costs. Regulatory bodies and the industry have to find common ground and rules to reduce the time for the re-approval process and associated costs. Until rules of the game for incorporation of process of continuous improvement in pharmaceutical manufacturing are simplified there will be no game. We just have to do the right things to meet the demand of over six billion people on this planet. It is an opportunity for technology innovation, reduction of carbon footprint, sustained profits and serving human needs. Why should we not do it? References 1. Bristol-Myers launches Onglyza. Sanofi-Aventis, Merck and Novartis other players in same category http://www.businessworld.in/bw/2010_04_02_4_Pharma_Cos_Fight_For_Rs_1400Cr_Drug_Mkt.html Accessed April 22, 2010 2. Eisai Cuts Aricept Price, Joining Glaxo, Sanofi Race, Bloomberg.com, April 15, 2010 accessed April 19, 2010 3. Merck’s low-priced diabetes drug might change a few rules Accessed April 19, 2010 4. AstraZeneca Extends Branded Generics Capability With Torrent Agreement, www.astrazeneca.com, Accessed April 30, 2010 5. GSK enters transformational agreement with Aspen to drive growth in emerging markets, www.gsk.com, Accessed April 30, 2010 6. Pfizer Expands Portfolio of Generic Medicines in the U.S. and Europe Through Licensing Agreements with Aurobindo, www.pfizer.com, Accessed April 30, 2010 7. Glaxo extends Asia push with Dong-A alliance, Forbes.com Accessed May 11, 2010 8. Abbott in new deal to target emerging markets, Forbes.com Accessed May 11, 2010 9. Abbott to Buy Piramal Unit for $3.72 Billion, The Wall Street Journal Accessed May 21, 2010 10. Rising Nations Are No Remedy for Big Pharma, The Wall Street Journal Accessed May 24, 2010 11. Drug Makers Raised Prices Sharply in ’09, The Wall Street Journal, Accessed April 21, 2010 12. Malhotra, Girish, Pharmaceutical Costs, Technology Innovation, Opportunities & Reality, Pharmaceutical Processing, February 2010. 13. Guidance for Industry: PAT – A Framework for Innovative Pharmaceutical Development, Manufacturing and Quality Assurance, US FDA, September 2004 14. Diabetes Facts World Diabetes Foundation, Accessed April 19, 2010 15. US patent 3,819,706 16. US patent application 2009/0012328 A1 17. PROCESS FOR THE PREPARATION OF BUPROPION HYDROCHLORIDE, WO2004024674 18. Malhotra, Girish, Alphabet Shuffle - Moving From QbA to QbD. Pharmaceutical Processing, February 2009 pg 12-13 19. US patent 5,386,032 20. US patent 6,268,502 21. IMS Forecasts Global Pharmaceutical Market Growth Accessed April 29, 2010 22. Fox, Justin, The C.K. Prahalad Fortune at the Bottom of My Backpack, Harvard Business Review accessed April 29, 2010 23. Malhotra, Girish, Process Centricity is the Key to Quality by Design, Profitability through Simplicity, April 6, 2010
About the author: Girish Malhotra is the president of EPCOT International (www.epcotint.com). He can be reached at
[email protected] or 216-292-0626
