Lately new terminologies e.g. flow chemistry, continuous manufacturing, plug flow reactors, process intensification to name a few have become the new way or the buzz words to develop, solve and/or commercialize active pharmaceutical ingredients (API) and some fine/specialty chemical processes. I am not sure everyone understands them and whether these will create economic and environmentally friendly processes. Based on my experience, names do not create excellent processes but fundamental understanding and their application heat and mass balance, physical and chemical properties, reaction kinetics, thermodynamics and mutual behavior of chemicals used and produced have and going forward will continue to assist. To develop and commercialize any chemical process the chemist and the chemical engineer has to have complete command of the physical properties and their mutual behavior and interaction of chemicals with the processing equipment i.e. the “sociochemicology” (1, 2, 3, 4, 5, 6, 7, 8).
Questions raised in the following analysis could and/or should be asked from the onset for each product and process that is commercialized. My observation is that each process development chemist and chemical engineer has to act as a “Village” (3, 7), spend time and understand the chemistry and chemical engineering of each reaction that is commercialized. Such an effort has and will continue to create optimum processes that are very profitable at active pharmaceutical ingredient stage (API), fine/specialty chemical and are also environmentally friendly. These practices/learnings are also applicable to formulations. I have no vested interest and/or relationship with any nonprofit and or nonprofit organization.
Following is my analysis and observations related “A Continuous Process for Manufacturing Apremilast. Part I: Process Development and Intensification by Utilizing Flow Chemistry Principles (9). There are 23 authors and some of them have left Amgen before the article was published.
Figure 1: Schematic of production of Otzela
The following abstract gives authors perspective. However it raises multiple questions at least for me.
Abstract: Herein, we report the development of an integrated continuous manufacturing (CM) process for the penultimate step in the synthesis of apremilast, the drug substance (DS) of the commercial product Otezla. This development effort was motivated by the desire to create an alternative manufacturing configuration with a significantly smaller footprint and to impart intensification resulting in a more sustainable process. Three primary aspects of the existing batch process had to be addressed to achieve this goal: (1) long reaction time (2) low solubility of the starting materials and intermediates in the primary reaction solvent (THF) 3) extensive postreaction unit operations contributing to significant solvent waste Key features of the intensified CM process include the following: · use of a plug-flow reactor (PFR) to access increased reaction temperatures (130 °C), resulting in a shorter reaction time to reach the target conversion (>18 h in batch to 30 min in flow); · replacement of THF with DMSO to solve solubility issues related to starting materials and reaction intermediates, and · development of a multistage continuous MSMPR (mixed-suspension, mixed-product removal) crystallization upon addition of water as antisolvent to the end-of-reaction stream containing apremilast. This intensified CM process reduced the number of primary unit operations from nine to three (67% reduction). Moreover, it can be executed at commercial scale using a compact manufacturing skid. Part I of this manuscript series highlights the effort to develop the novel process and the corresponding kg-scale demonstration of the optimized process. Part II describes the process characterization and development of a control strategy in detail to ensure process efficiency and robustness of the small-footprint continuous skid. Copyright © 2024 American Chemical Society |
Table 1: Abstract
Questions:
1) Global sales of Otezla were ~2.2 billion dollars in 2023 and are decreasing. U.S. Patent 7,427,638 expires September 2025 and has been extended (10). This would make the product generic and its global sales, dollars, would be an unknown. Most likely they will decrease. Companies other than Amgen would produce the product. Why there was no effort made to improve the chemistry and manufacturing process before the product was commercialized. Since the patent ‘638 does not expire till 2028 my question is “Is any of the information published confidential?” I am not sure about the intent of this paper (9).
2) Based on dosage (2 tablets of 30 mg per day) and the current selling price $6,000 per month, annual API needed to serve the global market would be about 750 kilograms per year. What would be the generic sales is an unknown?
3) This article claims the process to be a “continuous process” and this raises many questions when there is an established definition of “Continuous Manufacturing (11)”. It clearly states continuous usually means operating 24 hours per day, seven days per week with infrequent maintenance shutdowns, such as semi-annual or annual. In hours this would be close to 7,400 hours per year taking about 1000 hours for maintenance.
Based on published sales and selling price of Otzela (12) total production of Otzela API would be about 750 kilograms per year. If it were to be produced using a continuous process (11) its production rate would about 0.1 kilo per hour. This would be uneconomical. Every company will produce their product using batch a process. My best estimate is that production of Otezla has to be a batch process and the product would be campaigned. Claiming that Otzela production is a “Continuous Process” is an outrageous claim when it is produced using a batch process. World would like to have a clear explanation from Amgen for its claim. If a batch process is called a continuous process then it suggests that there is total disregard and mockery of established definitions which have been accepted by American Chemical Society, publisher of this article and American Institute of Chemical Engineers.
Even if the Apremilast (Otezla) yearly production doubles up when the product becomes generic, with the current landscape of API manufacture there will be two options. Fit the process in existing equipment that is most suitable or design a dedicated plant for its production. In the latter case, the equipment will sit idle when not used to produce Otzela. Company will have to decide. Having a dedicated equipment for this product would be an unproductive investment.
In my career I have developed, commercialized, managed fine/specialty chemicals, older cousins of API operating at production rates of about 100 Kilos/per hour to 1,800 kilos per hour operating about 7,200 hours per year. Batch products had annual production that ranged from 100,000 to 400,000 kilos per year. Many of these batch produced products were campaigned in designated equipment. I am not sure about the euphoria or advantage behind calling a batch production a continuous production. I am sure everyone would like to know.
3) Based on number of authors (TWENTY THREE) (9) and their time, my conjecture is Amgen has spent in excess of FIVE million dollars. I wonder what is the return on this investment. Based on the article I am not sure if the process in its current state is ready for commercialization. Additional manpower i.e. investment would be needed for the process that still needs to be developed, tested using alternate equipment and commercialized. If an alternate process is commercialized, there could be significant changes from the original process. That could mean the final production process would need necessary regulatory approval.
I don’t believe Amgen is ready to invest that money when it knows that by late 2028 this product will be a generic. With Amgen’s patent expiring in February 2028 (10) Amgen’s strategic advantage would be anyone’s guess. All this work raises questions about the rationality of this paper (9) and related work. Fundamental question is why Celgene did not do the necessary work before the product was commercialized. What is the rationale and incentive for the current work and its “return on investment”?
4) Why did Celgene/Amgen wait this length of time (four or five years before the patent expiration) and did not do the necessary work to reduce solvent use from the onset of its commercialization? It seems there was no effort to get to “Net Zero” (13).
5) This paper (9) suggests use of alternate equipment. Does Amgen know what that alternate equipment is and has it been tested. My conjecture is that significant work would be needed to prove the viability of the suggested equipment. Would/are the alternate equipment that have been available and traditionally used in fine/specialty chemical industry be considered (1, 3)? It is possible that chemists and engineers at Amgen may not be familiar with such equipment.
6) Some of the raw materials are from the supply houses i.e. Sigma Aldrich, rather than from commercial suppliers. This suggests that Amgen is still depending on high purity raw materials that are high priced vs. commercially available raw materials.
7) I wonder if Amgen personnel had the time to define the solubility and other physical properties and the mutual interaction of the chemicals used, intermediates and produced. This information is critical and should have been developed and used in the development of an optimum process. They could have developed the necessary data like we had/have done for the chemicals we commercialized (1, 2, 3). Yes it is a challenge but the rewards are there as the information can be used to trouble shoot the process. Before the Internet came around suppliers readily provided chemical property data (14).
8) It is very interesting to note that Amgen suggests that the reaction time was reduced to about 30 minutes from 18 hours. Question needs to be addressed how and why this was not addressed before NDA (new drug application) was filed. Same question holds for reducing number of unit operations, long reaction time and low solubility of the starting materials and intermediates in the primary reaction solvent (tetra hydro furan, THF) and extensive postreaction unit operations contributing to significant solvent waste. If Amgen is still using their inefficient process, then what and where is its environmental responsibility. Did developers ever think and/or consider environmental conservation (13)?
9) There is mention of PFR (plug flow reactor) use. Has Amgen used such devices in the manufacture of any of its APIs? Has Amgen considered other equipment that is commercial and is used in the manufacture of fine/specialty chemicals (1, 2, 3, 7)? Are the chemists and chemical engineers at Amgen familiar with use of back mix flow reactors? They have been and are used in the manufacture of fine/specialty chemicals, older cousins of APIs.
10) Lots of “fancy verbiage” has been introduced by the pharmaceutical industry to state that is cut above its older sibling fine/specialty chemical industry (e.g. substoichiometric, process mass intensity etc.). No matter what the pharmaceutical industry claims 2+2 will always be 4, Sun is always going to rise in the East and male of the species is will not be pregnant. Introducing new verbiage is not going to change established facts. Have we forgotten simple verbiage like mass balance, heat of reaction, value of mutual behavior of chemicals and chemical reaction kinetics etc. to develop and simplify the chemical processes. Another term is “flow chemistry”.
11) Searching “flow chemistry” does not give any clear definition of what it means. When chemists and chemical engineers are asked about the “flow chemistry” or “continuous process” (11) they are not able to give differentiating definition of either. What they recite is not any different from the established definitions of “batch process” (15) or “continuous process” (11). If “flow chemistry” is the new way for all of the chemical reactions then the question is what were the chemists and chemical engineers doing/teaching who built the foundation of the chemical and pharmaceutical industry. They created and commercialized very many useful processes and products which included disease curing molecules called API (3). May be my generation and authors of McGaw Hills Chemical Engineering Series (16) are all wondering what does it mean. New verbiage does not mean anything if one does not understand the fundamentals.
12) One very basic question is “if what has been suggested in some of the articles is an improved and efficient process” then why are inefficient processes was commercialized and monies are being invested in equally cumbersome process where specialized equipment might be needed and it might not be used to produce any other product.
Based on reviewing the paper I have serious doubts that any further work would be done to improve and commercialize this chemistry especially if the process needs regulatory approval. If Amgen just publishes the data it created it could be of value for future generations.
COMPARISON OF SIMILAR CHEMISTRIES:
When developing a process it is critical and necessary that the “Village” (1,2,3, 5) be involved from the onset. Literature gives us lots of knowledge for free. Following is a comparison of two chemistries. Reason for the examples is “did Celgene do sufficient literature search to explore prior art?” Stoichiometry USP 7,109,203 (17) and a similar chemistry that has been commercial since 1970 tell us that for Product “X” no effort was made from the onset to have an excellent process. Product “X” operated about 7,200 hours per year for many years. Product from patent ’203 would require significant work prior to its commercialization.
Figure 2: USP 7,109,203 NOVARTIS
Mole Ratio | Patent ‘203 | Product “X” |
Aromatic Amine | 1.0 | 1.0 |
HCl | 14.0 | 2.4 |
NaNO2 | 0.97 | 1.1 |
SO2 | 25.9 | 4.5 |
Solvent | Acetic acid | None |
CuCl2 | 0.50 | 0.068 |
Figure 2: Comparison of two steps of a reaction schemes
A review of the patent ‘203 illustrates that 2-chloro-4-bromobenzensuflfonyl chloride is produced in a single pot and processed further. One can see that the process for Product “X”, similar chemistry, is synthesized in two steps in a very eco-friendly back mix flow reaction process. It is difficult to understand use of acetic acid in ‘203 as a solvent as it will have to be neutralized and takes up productive reactor space. Purpose of the illustration is that proper process can be developed and commercialized if an effort is made.
Again the emphasis is that lab synthesis shows the pathway and it can be complex. Breaking down the reaction for 2-chloro-4-bromobenzensuflfonyl chloride in two steps simplifies the process and directs us to create an ecofriendly process. However, one has to recognize that breaking the reaction in two steps is not feasible for every reaction in the lab. It can only be envisioned and tested by a chemist and/or chemical engineer who is familiar and experienced in creating such processes.
In addition chemists and chemical engineers have be familiar with other process schemes like “back mix flow reactors” (12, 13 ) and what other methods across the industrial fields are available and can be used in the manufacture of products under discussion.
Product “X”’s reaction scheme and similar reactions have been commercially practiced in the fine/specialty chemical industry older cousin of API for more than fifty years (1,2,3,4,5,7,13). Some have been briefly reviewed (5, 7). We have to recognize and acknowledge that we are taught principles of chemistry and chemical engineering but are not trained to create processes that have minimum ecological impact. This comes from “hands on experience” when chemists and chemical engineers who are involved in scale up and process design have the knowledge and the command of every nuance of sociochemicology (6,16)reaction kinetics, solubilities etc. of the chemicals involved and produced.
By calling the same widget by different name is not going to make it a new widget or technology i.e. flow chemistry. It just tells us that we have not paid attention to the history and fundamentals of chemistry, chemical engineering and the resulting technologies that have been practiced at least for the last 80 years.
Naming fundamentals as stated above whose ancestors have been well practiced since mid-nineteen sixties does not make them new. Using an inline static mixer or adding a fluid at the inlet of the pump improves mixing and ensuing reactions is capitalizing on creativity and imagination (1,2,3). One has to admit and recognize that without “flow” of fluids no reaction, batch or continuous, will ever takes place. Thus “Flow Technology” is nothing new, at least in my book.
It is necessary that the “Village” (1,3) be involved from inception and development of the process. Without such an effort we will continue to “after thought” process improvements that may never be commercialized especially for the manufacture of brand APIs.
Purpose of the post is not to find faults of Amgen or any other company but we all should collectively apply the fundamentals of science and engineering technology, do the right things to have economic and environmentally friendly processes. We have it all what it takes. Let us unleash our creativity and imagination. It is time.
Girish Malhotra, PE
EPCOT International
References:
1. Malhotra, Girish: Chemical Process Simplification: Improving Productivity and Sustainability John Wiley & Sons, February 2011
2. Malhotra, Girish: Chapter 4 “Simplified Process Development and Commercialization” in “ Quality by Design-Putting Theory into Practice” co-published by Parenteral Drug Association and DHI Publishing© February 2011
3. Malhotra, Girish: Active Pharmaceutical Ingredient Manufacturing: Nondestructive Creation De Gruyter April 2022
4. Malhotra, Girish: Capitalizing on Mutual Behavior and Chemical Reactivity of Chemicals, Profitability through Simplicity, May 29, 2013
5. Malhotra, Girish: Considerations to Simplify Organic Molecule (API) Manufacturing Processes: My perspective, Profitability through Simplicity, April 20, 2019
6. Malhotra, Girish: Sociochemicology May 30, 2013 Accessed January 13, 2023
7. Malhotra, Girish: Profitability through Simplicity
8. Malhotra, Girish: USP 11,267,798 B2: Manufacture of Piperine (1) An Excellent Teaching Tool. Profitability through Simplicity, June 17, 2024
9. Hsieh, Hsiao-Wu et al: A Continuous Process for Manufacturing Apremilast. Part I: Process Development and Intensification by Utilizing Flow Chemistry Principles https://doi.org/10.1021/acs.oprd.3c00400 Org. Process Res. Dev. 2024, 28, 1369−1384
10. Amgen wins patent appeal on Otzela (Apremilast) https://www.amgen.com/newsroom/press-releases/2023/04/amgen-wins-patent-appeal-on-otezla-apremilast Accessed November 20, 2024
11. Continuous Production/Manufacturing
12. AMGEN REPORTS FOURTH QUARTER AND FULL YEAR 2023 FINANCIAL RESULTS https://www.prnewswire.com/news-releases/amgen-reports-fourth-quarter-and-full-year-2023-financial-results-302055131.html Accessed November 27, 2024
13. Malhotra, Girish: NET ZERO for Active Pharmaceutical Ingredient & Fine/Specialty Chemicals: Nondestructive Creation, Profitability through Simplicity, November 7, 2024
14. Malhotra, Girish: Information Challenges for Product, Process Development and Process Design: A Reality Check, Profitability through Simplicity, April 10, 2011
15. Batch Production http://bit.ly/31dzpo3
16. McGraw Hill https://www.scribd.com/document/318719966/Docs
17. USP 7,109,203 Sulfonamide Derivatives Novartis AG