All opinions are my own.

Wednesday, November 7, 2018

Quick Review of Chemicals Related Process Development, Design and Scale up Considerations

Chemical engineers and chemists are the Picasso, the Michelangelo, the Zaha Hadid or the Gaudí or other artisans of their process designs. Their designs are emblematic of their learning, creativity and imagination. They use their training including their capabilities to exploit mutual behavior of chemicals and creatively extend  equipment performance to design excellent economic processes. 

Process development from the laboratory to a commercial operation goes through many similar steps. There is a process we all are aware of and go through, some cited (1,2,3). However, it is always good to re-visit the process and brush up on what we have learnt and are practicing (4). Such reviews lead to creativity sparks and eureka moments to do better. Learning never ends and has been covered by many to cite.

Below is a quick review from the lab to commercial path and steps we took and have since continuously improved. Steps are simple but as said earlier need to be revisited and improved. There is not enough space to cover everything. Please don’t be offended if I have overlooked few things. There is no financial involvement or remuneration from the companies cited. They are just examples.   

Most of us go through all or some variations of the following generic steps. 

1.    Commercial value of a chemical or formulation product is identified.
2.    A cost analysis is done and compared to any similar product on the market. If the cost and performance are better than the existing product/s, commercial value is explored/developed. If there are no similar products companies work fast to establish their foothold. 
3.    Product’s synthetic/formulation route is developed on paper, costed and explored in a laboratory.
4.    Time is spent to come up with the best operating conditions and the highest yield. Process is optimized. Additional cost analysis is done while the process development work is underway.  

With each positive in the above, race accelerates to commercialize the product. In process development subtle simplification steps can be overlooked. They can have significant impact on the product’s success. List is long but I am sharing some of the key steps that I have practiced. I am sure there are other ideas that are used. 

Raw Materials: 

Well accepted common practice in the lab is to get the raw materials from known research supply companies. For the initial steps, this might be alright but for the long-term success, it is of utmost importance to source the raw materials from commercial producers. Raw materials from the research supply companies are expensive and too pure to give any credible data and value for a commercial product. By getting the raw materials from commercial producers not only the developer company establishes a relationship and a long-term value but is also dealing with real raw materials that give us a dose of reality to develop an optimum and economic process. 

Acquisition of raw materials is different for batch and continuous process. Raw materials of batch process can be acquired when needed but for every continuous process they have to be available all the time. In either case raw materials don’t have to be tested prior to use if the laid-out needs are clearly defined and there is trust in supplier’s quality. Any deviation in raw materials from the established standard can spell significant financial loss especially for the continuous process.   

Physical State:

We all know that each raw material has its established physical properties and has its own specific behavior towards other chemicals including reaction by-products and products. How we understand these nuances and capitalize on their mutual behavior is critical for process development. Even with this acknowledged, still there are challenges we have to deal with. 

Physical state of each raw material and each reaction has value. We need to think how the raw materials, intermediates and the product will be handled in a commercial plant. They influence every processing step and disposal.  

Of the three states, liquids are the easiest to handle as they can be poured, metered and pumped with ease. Solids can be weighed but their handling in the lab vs. bulk can add their own challenges especially when the solids are hazardous. Hazard and handling can be somewhat improved or reduced/eliminated if the solid can be used as a liquid (solution or melt). Dilution to improve handling lowers process productivity. Economics has to be considered.  

Reaction products, if liquid, have a hidden value. If their melting point is compatible with the process operating temperature, reaction product can replace or minimize solvent use, there by improving process productivity. Most of us overlook this simple nuance but who recognize it, know its value. 

Raw materials that are gaseous at room temperature are best handled as liquid. Their use in the laboratory as liquid is limited and safety precautions are needed for safe handling. Their use in a commercial plant can be done well if the gas is metered in as a liquid. Exotherm of such reactions can be absorbed as the liquid converts to gas. An interesting example is ammonia as liquid vs. ammonia as 35.6% solution. Productivity and process conditions are significantly different in each case. 

Solids have their own nuances and also can create challenges. If the reaction can be carried out as a melt it can minimize or eliminate solvent use and improve process productivity. However, due to volume of material to be handled and produced such opportunities are generally limited to continuous processes.  

Hazardous chemicals require extremely careful handling. Batch vs. continuous process most likely will require different handling equipment and process. What might work for a batch process might not work for continuous process and vice versa. If hazardous solid raw material can be dissolved in a solvent or suspended as a uniform slurry, their metering can be better managed and controlled. Inert thickening agents like the kind that are used in the coatings industry can be used to create a pumpable uniform slurry. This is an unconventional approach and minimizes productivity loss that can happen due to excessive solvent needed for dissolution. Concept sounds simple, but can be a challenge. Ideas like this might seem farfetched. When they work they are called innovation. 

Solids when blended to create uniform powder behave very differently due to their densities, particle size, shape, viscosity, angle of repose, mixing characteristics (5)and processing conditions including temperature etc. As discussed later, they effect equipment design and selection for a batch or a continuous process. As suggested earlier what might work for a batch process might not work for a continuous blending process. 

Intermediate reaction products, if solid can create challenges. Their solvency in a solvent of choice facilitates the process. They can be solid and will have to be handled in a manner that will maximize process productivity.  

Liquids due to their mutual miscibility or lack of it can create opportunities e.g. in separation. Lack of miscibility can be changed by using different solvents. Many do not recognize and as explained later can capitalize on this solubility and their density differences. Different solvents can facilitate processing, improve process productivity and have financial and environmental impact. It is most productive if one additional solvent besides water can be used in the process. Disposal costs are thus limited. More than one organic solvent used in the same process add complexity and cost to the process. Experience of the chemist and chemical engineer and their ability to recognize such subtleties matter.  

Physical and Chemical Properties:

Once we know and understand different properties of chemicals, we can capitalize on their mutual behavior to create a process architecture that is economic and optimum for the needs. First time I saw physical and chemical property data details, their value did not register till I needed the information for process design and debottlenecking. They were a godsend. 

As has been discussed, physical and chemical properties are the building blocks of every process. Many of the physical and chemical properties may not be readily available. Some can be obtained from product suppliers and also retrieved from different chemical databases(too many to cite). Getting some of the mundane information e.g. solubilities, specific heat, viscosity, heat of formation etc. can be a challenge. If some of the values are not available they have to be developed. Solubilities in different solvents or different temperatures are not available from databases and they might have to be generated the old fashion way. Azeotropic behavior along with solvent miscibility/immiscibility is of values can be used to advantage.

Harnessing and creatively manipulating physical properties of solids to produce a consistent blend, that will not be prone to separation, can change the solid formulation landscape, especially for pharmaceuticals. Effort that is very different from what has been considered/tested and practiced would be needed. Some could say that it would be “impossible” to come up with a solution where separation can be minimized. However, I would say “we would be significantly challenged”. Formulators will have to think and step outside their comfort boxes. 

If we come up with a viable solution, my conjecture is that formulated products that have over trillion sales dollars per year could be impacted. Success could result in continuous formulation of many products especially pharmaceuticals (operating 7,140 hours operation per year per product per dose) would become a reality. We should see significant cost reductions, improved profitability and increased affordability. Existing equipment can be easily used. Few examples are referenced(1, 2, 3, 4)

Reaction kinetics can be manipulated to simplify design and process yields i.e. process economics. It can be of significant advantage. Raising the reaction temperature by about 10 0C, a simple quirk, doubles the reaction rate is of considerable value and works. It improves the whole process, design and economics.  

Equipment Design:

A critical element in the process design is the volume of product that needs to be produced. Product volume is the fundamental building block and a critical differentiator between a batch and a continuous process. Product costs, process economics and ROI of the investment depend on product demand. 

Combination of unit processes and unit operation make processes that produce the needed quality products. Physical and chemical properties of chemicals involved influence process design. Many books have been written on the subject. However, how different equipment and configurations can be creatively used depends on process designer’s imagination. They are difficult to write and only can be documented after their successful use to produce products. Many are considered proprietary. Volumes have been and can be written to document creative applications. Some examples (1)are discussed.  

Engineers/inventors at various equipment manufacturers, based on the needs they perceive for their customers, create and design different equipment and applications. However, the designed equipment could and does have value in other process applications that are not on their radar. Many times such uses and applications are not obvious. Engineers, do recognize their value and use them to simplify their processes. Selling application of such cross-fertilized equipment and their value internally takes out of the box thinking and sometimes skunkworks. Such innovations are considered proprietary and least documented.    

Plate and Frame heat exchangers e.g. Alfa-Laval (6)and others provide versatility of not only a heat exchanger but as a compact micro-reactor. Unknown to the suppliers we used such heat exchanger in early seventies with exceptional success. Creativity and imagination are needed to capitalize on such possibilities. Different breed of tubular heat exchangers (7)that are not used in the chemical industry are available and provide excellent possibilities of not only heat exchange but reaction space also. Such exchangers are significantly less expensive than the micro-reactors that are being tested in different laboratories and may be used in very specialized processes. Their use is ROI dependent.  

In certain applications, mostly due to tradition, high speed dispersers of high horse power are used. They pose their operating challenges. Inline dispersers (8)are economic, available and can be used. In certain applications they deliver similar results. They also provide significant operating flexibility e.g. a batch to continuous process conversion. One has to figure out their applicability and value. These along with an inline eductor can be used for dissolution and dispersion. 

As mentioned earlier, density differences of liquids can simplify processing. Value of density differences and hydraulics is of value in the design of gravity phase separators (decanters). If used thoughtfully, they can eliminate investment in exotic process controls that are generally considered a first option. I have seen decanters with fancy process controls but many become an operating bottleneck if their design is not understood by operating personnel. A decanter(9) based on specific gravity and hydraulic balance, if properly designed, is a simple and poor man’s elegant device. Most of the fancy process controllers (investment) disappears.   

Process Design:

Mass balance of every chemical reaction is important as it gives us the knowledge of what to expect. Information can be used to exploit mutual behavior of chemicals in process design, optimization, and waste disposal. Sometimes the reaction side products can pop surprises due to their toxicity and hazardous nature. They have to be dealt with. Scale up from lab to pilot plant gives us a better picture of how the process will behave chemically and physically. All of the generated information along with product volume can be readily used to commercialize a batch or a continuous process. However, many a times what works in the lab does not necessarily work in the pilot plant and scale up to a commercial plant. Changes have to be made and that is where the experience is put to test. 

In every manufacturing situation there are two kinds of processes: Batch or Continuous. Each has a well-established definition(10, 11). Routes are product volume and time dependent and there are no two ways about it. At times many forget the time element part of the continuous process. Time is an essential part of the definition. Batch processes are time interrupted. In continuous process, there is no stop and go except for the designated downtime for preventive maintenance and any production hiccups. 

There are few other very striking differences in the two routes. Since batch processes are stop and go, intermediate product after every step generally by force of habit is sampled, tested and the process corrected if out of spec. Intermediate sampling is an addiction and extends batch processing time. Stoichiometry deviation in reactive and formulation batch processes can happen if the operating instructions are not followed precisely. This generally results in lower overall yield and a financial loss. Such processes producing quality by analysis could also be called quality by aggravation (QbA).  

Stoichiometry in continuous reactive as well as formulation processes cannot deviate from design parameters. Designed process has to produce expected product quality every instant of the operating time. Compared to batch processes continuous processes are not at all forgiving to any deviations. Excursions outside the specifications can be an economic disaster. Uniform distribution of components in formulated products is very critical especially in formulations. Thus as stated earlier complete command is a must. Product quality has to be built in the process design and cannot be tested in.

Companies have to select the process to be used based on their business, short and long-term strategy, competition and expected return. Since each process, batch or continuous, is an independent route to produce the product, scale and method of execution of same unit process and unit operation can be different. Engineers and chemists are creative and do their best. We have to recognize that the equipment selected for a batch process can be retrofitted to produce many other products. However, such luxury does not exist for any continuous process as their design is product specific.   

Process controls based on process logic can make some changes within the designed limits in the stoichiometry but still the resulting product cannot deviate outside the product quality expectations. Today’s process controllers and logic are extremely robust and sophisticated to deliver the designed quality product. As said earlier if the product volume is there, continuous processes are generally method of choice. 

Batch and continuous processes have their own nuances of product supply and inventory. How they are handled and managed impact cash flow and cannot be overlooked. Manufacturing folks at times have to simultaneously wear many hats: manufacturing, accounting, marketing and purchasing.   

It is also important for the creators and the designers of process to document their thinking, rationale and the design basis. This includes mass and heat balance, equipment selection rationale and design calculations etc. It might be considered a mundane task but is a gift for the upcoming engineers and chemists. They would know how, what and the why of the design. This information comes in handy for trouble shooting, debottlenecking, filings and discussion with any regulatory authority. First time I had to document my design basis, rationale and calculations, it seemed to be a pointless exercise. However, was thankful for it being available when I needed the information for the task at hand and later work by others. 

In the annals of chemistry and chemical engineering there are many cases where equipment, physical and chemical properties have been married to create very simple and elegant designs and simplify existing batch and continuous processes. Chemists and chemical engineers at every company are the innovators and creators (12), and as stated earlier are the Picasso, the Michelangelo, the Zaha Hadid or the Gaudí or other artisansof the most innovative technologies and economic processes. Imagination and creativity are the two gifts they have and exploit. We have to let them do the best they do.     

Girish Malhotra, PE
EPCOT International 
  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: Focus on Physical Properties To Improve Processes: Chemical Engineering, Vol. 119 No. 4 April 2012, pgs 63-66
  4. Malhotra, Girish: Process Simplification and The Art of Exploiting Physical Properties, Profitability through Simplicity, March 10, 2017
5.    Tekchandaney, J. Material Properties Affecting Solids Blending and Blender Selection, August 22, 2009 Accessed October 8, 2018
6.    Alfa-Laval
9.    McCabe W.L. & Smith J. C., Unit Operations of Chemical Engineering, McGraw-Hill Book Company 1967, page 40
10. Batch Production: https://bit.ly/2ptp0kS
11. Continuous Production:https://bit.ly/2qtSYY6https://bit.ly/2POmN3Ghttps://bit.ly/2qAyc9f

Wednesday, October 24, 2018

What Is Needed for a Regulatory Approval of NDA/ANDA Filings in 90 Days?

Approving pharmaceutical manufacturing NDA/ANDA [New Drug Application/Abbreviated New Drug Application] applications at FDA in three months could be considered a Herculean task for the application filer (company) and the regulators (FDA) and it would be. But, what if, it could be done. If achievable, we would see significant changes on the pharmaceutical landscape (1). Monies would be saved in dealing with regulatory affairs, would reduce the time to market. That would mean higher revenues and profits for the brand and the generic companies. It could also lead to increased competition and potentially improve drug affordability though lower costs. There would be challenges to create and incorporate such a modified filing and approval process but the benefits achieved should outweigh them.  

The existing process of NDA/ANDA filing and approval is cumbersome and a challenge for the regulated and the regulators. I have not been involved but just reviewing the documents, what is expected and needs to be done, the process is cumbersome. I have to admit that I got lost in acronyms (alphabet soup), numbers e.g. xxx (w) (a) etc. and the legal jargon. No disrespect for the creators of the jargon, this alphabet soup gave me indigestion. 

I am sure there is process to accomplish approval in ninety days. For it to happen a precise roadmap (2)has to be created and it has to be so good that it can be followed even with eyes closed.  

Expectations from the Road Map and What would be Expected from the Regulators and the Regulated:

Putting my process development, manufacturing and management hats simultaneously I am proposing a process that would have to be developed by the regulators (FDA) and the companies for their NDA/ANDA approval. I consider filing an NDA/ANDA filing equivalent to a manufacturing process that has to be followed precisely by both sides to create a quality product (3), an approved NDA/ANDA filing. 

Regulators based on their years of experience will write down precise step by step recipe that has to be followed by the NDA/ANDA filing company to get their approval in the slated time. I need to emphasize that if the regulators (FDA) precisely define their expectations, the process will be smooth. There will be no doubts about the expectations in the mind of the filers. Again, information asked has to be precise and cover every detail of type of manufacturing, synthesis as well as formulation, process. It is not going to be easy and can only come from chemists, chemical engineers and other engineering disciplines who have developed, designed, commercialized and reviewed such processes and products from head to toe day in and day out.

Yes, the companies might take their own time to file, their own choice, but the expected information has to be such that there is no finger pointing suggesting “we did not know what was expected”. If companies fulfill and deliver the information as expected and every deliverable is there, company should get their approval/disapproval decision in ninety days or less. That would be a huge win for all involved. There could be some back and forth but it would be minimal and the ninety-day clock, from the day the filing was submitted, will keep ticking. i.e. no stoppage. 

I would call regulator’s road map a quality by design recipe for a perfect process created by the regulators that the filing company will follow. If for some reason the regulator’s (FDA) recipe, the defined road map, cannot be followed or understood by the company, the application process will go from Quality by Design (QbD) to Quality by Analysis, aggravation, (QbA) and that means delays and added costs.  

Once the regulators have defined their needs to approve a filing, chemists and chemical engineers at the companies have to understand the questions and deliver the answers through their submissions. Both entities will have to understand each other’s paper conversation about every process. Only their dialog will produce repeatable quality product through the asked questions and the delivered answers. If this happens, the review process will go smoothly and time reduction mission will be accomplished. 

Companies submitting information will have to have an absolute command and knowledge of their processes and will have to present information so that it will convince the regulatory examiner/s that the product is going to follow every claimed expectation and regulation that has been laid out. In short product will be of consistent and repeatable quality. If this happens I envision that the approval process in the long run will be simpler and quicker than the current process.

Examiners at the regulatory bodies will have to understand the conversed information. In some instances clarification and or additional information might be needed. My expectation is that with time such cases might be few and far in between and the expended time would be significantly less than the present process. 

What would the Showstoppers say or think? 

There will be many who will be apprehensive of the suggested process. Their perspective would be mis-information could be submitted to get fast approval. Yes that is possible but the regulators have a recourse of barring or shutting down such rouge operations. That privilege will have to be used if the current legal system stands in the way. 

I am conjecturing that for the success of this proposal regulators will draw a precise map of what they need to have from the filers. It is going to easier said than done. Regulatory staff will have to be envision and document from their past experiences what all they need for a first-time complete submission. We have to recognize that it can be done but no one is going to welcome a voluminous dissertation. 

Let’s assume that every possible information FDA or any other regulatory body asks/needs for the approval is submitted to the utmost correctness the first time and the regulator have no questions. Would the regulators accept such an application without any apprehension and approve the filing in the 90 days or sooner from the first day of filing? I am sure that there will be doubters on both sides. It will take time to accept this process as it is outside the comfort zone of filers and the regulators.  

We have to accept that will not have a perfect road map from the get go. To get to the perfect stage, I am envisioning there will be lot of obstacles from the legal and consulting firms and even the manufacturing companies. Pilot program might have to be developed by the regulators and tested and over tested. Naysayers and doubters will have their days. However, if successful pharma’s landscape will change. I believe the best minds can create a process that is much simpler than what is being followed. 

I don’t have every “t” crossed and “i” dotted for the proposed process. I may sound repeating myself at times.  Objective here is to use the proposed process to build a system that will expedite the approval process. As stated earlier results will be higher revenue for the brand and the generic companies through speedier introduction of new therapies and out of patent generics, additional competition, possibly lower drug prices and inclusion of better technologies after the products are commercial. Process of continuous improvement would be included more than ever. 

We have an opportunity to play outside the box. It could be a hip-hip hurray moment. 

Girish Malhotra. PE
EPCOT International

1.     Malhotra, Girish: Can the Review and Approval Process for ANDA at USFDA be Reduced from Ten Months to Three Months? Profitability through Simplicity, March 25, 2017

2.     Malhotra, Girish:ANDA (Abbreviated New Drug Application) / NDA (New Drug Applications) Filing Simplification: Road Maps are a Must. Profitability through Simplicity, May 11, 2017

3.     Juran, Joseph M., Juran on Quality by Design, Simon and Schuster, ISBN 9780029166833, May 1992