The Good, The Bad, The Ugly (1) Complexities of Pharmaceutical Manufacturing

Most of the pharmaceutical products irrespective of product demand are manufactured using batch processes. Continuous manufacturing of pharmaceuticals is possible. However, process viability and product demand correlation need to be microscopically examined. Lately there has been an increasing fervor among the regulators, equipment suppliers, some research institutions/universities and consulting houses that the pharmaceuticals companies should use/adopt continuous processing for the manufacture of drugs. Use of continuous processes where relevant, applicable and economically justified is not a bad idea. It is a great idea. However, it is necessary that such processes are authenticated according to the established definition².

Some batch pharmaceutical formulation processes are being called continuous processes. I hope we are not twisting fundamentals of science and engineering, economics and common sense to claim that we are on the leading edge of innovation when in reality, we are not. I am afraid that some of the technocrats and bureaucrats may be basking in false sense of accomplishment. Considerable monies are being spent. Such situations can impede real opportunities when they come along. I hope there is economic and science-based justification for the effort and the industry is not being led along an unreal path or being given an optical illusion.  

   

I have shared my perspective about what is involved and what all it takes to develop, design, scale up, commercialize and manage continuous processes especially in pharmaceutical manufacturing^3-16. For my own benefit I thought it would be helpful to re-review and share what I have learnt and practiced and make sure I crossed every “t” and dotted every “i”. My observations are based on my experiences and are in no way intended to criticize or challenge opinions and perspective of others who are involved in process design, development, commercialization and management of manufacturing operations. It is very possible that I might have missed some process design considerations.

Based on my experiences it takes a village (team of chemists and chemical engineers, marketing, financial analysts, supply chain professionals, quality control, maintenance and manufacturing) at a company to think about and commercialize manufacturing processes for any chemical and related industry. Some may not want to accept it, but pharmaceutical manufacturing is a subset of the fine/specialty chemical industry. Unit processes and unit operations used in the chemical industry are also used in API and Formulation processes. API are fine/specialty chemicals that have disease curing value and excipients are inert additives that along with binders create the dose that can be easily ingested. 

Like any manufacturing, pharmaceuticals have their good, bad and ugly complexities. They are not per say identified here but anyone familiar with process design, development, manufacturing and profitability would understand them. It needs to be recognized that the technologies and equipment that can significantly simplify processing and lower manufacturing costs are well developed and practiced in the chemical industry. However, I feel that in pharmaceutical manufacturing regulatory constraints slow down process simplification and innovation. 

Product demand dictates type of process used and this applies to every business. Since pharmaceuticals are for human consumption there are regulatory compliance requirements. They add additional complexities of why, what and how the manufacturing will be controlled. 

Discussion here is focused on small molecules actives and their formulations. Biopharma are not discussed. They are in their infancy. Some concepts used in the small molecule processing can be applied in biopharma processing. They have to become affordable to capitalize on values of economies of scale. To get there, they have long ways to go. Different business model may be needed. 

Pharmaceutical manufacturing has two components and each has to be treated differently. 

•      API Manufacturing 

•      Formulations  

API Manufacturing:

Active pharmaceutical ingredients (API) are toxins and are needed in small quantities to cure various diseases. Due to the small quantity needed, their dispensation in pure form is difficult. The best dispensation method is to convert them in tablets or solutions. Tablets are generally the most convenient form. 

Table 1¹³ is a hypothetical illustration of requirement for different APIs needed per patient at one tablet per day at variable doses. 

Dose, mg	Patients, millions	API Kilograms needed/ year @ one tablet per day per year	API Production
			Preferred process Type	Number of plants
1	500	182,500	Batch	One or More
200 *	0.1	7,300	Batch	One
10	100	365,000	Batch	One or More
100	50	1,825,000	Batch or Continuous	Could be a single continuous plant but generally batch
500	20	3,650,000	Continuous	Could be a single continuous plant but generally batch due to multiple sites

* Orphan drug

Table 1¹³: API Manufacturing Options

Table 1 may not look of much value but is extremely important for process design, production planning and scheduling, inventory control and product management. They tell us the good, the bad and the ugly complexities of manufacturing as they have financial impact. Actually, profit and loss of the products depends on the how they are produced, commercialized and managed. 

Since one kilogram of an active ingredient theoretically produces one million of one milligram tablets, small amount of API can fulfill demand of a large patient base. Dose and population in Table 1 determine the product demand. Chemistry and economics dictate the type of process used. Batch processes are the tradition for API manufacturing. Ways to improve productivity, product quality, profits and affordability have been discussed elsewhere^3-16.  

Two of the five APIs illustrated in Table 1 (let’s assume these are generics) could be produced using continuous processes. As stated earlier batch processes are generally the first choice till the market demand increases. However, under pharma’s current business model, even if the demand increases most APIs will be produced at multiple sites by batch processes. Efforts to improve or simplify existing batch processes or transition to continuous processes are shunned. Regulations do not facilitate innovation either. Existing regulations that are focused for batch processes cannot be optimally applied to continuous processes. 

We have to acknowledge that APIs are fine/specialty chemicals. Equipment and processes used are no different from non-pharmaceutical fine/specialty chemicals. Same equipment is generally used to produce different actives. Since the equipment is not dedicated to any certain API, processes are modified¹⁷ to fit the equipment. This may be the most convenient method with least investment but generally such processes are inefficient and not the lowest cost. 

APIs are toxins. Thus, thorough equipment cleaning is a must to assure that there is no cross-contamination. Due to cleaning requirements and many products being produced at the same site, asset utilization for the API batch processes is less than optimum^18-20at best. Global overcapacity does not help either. Chemistry similarity can improve asset utilization. Sometimes it is overlooked. Judicious review is necessary and potentially necessitate a different business model¹⁵. 

Formulations:

Table 2¹³ is a hypothetical illustration of number of formulation plants or parallel trains that would be needed to produce at the 200,000 tablets per hour for different drug doses.

Like Table 1 this table might not look of much value. Besides giving us API quantity needed it also gives us the amounts of excipients needed once they are finalized. Combined information gives us process design, production planning and scheduling, inventory control and product management parameters. 

Theoretically Table 2 suggests that continuous formulation plants/trains could be used to fulfill the demand. Some processes need to be continuous as economies of scale will improve profits, cater to fluctuating demands and improve affordability. Since continuous formulation operations have never been on pharma’s plate, the product requirements are generally filled by batch processes.  

Table 2 Formulation Options

As stated earlier process development, equipment sizing and command would be extremely critical. One would have to have complete control of the stoichiometry, mixing and component distribution to assure tablet uniformity. Single formulation train operating 7,140 hours per year (50 weeks x 7 days/week x 24 hours/day x 0.85: allowing 15% downtime) requires a mindset that is very different from batch operations. Such operations are very possible but my conjecture is that due to stringent regulatory, product quality demands and process development challenges companies most likely opt to stay with batch formulation processes. Actually, batch processes due to their continued sampling, analysis and cleaning between products can be more complex to manage compared a properly designed continuous process. 

Last drug dose in Table 2 needs attention. Due to number of patients it is an orphan drug. Yearly need can be produced in a short time e.g. less than 10 days at 200,000 tablets per hour rate with very little down time. Such processes per established definition²would not be called a continuous process as the equipment would sit idle till the next run. If companies or the regulators want to call such process a continuous process are they essentially changing laws of science and engineering to propose a different definition that have not been clearly stated/proposed for public scrutiny and comments. 

It is ironic that certain extremely high-volume products (NSAID) could have been re-engineered to continuous process but stayed with batch process option.  

Complexities:

Pharmaceutical companies like every other corporation have the goal to maximize their economic return. API manufacturing and their formulations are two distinct and different processes needed to produce a drug dose. API manufacturing are reactive processes whereas formulations in simplistic terms are blending and tableting operations. Every designed and commercialized process has to be the most economic. Economies of scale and process methodology (batch vs. continuous) change the product cost dynamics: batch costs generally being higher than the continuous process costs.

As stated earlier even with using same/similar unit processes and unit operations being used in batch and continuous processes, different thinking goes in their development, design, scale up, commercializing and management. Since the current regulations are geared for batch processes, different regulations will have to be developed and applied for continuous processes. It would be sensible if chemists and chemical engineers who have hands-on experience in development, design, commercialization and management of such processes advise FDA and other regulators in creation of the necessary regulations. Manufacturing Advisory Committee that existed once should be revived to facilitate and expedite the development of regulations that could apply to continuous process. I am not sure if FDA and similar regulators have the staff with hands-on experience.  

In batch processes, raw materials and intermediates are generally staged and tested for quality and use. Since continuous process are time independent, testing of every raw material and intermediates is not be possible and economic. Stop and go opportunity does not exist. Every unit process and unit operation has to operate as designed. Deviation from the operating conditions would mean poor product quality and significant waste and financial loss. Absolute command of process stoichiometry and operating conditions is a must.  

Due to inherent nature of batch processes, companies in every industry sample and test intermediate samples even if QbD (quality by design) methods are incorporated in the overall design. Companies do make every attempt to minimize sampling and testing. Benefits are higher profits. However, some habits die hard and companies have to be proactive and curtail these habits. 

Every chemist and chemical engineer incorporates fundamentals of science and engineering to design and commercialize the best and the most economic process. They use existing process control technologies that have been used in the chemical/fine/specialty chemical industries for more than fifty years to meter liquids, solids and manage reactive batch and continuous processes. These work extremely well. It is bit confusing when FDA asks companies to practice QbD methods when they are the very foundation of every process design.   

Since continuous processes are time independent and processes are under control, inventories should be minimum.  Intermediate inline testing in continuous process can be done to make sure that the process is operating within the designed parameters and all is going well. Excursions outside the design limits if not caught in time can result in significant quantities of off-spec product and a financial loss.  

Raw material and intermediate sampling and testing have significant impact on supply chain and production planning. Batch processes increase in-process inventories and influence cash flow. In-process testing also extends batch cycle times and negatively impact asset utilization^18,19. Collectively they increase product cost. However, raw material and work-in-process inventories can be held to a minimum in continuous processes. Production rates can be managed to meet variable product demand.  

Pharmaceutical companies like any other manufacturing companies have sufficient knowledge base and experience that has been used to produce dispensable tablets. Since they have been and are being practiced I firmly believe that they can be applied to overcome every complexity of pharmaceutical manufacturing. It seems that we are bogged down in “analysis paralysis” and wrath of regulators if changes are made to the existing manufacturing processes without their approval. 

Figure 1 is a review of a formulation option from Table 2. An assumption of 95% excipient per 5% API is made. 100% yield is used. 

Dose. Milligram	1
Tablets per hr.	200,000
API needed Kg. /hr.	0.20
Excipient Kg. /Hr.	3.85
API+ Excipient Kg. /Hr.	4.05
Tablets needed millions per Yr.	18,250

Figure 1

To satisfy the demand of, Figure 1, 50 million patients per year equipment and technologies to commercialize a continuous pharmaceutical formulation operation exist.  Why companies have not commercialized continuous processes for such cases is perplexing. Are the regulations in the way or are the companies afraid to test new equipment and methods to simplify manufacturing methods and processes? 

Companies can convert the bad and the ugly complexities to good by focusing on selected drugs thereby create economies of scale and capitalize on what all has been discussed above. If done, landscape will change significantly and affordability of related drugs will improve. Most likely different or new business model would be needed. Amazon, Berkshire Hathaway, JP Morgan alliance and Veteran’s Affair^{21, 22}initiatives, if successful, could make a difference. 

Regulations add complexity to any company wanting to take advantage of economies of scale, process simplification and transitioning from batch to continuous processes. My conjecture is that no one wants to spend monies on re-approval of the products. Regulators²³are making attempts to simplify approval processes but there are external and internal challenges. Companies might not want to simplify manufacturing for the simple fact that they are profitability even with inefficient processes.  

All said and done it is ironic that in the last four decades technologies have been developed and commercialized that are beyond our imagination but we have not used available technologies to simplify pharmaceutical manufacturing that would benefit better than 50% of the global population. It would be interesting if reasons and causes can be identified. I believe that through creativity and imagination most of pharma’s complexities can be simplified and overcome. 

Girish Malhotra, PE

EPCOT International

1. The Good and the Bad and the Ugly, 1966, Accessed March 29, 2018 
2. Continuous Production, https://en.wikipedia.org/wiki/Continuous production, Accessed July 14, 2017
3. Malhotra, Girish: Batch or a Continuous Process: A Choice; Pharmaceutical Processing, March 2005, Pg. 16 
4. Malhotra, Girish: Alphabet Shuffle: Moving From QbA to QbD - An Example of Continuous Processing, Pharmaceutical Processing, February 2009 pg 12-13
5. Malhotra, Girish: Hesitation In The Drive To A Continuous Pharmaceutical Manufacturing Process: Real or Imaginary? Pharmaceutical Processing, July 2009 pg-12-15
6. Malhotra, Girish: The Path Towards Continuous Processing, Pharmaceutical Processing, August 2010, pgs 16-20
7. Malhotra, Girish: Is Continuous Processing in Pharma’s Future? Profitability through Simplicity, July 24, 2012
8. Malhotra, Girish: Continuous Process in Pharmaceutical Manufacturing: Considerations, Nuances and Challenges, Contract Pharma, June 2, 2015
9. Malhotra, Girish: My perspective for Pharmaceutical Manufacturing Technologies/Processes and Continuous Improvements, CPhI 2015,http://epcotint.com/Pharmaevolution/GirishMalhotraSeptember2015CPhIreport.pdf  
10. Malhotra, Girish: Continuous Pharmaceutical Processes and Their Demands, Contract Pharma, pg 37-40, April 5, 2016
11. Malhotra, Girish: Strategies for Improving Batch or Creating Continuous Active Pharmaceutical Ingredient (API) Manufacturing Processes, Profitability through Simplicity, March 20, 2017
12. Malhotra, Girish: Reality and Un-Reality: Continuous Processing in Pharmaceutical Manufacturing, Contract Pharma, April 3, 2017
13. Malhotra, Girish: Batch, Continuous or "Fake/False" Continuous Processes, Profitability through Simplicity, July 20, 2017, American Pharmaceutical Review, Vol. 20 Issue 6 September/October 2017 pgs. 86-91, Contract Pharma, Nov./Dec. 2017 pgs 56-58, Chemistry Today, November/December 2017, Vol. 35(6) pgs 62-65
14. 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
15. Malhotra, Girish: Strategies for Improving Batch or Creating Continuous Active Pharmaceutical Ingredient (API) Manufacturing Processes, Profitability through Simplicity, February 2011, Revised March 20, 2017 
16. Malhotra, Girish:  Chemical Process Simplification: Improving Productivity and Sustainability John Wiley & Sons, February 2011
17. Malhotra, Girish: Square Plug In A Round Hole: Does This Scenario Exist in Pharmaceuticals? Profitability through Simplicity, August 17, 2010, Accessed March 28, 2018
18. Keeling, David, Lösch, Martin, Schrader, Ulf: Outlook on pharma operations, McKinsey & Company, 2010 Accessed March 28, 2018
19. Benchmarking Shows Need to Improve Uptime, Capacity Utilization, Pharmaceutical Manufacturing, September 20, 2007, Accessed March 28, 2018 
20. Kodipyaka, R: OSD: Challenges & Improvement Opportunities, Pharma Horizon, Vol.2(1) 2018 pg 21-22 Accessed April 2, 2018
21. Leading U.S. Health Systems Announce Plans to Develop a Not-for-Profit Generic Drug Company, www.businesswire.com, Accesses March 1, 2018
22. Malhotra, Girish: Could Amazon (A), Berkshire Hathaway (B) and J.P. Morgan Chase (M) be the Anti-Ballistic Missile (ABM) Needed to Control/Curb Rising Healthcare Costs? Profitability through Simplicity, February 9, 2018, Accessed February 27, 2018
23. 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 Accessed April 4, 2018

          

Pharmaceutical Sales Forecasts: Fiction, Facts or Alternate Facts

Every year many analysts and magazines publish their forecasts for the new drugs that have been approved by the US FDA the previous year. Every thing looks glorious and sales are projected in billions. An extremely rosy picture is painted. However, I have not found a “fact check recap” of the projected numbers three or five years after the sales projections have been made. In today’s new political lingo no one has presented “alternate facts”. Real numbers come from how many patients are benefitting from the new drugs and that is a missing piece of the reality.

It seems that the sales projections are always Up Up and Away ⁽¹⁾, a wishful thinking. Even the yearly total global pharma sales are projected to grow about 5-7% per year ending around $1.3 -$1.5 trillion dollars by 2020-2022 ^(2,3). I have serious doubts about these numbers and wonder have these projections been questioned in a public forum. Most likely no one cares, at least that is my conjecture. Since businesses design their plants based on projected sales, I am comfortable in saying that we will consistently have “mega over capacity” at plants. In simple terms overspent capital, under-utilized assets and inefficient plants. This sentiment was recently echoed ⁽⁴⁾. I am afraid poor patients pay for this largesse through higher drug prices.

Trying to understand these forecasts, a recent article ⁽⁴⁾ was brought to my attention. Excerpts from the article enlightened me and reconfirmed my apprehensions. It seems to be telling us things many of the forecast pundits most likely do not want to know. 

“First, most consensus forecasts were wrong, often substantially.”

 “Furthermore, a significant number of consensus forecasts were overly optimistic by more than 160% of the actual peak revenues of the product.”

“Although the conclusion that most forecasts are poor is not surprising in our view, the magnitude and extent of the error in forecasting is striking and troubling as it suggests a large-scale and systemic misallocation of capital and destruction of value in the industry. It also suggests that there is a substantial opportunity for companies and investors who develop a competitive advantage in forecasting.”

 “The ‘consensus’ consists of well-compensated, focused professionals who have many years of experience, and we have shown that the consensus is often wrong.”

A recent analysis of the projected global pharmaceutical sales ⁽⁵⁾ shows their irrationality, at least to me. 

If the world knows that these forecasts are “often wrong”, the question is what is the rationality and value of publishing such numbers year after year. Who benefits? Why are these supposedly “ostrichian” numbers published? Are these numbers just exuberance based or are telling us that the companies are doing great things when the reality is otherwise or is this a way to impress investors or much ado about nothing? Shouldn’t we be associating new therapies with people rather than dollars when we know these therapies have a very limited patient base?

The most important fact or a reality, which matters the most is never published or for that matter known to anyone and that is how many patients (+/-10%) are benefiting from the drugs or using it. Forecast numbers per drug can be checked, published and compared against forecast using various company Annual Reports.

I believe that number of patients would be an excellent indicator of each drug’s efficacy and ability to treat patients. Increasing numbers will tell us that drug’s mode of action is of value and could be modified to create better treatments for larger population. Future treatments could be based on this learning. If any new drug is just marginally better than an existing drug, lack of its sales will also be a deterrent for R&D investment in such therapies.

As I have not seen reality checks against forecasts and it begs the question “are we afraid to ask or know the reality?” My conjecture is that reality checks will allow the companies to have better R&D, manufacturing processes and offer us an opportunity for efficient asset utilization, a small step for making drugs affordable. In addition, these numbers will also be of value to the investors to discern and direct value of their investments.  

Girish Malhotra, PE

President

EPCOT International

1. The 5th Fifth Dimension, Up Up and Away, Accessed February 8, 2017
2. World Preview 2016, Outlook to 2022; Accessed February 8, 2017
3. Global Medicines Use in 2020: Outlook and Implications, Accessed February 8, 2017
4. Cha, Myoung, Rifai, Bassel and Sarraf, Pasha, McKinsey & Co.: Pharmaceutical forecasting: throwing darts? Nature Reviews, Drug Discovery: Volume 12, October 2013,  737-738, Accessed February 7, 2017
5. Malhotra, Girish: Manufacturing technologies and their part, Chemica Oggi Chemistry-Today, September/October 2015 Vol. 33(5) pg. 28-31

Reading the Tea Leaves: Predictions for Pharma's Future

The following was posted on January 15, 2014 on www.pharmaEvolution.com now defunct. 

Based on what has transpired in the last two years, I am writing an update that will be published in few days. I thought an old post would be a good preamble for the future. 

Wednesday, January 15, 2014

The human race has one significant weakness. We all want to know what the future holds for us, and we want to know now -- hence the popularity of fortune-telling and such "fields" as astrology. In our desire to predict tomorrow, however, we often forget simple cause and effect. What we do today helps determine what happens tomorrow. For instance, every new developed and approved drug, in efficacy and in price, directly impacts company and industry profits.

Despite ongoing product-quality issues, drug companies have been profitable. However, this picture is about to change. Some say that it has already begun to change. Clearly, what we do today will shape pharma's future.

Here are my predictions for the industry over the next five to fifteen years. How do they compare with yours? As you'll note, I see the status quo prevailing in some cases, forcing change to come from unexpected places, such as regulatory agencies or outside forces. But what do you think?

1. Pharma companies will have to figure out how to reach the world's masses, rather than small patient populations, if they are to succeed in the long term.

2. On average, the industry's success rate of development and commercialization, for both small molecule and biotech drugs, will not change much from its current level. “Eroom’s Law” still limits pharma's destiny.

3. High drug pricing soon will be frowned upon and discouraged by PBMs (pharmacy benefit managers) and hospitals.

4. Instead of becoming “process-centric,” pharma will stay “regulation-centric.” Process-centricity, if adopted, would allow companies to exceed regulatory requirements, which could also avoid many issues that have created public relations and financial headaches. Current regulatory guidelines and requirements discourage change. Since changing an existing process requires approval, it is still perceived to be a long and expensive step, and something to be avoided.

5. Industry will hedge in adopting many of the internal changes (manufacturing methods, technology, and supply-chain improvements) that could improve profitability and move from the current quality by analysis/aggravation to a quality by design approach. Again, this is due to regulatory constraints. Short patent lives for the ethical (brand) drugs will impede innovation in manufacturing technologies. At the same time, most generics manufacturers have been in business too short a time for their managers to grasp the value of better technologies and methods.

6. Since current or higher level of profits can be achieved with the current inefficient practices, industry leadership does not see any value in improving its product development, business, and technology practices. External forces will drive change.

7. With not much forthcoming from the industry to improve its manufacturing and business practices, regulatory bodies will enact regulations that will force the industry to adopt better practices. This tug of war will continue, unless the industry takes the lead.

8. Continuous formulation processes could be a reality in the next five years for new drugs.

9. Continuous API manufacturing is possible, but it will require a different economic and manufacturing model.

A shift is taking place in the global pharmaceutical businesses, whether we like it or not, and is being brought on by generics manufacturers based in developing countries. Pharmaceutical companies will have to take radical measures to retain or increase their revenue streams and profits.

A shift to a new model from the current model in new drug discovery, development, manufacturing technologies, supply chain, conservation, and sustainability requires an “unreasonable man” as described by George Bernard Shaw. ("The reasonable man adapts himself to the world; the unreasonable one persists in trying to adapt the world to himself. Therefore all progress depends on the unreasonable man.")

Significant innovation is feasible in manufacturing and its support functions. For it to happen, though, the pharmaceutical industry will have to move its focus from “regulation-centricity” to “process-centricity.” The industry still appears unwilling to take the steps required for change.

The world and its inhabitants would clearly benefit from better and higher availability of lower cost drugs, while companies would be able to improve their profits from the current levels. I view this as a win-win situation. What do you think?

Girish Malhotra, PE

EPCOT International