COVID-19 has changed the world like nothing most of us have ever experienced in our lifetime. The world is dealing with a greater number of critical raw materials and drug shortages.
Looking back at the FDA Science Board Meeting November 16, 2001 which led to the Guidance for Industry PAT - A Framework for Innovative Pharmaceutical Development, Manufacturing, and Quality Assurance September 2004. Key drivers for this initiative were drug shortages, drug recalls, overall product quality concerns and the concern regarding the potential effects on the nation’s safety (9/11).
The FDA considers PAT to be a system for designing, analyzing, and controlling manufacturing through timely measurements (i.e., during processing) of critical quality and performance attributes of raw and in-process materials and processes, with the goal of ensuring final product quality. Quality cannot be tested into products; it should be built-in or should be by design. However, aren’t the QC labs still doing this? Reports over the years estimate the industry wastes approximately $50B a year in manufacturing costs.
Subscribe to our e-Newsletters
Stay up to date with the latest news, articles, and events. Plus, get special offers
from American Pharmaceutical Review – all delivered right to your inbox! Sign up now!
Process Analytical Technology or PAT, is intended to support innovation and efficiency in pharmaceutical development, manufacturing, and quality assurance. Since 2004 technological innovation has made the FDA goal closer to reality. This has been done by a combination of pharmaceutical technology vendors that provide specific solutions to pharma’s analytical needs. Advancements have been accomplished partially through collaborations and partnerships formed to solve specific workflow issues. Sometimes it was through a group's innovative approach to solving a workflow challenge. But finally, in part through the pharmaceutical scientists and engineering teams that, based on their knowledge of the molecule they wanted to design, was the understanding of a workflow issue and coming up with their own tool to solve that problem. Other factors of great impact were the move to Flow Chemistry as an alternative to batch manufacturing allowing for a more scalable approach, and the use of new reactants to develop new products both on a scalable model and the discovery of a new drug candidate faster. Note that all of these require either new technology and/or upgrades to existing systems. In the biomolecule world utilization of single-use bioreactors and ongoing breakthroughs that provided not only a scalable solution but also the development of a faster drug development cycle. These innovative approaches provided scientists with a suite of tools that enhanced their process learning, understanding, and knowledge which leads to the goal of process control.
Fast forward to 2020. Dr. Huiquan Wu (FDA) in American Pharmaceutical Review October 2020 Integration of PAT and Data Systems for Manufacturing Process Control and Quality Improvement: A Brief Review of Past Developments, Current Status, and Certain Future Considerations wrote, “apparently it is necessary to integrate PAT and data systems for effective manufacturing control and product quality improvement.”
This quote by Dr. Wu states perfectly where things are or need to be. By clarification, if we look at the general role of ‘laboratories’ as the monitoring stations of the process where things need to go for confirmation. It is the lab’s role to control stations in manufacturing and provide almost real-time and actionable controls to ensure process control and product quality.
It is essential for successful implementation in bringing together the key stakeholders (Process Development, Quality, Regulatory, and Manufacturing) to be one team with the expertise comingled - not four individual organizations as was done in the past. This allows for a smoother new process/product implementation due to cooperation, collaboration, knowledge sharing and transfer, with ongoing oversight during the implementation and product filing.
It’s all about the workflow. The elements are broken down into three categories:
1. Sampling – how it’s done, how its handled, (like in a mathematical formula if a mistake is made in plugging in the values along the way this will potentially lead to a wrong answer)
a. Critical to the choices of sensors
i. Sample Matrix: powder, slurry, emulsion, pills, tablets, capsules liquid
ii. Drug Substance or Drug Product
2. Sensors (tools) run the gamut of simple to complex (Thermometer, pH, Spectrophotometric, GC, Particle size, LC, MS, NMR, etc.) for brevity we’ll stop here.
a. As a means of a quick example would be the monitoring and now measuring Glucose.
i. From the tool bag of sensors (RAMAN, FT-IR, NIR, GC, LC, NMR)
1. Your choice of tool is dependent upon the sample/matrix
2. The time and amount of information needed to decide on an action (milliseconds or hours)
b. Additional considerations for tool choice (purity, concentration, levels of impurities detection (LOD) this will determine the tool required.
3. Data also runs the gamut from the individual ‘sensors’ and their overall operational communication, analysis, and control of each individual vendor’s solution.
a. This is to Dr. Wu’s point – you have many types of sensors - for sake of discussion let us say there are 10 of these tools and they are from 10 different vendors. Each one of them has a communication, analysis, and control software specific for their sensor. Each one of them has a unique software language. The critical data from each individual sensor is providing Critical Process Parameters (CPP’s) and Critical Quality Attributes (CQA’s) that provide specific data points about the process. The data/information from all these sensors must be gathered and translated into a centralized data hub for further distribution to other specialized software products. These will be utilized from chemometric, multi- variate data analysis, historian, and control software packages that bring together the specific data points from each sensor. Based on these values the overall operation, monitoring, and control of the manufacturing process can deliver a higher product yield and assurance of product quality.
Since the beginning of the FDA initiative the early sensors that met the need were primarily NIR, FT-IR, UV/Vis, and Raman spectrometers. Early drug approvals were as follows:
Orkambi- Vertex (Cystic Fibrosis)
Janssen -Prezista (HIV)
Verzenio – Lilly (Breast Cancer)
Symdeko – Vertex (Cystic Fibrosis)
In the laboratory where all the studies, DOE’s, and testing, which now has provided the learning knowledge and understanding of the CPP’s and CQA’s for an optimized process, the optimum conditions and results have been determined and this information is now embedded into the control system in manufacturing.
The sensor technologies in some cases had to be adapted from laboratory use into being embedded into the manufacturing process. It’s a combination of physical proximity to the process as well as the communication of spectral output to the control system. This provides the monitoring information upon which the control of the process yields the confidence, repeatability, and continuous delivery of a high- quality product.
For all sensor technologies many must be modified in some way as to make the sensor fit for purpose in the manufacturing space. The manufacturing environment in combination with the product requirements for manufacture play a key role in these modifications. One of the key drivers in pharmaceutical manufacturing is flexibility and mobility. A major factor is a reduction in the size of the manufacturing footprint. This applies to both small and large molecule development and manufacture. The enabling technologies allow for this drive to newer and smaller manufacturing sites and a derivative of this is also mobile manufacturing. As things get smaller the supporting sensor technologies must also become smaller to fit the model as they may be part of the manufacturing monitoring and control system. Another concern in the development and manufacturing of a small molecule product is if the manufacturing space is a classified/hazardous environment, the sensors that are deployed must accommodate that space. The sensor cannot be an element in that process that could lead to a catastrophic event. Many of the sensors being used today have a much smaller footprint while offering improvements in output enabling greater discovery and implementation into manufacturing. In biopharmaceutical manufacturing the sensors that are being used have different considerations. In this case sterility is paramount.
One of the other sensor technologies that had a major technological during this time frame was High Performance Liquid Chromatography (HPLC). For years HPLC has been and still is considered the gold standard regarding the degree of high-quality quantitative information about a sample. The draw backs with this technology have been the speed and level of operational complexity. In the world of PAT and Continuous Manufacturing depending upon the speed of the reaction HPLC might not be the correct sensor for that part of the workflow. The breakthrough came with the advanced technology of UPLC and/or UHPLC depending upon the vendor. Ultra-Performance Liquid Chromatography could now allow for the use of separations technology in the world of PAT. This technology came about through the design of new smaller particle chemistry in combination with newly designed hardware which was needed to operate with the high-pressure column chemistries. The combination of new column chemistries and systems hardware provided the scientist with chromatographic data that was faster, more sensitive, and with greater resolution. What the scientist now had was more chromatographic data in seconds to minutes depending on the sample complexity.
With this new technology the design and development teams had an expansion of most of their old sensors which were now able to provide greater ranges and sensitivities, and with new additional detection techniques added to their knowledge base. The increase of higher quality data that is delivered much faster allows for faster process knowledge and process optimization for the next potential drug.
So, imagine all these sensors deployed to look at all aspects of the process. Each one delivering its expert interpretation of the process at specific time points. Each sensor providing specialized information as well as the scientist knowing the ranges and/or limits of where it can be the most beneficial in controlling the process. This process knowledge can now layout the control process based on the need of measurement time, measurement tool, and control action that is needed based on look-up table values in determining what, if any, action must happen to the process. As part of this process looking at the output of the individual sensors is the utilization of additional software tools (chemometric, MVDA, Data Historian, Control software). Again, going back to Dr. Wu’s statement – the need for the integration of all sensor data and control so that it can linked to the additional analysis that the additional software tools provide. Depending upon the company approach about control packages as well as the vendor they choose I’ll refer to them as a PAT software hub. These PAT software hubs can integrate into the various sensor vendors hardware. The point is that based on their choices the scientist has a wide range of software and hardware choices that have to be integrated into the suite of communication and control packages.
Summary: Workflow = Sampling (#of steps/hardware and software) + Sensors (# utilized hardware and software) + Data (# utilized) integrated and incorporated into the entire process from development to manufacture.
Does this make business sense?
If done properly raw material and inventory costs decrease, you can also see a reduced facility footprint, and reduced energy cost.
Plant utilization, efficiency, quality, yield, flexibility, mobility, manufacturing expansion through reduction – all increase. Another COVID-19 example is if a pharmaceutical manufacturer had a critical drug that was needed as one of the medications that was utilized as a treatment – they could either changeover to just manufacture that drug faster, safer, and with higher quality if they had implemented PAT CM into their operation.
Technology continues to evolve in small ways where sometimes only a tweak is needed. But when the scientist comes up against a new challenge in the workflow sometimes a major innovation is required. In this brief discussion over the past 16+ years much has happened. The FDA dream in some instances has already become a reality. For most the journey continues but due to the insight of those in the FDA and those in the industry that had a vision of how it might be done they laid the groundwork for PAT and CM in becoming a reality. We opened the discussion with the mention of COVID-19. As we look back on this horrible year for the entire world and then you think of what the many scientists and engineers accomplished in the development and manufacture of new vaccines it is truly remarkable.
In closing, the discussion of drug shortages, drug recalls, raw material shortages what you are seeing is an expansion of new companies that are tasked with solving this problem which has been plaguing the FDA and the industry for years.
Going forward these things can now be accomplished because of much of the innovation we have briefly discussed in this article. Due to to all who have made these breakthroughs and discoveries a new reality we can celebrate in 2021!