Process Analytical Technology: Charting Progress in Data Analysis

When FDA first announced in 2002 a new initiative, Pharmaceutical Current Good Manufacturing Practices (CGMPs) for the 21st Century, and later issued its report, Pharmaceutical cGMPs for the 21st Century—A Risk-Based Approach, in 2004, it put into motion an effort to enhance product quality and modernize pharmaceutical manufacturing through a science- and risk-based approach under quality-by-design (QbD) principles (1). That effort was further encouraged by the issuance of guidance on process analytical technology (PAT) in 2004 to facilitate the introduction of new technologies that would enhance process understanding and assist in identifying and controlling critical points in a process (2). These technologies include: appropriate measurements devices, which can be placed at-, in-, or on-line; statistical and information technology tools; and a scientific-systems approach for data analysis to control processes to ensure production of in-process materials and final products of desired quality (3).

Illustration by Dan Ward

So how far has the industry come in advancing PAT and where may future innovation lie? Pharmaceutical Technology conducted an industry roundtable to gain perspective on advances in analytical instrumentation and methods development. Participating are: Tim Freeman, managing director of Freeman Technology and past chair of the Process Analytical Technology Focus Group of the American Association of Pharmaceutical Scientists; Kevin Aumiller, TOC product manager at GE Analytical Instruments; Andy Salamon, senior staff scientist and customer advocate at PerkinElmer; Chris Heil, product specialist, Antaris NIR analyzers at Thermo Fisher Scientific; and Scott Samojla, senior director of PATROL process systems at Waters.

Advances in PAT

PharmTech: What would you identify as the most significant advances in PAT used in the pharmaceutical industry during the past five years?

Freeman (Freeman Technology): Before focusing on specific technologies, I'd like to highlight what I believe is an important shift in the perception of PAT over the last five years. Initially introduced by the regulators in 2004 to support efficiency in development, manufacturing, and quality assurance, PAT has, in the intervening years, progressively proven its potential value. Collective and individual experience has led to a widespread appreciation of the integral role that PAT will play in achieving new levels of manufacturing efficiency and product quality, and as a result, the uptake of PAT is now increasing significantly, most notably for the information-gathering needed to support the development of more efficient processes.

Industry roundtable participants, from left to right: Tim Freeman, managing director of Freeman Technology and past chair of the AAPS Process Analytical Technology Focus Group; Kevin Aumiller, TOC product manager, GE Analytical Instruments; Andy Salamon, senior staff scientist and customer advocate at PerkinElmer; Scott Samojla, senior director of PATROL process systems, Waters; and Chris Heil (not pictured), product specialist, Antaris NIR analyzers, Thermo Fisher Scientific

This focus on information-gathering and process monitoring is reflected in the technologies that have advanced most during the past five years. The use of on- or in-line systems with a proven track record has increased considerably but so too has the application of at-line methods that provide unique information and insight.

In- and on-line systems deliver value by enabling the real-time tracking of process behavior and the changing properties of in-process materials. Today, the pharmaceutical industry is making greater use of some traditional on-line techniques, such as pH measurement, and also newer real-time technologies, such as particle sizing, and, of course the spectroscopic methods such as NIR. In the at-line arena, I'd recognize particle-imaging and bulk powder characterization systems, as examples of technology that are really advancing understanding. By providing the information needed to rationalize particle and powder behavior, these technologies support the attainment of better powder processing performance, which is vital for greater manufacturing efficiency.

Heil (Thermo Fisher Scientific): NIR spectroscopy has matured into a common technique for PAT analysis across the whole pharmaceutical manufacturing life cycle from raw material identification to granulation and drying to blending and tablet production. The advent of fit-for-purpose and total solution NIR analyzers has unlocked the full potential of NIR spectroscopy as a process analytical technology for monitoring and controlling pharmaceutical production processes. We have witnessed the development of dedicated PAT NIR analyzers during the past five years, which offer total process analysis solutions. The evolution in design was witnessed not only in the NIR analyzers but also in the software, accessories, and probes required for a total analysis solution. Software integration and process communication also are key aspects of a total analyzer package.

Experience has shown that the most challenging and often most important facet to a successful process analysis is representative sampling. The interface of the analyzer to the sampling point is critical since this is where the NIR spectrometer interrogates the sample for analysis. A perfect example is the challenge in getting a representative sample on a probe installed in a fluid-bed dryer when the product is being fluidized and the moisture level dictates whether the product will coat or foul the probe window. Advances in fiber-optic probe design have led to self-cleaning probes with purge air or retraction mechanisms to prevent probe fouling. In addition, probes that use side-view windows, angled tips, and curved, cupped design have been developed to ensure that representative samples are in contact with the probe window. These advances in NIR analyzers, software, and accessories have furthered the application of PAT for production process optimization, reduction in production costs, and improved product quality through timely measurements of critical quality attributes.

Samojla (Waters) : The need for an improved scientifically based understanding of the manufacturing process has been and will continue to be key to future success in the pharmaceutical industry. The advent of ultra-performance liquid chromatography provided the power to not only meet these needs within the laboratory, but enables on-line chromatographic use for process monitoring and control. The benefits of this approach reduce process variability and risk while adding automation and information not previously available on-line.

Aumiller (GE Analytical Instruments): A prime example of PAT advancement has been associated with equipment cleaning and release. Clean-in-place (CIP) systems, which have historically been used in the biopharmaceutical space, have gained traction within traditional pharmaceutical facilities and contract manufacturing organizations. This has been promoted by the unprecedented merger and acquisition activity seen in the past three to five years. With an industry focus on increasing plant capacity and reducing overall operating costs, facilities are taking on more products, which in turn, spawns the need for rapid equipment changeover. The implementation of CIP to replace manual cleaning has provided significant opportunities for efficiency in this area.

Offering the benefits of automated, predictable, and repeatable cleaning cycles, CIP systems satisfy the main objectives of PAT by building quality into the process and providing real-time assurance of quality. The critical quality and performance attributes associated with cleaning can be monitored and controlled continuously. For example, the temperature and concentration of cleaning solutions can be monitored with in-line conductivity probes and temperature sensors. The agitation of the vessels can be actively controlled with flow meters and pressure transducers. With the use of process logic control (PLC) systems, sequences can be timed to provide consistent exposure to the cleaning agents and rinse water.

Once cleaning is complete, indirect sampling of final rinse water can be performed in near real-time with at-line total organic carbon (TOC) analyzers and in-line conductivity sensors. Direct swab samples can also be analyzed with at-line TOC instrumentation. TOC instruments provide detection of residual product and other non-ionic species present in the process. The conductivity sensors can detect the presence of ionic contaminants, such as cleaning agents in the rinse water. The combination of test methods provides expedient feedback to release the equipment for use.

The future of PAT

PharmTech: Looking ahead five years, how do you see PAT evolving in the pharmaceutical industry?

Freeman (Freeman Technology): While the use of PAT for information gathering and process monitoring, particularly in development, has grown substantially during the past five years, there is still some way to go in terms of optimizing its application across the whole manufacturing cycle and its utilization as a mechanism for process control. In the next five years, I expect PAT to penetrate further into commercial manufacture as confidence in its use and capabilities expand.

The automation of process control is an ongoing trend and will continue to motivate suppliers to bring new technologies on line. On the other hand, experience from other sectors would suggest that the sophisticated application of the most relevant at-line techniques is an extremely complementary and productive approach. It, therefore, seems likely that the focus should and will remain on technologies that truly deliver in terms of relevant information whether at-, on- or in-line. Parameters, such as particle morphology, and powder flowability, compressibility, shear strength, and surface area are all likely to remain pertinent for powder processes.

Heil (Thermo Fisher Scientific): In the pharmaceutical industry, PAT is well adopted among the largest pharmaceutical companies, especially in the more developed countries. In the next five years, I see the evolution of PAT into smaller pharmaceutical companies, contract manufacturing organizations, dietary supplement manufacturers, and companies in the less-developed world. Companies in these regions and industries have seen the progress that the larger pharmaceutical companies have made and see the value of PAT tools, such as NIR spectroscopy, for monitoring and controlling their processes to improve quality and lower production costs.

Advances in data analysis and data management are critical to the future success of PAT. Managing the sheer volume of real-time, multivariate data generated from a production process so it can be used for real-time process adjustments or off-line production analysis is a challenge. During the last 10 years, significant knowledge has been gained on how to properly implement process instrumentation for on-line analysis.

For many companies, the starting point for implementing PAT tools was moving familiar laboratory spectroscopy techniques from laboratory to line. The value of PAT is built around real-time analysis of critical quality and performance attributes to improve process quality. For a pharmaceutical process, this means combining data from multiple sources to get a complete picture of the overall quality of raw, in-process, and final products. This approach requires advanced data-management systems capable of receiving information of various formats from multiple analyzers and production automation systems. Take for example, the data management required when applying PAT to the pharmaceutical hot-melt extrusion process. Real-time data of multiple formats and sources from a NIR analyzer, barrel temperature probe, pressure sensor, extruder-screw speed, and ingredient feed rates would need to be archived in a common location for real-time, multivariate data analysis as well as off-line post processing analysis.

With future advances in data-analysis and data-management software, companies will be able to focus their PAT efforts on optimizing their sampling at point of analysis, chemometric models, and multivariate analysis tools, thereby allowing them to apply PAT to more challenging production processes.

Samojla (Waters): Real-time chromatographic analysis will increasingly enhance manufacturing processes—development, transfer, monitoring, and control. Integrating these enhancements within a unified information access model capable of very high-speed analytics will yield improved decision management, predictive analytics, and comprehensive analysis—without the paper. This integration can be achieved by deploying real-time analytics while eliminating the time delays, analytic, and process variability associated with traditional off-line analysis.

Aumiller (GE Analytical Instruments): In the next five years, the adoption of PAT will likely be very evolutionary in nature. Existing processes will be incrementally changed to provide for better efficiency. It is unlikely that a disruptive technology will be readily adopted that totally replaces an existing process.

Within the context of cleaning systems, we will likely see the development of better tools and analytics that allow for more optimal control of the processes. As indicated earlier, many cleaning processes currently employ the use of conductivity and TOC measurements as a check of equipment cleanliness. This testing is most often performed in the laboratory but also can be performed at-line to speed the time to result. With the use of robust and fast-responding on-line instrumentation, CIP cycles could be designed to run until contaminants drop below a predefined threshold rather than running for fixed periods of time. With the immediate feedback afforded by integrated conductivity and TOC monitors, processes could become quicker and require lower utility costs to achieve the same or better quality. This transition will require the use of instrumentation that can reliably detect the contaminants of interest within the complex matrix of a cleaning process.

Salamon (PerkinElmer): To consider the potential advancements in analytical instrumentation, I would look at other significant advances in pharmaceutical development and not just at PAT advances. Nanotechnology drug delivery is the biggest advancement in the past five years and will be the biggest advancement in the future five years. Currently proven scaled-up developed manufacturing practices and processes do not exist to consistently produce, on a large scale, nanopharmaceuticals. In the development laboratory, single particle-inductively coupled plasma–mass spectrometry (SP–ICP–MS) of nanomaterials is the most significant improvement. This is not in-line, on-line, or at-line testing, but is a significant advance in analytical testing.

References

1. FDA, Pharmaceutical cGMPs for the 21st Century—Risk-Based Approach: Final Report (Rockville, MD, 2004).

2. FDA, Guidance for Industry: PAT—A Framework for Innovative Pharmaceutical Development, Manufacturing and Quality Assurance (Rockville, MD, 2004).

3. FDA, Progress Report on Process Analytical Technology, www.fda.gov/Drugs/DevelopmentApprovalProcess/Manufacturing/QuestionsandAnswersonCurrentGoodManufacturingPracticescGMPforDrugs/ucm072006.htm, accessed June 18, 2012.