AutomationXchange Lunch Speaker: Pfizer's Lou Pillai

The thought for the day, as Lou Pillai started his discussion of how Pfizer is dealing with PAT and the FDA's new focus on continuing process improvement, is "Be the change you want to see in the world." Gandhi's phrase sums up Pfizer's change management around GMP for the 21st century. The top issues facing the pharmaceutical industry are: --Patent expiration (14 major drugs will expire next year) --Pressure to reduce drug costs. (The US spends over 10% of its GDP on health care, and the fastest growing component is drugs) --Challenges to Intellectual Property-- It costs an average of $800 million to bring a drug to market, and legislatures are considering reducing patent protection. --Counterfeiting and Piracy are rampant throughout global supply chains and it is very hard to "track back" and find out where they came from, or who made them. --Unmet needs: small populations needing more exotic drugs: The blockbuster drug era is over. --tremendous regulatory pressure The FDA has been complaining for years: --pharma manufacturing is inefficient, --the industry failed to innovate --compliance record is unacceptable There has been a new recognition on the part of the FDA that barriers to adoption of innovations and continuous improvement are at least partly their own fault, since they provide no incentive for process optimization or process innovation. The PAT framework (at Pfizer, at least) aims toward advances in quality systems and science, even though the GMP regulations themselves have not been changed since 1979. The FDA has started taking a process centered approach, rather than a compliance regulation centered approach. GMPs for the 21st Century was published in September 2004, but didn't update the 1979 Good Manufacturing Practices. FDA has a focus on risks to public health. They are now operating from a risk-based orientation, and are encouraging manufacturing science-based policies and standards. They are promoting an integrated quality systems approach, and even harmonization and alignment with other regulations. GMP for the 21st Century basically promised a "safe harbor" for innovation. No longer does the drug company fear being dinged because they changed from a system that is substantially in compliance to one that is better. But the regulations themselves still haven't been changed. These are fundamental shifts, Pillai contends. The FDA is shifting from corrective action to continuous improvement. They are encouraging a change to continuous quality verification. They encourage a shift from Quality by Testing to Quality by Design. Pfizer needs to deal with diverse global supply chains, new measurement methods, and provide more information, both role based and event based, about the process. "A lot of the information," Pillai pointed out, "is more valuable and longer lasting than the batch of product itself." Scalability is going to be critical for PAT processes in the future, for Pfizer. PAT at Pfizer is analytical technology used to gain more informaton on the process to identify sources of variability. It is NOT lab based, but rather based on continuous analysis and online analyzers. "We used to run the blender for eight minutes, and then send a sample to the lab," Pillai says, "and wait until we could certify the lot. Now we will use NIR in the blender to characterize the blend in realtime." PAT is a key enabler of Manufacturing Science at Pfizer. It is being used to establish product/process knowledge: CtQA. It gives Pfizer a process capability that translates into the ability to perform CtQA. And it provides significant process understanding. Technical barriers are dissolving and the ability to get realtime quality feedback from the process is at hand. There is the typical gulf between process and IT at Pfizer. Pfizer has islands of automation, and many automation systems have been installed as the result of platform-based initiatives ("Let's put this one in, and see what we can do with it.") Infrastructure planning is usually driven by point application planning. There are the typical domain silos, and the solutions are almost always tactical, not strategic, in nature. What Pillai hopes to create is a new strategic approach, new measurement system capabilities and collaborative processes. "In the MES space, it isn't unusual," he says, "for there to be twenty different solutions within one company." Yet, is isn't about the technology alone. Pillai avers, "It is more about the intersection of people, technology and process." What is there to do? First, standardize measurements, then standardize data. Identify process owners and users for the data. Put together a master plan, with key impact assessments on applications, architecture, and infrastructure. leverage existing infrastructure and "slipstream new capabilities." "Lay the foundation first, then build as you go," Pillai says. "Remember that you can't cross the chasm in two jumps. You have to build bridges to cross that chasm." What is needed is common PAT software, with standard modelers, interfaces and software systems. Also needed is standardized data: recipes, item masters, measurement data, production info, and to develop a lifecycle management model for data. "The analyzer vendors have organized their products into 'virtual stovepipes' of information," Pillai complained, "and we need to standardize schema for information exchange. Fundamentally we need to migrate to a single integrated architecture for Process Analyzers." "Remember, we aren't in this alone." --