Broadening PAT's Manifesto

Nov. 17, 2004
The last word on PAT extends its use from a manufacturing to a cradle-to-grave tool for gaining process knowledge, FDA’s Ajaz Hussain says.
Anyone who has viewed process analytical technologies (PAT) as primarily a manufacturing tool is on the wrong track. The FDA made that point clear last month, when it released its final guidance on PAT (available on www.pharmamanufacturing.com). The document may appear little changed from the draft released two years ago. In its quiet way, though, FDA’s final word on PAT creates a mandate for revolution. “FDA is trying to turn the industry on its head,” says Gawayne Mahboubian-Jones, product development manager for OPTIMAL Industrial Automation (Bristol, U.K.) One explicit change is that the final guidance also covers biological products traditionally regulated by FDA’s Office of Biological Products, says Chris Watts on FDA’s PAT Policy Development Team. Other key differences are more subtle, however. One is the new document’s emphasis on product development. “Development” has even joined Manufacturing and Quality in its title: “PAT—A Framework for Innovative Pharmaceutical Development, Manufacturing and Quality Assurance.” The new document stresses even more than its predecessor the concepts of “process understanding,” which is addressed in its own separate section, and “innovation.” These shifts in emphasis provide a good clue as to where PAT is, or should be heading, says Mahboubian-Jones. “Process understanding is at PAT’s core, and you can’t get that understanding unless you implement PAT at the developmental stage of any product,” he says. Companies such as GlaxoSmithKline and AstraZeneca, where FDA’s PAT policy development team’s Ali Afnan worked before joining the Agency, have already taken this systemic approach to PAT. AstraZeneca’s research team in Mölndal, Sweden, for example, uses PAT “design of experiments” and multivariate analysis to evaluate potential drugs.For companies that have already considered PAT as a holistic product and process development tool, the final guidance provides affirmation. For others, incorporating PAT into the drug development process may involve difficult cultural change, Mahboubian-Jones suggests.There were other minor, yet telling, changes in the document as well. For example, the final guidance does not single out or even mention any specific type of technology such as NIR spectroscopy, and eliminates the discussions of vibrational spectroscopy-based sensors found in the draft guidance. This change was important for equipment vendors, says Nancy Mathis, president of Mathis Instruments, Inc. (Fredericton, New Brunswick), which manufactures thermal effusivity sensors. But, even for end users, she says, “those passages in the previous document dated it, since so many different types of sensors and technologies can be integrated into PAT programs.”Standards under developmentThe new document also refers to ASTM’s E-55 Committee, which is developing standards for PAT and whose members include drug manufacturers, equipment and software vendors, implementation specialists and pharma ingredient suppliers. So far, E-55 has completed its first standard, defining PAT terminology, in record time, and is working on a number of others. The committee’s work will establish standards that are relevant from a process perspective, says Ajaz Hussain, deputy director of CDER’s Office of Pharmaceutical Sciences and chair of the PAT steering committee. These standards, in turn, should evolve to support process analytics and PAT that will bring the deep process understanding that will be required for regulatory flexibility. “With standards, we won’t have to impose the old analytical or process validation criteria on a new technology,” he says.The fact that the guidance was finalized at all was extremely important, Mathis says, and sends more than a symbolic gesture to drug manufacturers. FDA had issued a draft guidance on blend uniformity a while ago, but never moved forward to a final guidance. Some people in the industry may have expected the same to happen with PAT, she suggests. “The final guidance represents a strong, positive step in the right direction,” Mathis says.No micromanagingDrug manufacturers hoping for explicit guidelines on how to prove the equivalency or superiority of PAT to traditional quality testing systems will be out of luck. Like the draft guidance, the final document merely provides a framework to allow companies to come up with their own solutions. “Instead of telling them how to do things, we’ve given companies all the flexibility they need,” says Hussain. “If they’re not willing to innovate, we cannot help them.”However, the document clearly stresses FDA’s desire for open communication, leaving it up to individual drug manufacturers to address their questions or concerns with the Agency. “Close communication will be a key component in this approach,” the document reads. “We anticipate that communication may continue over the life cycle of a product, in the forms of meetings, telephone conferences and written correspondence.”Hussain, a guiding force in the PAT movement, cannot understand why more pharmaceutical companies aren’t embracing analytics. “The critical part of the PAT guidance was very simple and I don’t think people read it often enough,” he says. Revalidation is not an issue, he adds. “With PAT, we said, ‘you don’t validate it, you control it, using validated controls.’ This approach makes the whole system modular and based on the level of understanding. It allows you to distinguish between what’s important and what isn’t, so you can focus on what is.”While the Agency can’t be responsible for innovation, Hussain says, it does want to hear drug manufacturers’ concerns. “Our team can meet with you, or you can have a preoperational visit so that, before you agree to anything formally, we can sit and talk about it,” he says. “The guidance is very clear about this and companies have already taken advantage of this—that is, some have and most have not. What is missing from our guidance is hard to comprehend.”Changing the regulatory cultureFor its part, FDA has been working hard to integrate PAT into its own culture. Its first PAT training program ─ not just for investigators but for reviewers, compliance officers and field staff, has progressed well so far, Hussain says. A second training program will start soon involving FDA’s National Science Foundation (NSF) center affiliates. These include Purdue University’s Center for Pharmaceutical Processing Research, the Center for Process Analytical Chemistry (CPAC) and the Measurement & Control Engineering Center (MCEC). “We’re going back to see what didn’t work, and we plan to expand the program to include more biotechnology and more control engineering concepts,” Hussain says.The PAT movement is also “going global.” “We’re talking with our regulatory colleagues in Europe, which now has a PAT team similar to ours. Once we analyze results from our first training program, we’ll share the information with them,” Hussain says. "As we move to the next training program, we’d like to invite Health Canada for training, and we will soon start a collaboration project with Japan."PAT’s not an elitist movementHussain dismisses the idea that PAT technology is for an elite group, or accessible only to larger companies. “If you don’t understand your processes, regardless of your company’s size, your manufacturing costs are already high,” he says. “Striking a balance between the two will depend on how innovative each company can be and how it makes business decisions.” In most cases, he says, companies that work out the short- and long-term cost projections will find that PAT can actually be much cheaper than what they already have in place.Stuck in corrective action modeThe industry needs to change the way it looks at quality, Hussain says. He points to QS 9000 and the way the automotive industry defines continuous improvement. “If you use attribute data to describe quality standards or specifications, you cannot have continuous improvement. By definition, the result will be corrective action,” he says. “This industry has been stuck in a corrective action mode.”“We have been so inward looking, or inbred in our thinking, that we’ve missed opportunities that have been there for 60 or 70 years,” he says. With PAT and its other recent initiatives, FDA is trying to push the industry closer to Deming. "Improvement is either continuous, or springs from innovation or corrective action." PAT focuses on innovation, the Manufacturing Science piece of the GMP initiative focuses on continuous improvement, and, collectively, they focus on corrective action. PAT provides a bridge to Six Sigma and other quality programs.For an in-depth analysis of the guidance document and industry’s reaction, see the first edition of Pharmaceutical Manufacturing’s new electronic newsletter, PAT Insider.