Integrating PAT into Processes

Published on: 

ePT--the Electronic Newsletter of Pharmaceutical Technology

San Antonio, TX (Nov. 1)-Though the US Food and Drug Administration's final guidance on process analytical technology (PAT) was published in Sept. 2004, companies are still unsure about how exactly to implement PAT in their processes.

AAPS, San Antonio, TX (Nov. 1)-Though the US Food and Drug Administration’s final guidance on process analytical technology (PAT) was published in Sept. 2004, companies are still unsure about how exactly to implement PAT in their processes. A symposium at the 2006 Annual Meeting of the American Association of Pharmaceutical Scientists in San Antonio, Texas entitled “Process Analytical Technology for Pharmaceutical and Biotech Products: Current States of Implementation and Future Trends” aimed to offer guidance for implementing PAT.

PAT: a framework

PAT is a system for designing, analyzing, understanding, and controlling manufacturing processes using timely measurements and critical quality and performance attributes. “The two key words are ‘understanding’ and ‘control,’” noted D. Christopher Watts, PhD, team leader of standards and technology at FDA (Rockville, MD,

“In my view,” said James Spavins, head of regulations and chemistry, manufacturing, and controls (CMC) at Pfizer (New York, NY,, “PAT fits into building an integrated pharmaceutical quality system. Everyone benefits from integrated systems.”

Gerd Fischer of Sanofi-Aventis (Paris, France, added that the PAT system approach comprises four pieces: design for quality, sensors, data analysis, and process control, which is the “ultimate goal.” The objective of PAT, he said, is unit-to-unit consistent quality.


Three challenges often stand in the way of PAT implementation, Spavins noted in his presentation, “Meeting the Regulatory Challenges Associated with Process Analytical Technologies.” First, companies must “get over the barrier of learning” and become comfortable with the technology. Second, datasets are difficult to visualize. “There are lots of ways to look at data. There’s no one way,” said Spavins. “It depends on what it is and what factors you’re looking at.” Third, transition and translation can be problematic.


To help address confusion about how to implement PAT, Fischer provided examples in his presentation, “Experiences from a PAT Project and Implications for Industry Continuous Improvement Strategies,” of how Sanofi-Aventis has implemented PAT in processes such as fluid-bed process monitoring. He suggested that a project-realization methodology for PAT includes (from laboratory to production conditions): feasibility studies, scale-down studies, model validation, deployment, data collection, and process control.

Moving forward

Watts explained that several approved manufacturing processes have incorporated PAT such as those for branded, generic, and over-the-counter products. Approaches for regulatory approval ranged from comparability studies to annual reports.

Throughout his presentation, Watts stressed the following points for clarification:

  • Implementing PAT is voluntary, and FDA will be flexible about how companies integrate the technology into manufacturing processes.

  • Many options for PAT implementation exist, and FDA is open to many processes. Though many companies are looking for a “how-to” guidance, the agency did not make specific guidelines for implementation because a “cookie-cutter” approach to implementation will not be successful for all processes.

  • Automation does not necessarily equate to control.

  • Research data should be collected without the fear that FDA will examine it at an inspection.

Watts also outlined a strategy for achieving PAT implementation that includes the next round of training, continuing education (through seminars, workshops, and lecture series), an Interaction Working Group (with the objective of fostering interaction between CMC reviewers and CGMP investigators), and continued advanced training for the Pharmaceutical Inspectorate.