Process analytical technology (PAT) has been the subject of regular discussion since the FDA released its guidance for the industry in September 2004. The goal, then and now, was to provide a mechanism to design, analyze, and control pharmaceutical manufacturing processes through the measurement and control of critical variables such as moisture content, particle size, temperatures, and a host of others.
Quality by design (QbD) and Real-Time Product Release (RTPR) are more recent variations on the theme, but PAT is vital to these, according to Brian Stephens, North American PAT coordinator at ABB, who was a participant in the recent IFPAC/INDUNIV “PAT Summer Summit” held in Puerto Rico.
PAT is a vital tool for drug companies, according to Stephens, as they work to achieve the goals that the FDA describes in its publication “Pharmaceutical cGMPs for the 21st Century—A Risk-Based Approach.” Among FDA’s goals is to encourage the use of the latest scientific advances in pharmaceutical manufacturing technology and to employ the latest quality management techniques in the process.
In short, the FDA wants pharma manufacturing technology to catch up with other industries, and to further assist industry in their efforts, the FDA has released “PAT—A Framework for Innovative Pharmaceutical Development, Manufacturing, and Quality Assurance” and other guidance.
With a notable sense of irony, Stephens recalls a Wall Street Journal article that claimed potato chip manufacturers use more advanced manufacturing technology than pharmaceutical manufacturers. Silicon chip makers, he noted, have increased their manufacturing efficiencies by more than 10,000% in the past few decades by using new technology. In contrast, the pharmaceutical industry has used the same basic manufacturing processes for the past several decades, with the only improvements being made in the equipment, not the process.
Currently, Stephens says, pharmaceutical products being manufactured using traditional control systems require immense amounts of documentation to be released for sale. Laboratory-based QC systems now in use must process numerous variables, which must be interfaced, calculated and reviewed on a batch-by-batch basis. This is usually performed by hand, using printed data, resulting in a process that is much slower than the actual manufacturing steps.
“There are some comments being made that we don’t have this function or that we don’t have that operation to make these improvements, but the tools exist to complete a closed-loop control scenario today,” Stephens stated. He went on to describe how data management can fit into the process control architecture to provide a fully functional PAT system.