PAT — a place for separation science

FDA's PAT guidance is, no doubt, increasingly important in today's pharma and biopharm industries. The fact that there are still more questions than answers on driving broadscale change and reaping the benefits is a very good sign that things are advancing. There is greater added value in process understanding and consequent process improvement, rather than implementing strongly recommended or legally enforced procedures.

Dr Michael Kraft

A stepwise approach. Some workflow steps in the biopharm value delivery chain process can be addressed to implement the principles of quality by design (QbD) within a PAT framework. As the industry makes the appropriate choices, measurement solution suppliers are deciding how to most effectively support their customers' workflow and achieve their business goals by offering hardware, software, consulting and support.

Remove variability, then optimize. The increasing numbers of Six Sigma projects indicate that they can become the chosen method in PAT biopharm production investigation projects. This approach systematically identifies the most important process step; the one creating the greatest variability in the overall process. The need for measurement may be reduced in the long-run with the help of chemometrics, which defines the process steps and supports a predictive model. During the implementation of Six Sigma, a short-term positive side-effect may result in finding 'quick wins', which may be assessed and implemented without major changes, and are reversible — should they fall short of expectancy levels.

Do not try to boil the ocean. The Pareto principle (80% of problems are caused by 20% of root causes) already has many quality control applications. Dramatic improvements can often be achieved by identifying the core root causes. Where to start? There are many degrees of freedom in the process steps. A decent number of root causes that lead to variability in yield could be identified by measurements on a macroscopic level. So in the PAT real world of measurement it could be asked: "Will PAT implementation exclusively permit the use of process spectroscopy, for the purpose of measurement?"

Know your limitations. The advantages seem obvious for the biopharm industry: noninvasive testing equals no contamination problems. Looking into a fermentation vessel through a window or with the help of a fibre optic seems to be the method of choice. Yet, complex mixtures may require congruous and sensitive measurements to produce a comprehensive picture and an accurate answer. Realistically, there are probably many parameters in a bioprocess that contribute to the 80% impact on the fermentation process, which cannot be detected by noninvasive measurement technology.

Walk before you run. A large number of process and quality related measurements are still performed by drawing a representative sample from the process, and separating and determining compounds by chromatographic or electrophoretic methods. PAT will not eliminate this process in the long-run. Although it takes time, it is definitely worth the effort. The actual speed of analysis, once the sample reaches the chemist, is usually not the limiting factor. HPLC, which has been the workhorse technology for many years, has undergone a metamorphosis to become a race horse.

Which risks to take in a risk-based approach. There is huge potential in implementing PAT to understand processes and to drive QbD — and it has not yet reached its peak. Both noninvasive spectroscopic methods, as well as online sampling, combined with chromatographic separation and subsequent UV or MS detection will find their place in PAT. Combining data and information will be instrumental in developing and fine-tuning the chemometric models to support predictions. For example, unattended sterile sampling will replace manual sampling; eventually other analytical techniques, such as rapid resolution liquid chromatography, LC/MS and even HPLC-chip technology will be simplified and become more robust. With appropriate standardization from an interfacing point of view, separation science will find a broader audience in PAT. This will realize the benefits of moving instruments in or closer to production to eventually achieve the ultimate goal "realtime release", yet leaving appropriate space for flexible and sophisticated analysis in the laboratory for the time being.

Dr Michael Kraft is industry marketing manager pharma/biopharm solutions at Agilent Technologies R&D and Marketing GmbH & Co. KG (Germany).