Response by Ian Clegg and David Hughes of the ABB Centre of Excellence for PAT, located in Daresbury, UK
The term PAT can be (and is) interpreted in many different ways by people in the industry who are typically from a variety of technical backgrounds working for a variety of companies. Such diversity is understandable and, indeed, should be encouraged since individuals will apply a specific interpretation that is most appropriate for them in their individual technical context and within the constraints of their manufacturing environment.
The industry can be categorized in many different ways, such as by therapeutic area, geographical location, business area and process technologies. Perhaps the most important division (at least in the context of PAT) is between chemical synthesis and fermentation / biopharmaceutical production. The division is important in the sense that processes that rely upon chemical synthesis (and subsequent formulation) are, in principle at least, similar to the production of chemical and fine chemical materials and these have been a fertile proving ground for the techniques and methodologies that underpin the PAT initiative.
In traditional organic synthesis and the more straightforward bioprocesses such as penicillin production, much work has been done to establish the fundamental mechanics of the processes and provide optimized production conditions. However, for the more complex bioprocesses, there has been insufficient time and experience to allow the subtleties of the problem space to be explored. Therefore, process performance is dictated (in general) by empirically derived operating parameters that may be sufficient to ensure production, but are insufficient to ensure consistency and high yield.
Many factors may change the operating performance and trajectory of any batch process and controlling these at a macro level will inevitably result in processes that perform with inherent variability. To improve, we must see through the readily apparent process information to set combinations of conditions which select the metabolic routes and biological response which optimize process performance and this includes the whole chain from bioproduction throughout separation and purification.
This will require a complex blend of advanced process understanding to expose the critical and measurable biological responses and pathways to encourage the behaviors required, advanced measurement science to provide robust measurements (or analogs that describe the behavior), and sophisticated control to execute the multivariable blend of conditions which from moment to moment describe the correct trajectory of the batch.
This is a challenging problem and, although the activity may be sophisticated in scope the bioprocess industry is exploring the possibilities with the same vigor as the traditional pharmaceutical industry. Although PAT-based processes may be seen first in traditional pharmaceutical production, there can be little doubt that the same techniques will be adopted by biopharmaceuticals and offer clear opportunities in terms of yield, purity, throughput and efficiency.
