PAT - The plant designer's view

The introduction of Process Analytical Technology (PAT) can be viewed from various aspects depending upon your role within the regulated production of life science therapies or more general biotech products. Bill Treddenick (Life Sciences Director) and Marcin Krawiec (Process Control Engineer) at Lorien Engineering Solutions offer the plant designer's view of incorporating PAT into a regulated production facility.

The majority of articles that have been published on the subject of PAT have centred around the QA/QC aspects of the regulated environment - and the effect of PAT within already established compliance frameworks. This article is written from the aspect of manufacturing plant design, and it reflects on how these changed working practices might affect the design of the facility and its services infrastructure. It looks at some of the issues in installing PAT within an existing regulated facility, and also some of the practical steps companies can take to ensure the full benefit to companies and to public health is delivered.

For the manufacturers of both ethical medicines and GMP-licensed biotech products, regulatory uncertainty tends to inhibit the introduction of change; whether the change involves a new technology or it is simply a prudent adjustment within the validated process, and this limits the scope for innovation and improvement.

Taking the life sciences model as an example, manufacturing processes, including the Standard Operating Procedures (SOPs) are often finalised around the Stage 2 clinical trials (or 1st in-man studies), and these processes become fixed and frozen from this point forward. The Registration File (the licence) for a medicinal product, or the Device Dossier for a medical device, lists and defines process parameters and quality attributes. There is a similar set of requirements when manufacturing GMP-compliant products that become the starting materials for regulated products. Most post-approval variations to that file require defined change control protocols to be followed, and this can involve a great deal of effort (cost and time), and will usually involve gathering additional supporting data, additional safety and efficacy testing or even new clinical trials. While this policy has produced good quality and safe products, inhibiting efficiency within drug production is not good for public health with regard to affordability.

During the product lifecycle, there are opportunities for process innovations to improve the product quality and cost performance, however they have normally been associated with high costs to change the registration file with the regulators, and for global products, many different regulators. Also, regulators have traditionally had an authoritarian approach when it comes to dealing with manufacturers, happy to state what they will not pass, but not so keen on contributing to the solution.

In more recent times, regulators such as the US Food and Drug Administration (FDA) have become more proactive, which is crystallised in their PAT Guidance (first published in 2004). This followed an initiative in 2002 entitled "Pharmaceutical CGMPs for the 21st Century: A Risk-Based Approach." This altered paradigm recognises that medicines are often too expensive, and it signals a shift from a culture of 'end of line testing' to one of 'in process analytical testing' which would be more familiar to the rest of the process manufacturing community. The word 'analytical' is explained in the FDA guidance as, "broadly to include chemical, physical, microbiological, mathematical, and risk analysis conducted in an integrated manner". The benefits of this approach are fairly obvious: avoidance of adding value to a batch that may not be released; the batch can be 'real time released' as opposed to spending time in quarantine pending release tests; detailed process data is gathered enabling process optimisation.

The key change required within the manufacturing environment is that production staff fully understand the quality needs of the product, rather than developing and following SOPs that solely capture the processing requirements, and leaving quality to the QC function. For example, the FDA is minded that if greater understanding of the critical issues is achieved by everyone involved in the manufacturing phase (and why they are critical to product efficacy and quality), then the manufacturing process will move from being task driven to quality driven, where the word 'quality' incorporates drug efficacy attributes. This lies at the heart of PAT.

As with any change, moving to PAT needs careful consideration, with input from specialists outside of the QA/QC team in order to produce a 'right first time' solution. An issue that can occur is that, in leading the project, the QA/QC team will consider what product attributes can and will be tested, along with the performance criteria (tolerances). Careful design work and application knowledge of instrumentation is needed to ensure where, when and how the process attributes can be monitored, along with practical tolerances. A second issue can occur with regard to the cleaning and sterilisation cycles, where in-line/ on-line instrumentation will need to cope with chemical action and elevated temperatures. A third issue comes with the physical positioning of monitoring instruments (within reactors, mixing vessels and pipelines) - sometimes there just is not enough physical space to get everything mounted correctly in order to obtain accurate and consistent readings. The final problem to handle is one of validation - ensuring there is a means to prove the function and accuracy of the instrumentation both at start-up and throughout the life of the plant.

The only way we have found to address all of the above issues is to both undertake a classical system validation route (emphasis upon the User Requirement Specification and the Design Qualification), and to conduct appropriate design risk reviews such as HAZOP (Hazard and Operability review) or an FMEA (we use the EN60812:2006 model for life sciences work), along with the normal GMP risk assessment.

So what are the positive effects of introducing PAT?

The first benefit is efficiency of manufacture, which will vary depending on the process and the product, and are likely to come from:

 Reducing production cycle times;
 Preventing rejects, scrap, and re-processing;
 Real time release;
 Increasing automation to improve operator safety and reduce human errors;
 Improving energy and material use and increasing capacity;
 Facilitating continuous processing to improve efficiency and manage variability.

Efficiency also comes from 'real time release'. 'End of line testing' drove the need to create space to store finished product in 'quarantine' pending QC testing and QP release. PAT and 'real time release' sees the product moving directly from the process to the distribution centre.

The next impact is on the QC function. Moving QC into the live manufacturing process should drive a change in QC function and in turn a different location for the QC facility. The current QC output is a PASS or FAIL against lab analysis, whereas in-line process measurement and any required corrections (within the allowable limits) are captured on the automated batch release record, and the finished product can be released upon batch completion. As more operational data is collected, process optimisation becomes a real possibility.

Another consequence of this new reactive manufacturing environment is the need for staff who are more attuned to in-process adjustments. Such people will ideally also have process technology experience, and can be central to making further process improvements. PAT offers a better platform for establishing a shared dialogue between Research, Quality, Manufacturing, Engineering and Business Systems functions.

In summary, the implications that flow from the introduction of PAT into an existing facility are:

 An altered approach to QA/QC, which drives a change in headcount, location and facility area;
 Real time product release (and a reduction in space currently assigned to pre-release quarantine storage);
 The introduction of operatives of a process automation/ engineering calibre;
 Changes to the facility with respect to process flow;
 Through headcount reduction, and product testing changes, the operation and design of cleanrooms may be altered, including HVAC design (as heat gains from people in cleanrooms and other GMP areas will be lower).

The above factors can lead to a review of the facility infrastructure with respect to media systems (heating and cooling water, purified water/WFI, drainage, process gases and electrical power distribution).

When fully embracing the world of PAT, we should not forget to include the facility design engineers in the dialogue to ensure that a thorough risk-based design process can be evidenced, and that the maximum facility benefits can be realised.

Bill Treddenick, Life Sciences Director
Marcin Kraviec, Process Control Engineer