By Terry Redman, Principal Scientist, Mettler Toledo AutoChem

There have been a number of recent articles highlighting the disappointing adoption rates of PAT and QbD within the pharmaceutical industry. With most new ideas and new technologies, making the leap from innovators and early adopters to the early majority requires the complexity of the product to be significantly reduced to make the product accessible and the value obvious. With PAT and QbD, it does not seem we have not been able to sufficiently simplify the guidelines, the technologies, the instrumentation, and the overall concept to the point that it can be readily adopted throughout the industry. Rapid industry-changing innovation requires both top-down and bottom-up acceptance–and so far I’m not sure we have either.

And while there are success stories highlighting benefits of PAT and QbD, there are still many challenges to getting buy in–including the high level of technical specialization required to implement these solutions, and not least of all the regulatory shadow that continues to make the risk vs. benefit analysis extremely difficult. Economic justification based on gains of process efficiency and improved yield and throughput, does not usually take into account the cost of changing the deep-rooted industry culture.

So in an understandably risk-averse industry, it is seems natural that widespread adoption has not happened overnight. At the same time, the trends of fewer blockbusters and increased generic competition (for both small molecules and biologics) scream out the need for improved process efficiency to remain competitive in the long run.

So where does one begin?

An article by Dominique Hebrault proposed that scaling up a chemical reaction without reaction monitoring was the equivalent of driving a car with your eyes closed. Let me push that further by asking: Have you ever planned a road trip using a map; lined out a detailed route; and then jumped into your car and drove the route from memory with your eyes closed?

Probably not, and though the analogy may be exaggerated, it does help illustrate how QbD and PAT together  are intended to get us to our destination of a quality final product as quickly and efficiently as possible.

A New Analogy

Actually, I find an even better analogy is that of PAT providing the benefits of a global positioning system (GPS) in the manufacturing process. (I used this analogy in a white paper on crystallization scale-up, but I think the story fits the general application of PAT as well.)

Consider that QbD is primarily focused on laying out a detailed map of the manufacturing process–mapping the complete operational design space and highlighting critical features that must we must pay attention to on our journey. It should provide information on the optimal path and on potential roadblocks. Ultimately, it may describe the precise path we need to take to get to our destination without fail under normal circumstances.

Quality by Design, properly applied in the development process, should give us the process map needed to get from point A (raw materials) to point B (finished product).

But what happens if the unanticipated happens? The known unknowns and the “unknown unknowns” such as undetected impurities in feed materials, operator error, and the inherent variability in any physical or biological process inevitably mean deviations in our process route. In the end, even a slight undetected shift in direction can leave us a long way from our desired final destination.

It is difficult to conceive of QbD without connecting it to the application of PAT from development through to the manufacturing process.

If QbD  provides a map and PAT can potentially act as a GPS, consider that before you begin any trip you need to know where you are starting from, where you want to go, and what the possible routes are that can get you to your destination.

To blindly follow a route (our batch recipe), we need to assume our map is correct and we must be confident in our starting position–or the process will fail and we might end up in a ditch.

We can carry out QC on starting materials to provide assurance of our starting point. Once we know where we are starting from we can begin our journey.

As we proceed, we have the choice of many different instruments to help us get to where we want to go–potentially a speedometer, an odometer and maybe even a compass. These are measuring our critical process parameters, helping to ensure we proceed safely and stay on the right track.

Provided the map is accurate and we have accurate measurement and control of these critical parameters, we should end up close to our destination–possibly within acceptable limits.

But the only way to be absolutely sure we are on track is to be able to directly measure our true position in real time-our critical quality attributes.

(I should also point out that sampling and analysis during the batch is like stopping every once in a while to ask for directions . . . that can also keep you on course, but it will almost certainly slow you down and the quality of information can vary greatly depending on the source of the information.)

Where You Need to Be

The use of PAT instrumentation that can measure quality attributes in real time is like the use of a GPS during your road trip. Used in conjunction with QbD, the application of PAT instrumentation contributes to a more accurate map to begin with. During the manufacturing route, PAT provides assurance that you are where you need to be. It can even help guide you to where you need to go next if there is an unexpected disturbance or detour.

Admittedly, I have stretched the analogy pretty far, but I like to note that most automobile navigation systems actually combine a map, compass and GPS to be an all-in-one solution for getting you where you need to go. Likewise, PAT, QbD, and the regulatory environment need to work seamlessly together to drive efficient manufacturing of quality pharmaceuticals.

The funny thing is, it probably won’t be long before we have widespread access to cars that can drive us safely to our destination with little manual intervention. Will we see automated control of pharmaceutical product quality on a routine basis by then?

Related posts:

1. How Can QbD’s Progress Be Measured?

   Survey data say that PAT and QbD are not gaining traction, but are there other ways to measure drug manufacturers'...

2. Dangling a Carrot: Should PAT Be Mandated, or Incentivized?

   At a May 11-12 conference on PAT and QbD in Bethesda, CDER deputy director Keith Webber responded to questions about...

3. QbD’s Conundrum: Too Many Cooks?

   Novo Nordisk’s Erik Skibsted takes a whimsical look at what ails QbD: the diversity of its practitioners. That is, its...

4. Notes from ISPE: Abbott, Pfizer Present PAT Applications

   At this week's ISPE annual meeting, Abbott’s Eric Correa and Pfizer’s Sarah Mancini shared PAT applications and successes. ...