No, I am not going into a political discussion . . . “lipstick on a pig” is just a saying that we all heard during the latest presidential campaign. I take it to mean dressing something up to look better or merely giving it a different name to change peoples’ opinion of it. This could be true for PAT (mostly known as Process Analytical Technology or Technologies, as per the FDA Guidance), if you think about it for a second. Over the years, I have heard roughly 75 different interpretations of what PAT actually means; and the opinions run from “this is the greatest thing since sliced bread” to “it isn’t anything new.”
The dissention started right after the term was introduced by Ajaz (like Madonna, do we really need a last name?); the “analytical” term was debated as much as the recent Health Care Reform bill. Ajaz insisted that it should be used in the broadest of its meanings and that we shouldn’t dwell on the instrumentation used for any particular measurement. He also (still) insists that the measurements, per se, are not PAT. However, as I go to meetings around the world, I meet people who have many interpretations of PAT.
One client insisted that his company wanted to do QbD (Quality by Design, for those of you recently returned from Antarctica), but didn’t think PAT was needed. Of course, we discussed the question (heatedly) of how one’s well-planned QbD process was running according to plan if there were no measurements along the process. And, after all, isn’t that all that QbD and PAT are about? Control?[pullquote]
Once, during a course in the Netherlands, the head of the Formulations group from a large manufacturer stated, “We don’t need PAT; we make good products now.” My answer was also in the form of a question. I simply asked, “Have you ever had a product lot fail specs?” When he answered, “Of course,” I simply said, “Then you are not in control.” He thought for a minute and then told me to continue. That is the whole concept that we are seeking: CONTROL! And control means no mistakes.
Many a bride and groom may get lost in seating arrangements, flowers, dinner plans, the “correct” band, the dresses, and honeymoon plans and forget that they are setting a “process” in motion. A process where their lives are joined (much as in making a solid dosage form) into a new and different “product”. The “PAT” here is an understanding of what is being gained, lost, and compromised . . . not what the first dance is or who sits next to whom.
In a similar vein, the focus of most PAT talks is the instrument used and data gathered. There is seldom a connection in the talk with the full product process. What I have noticed is a two-tier system: the holistic, theoretical approach by the FDA and other agencies, full of platitudes and encouragement (some critics say “smoke and mirrors”), but seldom specific points; the second is the microscopic approach where the overview is lost in the specifics. One sees only a forest while the other only sees pine cones, and never the twain seem to meet.
Despite all the conferences and meetings and publications, we are still in a silo mode. The R&D, formulation, scale-up, production, and stability phases are still treated as if they were attached to completely different products. At some point, more than the few early adopter companies will recognize that the instruments used to generate the data used to control the process are just tools. What I keep hearing makes it sound as if PAT is a stand-alone, separate entity, divorced from QbD. To help this misunderstanding, I offer this simple solution: let’s change the words associated with PAT. [A similar thing happened with HPLC: the “H-P” is interchangeably called High Pressure, High Performance, and even High Priced.]
In my PAT/QbD courses, I have renamed PAT, Product Assurance Technology. It sounds hokey, but notice that “analytical” is now gone? Now, all we have is a concept to make an acceptable product. So, now, under the new nomenclature, we can ask, “Is ______ PAT?” By _______, I can fill in a myriad of tests. If I place a Raman or NIR probe into an API synthesis to monitor/control polymorphism, is it PAT? Now, under my definition, it certainly is! What about qualifying incoming excipients by NIR? Of course, it is. Placing a monitor on blending or drying or granulating? Yes, yes, and yes.
Now that we used the title cosmetic on that ungainly creature (PAT), is it pretty enough for all of us to want to date it? Instead of trying to define PAT as a stand-alone project or entity, we can simply recognize it as a TOOL to enhance the quality of our products, to speed up processing; and to reduce costs/waste. In other words, a means to implement a QbD project.
Now, pucker up and start measuring.