Back in the late 1970s, I was forced, as the AR&D manager, to assume control of a “dicey” QC department at a non-Pharma company. I had to surrender my best “sous-chemist” to become the new supervisor, so I was a little annoyed. I addressed the QC group, telling them that they could insult me, my family, my pets, but two phrases would get them terminated: 1) “We’ve always done it this way” or 2) “We’ve never done it that way.” I pointed out that, if “this way” had worked, I wouldn’t have been forced to take over and have them do it “that way.”

How does this tale relate to the manner in which small molecule Pharma formulators make their products? Well, I have stated on more than one occasion, that someone who retired from a production job in 1965 could easily come back to work today and, with a few days’ training, be up to speed. If you think religious doctrine or Congress moves slowly to change, just observe pharmaceutical formulators, where it seems to be: “if it was good enough for dad, it’s good enough for me.” Spoiler alert: Why change any procedures when Pharma can charge whatever it wants to cover costs?

When the Pure Food and Drug Act, which created the FDA, came into existence, it made two assumptions: 1) physicians knew about drugs and 2) pharmacists knew how to put them into dosage forms. Considering the number of commercial drugs available at that time, the first part was true enough. Although today, an honest physician will admit they mostly know about drugs from company-sponsored seminars—trips to the Bahamas or Puerto Rico—and sales brochures. And, as for the personnel in a production facility, it is staffed and run by pharmacists.

That second part made sense when most drugs were made in the back rooms of corner pharmacies. They were rolled into little balls (pills) or tapped into capsule shells, both hand-prepared and labor intensive. The earliest real tablets were made in commercial facilities, but at a slow enough rate and in small enough numbers and formulations that the converted “corner pharmacist” could formulate via “hunt and peck” R&D. When I started in 1970, I used to joke that “method development” consisted of “if 0.5% magnesium stearate doesn’t work, try 0.75%.”

As more and more drugs were discovered and more and more doses were needed—a few thousand to a few million—the frenzied world of formulations was a “wild west” show. As pharmacy degrees went from two years in the “old days” to four years or five years, the pharmacists were trained, more and more, in school, not “on-the-job.” However, many of the same ideas persisted: Start small, work up, guess and pray for pilot studies to work; reformulate when they didn’t. And, along the way, some actually began using “Physical Pharmacy.” That is, some began to worry about what went on during the steps—blending, granulating—in a more materials-savvy way. 

Nonetheless, even today, you can step into a production facility, swing a dead possum while wearing a blindfold, and strike a dozen pharmacists, yet not be able to find a Materials Scientist with a small nuclear charge. Yes, this is slowly changing with the introduction of PAT and QbD, but mostly in the larger, proprietary companies. Generics and CMOs are still trudging along the well-worn path (rut?) as in years past.

Fortunately, the “path to enlightenment” (a Zen pharmaceutical approach?) is pretty much the same, no matter how large the company. I did not use the term enlightenment lightly. There are large gaps in the knowledge of the products we make and sell. The first is, we do not have control over our excipients and, in cases where the API is supplied by a client company, that, too. We still receive a shipment of, for example, lactose. We can either accept the certificate of analysis (CofA)—bad choice—or do some/all tests in-house. In either case, we can only accept or reject the batch. In very few cases, the company has some type of agreement with the supplier, based on knowledge of the process involved.
What can be done to increase our chances of getting a “proper” lot of excipients? Well, as mentioned previously, upon generating your PAT/QbD committee, include some RM suppliers. They need not be in any proprietary discussions, just be told that it will be, for example, a 5% API, immediate release capsule or a 15%, continuous-release tablet. As your formulators understand via design of experiments the parameters needed for the dosage form, the suppliers can become “partners” and work to supply the needed material parameters. As an example, if “flowability” based on shape, surface tension and morphology is important, the supplier(s) can (for a fee?) put aside any batch that meets the specification required. Worst case, the CMO has too many batches of “prime” excipient, which could probably be used for other, less excipient sensitive products. But, you would never be out of the best material or have to use a “lesser” material for your critical product.

As for the formulations portion of this infomercial, we need to examine the modus operandi, thereof. Assuming, for the moment, that we have experienced formulators building the dosage form. These people do preliminary pre-formulation studies, choose the release mechanism—immediate or continuous release tabs—dose size/API level, coating or no coating, etc. They proceed to perform bench-level experiments, then move on to pilot plant-sized batches, and finally, move to production-sized lots. There are several problems with this approach, namely, it is time-consuming and costly, may not develop the best dosage form, and you may not have the longest expiry date.

There is a tendency in our industry to have “silos.” One effect is that we treat the API as a fait accompli, with only one possible form for doses. The organic chemists who synthesize the molecule are not interested and seldom know about the destination of their product; they simply try for the highest yield. They seldom give formulators choices of polymorphs, salts (Cl-, SO4-2, as co-ions), particle sizes, etc.

When formulators perform pre-formulation, practicality forces them to merely make 50:50 mixes of each excipient with API, limiting the potential effects of various salts/polymorphs in the data bank. If formulators were allowed to “dream,” they would likely wish to see the effects of multiple interactions—e.g., API plus lactose, plus starch, plus talc. Sadly, using traditional pre-formulation techniques, involving sacrificing samples at time points and analyzing with TLC or HPLC, causes time constraints that do not allow multi-excipient or multi-API form testing.

If we extend the concepts of PAT/QbD, namely design of experiment (DoE), coupled with modern spectroscopy, we can, indeed, perform meaningful pre-formulation studies. If we use, for example, NIRS or Raman to examine samples, we can increase the number of mixtures by a large factor. The multiple mixtures can be examined, non-destructively, daily, searching for potential interactions. Any sample that shows changes can be analyzed and any degradation plotted versus time. The good news is that we may also see some excipients “protecting” the API from degradation. This approach will give far more information to skilled formulators to build the best formulation. It will also give hints as to the long-term stability of the eventual dosage form.

Assuming a full QbD program, this data almost obviates much of the work for a DoE on the finished dose. When the Design Space—parameters of pressure, mixing, etc.—is established, the exact production conditions of any lot may be coupled with results of stability tests for that lot. If we can show (validate) that a narrower set of parameters produces more stable products, we could apply for a longer expiry day, potentially saving a small fortune in product recalls. To fully utilize this approach, the same measurement technique used in preformulation may be applied here. Instead of merely testing 10-20 tablets for each bottle, all 50-100 or more may be examined by Raman or NIRS and a better picture of stability produced. Namely, we can state that, for example, lot A had 10 “different” tabs at one year, lot B, 5 “different,” lot C, 20 “different,” and so forth. These can be assayed to confirm whether the differences were physical—affecting dissolution—or chemical—affecting safety.

Of course, the actual DoE, itself meant to be run on “production sized” lots, is far faster possibly by months using continuous manufacturing (CM) equipment. Also, since there is no pilot plant size in CM, all batches are “production level-sized” by merely running the set-up longer. These smaller lots are both less expensive by a factor of 10 and can be completed in days, not weeks. A win-win-win situation, no? 

---

Emil W. Ciurczak
DoraMaxx Consulting

Emil W. Ciurczak has worked in the pharmaceutical industry since 1970 for companies that include Ciba-Geigy, Sandoz, Berlex, Merck, and Purdue Pharma, where he specialized in performing method development on most types of analytical equipment. In 1983, he introduced NIR spectroscopy to pharmaceutical applications, and is generally credited as one of the first to use process analytical technologies (PAT) in drug manufacturing and development.