Buying Excipients and APIs Could be Done Better

By Emil W. Ciurczak of Doramaxx Consulting

I have spent years dealing with the transfer of methods throughout a greatly extended supply chain. I considered both assuring that methodology is standardized and performed properly by contracted CROs and CMOs and, of course, various venues within a corporation, from the next building to the next continent. After you have navigated the arcane and complicated waters of method/methodology transfer, we need to consider whether the methods being transferred actually did what we expected to give meaning useful information, not just data for QA to cherish.

It would appear that we are finally attempting to characterize the materials we place into a pharmaceutical process (beyond the older USP, BP. etc. monographs) such that any PAT/QbD project has a reasonable chance of succeeding. That is, we should be measuring the morphology (note: we did thermal analysis at Ciba-Geigy of every lot APIs synthesized as far back as 1970 to be sure they were all the "proper" polymorph), crystallinity, particle size, and density to assure proper blending, granulation, and tablet compression for the QbD process. These data allow us to establish and maintain a working design space for our operators to be able to modify the process to (legally) obtain quality products. [The FDA and Janssen Pharmaceuticals have pledged several millions to a cooperative project at Rutgers, New Brunswick, USA, whose purpose is to characterize which parameters of excipients should be measured/standardized for "proper" continuous manufacturing conditions.]

Unfortunately, even the most diligent screening or characterization program is often merely after-the-fact (lab-based, using "retain samples") and is designed to either use or reject an ingredient after it has been delivered and stored in the warehouse, often for more than a month. While this vigilance is admirable, it begs the question of what do we do with materials we deem not suitable for our process (parameters outside of what may be controlled by design spaces limits), yet not returned to the vendor? Or, more to the point, how do we even determine "good" from "not so good?" We could just dispose of them (something as inexpensive as lactose is no problem, but some drug substances are very, very expensive) or try to return them for credit.

Disposing of any material is a waste of money and, unless there is a preexisting agreement with a vendor, trying to return a lot of raw material for credit is problematic. [If the material destroyed or ruined because of improper parameters is a controlled substance – narcotic -, the paperwork needed for the DEA is horrendous.]

The not-so-secret secret is that most suppliers of excipients (and APIs) do not bend their production parameters to our needs for several reasons:

1. The Pharma/Biopharma industry may generate an incredible level of income, but they do it on a relatively low amount of incoming raw materials. We are "small potatoes" in the eyes of excipient suppliers, so we wait in the line of customers who purchase these excipients. I would wager the candy industry buys more cocoa butter than we (suppositories) and the food industry uses far more lactose (tablets) than Pharma companies and beer manufacturers use more yeast than Biopharm companies. So, why would the suppliers veer from the process they are (profitably) doing at present just to make a "minor" customer happy?

2. Unfortunately, Pharma companies are all paranoid (largely, thanks to the FDA and EMA). Expanding on that statement: when a company formulates its dosage forms, despite all the "reverse engineering" tools available to generic competitors, they treat the excipients, APIs, and their ratios (and, of course, the mechanics, themselves) as state secrets, never to be discussed outside their walls. So, even if an excipient/API supplier wanted to cooperate and supply materials with ideal physical and chemical parameters, they are not let in on the application(s) for which the excipient is intended. So, they couldn’t help us, even if they wanted.

3. Sadly, most Pharma companies actually don’t know which parameters they want their APIs and excipients to possess; all specs are based on "past history." This is principally because the majority of Pharma/Biopharma companies do not have an active PAT or QbD/continuous manufacturing program in place. Thus, while there is much preliminary R&D work being done on defining what is and what is not needed in excipients, the majority of products produced are "old-school?" That is, we depend on seat-of-the-pants experience of formulators and operators to produce a workable product.

Not wanting to merely dwell on shortcomings, I have some suggestions that might help. While I do not personally formulate or make solid dosage forms, I have observed and helped formulate, produce, and analyze them for nearly five decades. (Does that count as experience?) I look at this "situation" an opportunity to advance PAT, QbD, and Continuous Manufacturing (CM). (Although, for some, it may be an "insurmountable opportunity.")

This may seem like heresy to many career pharmaceutical people, but why not include the suppliers in some of your PAT/QbD committee meetings or pre-formulation/formulation discussions? Not the ones that are about a new API manufacturing technique or financial considerations, but the ones with the formulators and production staff. You could simply refer to an API as "the active" and disclose a few details, such as the crystallinity, particle size, percent of formulation, hydration state, and some processing traits, like shape and flowability. If and when the vendors are aware of what their excipients must do in a mixture, they can work with us, not just be a passive "supplier" supplying us with a "take it or leave it" attitude.

Just how would the new partnership work? Well, several changes could be instituted in the way we buy/specify materials. I previously stated that suppliers will favor the bigger spenders (food/confectioners) and the smaller (volume) buyers (Pharma) gets what it gets, But, as the saying goes, "money talks."

So I will offer several suggestions for change:

1. If the originator (proprietary API owner), CMO/CRO (or generic company) spells out what is needed (i.e., specific material specs) to produce a dosage form and the vendor has been privy to the formulators’ reasoning, the supplier can continuously monitor their lots for the ones having specs that the a particular buyer wishes.

2. If the monitoring adds a cost to production, the buyer can agree to paying a premium for lots that are in the "sweet spot" of their QbD requirements for a working Design Space. These could either be deliberately produced (e.g., adjustments to the spray dryer for lactose) or, when the supplier does product QC to generate a certificate of analysis (CoA), he sets aside any batches that fulfill the desires of your company. [Perhaps you will need to purchase one or two "extra" batches per year, but that is preferable to a "hit or miss" purchasing program, as we have now, where lots are either returned or destroyed.]

For this concept to work, we may need a paradigm shift. Check that: we will need a paradigm shift. Currently, by the time an established (code word for nearly off or already off patent) product is farmed out to a CMO or picked up by a generic house, there is almost zero desire to streamline the process. It is far simpler to assume that shipping the old process to a CMO in a country with lower labor costs is a good idea.

That may not work for long, due to the fact that wages (therefore costs) are rising in many "developing" countries. And, as my dad used to say, "What’s cheap is cheap." He was speaking about food, furniture, and tools, but it pertains to drug products as well. Making a tablet/capsule the way it was made 17 years ago is (mostly) like making an auto the way it was made 17 years ago. Personally, while I enjoy classic cars, I don’t want a 1999 Ford as a new car (strangely, I prefer more than 13 mpg in my car). Similarly, a dosage form, made by an old method, doesn’t get the best "mileage." In short, the month-long batch approach is the largest part of COGS (cost of goods sold), so applying modern approaches (PAT/QbD/CM) is not wasted on an "old" product, it makes it profitable again, without charging "new drug" prices.

When a company understands what makes the (existing as well as new) product shine, batch after batch, a number of things happen.

1. The CMO doesn’t need as long for process transfer. As with product scale-up, getting the new location (or the full-sized batch) up to speed could take a year or more, especially considering that the CMO/generic will have a different supply chain from the initiator, meaning different raw materials and APIs from the ones used to originally develop the product.

2. As hinted at in the above paragraph, the costs of production could fall low enough for the originator to continue selling the product under its own name. The difference is that the wholesale price could be dropped, making it competitive with generics. This could be permanent or a stopgap until your CMO or generic company is up to speed.

3. Lastly, generating a QbD for an existing product (with tons of production and QC data) will allow a company to practice for its newer products. In truth, most companies only have a few "recipes" for solid dosage forms, so understanding an older product will lead to having a step up on the next "new" formulation.

So, as a rapid wrap-up, it makes good business sense to fix up your "old car" (existing product) before putting it out on the market (assigned to a CMO). By preparing it for a proper PAT/QbD, you make a better product, more consistently, AND make a few more dollars, euros, yuan, etc.