The emerging paradigm for pharmaceutical manufacturing is to have more and more of the “pieces” of the final product fabricated off-campus. In the “good old days” (1950s-80s), the excipients were, of course, bought from a supplier. Can you imagine a major pharma company with a farm, where they harvest milk for lactose or grains for starch? The difference between then and now is that we knew from whence came the materials. Talc was a product of Alabama, cornstarch was from our mid-west, most APIs were made in-house or by a nearby contractor, etc.
We still purchase these excipients and APIs, but now, we are called upon to trust other Agencies, other than FDA, USDA, or EMA, in countries where the agencies seldom have the personnel, force of law, or resources to monitor production processes—raw materials, intermediates, or finished products. In the past—1940s through the 1990s—the way in which we cleared these raw materials for use in our products was two-fold:
- We accepted certificates of analysis (CoA) and “spot-sampled” some containers of some, if not every, lot(s) for simple tests (ID by USP methods, visuals, organoleptic tests), or
- We took some samples (e.g., √N + 1, meaning 11 of over 100 containers tested) and performed a full monograph (USP, BP, EP) of tests.
- Knows what the final product looks like (has experience supplying Pharma companies);
- Has GMP training;
- Has competent regulatory oversight; and
- Actually cares about providing a quality product
In 1984, with the supply chain only extended from the Atlantic to the Pacific and Canada to Mexico, this “due diligence” was sufficient. However, these were simpler times. In recent years, with the concept of global outsourcing of both raw materials and manufacturing (as well as R&D), these naïve checks cannot be counted upon to protect the supply chain/patients. As with any international treaty, we need to follow the “trust, but verify” doctrine. As a clear example, I refer to the heparin disaster of a few years back.
To refresh any memories that need refreshing, heparin from China was contaminated (if 50% impure can be called contamination) with over-sulfated chondroitin (OSC). OSC is structurally is chemically similar to heparin, with the same pieces, merely in a different order. Taken orally, OSC is used for treating arthritis but, intravenously, it is toxic and killed a number of people before the cause was discovered. Why? Standard USP tests could not discriminate between the two materials. Why? Remember trust? There was never a need to tell the difference. The “usual suspects,” I mean, traditional suppliers would never substitute OSC for heparin, so we never needed to check for OSC before. Then there was the contamination of pet food with melamine, killing numerous animals. The “usual” tests could not determine its presence. Sadly, the USDA at U.S. ports tested for its presence in OUTGOING grains since 1980, but not for imports.
Now, with a 15,000-mile supply chain, with hundreds, if not thousands, of inputs, we need an entirely different paradigm. We must find the most efficient, yet cost-effective methods of supply chain control. Happily, there is a “good news-bad news” convergence of needs here. Simultaneously, we find a need to ascertain the quality of an incoming material or (externally) finished product for safety and provenance (authenticity) and the need to know much of the same to be able to run a proper Quality by Design (QbD) program.
Having already established that a compendial monograph is not, in any way, helpful as a source of data on whether or not a given material will aid or hinder a production run, it is also disconcerting to realize that it may also not be capable of determining false or out-of-specification (OOS) materials. If we take the USP monograph for lactose, for example, we can see the lack of specificity:
- Boil the material in a copper salt. If it turns red, confirm a reducing sugar.
- An ammonia solution has a specific optical rotatory value, which actually changes with time and yet there is no time component in the monograph.
- Particle size is “determined” by how much material is retained on, say, a 100-mesh screen. (No range, no accommodation for static clumping, no actual lower limit, even though micronized particles could pass the same test as 100-mesh material.)
- No polymorphic or crystallinity parameters.
- Yes, a mid-range IR is run, but, unless a sophisticated program is run (USP only specifies “looks like”), most sugars resemble one another.
These data not only show adherence to incoming product specifications, which are, of course, a range, not a single value, but act as a fingerprint of a “quality” product. Coincidentally, this fingerprint occurs with properly made finished products. The “process signature” is what, in fact, is used by the controlling software to estimate the dissolution parameter, hardness, content uniformity, and so forth… the process signature is the “passing product signature.” Clearly, this is not just a way to pass a product, not just a way to show a quality product, but to show that it is properly made by a contract manufacturing organization (CMO).
Since my first full-time position in Pharma in 1970 (yes, AD!) was to search for potential perform drug-packaging interactions, I learned:
- How important proper packaging can be for the protection and dispensing of products; and
- How to properly assay/qualify the packaging material.
I can also assure you that it took as much R&D to develop the plastic bottle and blister packs in which the product is stored for delivery (e.g., 2% TiO2 and the light gets to the drug; 4% and the bottle crumbles over time; 3% and we have the proper light protection with no negative impact on the polymer). An improper additive in a polymer could leach into a parenteral solution or vaccine; an improper polymer could extract a preservative from a similar solution. Thimerosal, for example, was rapidly absorbed by the polypropylene dropper for nose drops, rendering the solution open for breeding bacterial, helping to end its use.
We tend to accept lab results for things such as blister packaging polymer. It is usually a bi-layer “sandwich,” with the moisture barrier on one side and the lower melting material on the other. We may check the piece of material, but seldom check the actual roll to show that the proper side meets the heated rollers, thence to the case. Oh, the horror of a bad roll.
In addition, numerous batches of drug are recalled yearly for improper material in containers. We accept the labels without doing a last check before filling. Yes, we really need to examine EVERY part of the supply chain, beyond what we do today.
Emil W. Ciurczak
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.