Prior to current international events—namely, the COVID-19 pandemic—I was contemplating how extended and global the pharmaceutical industry has become. Concurrently, so has the supply chain needed to fuel a functioning industry. Looking back at 1970, when I joined Ciba Pharmaceuticals, the paradigm for producing and distributing pharmaceutical products was quite simple:

1. All raw materials, such as lactose, starch and talc, and organic chemicals—API starting materials and preservatives—for synthesis were, as they now say about produce, “locally sourced.” That simply means made in the USA and overseen by the US FDA and US DA, so, in short, safe and readily available. Lactose came from Midwestern dairy farms, talc from a mine in Alabama, etc.

2. Packaging materials (bottles, caps, cotton, strip-packs) were also locally manufactured by numerous smaller, specialized companies and readily available.

3. APIs were (mostly) generated in-house; then purified and assayed, in-house. So, they were always available and known to be safe and pure.

4. The commercial product was made in-house using self-made APIs, was bottled (and labeled) in-house, and stored and shipped from in-house warehouses, and shipped directly to pharmacies, doctors, and hospitals.

So, what changed the paradigm? Many factors, but two of the largest were the 1) rise of mergers and acquisitions. I experienced this when Ciba merged with Geigy in 1972-1973; and 2) blockbuster drugs became the “soup-de-jour.” The mergers began in earnest with the Ciba-Geigy merger, and continued for well over two decades. Some of the reasons, and consequences, of M&A’s and blockbusters were:

1. Mergers happened to, among other reasons, avoid having to bring products from Europe (Basel, largely, for Ciba, Geigy, and Sandoz; now all part of Novartis). Making all the products at smaller U.S. sites, in light of increasing sales and increasing populations, was difficult, so companies wanted to add production and R&D capabilities—merging is less expensive and faster than building from scratch.

2. Multiple locations became available to manufacture common products, so any break-downs or weather conditions did not delay deliveries from any single-sourced location. This came in handy when the Ciba-based union went on strike, so critical products could be produced at non-union former Geigy locations.

3.  As the production of “product A” spread to multiple sites, the demand for API synthesis precluded organic synthesis labs at each location and existing production capacity at existing sites was strained to the point of having short-fall, so we engaged third-party synthesis labs to fill the need.

4. As production sites spread around the world (well over 80 sites for Pfizer, alone, at the height of Pharma’s growth cycle), local raw materials were incorporated into the products. Pfizer presented a paper on the amount of work needed to assure comparative magnesium stearate supplies. And that is merely one small component.

5. When blockbuster drugs, e.g., Lipitor, exploded on the scene—followed by the many generics jumping into the game, when patents expired—there was no question of single sources for raw materials. Production facilities needed to keep working, so “equivalent” sources were found in whichever country they existed. This proliferation included labels, cotton filler, bottles and sprayers, and blister packages; all locally sourced (at first).

“OK,” you might say, “is this a good or bad thing?” My answer is “yes.” While that may sound like a “Schrödinger’s cat” to you, it simply means that very few things, meant to help an industry, are either, or both, depending on circumstances. There are some good points (obviously, or they wouldn’t be in business) to a specialized, extended supply chain:

1. A “widget” made by a specialist will almost always be better-made and less expensive than were it made by a “Swiss Army knife” company, which makes many widgets. Why?
  a. Not having to change tooling between products adds to speed of delivery.
  b. Making one item allows the factory to streamline the process and make a superior product.
  c. Experience with that single widget tends to make them better and faster as time (and many lots) goes by.

2. Having a constant supply of the dozens of the ingredients of a pharmaceutical product makes them more readily available and made to specification:
  a. Powdered API and excipients for tablets, capsules, creams, and solutions
  b. Containers for solids and liquids
  c. Specialized packaging
    i. Monthly distribution packs for birth control tablets
   ii. Spray pumps for everything from throat sprays for sore throats to dosing inhalable insulin.
  iii. Packages where parts or components of a working applicator are enclosed.

3. Making the “same” or nearly the same parts for an entire industry keeps the unit price well below what building them in-house would cost. Also, the myriad number of tools and skills required for self-construction would add a great deal to the overall cost of the product.

4. Potentially, more than one company can make a particular unit for a complex pharmaceutical package, allowing a steady stream to the Pharma company and potentially allowing for price negotiations.

Now, we get to the flip-over point towards our good/bad scenario. There are very few parts of the supply chain that are per se “bad” (a black or white designation), instead we have over 50 shades of gray. To address some of the more obvious:

1. Even small companies expect to make a profit. As a consequence, they cannot be expected to share their designs or proprietary manufacturing technologies with competitors. As a consequence, the interchangeability of parts between vendors may not be as seamless as we believe. This means we are more dependent upon a single-source than we might believe. A source 5-10,000 miles distant!

2. I addressed this point previously, but worth repeating. If we wish to set up a PAT/QbD system for complex products, it is hard enough with simple tablets (controlling/continuously evaluating raw materials, APIs, and auxiliary items, such as capsules or coating liquids/powders), but when you are dealing with dozens of suppliers, potentially in several countries (none of whom are sharing specs), building a PAT or QbD paradigm is very, very problematic.

The testing and quality programs, in light of an extended supply chain almost obviate any advantages in a working QbD program (almost). Many more safeguards and enforced specifications are needed for imported materials than domestic materials, but even with additional tests on all materials—100% of raw materials has been possible since 1985; I know because I started it—PAT is always worthwhile, if only from a safety perspective, and the bottom line isn’t hurt, either.

We need to examine all the downsides of an “elastic” supply chain, stretched across the planet. What could possibly go wrong with a 12,000-mile long supply chain?

1. Political turmoil comes to mind. An uprising in Turkey or a financial crisis in Argentina can cause delays or curtail needed materials for Pharma companies, both large and small. While some events overseas can be predicted and addressed, many are unexpected. For example, a missile from North Korea, flown over Japan, can upset deliveries from South Korea, Japan, or Vietnam.

2. Then there are “natural” disasters. The volcano in Iceland a few years ago disrupted trade and travel between North America and Europe. Flowers from the Netherlands, unable to reach the U.S. for Valentine’s Day, cost the Dutch flower industry and U.S. florists millions of dollars and resulted in many postponed airplane flights, also costing the industry millions of euros and dollars.
While this was “localized” to northern Europe and only lasted for a month, the economic effect was considerable. The “ripple effect” in the hospitality industry (restaurants, hotels) was significant.

3. Then there are longer-lasting disasters and disruptions. The Nakashima power plant meltdown, as an example, has been a long-lasting economic drain and will be so for years. The disruption has been local, thanks to government efforts in Japan. But this illustrates how natural disasters can cut any supply chain.

4. Then there is a force majeure, as we are seeing with the COVID-19 pandemic. Apart from the loss of life and burden on the healthcare system, a general downturn in our economic system, we are seeing the effects of an over-extended supply chain:
  a. The ventilators and facemasks so needed for U.S. patients are in short supply and the major source of these components is, CHINA! And, since the virus first hit China, closing all industry, not only are these not being produced for export, but all existing supplies in China have been diverted to the local population, further exacerbating the world-wide shortages.
  b. Many of the raw materials and APIs needed to fight the effects of COVID-19 also come from areas hardest hit and with travel and work restrictions. So, not only are there health costs, there are monetary ramifications.
    i. If you don’t have materials, you cannot make a product. For example, Apple had to suspend production, because China couldn’t supply screens.
    ii. If you can’t sell products, you cannot make a profit, pay employees (who in turn cannot buy your product), or even keep the lights on.

So, while long supply chains help bottom lines, we see that any construct—plant, animal, or mechanical—follows a simple rule: the more complex something is, the more easily it can break. 

Note: China controls most of the rare earth metals used for iPhones, computers, and most defense systems in, for example, Air Force planes and Navy submarines. Should there be a break with them, politically, the full impact of a supply chain would be felt.

---

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.