In a recent column, I addressed the effects of COVID-19 on the multi-country, multi-continent supply chain(s) for both raw materials and finished products. Spoiler alert: it didn’t get much better this year. It is almost like a good news/bad news joke: as the COVID pandemic began to subside, consumer demand increased. Wait! Unfortunately, that was both sides of the joke.

One perfect example of good news/bad news can be seen in car sales. The major rental chains routinely sell their stock and renew them to keep them fresh and modern. Well, the sell-off went well, but with nobody travelling, the fleets hesitated to buy thousands of new cars to restock. So, that, coupled with a slowing in consumer buying, the major automotive producers had a slow-down, causing a concomitant drop in parts demand. One major example of change were the suppliers of computer chips. In short, cars use a bunch of chips in each vehicle, so the lack of orders put strain on the chip manufacturers, who then re-tooled to make chips for home game units instead—sales were soaring, since almost everybody was at home and bored.

The profit was higher and the specs less stringent, so why not? Well, now we have a surge in car orders as people start going back to work and there are almost no computer chips available. Have you tried renting a car recently? No? Good luck when you do. That means there are huge backlogs of cars on order with concomitant price increases (supply/demand, etc.). In fact, thousands of cars are being stored in large lots, waiting for the chips to arrive and allowing them to be drivable.

Worse yet, the supply chain breakdown was forced to improvise: exponential increases in mail-ordered materials meant all purchases going through many more sets of hands. Also, “middle-men” (read: Amazon, etc.) were less picky about the provenance of items they were providing, if only from the burden of carefully checking millions upon millions of items. This led to counterfeits and substandard materials entering the legitimate marketplace, including pharmaceutical actives, raw materials and finished products.

Last month I spoke about anti-counterfeiting methods (“chain-of-custody” protection) for finished products. I will allow an autopsy on that equine corpse and concentrate on the sanctity of raw materials. As an example, according to new research from the World Health Organization (WHO), an estimated 1 in 10 medical products circulating in low- and middle-income countries is either substandard or falsified.

One example of necessity being a mother, is a mini-supply-chain interruption: it wasn’t that long ago that there was a scandal with tainted heparin. Tainted is a mild word for “massively corrupted.” There was a drought in China, our biggest supplier, and fewer pigs were slaughtered—heparin comes from pig stomach linings and is collected by farmers, not pharma techs. The smaller yield of heparin was “supplemented” by adding over-sulfated chondroitin (OSC). OSC is excellent for helping with arthritis, when taken orally. Unfortunately, when injected, it does massive damage to kidneys. Samples of commercial lots of heparin, sometimes containing up to 50% OSC were deemed fine, using compendial (e.g., USP) tests.

Why was that not a good way to test? Basically, because all the nations’ compendia are both old and based on trust. The USP was designed as a guide for corner pharmacies and smaller industrial labs as there really weren’t any big labs back then, of course. UV/Vis instruments were somewhat primitive and infrared, when available, took up to twenty-two minutes to produce a reasonably good scan. As a consequence, there were a number of “spot tests” to do some quick and dirty checks:

1. Lactose ID? Boil some in a copper salt solution. If it turns red, it is a reducing sugar. That was often the only test performed by small labs and pharmacists.
2. Heavy metals? Why bother in most cases? Some hydrogen sulfide in water and look down a test tube vs. a standard. Eyeballs were the most common spectrometers back in “the day.”
3. Particle size? Tap some sieves and weigh what rests on each sieve. Good, unless you have a materials like caffeine which has such a static charge that the micron-sized crystals clump enough to be sold as 80, 100, or 120 mesh, even though they are all the same and do some strange things upon mixing.

Unfortunately, almost all the original tests in the pharmacopeias were quick and easy identity tests. The methodology told you a material was, for example, lactose. Not freeze-dried, spray-dried, free-flowing, etc., just what it was, not the polymorphic form or such. That was enough when you purchased your materials from a distributer in a nearby state. Considering that the local pharmacist made his/her pills or capsules by hand, he/she didn’t worry about “grades” as long as the chemical was correct. The single punch through four punch tableting machines were close enough to “hand-made” to also not worry so much about “processability” of the raw materials. Another example is PEG (polyethylene glycol); the materials are sold as mean molecular weight, e.g., 3000. It is often made by combining two or more lots to give the proper “mean” molecular weight. Mixing 1000 and 5000 could give an average weight of 3000, but the material would not behave as if it were a mix of 2500 and 3500.

A pharmacopeia is not, however, a guide for qualifying raw materials (actives and excipients) for use in modern, high speed, highly technical processes. Add to that the potential for suppliers to er, um, shall we say “enhance” or “embellish” the values on the certificates of analyses—often accepted by smaller manufacturers, lacking a large analytical lab to check every lot of every material. So, what can be done to assure safe materials—raw materials and finished products—that are what we think they are?

For those of you that have never seen this column, it might be a shock that I will suggest some modern, instrumental means of measurement. In most of my discussions, I address process analysis and final dosage form quality. Some of the same devices are applicable, but, since we are not as constrained by FDA rules, we can employ some more inventive tools for screening purposes. By that, I mean we can still “release” the raw materials with compendial methods, but “pre-screen” them with disparate technologies.

We can use both Raman and NIRS to both identify and polymorphic form of excipients and APIs from every container of every lot, with less effort than some standard “wet” lab tests. The analysis speed of testing assures the ability to return/destroy OOS materials and quickly replace them, avoiding out-of-stock inventories. Moisture and average particle size can also be estimated/analyzed with the Near-Infrared, as well.

While using NIRS or Raman for validated analyses of API in dosage forms would take a fair amount time, it is possible to perform “quick and dirty” tests to ascertain both the API(s) and the approximate dose level (i.e., ~10, ~20, ~30 mg). While not proving that the tablet/capsule is from the indicated pharma company, it shows 1) that there is the correct active resent and 2) it is approximately the level written on the bottle.

As far as sterility testing of raw materials and retention samples, there are a number of rapid microbial tests available. Of course, “rapid” in some cases may be from an hour to 24 hours, but still faster than waiting a week, as with former tests. For the near future, these will be the longest tests and slow the pass/fail testing of materials.

Packaging materials can also be bogus or sub-standard, unfortunately. Plastics are not pure polymers, they contain chemicals for strength, antioxidants and numerous other physical properties. These are more carefully controlled for containers for pharmaceutical products. For example, in non-food or pharma bottles (household cleaners, etc.), organo-zinc compounds may be used as stabilizers and could leach into hand creams or cough medicine. One manner of testing for these could be x-ray fluorescence (XRF). Handheld models are available and easily used. Some phenolic compounds, also used, can be detected by classic fluorescence methods.

In an effort to get delayed deliveries out the door, there is always a possibility of a mix-up, whether it is a packaging component, raw material or finished product. Something as simple as the wrong adhesive on a label could cause a recall. A simple and quick IR of the adhesive side would be recommended so, in a few minutes, a serious problem can be avoided.

Finished products, in a facility that generates dozens of various products, can sometimes get mixed up. There are numerous cases when even the largest pharma houses have had to recall lots because wrong dosage levels or even different products were in shipped containers. This has happened for decades, in my opinion, because of the “normal” way we have been doing start-to-finish production.

Whether we use good-old GMP batch production or PAT/QbD, where we continuously monitor and adjust the production, based on measured data. In either case, the finished product is placed into cardboard containers or bins and brought to the packaging area. After that we trust the labels on the containers and place the product into what we believe are the proper containers (bottles, blister packs, etc.). In light of what was just stated in the previous paragraph, is it any wonder there are packaging errors?

It is very, very simple to have one or more hand-held NIR or Raman instruments in the packaging area, capable of being run by the packaging staff. The simplicity of modern instruments is such that, once they are validated—for companies employing PAT/QbD, this data already exists—any intelligent person can successfully operate them. This last-minute check would prevent many expensive errors. AND, it circles back to the title of this column.

---

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.