I recently saw a documentary about computers. A few years ago, the most powerful computer was a Cray. It was about the size of a Mini Cooper (on end): 77 in. x 263 x 104 in. and weighed 5.5 tons and cost $7.9 million (1977). With 8.4 Mbytes memory and 303 Mbytes memory, running at 80 MHz, it was the “screamer” of the time. Each couple of years, computer advances made these specs better and better. However, an Apple Watch (series 4) has a dual-core 64-bit chip, 16-GByte storage, and Wi-Fi connectivity—for several hundred dollars.

What, might you say, has this anything to do with producing a premier pharmaceutical product? Well, unless you have been training for the first Mars expedition, you know I am a very strong proponent of PAT (Process Analytical Technologies) and QbD (Quality by Design) and, in logical progression, to CM (Continuous Manufacturing). To-date, there have been a number of reasons why both large and small Pharma companies and, by extension, their contracted partners (CMOs and CROs) have remained firmly in the 1960-70 paradigm of manufacturing their products. Why is that so?

1. There is a lack of experience. To be fair, the universities are just catching up with the changes in manufacturing methodology, thus, a shortage of skilled, much less experienced, pharmacists, pharmaceutical process engineers, and process instrument gurus.
2. (A supposed) Lack of liquid assets. One example of a PAT tool was the first NIR chemical imaging device. While incredibly versatile, allowing a 3-D map (X-Y picture based on >80,000 pixels, with each PIXEL containing a full NIR spectrum) to be drawn, showing a wealth of information. Sadly, the final cost just before it ceased to be sold was North of $500,000. All the larger Pharma companies had ONE, while few, if any, smaller companies could afford one.
3. Lack of incentive (Governmental). Under the “batch mode” form of manufacture, there are, in reality, few to no controls (assays/monitors) before the batch is finished and final tests performed. If there is a chance that the batch contains outliers, it is not in the best interests of the company to analyze more than the minimum 20 units from a batch that could be as large as 5 million tablets or capsules! The PAT and following FDA and ICH Guidances were distinctly labeled “voluntary,” allowing the companies to ignore them and shout, “Damn the torpedoes. Full speed ahead!”
4. Lack of monetary pressure. This is because the industry lobbyists have successfully prevented any price controls or the ability of Medicare to negotiate prices, country-wide. So, even if (when) a number of batches fail or need to be recalled, the financial loss can be covered by raising the price of other lots, therefore, making change(s) in testing/control methodology uneconomical. Why change your way of producing product when nothing seems to diminish profits?

However, to quote the Nobel Laureate Robert Zimmerman (Bob Dylan, to you), “The times, they are a changing.’” Numerous countries have made moves to lower the cost of medicines to their “National Health Services.” Germany, for example, has a law that mandates a new drug for a particular illness (e.g., cancer) cannot cost more than existing drugs unless a demonstrable improvement or advantage exists. Numerous states in the U.S. have rules that allow a pharmacist to substitute a generic product for a name brand. In some states, this is mandatory, unless the physician specifically states the name brand be used.

If Congress finally allows Medicare to negotiate drug prices, the pressure on price levels and profits will increase. There is some room for lowering prices, but, at some point, major companies will have to cede all sales of name products to generics. The answer, of course, is improve the way that the products are manufactured. Let me see, is there a way to speed production, lower costs, AND improve quality? Yes! With PAT, QbD, and, eventually, CM. But it costs money and time to convert. However, there are a number of lights at the end of the tunnel.

Starting a few years ago, smaller instruments began to appear at trade shows. When the telecom industry imploded in the late 1990s, the technologies used to sense and re-transmit multiple voice and data streams turned out to be exactly what was needed for rapid and sensitive spectroscopic instruments:

1. Sensitive detectors. The most popular (both with spectroscopists and telecom technicians) material for low noise and high precision when impinged with a multi-frequency mix was indium gallium arsenide (InGaAs). However, the high demand caused the price to be very high. I asked the president of a NIR company how much a diode-array of InGaAs would cost. He said $10,000. When I asked about buying in bulk, he said $10,000 apiece. This at a time when a typical high end NIR spectrometer could be purchased for ~$45,000. When you figure you need to sell an item between 2.5 to 4 X the cost of manufacture, adding one of these superior detectors could have added $25-40,000 more to the price. Assuming you could actually find InGaAs, since the telecom companies were hoarding it. When the telecom bubble burst, there were warehouses of InGaAs available, so the cost of a detector dropped to the $100 range.
2. Novel wavelength selectors. In order to accurately resend undersea signals after detecting them, telecom used various MEMS (micro-electro-mechanical systems) to differentiate between numerous phone calls and data streams. These are also excellent for dividing a spectrum into individual wavelengths, so, when the manufacturers of these units had a choice of closing or changing direction from telecom to analytical monitoring, the choice was clear.
3. Faster/smaller computers. Already mentioned above, the cost per gigabyte of storage and speed of calculation has dropped so greatly that placing monitors in a process stream is almost trivial. In addition, almost all new equipment/monitors are equipped with Wi-Fi, allowing networking between various monitors, allowing immediate feed-back/feed-forward control of the process.

More new “toys” are also available for immediate injection into the process stream, every month. When we had a number of patients die from adulterated heparin from China, the FDA and USP did intensive work and concluded that 2-D NMR was the best way to ID and quantify heparin. The biggest set-back to smaller companies and contract manufacturers was the cost of purchase and operation of the unit used for the research. The unit could be more than $500,000 and the liquid helium needed to maintain the magnet and the liquid nitrogen used to slow the evaporation of the helium can cost up to $10,000 per day.

There are now at least two models on the market that have some very attractive parameters: 1) low price (under $70,000) and 2) the ability to work at room temperature. With this tool, alone, smaller companies are able to compete for the heparin market. It is also good to know that these are quite good for any company that synthesizes APIs (small molecule actives) or purifies biomolecules, allowing them to do internal testing and not depend on expensive third-party analyses.

Low cost mass spectrometers (MS) are also available and are readily available for freeze-drying and headspace analysis of bioreactions. The purpose of following a lyophilization (freeze-drying) process is to assure that the off-gassing is primarily the solvent(s) (water and or organic solvents). In addition, the MS follows any nitrogen peak; the presence of nitrogen indicates a potential leak in the gaskets, preventing a good vacuum from being achieved. The headspace analysis of a fermentation or other bioreaction is a good indicator of the bioreaction’s progress. Formerly, expensive and large MS instrumentation precluded ease of use in a process stream. Small, more mobile units are both less expensive and process hardened.

Ion Mobility “Spectroscopy” is another tool for fast analysis. Previously used for residues (cleaning validation, surfaces at airports), there have been many papers published where it has been used as a replacement for either gas or liquid chromatography, generating “chromatograms” in roughly 20 milliseconds. This allows it to be used as a process tool for liquid reactions (bio-fermentations, API synthesis) in nearly real-time.

As spectrometers (NIRS, Raman, TeraHertz) become faster, smaller, and less expensive, they can be used for real-time analysis of dosage forms or following the coating process. NIR-based units are already available and being used that are capable of measuring tablets for content uniformity and predicting dissolution profiles for 100,000 tablets per hour per line—multiple lines may be run, using fiber optic probes. One unit also uses a directed air, or nitrogen, jet to reject any tablet/capsule immediately after it being analyzed as “failing” either assay or dissolution prediction. This means that every dosage form arriving at the packaging line has been examined, eliminating rejected batches and recalled lots. The unit price is in the $100,000 range; far less than the cost of a recalled batch.

The bottom line to all these advances is that they have levelled the playing field—PAT monitoring and control and QbD for the flexibility of modifying process parameters, on the fly (based on data from monitors). These technologies are well within the budgets of smaller generics or contract manufacturers, allowing them to continue to lower prices while simultaneously improving the quality of the products.

As usual, I will point out a financial benefit of these units’ application: since each step has a potential to destroy the product quality, assuring each step and preventing even one lot would immediately present the company with a ROI, higher than the investment. All we need is the will to use these new mini-wonders of technology.

---

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.