Any number of great discoveries have come about by serendipity. The possibly apocryphal story of Isaac Newton and “the apple” comes to mind in physics. In commerce, when Henry Ford was only making black cars, the tires were the color of natural rubber: pink. Mr. Ford asked his friend Charles Goodyear to find a way to make them black and he added enough lampblack (soot) to color them black. In doing so, he also softened them and made them more durable, affording a smoother ride and longer tire life.

Vulcanized rubber was an accident where molten sulfur and rubber were spilled together and we know the products as “tires” today.

The other side of this particular coin is summed up by Sir Humphry Davy when he began making discoveries of new elements with Volta’s new invention, the voltaic cell. Nothing spurs new discoveries like the introduction of new technology. New technologies range from electric lights enabling three shifts of work and automobiles allowing people to live away from where they work to smaller machines, such as tablet presses and other wonders we use in our industry.

It was not all that long ago that all our tablets and capsules were produced by hand in the back rooms of local pharmacies. Remedies were few and somewhat expensive and maybe not made reproducibly, when made by hand. Since there were fewer prescriptions sold, there was no large cash reserve for new drug research, which ensured few drugs to be sold. So, when, in the first half of the 20th century, an industry was formed to produce dosage forms mechanically, not by hand and decades after Ford did the same thing for cars, a number of things happened:

1. More product was available at local pharmacies; all they had to do was count out tablets and capsules. No need to come back in a day or so for your prescription;
2. With greater amounts of a product sold, there were greater profits in one location;
3. With a larger income, companies could support research and development departments; and
4. The pharmacy degree grew from a 2-year diploma to a 5-year program, including all kinds of industrial topics, such as formulation, tableting, coatings, as well as academic information about the pharmacology of the substances.

Thus, the industry grew and prospered until it started to discover blockbusters. Even in the early 1970s, a company making $1-200MM a year was solvent and safe.  When they started selling blockbusters, several things happened:

1. As sales income increased, companies grew even faster, attempting to find even more big sellers;
2. As they grew, they needed to have all blockbuster-type drugs. Sales of $50-250MM were almost considered losses; and
3. To make up for the cost of their huge companies, many companies got larger through acquisitions. The hope was that the company they were buying had something that would sell, although, if the second company had such a product, it likely wouldn’t be up for sale, would it?

Now, in the late 20th century and into the 21st century, we see a diminishing number of large-selling drugs—bio-pharm products take much longer to get to market and often have limited sales potential—and, simultaneously, massive competition from lower priced generic versions of their own blockbuster of yesterday. The giants were finding that they could not fire their way to profitability; merely losing headcount was a limited solution. It appears that someone still has to do the R&D and actual production. So while approaching zero workers asymptotically seems a good idea on paper, it doesn’t work.

A number of interesting things happened in the 1980s-90s that weren’t seen as all that earth shaking at the time. Several of us began introducing near infrared into the industry in the U.S. and UK. Originally, it was seen as a fast qualification method for raw materials and employed mainly for that, followed quickly as a substitute for loss on drying. While a definite time-saver, it was still a lab-based method. A few companies worked on hardening the instruments, allowing them to be placed on a cart and wheeled into the warehouse, but the price-point was still high. And, of course, the units were both large and not explosion-proof still running on 120-220 volts and requiring a power cord.

Another, seemingly unrelated, phenomenon also occurred in the 1990s: the telecommunications industry was imploding after years of unchecked growth. Several things immediately impacted the measurements industry. For one, the massive amounts of Indium Gallium Arsenide (InGaAs), stockpiled by the telecomm suppliers, which is excellent for very sensitive detectors in fiber optic lines, became an excess commodity. The price of an InGaAs diode array detector immediately dropped from $10,000 to $1-200 apiece, allowing them to be economically placed in NIR instruments. Having a more sensitive and, more importantly, faster responding detector was a big step in process measurements. Another victim of the telecomm crash was the line of component suppliers to the industry. Since there were special challenges in keeping, for example, a trans-Atlantic phone/data line working, the industry depended on specialized suppliers for their components.

The data/photos/voice transmissions were carried by light, of course, through the optical fibers. This required, first of all, a strong, long-lived light source. The LED (light emitting diode) came to the rescue and, in fact, a new breed of SLEDs (super-luminescent light emitting diodes) were developed. Since they could not be easily replaced in underwater scenarios, they were designed for at least a 25-year life. Not too shoddy for analytical equipment, either.

The light carrying the information needed to be translated back into electrical signals and either sent through wires, though wireless transmission, or re-converted back to light, to be sent along another fiber. The InGaAs was developed for this purpose, being both fast and accurate. Another piece of the equation was the development of small and rapid devices that can produce/sort/direct wavelengths. Can we see monochomators in this? Small MEMS (micro-electric mechanical systems) of varying properties were built for sorting conversations from pictures from data coming over the fiber cables. The question was, what would these small companies mostly located near Boston do now? The two choices were to either close or make analytical instruments. We’ll be back to this in a minute.

Simultaneously, Pfizer was working with Zeiss to develop yet another, seemingly innocent instrument breakthrough. With only a blend uniformity application in mind, they developed a self-contained, WiFi communicating, battery-operated NIR spectrometer. It wasn’t the best, but it was more than good enough to optimize blending and eliminate lab testing of samples.

A few years later, the U.S. FDA put together a committee, chaired by Ajaz Hussain, to look into the way to guarantee better products through monitoring the process in real time. I was fortunate enough to be asked to work on the Validation Subcommittee of what was the gathering of information for the PAT (Process Analytical Technologies) Guidance. When the draft version was released in 2002 (final in 2004), all the developments mentioned above came together in a wonderful convergence.

Along with the newer, faster computers, the later generations of Chemometric software was coming out. Third-party suppliers greatly improved upon the instrument vendors’ versions of PLS (Partial Least Squares) and PCA (Principal Components Analysis) and supplied sophisticated programs to tie together disparate inputs (i.e., temperature, pH, NIR/Raman spectra, and energy consumption) to allow feedback and feed-forward control of processes.

With PAT in mind, new and existing instrument companies began providing mini-instruments constructed around MEMS technology. These were proven to be a boon to measuring incoming API and excipients at the point of delivery, in real time. Smaller, ruggedized monitors—NIR, Raman, Light-Induced Fluorescence, and TeraHertz—now abound and, since they are usually battery-powered, they are inherently explosion-proof, saving up to $80,000 per unit for explosion-proof cabinets.

The ideals in the PAT Guidance were expanded by ICH Guidances Q8, 9 and 10, where risk and lifecycle management were added to monitoring and control. That is, we now looked for critical points of control and qualities that needed to be controlled for a proper product. These small, rugged monitors were installed in numerous production steps: blending, fluid-bed granulators, tableting machines, and pan coaters. Units were also used to measure the accuracy of blister-packed clinical supplies.

Here is where the unforeseen part returns. Since so many of the larger Pharma companies are cutting staff and relying more and more on CMOs and CROs, it has become a cruel good news/bad news joke for the contract suppliers. With the added burden, THEY now have to expand as did the parent companies, which caused the overhead problems, in the first place.

Fortunately, all the pieces of the puzzle have come together at the right time. We have small, rugged monitors, proper software for controls, Chemometrics and computers give instant analyses/predictions, and, best of all, we have the regulatory agencies supporting real time release. We can now use the data, gleaned from our PAT process, to release product without laboratory results. Finish the run; ship he product.

And, saving the best for last: cost savings are even more substantial when companies convert to continuous manufacturing. Production floor space drops by a minimum of 60%, HVAC costs drop, warehouse space required is at least halved, and lab time drops to nothing. Since the process is continuous, there is no need for cleaning and cleaning validation between smaller batches. All because these tiny monitors allow us to monitor and control each successive process and do away with batch mode as suggested by Janet Woodcock of the FDA last June.

---

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.