By Herman F. Bozenhardt, Bozenhardt Consulting Services LLC
This article is a brief retrospective on aseptic technology to help the current pharmaceutical generation understand the legacy of our business and provide a snapshot of where we have been and where we are going. The reader can decide if the changes reflect revolution, evolution, or simply adaptation.
This article is also a testament to the progress we have made, not in one technology, but in many, while simultaneously delivering several generations of life-saving drugs and extending lives worldwide. Our business has many visionaries, and their dedication to this industry forced the evolution and growth of aseptic technologies that many take for granted.
In 1976 at Pfizer, I asked my boss, Ron Pomerantz, what makes an aseptic operation successful. He answered, “Good water, good sterilization, good air, and good people following good SOPs.” His response refers to a few key disciplines that we, as an industry, have evolved the most. I want to use these words as a focus for this discussion. I also use the word “discipline” because we executed successfully in those days with cruder equipment, less efficiency, and many more tests and cross-checks to compensate for the equipment available at the time. Our standards never wavered, and we took no chances; in most cases we simply worked harder. Let’s explore a few aspects of then versus now, and maybe glimpse a vision of what is to come.
HVAC And Bioburden
If you look at the aseptic facilities of the 1970s and examine them from an architectural layout, you would see a very busy work-centered layout, without any unidirectional flow patterns. In addition, there were interior wash stations, common product subdivision rooms, and no separation of clean and dirty corridors. This, combined with a commercial-style HVAC HEPA supply and return orientation, might cause you to wonder how the bioburden and cross-contamination was controlled and why the environmental monitoring results were surprisingly good.
- As a matter of daily procedure, the entire classified interior (inclusive of airlocks, formulation rooms, filling suites, and support rooms) of the facility was washed with a disinfectant solution, including ceilings, walls, and floors. Then, after the washing, the entire classified area was “gassed” with aspirated formaldehyde. Bioburden was meticulously monitored but was never a serious concern for the facilities that exercised decontamination at this level. This level of effort required an entire third shift of personnel to execute the repetitive and tedious work.
- The HVAC air handing unit (AHU) systems in the 1970s were large-scale drive units serving many rooms within the aseptic complex, with a series of air dampers set manually to maintain the required pressure differentials. Like today, the pressure differentials were set to protect the interior rooms; back then, however, differential pressure gauges were placed on a large wall display. Daily airflow directions and pressure differentials were checked with smoke sticks and manually documented by the first shift. Pressure cascades always flowed outward, as the concept of potent compounds was not a design criterion yet.
- Today, AHU drives have evolved into much smaller, more efficient VFD (variable frequency drive) motors, and are segregated to control only specific rooms. The modern VFD enables lower power consumption, and having each AHU maintain a smaller part of the suite provides lower risk, as the failure of any one AHU will not shut down an entire operation.
- Modern process control systems — with 100-millisecond data sampling, precise differential pressure control, alarm annunciation, room-by-room interior status lights along with the control consoles, and 3D vibration monitoring systems — have improved the reliability and performance of HVAC systems. These modern systems have reduced the incidence of HVAC failure and its catastrophic consequences, provided a greater likelihood of compliance, and replaced the daily smoke stick.
- The HVAC units of the past had varying locations of the return registers, including ceiling returns. The value of low wall and floor-level returns was not understood until analysis with computational fluid dynamics was done in the 1990s to determine dynamic particulate distribution and the impact on bioburden reduction.
- One of the most significant developments in bioburden reduction was in construction materials and interior finishes. In the 1970s, epoxy-coated cinderblock and gypsum board surfaces for walls were standard. The ceilings were a mix of coated/painted gypsum board and coated tiles with poorly integrated or hanging lighting fixtures. Any modification to the walls and ceilings required significant labor for preparation, installation, painting, and finishing and resulted in shutdowns/outages requiring a significant amount of time.
- The floors of the 1970s were industrial coated aggregate systems or epoxy-coated concrete. These systems, although hard, chipped and wore down in the trafficked areas, requiring frequent repair and recoating. The problem occurred if the floors were not maintained, and the chipped surfaces became sites harboring bioburden and expelling particles.
- Today’s walls have a monolithic polymer surface that is cleanable and chemical resistant, with special adhesives and caulking to seal sections completely. Today’s AES wall panels and Kydex sections, among many others, provide surfaces that resist bioburden adhesion and withstand decades of cleaning. These wall and ceiling systems are all prefabricated and can be quickly installed and ready for use.
- The modern epoxy terrazzo and Stonehard systems are polymer-based surfaces and have a great resistance to chipping, and if they do chip, they present a polymer undersurface that does not readily shed. In addition, the development of Nora flooring and its tile system has addressed floor areas that are prone to impacts by providing more give in the floor, and allows for easy replacement of sections, which can be done during routine maintenance.
- Overall, the older plants required substantially more effort and cost (relative) to maintain. HVAC outages were significant events that typically caused capacity loss for a long period of time, since an entire operating suite would have to be resanitized. The larger drives required more repair time with larger, expensive parts (that hopefully were in stock). Interior finishes were often not frequently addressed during turnarounds due to the time required to repair a floor or wall sections, as well as the fear of particle generation caused by the construction.
- Today’s facilities are more modular from a mechanical standpoint, and parts and repairs can be done more readily and more quickly. Polymer systems for interior finishes allow more routine upkeep, less construction particle generation, quicker turnaround, and higher production capacity.
- The constant drive to reduce both cost and bioburden/particulates has evolved with the introduction of materials and technologies. Today’s leading edge of fanwall AHUs with multiple VFDs programmed for the daily operational loads is just another milestone in the utility area. This improvement must be recognized in the face of ever shrinking classified space, as we use more “gray space” processing, closed process skids, and filling isolators, including closed vial filling.
Water For Injection
United States Pharmacopeia (USP) water for injection (WFI) has been a universal constant in the business over the many decades. WFI must always be kept flowing, hot, and in a line/vessel that is free of dead legs. The most significant changes have been in generation systems, testing, and volume of use.
- In the 1970s, generation was done via several methods, all deriving from boiling treated and filtered water. It is the method and location where the generation takes place that has changed.
- The quickest and easiest way to generate WFI in the 1970s was to condense plant steam in a stainless-steel heat exchanger. This water was exclusively used for formulation and compounding, because it was directly fed into the process vessel, and compounding commenced directly thereafter. The only weakness in this system was that the generation source usually contained boiler feed water additives to reduce corrosion. In addition, the pressure in the condenser was typically lower than the cooling water. This presented a problem if a pinhole breach across the tube sheet existed, as it could bleed cooling water into the condensed steam/WFI. The FDA changed its regulations to combat those potential problems.
- Another generation technique was a simple single-pass still. This still was fed with filtered potable water, and the steam coil in the bottom boiled it off and condensed it in a tin-plated condenser. This was essentially a local “boiler” that had no treatment chemicals. This water was typically sent to a holding tank (with steam coils) and used for washing vials, parts, and other utility functions. Holding water in any vessel (including the holding tank/surge tank) was always a temporary state, and even if it was just used for wash water, it was purged/emptied on the third shift. The downside of this generation method was a lack of continuous purge and heat recovery. Generally, once or twice a year the still would have to be opened up and “mined,” wherein the mucklike residue that built up from the volume of water boiled off was physically/mechanically removed. In those days, city water systems were believed to provide sufficient treatment other than filtration. In reality, there were substantial microscopic material and dissolved chemicals that precipitated out of the evaporation process.
- Purging lines before water use or collection was a ritual governed by SOPs, and many thousands of gallons were sent to the sewer. Overall, the energy efficiency was not measured, but it was probably incomprehensibly low by today’s standards. In addition, maintenance was a nightmare and productivity/capacity was poor.
- In the early 1980s, multiple effect stills became a necessity to improve energy efficiency via heat recovery and to eliminate the steam condensation compliance issue. Coupled with the installation of new stills, improvements continued with the incorporation of more elaborate pretreatment systems such as deionization beds, softeners, chlorination/carbon bed systems and reverse osmosis for both well and non-well water systems. The initial reverse osmosis systems were not very robust, had service gaps, and did not perform as hoped.
- By the late 1980s, reverse osmosis (RO) systems became the pretreatment system of choice, and the industry embraced RO, to the point that some very innovative organizations began using double pass RO to produce “cold” WFI. It did take the regulatory agencies a substantial time to become familiar with this method and accept it.
- Today the vapor recompression stills require only softened water as feed, and they produce at extraordinary efficiency. This is at a time, surprisingly, when WFI demand is declining, as the use of single use systems (SUS), various disposable process units, blow/fill/seal plastic product containers, ready to use (RTU) stoppers and seal systems, and closed vial filling is increasing steadily. These systems essentially eliminate the need for the operating company to wash and flush, which is the primary consumer of WFI. This is another example of a technology that is efficient, is low-maintenance, and has found an optimum use, but is at the same time disappearing.
- The last point that should be noted is the ability of the operations organization to get a lab result from the daily water sampling program or for compounding water prior to adding the API into the vessel for formulation. Each day, many samples are taken from the WFI drops around the plant and analyzed for sterility, endotoxins, and occasionally organic/inorganic content. The industry did not have LAL (limulus amebocyte lysate) technology until 1978, and endotoxin was determined by using live rabbits and measuring the body temperature increase when injected with the WFI (or product). This crude test was the industry standard for decades for pyrogens. This required a long lead time for the sample processing/test turnaround and introduced substantial hold times in the process, which in turn reduced plant fill line capacity. In addition, it required each aseptic plant to maintain large colonies of rabbits.
As much as we think sterilization technology appears to have changed, it actually hasn’t, because we still use primarily moist heat to destroy microbial cells and dry heat to incinerate any cellular residual. The use of vaporized hydrogen peroxide (VHP) as “cold” sterilization was truly a revolution, albeit adapted for a very specific use. In the 1970s we had dry heat ovens for batch sterilization, sterilization (or depyrogenation) tunnels for continuous feed, ethylene oxide gas for heat sensitive batch sterilization, .2-micron filtration for liquids, and autoclaves for just about anything. The theory behind sterilization also has not changed, including the thermal death curve.
- In time we have virtually eliminated ethylene oxide, due to environmental concerns, and replaced it with gamma irradiation.
- The selection and availability of .2-micron filtration systems is nearly endless, including inexpensive disposable units. We still bubble point check the filters as we did in the past.
- The modern disposables market should push the need for more gamma irradiators and smaller autoclaves.
- While glass (mostly molded) dominated the fill of yesterday, plastic vials — blown on demand in front of the filling systems — are naturally sterile and pyrogen free.
- Vaporized hydrogen peroxide, ionized hydrogen peroxide, chlorine dioxide, and nitrogen dioxide were not contemplated until the isolator became a reality, provided a sealed/contained processing area, and became mainstream. Although all these gas systems are viable, the market will belong to the easiest to use, the most reliable, and the safest. That does not necessarily mean the market leader, VHP, remains dominant. The issues of degassing in the presence of a low-potency drug products and the relentless march of the use of SUS may force the market to segment.
- Ultimately, sterilization and depyrogenation may evolve into cold methods. Finally, plastic vials will eliminate the need for washers, tunnels, and pretreatment of containers.
People, SOPs, And Batch Records
As I look back over the years, the men and women who taught me about the world of aseptic systems were unique, and we will never see their kind again. In Pfizer Brooklyn, the vast majority of the folks were from the “greatest generation,” the people who fought in World War II and who sacrificed, suffered, and brought a unique dedication to our workplace. It almost seemed like they never left military service — they brought that discipline to the pharmaceutical business. We, in turn, treated them with respect and regarded them as the largest group of company stockholders.
The nature of what we regard as an SOP and a batch record has gone through several evolutions from the completely manual execution and documentation of the 1970s to today’s myriad of computer systems, often disparate and not integrated. The most important aspect over time is that there is still a person in the center of the SOPs and records: the operator. The operator has always been the mechanism of collection, verification, and enablement.
- What we think of SOPs and batch records has not really changed in terms of content; what has changed is the method of data collection, the precision of control, and how the valves are manipulated. The need for dedicated execution at many levels is required, and our people are the key. This was true in the 1970s as it is today.
- Automation was nonexistent in the 1970s, and our industry has always lagged behind other industrial sectors because of the batch nature of the business and the material handling required. Everything was manual, from fitting the piping to running the many machines.
- The real dawn of automation in our business began with the packaging lines, where metal cams were replaced with programmable logic controllers (PLCs). Then the sterilizing systems, such as autoclaves and European process skids, became automated, and the industry realized how we could improve. Finally, with the implementation of isolators and restricted area barrier systems (RABS), pharmaceutical operations became more computer-dependent and automated.
- Today, all our utilities systems (WFI, air, nitrogen, etc.), HVAC, skid-based process systems, and, most importantly, our isolator and RABS filling systems are heavily automated. In most cases our batch/formulation tanks and systems are also automated.
- With all the current automation, our operators are still the single largest group of personnel in our plants, and they still function as our “drug product integrators.” They connect the pipes with the tri-clamp fittings, connect up the disposables, prep the rooms, monitor a myriad of computer systems, master the logistics, and shepherd the drug product out the door. Everything we have discussed here depends on them. We must make every effort to develop a culture of dedication, education, loyalty, and sense of ownership of our customers’ lives in our operators. The operator of 2017 does not come with the background of the greatest generation, so it is incumbent on us to create a new culture of loyalty, dedication, and success. Our investment in and compassion toward our operators assures us of success today — and provides a mechanism for them to train the next generation tomorrow.
This article is dedicated to the original “grand master” of aseptic systems, Ron Pomerantz, and the many innovators and visionaries of our business, including several that I have had the pleasure of working with: Jim Agalloco, Phil DeSantis, Dave Maynard, and Steve Ostrove. This is also in remembrance of those of the greatest generation who taught me my first lessons in aseptic technology: Sal Catanese and Nick Marone.
About The Author:
Herman Bozenhardt has 41 years of experience in pharmaceutical, biotechnology, and medical device manufacturing, engineering, and compliance. He is a recognized expert in the area of aseptic filling facilities and systems and has extensive experience in the manufacture of therapeutic biologicals and vaccines. His current consulting work focuses on the areas of aseptic systems, biological manufacturing, and automation/computer systems. He has a B.S. in chemical engineering and an M.S. in system engineering, both from the Polytechnic Institute of Brooklyn.