Contamination issues continue to plague biomanufacturing even now, years after the Chiron (www.chiron.com) debacle in 2004 and 2005. In December 2007, Merck & Co. (www.merck.com) voluntarily recalled one million doses of two vaccines: 11 lots of its Haemophilus influenzae type B vaccine (Pedvaxhib) and two lots of the company’s combination Haemophilus influenzae type B/hepatitis B vaccine (Comvax). In a press release, Merck stated that it could not “assure sterility of these specific vaccine lots.”
The recalls have already caused shortages of these two vaccines in some states. How they will affect long-term utilization of these products is unknown.
Sterility problems are not limited to high-tech vaccine and protein products, as Bausch and Lomb will attest. In 2006 the company initiated a worldwide recall of its contact-lens solution after the product was linked with cases of ocular fungal infection. Subsequently Bausch announced it would no longer sell the once-lucrative product.
[Bausch & Lomb maintains that it is inaccurate to infer that the 2006 recall of MoistureLoc contact lens solution was the result of a sterility issue. “In testing of bottles taken from manufacturing, from store shelves, and from patients, MoistureLoc continued to demonstrate that it was sterile. As supported by the FDA and CDC, there was found to be no contamination of the MoistureLoc product and no tampering,” according to a company official.]
The Ubiquitous Microbe
By far, the greatest source of contamination in bioprocesses is humanity. “It’s all about operator variability,” says Mike Caltagirone, an independent industry consultant based in Nevada, whose specialty is the evaluation of facilities and processes, especially cleanrooms and aseptic facilities.
“The more automated the process, the less the likelihood of contamination,” he insists. Clean-in-place techniques do a great job with bioreactors and large equipment, he adds, but some accessory process components must still be cleaned by hand. “That’s the weak link in the chain.”
To reduce the possibility of contamination, Caltagirone recommends “strong facility design” and procedural steps to control that human variability. Another suggestion is to avoid large all-or-nothing manufacturing batches, where one slip can ruin months of work, and to employ up-to-date manufacturing methods.
Enforcement of regulations related to contamination has taken on a life of its own for good reason. Of late, regulators appear to be magnifying low-risk contamination concern, such as variability of excipients at the expense of potentially more serious contamination issues. “There’s a lot of subjectivity with enforcement compared with seven or eight years ago,” Caltagirone notes.
Regulators, for example, may over-emphasize deviations from sterility specifications that have little or no impact on a product or which may be due to sampling errors. “Many times a swab or water test is out of spec because of operator mistakes. The sample is indeed contaminated, but the surface or water may not be.”
Similarly, labs that still use old-fashioned glassware may not clean or store their equipment properly. “Microbes are ubiquitous,” says Caltagirone. “Many times lab reports will eventually come back with the message that ‘sample was not indicative of bad water, but of bad sampling’.”
Matter of Choice
The choice between cleanrooms, isolators, and restricted-access barrier systems (RABs) remains a burning issue in contamination control, according to Scott Rudge, Ph.D., COO at RMC Pharmaceutical Solutions (www.rmcpharma.com). “Finding the appropriate balance between risk and operability among these technologies is of high concern, especially for biotech firms.”
RMC conducts analytical and process development consulting for pharmaceutical and biotech clients especially in the areas of chemical, manufacturing, and controls and quality assurance. The company is often asked to provide advice on choosing appropriate fill/finish equipment for aseptic processing, particularly with respect to isolators, RABs, and related fill-line equipment. Another area of expertise is trouble-shooting sterilization projects such as steam sterilization of tanks and lyophilizers.
Which barrier/containment system works best depends on the type of sterile operation, its scale and maturity. Cleanrooms and RABs tend to be more flexible to process changes than isolators since the latter tend to be custom-built for equipment or process. More flexible options, however, provide much less control and demand tighter procedural rigor than structured, automatable solutions.
Maturity of a process is particularly determinative of which anticontamination course is followed. Demand for new products is hard to predict, as even top-class biomanufacturers have found. “Once you know you’ll need to run a process at, say, 50% capacity on a certain line, and can more easily predict a certain number of batches, when they’ll be run, their duration and volume, it becomes easier to automate and isolate them,” Dr. Rudge observes.
Conversely, the advantages of isolator technology diminish with greater process uncertainty, although switching from flexible to fixed enclosures after validation and approval is costly and not without its own risks.
Much has been spoken and written about single-use products and their potential to eliminate most cleaning and cleaning validation—for the disposable equipment. It may sound obvious, but the necessity to clean nondisposable components such as fittings, probes, tubing, housings, and the like puts a damper on disposables’ ability to reduce contamination and cross-contamination, not to mention associated cost savings.
“Disposables are, of course, important and still maturing, but my experience with them indicates that we need to get to the point where 100% of product contact surfaces are disposable, with no leaching, extraction, or degradation,” Dr. Rudge explains. “What we see today are lots of hybrid disposable processes.”
One might well ask what could possibly be wrong with a process that is 90% disposable. After all, it should provide 90% of the benefit of completely disposable processes. While they do eliminate 90% of what Dr. Rudge terms the “structured problem,” and are much more operationally flexible, such processes tend to demand nearly the same infrastructure as fixed-tank processes.
In other words, conquering the last 10% might bring about a 20% or 30% improvement in costs, and perhaps in contamination control as well.
“What we need to do is eliminate those last pieces of crossover equipment, so we can remove cleaning validation from our vocabulary entirely. Clearly that is not the case today, as one can still attend a major conference every other week on cleaning validation.”
A good deal of contamination control effort is devoted to assuring the quality of ultrapure water. Water contamination can take the form of infectious agents, particles, minerals, or chemicals. Distillation removes almost everything, but minerals have a way of creeping back in, especially after process fluids pass through purification columns and certain membranes.
Silica in itself is not necessarily detrimental to bioprocesses, but its presence indicates that something has broken down in a demineralization step, according to Dan Kroll, principal investigator for advanced technologies at Hach (www.hach.com). The presence of silica suggests that demineralization media should be changed, or that purification membranes are suffering from breakthrough. “Think of silica as a canary in a coal mine,” he says.
Hach markets test equipment for ultrapure water applications as well as municipal waterworks. Among its offerings are the Series 5000 Silica Analyzer, which measures silica levels in water continuously and in real time.
One of the most versatile instruments for contamination monitoring is the PallChek™ Rapid Microbiology System by Pall (www.pall.com). PallChek works primarily with liquid analytes—product, media, and buffers. It can also detect airborne and surface bacteria through sampling work-arounds. For example, air monitoring is done by impacting air onto agar and dissolving the agar. Bacteria on surfaces are detected by swabbing and placing the swab into reagent.
The instrument is built around a luminometer and reagents that detect adenosine triphosphate (ATP) given off by as few as 10 microbes. The reagents convert ATP release to light.
PallChek was reportedly the first instrument of its kind approved by FDA and Europe’s EMEA to be used in a protocol for detecting microbiological contamination in drug-manufacturing processes. Tests for bacterial culturing are completed in minutes, versus hours or days, according to Pall.
Among the many cleaning options for enclosures are bleach (sodium hypochlorite) solutions, hydrogen peroxide vapor (HPV), and chlorine dioxide gas. The latter technology was originally developed at Johnson & Johnson for isolators, reports Paul Lorcheim, director of operations at ClorDiSys Solutions (www.clordisys.com).
ClorDiSys Solutions sells equipment that generates chlorine dioxide gas for sterilization of biological safety cabinets, incubators, and cleanrooms. The company also provides sterilization services based on chlorine dioxide. One of the company’s recent projects involves designing a cleanroom with chlorine dioxide inlet and outlet ports in place.
Like any good antiseptic, chlorine dioxide kills just about everything, including bacteria, viruses, and spores. The advantage of chlorine dioxide, according to Lorcheim, is faster cycle times compared with HPV. Chlorine dioxide requires between 12- and 15 air exchanges to completely remove it from an enclosed structure. Removal is monitored spectrophotometrically.
Hydrogen peroxide is also a vapor, but easily condenses out. Eliminating the liquid involves many more air exchanges to revaporize the liquid. Compared with chlorine dioxide, peroxide is a lazy gas that distributes into nooks and crannies with difficulty. Plus it adsorbs into plastics, which adds to cycling time. “As long as you can seal the chamber up, chlorine dioxide works anywhere,” says Lorcheim.