“Sterility Assurance” is a phrase often encountered in literature, regulation, guidelines and conversations involving sterile product manufacturing. But what does that term mean? Luis Jimenez (2026), reports that his analysis of enforcement data from the years 2004 to 2025 identified “lack of sterility assurance” as the primary cause of sterile drug recalls. In this communication we are going to look at the factors often cited as central to the demonstration of sterility assurance, consider their history, and their position in the regulatory hierarchy.
There are numerous references to “sterility assurance” in EU Annex 1 and FDA’s 2004 Aseptic Processing Guidance that allude to how difficult it is to achieve or prove sterility in aseptic processing. Section 2.2 of EU Annex 1 tells us that we can’t monitor our way to sterility assurance (European Commission, 2022):
“QRM priorities should include appropriate design of the facility, equipment and processes, followed by the implementation of well-designed procedures, and finally application of monitoring systems as the element that demonstrates that the design and procedures have been correctly implemented and continue to perform in line with expectations. Monitoring or testing alone does not give assurance of sterility.”
FDA’s 2004 Aseptic Processing Guidance tells us that we cannot sterility test our way to sterility assurance. This passage from the Guidance tells us something that has always been obvious: the sterility test is unable to prove sterility, which means it is not a test for sterility it is really a test for “safe enough” (USFDA 2004).
“Sterility tests are limited in their ability to detect contamination because of the small sample size typically used. For example, as described by USP, statistical evaluations indicate that the sterility test sampling plan “only enables the detection of contamination in a lot in which 10% of the units are contaminated about nine times out of ten in making the test. To further illustrate, if a 10,000-unit lot with a 0.1 percent contamination level was sterility tested using 20 units, there is a 98 percent chance that the batch would pass the test.”
Taken together, Annex 1 and the 2004 Aseptic Guidance leave the impression that no single in-process sample or final product test can independently prove sterility assurance. What is not argued is why one should think that taken together they could prove sterility assurance; the reader is left to assume that proving sterility assurance is possible. Can we really add several subjective observations and a couple of highly variable microbiological analyses together and arrive at a firm conclusion (assurance) of sterility? If we cannot than our regulatory inspectional process becomes highly subjective and qualitative not quantitative.
Aseptic Process Simulation (APS or Media Fill Testing) and SAL
Aseptic process simulation (APS) is the only fully integrated assessment of aseptic process capability but it is not a product associated test. It gives us a result which is not an SAL but rather a contamination rate. Annex 1 uses the word “failure” to describe an APS with any unit positive for growth. The “failure” should result in investigation, appropriate CAPA and repeat APS. However, in response to a “failure” Section 9.46.iv directs a review or all records related to aseptic production in the field since the last successful APS. This review, along with the results of the investigation, must inform decisions regarding recalls of impacted batches since the last successful APS.
It seems obvious that since the APS is the best assessment of process capability it is also the most useful piece of data we have regarding the ability of a production line to demonstrate a contamination rate approaching zero, but it is not assurance of sterility.
Aseptic Manufacturing Regulation: How Much Has it Changed?
We will posit here that regulation of aseptic processing is difficult. As EU Annex 1 and the FDA 2004 Guideline both assert, our analytical microbiological data (environmental monitoring, sterility testing, APS) are insufficient to prove, much less assure sterility. Given the errors in sampling and testing (Hussong and Madsen, 2004; Sandle 2023), the capability limitations in aseptic processing continue to box regulators into a corner from which there is no rational escape.
Are the Terms, “Sterility” and “Aseptic” Synonymous?
Sterility and aseptic are not synonymous, but to people not trained in microbiology (and some that are) the terms are often used interchangeably. The truth is that assuring sterility in an aseptically manufactured product is impossible. In 1986 at the PDA Annual Meeting in San Francisco Edmund Fry, who was at the Center for Drugs and Biologics at FDA at the time, presented an update on the progress of FDA ‘s Guideline on Aseptic Processing that would ultimately be released in 1987. Mr. Fry was very much aware of the fact that aseptic and sterile were not synonyms and he knew very well the differences created a regulatory conundrum.
In 1986 APS positives were common. At the time, the actual process capability (contamination rate) of aseptic processing was estimated to be ~1/1000. Mr. Fry recognized that aseptic processing could not guarantee the sterility of each container, even though the regulatory expectation was that every container must be sterile. We think Mr. Fry stated the truth, which is, the scientific reality and the legal expectation do not align. This of course creates a legal and regulatory problem including potential liability issues for manufacturers.
The result was an attempt to do through regulation that which could not be proven at all. Regulations took an absolute perspective and asserted that several factors taken together provided not the assurance of asepsis (absence of infection or “safety assurance”) but rather sterility assurance. Sterility of aseptic products is unprovable and only through technology can risk be reduced, thereby ensuring greater patient safety.
The aseptic/sterility quandary made itself clearly visible when Abbreviated New Drug Applications were filed for generic sterile products which the innovators had terminally sterilized. Some generic firms looked at sterile generics and saw two paths to what they thought was the same end. Aseptic processing and terminal sterilization both were used to make “sterile” products, therefore why not assume they are functionally equivalent?
The Way Asepsis is Achieved
The Pre-Clean Room Era
Asepsis prior to the industrialization of clean rooms required maybe 50 or more seated technician/fillers in the same room, manually filling vials, bottles, and ampoules with little or no mechanical assistance. No doubt the contamination risk was high compared to modern process capability, but some of us born in the 1940s and 1950s who were vaccinated as children are living proof that those vaccines were not only generally effective, but they were also generally safe. There may have been some environmental monitoring done during these manual fills, but details are murky. Sterility testing was introduced in 1940 giving something of a safety “backstop”, but a limited one.
The Clean Room Era
The modern clean room was invented by Dr. Willis Whitfield in 1961, working at Sandia Laboratories in New Mexico (Holbrook, 2009). Like many technological innovations, it was driven by military requirements as it was recognized in the late 1950s that complex parts, particularly those required for nuclear weapons, would not function due to particulate air contamination in manufacturing operations. Whitfield’s idea was simple he wanted to as he put it, “Let air be the janitor”.
Whitfield’s idea was to recirculate unidirectional air flow through HEPA filters, which was based on work done to control radioactive particles in the Manhattan Project. He called his principle “laminar flow” which wasn’t technically correct, but it stuck. In 1961 he built a prototype that could achieve 750 dust particles in a cubic foot of air, which was 1000 times cleaner than rooms in use at the time and he kept working on improvements. By 1963 the clean room designs had evolved enough for a regulatory standard to appear as United States Federal Standard 209 (FS209), thus the Class 100 clean room was officially born. It is basically the same as modern ISO 14644 Class 5, and therefore the same as EU Annex 1 Grade A. Also codified was 90ft/minute air velocity (0.45m/s) same as Annex 1. FS209B in 1972 introduced the pressure differential specification of 0.05” water column, which is 12.5 Pa. None of this initial work had any real relationship to aseptic production of drugs or biologics and it is likely that most industrial and clinical microbiologists and hospital pharmacists at the time knew little about it.
Particulate air quality in numerous aseptic manufacturing guidelines, ISO 14644 and EU Annex 1 are based on FS209 and are now 54 years old. We adopted US Class 100 (ISO 5) as the aseptic processing critical zone air quality sometime in the 1960s and it is essentially unchanged in 2026. The 90 ft/minute is a light breeze of 1.02mph and was chosen because it did not cause discomfort for workers. It is important to note that none of Whitfield’s cleanrooms was designed to achieve or prove asepsis, but it was implemented rather quickly in aseptic manufacturing and was a massive improvement.
The Modern Era
The core achievement of the modern era aseptic processing was the introduction of automation such as weight checks, filling and capping, meant to reduce reliance on clean room operators. Later separative technology removed humans from the critical zone entirely. EU Annex 1 recognizes that personnel in aseptic manufacturing should be kept to a minimum:
7.2 Only the minimum number of personnel required should be present in cleanrooms. The maximum number of operators in cleanrooms should be determined, documented and considered during activities such as initial qualification and APS, so as not to compromise sterility assurance.
A recent publication by Wang, et al. (2026) is an interesting three-year longitudinal study looking at the spatial distribution of a total of 1,117 microorganisms recovered in traditional cleanroom environments. Not surprisingly, Grade A not only had the least number of recoveries, but also the narrowest range of recovered genera relative to the B, C, and CNC areas. There were a total of ten (10) recoveries in the Grade A area for a contamination rate of 0.9%, which compares favorably with the <1% recovery rate for ISO 5 areas suggested by USP <1116> (USP, 2026). Not surprisingly, 50% of the isolates in Grade A were Staphylococcus or Micrococcus species. The authors conclude that microorganisms constitute a “reservoir” of adaptability to cleanroom environments and therefore they can demonstrate the ability to withstand routine cleaning including biofilm formation, selection for facultative anaerobes, spore formation, and motility. They conclude 1) that a “grid” system often used to determine environmental monitoring sites in cleanrooms is of little value given the results of their distribution studies, and 2) that the overemphasis on microbial counts while ignoring spatial relationships to systematically determine sources and routes of contamination can lead to indiscriminate “one size fits all” cleaning without truly addressing the sources and locations responsible for the lack of sterility assurance.
Given these data and the conclusions that were drawn, why would a company choose to continue to make “sterile” products in a “manned” cleanroom? To say that automation followed by separation, meaning getting people away from the process, has worked well is an enormous understatement. The acceptance of modern technology has resulted in a dramatic improvement in contamination control as measured by APS.
What is cGMP- How Close are We Currently to “Absolute” Asepsis?
A dilemma exists today because we have a continuum of aseptic processing technologies, as well as terminal sterilization for making products all of which are labeled sterile. We still have 1980s cleanroom technology evident; we have several varieties of separative technology from simple curtains to various types of RABS and full-on isolator technology. We even have isolator systems so automated that we can dispense with gloves entirely. One can only assume that how close we are to absolute aseptic is really a function of technology, meaning getting people away from the process. The pinnacle of modern separative technology coupled with highly automated machinery gets us quite close indeed.
Final Thoughts
How is it possible that multiple generations of aseptic manufacturing are simultaneously cGMP? The simple answer is they cannot be. For any technology there is at any point in time where a pinnacle of performance can be identified by looking at the data. We understand that it costs money to reach the aseptic manufacturing technology pinnacle and it takes time to make paradigm shifts, so we understand the need for regulatory flexibility. However, there comes a time when processes that fall well below the current standard should be discouraged, phased out, and incentives should be provided to adopt improved technology. It’s a tough call to be sure, but one that should be made. Isolators were first approved for aseptic processing in the 1980s. It should not take 46 years to achieve a paradigm shift.
Technology and data have taught us that aseptic safety is an engineering problem! Neither human eyeball evaluations nor highly variable and insensitive microbial analytics nor voluminous “contamination control strategies” enable us to use the term “sterility” to define aseptic processing. As EU Annex 1. tacitly admits, our monitoring and batch release tests are unable to prove sterility; what it should recognize is that as a result they logically do not assure sterility either. We’ve heard people way “It’s OK. We know what they mean.” Actually, it’s not OK, and it is not at all clear what they mean. Science always requires the truth; it should never rely on a dogmatic interpretation of an inappropriate definition.
References
- European Commission. 2022. The Rules Governing Medicinal Products in the European Union. Volume 4 EU Guidelines for Good Manufacturing Practice for Medicinal Products for Human and Veterinary Use. Annex 1: Manufacture of Sterile Medicinal Products
- Holbrook, Daniel. 2009.Controlling contamination: the origins of clean room technology. History and Technology. 25(3): 173-191.
- Hussong, David and Russell E. Madsen. 2004. Analysis of Environmental Microbiology Data from Cleanroom Samples. Pharm Tech Aseptic Processing. Pages 10-15
- Jimenez, Luis. 2026. Microbial Landscape of Pharmaceutical Failures: A 21 Year Review of FDA Enforcement Reports. BioTech. https://pmc.ncbi.nlm.nih.gov/articles/PMC12821441/pdf/biotech-15-00008.pdf
- Sandle, Tim. 2023. #6: Pharmacopeial sterility test: the statistical limitations of sampling. https://www.linkedin.com/pulse/pharmacopeial-sterility-test-statistical-limitations-dr-tim/
- United States Food and Drug Administration. 2004. Guidance for Industry: Sterile Drug Products Produced by Aseptic Processing- Current Good Manufacturing Practice.
- United States Pharmacopeia. <1116>, “Microbiological Control and Monitoring of Aseptic Processing Environments.”
- Wang, Jiaji, Pan Jiang, Junhui Yan, Huili Shen, Lajie Wu, Fen Wei, Xiaozhen Lin, Leiming Xu. 2026. Spatial ecology meets quality control: A GIS-integrated strategy for visualizing and managing contamination in sterile pharmaceutical cleanrooms. ASM Microbiology Spectrum https://journals.asm.org/doi/10.1128/spectrum.00268-26
Subscribe to our e-Newsletters
Stay up to date with the latest news, articles, and events. Plus, get special
offers from American Pharmaceutical Review delivered to your inbox!
Sign up now!