A More Rational Approach for Sterile Product Manufacturing

For decades, two different approaches have been used to manufacture sterile products. The most common approach, aseptic processing, relies upon separating potential microbial contaminants from the product through a combination of filtration of the product stream, individual sterilization of product-contact packaging components, and environmental controls. Terminal sterilization relies on a more limited contamination control of the product, components, equipment, and environment followed by a lethal sterilization process applied to the fully assembled dosage form. From a macro perspective, these two approaches have been treated as equivalent means to the same end, but they have never been considered equal in terms of process capability. In the jargon of sterile-product manufacturing, they are not considered fully equivalent in terms of sterility assurance.

Regulatory perspectives

The world's regulatory authorities have long expressed a preference for terminal sterilization, but have rather inexplicably set up expectations that provide little real benefit to firms that implement terminal sterilization. The foremost shortcoming is requiring firms to make a regulatory filing to delete the compendial sterility test as a requirement for product release, a practice known as parametric release. Regulatory pressure exists to manufacture products for terminal sterilization in ways increasingly similar to aseptic processing. This change involves added facility controls, stricter gowning requirements, and substantial environmental monitoring that further deteriorate any financial benefit that would accrue to firms using terminal sterilization. In Europe and to a lesser extent in the United States, another blow against terminal sterilization has been the increasingly commonly enforced idea that terminal sterilization should require lethalities in the range of F_° ≥15 min. (F_° refers to lethality in moist-heat sterilization; by convention, F_° is equal to 1 min of exposure to moist heat at 121.1 °C.)

In Japan, a different and far more scientifically logical approach has evolved and is undergoing further refinement. The idea is that with reasonable control of presterilization bioburden, lower levels of lethality can afford high levels of patient safety. Specifically, for reasonably heat-stabile products F_° values of 2 min or less can provide levels of end-user safety substantially higher than those attainable in aseptic processing. The logic of this approach is irrefutable as can be shown by simple microbiology.

Although spore-bearing gram-positive rods with substantial heat resistance are used to develop and validate sterilization cycles, the most common human and animal pathogens are bacteria, mold, and virus with substantially lower heat resistance than these spore-bearing organisms. Protein, a key structural and functional component of microorganisms, is typically denatured in vegetative mold, bacteria, and virus at temperatures only slightly greater than 56 °C. Unsurprisingly, these organisms die quite readily at temperatures in the range of 65–75 °C. This difference means that substantial increases in user safety could accrue at temperatures that are substantially lower than those at which spore killing begins and at which F_° begins to accumulate, which is in the range of 100–105 °C. Thus, the vast majority of medically significant organisms would be dead well before a postassembly heat process reached F_° = 0.1 min. The improved patient safety that this approach could add to moderately heat-stable products is so obvious that one wonders why it is so steadfastly ignored. In fact, a process need not get to 100 °C to add substantial safety, five minutes or less at 70 °C or so would be appropriate.

One used to applying the common pharmaceutical approach to sterilization might ask whether a process that did not kill spores had value? Aseptic processing is a completely nonlethal process and yet is widely accepted. Of course, a moist-heat sterilization of F_° = 2 min would effeciently kill many spores of many spore-bearing species, including those known to produce human or animal disease. (The recommended means to deliver this process is at temperature less than 121 °C, whereby more uniform conditions can be delivered through the use of a much shorter time at an elevated temperature). Most spores isolated from industrial environments have D₁₂₁ values of < 0.2 min. (D₁₂₁ is the time required to reduce a microbial population by 90% when exposed to moist heat at 121 °C.) Given the conservative assumption that the presterilization bioburden following an aseptic processing would be no more than one organism per container and the organisms present had a D₁₂₁ value of 0.2 min, a process yielding an F_° = 2 min would produce a probability of nonsterility of 10^-9. In medical terms, this calculation dramatically understates the actual safety because the majority of the bioburden (especially those considered human pathogens) would be non-spore forming and would therefore be dead while the process was still in heat-up mode.

Parametric release

Dr. Tsuguo Sasaki, GMP Expert, Office of Compliance and Standards of the Japanese Pharmaceuticals and Medical Devices Agency (PMDA) advocates that F_° ≥ 2 min would be a reasonable expectation for parametric release. The authors strongly agree and would favor a harmonized regulatory approach in which any firm with acceptable GMP compliance that submitted an application for a product sterilized at a process lethality of F_° ≥ 2 min would be granted parametric release for its product without any added expectations. This proposal is radical only because existing regulatory policy fails to properly consider true medical risk.

Parametric release simply means releasing product without having subjected it to a compendial sterility test. The real question is what value the performance of a sterility test on a process offers. Given the low sensitivity and statistical sampling limitations of the sterility test, it has limited value and really should be called a test for gross product contamination. It is rarely failed when applied to aseptically produced products and simply has no value when used to test products sterilized in their containers.

Postfill treatments

Lower temperature heat treatments would provide significant safety benefit to products that are filled in human-scale cleanrooms. The benefit of these treatments would diminish in instances where advanced aseptic processing technologies, such as isolators, closed restricted access barrier systems, or some blow, fill, and seal processes are used. The authors postulate that comparable approaches could be used with radiation sterilization where low-dose exposures would provide similar advantages to patient safety. ISO 11137-2 provides for developing relatively low does processes based upon bioburden number and resistance (1).

Recommendations and experience from Japan

PMDA's Dr. Sasaki shared with the authors some basic concepts from Japan's Guideline on the Manufacture of Sterile Drugs by Terminal Sterilization, which is expected to be issued in 2012. This document will consist of a narrative section and annexes. The following information reflects general recommendations for low F_° terminal sterilization and will be presented in one of the guideline's annexes:

• Minimum F_° is 2.0.

• Cleanliness of filling area is Grade A, and surrounding areas are Grade B. (Grade C may also be used but personnel microbial monitoring would be expected to meet Grade B levels.)

• Average presterilization bioburden should be less than 1 colony-forming unit (cfu)/product unit.

• Bioburden should be evaluated for heat resistance.

• Grade A is used for filling, but media fill testing is not required.

Low-dose terminal sterilizational is in widespread use in Japan. Dr. Sasaki provided some data regarding its application in Japan during 2005–2010 from 13 different manufacturers (see Table 1). Nearly 80% of the roughly 2.8 trillion units manufactured by terminal sterilization were subjected to F_° of two minutes or less. Quite clearly, these low lethality sterilization processes would not be allowed were they not safe and efficacious.

Table I: Application of low-dose sterilization from 13 manufacturers in Japan from 2005–2010.

Looking ahead

The re-examination of the regulatory philosophies that underpin sterile-product manufacturing is long overdue. There is a failure to recognize that postfill heat and radiation treatments could provide additive safety benefits and that these adjuncts to aseptic processing may be applicable to a significant range of products. A process need only be as lethal as necessary to destroy the bioburden present. The current gulf between aseptic processing and terminal sterilization can be bridged as the information from Japan clearly illustrates. At the same time, these data indicate that true terminal sterilization can be applied far more frequently than it is currently by the industry if one considers that low lethality processes can yield safety levels that are so good that further improvements would provide only theoretical medical benefit.

What has to happen other than this re-examination of the fundamental regulatory philosophies for sterile-product manufacturing? The authors suggest the following:

• Item 1. The realization that most medically significant organisms die at temperatures and conditions far lower than where modern industry considers that sterilization begins.

• Item 2. The universal recognition that the F_° principal has tremendous value and relevance and that there is nothing magical about 121.1 °C as a minimum target temperature for sterilization.

• Item 3. Increased awareness that parametric release should require evidence only that a process does not benefit from the application of the compendial sterility test as a quality-release requirement.

• Item 4. Awareness that Item 3 requires disavowal of the notion that there is a test to prove sterility or that such a test could be devised. No microbial test, be it growth-based or rapid, has a limit of detection of zero. It is impossible to prove a negative absolute.

• Item 5. Understanding and accepting the clear scientific principal that neither safety nor sterility hinge upon the ability to kill 10⁶Geobacillus stearothermophilus spores per article. Organisms isolated from extreme environments, such as hot springs and other geothermal features, are not and cannot be medically significant.

• Item 6. We can learn from the extensive experience of Japan with respect to lower lethality terminal sterilization.

The authors urge broad international discussions on sterile- product manufacturing and strongly suggest that these discussions be undertaken with the understanding that patient safety must be evaluated from a medical perspective only and not from the viewpoint that begins and ends with current regulatory paradigms. Many modern-day sterile-product manufacturing paradigms are mistaken and unduly inflexible. The result is a stifling of innovation and an overreaction to misperceived risk while thwarting simple steps that would add safety and reduce cost.

Acknowledgment

The authors thank PMDA's Dr. Sasaki for the information he provided for this article and for the many discussions that have been instrumental in exploring the options examined in this article.

James E. Akers*, PhD, is president of Akers Kennedy and Associates, PO Box 22562, Kansas City, MO, 64113, AKAincKC@aol.comJames P. Agalloco is president of Agalloco & Associates. He is also a member of Pharmaceutical Technology's editorial advisory board.

*To whom all correspondence should be addressed.

Reference

1. ISO, ISO 11137-2 Sterilization of Health Care Products. Radiation (Geneva, 2012).