Introduction: the problem of the Annex 1 draft around VH2O2 to ‘sterilize surfaces’

As many of you may already know, the draft of the new Annex 1 (v12) questions vaporized hydrogen peroxide (VHP/VH2O2) cycles as a method of surface sterilization for parts of indirect contact with the product. It is argued that VH2O2 has no penetration capacity, and that sterilization implies being able to penetrate to ensure the level of sterility. The truth is that the VHP/VH2O2 cycles show some fragility and sometimes positive indicators are detected unexpectedly. In the EU pharmacopoeia there is no explicit reference to VH2O2 as a method of sterilizing surfaces. On the other hand, there are references in the USP or even articles of the PDA that support the use of peroxide to sterilize or decontaminate surfaces.

Section 5.5 “Direct and indirect contact parts should be sterilized. Direct contact parts are those that the product passes through, such as filling needles or pumps. Indirect product contact parts are equipment parts that come into contact with sterilized critical items and components”
Section 8.12 “For sterile products that cannot be filtered, the following should be considered: i. All product and component contact equipment should be sterilized prior to use.”
In the case where vH202/VHP cycle is applied as the only bio-decontamination process for the barrier and process equipment non-product contact surfaces and in the same cycle ‘surface sterilization’ of in-direct product contact parts this would have to be justified using QRM principles. Such principles would need to recognize the different requirements for cleaning and bioburden control for bio-decontamination of non-product-contact process equipment/ barrier surfaces and ‘surface sterilization’ of in-direct product contact surfaces.

In the context of aseptic filling, eliminating this practice leads necessarily to an “out of place” sterilization: disassembling feed hoppers, guides, and other elements from inside the isolator, taking them to an autoclave chamber for sterilization and then reassembling them trying not to contaminate them.

This situation can involve problems such as that the autoclave chamber does not have enough capacity, the development of new SOPs, training and qualification of operators, productivity losses due to line stoppage, etc.  A real headache.

Photo: Courtesy of Dara Pharmaceutical Packaging

Differences between sterilization and biodecontamination

This article tries to clarify the concepts of sterilization and biodecontamination to understand the problem and try to find solutions.

1. Sterile, sterility and sterilization concepts

The first term is a binary concept. One thing, it’s either sterile or it’s not. It cannot be “half sterile”. By sterile we mean that product with total absence of viable particles.

As for the second term, sterility, or level of sterility assurance (SAL), it is a microbiological concept that designates the probability that a product undergoing a sterilization process contains a viable microorganism. This probability is, by definition, 10e-6.

And what does sterilization mean? According to the ISO 14161 standard for Sterilization of medical devices, sterilization is that process that can be validated starting from a microbial load of 10e6 until reaching the SAL of 10e-6. This literally means that a sterilization process must achieve a 12-log reduction.

In practice, it is impossible to prove such effectiveness. Why? Because to demonstrate this empirically, 1 million samples should be manufactured, each of them inoculated with populations of 10e6 spores and sterility tests of each one of them should be carried out. As this is virtually impossible, microbiologists use empirical data and mathematical formulas to calculate the lethal effect of heat and time needed in a sterilization process to secure the SAL.

2. Biodecontamination

Some studies have determined that VH2O2 has a penetration capacity of only 10-20 nm. A spore of Bacillus stearothermopillus is up to 3-μm long and 1-μm wide. Sometimes, during the manufacture of the indicators, agglomerations of spores (“clumping”) can occur and then the peroxide will not be able to penetrate beyond the first layer. This results in the indicator being positive, regardless of whether the process achieves a good distribution and has got contact with all surfaces. These indicators are called “rogue BI” because they give a false result. False as to the fact that the peroxide has contacted the surface, but true as to the positive is because it has not been able to penetrate inward. Hence the fragility of VH2O2 cycles and why it cannot be considered a sterilization process. In the case of autoclave sterilization, this would not happen since, under conditions of pressure and temperature at 121°C, the steam penetrates perfectly to reach the surface and these indicators would result negative. For this reason, in VH2O2 processes or cycles we should use the word ‘biodecontamination’ and not ‘sterilization’, which is reserved for penetrating processes.

3. Cleaning, bioburden and risk analysis

Does this mean that on surfaces it is impossible to achieve a SAL level (10-6) with VH2O2? Well, in theory, no. To achieve this, the process of cleaning and sanitizing surfaces should reduce biological contamination (bioburden) to a monolayer of microorganisms. The new annex seems to tells us that, if the VHP/VH2O2 cycle is used for decontamination of the isolator and filling machine of non-product contact parts and we want to take advantage of it as surface sterilization of indirect contact parts, we should use QRM or risk analysis techniques (see the quote above in bold) for those most critical parts, putting effort in the study of cleaning and bioburden.

Looking for solutions

Therefore, if we manage to sanitize at bioburden levels <10e3 and we design a 9-log reduction VH2O2 biodecontamination process, the process will achieve the SAL (10e-6). Unfortunately, there are no indicators with populations of 10e9 to validate these processes. Quite recently, at the Congress of Rapid Microbiology organized by A3P Spain Committee, the use of innovative microbiology techniques to better ensure aseptic processes was discussed. One of the speakers, the consultant Gilberto Dalmaso, defended that the pharmaceutical microbiologists should take the initiative and innovate in this aspect. He pointed out that the same Draft V12 of Annex 1 encourages the use of rapid microbiology methods with this sentence: “The use of appropriate technologies (e.g. RABS, Insulators, Robotic systems and environmental monitoring systems and / or rapid microbiology tests) should be considered to increase the protection of the product against contamination of personnel, materials, …”. We will see how it finally turns out, but it is time to take advantage of innovative technologies of rapid microbiology. One of them is the enzymatic indicators.  These indicators allow quantifying the lethal effect of peroxide by correlating the mortality of a population of spores with the remaining activity (RLU) of the enzyme Adenylate Kinase (tAK). Certainly, implementing such a process requires multiple tests and generating data over a period to validate it. However, the benefit in the form of productivity and risk reduction more than outweighs such effort and some multinational companies are already taking advantage of this tool.

This graph, taken from the article “Real-World Vapor Phase Hydrogen Peroxide Decontamination” by James Agalloco, is very illustrative to clarify concepts. As mentioned, neither 9-log reduction nor 12-log reduction to achieve a 10e-6 SAL can be tested experimentally. In the case of sterilization, we rely on the penetration capacity of the steam to ensure that SAL 10e-6 has been reached, while, with biodecontamination, assuming a maximum biological load of 10e3 we should develop a cycle that reaches a reduction of 9-log to have a lethality level of 6-log (or SAL 10e-6).

Conclusion

Biodecontamination and sterilization are analogous processes that involve killing populations of microorganisms but have different connotations. Sterilization is penetrating and this offers the necessary confidence to achieve the level of sterility (10e-6) according to the calculations of microbiologists. In contrast, biodecontamination is a more fragile process because it is superficial. It offers no such confidence and requires further attention. Therefore, it is necessary to carry out bioburden studies in the most critical areas and develop robust cycles. To validate a level of biodecontamination close to sterility, we must demonstrate that our cleaning and sanitization process achieves a low bioburden (<10e3) and then perform a VH2O2 cycle that allows reducing up to 9-log. Enzymatic indicators, which are part of the group of rapid/alternative microbiology techniques, are the perfect tool to support logarithmic reduction studies greater than 6-log and help validate such processes.