Biosafety of drugs, especially blood or biotechnologies derivatives, and medical devices is now a fact well rooted in the minds of scientists in charge of the development of those therapeutics and diagnostics. The succession of recent sanitary crisis, mainly linked to viruses such as HIV/AIDS, flu, SRAS, chikungunya or other agents, previously believed to be related to prions (also known as transmissible spongiform encephalopathy (TSE) agents) probably explains this raise in awareness.

The viral safety (for convenience, we associate prions to viruses in this document) of drugs and medical devices lies on at least 3 of the following issues:
-    For drugs:
1)    Safety of raw material and additives or excipients;
2)    Steps susceptible to eliminate or inactivate agents during manufacturing processes, and;
3)    Control of intermediary or final products.
-    For medical or surgical devices:
1)    Knowledge of the infectious risks and its load;
2)    Efficacy of the product or the sterilization process, and;
3)    Control of medical or surgical devices.

The regulatory agencies’ guidelines, national and international, state those issues, but others, closer to the methodologies to be implemented to demonstrate the safety or to the routine use, should also be taken into account.
1)    It is important to make sure that the miniaturized process (scale-down used to test one or more steps of manufacturing process) mimics in the best way the industrial scale reality. Similarly, for a physical or chemical process aimed at decontamination of medical devices, it is appreciated, or even recommended, to test the processes being routinely used in hospitals.
2)    Prions along with viruses are sensitive to environmental factors and to copy « field reality » is also of interest for the spiking of pathogen agent.

As a consequence, within the evaluation of processes able to eliminate or inactivate agents, conventional or non conventional, it will be of interest to test first the efficacy toward the « model » strain, then to confirm toward a strain as close as possible to the infectious risks i.e. risks for public health. So for a conventional agent like the human immunodeficiency virus (HIV), the sensitivity of clinical isolates i.e. as opposed to virus strains amplified in Labs using cell lines to the test product seems logical after an initial evaluation using the reference HIV-1-LAI strain. About prions, considering their extreme resistance, this concept is looked at for medical devices, at least. In fact, the “Standard Prion Protocol” validates the consensual side of the experimental model using golden Syrian hamsters infected with the scrapie 263K strain and the interest as confirmation of a further evaluation using a second animal model like C57Bl/6 mice infected with the 6PB1 strain, a bovine spongiform encephalopathy strain (BSE; pathogen for humans and the origin of vCJD) adapted to mice.  

The eventual multiplication of tests, as suggested by point 2), must imperatively go with technological alternatives, susceptible to lower their costs. On the contrary, it is disadvantaging the emergence of new efficient processes, and it is also questionable by the general public. The interest of in vitro approaches for prions is fully justified even if they do not measure infectivity but only the abnormal pathologic isoform of the prion protein (PrPTSE). Not knowing the exact nature of prions, these in vitro approaches are actually insufficient. But linked to an in vivo approach (bioassay), they allow identifying the efficient processes, and constitute help to the decision to pursue or not the explorations toward the in vivo assay and to break down the mechanism of action of the process.


Results obtained by the Advanced Sterilization Products (ASP) company to prove the efficacy of hydrogen peroxide plasma sterilizers against prions demonstrate the interest of juxtaposing the in vitro and in vivo approaches. Indeed, in order to document the efficacy of diverse generations of STERRAD sterilizers, this company has engaged, since 2002, series of researches in collaboration with three laboratories specialized in the field of medical devices and prions.

As a reminder, prions are responsible for neurodegenerative diseases with slow evolution and always fatal, known as TSE. Transmissible spongiform encephalopathies are described for animals and humans; human TSE are Kuru, fatal insomnia, Gerstmann-Sträussler-Scheinker (GSS) syndrome and the Creutzfeldt-Jakob disease (CJD) with its variant (vCJD), better known as « Mad cow disease». Prions are characterized by their composition probably purely of proteins and by their high resistance, unusual, to inactivation processes used in microbiology.

This extreme resistance is a major handicap in the decontamination of medical devices, eventually infected. Today, in France, the circular DGS/5C/DHOS/E 2001-138 of March 14th, 2001 remains the reference even if the update is in progress and, classifies processes into 5 categories with regard to their efficacy. So, aside from the destruction of medical devices (Group V), this circular positions important efficacy products or processes in Group III (steam at 134°C for 18 minutes, sodium hydroxide 1N for 1 hour, or 2% of active chlorine for 1 hour) and those of maximum efficacy in Group IV (combination of processes of Group III, sodium hydroxide or chlorine associated with steam at 134°C). Those efficient treatments are poorly appropriate for thermosensitive materials, constituting endoscopes, for example. As a result, it was important to identify “soft” alternatives in order not to deteriorate endoscopes and enhance the safety of users. Hydrogen peroxide could, for many reasons, appear to be the best solution; on one hand, it possesses a large range of efficacy regarding conventional agents and on the other hand, it is poorly abrasive toward thermosensitive materials.

In this idea, ASP initiated research using sterilizers strictly identical to those used routinely in hospitals. The results of 3 laboratories engaged in this program establish that the efficacy level of hydrogen peroxide depends on these conditions of study design. In fact, the ancient generation of sterilizers, tested in the absence of any pretreatment, has been shown to be less efficient than the autoclave at 134°C for 18 minutes, positioned in Group III of the circular DGS/5C/DHOS/E 2001-138 and as a reference in France for medical devices. However, the new generation of ASP sterilizers is comparatively more efficient regarding prions than this same autoclave at 134°C.
Those results have been obtained using the consensual experimental model of golden Syrian hamster infected using stainless steel wires artificially contaminated with the 263K scrapie strain. Other results more recent, obtained in vitro, permitted to confirm this efficacy. In fact, the new generation of ASP sterilizers happened to be efficient during those in vitro tests toward all strains of prions tested, especially a vCJD strain, and for the 3 supports tested (stainless steel, polyethylene and polypropylene). In parallel, parameters that seem to govern ASP sterilizer efficacy have been identified.

For prions decontamination, if steam at +134°C remains the absolute reference, processes such as hydrogen peroxide seems to constitute an interesting alternative, to guarantee the safety of heat-sensitive devices and their integrity throughout their routine use in hospitals. These new in the prions field highlights the fact that the viral safety, either for drugs or medical devices, must then rely on reference elimination or inactivation processes, while integrating new emerging technologies. New are not only the technology side but also on the side of microbiological risk. The problematic is then: “how to manage a new microbiological risk?” or more precisely “is it possible to guarantee sufficient security regarding a newly identified agent by relying on the pre-existing scientific data?”. The experimental approach aiming at testing the susceptibility of the new agent for products or processes is still the best guarantee. However, if no validated experimental models are available and if the structure of this new agent are unknown, an extrapolation of data collected from tests using a « model pathogen » is a temporary alternative, if possible.