“Biological Indicator (BI): A standardized preparation of bacterial spores on or in a carrier serving to demonstrate whether sterilizing conditions have been met. Spores of different organisms are used for different methods of sterilization”.

The key word in this definition is “standardized”. BIs have been used, and are used, to validate sporicidal gassing cycles, and it is not uncommon to find that a sporicidal gassing cycle that has been validated using “standardized” biological indicators fails when revalidated at a later date.

The failure of a revalidation cycle may have been caused by a failure of the sporicidal gassing process, a change in the system being gassed, an environmental change or the biological indicator.

The Parenteral Drug Association (PDA) has formed a Task Force to prepare a monograph on the “Recommendations for the specification, manufacture, control and use of biological indicators”, to assist users of BIs with the selection, application and quality control measures that are necessary to increase the reliability of BIs. However, at the present time the BI is probably the most variable factor in any sporicidal gassing process.

For the purposes of this paper I shall confine myself to discussing biological indicators as commonly used for the validation of hydrogen peroxide sporicidal gassing.

Construction of a Biological Indicator.

The carrier material is most commonly stainless steel, although it should be understood that the material from which the carrier is manufactured, its shape and surface finish will all have an impact on the performance of the finished BI.

The chosen microorganism should have a high resistance to hydrogen peroxide vapour and a higher resistance to the sporicidal gassing process than the naturally occurring microorganisms.

The most frequently used organism for BIs for hydrogen peroxide sporicidal gassing is Geobacillus stearothermophilusspores. They have a proven resistance to hydrogen peroxide vapour, are non-pathogenic and incubate at 55 to 60ºC, thus reducing the risk of cross contamination.

The primary pack is intended to protect the inoculated coupon from damage and prevent the escape of the spores. It should also be pervious to the hydrogen peroxide vapour, allowing the vapour to reach the spores that have been inoculated onto the carrier.

Decimal Reduction or D value

Before discussing in more detail the construction and performance of BIs it is necessary to define D value. The Decimal reduction or D value is the time it takes to reduce the viable population of the microorganisms by a factor of 10, when subjected to a constant stress level. Thus if the D value is 2 minutes, and the initial viable population is 1,000,000 then after 2 minutes the viable population will have been reduced to 100,000 and so on. Therefore after 10 minutes the viable population would be 10. The D value concept is useful in predicting the performance of a BI, and making comparisons from one source of BIs to another. The other quality control measure that will be discussed later is the number of spores on the carrier.

The Carrier Material

Several studies have been published to establish if the performance of the BI is affected by the material of the substrate. These have produced conflicting results. A study undertaken by Davenport and reported at an ISPE Barrier Conference showed a variation in D values from 1.1 for Hypalon to 1.7 for Stainless Steel. Similar studies have different results probably because of different vapour concentrations.

The variation in D value seems likely to be the result of a reaction between the substrate and the vapour causing decomposition of the hydrogen peroxide, or a change in the mechanisms causing condensation. It is well known that condensation forms more readily on a nucleus such as a dust particle.

It has also been shown that condensation will form preferentially on microorganisms when presented on smooth surfaces, preventing condensation in the immediate surroundings. If the surfaces surrounding the microorganism are rough then the condensation may equally form on the surface, and not on the microorganism, thus prolonging the time required to achieve a kill.

At the present time we do not have an understanding of the mechanisms that cause an apparent change in the D value on different substrates, it is therefore necessary to carefully control the quality of the material used in the manufacture of the carrier, and the surface finish.

Microorganism

There is a recommendation in the European Pharmacopoeia that Geobacillus stearothermophilusshould be used for validation of hydrogen peroxide sporicidal cycles.

Under some circumstances it may be expected that other microorganisms are likely to be encountered that may have a greater resistance to hydrogen peroxide vapour than G. stearothermophilus It is then necessary to conduct some studies to establish the relative resistance of the suspected microorganisms compared with G. stearothermophilus, after which an informed decision may be made as to the correct microorganism to use.

Primary Pack

The primary pack is intended to protect the inoculated carrier, and also to prevent the escape of the spores. It should also not prevent the vapour reaching the target microorganisms. A number of tests have been conducted to establish the affect of the primary pack on the apparent D value. It would be more accurate to discuss the change in stress level experience by the microorganisms, rather than a change in D value, since at any given stress level the D value should be the same.

What the primary pack provides is a barrier that the vapour must pass through in order to reach the microorganisms, and if the process is one in which condensation is expected then it is probable than condensation will form on the primary pack inhibiting the free movement of the vapour to the inside of the pack, thus significantly reducing the stress level.

It is interesting to note that even with this barrier that the condensing process is still much faster than the dry gas process as indicated by the length of time taken to achieve a complete kill of the BIs.

Temperature Affects

At the time that hydrogen peroxide vapour bio-decontamination was introduced in the early 1990’s it was thought that it was a dry gas process, but contrary to expectations it was found to be more difficult to achieve a kill in areas within chambers that were at higher temperatures. This anomaly was originally thought to be because the hydrogen peroxide decomposed more rapidly as the temperature increased.

In fact an increase of only a few degrees in the surface temperature caused a problem. This is counter intuitive because chemical reactions proceed more quickly as the temperature rises and hence it may be expected that at higher temperatures the kill time would be reduced. The explanation of the extended kill time at higher temperatures is that the process relies on condensation to be effective and areas within the chamber at higher temperatures will be the last to attract condensation.

It follows that when BIs are used for process control they should be at a temperature equivalent to the surfaces that they represent. Any increase in the temperature of the BI compared with its surroundings will increase the time required to achieve a kill.

Most BIs are manufactured by inoculating spores onto small discs of stainless steel. The BI is then suspended inside the chamber to be bio-decontaminated, not in thermal contact with the surroundings it is intended to represent. The gassing process adds heat to the chamber as the vapours are introduced at an elevated temperature.

Items within the chamber that have a low thermal mass and are not in thermal contact with the surrounding surfaces will heat up faster than the surroundings. The BIs are a prime example of low thermal mass not in contact, and as a result exhibit extended killing times over the surfaces that they are intended to represent.

Standard Commercial BIs

Commercial BIs are available from a number of sources. Each lot of BIs should be accompanied by a test certificate which as a minimum should state, the D value and the conditions under which it was determined, the mean population, the expiry date, the lot number, the organism, the storage conditions, and recommended incubation conditions.

This information may, however, mask the poor quality of a BI. The PDA Task Force had scanning electron microscope photographs taken of a number of commercially available BIs, all of which had a claimed population of about 10⁶Geobacillus stearothermophilusspores on stainless steel carriers; four of which may be seen in Fig 1, and as can be seen are of very different quality.

The top left photograph shows clean spores from a clean suspension allowing the vapour to come into contact with the microorganism. The other three photographs all show considerable amounts of debris and clumping which the vapour must penetrate before reaching the spore. This will inevitably lead to variations in the performance of the BI and difficulty in process control, and will not be indicated in the manufacturer’s test certificate.

Scanning Electron Microscope Photos of 4 Biological Indicators - By Kind Permission Of The PDA Biological Indicator Task Force

It is not uncommon to find D values that vary from 1.0 to 2.0 minutes from lot to lot, even from manufacturers using high quality clean spore suspensions. This variation in D value would indicate that a log 6 reduction would be achieved in either 6 or 12 minutes. Not a very wide window when considering a bio-decontamination cycle. But looked at another way this is a variation of 100% when measured against the 1.0 minute D value at the stress level used by the manufacturer.

In practice the stress level may be much lower in a real chamber and it has been shown that at lower stress levels the percentage difference in D values increases. Thus an increase of 100% between batches may become an increase of as much as 180% at a lower stress.

If BIs with the lower 1.0 minute D value are used in a validation of an isolator with a gassing cycle of 30 minutes and the re-validation is undertaken with a BI having a D value of 2.0 minutes it is likely that the process will fail, as the safety margin may not be sufficient to deal with the increased resistance of the BI.

A test was performed using three lots of BIs, with D values reported by the manufacturer of 1.2 (Lot 53), 1.4 (Lot 35), and 1.6 (Lot 24) minutes. Samples of these three lots were gassed with Hydrogen peroxide in a 110 m³ room with a maximum vapour concentration of 450 ppm and a starting RH of just above 45%.

The BIs were placed in TSB at timed intervals and incubated. The time taken to achieve full kill is shown in Fig 2. Whilst the difference in D value between Lots 24 and 53 is only 0.4 minutes or 33% the difference in the kill time in the room is 100%.

Because of the sampling time in the room is relatively long, it is possible that Lot 53 was killed after 6 minutes and Lot 24 after 16 minutes. It is therefore arguable that the difference in kill time between these two Lots was 16 minus 10 minutes i.e. 6 minutes, giving a minimum difference in the kill time of 60% compared with a difference in the D value of 33%.

It is likely that the very large differences in the reported D value by the manufacturer, and the kill time achieved in the room, are the result of the differences in the test conditions. The most important of which is the vapour concentration; the manufacturer’s D value was established at about 1350 ppm compared with the room test at a maximum concentration of 450 ppm.

Stress Levels.

Biological Indicators are compared using the D value and the population, but D values are only meaningful at a stated and constant stress level. We have already seen that as the stress level changes so does the D value, but in a real bio-decontamination cycle the stress level is never constant.

If we consider the supposed dry gas cycle as the gas concentration increases so will the stress level, but in the condensation process the situation is even more complex. (A study of the physical chemistry of the process will rapidly demonstrate that all hydrogen peroxide bio-decontamination cycle cause condensation).

In a condensation cycle the stress level will increase as the vapour concentration rises to the dew point, at which time there will be a sudden increase in the stress level as the first droplet of condensation forms. The concentration of the condensate also changes with time, and under the circumstances usually encountered will decrease as the cycle proceeds. However, as the mass of condensate increases so will the stress level until it reaches a maximum for that concentration of condensate. Thus for a real cycle the stress level is constantly changing.

Population

Because of the difficulties encountered in the manufacturing process the generally accepted criterion for the population of a BI is -50% +200%. This by engineering standards is very wide. Verification of the population also causes difficulties because different techniques will release different numbers of spores from the substrate, and subsequent enumeration methods require very good laboratory technique to obtain consistent results.

Summary

It is difficult to conceive of a validation process that relies on an indicator with such variability. In the space available I have discussed some, but not all, of the issues that will cause variation in the performance of a BI. Much more detailed information will be available as soon as the PDA Task Force publish their report. There is no doubt that considerable time and effort has been expended as a result of the failure of a validation cycle, not because the cycle was at fault, but because of variations in the BI.

To minimise the problems at re-validation it is essential that batches of BIs are subjected to stringent quality control checks. This should include an enumeration of the spore population and a measure of the resistance of the BI. The resistance measurement should be made under conditions that as near as possible represent the stress conditions that the BI will experience in use.

Lastly, if a re-validation cycle fails consider all of the factors that may be the cause, the BI, the cycle, changes to the equipment, changes in the environment, and whilst I believe that the most frequent cause of failure is the BI that is not always the case.