Test for sterility

 

In the European pharmacopoeia (Ph. Eur. 4), there are various monographs concerning the requirement for sterility (see figure 12.H-1).

It is not specified that a test for this must be carried out. It is only asked that, if the test is carried out, the substances/preparations/products must conform to the test for sterility. This is expressly referred to in the "Guidelines for Using the Test for Sterility" ("A manufacturer is neither oblieged to carry out such tests..."). This pharmacopoeia method is the arbitration method that is to be implemented in the event of disputes ("The purpose of the test for sterility, as that of all pharmacopoeial tests, is to provide an independent control analyst with the means of verifying that a particular material meets the requirements of the European Pharmacopoeia") and that serves as a standard for changes to the method (the manufacturer is not precluded from "using modifications of or alternatives to the stated method provided they are satisfied that, if tested by the official method, the material in question would comply with the requirements of the European Pharmacopoeia").

12.H.1 Parametric release

Although the introduction to the "Sterility test chapter" 2.6.1 in the pharmacopoeia states that sterility testing must be carried out on substances, preparations or products for which sterility is specified, this does not in any way mean that testing must actually be carried out - as already mentioned. Because of the statistical uncertainty of the "sterility testing", parametric release is preferable for products sterilised in the final container. The pharmacopoeia specifically agrees to this (see figure 12.H-2).

However, chapter 5.1.1 does not contain any further details about parametric release other than the instruction that the process must be validated. For this, we can turn to the PDA Technical Report No. 30, which describes parametric release for products sterilised with steam in the final container (see figure 12.H-3).

However, it is not so easy to switch from the sterility test to parametric release; the authorities responsible must agree. This is particularly problematic if the product has been registered in several countries and every authority with which the sterility test is recorded for the drug product concerned must be notified and give their consent.

There are now papers relating to this on the "official side". Annex 17: Parametric release to the EU-GMP-Guideline, is closely related to a relevant EMEA Note for Guidance. The Pharmaceutical Inspection Convention (PIC) has also issued a relevant guideline. These papers go further than the PDA paper: as well as parametric release of sterile drug products, it is also generally discussed for other tests.

12.H.2 Sterility test

If the sterility test is selected for products sterilised in the final container, or if parametric release is not possible, e.g. for aseptically filled products, it is advisable to carry out the sterility test in accordance with the specifications in the pharmacopoeia. A different procedure would also be possible here (page 1), but it may result in endless discussions in the event of an inspection.

The following section covers the most important points of the test. The sterility test included in Ph. Eur. 4.6 is the result of an international harmonisation to the end that a test exists that is recognised in Europe as well as in the USA and Japan. In addition to a simple explanation of the test, the most important changes in comparison to the previous test should be highlighted. As not all points could be harmonised, the points that will be handled differently in the future in the USA and Japan than in Europe will also be mentioned.

12.H.2.1 Environmental conditions

The sterility test must be carried out under aseptic conditions in order to prevent secondary contamination. In order to achieve this, the test environment must be adapted to the test procedure. However, the measures implemented must not affect any existing contaminants.

In the non-harmonised, non-binding Guidelines for Using the Test for Sterility, for example, the European pharmacopoeia mentions the use of a laminar flow bench that should be present in a class B clean room (according to the requirements of the EU-GMP-Guideline for the manufacture of aseptic products) or the use of an isolator. There is a number of points that speak in favour of an isolator:

• High level of result reliability due to maximum product protection (reduction of the secondary contamination rate)
• FDA investigators recommend the use of isolators for the sterility test.
• Tried and tested technique in North America (in <71> Sterility Tests, the USP expressly refers to the use of isolators for the sterility test).
• High degree of flexibility for personnel deployment because the air lock procedure is not required.
• Low space requirements (normal sterile test isolator approx. 1 x 2 x 2 m)
• No requirement for an expensive HVAC system for the sterile test rooms and their maintenance (legislators do not require any special environmental conditions for isolators. In Ph. Eur. there are no background requirements for isolators, USP Sterility Testing - Validation of Isolator Systems: Isolators for sterility testing need not be installed in a classified clean room)
• No costs for sterile clothing (purchase, care, laundry and sterilisation)
• Minimal routine monitoring in the isolator, no monitoring of the environment or personnel.

However, it must be made clear that the cost for a sterile test isolator with validation will cost around 400,000 euros. This cost quickly proves its worth though, when we consider that "false-positive" results can be avoided. This usually means that batches that are actually sterile are not quarantined because of the impracticality of carrying out follow-up tests (see chapter 12.H.6 Reading and evaluating).

12.H.2.2 Environmental monitoring

The work conditions specified for carrying out the tests must be regularly monitored with suitable environmental checks in the working area and using checks of the aseptic work methods.

Monitoring of the hygiene status in sterile test rooms

In figure 12.H-4 and figure 12.H-5, there is a suggestion for the implementation of the requirement for environmental checks in the sterile test room which is based on the limits and frequencies for monitoring production.

The usual methods for aseptic processes are used for monitoring a clean room (chapter 12.G Microbiological monitoring). There are no far-reaching requirements as in aseptic processes.

Monitoring of the hygiene status in isolators

A main advantage of using sterile test isolators is that the tests are carried out in a sterilised environment free from personnel. This reduces the contamination risk to zero. However, it is a mistake to think that you can get by without monitoring. This is clearly stated in Annex 1 to the EU-GMP-Guideline: ("Monitoring should be carried out routinely and should include leak testing of the isolator and glove/sleeve system.").

The question of which limits to set then arises. It seems logical to apply the normal air limits for room grade A, but it has become accepted to require 0 CFU/m³ air in the sterile test isolator. This is also in line with the suggestions in the FDA Aseptic Guidance, where no microorganisms are tolerated in the air for aseptic processes (Samples from Class 100 ... environments should normally yield no microbiological contaminants). 0 CFU/25 cm³ must be achieved for surfaces as this is of course a true, validated sterilisation (see figure 12.H-6).

There is however uncertainty about the scope of the test for checking isolators. A suggestion is made in figure 12.H-7.

The monitoring methodology is particularly problematic. One must anticipate residual H₂O₂ in the room, which is emitted from the damp agar surfaces and which prevents any microorganisms present from growing. Proving growth promotion using control organisms (positive controls) takes on particular significance here. For this, max. 100 CFU Staphylococcus aureus ATCC 6538 is dripped onto the plates or strips with no growth following incubation. This demonstrates that no organisms would have been able to grow.

For air testing, it is possible to try to come to a satisfactory solution using multiple extractions (e.g. 2 x 500 l with RCS). For the RCS unit, there are now agar strips on the market with a suitable neutralisation additive which alleviate the problems of proving air microbial counts. For surface contact testing, Becton-Dickinson have brought a suitable contact plate onto the market which contains D/E neutralising agar. Plates with a suitable neutralising additive should also be used for a collector with solid culture medium petri dishes.

Aseptic working methods

The aseptic working method testing required by the pharmacopoeia (negative controls) must be regularly monitored using proven sterile preparations. These controls must be carried out at the same time as the testing, as the result must be used when making a decision about any follow-up tests (see chapter 12.H.6 Reading and evaluating). If different methods are implemented (membrane filter/direct inoculation of the culture medium) and drug products are tested that require different preparation types (e.g. aqueous ampoules, vials with lyophilisates, etc.), the test must be carried out for the methods and the preparation types. If several people carry out the test, negative controls are also required for each person.

12.H.3 Method description

The pharmacopoeia contains method descriptions for the membrane filtration and direct inoculation methods. Membrane filtration is stipulated as the method of choice (The technique of membrane filtration is used whenever the nature of the product permits ...)

The membrane filter method is described for the following dosage methods:

• Aqueous solutions
• Soluble powders
• Oils and oily solutions
• Ointments and creams

The direct inoculation method is described for the following dosage forms:

• Oily liquids
• Ointments and creams
• Catgut and other surgical suture materials for veterinary use

There are a few small modifications here that may be extremely useful. For aqueous solutions, for example, a solution quantity is not stated, but validation is suggested. The wash cycles are limited to 5 x 200 ml, unless a validation study shows that this is not sufficient to wash out antimicrobial substances. (Japan requires 5 times 100 ml per filter for this.)

In addition, there are detailed descriptions for testing parenterals, ophthalmics and other preparations not intended for injection. These have been adopted virtually unchanged. There are only two changes. The instructions "The total volume filtered through a membrane must not exceed 1000 ml, except in justifiable and authorised cases" (for membrane filtration) and "If the filling volume of the sample container is more than 100 ml, the membrane filter method should be used, except in justifiable and authorised cases" (for the direct inoculation method) were difficult to justify objectively and have been removed and not replaced.

12.H.3.1 Incubation

The harmonised text stipulates a minimum of 14 days' incubation. Liquid thioglycolate medium (for determining bacterial contamination) 30-35 °C and soya-bean casein digest medium (for proving fungal contamination) 20-25 °C. This means that the previous option in the USP for "seven days' incubation for products sterilised in the final container" can no longer be used.

12.H.4 Number of samples

The numbers of samples were agreed as part of the harmonisation, but only USP and JP want to include these numbers in the binding part. In Ph. Eur., they are in the non-binding "Guidelines for Using the Test for Sterility" as before (figure 12.H-8). It is certainly a good idea to abide by these suggestions. Otherwise the table is unchanged; only regulations for testing bulk materials have been added.

When determining sample numbers, the minimum test quantities must also be taken into consideration (see figure 12.H-8), as the number of samples must be doubled if the entire content of a container is to be tested.

12.H.5 Sample quantity

The required test quantity is contained in a table in the main part of the pharmacopoeia (and is therefore binding). As part of the harmonisation, some of the quantities to be tested have been changed (which may mean revalidation of the methods used). An overview of the requirements is given in figure 12.H-9 and figure 12.H-10, which also shows the previously required quantities, so that it is easy to see whether validation is required for drug products that have already been introduced.

12.H.6 Reading and evaluating

Several times during and after the incubation time, the cultures are checked macroscopically for visible growth of microorganisms. If there is growth, the test can be interrupted and then evaluated.

If no growth is determined, the tested product complies with the sterility test. However, if growth is detectable, the product does not satisfy the requirements, unless the test can be proven to be invalid for reasons that are not related to the product itself. The test can only be viewed as invalid and repeated if one or more of the conditions summarised in figure 12.H-11 are fulfilled.

Point four in figure 12.H-11 ("after determination of the identity of the microorganisms isolated from the test, the growth of this species or these species may be ascribed unequivocally to faults with respect to the material and/or the technique used in conducting the sterility test") is only sufficient grounds for a repetition if the organism species can be clearly assigned. A similar species with conventional identification methods is not sufficient in this case. This is expressly stated in the non-harmonised, non-binding "Guidelines for Using the Test for Sterility". If the identification of the microorganisms is to be used as the only criterion for declaring the sterility test invalid, more sensitive techniques must be applied as microbiological/biochemical methods, e.g. molecular typing with RNA/DNA homologies.

Other reasons for the sterility test being invalidated are conceivable, for example, if a test was carried out in the isolator and there were problems with the integrity of the system. This instruction was contained until now in the USP.

 

12.H.7 Procedure in the event of culture medium turbidity

If the material to be tested clouds the culture medium so that the existence or otherwise of microbial growth is difficult to determine at the end of the incubation time, suitable quantities of the culture medium must be transferred into fresh containers with the same culture medium. The main restraint when carrying out a test, as stated by both the USP and Ph. Eur., was how to proceed in the event of culture medium turbidity. The Ph. Eur. required: The incubation of the original container and the container with the transferred culture medium to be continued for a total at least 14 + 7 days, calculated from the original inoculation.

The USP required sub-cultures to be created between the third and seventh day and the sub-cultures to be incubated for 11 to 7 days, so that they had a total incubation time of 14 days. Although this incubation time is pleasantly short in comparison to the Ph. Eur., this procedure is somewhat problematic. Firstly, when the incubation time was determined by the USP for the original test, it was assumed that organisms require 14 days to multiply to a quantity visible in a liquid medium with the direct inoculation method. If there is turbidity (which generally occurs with the direct inoculation method), it is now assumed that after the third day there is a sufficient quantity of organisms for enough organism material to be detected when transferred into a new culture medium container. This is an unfortunate misconception. This long incubation time of 14 days is not necessary because the organisms grow so slowly but because the organisms often have a very long lag phase in which they do not grow. A problematic organism for aseptic filling is Propionibacterium acnes, for example, which in our experience often does not develop into a visible organism concentration in the culture medium until after the twelfth day. These organisms are certainly not detected with the procedure suggested by the USP.

This major difference has now been rectified. The harmonised text now states that 14 days after the start of incubation, portions (each no less than 1 ml - JP requires transfer suitable portions and omit here) of the culture medium should be transferred to fresh containers with the same culture medium. The incubation of the original container and the freshly inoculated container should then continue for at least four further days so that there is a total incubation time of 18 days.

12.H.8 Culture media

The two pharmacopoeias name the following as culture media:

• Fluid thioglycolate medium
  This is primarily used to detect anaerobic bacteria, but can also be used to detect aerobic bacteria.
• Soya-bean casein digest medium
  This is mainly suitable for detecting aerobic bacteria and fungi.

Other culture media can be used on the condition that there is evidence of its growth promoting properties and validation tests have been carried out. It is not advisable to make use of this opening clause, as neither Japan nor the USA will adopt it.

Very detailed information is available about storing culture media:

• Fluid thioglycolate medium
  This must be stored in sterile, air-tight containers between 2 and 25 °C. If more than the upper third of the medium turns pink, it can be made useable again by heating until the pink colour disappears then rapidly cooling it. The medium may only be stored in the long term with the relevant validation (JP and USP will not adopt this clause). The question of what "long term" means is unanswered - but certainly longer than 1 month. If the medium is stored in unsealed containers, it may be used for a month with the prerequisite that the growth promoting properties have been tested within two weeks of the medium being used and the colour indicator does not cover more than the upper third.
• Soya-bean casein digest medium
  This must be stored in sterile, well-sealed containers between 2 and 25 °C, unless the medium is to be used immediately. The medium may only be stored in the long term following the relevant validation.

12.H.9 Culture media controls

Every batch of the culture media used must comply with the tests listed below, which can be carried out before or at the same time as the test of the product being investigated:

• Sterility: Samples of the culture media are incubated for 14 days according to their intended use, i.e. thioglycate medium is generally incubated at 30-35 °C and soya-bean casein digest medium at 20-25 °C. No microbial growth must be detected.

The European pharmacopoeia stipulates that every batch manufactured must be tested on ready-to-use medium. This requirement is not so high in the JP and USP. The two pharmacopoeias wish to point out that, in suitable cases, periodic testing of the various manufacturing batches manufactured from the same batch of dehydrated medium is acceptable.

Because there is always the risk that a medium is over-sterilised and thus that growth promotion properties are reduced, for example due to caramelisation of the sugars, this option must not be used under any circumstances for tests for Japan or the USA. This would be problematic for the USA at least, because the problem of over-sterilisation is expressly referred to in the FDA-Guide to Inspection of Microbiological Pharmaceutical Quality Control Laboratories.

In order to test the growth promotion properties, samples of the culture media are inoculated with a small quantity (10 to 100 CFU) of microorganisms and incubated. The microorganisms must create a visible turbidity within the specified time.

The following microorganisms are specified:

• Fluid thioglycolate medium: Clostridium sporogenes, Pseudomonas aeruginosa, Staphylococcus aureus
  FDA investigator also expect wild strains to be used (organisms that have been found as part of the monitoring in the units).
• Soya-bean casein digest medium: Aspergillus niger, Bacillus subtilis, Candida albicans
  Incubation is for up to three days for bacteria and five days for fungi. Reference strains are given in Fig. 12.H-12. The number of passages from the original culture is limited to 5, as before.

  Figure 12.H-12 Reference organisms for the sterility test for method validation and culture media controls

  Reference organisms for the sterility test for method validation and culture media controls
  Organism	Strains	Newly recorded
  Aerobic bacteria
  Staphylococcus aureus	ATCC 6538, CIP 4.83, NCTC 10788, NCIMB 9518	-
  Bacillus subtilis	ATCC 6633, CIP 52.62, NCIMB 8054	-
  Pseudomonas aeruginosa	ATCC 9027, NCIMB 8626, CIP 82.118	-
  Anaerobic bacteria
  Clostridium sporogenes	ATCC 19404, CIP 79.3, NCTC 532	ATCC 11437
  Fungi
  Candida albicans	ATCC 10231, IP 48.72	NCPF 3179, ATCC 2091, IP1180.79
  Aspergillus niger	ATCC 16404	IP 1431.83, IMI 149007

12.H.10 Method validation

All methods must be validated for the specific substance and/or drug product using all the microorganisms named in figure 12.H-13. The strains stated in figure 12.H-12 must be used. If use has been made of the previous option of using one type of aerobe, one type of anaerobe and one type of fungus, this will mean revalidation in all cases. FDA investigators sometimes expect validation with organisms that have been found in the unit

For the validation, the lowest number of organisms possible (maximum 100) are added to the last portion of the rinsing fluid for membrane filtration or to the drug product/culture medium mixture for the direct inoculation method and incubated for up to five days (previously three days for bacteria) at 30-35 °C (liquid thioglycolate medium) or 20-25 °C (soya-bean casein digest medium). After incubation, growth must be detectable comparable to the accompanying control containers. The method can then be viewed as valid. If organism inhibition is observed, this must be eliminated, for example by neutralisation, dilution, increase of the rinsing volume, etc. Especially for preparations with a high content of antimicrobial substances, it can be useful to test the growth promotion again following the incubation.

The question of when a validation must be carried out is clearly regulated:

• If the sterility test is carried out with a new product
• If there is a change in the experimental conditions of the test, which does not make clear when one must assume changes in the experimental implementation (this should not understood in too narrow a sense, in my opinion).

Another important question is how often and with which product batches this type of validation must be carried out. There is no direct information about this in the sterility test regulations in the pharmacopoeias. However, in the Validation of Microbial Recovery From Pharmacopeial Articles chapter of the USP, it states that three independent validations should be carried out. The FDA explains that the validation must be carried out with three different batches in order to rule out any influences of drug product formulation, etc. ("The three tests should also be performed on three different lots to demonstrate that the test is not affected by within specification variations of different lots of the formulation"). Although this only relates to the validation of microbial count determinations, it is a good idea to use it for the validation of sterility test batches.

There are always discussions about whether validations are transferrable to other active substance concentrations. The European Pharmacopoeia states that validations must be carried out if the sterility test is performed for a new product or if there is a change in the experimental conditions. This means that a validation must be carried out for every new drug product and in the event of changes in the experimental implementation of the test. The only conceivable exception to this is if a drug product with an identical composition has several different dosages of the active substance. In this case, the validation of the highest and lowest concentration is transferable to the dosages in between on the condition that the two validated methods are identical.