The big advantage of the dose criteria is that the therapeutic dose of the drug substance is considered in the limit calculation. The different pharmacological properties of different drug substances can therefore be accommodated to an extent.
The disadvantage of this procedure is that there is a large number of different subsequent products for the worst case product in practice, and strictly speaking, an individual calculation should be made for each possible combination of previous product/subsequent product. The limit calculation is costly and complex as a result.
However, it is not acceptable to calculate the limit while the validation is carried out using the respective subsequent product, because according to general understanding of the validation, the acceptance criteria (and therefore the resulting limits) have to be defined in advance, in other words, the limits must already be specified in the master plan, or at the latest, in the validation protocol.
For the analytical method validation which has to be carried out prior to the cleaning validation, knowing the acceptable limits for the critical active pharmaceutical ingredients is also essential, because all tests for reproducibility, linearity and recovery should be within the range of quantities expected to be in the samples respectively in the range of the maximum acceptable residue.
As a possible solution, a fictitious "worst case" subsequent product could be used as the basis for calculation. So, based on the product range of the respective company, the smallest possible batch size, the highest intake frequency and the largest mass of an individual dosage form are to be used in the formula - regardless of whether these values are assigned to an individual product or several products.
If the limit for this procedure falls so low that problems regarding the practicability (effectiveness of the cleaning process) and/or the verifiability (limit of quantitation of the analytical procedure) can be expected, the following solution can be chosen, for example:
The limit is calculated for two real "worst case" subsequent products, namely the product with the smallest batch size and the product with the largest daily dose. The smaller of the two resulting limits is used for the cleaning validation.
A further disadvantage of the dose criteria is that the calculated limits are proportional to the therapeutic dose, in other words, active pharmaceutical ingredients with a low therapeutic dose result in low limits, for active pharmaceutical ingredients with a high therapeutic dose, the limits are high. These can sometimes be so high that the residues are visible with the naked eye and the visual criteria is no longer fulfilled. This may be acceptable from a pharmacological point of view but it contradicts the general GMP requirements.
The maximum acceptable residue in the subsequent product is calculated using the ppm criteria in accordance with the formula specified in figure 8.E-2
The ppm criterion has the advantage that it can be used regardless of the previous product/subsequent product combination, which makes the limit calculation within a complex validation project considerably easier.
The worst case for the limit calculation can be considered by using the smallest possible batch size for the subsequent product as the basis for calculation.
On the other hand, this method of limit calculation has the disadvantage that it is substance-independent, in other words, the therapeutic dose of the active pharmaceutical ingredient is not included in the limit calculation. This means that the various pharmacological properties of the different active pharmaceutical ingredients are not taken into consideration.
For highly potent drug substances, the limit calculated according to the ppm criteria may therefore not be acceptable from a pharmacological point of view.
Selection of the appropriate acceptance criteria
With view to these considerations, the recommendation from the PIC/S guideline to use the most stringent criterion to determine the limits is comprehensible.
In the planning phase of the cleaning validation, it is therefore necessary after identifying the worst case products to calculate the acceptance limits using the dose as well as the ppm criteria, and, based on this calculation, to define the corresponding acceptance criteria for each critical active pharmaceutical ingredient. Based on this comparative calculation the above mentioned rationale can be established. The following example (see figure 8.E-3) shows the limit calculation for the critical products A, E and F.
In this example, it is assumed that the smallest batch size manufactured in the company is 100 kg. According to the ppm criteria, the maximum acceptable residue in the subsequent product is 1,000 mg for all three critical products.
For calculation in accordance with the dose criteria, the therapeutic dose of the critical product and the daily dose of the subsequent product are still required. The latter is calculated from the intake frequency and the mass of the individual dosage form. Following the solution approach described above, a fictitious subsequent product is used as the basis for calculation, whose batch size is 100 kg (100,000 g) and whose daily dose is 5 x 1 g. The resulting limits in the subsequent product lie between 1,500 and 10,000 mg.
A comparison of the limits shows that the dose criteria must be used for product E, whereas the ppm criteria delivers the stricter limit for products A and F.
Calculation of the acceptable residue in a sample
Knowing the acceptable residue in the subsequent product does not actually help during evaluation of the results of individual samples.
In principle, there are two solution approaches here again:
The use of the method described under a) requires that for each test run, the results from all of the production equipment are available at the same time. This method is therefore more suitable for the product-related execution of the cleaning validation. For the equipment-related execution, organisational measures must be taken to ensure that the cleaning validation is carried out for all production equipment used after manufacture of the critical product. Also, each validation report must refer to the results from the remaining production equipment, because the total of all residues is used to assess the validation.
A further difficulty of the method described under a) is in the extrapolation, in which the residue status of the entire product contact surface is derived from the results of the sampled surfaces. Because sampling usually takes place at the critical parts of the equipment, it can be assumed that the highest residue levels are measured there. If this residue concentration per surface unit would be extrapolated onto the entire product contact surface, a relatively high residue quantity would result which does not reflect to the actual circumstances and may even exceed the acceptable limit. On the other hand, it would be a crude simplification to set to zero the residue concentration per surface unit for the non-sampled equipment surfaces which are graded non-critical. The difficulty lies in defining a realistic residue concentration in the extrapolation for the non-sampled - and surface-wise much larger - equipment surface.
In contrast, the method specified under b) can be used without problems both for product-related and for equipment-related execution. When calculating a surface-related limit, it can firstly be assumed that there is a uniform distribution of the residue on the entire surface of the equipment. Sampling is carried out at the critical areas of the production equipment, as this is where the highest residue values are expected. If the sample results of the critical areas lie below the calculated limit, then it can be assumed that the remaining product contact surface has an even lower concentration of residue. If the test results from all critical areas lie below the surface-related limit, this guarantees that the limit in the subsequent product is also met.
The way in which the surface-related limit is calculated is demonstrated in the following example (see figure 8.E-4 and figure 8.E-5).
Firstly, the sum of the product contact surfaces is determined for each product. To do this, all production equipment used during manufacture is to be considered in its actual sequence, even if individual types of equipment - e.g. containers - are used repeatedly.
This already shows that the product contact surfaces can vary considerably according to dosage form and manufacturing process.
To calculate the acceptable residue quantity in an individual sample, the acceptable residue in the subsequent product is divided by the product contact surface and multiplied by the sample surface.
For products A and F, whose acceptable residue in the subsequent product lies at 1,000 mg according to the ppm criteria, very different limits for the swab test are the result, due to the product contact surfaces varying considerably in size. Product E has the smallest limit in the swab test and according to the dose criteria, only 200 mg may be transferred from this into the subsequent product.
Visual criteria (visually clean)
Both the FDA inspection guideline for cleaning validation as well as the PIC/S guideline PIC/S PI006 name the visual criterion as acceptance criterion for assessing the cleaning result. Whereas the FDA guideline simply specifies ppm criterion, dose criterion and visual criterion as alternative possibilities, the PIC/S guideline requires that the acceptable residue is defined following the most stringent of the named criteria.
In conjunction with the visual criterion, the PIC/S guideline also requires that spiking tests are used to determine the concentration in which the majority of the active pharmaceutical ingredients are visible. Fourman and Mullen specified the visual limit of detection for most active pharmaceutical ingredients with 100 mg/25 cm2 in an article published in 1993.
According to this, the visual criteria would always represent the strictest acceptance criterion if the limit in the subsequent product calculated according to the ppm criteria or dose criteria leads to a surface concentration of more than 100 mg/25 cm2 as acceptable residue in the sample.
The fact that production equipment should be free from visible residues after cleaning, is - independent of the cleaning validation - a basic GMP requirement. This assumes that the positive visual inspection of cleaned equipment in conjunction with the cleaning validation is seen as a prerequisite for further tests such as a swab test or rinse test.
What should be the procedure for the cleaning validation, however, if due to previous calculations, the visual criteria had been determined as the strictest criteria? The question of whether further analytical tests such as swab or rinse tests can be omitted in this case, results in controversial discussions.
Following the requirements of the PIC/S guideline, for the active pharmaceutical ingredient to be detected, the visible concentration of residue on the surface has to be documented with spiking tests. For this purpose dilution series of the affected active pharmaceutical ingredient are made in a solvent as volatile as possible and defined volumes applied on a marked test surface which corresponds to the sample surfaces (e.g. 25 cm2). To guarantee even distribution of the applied active pharmaceutical ingredient quantities on the respective test surfaces, the application and drying of solutions with different concentrations should be carried out under the same conditions as far as possible, in other words, the applied volumes should be approximately the same. An implementation example is shown in figure 8.E-6.
Once the solvents are fully evaporated, the visual limit of detection is to be determined by comparing the individual test areas. However, the problem adhering to this procedure becomes visible here. The determination of the visual limit of detection is essentially dependent on the following factors:
In order to determine the visual limit of detection with high reliability, the test conditions should match as far as possible the conditions later to be met during the practical execution of the cleaning validation. This can be achieved in the following way, for example:
Whereas these test conditions can be adapted easily to real conditions, the manner of active pharmaceutical ingredient distribution, however, is a permanent weak point of the visual detection method. Whereas in trials, thanks to the suitable choice of solvent and the volume applied, an even distribution of residue on the entire test area can be achieved, in practice - i.e. after cleaning - the kind of residue distribution is determined by chance. There ist no doubt however that it makes a difference for visual detection of residues, if e.g. 50 mg of a substance are homogeneously distributed on 25 cm2 (invisible), or if they are cumulated as particles with a size of a few mm (visible).
The question of whether analytical tests for determining the residue can be omitted under reference to an experimentally determined visual limit of detection, cannot be suitably answered in general. The decisive factor is certainly how far apart the surface-related limit calculated according to the ppm or dose criteria and the visual limit of detection lie. If the visual limit of detection was determined for example at 100 mg/25 cm2, further analytical tests are required for a calculated limit of 150 mg/25 cm2 rather than for a calculated limit of 500 mg/25 cm2.
If the cleaning validation is only to be carried out based on the visual criteria, a thorough documentation of the experimental determination of the visual limit of detection is necessary. To support the argumentation for an exclusive use of the visual criteria, it is also recommended to carry out a few examples of swab or rinse tests on visually clean surfaces and to determine the actual available residue analytically.
8.E.2 Calculation of cleansing agent residues
"The efficiency of cleaning procedures for the removal of detergent residues should be evaluated. Acceptable limits should be defined for levels of detergent after cleaning. Ideally, there should be no residues detected. The possibility of detergent breakdown should be considered when validating cleaning procedures." (PIC/S PI 006, section 7.9.1)
Similarly to limit calculation for active pharmaceutical ingredient residues, it is also possible here to use either a purely quantity-related criteria (ppm criteria), or to include the various properties of the cleansing agents used in the limit calculation.
In the same way as the therapeutic dose is used for the characterisation of the pharmacological properties of active pharmaceutical ingredients, the LD50 can be used for the characterisation of the toxicological properties of cleansing agents.
The acceptance criteria derived from this can be formulated as follows:
"The largest daily dose of the subsequent product may contain as a maximum the acceptable daily intake quantity of the cleansing agent used".
The calculation is made according to the formula in figure 8.E-7.
The LD50 specifies the dose of a substance (in mg/kg body weight) that causes death for 50 % of the laboratory animals after oral intake. The transferability of toxicity data determined on laboratory animals to humans was the purpose of investigations published by Layton in 1987. He realised that the 5 x 10-4-times LD50 no longer has measurable pharmacological effects on humans and deduced the so-called NOEL value (no observed effect level), which refers to the average body weight of an adult (5 x 10-4 x LD50 [mg/kg] x 70 [kg]). To determine the acceptable daily intake (ADI value), the NOEL value must be divided by a safety factor, taking into consideration the type of application of the administered substance.
The following guideline values apply to the safety factor:
Cleansing agents which are approved for the pharmaceutical and food-processing industries generally feature relatively high LD50 values, in other words, these substances are rather uncritical from a pharmacological point of view. Due to the high numerical values for the LD50, the acceptable residue quantities calculated in accordance with this formula are rather high.
Similarly to the limit calculation for active pharmaceutical ingredient residues, a comparative calculation can be made following the ppm criteria and the LD50 criteria, where the stricter limit is used for cleaning validation (see chapter 8.E.1 Calculation of active pharmaceutical ingredient residues).
It is also possible to include the capacity of the cleaning procedure as well as the toxicity of the cleansing agents in the determination of the limit. Preliminary tests can reveal the actual amount of residues after cleaning. A pragmatic approach of establishing limits will then lie between what is feasible in terms of cleaning and what is acceptable in terms of toxicity.
8.E.3 Determination of the microbial status
Checking the microbial status of product contact surfaces can be carried out as part of the cleaning validation, as long as this is not already part of hygiene monitoring (see chapter 11.E Environmental monitoring).
For the determination of limits, the "Supplementary and reviewed guidelines for the manufacture of sterile medicinal products" in the appendix of the EU GMP Guideline can be used. Limits (CFU/plate) for cleanroom classes A to D are specified there, but details are missing about whether they concern personnel or surfaces. Because limits for the microbial status of the hand are also specified, it can be assumed that the requirements refer to the surfaces.
In a recently published overview article, Seyfarth derived proposals from this for implementing the requirements of the "Supplementary guideline for the manufacture of sterile medicinal products" for the microbial status of surfaces. Alert and action limits accepted by the FDA were also taken into consideration in this.
The values given in this guideline were graded as guideline values or recommended limits which should be reached on average for a larger number of measurements. This only concerns guideline values for sterile production (clean room grades A-D). Data about the microbiological status of surfaces in areas with requirements E and F (GMP area) are not included in the EU Guideline.
The values proposed by Seyfarth are presented in the table figure 8.E-8.
The guideline value of 50 CFU/plate specified for product contact surfaces in areas of clean room grade D can also be used for product contact surfaces in the GMP area. According to experience, it is not a problem to keep this value for freshly-cleaned equipment surfaces presuming correct and hygienic execution of cleaning and drying.