Cleaning Validation Queries

GOOD VALIDATION PRACTICE (cGVP) CLEANING VALIDATION
1

Dear All,

Can any one clear my Queries??

  1. How to adress if swab/ rinse sample fails in a piece of equipment in a train during cleaning vaildation batches. should we perform re cleaning??
2

Dear Mr Ravi,

Failure of any swab/rinse sample to comply with the acceptance criteria indicates failure of cleaning validation (i.e. the cleaning method is not effective enough to remove contaminant residues and hence cannot be validated). Although this is very straight-forward answer to your query, I would advice you to investigate the failure first before concluding anything. Try to find out the reason for failure: Is it because of sampling error? Is it because of personnel involved in sampling (e.g. if they are not trained)? Is it because of analytical method? Is it because of acceptance criteria (the criteria set is not correct)? Is it because of cleaning method itself? If by thorough investigation you conclude the failure is because of cleaning method, then you need to modify the cleaning method to make it more effective. Carry out the validation for the modified cleaning method. If the failure is not because of cleaning method, then include the finding in the final cleaning validation report (with documentation of deviations).

3

Dear ovais

upon investigation Analytica error was ruled out, and even sampling error was also ruled out, and finally it was concluded that re-cleaning shall be carried out for that part of that equipment & reswab shall be taken (Only Sieve), in this case whether we have to re-think upon cleaning procedure?? cleaning procedure ok with all parts of equipment and only failing in Sieve.

Can we justify in our report during conclusion??? if yes pls sugesst me how??

Regards,
RAvi.B

4

Dear bravi,

in my opinion is depends on how you set your acceptance criteria. The usual (and most conservative) way is to set the same maximum level for every sampling point. In that case your validation failed because one point is out of spec. The other way (which although I do not use, in my opinion is much more common sense) is to run a prevalidation study and set acceptance criteria for each sampling point taking into account the contamination distribution found in that study, provided that the overall absolute contamination of the whole piece of equipment is below the calculated target contamination. This way not only accounts for the non uniform distribution of contamination in the equipment, but allows for higher limits for discrete parts. Condition is that these parts are discrete (that means that they are sampled entirely or the contamination values of the sampled point can be extrapolated over the entire surface of the part) and that the contamination can be regarded as homogeneously distributed in the “contaminated” batch, which as a matter of fact is only true up to the last production step with mixing action. In your case, and again in my opinion, a factor that is against you is the fact that swabbing is not the best method for a sieve because the swab is unlikely to penetrate the mesh, and probably if you sample the same sieve by rinsing with an appropriate solvent you will get even higher contamination values.

Best regards

Alfred

5

Dear Mr Ravi,

Frankly speaking, I can’t advice you anything regarding “test until clean” OR “clean until pass” kind of approach for cleaning validation studies. This kind of approach is totally unacceptable from regulatory and compliance point of view. No matter it is just one part (in your case its sieve) which is failing (for one batch or all 3 the batches), the fact is “it has failed”. Probably we can justify the failure with some “reasons”, but what about subsequent cleanings (following cleaning validation study) where no samples are taken for residue analysis? How we can prove that this kind of failure is not going to occur in future and if it occurs, it will be detected? I have few suggestions for you:

[LIST=1]

  • [b]Make your cleaning method more robust (learn from mistakes)[/b]: Till now (based on CV results), you would have been able to identify the parts that pose the most difficulty in cleaning. I think you are using “manual cleaning” for your equipment, which I feel is highly variable and operator-dependent. You may control the manual cleaning by providing extensive training to operators and writing detailed SOPs with all the variables (i.e. operating conditions) covered and controlled. You may make cleaning procedure more robust by reducing errors caused by human factor, for achieving this, for example, you may limit the quantity of water/detergent used for cleaning, use hot water (if required, specify the temperature of washing & rinsing solvent), use spray-jet for cleaning surfaces (e.g. for cleaning sieve surface) specifying clearly the pressure and speed of jet, using brush for cleaning rough and uneven surfaces (include in SOP: for how long the operator need to rub the surface with brush), extent of equipment disassembly and instructing operator to visually inspect the surface during and after cleaning (repeating the procedure if “not clean”, reporting the changes). After modifying the method, repeat the cleaning validation cycle.
  • [b]Determine the acceptance criteria properly[/b]: If you don’t plan to modify the cleaning method (as suggested above) then one way to avoid the situation you are facing is to set the cleaning validation acceptance criteria properly. You may use the approach, as suggested by Alfred, and set different limits for different parts. Now, the question is how to do that? The simple way to understand it is with an example. Assuming that I have an equipment with a surface area of 1000 sq cm which can be divided into five parts A, B, C, D and E (with surface areas of 100, 200, 150, 300 and 250 sq cm respectively). I plan to collect one swab sample (covering 25 sq cm surface area) from each part. The MACO (in mg/sq cm) is obtained to be = 0.12 mg/sq cm. Now, if I multiply this value with 1000 sq cm, I would get 120 mg, this means that out of the total amount of Product A allowed to be carried over to Product B (whole batch of Product B), not more than 120 mg should come from the equipment concerned (i.e. the equipment being cleaned). Based on prior experience or pre-validation studies, I set limits for Parts A, B, C, D and E as 0.12, 0.05, 0.10, 0.12 and 0.18 mg/sq cm respectively. Multiplying the individual limits with corresponding surface areas and summing up the values would give a value of 118 mg (i.e. the equipment should not contribute, more than 118 mg to the total residue of Product A in Product B), which is lower than the target value of 120 mg. Probably, you can use this approach to justify CV results. However, the disadvantages of using this approach are that you have to determine the surface areas for each part, you have to calculate acceptance limit for each part (which I feel is more tedious to do than determining a single limit) and you have to provide justification for using this approach (may be by means of pre-validation studies). Please make sure that you calculate the surface area for sieve carefully, most of the times validation guys tend to consider sieve as a flat surface (and calculate the surface area of wire mesh using area of circle). Surface area for wire mesh should be determined taking into account the surface area of individual wires. Perhaps this is the reason, you got high residue value (surface area you considered to be = 25 sq cm, might have been more than 25 sq cm). I would like to ask you few questions regarding recovery studies; did you perform recovery studies for wire mesh? If yes, then how? Sieve is not only the part, which is the most difficult to clean but also a part from which it is difficult to recover residues. By no means you could spike accurate amount of residue onto the wire mesh. Another contributing factor to failure might be “[b]recovery factor[/b]”, the one you are using for samples collected from sieves.

    Thanks,

    Ovais

    • 6

    Dear ovais,

    we have not performed recovery studies for wire mesh.

    Regards,
    Ravi.B

    7

    Dear All,

    Can Any one clear my queries???

    1. Should we consider Exhibit batches for cleaning validation matrix??
      • If we consider the exhibit batch in cleaning validation matrix, then during this stage batch size will be always 1X, if it comes to commercial stage it may be up to 10X- My question is: for MACO calculation we have to consider smallest batch size of subsequent product (next product) then which one we have to consider the smallest batch. Should we again establish the acceptance criteria for worst case by taking other smallest batch size in the facility??
      • As per guidelines during exhibit batches cleaning verification shall be considered not a validation, then should I only consider commercial products for cleaning validation??

    2. After getting the swab results can we do the calculation as below, is this is acceptable?

    Highest PPM value observed x total Surface area
    100 x 100 x Recovery factor (in %)
    1.5 x 158986 = 27.41 PPM
    100x100 x 0.87

    1. Can we perform cleaning validation for each product in the facility?

    2. Should we perform swab for each time for exhibit product.
      For Example: after manufacturing of Product XYZ tab (exhibit batch) cleaning done and again I want to go to Commercial batch, should I wait for swab result to pass??

    Regards,
    Ravi.B

    8

    Ovais,

    I was under the assumption that you could spike sieve mesh using a Hamilton Syringe, allowing it to dry, and then doing a rinse recovery study…no?

    9

    Dear Mr Ravi,

    Here are answers to your queries:

    1. Should we consider Exhibit batches for cleaning validation matrix?

    If you are sharing the equipments for exhibit and commercial batches then my answer is “YES”. Please share with us which guidelines you are referring to . . . What does the said guideline mention regarding cleaning validation requirements under conditions when equipments are shared? Does it require the manufacturer to have a separate facility for manufacturing exhibit batches?
    Two situations may arise when you are sharing equipments for manufacturing exhibit and commercial batches, the exhibit batch may be considered either as Product A (product to be cleaned) or Product B (product where residues from Product A would end up). (i) If the product (for which exhibit batch is manufactured) is the most difficult to clean product (as determined after considering it in CV matrix) then CV studies should be performed using it as worst-case product (i.e. Product A). (ii) The product (with exhibit batch) may be considered as Product B, if and only if the exhibit batch is meant for human consumption (e.g. for BE study etc.) This means that if the product (with exhibit batch) is meant for human use then the batch size for the exhibit batch should be used for determination of MACO (otherwise you have to consider batch sizes of commercial batches only).

    1. After getting the swab results can we do the calculation as below, is this is acceptable?

    Simple formula to report the swab results is Actual Amount Present (per swab) = Amount detected (per swab) / Recovery Factor. Compare the individual results with MACO (per swab) and determine the acceptance of swab results. If you want to extrapolate amount of residue in total for a given equipment/equipment train then you can use this: Actual Amount Present = Highest Amount detected (among swab results) x Total shared Surface Area / Recovery Factor.

    1. Can we perform cleaning validation for each product in the facility?

    Yes, you can. But why do you want to do that if you can achieve the same objective by using a worst-case approach. Assume how much resources it is going to take if we had to validate each and every product.

    1. Should we perform swab for each time for exhibit product.

    Please find the answer above (Answer #1). If your exhibit product is the “most difficult to clean” product then you have to perform CV using it as worst-case product otherwise you need not to take swabs sample each and every time you manufacture exhibit product

    Regards,

    Ovais

    10

    which is more actuall in determine the residue in cleaning validation
    maco= TDD perMBS /SF TDDnext
    or
    analysed sample = 0.001* min dose pervious product AMBS SWAB AREA/SSAmax dose prod B AMONUT OF SOLVET DESORTED
    THANKS

    11

    Dear all,
    i want to start cleaning validation for a single product eg. Amlodipin 5mg tablet (uncoated) by direct compression method.
    So what are the procedure i need to follow from the initial step?

    wating reply of u all
    SUNITA