Compendial Methods Need To Be Qualified
September 12th, 2011 by Peter CalcottDuring a recent class I taught at University of California, Berkeley in their Quality and Compliance postgraduate certificate program, a question came up from one of the attendees that mirrored very closely observations I have made in audits over the years. And that question was “do we have to validate compendial methods we use in testing our commercial products?”
That question is, on the surface, very simple. And the answer is no. Neither the FDA nor the EMA requires us to validate compendial tests that we use with our products. But the answer actually has another layer to it which people do not venture too often to address.
While we do not have to validate a compendial method we use with our products, we must demonstrate that it works for the application with our product. ICH Q7 describes in section 12.80, this point clearly.
Analytical methods should be validated unless the method employed is included in the relevant pharmacopoeia or other recognised standard reference. The suitability of all testing methods used should nonetheless be verified under actual conditions of use and documented
Thus it is clear that while formal validation is not required, we must demonstrate that it works for our product for the intended purpose. This is very clearly shown in the original form 483 issued to Chiron Corporation for their flu vaccine product plant in UK in October 2004 that resulted in issuance of a warning letter in December 2004. They were cited for a lack of data to support the use of the USP sterility test for their product. Their response was clearly adequate since it was not cited in the warning letter. This action took them off the market in the 2004 flu season for their flu vaccine product
Many times on audit I have reviewed specifications and seen that compendial methods are used for a variety of tests including sterility, endotoxin, particulates and others. My first question is usually to ask for the data to support use of the method for their product. At least 9 out of 10 times, I get blank stares followed by the statement, “we don’t have to validate compendial methods!” I agree with them but follow on with the statement, “I know, but where is the data to support usage of the method?” They often look puzzled still but cannot provide any data. Too often, I believe, the first sentence of the ICH Q7 document is read but the second missed.
What do you have to do the “qualify” it? I recommend a simple qualification plan – rather like a validation protocol – with a testing plan to check the key parameters of the assay such as accuracy, precision, specificity and ruggedness, including interferences. That sounds like a validation and it is: but an abbreviated one touching only on the parameters of issues that make your product unique or challenging.
Good Afternoon Peter,
Recently at our biotechnology plant we started working with Bioburden methods qualification and have been having issues trying to filter cell culture samples specially Harvest Samples, method proposed was membrane filtration but any feedbacks/recommendations that will help us identify a better method?,
Thanks,
I.Rodriguez
You posed an interest challenge. Firstly, I am assuming that you are dealing with a mammalian cell culture sample. This conditioned medium sample does pose challenges to filtration. There are many ways to approach this including:
1.) Try to pre-filter the sample through a 10 micron filter, followed by 0.2 micron filtration to trap bacteria which are then placed with the filter on top of a nutrient medium eg. nutrient agar (NA). To qualify this you will have to prove that the pre-filter does not remove bacteria. You are probably safe with this.
2.) Use a detergent to solubilise the debris and lyse the mammalian cells. Pre-filter and then 0.2 micron filter to get count. Again you will have to assure that the sample preparation and filtering do not interfere with bacterial recovery.
Both of these are really using the basis for the sterility test to count bacteria. But there is an alternative approach. The sensitivity is less but remember that bacteria once in a cell culture do grow very fast and much more so than mammalian cells.
You can actually use a direct plating technique. Take a sample of culture fluid – say 0.2mL and spread plate it on NA. The mammalian cells will not grow – at least within the 2-5 days you need to incubate the sample to get bacterial colonies. That translates to being able to deterct <5 bugs/mL.
Using the agar pour method, take 2.5 mL of culture fluid and dilute with same volume of double strength NA to yield 5mL of normal NA. Pour onto a shallow plate of NA. In this case you can improve sensitivity about 1 log to 0.4 bugs/mL
These methods have been used in plants I have worked at and accepted by agency people.
In both cases you need to qualify recovery of bugs especially EM samples that might be contaminants as well as a few more. Write out the method and qualification and present to agency. No requirement I am aware of says you must use filtration method with its sensitivity. After all you are looking for gross contamination and we know if a bug gets in it will grow like wild fire.
Good luck and hope this has been helpful
We are working on glucagon bioidentity test. This is a recent revision of glucagon bioassay method, USP 35.
We are planing for verification of the test method USP glucagon bioidentity test. As this method is a indirect method to measure the relative potency of glucagon and the experiment having two measure part one is hepatocytes preparation and second is analysis of samples for glucose estimation by glucose analyzer.
To verify the method suitability in our laboratory condition first we validate the instrument i.e. accuracy, precision, specificity,linearity and range.
The acceptance criteria for hepatocytes preparation is already mentioned in the USP.
Is this required to verify the acceptance limits in verification experiments?
Which factor should be considered for verification of bioassay method.
Anticipating your reply
Regards and thanks
Durgesh
In general, this is how I would do it.
1.) Equiment validation. DQ, IQ OQ and appropriate PQ for the instrument – all components. This includes the software to calculate any data.
2.) Now to method. Examine the method as per USP and determine if it is appropriate for your product through research – ie no GMP conditions
3.) When you are con]mfortable that the mteethod works for you, either as is or with some adjustments, you are ready for qualification
4.) Consider what parameters are importnatn to confirm and incorporate into a protocol. It need not be as detailed as a full validation. It might include accuracy, precision, specificity and ruggedness but it does not have to include all. perform work and see how it turns out. Write up and Voila, there it is.
Hope this helped.
I have found people tend to either go overboard or ignore any qualification. Just define what you will do and why in the protocol.
Hello! I could have sworn I’ve been to this website before but
after looking at some of the articles I realized it’s new to me.
Anyhow, I’m definitely happy I stumbled upon it and I’ll be book-marking it and checking
back frequently!
Peter, I have a question regarding compendial tests for monitoring incoming raw materials. My question is “how can I know which method needs verification and which doesn’t”? An example is sodium chloride testing. If I follow the USP monograph for NaCl, there will be approximately 14 tests to perform to accept the material. Does each test require method verification? I understand that method verification is required when you are not certain how sample matrix affects the assay performance, but in this case, all test materials are pure as they are either USP or NF grade. Can you provide guidance on how to determine if a compendial method needs verification when it comes to raw material incoming testing?
Thanks a lot.
Scott
The main purpose of this post was to deal with using methods that are taken directly from Pharmacopoeias and used with products not listed in these reference documents. However, these methods are used for a number of situations. I will take three different cases that will address your question
1.) Using a pharmacopoeial method directly for testing (say Chloride ion in NaCl). Clearly, there is a monograph for NaCl that references the method directly. So if I want to qualify my source of USP NaCl, then I use the Cl method referenced. No verification needed. That goes for my active if it is described in a monograph and it references the exact Cl method. No verification or qualification needed.
2.) I have a new raw material that happens to be in a chloride form of the compound. That compound is not a subject of a monograph, so I look up Chloride method in the USP and decide to use that method. Before I can use it to measure Cl I need to verify or qualify that it works in my situation. Then a verification and qualification is needed. Obviously in the case of Cl I am sure you are pretty sure it will work.
3.) If I have a brand new active that is not in the USP, then if I want to measure Cl in it, I do the same as for example 2. I would qualify the method.
Obviously for Chloride, the verification will be very simple and not complex.
However when we get to other methods, the complexity can be more. Adjusting or qualifying the USP sterility test for a bio-active compound, I would have to prove that the method did not give false negatives or positives in the way it is run with my compounds. If I did see interference, I might have to modify the method to remove the interference. It is gathering the data to prove it works or the modification of the method that constitutes the verification or qualification. This is a classic case of a risk based approach that is used. f the burden of proof is small then less work. If more then more work is involved. The key to success is to think logically and document what you know and your assumptions and then perform the key experiments to assure the method is giving valid results.
As they say QED – in latin that is quid est demonstrandum or in English quite easily done.
Dear Peter, I have another question relative to appoaching compendial methods. After having sucessfully demonstrated product-specific method suitability for bioburden and sterility testing at one testing site, the methods shall be transferred (1:1) to an alternative additional testing site at a later point in time. This alternative site shall function as an alternative testing site for release testing of finished product in parallel.
What is the data requirement for such a transfer to be shown?
Thank you very mch for your advice.
Re, Dietmar
The answer is yes. If you develop the method and qualify it on site 1. You need to transfer to show it works on site Two. Since its a Compendial the transfer can be very simply done. Does not have to be like a secondary validation done for regular methods. But yes documentation is appropriate and required.