Regulatory Open Forum

 View Only
Reg Intel Conference 2024
Back to discussions
Expand all | Collapse all

Process validation

  • 1.  Process validation

    Posted 13-Mar-2013 15:48
    This message has been cross posted to the following Discussions: Global Regulatory and Products News (GRPN) and Regulatory Open Forum .
    -------------------------------------------

    Hi, 

    I have one query, if anyone can help. If we have perform process validation of Finish product with API from X vendor now after adding API vendor Y do we need to perform again process validation of FP batches Mfg with this Y API, if there is no difference in API quality or impact on FP or process etc. 
    -------------------------------------------
    Ashish 
    Senior Regulatory Executive
    United Kingdom
    -------------------------------------------


  • 2.  RE:Process validation

    Posted 14-Mar-2013 04:14
    Dear Ashish

    The answer is, yes you must. Validation is not transferable, nor is there evidence that the APIs are indeed identical. Just because they both meet specifications does not mean they are the same.
    It is very likely that you will see differences in either process behaviour, analytical results (e.g. for stability samples) and yields.

    Regards
    Siegfried

    -------------------------------------------
    Siegfried Schmitt
    Principal Consultant
    PAREXEL
    Braintree, Essex
    United Kingdom
    -------------------------------------------






    Original Message


  • 3.  RE:Process validation

    Posted 14-Mar-2013 04:51
    Hi Siegfried,

    I do not completely agree. It is possible to assess (maybe indirectly through its DMF) the manufacture and quality of the proposed API supplier and compare it with the current API supplier. I would say re-validation is not necessary, and usually concurrent long-term stability data and dissolution comparison should suffice. Again, this is from my experience with simple immediate release dosage form, i.e. tablets.

    Regards,

    -------------------------------------------
    Tom Namsavanh RAC
    Manager, Regulatory Affairs and Compliance
    Intelgenx Corp.
    Saint-Laurent QC
    Canada
    -------------------------------------------






    Original Message


  • 4.  RE:Process validation

    Posted 15-Mar-2013 09:35
    I agree with Tom.  Even for a biological, we would not revalidate the DP process due to a change in the DS/API.  We would, however, do studies to make sure that the new API would perform the same in the validated process.  We would also generally provide release data from at least one lot of DP made with the new API and put the lot on stability.

    However, it does depend on the specific type of product.  If there is an aspect of the API/DS that could impact the DP process, then the answer might be different.  But you already said that the API does not impact the DP process.

    -------------------------------------------
    Rachel Thornton
    Associate Director
    UCB
    Smyrna GA
    United States
    -------------------------------------------






    Original Message


  • 5.  RE:Process validation

    Posted 18-Mar-2013 04:24
    Dear Rachel and Tom

    I don't think we disagree, we merely use different language to express the same, i.e. have a different definition of validation To me, validation is the process by which you prove a process works given the various inputs (including API) and controls.
    If you look at the FDA's GfI on validation you will find that your activities fall well within the context of validation.

    Not performing validation would mean to merely use any API (as long as it meets the specifications) in your process without any additional studies (e.g. engineering runs, small scale synthesis, dissolution studies). There would also be no need for stability studies. Running stability studies simply states that you do not believe the DPs to be identical.

    Rachel, I find it hard to believe an agency permits change of an API for a biological without re-validation of the DP. I'd be most interested to hear which agencies allowed this approach.
    I presume you therefore do not seek prior approval from the agencies, but merely state the use of a different API in your annual report?

    Thanks for


    Original Message


  • 6.  RE:Process validation

    Posted 19-Mar-2013 10:26
    Siegfried-

    I think the situations are not completely equivalent between a chemical entity and a biological.  First of all, I am not aware of any biological where the Drug Substance can just be purchased from a supplier with a DMF and a CofA.  However, taking the example of using a new CMO to manufacture the DS, you would have to validate the DS process (meaning the usual 3-lot demonstration of the process in the commercial facility) at the CMO.  But all the validation (in that sense) would be on the DS process.  You would then generally need to provide batch release data on at least one lot of DP made with that DS, but you would generally not need 3 lots and you would not need to provide all the additional in-process data that is usually generated from a full-blown "validation" run.

    It would be a PAS, but the focus is on the DS, not the DP.

    There may be exceptions, but keep in mind that for most biologicals, the DP process is not very complicated.  The DS is a solution, usually very similar if not identical to the DP formulation.  If the DS passes validation, it is unlikely to impact the DP process, and the characteristics of the DS (release properties & stability) are predictive of the DP.

    I hope that is clear.

    -------------------------------------------
    Rachel Thornton
    Associate Director
    UCB
    Smyrna GA
    United States
    -------------------------------------------






    Original Message


  • 7.  RE:Process validation

    Posted 20-Mar-2013 11:43
    Hi Siegfried,

    I do agree that we do not use the same language to express ourselves. For me, running stability studies does not imply I do "not" believe the drug product to be identical but on the contrary that the drug product fabricated with the API from supplier A is identical to the drug product fabricated with the API from supplier B. Identical is viewed in term of product performance.

    As Rachel said, the API supplier will have already validated the manufacturing process for the drug substance (hence the availability of a DMF - I have yet to see a DMF where validation sections are omitted).

    At the end of the day, what regulatory agencies want is that the change from API suppliers does not affect the quality, and consequently safety and efficacy of the drug product. And I think re-running validation on the drug product manufacturing process is a bit overkill as there are more accessible and accepted alternatives in order to reassure the agencies.

    Regards,

    -------------------------------------------
    Tom Namsavanh RAC
    Manager, Regulatory Affairs and Compliance
    Intelgenx Corp.
    Saint-Laurent QC
    Canada
    -------------------------------------------






    Original Message


  • 8.  RE:Process validation

    Posted 21-Mar-2013 05:59
    Hi Tom and Rachel

    I really this discussion very interesting as it shows so well how we all have some preconceptions and simply assume we all mean the same when we discuss things here.
    For example, I didn't assume that a DMF exists, as often it doesn't. Now this may be very different to your operations, where a DMF may be standard practice. Of course, where you got a DMF and maybe even a supplier inspected by FDA or EU, then there should be little need to worry. But where do we stand if this is not the case?
    Another aspect to bear in mind is that even if you do produce exploratory batches, your analytical methodology has to be up to the task (of identifying any potential changes). 

    An ex-FDA inspector commented this week that in his thoughts the DP process for DS manufactured at a second site may not require formal PV, but would require as a minimum monitoring and evaluation of the first few batches.
    Now this monitoring is precisely what is described in FDA's process validation guidance in section D. Stage 3 - Continued Process Verification: The goal of the third validation stage is continual assurance that the process remains in a state of control (the validated state) during commercial manufacture.
    So we are back to validation.

    Really, I think it is not important if you call it validation or not, what matters is that the approach is compliant with the regulations and that one doesn't merely use expressions and terminology that is assumedly widely understood, but that one clearly defines in one's quality system the concept as applied within one's company.

    As an aside Rachel, you mention 3 lots/batches to represent process validation. At the moment the EU will simply accept 3 batches, the FDA however requires a rationale for any number selected.

    Your thoughts?
    Siegfried

    -------------------------------------------
    Original Message:
    Sent: 03-20-2013 11:43
    From: Tom Namsavanh
    Subject: Process validation

    Hi Siegfried,

    I do agree that we do not use the same language to express ourselves. For me, running stability studies does not imply I do "not" believe the drug product to be identical but on the contrary that the drug product fabricated with the API from supplier A is identical to the drug product fabricated with the API from supplier B. Identical is viewed in term of product performance.

    As Rachel said, the API supplier will have already validated the manufacturing process for the drug substance (hence the availability of a DMF - I have yet to see a DMF where validation sections are omitted).

    At the end of the day, what regulatory agencies want is that the change from API suppliers does not affect the quality, and consequently safety and efficacy of the drug product. And I think re-running validation on the drug product manufacturing process is a bit overkill as there are more accessible and accepted alternatives in order to reassure the agencies.

    Regards,

    -------------------------------------------
    Tom Namsavanh RAC
    Manager, Regulatory Affairs and Compliance
    Intelgenx Corp.
    Saint-Laurent QC
    Canada
    -------------------------------------------






    Original Message


  • 9.  RE:Process validation

    Posted 22-Mar-2013 10:37
    Siegfried-

    Re. your comment/question

    As an aside Rachel, you mention 3 lots/batches to represent process validation. At the moment the EU will simply accept 3 batches, the FDA however requires a rationale for any number selected.


    Yes, the FDA guidance says that the number needs to be justified, but my experience has been that industry still goes back to the standard 3 batches.  As far as I know, there has been no further guidance from FDA on what they think is reasonable, so in the absence of further guidance, companies tend to stick with the traditional approach, especially for new applications and new facilities.  And when that approach is used, I've never seen FDA ask for specific justification of the number. 

    I think we're all still feeling our way with how the PV guidance will work in practice, both in the industry and at FDA.  And for now, the old terminology persists, though at a previous company I was involved in a project where they made an effort to call those first full-scale batches "process qualification" batches instead of "process validation" batches.  I agree that we all need to be very clear and careful with our terminology to avoid confusion.
    -------------------------------------------
    Rachel Thornton
    Associate Director
    UCB
    Smyrna GA
    United States
    -------------------------------------------






    Original Message


  • 10.  RE:Process validation

    Posted 23-Mar-2013 01:52
    Thank you Rachel for this observation. The difference between theory and praxis...
    Siegfried


    Original Message


  • 11.  RE:Process validation

    Posted 20-Mar-2013 11:51
    Rachel,

    I can assure you that what you described is very familiar to me as well, i.e. it is similar for chemical entity.

    Regards,

    -------------------------------------------
    Tom Namsavanh RAC
    Manager, Regulatory Affairs and Compliance
    Intelgenx Corp.
    Saint-Laurent QC
    Canada
    -------------------------------------------






    Original Message


  • 12.  RE:Process validation

    Posted 21-Mar-2013 08:58
    Hi,

    In my view, understand the manufacturing process differences of the API provided by supplier A and supplier B will gain additional information and confidence of levels of validation required. Sometimes, synthesis route, solvent usage etc. could create differences which may not pick up by API specification, but enough to impact the finished product. API validation would not help different suppliers using different manufacturing processes. If both API suppliers use very similar manufacturing process for the API, Finish product process validation may be reduced.


    -------------------------------------------
    Jian Chen RAC
    RA Manager
    Parsippany, NJ
    United States
    -------------------------------------------






    Original Message


  • 13.  RE:Process validation

    Posted 15-Mar-2013 10:18
    Dear Tom

    Admittedly, there is no hard and fast statement in the regulations that mandates revalidation.

    In Europe, this is indirectly implied thorugh the variations regulations (classification guideline), which requires: batch analysis data (in a comparative tabular format) for at least two batches (minimum pilot scale) of the active substance from the current and proposed manufacturers/sites. This equates to validation batches.

    In the US the process validation guidance does not contain information as specific as a change in the API, nor does Part 211 specifically require validation for a change in the API. However, CFR 314.70 - NDA supplements - does require a supplement be filed for a change in the "new drug substance". It is clear that there is no explicit requirement to validate when changing an API source. 

    The Federal Food, Drug and Cosmetic Act section 506A reads as follows:
    (b) VALIDATION OF EFFECTS OF CHANGES.-For purposes of subsection (a)(1), a drug made with a manufacturing change (whether a major manufacturing change or otherwise) may be distributed only if, before distribution of the drug as so made, the holder involved validates the effects of the change on the identity, strength, quality, purity, and potency of the drug as the identity, strength, quality, purity, and potency may relate to the safety or effectiveness of the drug. 

    Having said that, validation is certainly a common industry practice, makes good science and good common sense. It is conceivable that sufficient lab experimentation, testing, supplier qualification, etc., could be conducted and support a decision not to validate. However, validation seems like a more sensible approach.

    It is often where you don't expect to see a difference that you find unexpected issues.
    Regards
    Siegfried


    Original Message


  • 14.  RE:Process validation

    Posted 14-Mar-2013 09:30
    Ashish,

    Are you asking about medical devices?  If so, the notification requirements depend on the clearance path of the original device.  If this is a Class III device that was cleared by way of a PMA or HDE the FDA must be notified by way of a PMA supplement.  If the product is Class II the requirements depend on whether FDA has put any restrictions on the product when it was cleared that would require advance notification of a vendor change.

    Either way the second vendor will be expected to comply with 21 CFR 820 regulations and your company is responsible for having adequate control of the manufacturing process.

    Regards,

    John

    -------------------------------------------
    John Minier
    Small Bone Innovations, Inc
    Morrisville PA
    United States
    -------------------------------------------






    Original Message