Subscribe
  
IPQ Inside the global regulatory dialogue

BLA CMC Problem Areas Include Lack of Transparency in Supplement Cover Letters, FDA Reports

FDA is urging sponsors filing Biologic License Application (BLA) supplements to clearly highlight the significant changes and their scope in the cover letter to expedite the agency review and approval process.

The advice reflects the agency’s experience that supplement filers are not always giving it a clear picture up-front, resulting in significant review delays or a Complete Response (CR) letter indicating that the application is not approvable in its current form.

In turn, providing too much information on aspects of existing facilities that have not changed – for example, through hyperlinks to prior submissions – can also slow down the process as the reviewer must take time determining what has changed and what has not.

Speaking for the Center for Biologics Evaluation and Research (CBER) Division of Manufacturing Product Quality (DMPQ) at the PDA Annual Meeting in San Antonio, Texas in April, David Doleski emphasized the importance of clarity around these changes in a presentation on regulatory expectations for BLA CMC filings.  Doleski left CBER’s Office of Compliance just prior to the presentation to join CDER’s compliance office as Acting Branch Chief of its New Drug Manufacturing Assessment Branch.

In his presentation at the meeting, Doleski highlighted “desirable and undesirable” attributes in BLA submissions trends in approvals, and examples of specific issues causing applications to be significantly delayed or not approved.

[The remainder of the story is normally only available to subscribers, but is being made available by special arrangement with Pearl Pathways to showcase the value of our coverage to non-subscribers. For subscription/license information, click here.]

Reviewers Need to Understand Scope of Filing

A key problem area in BLA supplements, the compliance official said, is “a lack of completeness or a lack of clarity about the scope of the changes involved in the submission.  It is very important for us to know up-front what the scope of the changes are.”

The cover letter serves as the basis for determining which reviewer should be involved.  Given the number of supplements the agency has to deal with, Dolesky stressed, “it is important for us to make quick decisions about whether the filing category is correct and which organization should be involved in the review of the supplement – for example, a product specialist or facility CMC reviewer.”

As supplements can be “voluminous,” their scope should be described in the cover letter, the compliance official said.

“Most sponsors do have a good sense of the level of significance of changes,” but they sometimes fail to communicate the most significant changes up front in the cover letter.  “So it is important to put the most important information up front – I want to be clear about that.”

When the changes are not described in the cover letter, the reviewer may think that they are not significant.  Later in the review cycle “they realize that there is kind of a bombshell at the end that could require a lot of clarification or some additional data.”

Cover letters also play a key role in original BLA submissions.  Doleski explained that an example of significant information to be emphasized in the cover letter may be the use of a QbD approach to manufacturing of the product.

For supplements, the changes described should only be those that have been implemented and that the applicant has data for, Doleski noted.

Changes that have not been implemented – “proposals” – are “not read” with the exception of comparability protocols. Comparability protocols are submitted as a prior-approval supplement (PAS).  “Aside from that one exception, we only want changes that have already been implemented,” Doleski explained.

The agency wants the changes to have been put into practice prior to submission even in the case of a PAS, although products cannot be shipped until the PAS has been formally cleared.

Submissions Should Not Be Reviewer Labyrinths

Providing background and references to associated submissions is important, but should be done at the appropriate level of detail, the CBER official noted.

“Sometimes companies will provide excessive information on the existing aspects of the facility,” he pointed out.  “This is particularly a problem with electronic submissions, which basically hyperlink to other prior submissions. So we may have hundreds of pages of something that basically says how the facility was before the changes. You certainly will want to provide background, but not do it excessively.”

Within the submission, there should be cross-references to appendixes, tables, reports, and attachments, Doleski noted, including “clarity” regarding what is new versus the existing aspects of the facility.

“Certainly it is important to give a background on what the state of the facility is prior to the changes in terms of the design of the facility, the equipment, and the process. But you should make crystal clear what was done – what the state was beforehand, and what exactly the changes were.”

There should be a reference to any associated submission – if there is another supplement that is either currently under review or has received recent approval and has “great bearing” or “immediate impact” on the submission under review.  An exhaustive listing is not needed as the reviewers can locate other applications that are in the agency’s system.

Doleski stressed that when a firm chooses to make related submissions separately, it is important to explain that up-front.

“Sometimes companies will break apart a rather large project into different parts,” he pointed out. “Let’s say, for example, they might upgrade a room classification or make a room modification to the qualification – that would be one supplement.  Then subsequently they will change the equipment and the process, and that would be a separate supplement.”

When taking this approach, it is important for the firm to explain its “project plan.”  If it is unclear, it will lead to questions that can cause delays.  If there are any major deviations or failures – which are “certainly going to happen” – they should be explained and highlighted, including how they have been addressed.

Doleski stressed that submitting “a data dump of disorganized information” is a “really bad idea” that can “undermine confidence in your company.”

He has seen this happen most often after a merger or acquisition where regulatory affairs personnel who prepared the submission at one of the pre-merger companies needed to interact with a post-merger group, “perhaps in a distant facility,” and although they did not “completely understand what the scope of the change was” they submitted it anyway.

Also problematic is “excessive reliance on related or associated validation studies,” the DMPQ team leader said.  He provided examples of when that reliance is appropriate and when it is not.

If a firm is using, for example, a 10ml and a 50ml vial with the same stopper and has conducted container closure integrity studies on both, it may be appropriate to introduce a 25ml vial of the same type without repeating the integrity testing.

“On the other end of the spectrum,” it is “not appropriate” for a if a firm put in place a filling line that is identical to an existing filling line and maintain that media fill studies do not have to be performed on the new line.  Although there is “no clear line in the sand,” Doleski commented, good judgment and scientific rationale needs to be used when referencing related validation studies.

DMPQ has also seen applications where validation studies were omitted. “Most of the time this is not acceptable,” he emphasized, although “in rare instances, something can be worked out beforehand.  But you should be clear at the beginning that that study is missing.”

First Submission Approval Trend Positive

Doleski provided an analysis of the trends in first cycle approvals, information requests (IRs) and complete response letters within the purview of his group over the last six years.  CBER’s DMPQ is involved with manufacturing issues such as the review of new facilities, changes to facilities and equipment issues.

During the past six years, the “vast majority” of submissions to DMPQ received first-cycle approval, and the relative percentage of CR letters decreased, which Doleski characterized as “good news.”

“I think the reason why is because our reviewers have been asking more questions up front,” he said. “We have been getting to the reviews earlier.  We have been engaging industry earlier.  We have been resolving issues during the course of the review process.”

Information requests have improved the review process compared to “years ago,” when “all of the issues would eventually be discovered at the end of the process,” Doleski commented.

Click to Enlarge

However, IRs have a downside – generating them requires significant time investments by both the applicant firms and the agency.  “We have to distill down the questions and transmit them to you.  There is tracking involved.  It adds to the complexity of the process, particularly if you consider that many of our reviewers may have 30, 40 or 50 different supplements at any one time they are reviewing.”

The compliance official also pointed to the risk that IRs inject into the process:  If the information requested does not arrive in time to allow for approval, that will increase the likelihood of a CR letter.

“You would get a CR letter and then the clock would be stopped,” he explained. “Now we are resolving them at an earlier point.”  He added that if the submission was of “high quality,” it would “not be a necessary to go back and forth with the information requests.”

Doleski provided examples of deficiencies that led his group to issue CR letters from January 2008 to the present, noting that inspections, test methods and stability testing encompassed the top three reasons for the letters.

The most frequent cause for issuing a CR letter is adverse observations made during a GMP inspection.  “Obviously inspections are related, but certainly different than our review of applications,” he pointed out. “But the GMPs can certainly affect whether you have applications approved.”

Test methods used for “everything except finished product,” including cleaning, sterilization and visual inspection methods ranked second. These include cleaning, sterilization and visual inspection method issues. “Visual inspection is a topic that I think we are more and more focused on, particularly as companies are going to automated inspection systems,” the compliance official commented.

Stability testing concerns generally related to the lack of stability data included in the submission.

_____________________________________________________________________

FDA’S DAVID DOLESKI ON DO’S AND DON’TS OF BLA CMC SUBMISSIONS

Good Review Management Principles and Practices

The goal of the Good Review Management Principles and Practices (GRMP) for PDUFA products is to achieve first cycle approvals for NDAs, BLAs, or efficacy supplements. That certainly should be your goal, and that is our goal as well.  It also describes best practices both for the applicant – for yourself – and for us, the reviewers.

Here are some of the important principals for the GRMP Guidance:

The foundation for good review management is created during product development. Certainly that is the foundation of all the information that you have. If you have done a very thorough job with product development, you should have the basis for a good quality submission.

The second point is that the applicant is responsible for submitting a complete marketing application – underscore the word ‘complete.’ I think that as I go through these slides you will see that one of the main issues we have is a lack of completeness or a lack of clarity about the scope of the changes involved in a submission. Certainly if you adhere to good review management, it will increase the likelihood of first-cycle approvals.

Attributes of a Desirable Supplement

I came up with a list of what I thought were attributes of a good supplement submission:

The major changes should be described in the cover letter. It is very important for us to know up-front what the scope of the changes are. That cover letter and initial quality summary are the basis of someone like myself determining who should be involved in the review of this submission. So it is important to put the most important information up front – I want to be clear about that.

Click to Enlarge

The submission should describe changes that have already been implemented – in other words, changes that you already have the data for. If you use the CMC format, the application should be well-organized.

Within the submission, there should be cross-references to appendixes, tables, reports, attachments, and so forth. There should be clarity regarding new versus existing aspects of the facility. Certainly it is important to give a background on what the state of the facility is prior to the changes in terms of the design of the facility, the equipment, and the process. But you should make crystal clear what was done – what the state was beforehand, and what exactly the changes were.

If you use any alternative approaches – let’s say for example, a matrix approach using a surrogate or alternate method – you should be very clear about your rationale for doing so and give some type of scientific justification.

There should be a reference to any associated submission. In other words, if there is another supplement that is either under review currently or has received recent approval and has great bearing on the submission under review, it should be cross-referenced. Obviously you do not need to have an exhaustive listing, because we can check our system and determine that – but reference anything that has immediate impact on the submission.

Sometimes companies will break apart a rather large project into different parts. Let’s say – for example, they might upgrade a room classification or make a room modification to the qualification – that would be one supplement.  Then subsequently they will change the equipment and the process, and that would be a separate supplement.

So what you should do is explain that up-front – explain your project plan.  Otherwise we are going to ask why you have not described the equipment changes and so forth.  If there are any major deviations or failures – certainly that is going to happen – you should explain them and be up-front about them and how they have been addressed.

Attributes of an Undesirable Supplement

Now here is what you should not do. Here are attributes or qualities of an undesirable supplement submission.

Major changes are not mentioned in the cover letter or they are downplayed – they are very deep in the submission. What happens then is you have a reviewer who is thinking that perhaps this is a less significant change.  They get most of the way through the review, most of the way through the cycle, and then they realize that there is kind of a bombshell at the end that could require a lot of clarification or some additional data. So be very clear about the scope of the changes in the beginning of the submission.

There are times when validation studies are simply omitted from the application. Most of the time this is not acceptable.  In rare instances, something can be worked out beforehand.  But you should be clear at the beginning that that study is missing.

Sometimes companies will provide excessive information on the existing aspects of the facility.  This is particularly a problem with electronic submissions, which basically hyperlink to other prior submissions. So we may have hundreds of pages of something that basically says how the facility was before the changes. So you certainly will want to provide background, but not do it excessively.

[Regarding] changes that have not been implemented – basically this means proposals.  We do not read proposals, with one exception – that would be comparability protocols. Comparability protocols come in as a prior approval supplement.  Basically that is a plan for future implementation of a validation. Aside from that one exception, we want changes that have already been implemented.

Do not describe changes without providing associated validation or qualification data – obviously you should provide complete data.

[Do not provide] a data dump of disorganized information. This is a case where more likely than not there has been a merger or acquisition, where the regulatory affairs group that prepared the submission had to interact with another group, perhaps in a distant facility, and they did not completely understand what the scope of the change was, but they submitted it anyway. This is a really bad idea because it kind of undermines confidence in your company.

Excessive reliance on related or associated validation studies – certainly there are appropriate times where you can reference other validation studies. Just to give you an example:  You have two container closures, let’s say a 10ml and a 50ml vial.  They have the same stopper. You have done studies for both on container closure integrity and you introduce a 25ml vial. Certainly you can make a rationale that it is not needed to repeat container closure integrity testing.

On the other end of the spectrum, you might have a filling line that is very similar or identical to an existing filling line and you say ‘we do not need to provide media fill studies because we have done so for the first one.’ That is not appropriate. There is no clear line in the sand.  But I would say there are clearly times when companies have every single validation study and they cross-reference other studies that have already been done in another situation.

You should provide the rationale for the approaches that you use, whether it is a surrogate or a placebo, etc.

BLA and Supplement Content

At the end of the slide, there should be a list of references (see box below right). These are CMC guidances which provide a description of what the content should be and the format for application and supplement.

Click to Enlarge

Obviously you should provide a manufacturing facility description, some of the major methods used, the container closure system, microbiology and lyophilization data.

Sometimes companies forget to include water systems. If you have a new BLA or new facility, obviously you should include the water systems.  [Also describe] cleaning, sterilization, depyrogenation, and computer systems. I would say that there is a fair likelihood that companies sometimes omit these descriptions.

And again go back to the CMC guidance. It will tell you what is required as well as the separate sterilization guidance that talks about the sterilization information you need to include.

Trending Information

I have done some trending of some of the submissions that have been received and approved at CBER and created a chart (see box).

Click to Enlarge

What you will notice is the blue bars are first cycle approvals. The purple bars are resubmitted approvals.  That would be a situation where a CR letter has been issued. So the clock has been stopped from our perspective. The company has responded to the CR letter and they have resubmitted it and now they received approval. The third, yellow bar is the complete response. So that is a CR. That is where we have closed it out and said we cannot approve it at this time.

So what can we say about this chart? The first thing is that the vast majority of first submissions are approved. You can see the blue bars are much higher than the others, which is a good thing from your perspective. But there is more to the story.

I did some further trending of our CR letters. The scope of this is just from my group, the Division of Manufacturing and Product Quality [DMPQ]. Just a quick description of the organization: We have different offices: the Office of Vaccines, the Office of Blood and Office of Compliance. My Group, DMPQ, is involved with manufacturing issues: new facilities, changes to facilities, equipment and so forth.

I counted the number of CR letters that were within the purview of my group for the past six years. Then I counted the total number associated with those.  That’s the starting point.  Then I went into the number of information requests associated with these submissions. I divided the two numbers so I could determine the number of information requests per submission. And this is an aggregate – the total for all CR letters (see box).

The maroon bars [represent] the number of CR letters. Then there is a blue line, and that is the average number of information requests per CR submission. So there are two trends that become obvious. The number of CR letters seems to be diminishing over time and even as a proportion – if you account for the variation to the total number of submissions, you will see that the percentage of CR letters is declining, which is good news for the industry.  The blue line shows an upward trend. So that means we are going to the applicant more frequently and asking more questions.

[Regarding] information requests (IRs) during submission reviews: We are clarifying the content of the submissions. We are asking you to assist us in the interpretation of results, if the interpretation is unclear. We may be asking for additional validation for the qualification studies. We may ask you to modify the scope of the submission. Let’s say that the data does not support the changes – we may ask you to scale back the scope of the submission. Lastly, we may ask you for post marketing commitment.

I am going back to observations about the chart. The number of IRs per CR seems to be increasing – so we are asking more questions. The number of CR letters seems to be decreasing.  I think the reason why is because our reviewers have been asking more questions up front. We have been getting to the reviews earlier.  We have been engaging industry earlier.  We have been resolving issues during the course of the review process.

It used to be, years ago, that all of the issues would eventually be discovered at the end of the process. You would get a CR letter and then the clock would be stopped. Now we are resolving them at an earlier point. However, if you have a high quality submission, it would not be necessary to go back and forth with the information requests.

The downside is that generating IRs requires a ton of expenditure on our part and your part. We have to distill down the questions and transmit them to you.  There is tracking involved.  It adds to the complexity of the process, particularly if you consider that many of our reviewers may have 30, 40 or 50 different supplements at any one time they are reviewing. There is a risk involved here. The information may not arrive on time to allow for approval, so that increases the likelihood of a CR letter.  Once again, it would be better to have a high quality submission right from the start.

Problems with Initial Submissions

I have basically distilled down from the submissions the many scenarios. The point is to get a sense of the back and forth. These are somewhat based on real submissions, but the scenarios have been changed slightly.

This was a submission on the addition of a nanofiltration step. There was some lack of clarity in the areas of the equipment used. We had questions. The response came back that the equipment was not new.  But upon further clarification, they were able to reveal that the equipment and the area were new to the product. So essentially what came out is that there are other products being used that are going through this nanofilter skid. So then cleaning and changeover becomes an issue.

This is a CBE [changes being effected supplement] for the addition of a manufacturing area. In hindsight, it should have been a PAS [prior approval supplement]. The submission was very vague about the changes. It was very scant in terms of the amount of information it contained.

The question was about what equipment and what manufacturing steps were involved in these new manufacturing areas. It turned out that filling was one of the operations in the new area, and there was a new filling line being used. This then necessitated a request for media fills, filling uniformity, container closure integrity, and so forth.

The last example [pertains to] the addition of a TFF [tangential flow filtration] unit. ‘Please clarify any facility changes associated with the new TFF.’ Those of you who are familiar with plasma-fractionated products, you know that it is important to have the pre-viral inactivation area separated from the post-viral inactivation area – particularly with respect to air pressures, personnel flow, equipment flow and so forth. It turns out that there were room modifications with this pre-viral inactivation area, so then we had asked for room qualification information

Analysis of CR Letters Issued by DMPQ

I performed an analysis of CR letters issued just from DPMQ and just involving manufacturing facility issues from January 2008 to present. I came up with different categories or subject areas. The different subject areas could be associated with the CR letters. So basically I went through the CR letters and determined if there was a cleaning issue mentioned in the CR letter or an equipment sterilization or filtration issue. I came up with a listing of the most common topics that are associated with our CR letters (see box).

Click to Enlarge

I came up just with counts. These are total counts. They are not adjusted for the type of submission. Obviously it would be more interesting to know what the baseline is – what are the types of supplements that we received and what is the incidence for those types of supplements.  But I was not able to do that [at this time].

What you will notice right off the bat is that inspections are the number one reason for CR letters. Obviously inspections are related, but certainly different than our review of applications. But the GMPs can certainly affect whether you have applications approved.

The second topic is test methods. In this context it is not necessarily finished product test methods, but rather test methods associated with cleaning and container closure integrity, for example – everything except finished product test methods. There are cleaning, sterilization issues and visual inspection issues. Visual inspection is a topic that I think we are more and more focused on, particularly as companies are going to automated inspection systems.

Stability testing – most of those issues simply relate to a lack of stability data.   The last item is simply a lack of clarity on the facility, equipment, and manufacturing process.

Container closure integrity testing is also something we have been very focused on and we are asking more and more questions about it.

Most of the filtration issues involve changing filters. Our questions involve the process parameters, like the time, the flow rate, and why there was filter clogging.  Oftentimes filters are changed because there is clogging or because there is a new type of filter, so we are going to ask a lot of questions about that.

Deviations, investigations, discrepancies and failures – basically there is an unresolved issue, there is some problem with the validation that has not been adequately resolved.

Media simulations most of the time simply involves a lack of media fill information that has been provided.  The same goes with the last three counts [storage/hold times, depyrogenation and validation/consistency of lots] – it is just a lack of information provided.

These are the tail end in terms of problems that we find in CR letters. Not to say that they are not important, but many of these issues have been resolved throughout the course of the review.

Examples of Issues Contained in CR letters

I am going to provide some examples of issues contained in CR letters. The purpose of this is just to get a sense of what the questions are.  I will not necessarily get into a lot of details about the issues themselves, because many of these issues can be the topics of complete presentations.  [Editor’s Note:  The complete list of examples from Doleski’s slides is provided below].

As for methods, a lot of the CR letter issues have to do with container closure integrity, use of the dye method, and basically determining the sensitivity of the method. For cleaning validation, there are a lot of questions we have or issues with the acceptance criteria where basically the limits are too high. And the detection of the cleaning agent – particularly if there is a new, unusual agent – we want to find out how you detect it.

These are just more examples of cleaning issues that were contained in CR letters. In the last one, for example, there were some failures of the cleaning validation that were not resolved.

Basically we do not often ask for a full recovery study. If there is an issue with it, you should give us information on what the recovery is. In the top bullet we basically wanted more information on the cleaning agent, the appropriateness for the type of soils encountered, and the effectiveness of the methods to detect the cleaning agent.

[Regarding] sterilization:  It used to be that we would have to find load patterns and those would be validated. In order to have more flexibility, companies are saying ‘we have a mixed load’ and they want flexibility in terms of the number of items they can put in there and the relative distribution of these items throughout the cart. So we are finding a lot of issues with sterilization.

This was taken from a situation where a company was failing their visual inspections. They had multiple re-inspections. Even their re-inspections had some failures.  They had a particular particulate that was identified, but they had not fully characterized what the substance was. They had not provided us with the test results. We wanted to know what the contaminant was.  So we asked a lot of questions about that visual inspection process. And certainly we wanted to know which lots were affected and whether they were released or not.

[Here] we wanted the justification of the reject limit. We wanted some clarification of their procedures for performing the visual inspections and a toxicology assessment for the particulate. This is a somewhat unusual situation in terms of having problems with these particulates, but in general visual inspections are something that we ask a lot of questions on regardless of this kind of unusual scenario.

Container closure integrity testing:  A lot of these questions have to do with the use of microbial ingress or dye ingress and how you have determined whether there is an adequate level of detection for that method.

Filtration:  Basically we are asking about time, filters, and flow rates. And in this particular scenario there was an argument made that there was clogging of the filter and that it was based on time, but when we did an analysis we did not see that the relationship existed.

These are more filtration questions. Basically this relates to storage of UF cassettes.  We were concerned about having unlimited storage time and whether there could be the impact in terms of having bioburden because the filters were then used without testing for the presence of bioburden and there was no validation on the maximum storage time.

This is just an example of aseptic simulations. The most common issue is that typically the media fill information is not contained in the submissions, so I would ask for the simulation data. We will ask how many vials were involved, how many operators were in the room, the time frames, and so forth.

Click to Enlarge

This is just an example where validation lots were not submitted. We wanted them submitted but we also wanted to know what the hold times were for these lots.

In order to be proactive and to prevent some of these issues, we would advocate the use of Type C Meetings if you:  ● have substantial revisions or modifications to your facility ● build a new facility ● are using some new technologies or some use of new approaches that you have some uncertainties about ● use placebos, for example – in lypohilization, ● use a matrix approach for experimentation ● use comparability protocols, or ● use quality by design.

.

CMC Problems Cited in CR letters

Method:

• The visual dye method used for container closure integrity testing has not been qualitatively validated.

• The cleaning validation acceptance criteria for TOC and conductivity were calculated by adding four times the standard deviation to the mean of the data. Please provide additional information and data to support this approach.

• Please indicate how you determined the adequacy of this cleaning agent and the effectiveness of test methods in detecting residual cleaning agent.

Cleaning:

• The submitted data does not support your cleaning validation acceptance criterion of ≤ 10 μS/cm for conductivity. Please note that the conductivity specification for final rinse water during cleaning validations should be closely associated with the USP Purified Water specification. Please revise your conductivity specification for periodic monitoring and for future cleaning validations.

• You indicate that cleaning validation acceptance criteria for TOC and conductivity were calculated by adding <number> times the standard deviation to the mean of the data in that this allows for fluctuation in the process and assay. Please provide additional information and data to support the use of this statistical analysis to set acceptance criteria for cleaning validation.

• Please submit the investigation(s) and any corrective actions for the failed conductivity values <numbers> μS/cm for the mixing tank test samples.

• Please provide more specific information on how you determined the adequacy of use for this cleaning agent, namely, information on the properties of the cleaning agent, its appropriateness for cleaning the soils encountered in the production areas and the effectiveness of the test methods to detect this cleaning agent.

• Please submit recovery studies conducted to support your swabbing procedures.

Sterilization:

• Your study did not adequately validate the random component and equipment load configuration you intend for production. Since your sterilization load configurations are mixed load with a variety of different equipment, you have not adequately defined which equipment is the most challenging for sterilization. In order to validate your random autoclave loading pattern, please provide a comparative sterilization resistance

study to determine which equipment is the most challenging for sterilization. This study should be performed with representative packages of the stoppers and equipment which have biological indicators (Bls) placed in worst case locations.

• Once the item that is the most difficult to sterilize has been identified, please provide a sterilization validation with a full load to demonstrate that position of equipment in the fully loaded sterilizer does not impact sterilization. Alternatively, you may define your load configuration for production and then provide additional sterilization validation information for your defined load pattern.

• Please provide for each autoclave model the microbial and thermocouple sterilization validation data which supports your fixed load validation approach for all intended load configurations.

• Please note that these load configurations should be documented and reflect the intended load configurations that will be used in production. Please provide the load configuration documentation, e.g., picture of each load, and all deviations encountered during the validation.

Visual inspection:

• Please provide a rationale for inspection reject limits, AQL acceptance criteria, and conditions when a re-inspection is allowed under your SOP.

• In review of the SOP, we note that there is no specific requirement or limit for first re-inspection when the lot fails 100% visual inspection. For example, Lot <number> failed the 100% visual inspection. However, the reject limit is 1%. In addition, the lots failed multiple re-inspections for the same defect. Please provide the justification for considering this lot acceptable for further re-inspection with almost 45% of the bottles rejected due to a single type of defect.

• The rationale for releasing lots potentially contaminated by <substance> particulates.

• Similarly, please provide the maximum allowable reject limit, your justification for allowing a second re-inspection, and conditions when the lot will fail and be deemed not releasable.

• As noted in your SOP, a maximum of two 100% reinspections shall be performed. One additional 100% re-inspection may be performed and shall require a documented rationale, approved by the quality assurance director. Please provide the documented rationale, supplied by data, approved by the QA Director, for each of the third re-inspections for the following lots: <lot numbers>.

• A complete investigation report, including the toxicology assessment.

Container closure integrity testing:

• Please submit your plan for demonstrating that the sensitivity of your visual inspection dye ingress test can achieve an adequate level of detection. The plan should specify tests and studies that will be performed, including the analytical procedures that will be used, and acceptance criteria that will be archived to demonstrate equivalency. If you elect to incorporate a new quantitative container/closure integrity assay, please provide your plan for validating the new method.

Filtration:

• You state that <product> lots with greater than <quantity> protein may result in an increase in UF filtration time and a decrease in filtration flow rate. There were six lots identified as having protein levels equal to or exceeding <quantity> protein (see table below). However, the increase in filtration time and the decrease in flow rate appeared to have no direct relationship to the protein content. For example, Lot <number> had a higher protein level, but did not experience a decreased flow rate or an extended filtration time. Please provide an explanation based upon your investigation for this discrepancy.

• Please provide a complete copy of the microbial retention study performed on filters used to sterilize the product. Please include a discussion of the results and your conclusions.

Filtration/cleaning:

• We do not agree with your statement that a defined maximum hold time is not necessary for storing the UF cassettes at room temperature in the UF skid based on your validated cleaning process and storage under sanitizing conditions. To date you have provided no data to support an indefinite hold time at room temperature. Please specify the maximum timeframe that the UF cassettes may be held (stored) in the skid prior to performing a re-cleaning of the UF system.

• In addition, you indicate storage in a 1% sodium hydroxide solution at room temperature has been confirmed by <Filter Manufacturer> as being adequate. Please provide documentation and data from <Filter Manufacturer> demonstrating that bacteria and/or fungal contamination does not occur when the UF cassettes are held at room temperature in the skid for an indeterminate time frame.

Aseptic simulations:

• Please provide a summary report and protocol for the last three media fill studies to support aseptic filling of the product on the filling line. Please include media fill study design parameters used during these studies – for example, number of units filled, size of the commercial batch, number of shifts covered, line speed, number of personnel present, environmental conditions, holding period, number vials filled vs. number vials incubated and interventions. Please summarize how these media fills represent the contamination risk factors that could occur during filling of your product on the filling line and accurately assesses the state of process control.

Validation lots:

• Please submit full characterization data for three new consecutive validation lots of <product name> Drug Substance manufactured using the new procedures to support consistency and comparability of drug substance manufactured using the new process. Please ensure that all holding times remain within the set acceptance ranges.

©2021 IPQ Publications