Subscribe
  
IPQ Inside the global regulatory dialogue

FDA’s ONDQA Bolsters its QbD Review Effectiveness

FDA is continuing to refine its quality-by-design (QbD ) review framework by creating QbD “CMC lead” and “liaison” positions within the Office of New Drug Quality Assessment (ONDQA), instituting an internal database of QbD application elements for use by reviewers, and increasing collaborative research between ONDQA and academia on QbD-focused topics.

Additional support of ONDQA’s review efforts is being provided by the implementation of a quality management system (QMS), which is well under way, with the first audit of the system taking place in November, 2011.

ONDQA Acting Director Christine Moore told attendees at the ISPE Annual Meeting in Dallas, Texas, in early November that her office has been “working very actively” for about the last year and a half on implementation of a QMS.  “We have foundations in place – for example, the first SOP that you write on how to write an SOP.”

Moore provided updates on her office’s efforts to: ● meet the challenges of applying risk assignment-based review approaches to new drug applications ● advance its internal QMS ● improve the review of QbD applications, and ● collaborate with academics to advance the science related to manufacturing and product quality.

[The rest of the story that follows is normally only available to IPQ subscribers.  By special arrangement between IPQ and PharmaQbD, the complete story is being made available  to showcase the in-depth analysis IPQ provides.]

QbD Application Numbers and Complexity Challenging Reviewers

More than 50 new drug submissions and supplements containing QbD elements have been reviewed to date by ONDQA, and the complexity of some of the applications is challenging agency reviewers.

“At this time,” Moore commented, “I feel like we have a good understanding of how to deal with what I would call ‘standard’ QbD approaches, the more simple ones – evaluating CQAs, quality risk assessments, design space, etc. But I will admit that we do still need more experience to standardize our approaches with the more complex processes, [such as] real-time release testing and continuous manufacturing.”

She noted that ONDQA has not seen many examples of the more complex approaches, and that its experience base in reviewing them is “still evolving.”

To help reviewers with the applications, ONDQA’s steering committee instituted an internal, searchable database for tracking quality-by-design elements in applications.

“So if a reviewer comes across a design space for a particular unit operation, they can go to the database and easily find the other applications where similar approaches were used.  And that can give them a good starting point on how to review the application,” Moore explained.

The steering committee also established two new positions to help manage the application process –  “QbD CMC lead” and “QbD liaison.”

The QbD liaison is an experienced reviewer who acts as a mentor to reviewers who have not had as much experience “specifically in QbD approaches and the thought process that we need to go through” to evaluate those applications, Moore pointed out.

Other activities undertaken to support the review of QbD applications include: ● “extensive” QbD training ● increased reviewer participation in inspections ● developing internal guidelines on review considerations for QbD approaches, and ● collaborative research between ONDQA and academia on QbD-focused topics.

FDA/Academia Collaborations Bearing Fruit

Collaboration projects between FDA and universities on modeling continuous flow reactors, characterizing complex products and drug dissolution have produced results that will help “advance quality by design and the various tools” that are used in its application, Moore explained.

“Essentially, what we are looking for here is to go from an expert analyst to an expert system – having the methodology and the system in place that can provide you with reliable, reproducible analyses.”

A project with the Center for Process Analytical Chemistry (CPAC) at the University of Washington is looking at critical response factors in a continuous flow reactor and coupling the sampling and monitoring technique through the use of Raman spectroscopy as the analytical method.

“They have already shown a proof-of-concept of this and are moving forward from just measurement into incorporating control aspects, so that adjustments can be made based upon the changes that happen in the reactor,” Moore pointed out.

Another recent project, spurred by the heparin crisis, is focusing on characterizing complex products using chemometric techniques to produce “fingerprints” that can be compared against the spectra of product that is suspected of being contaminated.

“By using chemometric techniques you can you pull out patterns from that analytical data. It has the ability to handle that in a multidimensional fashion,” the ONDQA official noted.  “We are looking to extend this methodology to other complex products” including botannicals.

A third collaboration is evaluating “FloVitro” technology to simulate gastric conditions and help predict dissolution profiles for oral drugs.

Barriers to QbD Implementation Need to be Overcome

At the APIC/CEFIC European Conference in mid-November in Munich, Germany, Moore addressed cultural, organizational and regulatory barriers that need to be overcome for improved implementation of QbD.

Noting that a general theme at the conference was the reluctance to share detailed confidential information in a drug application, Moore commented that “there is still quite a bit of a barrier to that in the mindset of many in industry.  Having a collaborative and communicative [relationship] between industry and regulators and building that trust” is an essential obstacle to overcome.

One effective means to work toward the institution of a quality culture is “using a personal responsibility to quality as a way of operation within your organization,” Moore commented.

Business challenges exist both in industry and in regulatory agencies, she explained.  These include silos between various units and “just getting people to work together and talk together effectively.  Budgeting issues and which unit pays for what can also be a challenge.”

Regulatory barriers to QbD implementation include: a lack of harmonization of requirements and compendial standards; and a lack of clarity regarding what information is expected in applications and where that information is reported.

The ONDQA official noted that she hears “more and more” about a lack of international harmonization of CMC requirements – especially in countries outside the ICH regions.  “That is definitely a concern that is commonly or frequently expressed by people in industry.”

Even within the ICH regions it is not always clear how or where to represent information on a drug application – for example, a comprehensive control strategy (IPQ “In the News,” July 18, 2011).  “In the current CTD format,” Moore explained, “it is in maybe four different places.”  The question is, “how do you get the story to come out?  Because really what we are most interested in as regulators is the story of how you are controlling product quality.”

In addition to where the information is included, it also remains “unclear” how much information is needed to support flexible regulatory approaches.  Although the ICH Implementation Working Group (IWG) has supplied some direction in this regard, “I think that more experience on the international level will help clarify that,” she said.

GMP expectations – especially in non-ICH regions – and compendial standards as laid down in the various pharmacopeias also require more clarity.

The ONDQA official pointed out that by improving its implementation of QbD, the potential benefits to the pharma industry include: ● improved manufacturing efficiencies ● ability to adjust rapidly to market demands ●enhanced innovation, and ● higher product quality.

Although other industries are ahead of pharma in being “proactive instead of reactive” and are achieving superior quality results, “I think the pharmaceutical industry can get there,” Moore emphasized.  “The science is there to do it.  It has been done in other industries.  Industry and regulators need to continue to work together to get to that point. And that will allow us to achieve comparable metrics to these other industries, such as cycle times, plant utilization, inventory, etc.”

One area in which better process design and control has “great potential” is the ability to adjust to changing market demands on short notice.  “That is useful not only from a business perspective but also, from a patient health perspective and to adjust for potential drug shortages.”

Enhanced innovation is “something we continue to see in the pharmaceutical industry,” the ONDQA official stressed.  “I recently did a review of the applications for the new drugs that FDA has approved recently.  I was really surprised by the high number of fixed dosage form combinations and combination products – products other than just tablets, capsules and parenterals.”

All taken together, QbD implementation can deliver “a higher assurance of pharmaceutical quality” to the patient.

“There is the potential for improved quality, efficacy, reliability and cost,” Moore maintained.  “The enablers for this are quite simple:  Sound science, risk-based approaches, effective quality management, quality systems and good collaboration and communication.  That applies both within industry, between regulators and between regulators and industry.”

___________________________________________________________________________

FDA’S CHRISTINE MOORE ON QbD IMPLEMENTATION IN ONDQA

ONDQA Quality Management System

FDA’s 21st Century Initiative:  Along with encouraging industry to establish quality systems and use risk-based approaches there was also a distinct effort to use the same approaches in FDA in terms of our management systems.  Out parent office, the Office of Pharmaceutical Sciences (OPS), has been taking this very seriously

What do we consider our quality management system?  It is probably very similar to those of you in industry.  It is there to ensure that our products and services are appropriate for their intended use.  That includes the reviews that we do and other interactions.

We want to reliably meet customer expectations, whether those be internal or external to FDA.

We also want to have a mechanism to evaluate our products and services proactively in our day-to-day activities and to improve those products and services.

So, where are we in terms of implementing our quality management system?  We have been working on this very actively for about the last year-an-a-half, and we have foundations in place – for example, the first SOP that you write on how to write an SOP, and etc.

The management and all our division directors and branch chiefs have been actively involved.  We have our quality manual – a kind of a roadmap for where we want to go – as well as a reviewer manual, which is a compilation of many different resources for our reviewers.  And we are in the process of drafting many internal quality procedures.

ONDQA is not the organization that is doing this.  This is an active effort throughout OPS….  Some offices are further along than others.

We will also be developing tools that support quality management systems, similar to those in industry, on metrics, training, and management of quality records.

Happening this week, or starting this week – the first week in November, 2011 – is our first system audit.

CDER-Level SOPs for Integration

Another thing we are working in CDER is integration between different offices.  We have been working on integration between OPS – the ONDQA parent office – and the Office of Compliance (OC).

Many of these activities started late last year with a series of retreats into early 2011 to elucidate areas that need better coordination within CDER.  Subgroups were formed from those retreats to work on putting together SOPs for specific critical points of collaboration, including: ● preapproval inspections (IPQ “In the News” November 15, 2010) ● postapproval change management ● recalls and shortages, and ● innovative technologies.

At this point, we have drafted those.  They are so-called in final form, but they will need revisions. Essentially we are going to go ahead with the drafts that we have and implement them on a limited basis – do an evaluation – and make revisions based on what we have learned.  So it is in an early pilot stage.

Going through these activities has really helped our offices to better understand what each other’s needs are and to make minor tweaks to meet each other’s needs.

FDA Track

‘FDA Track’ is a way for us to report metrics.  It is an initiative that was started in 2010 to increase transparency – it lets the public see the activities of various offices and our progress.

The categories that are tracked for ONDQA by FDA Track include the:

● number of correspondences with applicants related to CMC review

● number of ONDQA managed product quality meetings

● number of presentations

● percent of NDAs and NME NDAs containing QbD elements

● percent of CMC meetings including QbD discussions, and

● implementation of ONDQA QMS

ONDQA Pilot Program

Many of you are aware that six years ago we started accepting applications that contained elements of quality by design in a pilot program.

The purpose of the pilot was to provide experience to companies in terms of submitting this information to the agency and experience in terms of FDA reviewing it.

That pilot has been completed for a while. There were eleven approved applications and one that was withdrawn for non-CMC reasons.

It did provide a great learning experience for both the agency and for industry. Together, with the learnings that we had, those ideas were brought forward and put together in some of the new guidances is that we see – ICH Q8 R2 – as well as some things that we are seeing further down the line.

Trends in QbD Applications

The QbD applications that ONDQA received did not stop at the pilot. We have been continuing to receive them and now we have over 50 total submissions that take elements of QbD, and they are continuously increasing in number.

There are various trends that we are seeing, not just in things that we label QbD. It is always difficult to determine whether we have to call something QbD or not, and we get questions sometimes like ‘why do we even have to make that designation?’  One reason is that those have to be reported in FDA Track.

It is not necessarily easy because the science and risk-based approach that we are talking about can be there without additional regulatory flexibility.  But somewhere we have to draw the line and say whether we label it [a QbD application] or not.

Another thing that we are seeing is that these concepts, these science and risk-based approaches, are also crossing into other things that we would not consider QbD applications – things that would not ask for additional regulatory flexibility.

Almost all of the applications that we are seeing include more development information than what we saw 10 years ago. Many of the applications are including enhanced approaches, risk assessments, etc. without asking for design spaces or other types of flexibility.

We are also seeing some trends that we did not anticipate with QbD, where companies have come to us and applied the same sorts of approaches to areas such as analytical methods or container closures.

We are definitely seeing an increased interest in real-time release testing, continuous manufacturing, and implementing biopharmaceutical approaches into QbD. We see that during meetings with the applicants and sponsors, and we also see that through the meetings that we go to.

As far as how our office is responding and what experience we have gained through QbD, I showed you before that the number of applications that we are seeing has been steadily increasing over the past six years.

At this time, I feel like we have a good understanding of how to deal with what I would call standard QbD approaches, the more simple ones – evaluating CQAs, quality risk assessments, design space, etc. But I will admit that we do still need more experience to standardize our approaches with the more complex processes – real-time release testing, continuous manufacturing, etc.

We have not seen as many examples of those approaches and they are more complex, so those concepts are still evolving.   We encourage companies who are interested in pursuing those approaches to have dialogue with the FDA early.

Internal Support for QbD in ONDQA

There are several other things that we are doing within ONDQA to support QbD.

One of them is that we have established a steering committee to have these discussions on hot topics of QbD or new concepts that are introduced.

ONDQA’s steering committee instituted an internal, searchable database for tracking quality by design elements within an application. So if a reviewer comes across a design space for a particular unit operation, they can go to the database and easily find the other applications where similar approaches were used, and that can give them a good starting point on how to review the application.

We have established the position of a QbD CMC lead.

We have also had extensive QbD training.  One thing that we found is that for large group training, we can easily bring in academics, people from industry, etc., to talk about the scientific concepts related to QbD, but we really need to have training between our people to talk about the review aspects and how to do reviews. We have found small group training to be very useful for that.

We have also, in the same light, found a very useful new role called ‘QbD Liaison,’ which is a mentoring role. We will be pairing up more experienced reviewers of these applications with reviewers who have not had as much experience, specifically, in QbD approaches and the thought process that we need to go through with that.

Also, in recent years, we have increased reviewer participation on inspections.

We are also looking at, and are in the process, of developing some internal guidelines on review considerations for quality by design approaches such as design space, NIR, etc.

Finally, we have been doing some collaborative research between ONDQA and academia on QbD-focused topics.  I will be talking about more of that in a minute.

FDA/EMA Parallel Assessment Pilot

I think many of you are also aware that in the past year we started a parallel assessment pilot with the EMA to evaluate QbD-containing applications.

The objectives of this pilot include: ● ensuring consistent implementation of the guidelines towards those applications ● encouraging the already existing program for joint preapproval inspections ● providing opportunities for drug training, which it already has, and ● setting up a pathway for knowledge sharing.

The applications that would be considered in this program are primarily NDAs and marketing authorization applications – or what we would call supplements and which EMA calls variations – that are submitted concurrently or in a very similar time to both agencies.

We can also consider applications that are just submitted to one of the agencies.  But that is not a parallel assessment, it is a more of a consultative role.

We could also consider a ‘scientific advice,’ which the EMA calls it, or what we would call a ‘CMC-only meeting,’ to discuss some of these QbD type concepts.

In a more recent [development], at least one of our applications will include Japanese regulators as observers.

ONDQA Research Collaborations

I will now go in and talk about examples of ONDQA research collaborations.

I wanted to include this topic because I wanted to point out that we have made a lot of advances in terms of QbD and establishing the science, but there are some additional scientific advances that we can go into, as well as things that are currently established, and better disseminate them to a broader audience.

This is a project we have been working on with CPAC, the Center for Process Analytical Chemistry, at University of Washington.

What they are looking at here is to provide publishable examples in terms of how you get process understanding for a continuous flow reactor – for example, what are the critical response factors and how is the response in the levels of those? They are coupling the sampling and monitoring technique through using Raman spectroscopy as their analytical method.

They have already shown a proof of concept of this and are moving forward from just measurement into incorporating control aspects, so that adjustments can be made based upon the changes that happen in the reactor.

This would be a typical diagram – the colorful one of the data that comes from the system.  Once you apply that to calibration modeling you can get, in this case, a concentration versus time [plot].

Another project that we have done over the past year and a half or so has to do with characterizing complex products. This project started after the heparin crisis.

We had a lot of heparin samples and we had a lot of analysis by NMR of heparin samples. The question was asked, ‘can we use that data and this complex analytical methodology in terms of NMR, and what kind of data could be extracted from that?  What could we learn from characterizing this complex molecule?

As many of you probably know, heparin has many different points of heterogeneity in terms of degree of substitution, types of substitution, length of the chain, etc.

This is an example here of different heparin NMR data. This contains both good heparin – heparin that has excessive and natural impurities in it – and contaminated heparin. If you are not an expert, you are going to have a hard time seeing what is good and what is not.

That is the concept behind this sort of characterization. The characterization of complex products can be difficult.  And by using chemometric techniques, you can you pull out patterns from that analytical data. It has the ability to handle that in a multidimensional fashion.

It can lead to hidden and unexpected patterns using the fingerprint approach. This is an example of what the chemometric analysis might look like and how it could separate, in this particular case, the contaminated versus the good product.

We are looking to extend this methodology to other complex products.  We are starting a project on botannicals.

If anyone is interested in these papers, there have been four public papers that are published or are in progress and I would be happy to send you references [see links below].

Essentially, what we are looking for here is to go from an expert analyst to an expert system – having the methodology and the system in place that can provide you with reliable, reproducible analyses.

Another project that we have worked on is using the ‘FloVitro’ technology. This is a multi-chamber simulation device using three chambers meant to simulate the gastric conditions. We looked at multiple drugs, different dosage forms, fast versus fed conditions.  This is just a diagram of the kind of results that you get out related to what the drug would look like in the system and matching that up to in vivo parameters.

Essentially, we are looking at advancing quality by design and the different tools for pharmaceutical applications. It is a collaborative effort. It is really advances from industry and academia, with encouragement and support from regulatory agencies as well.

Conclusions

ONDQA is continuing to meet the challenges of applying risk assignment-based review approaches to new drug applications.

I have shown you today some of the advancements that we have had in our internal quality management systems.

I have also shared some of our continuing progress on implementation of QbD.

Finally, I have discussed how we are collaborating with academics to advance science related to manufacturing and product quality.

LINKS:

Class modeling of heparin NMR data

Heparin analysis using NMR and chemometrics

Heparin chemomteric pattern recognition

Impurities in heparin by NMR

Visit IPQPubs.com everyday for IPQ’s real-time tracking of the latest CMC and GMP developments from “Inside the Global Regulatory Dialogue”

IPQPubs.com

Sign up here to get IPQ’s FREE “Weekly Alert” email

IPQ Publications LLC, 7920 Norfolk Avenue Suite 900, Bethesda, MD  20814
Phone:  202-841-5027,  Fax:  301-913-0119

IPQ Subscription/Company License Information

Contact us at:  [email protected]

Contents Copyright 2011 IPQ Publications LLC. All rights reserved.

©2021 IPQ Publications