Confidential –– Page 1

Confidential and proprietary

Understanding Challenges to Quality

by Design

Final deliverable for FDA Understanding Challenges to QbD Project

December 18, 2009
 Confidential –– Page 2

TABLE OF CONTENTS

Table of contents 2
Executive summary 3
Introduction 5
Methodology 6
Overview of current state of QbD 7
Key adoption challenges 9
Business case for QbD 14
Implications for FDA 19
Appendix 23
 Confidential –– Page 3

EXECUTIVE SUMMARY

This report is the result of work done with the FDA, CDER - Office of Pharmaceutical
Sciences as part of its effort to facilitate adoption of QbD. Its purpose is to help
understand the challenges and opportunities for adoption of Quality by Design (QbD).

The work was conducted with input from several sources: an outside press search / scan
of relevant articles, conferences, databases, interviews with internal experts, as well as a
set of blinded industry interviews.

These inputs uncovered several key facts about the state of QbD adoption as well as
potential catalysts.
1) The understanding and practice of QbD is evolving, gaining momentum and
passion throughout the industry
2) More than half of companies interviewed identified the business case as
strong, however, there are still some skeptics
3) Despite the acknowledgement of a strong business case, companies are at
different levels of maturity –– characterized by four segments –– Novice, Pilot,
Rollout, and Fully implemented
4) 10 key challenges are the most problematic for QbD adoption. These
challenges are evaluated by their relevancy against different drug types as well
as different levels of adoption
y The first four challenges occur within companies
–– Internal misalignment (i.e., Disconnect between cross functional areas,
e.g., R&D and manufacturing or quality and regulatory)
–– Lack of belief in business case (e.g., ““There is a lot of uncertainty over
timing of and investment requirements for QbD implementation””)
–– Lack of technology to execute (e.g., Difficulty managing data, limited
understanding of Critical Quality Attribute (CQA) implications)
–– Alignment with third parties (i.e., How to implement QbD with
increasing reliance on suppliers and contract manufacturers?)
y The next six challenges are directly related to the FDA
–– Inconsistency of treatment of QbD across FDA (e.g., ““Although a
number of people in the FDA are supportive of QbD –– this is not
consistent””)
–– Lack of tangible guidance for industry (e.g., ““We understand what
you are asking for broadly, but there are hundreds of variables –– there’’s
got to be an end in mind –– a tangible one we can deliver on””)
–– Regulators not prepared to handle QbD applications (i.e., reviewers
at different levels of understanding and acceptance)
 Confidential –– Page 4

–– The way promised regulatory benefits are currently being shared

does not inspire confidence (e.g., ““At the end of the day it is still
unclear whether the FDA will actually back these filings””)
–– Misalignment of international regulatory bodies (i.e., Difficulty
gaining acceptance of QbD applications in other countries)
–– Current interaction with companies is not conducive to QbD (e.g.
““…… we are treated with suspicion, it does not feel like collaboration.””)
5) Although key challenges will differ by the level of maturity, there is
uniformity across the participants that FDA can increase the quality and
consistency of its reviews and degree of internal alignment
6) Broadly, lack of corporate alignment is a challenge centered on cultural
issues and the need for cross-functional processes and tools to execute - this
alignment is crucial to reap the benefits of QbD (i.e., an aligned operating
model)
7) Two commonly held beliefs about QbD are not true. If QbD is executed
properly, the cost to implement QbD is negligible, and it will not lead to a
longer development timeline
8) Interviews also identified several steps the FDA can take to accelerate
momentum around adoption of QbD. These steps should address three main
areas
y FDA policy. The FDA will need to think about what incentives, regulation
and direction they will give to companies in industry. These will likely vary
by drug type. There is potential to mandate in areas where a strong case can
be made for QbD supporting patient safety.
y Internal FDA change management. The FDA should also think about
how they can adjust internal alignment to ensure uniformity in QbD
understanding and actions within offices, across offices, and between the
center and the field.
y External change management. Finally, the FDA can consider the best
way to engage both with other international regulatory bodies, as well as
with individual companies.
 Confidential –– Page 5

INTRODUCTION
In 2003, the Federal Drug Administration released its Final Report on Pharmaceutical
current good manufacturing practices (cGMP) for the 21st Century. The cGMP initiative
described a ““Desired State”” for pharmaceutical manufacturing through QbD in which:

¶ Product quality and performance are achieved and assured by design of

effective and efficient manufacturing processes
¶ Product specifications are based on a mechanistic understanding of how
formulation and process factors impact product performance
¶ Manufacturers have the ability to affect continuous improvement and
continuous ““real time”” assurance of quality
¶ Regulatory policies and procedures are tailored to recognize the level of
scientific knowledge supporting product applications, process validation and
process capability
¶ Risk-based regulations are commensurate with the level of scientific
understanding of how formulation and manufacturing process affect product
quality and performance, and the capability of process control strategies to
prevent or mitigate the risk of producing a poor quality product
Recent years have amplified the case for change to the desired state. Safety issues have
highlighted the gap between the pharmaceutical industry’’s quality systems and those of
other industries. Public pressure to reduce pharmaceutical costs is adding to the case to
improve the infrastructure for drug development and continuous process improvement.

The industry has made progress on QbD since the FDA began its 21st Century Quality
Initiative, but this has been a limited number of QbD submissions to date. Despite the
compelling case, the widespread change that the FDA envisioned has not happened at the
pace that was envisioned. Achieving the 21st Century Quality vision will require a
transformative journey for the industry that demands a significant shift in its development
process. This transformation has not taken place due to challenges within companies,
within the FDA, as well as the international regulatory community.

This report is the result of a detailed look at the current state of QbD adoption, focusing
on the challenges preventing full scale implementation. The report additionally takes
preliminary steps to build and quantify a business case for QbD.
 Confidential –– Page 6

METHODOLOGY
This report was generated with inputs from three main sources.
Data mining within internal McKinsey resources
Extensive search of McKinsey database, including case study libraries, benchmarking
efforts, articles, survey databases, and conference presentations. Team identified and
analyzed relevant materials, including previously obtained industry data and surveys, to
act as a supplement to industry interviews. Team additionally conducted several
interviews with global leading McKinsey experts in strategy, R&D, operations, quality
and product development across Pharmaceuticals, Biologics, and Generics. Experts
shared their knowledge and perspectives from many years of experience dealing with key
decision makers in industry to help push thinking on the topic.
Public publication search
Extensive search for most relevant industry articles, publications, and web resources.
Team identified and analyzed relevant materials to act as a supplement to industry
interviews.
Interviews with industry leaders
The steering committee identified a set of industry leaders from leading Pharmaceutical,
Biotechnology, and Generic pharmaceutical companies across Technical Development,
Pharmaceutical Science, Chemistry Manufacturing and Control, Operations, and
Regulatory. Team conducted interviews with identified set of leaders focusing on four
main areas: (1) current state of the company’’s QbD adoption (2) perceived challenges or
barriers to QbD adoption (3) business case for QbD and (4) potential steps to catalyze
QbD adoption.
Exhibit 1: Outline for Industry Interviews
¶ Current state of company’s QbD adoption
o Company definition of QbD
o Current scope of application (mechanisms in place to support)
o Source of motivation to utilize QbD
o Characterization / Track record of company’’s QbD efforts
o For generics manufacturers, impact of Question Based Review (QBR)
¶ Perceived challenges or barriers to QbD adoption
o Technical / scientific barriers
o Internal / organizational realities
o External / regulatory / market driven factors
¶ Business case for QbD
o Perceived business benefits from implementation - anticipated / realized
o Costs to implement –– anticipated / realized
o If applicable, comparison of development costs of QbD relative to traditional
¶ Potential steps to catalyzing QbD adoption
o Internal to company
o Technical / scientific
o Regulatory (FDA related, International)
 Confidential –– Page 7

OVERVIEW OF CURRENT STATE OF QBD

Evolution of QbD over the past year

QbD has continued to gain momentum over the past year. One of the most striking
factors has been an increase in the codification and practice of QbD on a standardized
basis. More and more companies are experimenting with the concept and developing
mechanisms to support it. This is accented by an increase in the passion and recognition
of QbD. Several companies, although not the majority, expressed the opinion that this is
something they would do regardless of regulatory benefits from the FDA.
As companies invest and gain experience with QbD, there are also increasing demands
for the FDA and other regulatory bodies. Two major demands that have emerged are the
need for greater clarification of acceptable methods of execution and filing and or
concrete tangible examples, as well as better codification of the benefits FDA ‘‘promised’’
when QbD was originally advocated.

Perception of the business case

Exhibit 2: Strength of business case by companies examined

Percent of companies examined

Strong business No viable

case with year 1 business
payback case
8 8
Business case is
uncertain / neutral

33

50

Strong business
case with multi-year
payback

[Pie chart showing the belief in the business case for QbD. Strong case with multi-year payback = 50%,
Business case is uncertain / neutral = 33%, Strong business case with multi-year payback = 8%, No viable
business case = 8%]

Levels of adoption of QbD among companies

Companies are at very different places in terms of adoption of QbD. Some companies are
still skeptical about the idea of QbD and have not made much, if any, change towards a
QbD approach. While others have fully implemented the concept and are designing every
product in development along a QbD framework. Most are in between these extremes.
 Confidential –– Page 8

Where companies fall on this spectrum will play a large role in how the FDA should
interact with them.
¶ Novice: Company is skeptical about the value QbD can bring. Utilizes
conventional development and has no platforming.
¶ Pilot: Company is trying QbD, but still on the fence about the potential value.
Tends to apply QbD to a small subset of projects and processes and has
implemented limited or no platforming.
¶ Rollout: Company is convinced about impact of QbD and is beginning to see
some of the benefits. Uses QbD techniques regularly, but not universally. May
engage in some lifecycle management with integrated platform and network
strategy.
¶ Fully implemented: Company is completely convinced about the positive
impact of QbD and is realizing the benefits. Uses QbD in almost every
development program and almost every production step. Additionally, has a
systematic, comprehensive review and re-design of in-line products.
Exhibit 3: Level of maturity and drug type of examined companies

Fully
Group Novice Pilot Rollout implemented Total
New Drug
22% 33% 22% 22% 100%

Gx
40% 20% 40% -- 100%

Biologics
17% 67% 17% -- 100%

[Chart with drug type on y-axis: New Drug, Gx, Biologics. Level of adoption on x-axis: Novice,
Pilot, Rollout, Fully implemented. Level of adoption
–– New Drug: Novice (22%), Pilot (33%), Rollout (22%), Fully implemented (22%)
–– Gx: Novice: (40%), Pilot (20%), Rollout (40%), Fully implemented (0%)
–– Biologics: Novice (17%), Pilot (67%), Rollout (17%), Fully implemented (0%)]

The majority of examined companies develop new drugs. Of these, ~45% fall into the
categories of Roll-out and Fully implemented, another ~30% are piloting. By contrast,
Generics and Biologics manufacturers are at a lower level of adoption, with ~40% of
Generics manufacturers at the Novice level, and ~70% of the Biologics manufacturers at
the Pilot level. Overall, New Drug manufacturers are at a higher level of adoption than
Generics or Biologics.
 Confidential –– Page 9

KEY ADOPTION CHALLENGES

The key objective of this project was to develop a more nuanced understanding of the
challenges preventing adoption of QbD. Research brought 10 key challenges to light.
The first 4 are challenges companies face internally as they attempt to implement QbD.
¶ Internal misalignment. Internal misalignment within a company around if
and how to implement is a key adoption challenge and can take several forms.
These are the five most important manifestations.
1. Misalignment horizontally across the organization. An issue raised
by several companies was the misalignment horizontally across the
organization. This can either be a result of a misaligned organization
structure or misaligned incentives. This results in pockets of QbD
occurring in different areas. For example, while manufacturing is the
group that tends to see the most direct benefit from QbD application,
the resource investment of both time and money comes from R&D.
Thus, the traditional organizational design does not incent R&D to
invest in QbD. As we will see in the business case for QbD (outlined
later in the document), one of the most important factors in successful
implementation is application of QbD across the entire operating
model. However, our research showed that several companies that
have implemented QbD in R&D have not fully carried through and
extracted the potential benefits in manufacturing. This was due to
horizontal misalignment across the organization.
2. Disconnect between leadership and middle management. While
leadership sees QbD as the way of the future and has set an agenda in
the organization, it is not being executed effectively by middle
management. This is often due to a lack of processes understanding,
unclear goals, or lack of clarity on how to execute within an individual
company context.
3. Culture of conservatism. Historically, companies have benefitted
from being as conservative as possible when dealing with the FDA. A
representative comment was ““the best CMC strategy is to not show all
your cards.”” Additionally, in companies where people often have long
careers, there is a tendency to have a historic lens and reliance on
doing things the ““traditional way.”” Another cultural issue raised
around QbD was the question of hampering the freedom of developers.
Some feel QbD would hamper this freedom, believing ““development is
an art-form…… how can you make a platform out of an art-form?”” An
additional cultural challenge is the fear of doing something that is
unknown. Several companies commented that they ““didn’’t understand
QbD until [they] did it.”” QbD is learned by doing and fear to invest in
something that is unclear is holding several companies back.
4. The amount of change required within company is not feasible. In
order to really implement QbD, many companies will have to redesign
certain aspects of their operating model. This type of redesign will
inevitably take time, money, and involve a lot of corporate politics.
 Confidential –– Page 10

For a lot of companies, implementing QbD is ““not enough of a priority

to do that.””
5. For some, QbD remains low on the priority list. Most managers
have many things on their plates and a laundry list of priorities.
Unless the benefits are guaranteed, it is made a requirement, or is
specifically made a priority by senior leadership, QbD often falls low
on the list of things to do.
¶ Lack of belief in the business case. The majority of companies believe that
there is a strong business case for QbD, but there are some skeptics. Many
skeptics from Generics companies believe that QbD will slow time to file, and
that the ‘‘First to File’’ concept is not conducive to QbD development. Several
Biologics companies feel the amount of clinical trials necessary to implement
QbD for drug substance production steps (e.g. upstream) make the business
case negative until there are further advances in the actual science. Others just
didn’’t believe that QbD as a concept ““will have any benefits that will truly
change the safety or efficacy of a drug.””
¶ Lack of technology to execute
1. Insufficient solution for controlling variability of raw materials.
Industry is divided on what QbD means for raw materials –– some
claim since raw materials are not a part of the development process,
QbD does not apply. Since they often are ““the biggest cause of
variability,”” it does not make sense to apply the rigor of QbD
elsewhere. On the other hand, there is the feeling amongst others that
QbD provides an opportunity to apply more rigor to raw materials and
better understand their parameters and impact on process performance,
safety and efficacy.
2. Limited understanding of implications of quality attributes. A
cornerstone of QbD is an ability to define the Critical Quality
Attributes (CQA) of a product. In order to have a solid understanding
of the CQAs, one must also have a solid understanding of how the
drug works and what parts of the drug are most important for efficacy.
Some skeptics claim that ““you will never know enough about your
materials and process”” to be able to really understand. This problem
was raised frequently for Biologics, especially at the molecular or bulk
level. Scientifically, ““we are just not very clear about what aspects are
doing what.””
3. Knowledge of QbD techniques is limited. When implementing QbD,
a company will need to explore new experimental techniques and
make new types of measurements. Some companies feel they are
simply not at a point where they can handle these demands and will
have to ““develop new skill sets and get new technology.”” One
interviewee from a company that successfully implemented QbD,
commented that ““developing the technical tools and standardization
necessary was a big challenge.””
 Confidential –– Page 11

4. Difficulty with data management. In driving to gain a deeper

understanding of a product’’s CQAs, design space and monitoring the
product during development –– a lot of new data will be generated.
There is a challenge around the ““need to develop new skill sets and get
new technology”” to execute this data gathering and analysis. Not only
will this data need to be mined in a smart and understandable way, this
““huge volume of information”” will need to be continuously managed
and maintained. Companies identified the need for ““more statisticians
with different skill sets and more education to be able to handle all this
data.””
¶ Alignment with 3rd parties. Industry is also concerned with the role of third
parties. Questions arose around how to manage both suppliers as well as
contract manufacturers. Some companies are concerned that ““it will be more
complicated to bring [contract manufacturers] along since all resources are not
under complete control.””
The next 6 are challenges to QbD implementation that are specific to the FDA and other
regulatory bodies.
¶ Inconsistency in treatment of QbD across FDA. There is an inconsistency in
how QbD is understood throughout the FDA, both between different offices
and within offices. The offices define QbD differently. ONDQA thinks about
QbD as being based on scientific proof, while OGD thinks about QbD in terms
of Question Based Review 1 . Even within offices, there is inconsistency.
Several examples were raised where a QbD filed product made it ““all the way
to the end and then the biopharma side overturned what the reviewers had
already approved.”” This inconsistency leads to a lack of ““trust that the FDA
will actually accept this filing in a consistent manner.””
¶ Lack of tangible guidance for industry. One challenge that was raised by the
majority of companies, especially those in the novice and pilot categories, was
a need for more tangible guidance, or ““information…… about how to actually do
it on the ground.”” One interviewee commented that ““we understand what you
are asking for broadly, but there are hundreds of variables –– there’’s got to be an
end in mind –– a tangible one we can deliver on.”” Companies are seeking a set
of ground-rules from the FDA around such things as acceptable methods,
criteria to select and deselect critical quality attributes, standards to judge
adequacy of controls, and criteria for analytical method substitution.
¶ Regulators not prepared to handle QbD applications. This challenge has
been apparent for years, and the FDA is taking steps to remedy it. However,
this is such a prominent challenge that it is important to reiterate again.
Interviewees cited more specific problems than reviewers just being ““green””
including ““high turn over”” and the tendency for reviewers to ““fall back to what
their organization used to do [when they worked in industry].”” ““A disconnect
between compliance and field groups…… [also results in] different feedback.””
Additionally, companies are hesitant to flag these issues to senior FDA leaders
as they occur since they do not want to delay or hazard the filing in question.

1 McKinsey White Paper: Transforming to the Desired State of cGMPs for the 21st Century
 Confidential –– Page 12

¶ The way promised regulatory benefits are currently being shared does not
inspire confidence. This challenge is primarily a result of the lack of
codification of real regulatory benefits from the FDA. ““There is questionable
payback –– it is unclear how much regulatory flexibility will actually be given.””
Without clear benefits, proponents of QbD in industry have expressed difficulty
in promoting the idea within their companies, increasing ““skepticism and
reliance on the internal business case.”” There are a lot of strong feelings
around this challenge, interviewees commented ““the FDA has still not shown
us the flexibility we can get from this”” and ““we started this because we were
promised regulatory benefits, now we can’’t even talk about them –– it’’s like [the
FDA] has amnesia.”” Additionally, a couple of companies expressed
disappointment that they had not received as much flexibility as they thought
should be granted in their previous QbD filings.
¶ Misalignment of international regulatory bodies. A large concern,
consistently raised by companies at the Rollout and Fully Implemented levels
of adoption, is whether QbD applications would be accepted by other
regulatory bodies. For example, one interviewee commented that ““we are very
concerned about the response to this outside the US and Europe. We sell
products in 75 countries, most are not a part of the ICH.”” Although no
interviewees have experienced flat-out rejection, there were comments about
the increased time and effort required in other markets. ““It is much more
difficult in secondary markets –– it takes longer to review because they ask a lot
of questions several times.”” This increased effort has caused some companies
to implement QbD practices but not bother to submit QbD filings.
¶ Current interaction with companies is not conducive to QbD. There is an
admittance that ““historically, there is not a lot of comfort talking with the
FDA,”” however, companies are eager to open and improve this communication.
Companies that are passionate about QbD want this to be an open, collaborative
journey. One interviewee felt ““the pilots asked a lot of tough questions, there
were several tough interactions. When we are treated with suspicion, it does
not feel like collaboration.”” One interviewee suggested joint teams working
together. Whatever the solution is, a clear challenge is the current closed way
companies interact with the FDA.

Challenges by industry segments

In order to gain a more nuanced understanding of the challenges, it is interesting to

analyze which challenges are most relevant for the different types of drugs.
¶ New Drug. The challenge companies cited as the most important was the
consistency across the FDA. This consistency pertains to both quality of
reviews, in other words the ““amount of reviewer inconsistency,”” as well as the
understanding of QbD within and across departments. Another challenge that
was cited very frequently, although not highlighted as the most important, was
the lack of international harmonization. The increased effort and time to file
QbD outside of the USA, a result of those countries’’ regulatory agencies taking
““a while to ‘‘get it,’’”” negatively affected the business case. The next big
challenge highlighted was the lack of codification of regulatory benefits and or
 Confidential –– Page 13

what it takes to achieve them. It makes sense that New Drug issues center
around execution of actual applications; in general, examined New Drug
companies were at a more mature stage of adoption. Internally, alignment
across the company is a problem for many New Drug companies. In many
companies ““R&D is incentivized by shots on goal, not QbD”” –– a concept
contrary to good QbD practices of focusing on developing a solid
understanding of the product.
¶ Generics. The most prominent challenge identified by Generics manufacturers
was a lack of belief in the business case. However, there are two camps. One
half believes that there is a business case for QbD in generics and is
implementing. The other half believes that today, ““generics is all about file
first, figure it out later.”” These companies fear the potential additional time in
early development will disadvantage them in making it ‘‘first to file.’’ Generics
manufacturers also identified unclear regulatory benefits as a key challenge ––
likely closely related to the lack of belief in a business case. A final key
challenge from Generics manufacturers was lack of guidance for how to
actually implement.
¶ Biologics. The most prominent challenge identified by Biologics
manufacturers was around the lack of technology to execute. There is
consensus across the board that it is almost impossible to ““prove the molecular
parameters necessary in a QbD file since we don’’t really understand what
effects what”” at the molecular level. Despite the inability to execute in this
upstream arena, there is general agreement that QbD can be applied to
downstream processes, for example purification and formulation processes.
Biologics manufacturers also identified lack of codification of regulatory
benefits as a key challenge.

Challenges by stages of adoption

¶ Novice. Novices have very little experience with QbD. Most of their
challenges are perceived rather than experienced. Internally, they face a lack of
belief in the business case. Externally, Novices feel the regulatory benefits are
unclear and there is a lack of tangible guidance. They simply do not see the
case for pursuing QbD.
¶ Pilot. Companies in the pilot phase are beginning to implement QbD. They
begin to see more tangible internal challenges, such as internal misalignment,
and limited knowledge of QbD techniques. As with novices, there is still a
substantial problem with lack of clarity of the business case. They face the
same external challenges as Novices, including unclear regulatory benefits and
lack of tangible guidance for the industry. As companies in the pilot phase
have begun experimenting with QbD, they have started to realize other external
challenges, including inconsistency of treatment of QbD across FDA, and
misalignment of international regulatory bodies.
¶ Rollout. Companies in the rollout phase have solved, or are solving, most of
their internal problems in terms of implementation –– the majority of their
challenges are external. These challenges are focused on the lack of alignment
of regulatory bodies, we well as lack of clarity of regulatory benefits. They
 Confidential –– Page 14

also identified regulator’’s ability to handle QbD applications as a big challenge,

as well as the fact that the current interaction with companies is not conducive
to QbD.
¶ Fully implemented. Companies that are fully implemented face seemingly
more ‘‘advanced’’ problems. For example, they struggle with the question of
alignment with third parties such as suppliers and contract manufacturers.
They are also preoccupied with regulators’’ ability to handle QbD applications
and the sufficiency of the interaction with the FDA in order to foster QbD.

BUSINESS CASE FOR QBD

As part of this project the team gathered preliminary data and insights on building the
business case for QbD. The work is meant to provide a foundation on early answers and
a recommended path forward to continue to build and understand the business case.
Our analysis shows a strong preliminary business case for QbD. Interviews debunked
two commonly held beliefs surrounding the extra time and money required to execute
QbD. Additionally, preliminary modeling indicates huge potential long term value for the
industry as a whole, which could be as large as $30 billion incremental annual profit. Our
analysis also surfaced key factors which must be in place within companies, as well as
within the FDA, for this strong business case to hold true.

Debunking QbD myths

As a company thinks about executing QbD, many factors must be weighed and
considered. Two factors that commonly work against QbD are the fear that QbD is very
expensive and will drive costs up, and the fear that QbD will require a lot more time and
analysis. Our interviews provided evidence that the cost to implement QbD is in fact
minimal, and the increase in time, if at all, is negligible.
¶ The cost to implement QbD is minimal. Most people believe QbD leads to a
marginal increase during set up –– mostly from procedure development, a
change in human resources, as well as new analytical tools or knowledge
management capabilities (see additional exhibit 1 in appendix). After this
initial set up fee, most people feel there is no marginal cost, some interviewees
even claimed that QbD drives development costs down. Some polled claimed
that QbD has reduced their technical development costs by up to ~25% per
program.
¶ QbD will not increase the timeline for development. Interviews indicated
that QbD may add a negligible amount of time (~2 FTEs over 2 days) during
the early development phase, however there is no effect on time spent in critical
path, and time in tech transfer and scale up were often reduced. Although QbD
many cause minimal increases in time upfront, many companies experienced
time savings downstream.
 Confidential –– Page 15

Exhibit 4: Conceptual timeline showing QbD development vs. traditional

Clinical Pre-proof Post-proof

Filing
development of concept of concept

Traditional Tech Scale Launch

technical Early dev Late dev
Transfer up ready
development

Technical Tech Scale Launch

development Early dev Late dev
Transfer up ready
with QbD

+ 0 - 10% -10% - +10% - 10 - 0% - 10 - 0%

[Conceptual picture of three timelines, vertically stacked. Three timelines are clinical development,
traditional technical development, and technical development with QbD. Clinical development shows three
stages – pre-proof of concept, post – proof of concepts and filing. Traditional technical development is
below the clinical development and show 5 phases, Early development, Late development, tech transfer,
scale up, and launch ready. Below is the 3rd timeline for technical development with QbD. This shows the
same phases as the traditional technical development. The Early development phase shows a potential 0 –
10% increase in time for technical development with QbD. The late development phase shows a – 10% -
+10% difference in time for the technical development with QbD. Tech transfer and scale up both show a
potential 10% decrease in time for technical development with QbD.]

Potential benefits from QbD

Analysis identified many potential benefits from QbD –– some of which are quantifiable,
and some of which are not. In terms of quantifiable benefits, value comes from four main
areas –– a reduction of Cost of Goods Sold (COGS) and capital expense, increased
technical development productivity, improved quality (and lower risk), and increased
sales. These four combined could potentially provide $20 - $30 billion more profit to the
industry.
¶ Reduction of COGS and capital expense. QbD allows improved planning,
cycle time, yield and quality ––potentially driving down costs by as much as $15
–– $25 billion across the industry.
¶ Technical development productivity. Utilizing QbD techniques during the
product development process (e.g., platforming, taking advantage of growing
knowledge base) can lead to $4 –– $5 billion in savings.
¶ Improved quality and lower risk. Implementing QbD is really using ““good
science”” which leads to an overall better, more reliable product. Savings from
reduced risk of regulatory citation are estimated to be $0 - $2 billion.
¶ Increased sales. Products that utilize QbD have better launches and generally,
will have better product design leading to fewer stock-outs. This can lead to $0
- $4 billion in increase sales.
 Confidential –– Page 16

Exhibit 5: Potential sources of incremental profit from QbD 2

USD billions

Reduction of
COGS and
15-25
capital
expense

Tech Dev
4-5
productivity

Improved
quality –– 0-2
lower risk

Increased
0-4
sales

Total 24-35

[Build-down vertical waterfall chart showing potential sources of incremental profit from
QbD, as determined by McKinsey analysis. First bucket of potential incremental profit is
from reduction of COGS and capital expense, $15 - $25 B. The second bucket of
potential incremental profit is Tech Dev productivity, $4 - $5 B. The third bucket is
improved quality – lower risk, $0 - $2 B. The fourth bucket is increased sales, $0 - $4 B.
The fifth bucket is the total bucket, showing potential incremental profit of $24 - $35 B]

Additionally, QbD has potential to provide benefits that are more difficult to quantify.
Improved public image, standardized definitions within a company, and best practice
sharing were highlighted in interviews.
¶ Improved public image. Implementing QbD in development will lead to an
overall increase in quality of product. As the public learns more about QbD
and its positive implications for product quality, people will potentially trust
companies that are ‘‘QbD users’’ more so than those that are not.
¶ Standardized definitions. Several interviewees indicated an unanticipated,
but extremely valuable benefit in the standardization of definitions across their
companies. In implementing QbD, a company must ensure that the entire
operating model is aligned and that the same language is being used throughout
the company.
¶ Sharing best practices. One company commented that ““since our redesign
was global and we are all on the same system, everyone has access to the best
practices –– if experts in Sweden develop something, we can roll it out
immediately.”” Implementing QbD has enabled global best practice sharing.

2 McKinsey
 Confidential –– Page 17

Critical factors to enable business case

Our analysis indicates that in order for companies to experience the most business
benefits possible, two things must be in place. Companies must be aligned across the
entire operating model to support QbD, and the FDA must ensure a high quality of review
and deliver on the regulatory benefits many companies feel were ‘‘promised.’’
¶ Alignment across entire operating model. Companies that are seeing the
most benefits from QbD have aligned their operating systems to QbD in a
systematic and streamlined way. They have aligned across the three elements
of the operating model –– technical / processes, management system, and culture
and capabilities.
1. Technical Processes. Companies that are successful have the right
tools and processes to execute on QbD. They have standard
development processes built on QbD principles and may have
standardized equipment across development and commercial
manufacturing plants as well. They also have the data management
systems in place to really make QbD work. Additionally, they
participate in industry and regulatory thought groups.
2. Management system. Successful companies also have alignment to
QbD among leadership –– ““executive sponsorship is crucial to making
this work.”” Additionally, incentives must be aligned to support a
connected operating model (e.g., R&D developers have interest in
product throughout its lifecycle).
3. Culture and capabilities. The right talent with the appropriate
capabilities needs to be in place for QbD to work. This is about
ensuring a company has ““people with the smarts, motivation, and
sponsorship to drive it forward.”” This can be ensured through talent
acquisition, and capability and training programs. Beyond capabilities,
the company culture needs to support QbD. All of the companies who
are seeing tangible benefits believe QbD is ““a part of our culture.””

¶ High quality of reviews and delivery of regulatory benefits. In order to

allow companies to get value from the business case the FDA needs to ensure
reviews are consistent and smart. Additionally, there needs to be delivery upon
the ‘‘promised’’ regulatory benefits. Other potential steps the FDA can take to
catalyze QbD adoption will be covered in the next section.
 Confidential –– Page 18

Exhibit 6: Correlation between supporting mechanisms and level of adoption

9 implemented 9
Somewhat Implemented

Fully
Mechanisms to support QbD Novice Pilot Rollout implemented

9 9
Formal QbD pilot program / organization / special
project

9 9 9 9
Standard development processes built upon QbD
principles

9 9
QbD principles "built in" to our regular Regulatory
CMC processes

Stage-gate process for CMC program review

9 9 9
9 9
Incentive alignment amongst development &
manufacturing

Talent acquisition and management

9 9
Participation in industry / regulatory groups
9 9 9 9
Capability / training programs for personnel
9 9 9
Standardized equipment
9
[Chart listing 9 mechanisms to support QbD on the y-axis, and their alignment against stage of adoption
(i.e., Novice, Pilot, Rollout, Fully implemented) on the x-axis. A dotted check mark under the stage of
adoption and next to the mechanism indicates somewhat implemented, a full check mark under the stage of
adoption and next to the mechanism indicated implemented.
y Formal QbD pilot program / organization / special project – somewhat implemented for Pilot,
implemented for Rollout
y Standard development processes built upon QbD principles – somewhat implemented for Novice
and Pilot, implemented for Rollout and Fully implemented
y QbD principles "built in" to our regular Regulatory CMC processes – somewhat implemented
for Rollout, implemented for Fully implemented
y Stage-gate process for CMC program review – somewhat implemented for Pilot, implemented
for Rollout and Fully implemented
y Incentive alignment amongst development & manufacturing – implemented for Rollout and Fully
implemented
y Talent acquisition and management – implemented for Rollout and Fully implemented
y Participation in industry / regulatory groups – somewhat implemented for Novice, implemented
for Pilot, Rollout, and Fully implemented
y Capability / training programs for personnel – somewhat implemented for Pilot, implemented
for Rollout and Fully implemented
Standardized equipment – somewhat implemented for Rollout, implemented for Fully implemented]

Potential next steps to continue building business case

Moving forward, there are many things that can be done to continue to build and
substantiate the business case.
¶ Bring together industry leaders to share case examples –– encourage sharing of
real, specific numbers
¶ Build library of business impacts from pilot and other QbD applications
 Confidential –– Page 19

¶ Continue to invest in a deeper understanding of where value is coming from

and what must be in place to achieve this value
y This potentially could involve a project-by-project based analysis, building
an understanding of baseline development performance and how that is
different than when utilizing QbD (i.e., time and effort to develop and file
by phase, technical transfer and scale up, process outcomes in terms of
quality, cost and performance, one time costs)
y One could bolster this analysis with an examination of the various operating
systems and management infrastructures companies have in place to
determine which have the greatest impact on the business case

IMPLICATIONS FOR FDA

Our research has highlighted many challenges to the implementation of QbD. These
challenges indicate several areas that the FDA may consider addressing in order to speed
QbD adoption. The actions the FDA can take become more tangible when applied to the
segmentation of drug type and level of adoption.
Potential options to consider
We have identified several options the FDA can consider to enhance QbD adoption.
These options fall into several areas –– FDA policy, internal FDA change management,
and external change management.
¶ FDA Policy options
1. Define and codify incentives. Although the FDA may or may not
believe there are benefits to further defining and codifying incentives,
from our conversations with industry, it is clear many companies see
this as a powerful way to incentivize QbD adoption. We recommend
the FDA consider analysis to understand what such incentives could be
(e.g., a faster review process, easier to change processes later on), and
what benefits could come from adoption of these incentives.
2. Develop tangible guidance for QbD execution. Companies,
especially in early stages of adoption, have identified confusion around
what QbD means and how to actually execute as a huge challenge.
They strongly believe they could benefit from more tangible guidance.
Issues raised from the regulatory side against increased direction are
around a fear of creating guidance too early, while things are still
being ““figured out,”” as well as fear of being too prescriptive. Although
there are valid concerns around the dissemination of more direction, it
is a clear need within industry that the FDA is able to, and should
meet.
3. Mandate. Some drug areas, for example several areas within the
Generics industry (e.g., controlled and modified release drugs), have
had numerous safety issues. For drugs such as these, where there is a
public health risk due to lack of process understanding, mandating
 Confidential –– Page 20

QbD is a reasonable option. As long as ‘‘first to file’’ exists in the

Generics industry, there is a strong argument that unless the quality of
scientific understanding is raised for all applications, companies will
continue to ‘‘cut corners’’ to file as quickly as possible –– disadvantaging
those companies that seek to gain a deeper scientific understanding,
and potentially putting patient safety at risk.
¶ Internal FDA change management
1. Ensure consistency of review process in terms of scientific
knowledge and quality of review. The desire for a consistent review
process and well trained reviewers is unanimous throughout industry.
The FDA is aware of this need and has been working to train and
educate reviewers to prepare them to handle QbD applications. The
FDA is also exploring utilizing teams of reviewers to handle
applications rather than a single reviewer. The FDA should continue
to pursue consistency in the quality of its reviews.
2. Harmonize the approach to QbD across the FDA. There is a need to
harmonize QbD practices and requirements across FDA including OPS
and associated functions e.g. compliance and inspectors. The current
state where decisions around QbD are sometimes called into question
or even overturned when areas subsequent to OPS review are involved
has the chance of eroding or even stopping the momentum around
QbD adoption in industry. FDA should present a unified approach to
QbD regardless of the section of FDA that a company is interacting
with at a particular time.
However, opinions are split on whether or not the OPS should have a
uniform approach to QbD across its three offices. Since the offices
deal with such different products, it may be impossible, and potentially
harmful to try and assign a single approach. At the same time,
completely different approaches can be very confusing. The FDA
should establish one clear, broad definition of QbD to be utilized by
OPS. Each office can then tailor that broad definition with an
approach that is best suited to its drug type.
¶ External change management
1. Change the method of communication. Industry almost universally
asked for more frequent, ““no risk”” dialogs with the FDA. The
companies who participated in the pilot programs felt they benefited
from the increased, and often less formal communications. There are
feelings that ““no risk”” dialogs are impossible, as the FDA needs to
protect public safety and cannot afford to offer any immunity areas.
There is also a fear that anything said by the agency, even if in a
‘‘casual’’ conversation, could be taken as ‘‘official opinion.’’
Additionally, more meetings would put an even greater resource
constraint on the FDA. Despite the constraints for the FDA, there
needs to be a reassessment of the FDA’’s interaction with industry. A
system should be crafted that allows for more dialogs in a manner
comfortable to both FDA and industry.
 Confidential –– Page 21

2. Create more buy-in by disseminating case examples. Industry has

made a clear call for real, tangible examples of what the FDA has
actually approved or rejected and why. However, there are legal
barriers for the FDA in disclosing this information. The FDA can
work with companies who have had successful applications, to
develop more understandable examples, with which both sides
(industry and the FDA) are comfortable sharing publically.
Additionally, the FDA can continue to disseminate findings and tips on
successfully navigating challenges as they work with companies.
3. Improve international harmonization. There are questions around
whether or not the FDA is responsible for addressing international
harmonization of QbD acceptance. This is an area worth at least
examining as the FDA is currently one of the biggest proponents of
QbD and the lack of international harmonization is one of the biggest
challenges raised by companies. The FDA should continue to seek
opportunities to work with other regulatory bodies (e.g., joint reviews).
The FDA should work to ensure alignment on what QbD means,
what’’s required and how regulatory flexibility can be granted.
4. Utilize a 3rd party model as a means of catalyzing and
standardizing QbD within industry. There is a call for more
substantial direction from and interaction with the FDA. Clearly this
is putting a great resource strain on the agency. One way to alleviate
this strain would be to promote the use of a 3rd party model as a means
for further education and QbD development.
Implications for different drug types
¶ New Drugs: Rapidly drive adoption. New drugs manufacturers are the most
mature in terms of their adoption of QbD. There is potential for the FDA to
continue to push more New Drug companies to adopt QbD, utilizing learnings
from other companies’’ success (i.e., real case examples) to help facilitate
understanding and adoption. Although, we have not heard strong anecdotal
case for QbD changing the patient safety for New Drugs, it seems this is an
appropriate area to provide incentives for adoption.
¶ Generics: Consider mandate or policy change. In the generics arena there
have been a number of examples where rushing to file and a lack of process and
product understanding has led to serious patient safety issues. Although this
has yet to be codified –– if it was codified and accepted by the health community
–– there may be a good argument to make QbD a mandate for some types of
Generic filings. Even if full blown QbD is not made a mandate, there is
potential to make the process understanding required within applications more
detailed, and to make some of the ‘‘optional’’ parts of QBR mandatory.
¶ Biologics: Do not give up on QbD. Although many manufacturers site QbD
as being impossible, especially with upstream processes –– our interviews
indicate this is not true. Many biologics manufacturers are applying QbD to
downstream processes with great results. These manufacturers believe that
QbD is even more important for Biologics as the molecules are more complex,
a deeper understanding will lead to better product. There is no reason why
 Confidential –– Page 22

QbD should not apply to biologics. The FDA should continue to be patient,
disseminate success stories, and hold forums concerning QbD for Biologics
companies. Over time, the case for QbD for biologics will become increasingly
clear.

Implications for different levels of adoption

¶ Novice: Ensure understanding. The FDA should ensure Novice’’s understand
what QbD really means, and what the potential benefits are. There is potential
to distribute more case examples and ensure codification of benefits is clear.
There is also potential to invite novices to workshops or a ‘‘QbD Academy’’
where the basics are covered and the FDA can train ‘‘ambassadors’’ to bring a
more nuanced understanding of QbD back to their company.
¶ Pilot: Help with internal issues. As companies begin piloting QbD, they face
many internal issues, including internal misalignment, and limited knowledge
of QbD techniques. There is potential for the FDA to disseminate ‘‘best
practices’’ or a more tangible ‘‘how-to’’ guide. There is also potential to hold
seminars / workshops where more experienced QbD companies discuss the
challenges they faced in implementation and lessons learned. Additionally,
more casual interaction with companies in a collaborative manner will be
crucial as they develop QbD programs.
¶ Rollout: Work through external issues. Companies in the rollout phase have
usually begun to work out their internal issues, and are more focused on
external challenges. It will be crucial to improve the interaction with
companies to be more open. Companies at this phase of adoption are looking
for how to have a more collaborative role with the FDA in making the QbD
application process run smoothly. Additionally, it will be very important for
the reviews to be consistent and of a high quality.
¶ Fully implemented: Collaborate. There is potential for the FDA to work with
companies at a very mature state of adoption to build case studies and share
best practices. The FDA can facilitate industry leader forums where companies
debate some of the more complicated issues facing QbD, such as what is means
for suppliers and contract manufacturers.
 Confidential –– Page 23

APPENDIX

Comprehensive data mining efforts

¶ List of articles examined for this project
ƒ Avellanet, John, ““Why Quality by Design? An Executive’’s Guide to the FDA’’s
Quality by Design.”” Cerulean Associates, March, 2008.
ƒ Fuhr, Ted, Michele Holocomb and Paul Rutten. ““Why 'Quality by Design'
should be on the pharma CEO's agenda”” March 2009, McKinsey & Co.
Operations Extranet.
ƒ Ginsberg, Peter L., Sandeep Bhatia, and Rachel L. McMinn, ““The road ahead
for biologics manufacturing.”” January 2002.
ƒ McKinsey White Paper: ““Transforming to the Desired state of cGMPs for the
21st Century.””
ƒ Neway, Justin. ““How to Make the Business Case for Quality by Design.””
BioPharm, December, 2008.
ƒ Paskeit, Diane M., ““QbD and Parentals: Strategies for assessment of leachable
in parental drugs.” Contract Pharma, April, 2009,
[Link]
ƒ Pickett, Joseph, ““CMC pilot program succeeding in integrating concepts of
QbD.”” Inspection Monitor, February, 2007,
[Link]
ƒ Quality by Design is Essential in the New U.S. Regulatory Environment, Aegis
Analytical Corporation, [Link].
ƒ Rathore, Anurag. ““Roadmap for implementation of quality by design (QbD) for
biotechnology products.”” August, 2009.
ƒ Snee, Ronald D., Philipple Cini, Jason Kamm, and Chester Meyers, ““Quality
by Design: Shortening the Path to Acceptance.”” Pharmaceutical Processing.
Tunnell Consulting,
[Link]
&ISSUE=0802&RELTYPE=PR&ORIGRELTYPE=ATO&PRODCODE=000
0&PRODLETT=Y&CommonCount=0.
ƒ Snee, Ronald D., ““Quality by design: Four years and three myths later.””
Pharmaceutical processing, February 1, 2009,
[Link]
E~0902~RELTYPE~PR~ORIGRELTYPE~ATO~PRODCODE~4315~PROD
LETT~[Link].
 Confidential –– Page 24

ƒ Somma, Russ, ““How Quality by Design Is Changing Drug Development: Without

QbD thinking, sponsors’’ attempts to address FDA ““complete response”” letters may be
doomed.”” SommaTech Consulting, [Link]
ƒ Somma, Russ and Andre A. Signore. ““Embracing Quality by Design: Applying
concepts can help CMOs Create Value,”
[Link]
[Link].
ƒ Spurgeon, Tom. ““Quality by design in Solid Dosage Processes: How will QbD impact
manufacturing?”
[Link]
[Link].
¶ Conferences / Presentations
ƒ ““Implementation of Quality by Design in New Product Development,”” GSK, Peter
Watmough (senior development scientist), Laura Morris (process development
engineer).
ƒ Peri, Prasad, ““Quality by Design (QbD) Approaches for Orally inhaled and Nasal
Drug Products (OINDPs) in the USA.”” Office of New Drug Quality Assessment
(ONDQA), OPS, CDER. RDD Europe 2007.
ƒ PhARMA perspective conference: ““Potential barriers to QbD implementation.””
September, 2009.
ƒ ““Sandoz: Pioneering Global development of Biosimilars,”” Bio Super Session, June
17th, 2008, Freidrich Nachtmann, PhD. Head of Biotech Corporations, Sandoz
GmbH.
ƒ Somma, Russ. Using Quality by Design (QbD) in Designing Efficient, FDA
Compliant Pharmaceutical Manufacturing Processes and Facilities: What is the
Impact? SommaTech, LLC.
ƒ Winkle, Helen. ““Evolution of the Global Regulatory Environment: A Practical
Approach to Change,”” PDA / FDA Joint Regulatory Conference: Implementing
Quality by design. September, 2007.
ƒ Yu, Lawrence. ““Pharmaceutical Quality by Design: Regulatory Perspectives.””

Additional quotes supporting challenges

ƒ Internal Misalignment: Disconnect between leadership and middle management
–– ““People are at different stages within different functions in terms of figuring
out how to do things –– leadership agrees with the general mantra, but the
groups that need to execute don’’t understand how to deliver””
–– ““Middle management understands QbD –– but the lower or higher into the
organization you get there are misunderstandings””
–– We have pockets and silos of QbD going on but are not united overall””
ƒ Internal Misalignment: Misalignment horizontally across the organization
–– ““R&D is 100% incentivized on the quantity of filings they get””
 Confidential –– Page 25

–– ““The biggest challenge is getting alignment within a company to support ––

many players need to be involved for this to be executed successfully””
–– ““The groups that stand to get the most benefit from QbD are not the same that
bear the brunt of the (perceived) risks and costs””
–– ““There is no ‘‘credit’’ for helping out with these issues [understanding biologics
CQAs] and so the progress is slow””
ƒ Internal Misalignment: Culture of conservatism
–– ““We need to ensure everyone understands the paradigm shift –– today the best
CMC strategy is to not show all your cards””
–– ““Smart people get into a company and are trained to do things a certain way ––
they get stuck –– we have to open people’’s minds back up to understand QbD””
–– ““There is a resistance to eliminating QC resources””
–– ““People believe development is an art form and ask, ‘‘how can you make a
platform out of an art form?’’””
–– ““We have a history of playing by the old rules””
ƒ Internal Misalignment: The amount of change required within company is not
feasible
–– ““Adoption of QbD will require remodeling of the entire operating model –– it’’s
a huge change””
–– ““In order to get the benefits from this we would have to totally redesign our
entire development process –– QbD is not enough of a priority to do that””
–– ““For this to work it’’s more than just QbD –– you need lean alignment, 6 sigma,
and an organizational culture element””
–– ““We needed to restructure manufacturing and development to recognize the
multi-functionality necessity to support these initiatives””
–– ““Significant change is required at manufacturing site when you implement
QbD””
–– ““Although it’’s now implemented, this new radical approach [redesigning
development program] took a while to get an embed and get real acceptance””
ƒ Internal Misalignment: QbD remains low on the priority list
–– ““When a manager looks at the list of things he needs to do, QbD
implementation is not high up there””
–– ““Unless this is a mandate from above, there is no reason to do this before
something else””
ƒ Lack of belief in business case
–– ““Planning for a QbD filing requires a huge investment of time and effort early
in development, before you even know if the product will make it to launch””
–– ““There is a lot of uncertainty over timing of and investment requirements for
QbD implementation””
–– ““Re-usability of knowledge gained on one product in support of another is
uncertain””
–– ““I’’m not sure QbD will have any benefits that will truly change the safety /
efficacy of a drug””
–– ““We have some drugs for which the increased flexibility would not be helpful””
–– ““We need to build out scenic understanding –– this requires many resources,
many people, many hours, and many expensive studies””
–– ““This costs more because it really is around making more and more batches ––
amount of analytical testing””
ƒ Lack of technology to execute: Insufficient solution for controlling variability of
raw materials
 Confidential –– Page 26

–– ““QbD does not address raw materials –– that’’s the biggest cause of variability
for us””
–– ““A robust raw material management system needs to be in place to facilitate
QbD implementation…… Critical raw materials significantly impact product
quality and should therefore be thoroughly characterized”” 3
–– ““We need a better understanding of excipients, what are their properties?
Their functionalities?””
ƒ Lack of technology to execute: Limited understanding of implications of quality
attributes
–– ““You will never know enough about your materials and process to be able to
change as you go through””
–– ““This is a really big problem for large molecules, we are just not very clear
about what aspects are doing what””
–– ““We just don’’t understand the biological significance of CQAs –– there is
misplaced confidence in being able to figure this out””
–– ““Vaccines will be an issue, we have always built consistency into the process
without knowledge of what was going on, we need better technology to get
there””
ƒ Lack of technology to execute: Knowledge of QbD techniques limited
–– ““In order to make some of the new types of measurements we will need to
develop new skill sets and get new technology””
–– ““Developing the technical tools and standardized processes necessary was a
big challenge””
ƒ Lack of technology to execute: Difficulty with data management
–– ““In order to make some of the new types of measurements we will need to
develop new skill sets and get new technology””
–– ““There is a challenge monitoring a product in late development to understand
how a process is playing out since there is such a huge volume of information””
–– ““We will need more statisticians with different skill sets and more education to
be able to handle all this new data””
–– ““For older products we are having difficulty with knowledge management, the
data could be stuck on someone’’s hard drive somewhere””
ƒ Alignment with 3rd parties
–– ““It will be difficult figuring out the best way to provide guidance for outside
partners –– heavily engineer vs. brute force …… it will be more complicated to
bring them along since all resources are not under complete control””
–– ““QbD is a space to redefine relationships with suppliers –– hopefully it will
give more freedom in terms of changing suppliers””
–– ““Suppliers come more into play as we think about sources of variability””
ƒ Inconsistency of treatment of QbD across FDA
–– ““The FDA is not aligned, the only thing they are aligned on is where you file””
–– ““Although a number of people in the FDA are supportive of QbD –– this is not
consistent””
–– ““ONDQA has a very philosophical definition based on scientific proof ““ 4

3 Rathore, ““Roadmap””

4 McKinsey White Paper: Transforming to the Desired state of cGMPs for the 21st Century
 Confidential –– Page 27

––
““Office of Generic Drugs has taken a more pragmatic approach and provided
an explicit, required process for submitting via Question based Review
(QbR)”” 5
–– ““We had trouble with a recent drug where we put all the appropriate controls
in place, passing our own bio-equivalency tests, then, the bio-equivalence
department asked us to change our method since that was not what they
typically see””
–– ““We got all the way to the end and then the biopharma side overturned what
the reviewers had already approved. This was a major blow.””
–– ““There are many different opinions within the FDA on how to apply QbD to
solution, we came to an understanding with chem group that was overturned
by biopharm group at the end””
–– ““There is a big disparity between center and field. We worked with the center
on QbD filing, then in an inspection the inspector didn’’t know what QbD
meant””
ƒ Lack of tangible guidance for industry
–– ““The QbD lingo is cutting edge and sexy –– but there is very little information
out there about how to actually do it on the ground””
–– ““FDA has not explained what is expected from a QbD file that has to be
proven clinically””
–– ““We understand what you are asking for broadly, but there are hundreds of
variables –– there’’s got to be an end in mind –– a tangible one we can deliver
on””
–– ““The path the FDA would like us to take to get here is not clear””
–– ““We know DOE is not sufficient anymore –– we need more but it’’s unclear
how to invest””
–– ““We don’’t trust that the FDA will actually accept this filing in a consistent
manner””
ƒ Regulators not prepared to handle QbD applications
–– ““The FDA needs to increase the depth of expertise of reviewers and inspectors
in the various approaches to quality by design, for example chemometrics,
engineering and modeling””
–– ““High turnover seems to be an especially big problem””
–– ““Reviewers who come out of industry fall back to what their organization used
to do””
–– ““Dealing with reviewers who don’’t fully support this or understand this really
slows us down a lot””
–– ““How can we be comfortable that the recipient of our file will understand?””
–– ““You never know what feedback you are going to get these days, it’’s a very
inconsistent interface””
–– ““There is a disconnect between compliance and field groups, we often get
different feedback””
–– ““FDA is not leveraging the data we are giving –– reviewers are green and don’’t
understand the processes””
–– ““Regulators have different expectations and ask different questions for smaller
companies than bigger companies””

5 McKinsey White Paper: Transforming to the Desired state of cGMPs for the 21st Century
 Confidential –– Page 28

–– ““Not everyone at FDA understands, they need more education and

development of CMC staff…… I’’ve seen conflicts within the FDA –– especially
with post approval, once you get approval –– gap in terms of level of
understanding””
ƒ The way promised regulatory benefits is currently being shared does not inspire
confidence
–– ““There is a questionable payback –– it is unclear how much regulatory
flexibility will actually be given””
–– ““The lack of clarity makes it harder for us internally and increases skepticism
and reliance on the internal business case””
–– ““FDA has still not shown us the flexibility we can get from this””
–– ““At the end of the day it is still unclear whether the FDA will actually back
these filings””
–– ““It is a challenge to understand what the agency actually wants, and what sort
of flexibility it will provide””
–– ““People are wary –– is this really going to come to pass?””
–– ““What we really need are for the regulatory benefits and flexibility that have
been promised to be codified, delivered and then realized””
–– ““We started this because we were promised regulatory benefits –– not we can’’t
even talk about them –– its’’ like they have amnesia””
–– ““We do need more clarity on short term apparent return on investment vs. real
ROI””
–– ““FDA needs to show business benefits””
–– ““There is a lack of clear regulatory incentives; without it you are giving an
advantage to someone who just seeks approval at the minimalist cost and
hurting companies who take a holistic approach””
ƒ Misalignment of international regulatory bodies
–– ““Duplication of industry resources to satisfy international markets that are not
willing to accept the content of a QbD filing submitted to the FDA””
–– ““Lack of clarity on how QbD and associated flexible regulatory approaches
will be received by countries outside of ICH””
–– ““Beyond FDA –– who else will accept this filing?””
–– ““Since there is no global harmonization, why move away from traditional
filing?””
–– ““It is much more difficult in secondary markets –– it takes longer to review
because they ask a lot of questions several times””
–– ““We are very concerned about the response to this outside the US and Europe.
We sell products in 75 countries, most are not a part of the ICH””
–– ““This is an FDA initiative, the rest of the world is watching and waiting””
–– ““We managed a joint FDA –– EMEA audit –– there were too many people with
differing opinions and in different places with QbD””
–– ““You have the Japanese saying you are doing QbD so we want this stuff, you
have Europeans saying we want this other stuff””
–– ““Japan’’s structure of submission documents, is so that you can’’t apply for
flexibility on formulation””
ƒ Current interaction with companies is not conducive to QbD
–– ““There needs to be a different way to interact, may even be joint teams doing
parallel reviews together””
–– ““Historically, there is not a lot of comfort talking with the FDA””
 Confidential –– Page 29

–– ““This needs to be more of a collaborative process –– it should be a mutual

education activity””
–– ““Pilots asked a lot of tough questions, there were several difficult interactions.
When we are treated with suspicion, it does not feel like collaboration””
–– ““There is a feeling that the FDA has yet to act on any feedback that the
industry has been giving””
–– ““It would be helpful if we could make changes without writing a PhD thesis
every time””

Additional Exhibits
Additional Exhibit 1: Correlation between supporting mechanisms and level of
adoption
Projected total cost to
Cost drivers implement

ƒ Process / procedure development $100 –– $200 K

ƒ Human resources (e.g., new human $0 - $200 K

resources, trainings)

ƒ Infrastructure (e.g., analytical tools) < $200 K

ƒ Knowledge management $0 - $200K

capabilities

Total < $1 MM