Quality by Design
Recognizing the Value
The twenty-first century is shaping up as a difficult time for pharma. Patent expirations,
thin pipelines, soaring manufacturing costs, and downward pressure on prices are
brewing a perfect storm for many companies. While QbD may not be a panacea, the
improved process understanding and more robust processes it promises can translate into
significant business benefits, including:
Lower cost of quality:
Using the generally accepted figure of 25% of sales as the current cost of quality, the top
10 pharmaceutical companies spend, on average, $7.6 billion annually on quality (Kamm
and Cini, 2007). Because quality is produced through extensive control the result is high
cost. QbD, however, enables science-based understanding of processes, so that
manufacturers can focus their control efforts on those factors that are critical to quality.
Further, greater process understanding also enables more accurate and thorough
validation than is now possible through the three-batch standard. Greater process
understanding also means more robust processes that can accommodate the inevitable
variations in raw materials that occur over time.
Better allocation of resources:
With greater confidence in the ability to maintain in-specification operations, companies
can free resources for more productive investment. Reduced manufacturing costs: As
QbD informs more and more processes, with greater control and greater robustness,
immediate bottom-line benefits accrue from improved yield, increased equipment uptime
and plant and capacity utilization, capital cost avoidance, and reduced rework and fewer
rejected batches.
Greater speed to market:
By maximizing the probability that a product in development will make it smoothly and
effectively through scale-up, technology transfer, and validation, QbD can greatly reduce
time to market and speed up return on investment. In terms of revenue alone, every day
that a blockbuster drug (defined as having annual sales of $1 billion) is delayed in getting
to market, its manufacturer forgoes more than $2.7 million in lost or deferred revenue.
Reduced regulatory burden:
Because QbD enables the manufacturer to understand the design space, the
manufacturing processes within that design space can be continuously improved without
further regulatory review. The manufacturer gains more regulatory room in which to
operate and the FDA can be more flexible in its approach, using, for example, risk-based
approaches to reviews and inspections.
Understanding Trends
In addition to the business benefits driving acceptance of QbD, some other broad trends
are also likely to give it added impetus. For example, medicines and therapies have
become far more complex since the days when medications like antibiotics were taken
briefly to treat acute and relatively straightforward conditions. Today whole new classes
of drugs have appeared for chronic conditions. Further, with many of the easy therapeutic
targets having been hit, pharmaceutical companies now often focus on far more
complicated therapeutic areas like oncology, AIDS, and Parkinson's, requiring more
complex medicines. QbD, with its ability to scientifically establish the complex multi-
dimensional combination and interactions of input variables and process parameters that
determine the quality of a product, works particularly well in complex contexts.
In biotech, for example, where testing a product is far more complex than with small
molecules and where characterization of the final product is less developed and
understood for biologics such as protein therapeutics and vaccines, QbD offers great
promise. In biotech, the process is, in effect, the product; and because high levels of
variation are often seen in biological processes, developing robust and reliable processes
is inherently difficult. QbD, however, frees biotech companies to focus more on
analytical tools to understand and control process development and manufacturing, which
in turn leads to more information about the product (e.g. secondary protein structures,
glycosylation patterns, etc.).
Further, because the biologic end product is often of high-value, biotech companies often
incur greater manufacturing risk than is typical for small molecule pharmaceutical
manufacturers. Biotechs can reduce this risk by systematically applying these QbD and
PAT principles: * Define acceptable limits for the critical-to-quality attributes of the
product. * Identify the primary sources of variability in those attributes in the
manufacturing process. * Identify which of the sources of variability can be monitored
and adjusted for better control during fermentation/cell culture, active agent recovery
(cell separation, product extraction, downstream product purification), and filling,
lyophilization, and final packaging.
In other words, biotechs can use QbD to define the design space and in conjunction with
the tools of process analytical technology (PAT) keep the process within that space.
For example, the fermentation process must be monitored and controlled to maintain
optimal cell growth conditions and predictable, reproducible product production with
consistent, well understood impurity profiles. Modern approaches use, for example,
automated spectrometers with probes inside the fermentor(s) to measure growth kinetics
and substrate consumption. These are coupled with automated real-time nutrient and gas
delivery systems triggered in response to the collected data. Mass spectrometers are
typically employed to monitor and control gas streams while FT-NIR spectrometers are
used to measure product concentration, nutrient concentration and biomass. These
techniques, coupled with the measurement and control of input parameters such as
temperature, pH, impeller speeds and gas rates, aid the process optimization and control
requirements to ensure robustness in which the output of the process remains insensitive
to variations.
QbD is also likely to gain added momentum as the pharmaceutical industry continues to
globalize, thus giving added urgency to goals of harmonization embodied in ICH Q8
(2005), which lays out the framework of QbD and suggests adoption by the regulatory
bodies of the European Union, Japan and USA. Further, as more and more companies
adopt QbD and it increasingly becomes a prominent component of new drug applications
(NDAs), it will reach a tipping point at which both companies and the FDA will be fully
committed to the practice, allaying fears that the agency might stop short of fully
implementing it.
Overcoming Barriers to Adoption
Despite the value that QbD promises and the trends that are likely to give it further
momentum, significant barriers stand in the way. Those barriers and the means to
overcome them include:
Technical - it won't work here.
Statistically-based improvement methods like SQC, SPC, Six Sigma, and Lean have been
demonstrably effective in improving process performance in many industries (Snee and
Hoerl 2003, 2005). Moreover, in recent years, the power of these statistical methods has
been dramatically supplemented by a new breed of widely available, easy-to-use
statistical software that puts the ability to do sophisticated calculations at virtually
anyone's fingertips. Nevertheless, because such key statistical tools as design of
experiments (DoE) originated elsewhere, some people don't believe they will work in the
pharmaceutical industry. In fact, DoE, which has been used in the chemicals industry
since the 1950s, is perfectly suited for performing the kind of multi-variate analysis
required to uncover design space and reap the operational, regulatory and business
benefits such knowledge offers.
For example, a new solid-dose, 24-hour controlled-release product for pain management
had been approved but not yet validated because it had encountered wide variations in its
dissolution rate, which presented issues of safety and efficacy. The manufacturer did not
know whether the dissolution problems were related to the active pharmaceutical
ingredient (API), the excipient, or to variables in the manufacturing process - or to some
combination of these factors. Frustrated with the results of one-factor analysis and seeing
an opportunity to take advantage of the power of designed experiments, the manufacturer
narrowed the range of possible causes of the unacceptable dissolution rate to nine
potential variables - four properties of the raw material and five process variables such as
temperature, feed rate, and screw speed. From this technologic space - the possible
combinations of variables most likely to affect the dissolution rate for better or for worse
- the team used a DoE to screen out irrelevant variables and to find the proper values for
critical variables, thus accomplishing screening and optimization in a single step (Kamm
2007).
The analysis showed that one process variable exerted the greatest influence on
dissolution and that other process and raw material variables and their interactions also
played a key role. The company was then able to determine the design space: the various
permutations of the settings for the all of these variables that still result in an in-
specification rate of dissolution (and other product properties). They then used advanced
statistical modeling software to get a clear picture of that design space in a "Contour
Profiler Matrix Plot" (Figure 1) created using the optimum settings for each of the two
significant raw material (RM) variables and four significant process variables (PV). The
X and Y axes are made up of the DoE variables, and the Z axis (the contour curves)
represents dissolution (the response variable). In the red regions, dissolution is out of
specification and in the green regions - the design space - it is within specification.
Keeping the process operating "in the green" by using the flexible parameters that were
optimized in the course of the DoE study, which is precisely the kind of approach
envisioned in ICH Q8, the company successfully validated and launched the product.
Financial - we can't afford it.
As with any major change in the approach to the development of products and processes,
many companies worry that the cost will simply be too high. In fact, the cost of
implementing QbD is almost negligible, especially when measured against the return on
investment. Nevertheless, some large companies may view wide adoption of QbD as too
expensive because they have extensive resources sunk in the old methods. Meanwhile,
many small biotech companies, focused on getting to clinical trials as soon as possible,
give other priorities a backseat and hesitate to invest scarce resources in new
methodologies like QbD.
The key to overcoming this obstacle is to rigorously translate the cost of QbD into its
financial impact. Consider, also, the opportunity costs of not pursuing QbD. Until the
manufacturer of the pain management drug executed the DoE study, the dissolution
problem had delayed the launch of the product for several years. QbD can help deliver
the kind of speed to market that still constitutes a real competitive advantage in the
industry.
Psychological - it's too painful to change.
In some ways, this is the most formidable obstacle of all and yet it is the least concrete.
Change inevitably means loss - of familiar ways of working, of comfort, of stability - and
such loss is painful. And there is no question that the more holistic way of working
required by QbD entails great change. From the QC and analytical labs to process
development to manufacturing and regulatory submission and compliance, people across
the enterprise will have to work together with the concept of the design space as the
framework for their common efforts. Contract manufacturing organizations will also have
to be able to handle QbD or their more advanced customers will look elsewhere for
partners.
In the face of such sweeping change, many people may understandably resist. In fact,
change is so painful that technical and financial objections to change are often really
masks for what is at bottom fear. Overcoming it means examining those objections
rigorously and honestly and disentangling them from psychological motives. Further,
there are proven techniques of change management that not only efficiently implement
improvement initiatives like QbD but also help negotiate the tricky psychic terrain that
change brings. Organizations that hesitate to confront change may find themselves forced
to do so anyway when industry conditions make the status quo more painful than the
alternatives.
.
