But before jettisoning manufacturing, companies would be wise to understand its strategic contributions, including the ability
to respond to rapid increases in demand, thereby avoiding costly stock-outs; opportunities for close partnerships with product
development to ensure smooth, rapid product launches and create barriers to entry for generics firms; and the generation of
substantial tax benefits due to operating in tax-advantaged locations. Strategic use of third parties can provide access to
advanced technologies, improve flexibility, and reduce cost. While it is an easy decision, outsourcing is no easy answer.
Forecast Accuracy for Drugs
Responding to Demand Spikes
Predicting approval and commercial demand for a new product is particularly difficult. Unlike in other industries, pharma
R&D efforts can look extremely promising for years only to uncover late-stage side effects or be hit with an FDA "approvable
letter" requiring additional years-long studies. Vanlev (omapatrilat) is an unfortunate example of how negative safety signals
can ruin the best laid plans. Bristol-Myers Squibb spent $500 million on a chemical plant in Ireland before FDA's concerns
about the cardio drug led to the program's cancellation in 2002.
On the other hand, a product can become an unexpected success. Lipitor (atorvastatin) is the ultimate example of a drug selling
beyond anyone's wildest expectations. Launched in 1997 as a "me too" drug, the fifth entrant into the statin category, first
year sales of Lipitor were expected to be less than $100 million. So Warner-Lambert's manufacturing executives were taken
completely off-guard when what they thought would be a three-month supply vanished within days. They rushed to purchase a
plant in Ireland to keep up with demand as sales zoomed to nearly $1 billion in the first year and over $2 billion in 1998.
Sales reached a high of nearly $13 billion in 2006—30 percent more than analysts had predicted two years after launch.
Compounding the problem of inaccurate forecasting is the significant lead time required to add new capacity. It can take anywhere
from two years (for a formulation facility) to five years (for a biologics manufacturing shop) to come online. Starting production
of a drug in an existing facility also takes time—a year or more to make saleable product, including the time for qualification
and stability batches. Maintaining excess capacity is one way to ensure against escalating demand.
Naturally, third party outsourcers are loath to maintain spare capacity without significant up-front payments. Even for specialized
capacity, it is often more economical to have a third party custom-build capacity rather than building internally.
Partnering With R&D
One advantage of internal manufacturing is the ability to work closely with R&D on product and technology development. There
is an opportunity to develop common technology platforms so that new products can fit seamlessly into existing capabilities.
In that way, one product failure does not render a manufacturing asset unusable.
GlaxoSmithKline has adopted this approach at its Cork, Ireland, shop with an R&D pilot facility and a manufacturing facility
on the same site. Similar technologies are used at both facilities, supporting rapid scale-up and launch. Technical staffs
are rotated frequently between R&D and manufacturing, enabling rapid transfer of know-how. Another example is Merck's construction
of a development-and-launch platform for product formulation at its Ballydine, Ireland, facility. During the development stage,
formulation scientists use a standard technology, such as roller compaction, which allows smooth transfer to manufacturing.
Achieving these benefits with third party manufacturers is difficult. Contract manufacturers work with many customers, making
it difficult to synchronize technology with any one. In addition, they may not be at or even near the cutting-edge of technology.
For example, few contract manufacturers have invested in Process Analytical Technology (PAT), which can cause delays and additional
costs in product transfer.