Shire's $210M Single-Use Gamble

By Rob Wright

December 2013 issue of Life Science Leader magazine

In 2010, Shire (NASDAQ: SHPG) completed Project Atlas — a biologics manufacturing facility (building 400) on its Lexington, MA campus. At 200,000 gross square feet (GSF) and a total direct cost of $210 million, the facility represents a sizable investment. However, what makes this facility different from most pharmaceutical manufacturing buildings is that it takes significant advantage of single-use systems (SUS). In fact, its entire upstream line utilizes SUS technology. This story is made more interesting when you consider that at the time when Project Atlas was conceived (2007), the company’s annual sales revenues were $2.4 billion, with profit after expenses of $82.4 million — $127.6 million shy of the plant’s $210 million dollar price tag.

But what really made this a bit of a gamble was that when the plant was designed and construction began, it was with the knowledge that the technology necessary to operate the facility, namely a 2,000L SUS commercially available bioreactor, did not exist! “The closest thing we had was a prototype by one of the vendors,” says Bill Ciambrone, executive sponsor of the project. The EVP of global technical operations and 20-year industry veteran, Ciambrone recounts Shire’s decision-making approach, the process of gaining buy-in from the team and regulatory bodies, and the various risk mitigation strategies employed.

Necessity — The Mother Of Decision Making
In 2005 Ciambrone was working for Transkaryotic Therapies (TKT), which was acquired that year by Shire. He says even back then while he was at TKT the need for a Project Atlas-like facility was evident. “We were a small biotech focused on orphan diseases,” he explains. “We knew we would be dealing with mammalian cell cultures, operating largely in perfusion, but there was uncertainty about which products in development would successfully come forward.” This uncertainty about which product would get FDA approval proved to be a driving factor for the type of plant to build at Shire. “How do you best deal with uncertainty?” he asks rhetorically. “It is hard to deal with uncertainty if you build a bunch of hard-pipe reactors of a certain size when you aren’t sure which products will be approved and will require commercial manufacturing capabilities. If you do create a hard-pipe configuration, you may be operating suboptimally with higher costs than the product really needs.” Because Shire was faced not with the prospect of manufacturing blockbuster products for millions of patients but with specialty drugs, the need for increased manufacturing capacity was superseded by an even bigger need — manufacturing flexibility. “The flexibility we needed was for producing multiple products, possibly even multiple scales, but all within a small scale,” Ciambrone explains. In addition, the company had a tremendous need for getting the plant up and running quickly. It was even more important to do so considering Shire’s competitor, Genzyme, was experiencing manufacturing deficiencies at its Allston, MA plant. An FDA warning letter in March 2009 led Genzyme to shut down the plant just three months later — resulting in a severe drug shortage for two life-sustaining, rare-disease medicines (Cerezyme and Fabrazyme). In addition, the Cerezyme shortage resulted in Shire getting FDA fast-track designation for its experimental Gaucher treatment, VPRIV (velaglucerase alfa for injection). “We thought we had an accelerated schedule of bringing a plant up and running by going with SUS,” he states. “The competitor’s supply crisis resulted in our having to accelerate everything even faster.”

When the Shire team looked at the various approaches to building a new plant, the positives of single-use technology were significant. “There were almost no negative points except for the uncertainty of the technology not yet being in existence,” he affirms. “If you believed in what the technology promised, you could build ‘a box’ [a building with an open floor plan] that could support perfusion over a variety of scales. And then you could decide which reactors to install later without changing the speed at which the plant could be built.” To support the decision-making process of building such a flexible plant, Ciambrone’s team conducted an algorithm analysis of a number of potential product-mix scenarios. “The algorithm showed the impossibility of the solution,” he explains. “There was no hard-and-fast outcome saying, ‘install two 2,000s and two 500s, and those reactors will cover enough capacity.’” Further, none of the configurations addressed the speed question. For example, when you build a traditional plant, install utilities, reactors, and equipment, only then can you begin the validation process. “You really cannot learn anything about your reactors until they are installed,” says Ciambrone.

That conclusion led Shire to a unique proposition for its bioreactor vendor, Xcellerex. To prove that the solution could solve Shire’s speed concerns, the pharma company wanted Xcellerex to provide a place where meaningful work could be done on the reactors by Shire staff along with technical support from Xcellerex. The resulting idea became known as the Atlas “Sandbox” Process Line Mockup, which eliminated a number of costly errors and inefficiencies from the finished process line. For example, during tests with Shire at larger scales (1,000L, 2,000L) and long periods of time, it was discovered that the mixing at 2000L was inadequate. Some design changes were needed, including a more robust impeller bearing for the bottom-mount agitator to provide adequate mixing. Had this deficiency not been detected in advance, the bioreactor would not have functioned properly — resulting in costly delays to getting the plant up and running. According to Ciambrone, once Shire and Xcellerex began generating data together in the Sandbox, it became obvious there would need to be some design changes to the bioreactors before the company could take delivery. Furthermore, this testing environment gave the two companies experience with making a lot of sterile connections because the Sandbox was not ISO 7 or ISO 8 certified. In addition, the Sandbox familiarized operators with the systems long before they went live at the plant and facilitated the start of early, offsite commissioning and qualification (C&Q) and SOP development.

Being able to create the Sandbox was the clincher for Ciambrone to choose SUS for the Atlas Project. But when you are boldly going where no one has gone before, you need to gain buy-in not only from your internal team but from regulatory agencies as well.

Don’t Fear Failure — Hate It
Ciambrone reports directly to the president of the company. So, he only had to convince one person above him once he decided to build an SUS plant. However, the process of gaining buy-in from Shire employees below him, especially those intimately involved and actually helping to make the decision, proved to be a bit more difficult. “It was a matter of freeing them from the fear of failure,” he explains. Ciambrone’s approach to removing the fear of failure first involved communicating that he, and only he, would be accountable for the decision to build an SUS plant if things didn’t go well. “If you stop people from thinking their heads will be on the chopping block, you create excitement,” he affirms. In addition, he pointed out that the company not only had the opportunity to do something very different, but also that there were good reasons supporting the decision (e.g. lower costs [as compared to a stainless-steel facility], faster setup time, and the flexibility to change the plant for manufacturing other products in the future). “We cannot let fear push us in another direction, especially if we have a contingency plan,” he explains. Having a contingency plan was another key to gaining buy-in from the team. Though the original plan required speccing for SUS, the Shire team also specced a hard-pipe solution. This allowed Ciambrone to communicate that if things didn’t go according to plan, the company could still remediate back to a hard-pipe solution with minimal business impact. “It was a matter of letting people know there were a lot of good technical and business reasons to do it this way, but fear of failure wasn’t one of them.” Ciambrone’s corollary to fear of failure is, “You have to hate failure. This makes you plan more and do everything you can so you don’t fail.” He says fear of failure leads to bad decision making and less risk taking. “Too many people, when setting out on a risky endeavor, feel they have an easy out and are quick to place blame on the decision as being the reason for failure. I utterly reject that,” he states. “The decision and risk didn’t cause you to fail. It was the fear of failure.”  Ciambrone ascribes to recognizing and understanding the risks a decision entails, and to plan accordingly.

Just as Shire employees needed to be alleviated of the fear of failure, regulatory agencies needed a thorough understanding to feel comfortable approving a new manufacturing process. When doing something new, Ciambrone advises working with regulators proactively. “We met with the FDA and EMA (European Medicines Agency) in the middle of designing the plant,” he relates. “Some of the feedback we received was expected, such as challenging us to make sure we had the right protocols in place for leechables, extractables, and other necessary controls.” According to Ciambrone, because the company already had been using the same material for storage and buffer, the team felt confident about the protocols in place. What surprised Ciambrone, however, was feedback from the EMA. “They loved the idea,” he says. “They thought we were actually doing what should always have been done in terms of contamination control. They liked the idea of getting rid of sterilization and validation and using disposable technology to prevent cross-contamination and carry over.” Ciambrone admits that working with regulators was the biggest challenge behind designing the plant and helping the providers of the reactor to develop their prototype into a commercial product. “We needed to demonstrate to the FDA that what we were planning not only made sense, but also that we were able to communicate the technology and due diligence behind the decision with regulators for whom this would be new,” he states.

Key Factors Of Shire’s Success
Shire broke ground on the new facility and got its first regulatory agency drug approval within three years. “No one had ever done this before,” Ciambrone claims. “The first approval was based on the purification site with the approval of the bioreactors coming later. But the plant would not have been running if we had used the old paradigm of plant building.” Shire deemed Project Atlas a success, and so did the industry. In 2011, ISPE recognized the pharma company with an honorable mention in its annual “Facility of the Year” awards program. For Ciambrone, one of the keys to the project’s success was effective project governance. “The owners of the project were the decision makers,” he affirms. “We didn’t spend weeks and months languishing for decisions.” Ciambrone admits this may be the benefit of being at a fairly small company. But does it have to be? Leadership governs best when it governs least. Ciambrone subscribes to removing as many layers of bureaucracy as possible. “You cannot have a process by which things have to go through 15 levels in the organization before they get blessed by someone at the top,” he explains. “It has got to be quick. If you are asking the team to move really fast, you need to support the team with quick decision making.” Ciambrone suggests if you are managing a similar project, put in place effective governance, with an empowered team and ready access to decision makers.

Since the completion of Project Atlas in 2010, which came in five months ahead of schedule and $10 million under budget, product sales revenues for Shire have grown by $1.2 billion. The gamble seems to have paid off. For example, the SUS design resulted in a 38 percent reduction in total facility size and approximately a $50 million reduction in initial capital costs. In addition, the facility uses 87 percent less water, consumes 30 percent less energy, and has resulted in a 95 percent reduction in caustic cleaning chemicals used. Combine these metrics with the fact that the facility has 26 percent less carbon emissions, and you can understand why the Lexington facility stands as a shining example of what it takes to be U.S. Green Building Council LEED (Leadership in Energy & Environmental Design) Certified. Shire seems to be playing its cards right, parlaying its business to 2012 year-end total revenues of $4.68 billion — vaulting the company into the high-roller ranks of Top 50 Big Pharma. Though the folks at Shire are proud of being first, Ciambrone cautions you not to let the perception of being innovative to be the motivation behind any decision, but to be based in logic and the facts particular to your company.

Why Not Simply Outsource Manufacturing To A CMO?
Given the rise in popularity of outsourcing pharma manufacturing, it would be natural to ask Bill Ciambrone, EVP of global technical operations at Shire, if using a CMO was considered instead of building its cell-culture manufacturing facility in Lexington, MA.  “I don’t think there was a single CMO that would have fit the bill because of the small volume of multiple products,” he states. “It is very costly to partner with a CMO when you cannot promise exactly what is coming.” Ciambrone notes prior to the recent formation of Gallus BioPharmaceuticals, there wasn’t a CMO with the SUS capability he was seeking. “We would have had to invest a lot with any CMO, and economically it would not have made sense,” he states. “They get different margins on small volumes. With perfusion, you are occupying whatever space they have for a long period of time. No one was able to cost effectively meet the demands our product lines required.” Even though it could not find one CMO with the actual capacity for the time period required, Shire conducted a detailed financial analysis to assess outsourcing as an option to further support the decision to build. “Even with the investment of building a $210-million plant, we were able to more cost effectively produce the range of products we needed on a cost per liter, cost per gram basis when compared to CMOs.”

Another option considered involved developing a strategic partnership with a CMO. “When we looked at potential partners, they already had a lot of clients occupying space,” he states. “In some cases the best deal we could get was to build something for them. This was how we decided having control justified the expansion and expenditure.” According to Ciambrone, when it came to trying to broker a strategic partnership, there was not enough willingness for an equitable sharing of costs and risks on the part of CMOs to make a strategic partnership economically feasible. In the end, the need for flexibility became the ultimate deciding factor. “Because the CMO is not the client, you never really have complete flexibility when outsourcing,” he notes.

What You Need To Know If Thinking About Adopting Single-Use
“Everyone who sells disposable reactors will tell you your capital cost will be a lot less,” says Shire’s Bill Ciambrone. “You don’t have to CIP (clean in place) or SIP (steam in place). You can do SUS in a closet without the air handling.” According to the Ciambrone, EVP of global technical operations, when Shire completed the design for Project Atlas, it found going with SUS did save some of the capital costs. “But it was not as dramatic as people would have liked you to believe,” he informs. “Operating expenses go down in some areas but go up in others because consumables are not cheap.” Ciambrone advises to be sure to take this into account when considering capacity expansion of traditional (i.e. stainless-steel) versus SUS. For example, with a traditional plant, once the piping is in place, it is a sunk cost with some variable operating and maintenance costs. However, with SUS you will be purchasing a greater number of consumables (e.g piping, tubing, bags) every time you use an SUS bioreactor, which will increase your variable operating costs. “The real economic savings which appealed to us were the speed with which we could get the plant up and running by taking advantage of the Sandbox (see main article for explanation) and being able to do some qualification on the bioreactors before they were installed.”

Something else you need to consider, which Ciambrone admits he did not have full visibility into until the plant was up and running, is the physical aspects of operating and moving around large totes and collection bags (bigger than the reactors themselves). He advises to be sure to take these factors into consideration when considering the adoption of SUS, as well as planning for adequate space for storing the various disposable consumables you will be using.

The Best Leadership Advice I’ve Received
“What’s the best advice I have ever received? Don’t try to be someone else,” says Bill Ciambrone, EVP of global technical operations at Shire. During the first nine years of his career, he held a number of lower-level leadership positions. Feeling he wasn’t a good leader, he decided to get out of leadership, seeking roles as an individual contributor. When a mentor asked why he was looking to leave leadership, Ciambrone was the benefactor of the best leadership advice he had ever received. “The reason you do not like it and are struggling with leadership is you are trying to be a leader that is not you,” he recounts being told. According to Ciambrone, if you want to be a great leader, you need to be yourself and not try to be like someone else. The message was to be more natural and genuine. “Authenticity is a great leadership characteristic,” he states. “I have never seen a fake leader maintain authenticity. You have to be leading naturally and not everyone does this the same way.” In addition to trying to stay true to his natural leadership style, Ciambrone takes care not to take himself too seriously. “Take your job seriously, but not yourself,” he says. “Your title does not make you special. The job is special, not you.”