Custom Manufacturers Take On Drug Firm Assets

Guy Villax, chief executive officer of Hovione, tells of receiving an ominous Christmas card in 2008. It came from Peter Pollak, a fine chemicals industry consultant who once ran Lonza’s custom manufacturing business. Along with season’s greetings, the card expressed misgivings about a recent acquisition made by Hovione, a family-owned producer of active pharmaceutical ingredients (APIs).

The company had just purchased a Pfi­zer chemical plant in Ringaskiddy, Ireland, that had produced the API for Lipitor, the biggest-selling drug in history. Pollak warned Villax about the dangers of buying a plant run by a big drug company to make a single product. “Multinationals run these as cost, and not profit, centers,” he wrote.

Villax knew that buying the plant was a huge risk because Pfizer was not going to stay on as a customer. Lipitor was coming off patent, and Pfizer’s hoped-for follow-up cholesterol drug, torcetrapib, had recently failed in clinical trials. Pfizer was finished at the Ringaskiddy site, but Hovione was willing to roll the dice on the purchase because it needed more production capacity.

Hovione was not the only company taking risks. Since 2008, a slew of manufacturing assets have come on the market as large pharmaceutical companies exited API production. Some drug firms put the “for sale” sign on plants dedicated to drugs coming off patent, and others exited manufacturing as part of efficiency regimes. Some stayed on as customers, and others moved out. In addition to Hovione, API producers that invested in these sites include Aesica, Evonik Industries, and Minakem.

Four years on, these companies say they have succeeded in converting factories from captive facilities, often dedicated to a single product, into viable contract manufacturing ventures—although most plants continue to supply the original owners. Reactors are filling up and work is gravitating, in some cases, toward finished-product manufacturing.

The strategies for success vary among the firms. Aesica, which has bought several plants, has developed a three-point checklist for success. A supply contract amounting to a strategic partnership with the previous owner is point one, says CEO Robert Hardy. The firm also needs to determine that it can run a plant more efficiently than the original owner did. Finally, Aesica needs to be confident that it will be able to attract a significant amount of business from new customers. “In a large majority of cases, it has worked,” Hardy says. “The key is that you have to have a relationship with the partner you are buying from.”

For Aesica, these partners have included BASF, which provided the plant in Cramlington, England, that launched the firm in 2004; Merck & Co., from which it purchased an API plant in Ponders End, England, in 2006; Abbott Laboratories, which sold it a site in Queenborough, England, in 2007; and UCB, from which Aesica purchased three sites in Germany and Italy last year.

The UCB deal marked Aesica’s first step outside the U.K. The firm has, for several years, been looking to expand into the U.S. So far, it has taken only one step back. The Ponders End plant had no room to develop new products, so it closed in 2010, when the API it made for Merck went off patent.

Hardy says the other locations are doing well. The Cramlington site has 35 customers, up from the 13 for which BASF manufactured. And half of the Queenborough site, which had been dedicated entirely to Abbott, now serves other customers. Hardy estimates that Aesica’s profits will be up by 20% this year on revenues of about $260 million.

According to Hardy, business has taken an unforeseen shift toward formulated drugs as customers extend outsourcing beyond API production. Aesica’s formulated products business, launched with the Abbot acquisition, was expanded with the 2010 acquisition of R5 Pharmaceuticals in Nottingham, England. “Our business is now 75% formulated products,” he says. “We started at 100% API.” Only the Cramlington site is still dedicated to API production, he says.

Aesica has several expansion projects under way, including a $40 million investment at Queenborough. Hardy explains that the firm differs from some of its competitors in its approach to managing risk. “There seems to be two models,” he says. “On the one hand, some companies say you invest to grow. We take the view that you grow to invest.”

In 2009, Evonik made the investment that has so far eluded Aesica—the acquisition of a U.S. API plant. The German firm purchased Eli Lilly & Co.’s Tippecanoe Laboratories facility in Indiana as part of a strategy to align with “megatrends” such as globalization, says Jean-Luc Herbeaux, senior vice president of Evonik’s health care business. The acquisition gave the company an API facility in the world’s largest pharmaceutical market and rounded out Evonik’s network of plants in Europe and China.

Evonik emphasizes staff training in converting a plant such as Tippecanoe from a single-customer pharmaceutical site to “a world-class manufacturing site,” Herbeaux says. “While the site may physically look the same, exciting changes are under way in the minds and skill sets of the people who work there.”

The company is in the midst of converting the plant to a multi­customer facility, with new business from U.S. drug companies as well as some European and Asian clients. The plant, which was predominantly a mid- and large-capacity, high-potency API facility when Evonik purchased it, has been adapted for smaller volume, highly potent compounds with the addition of a kilo-scale lab, Herbeaux says. The company has enhanced efficiency as well.

Just as API producers have various strategies for acquiring pharma industry assets, drug companies have varied reasons for selling them. Lilly, for example, said it sold the Indiana site because of upcoming patent expirations and because it had decided to purchase rather than make the late-stage chemical intermediates it manufactured there. The firm also cited its intent to pursue biologic therapies. For AstraZeneca, which transferred its Dunkirk, France, site to Minakem in 2009, it was a matter of a shift in corporate policy to outsource all of its API production.

The Dunkirk factory was built by Astra before it merged with Zeneca. Astra manufactured all of its chemicals, recounts Frédéric Gauchet, president of Minakem. “So they designed the factory to be very flexible,” he says, “with all kinds of chemistry, including steroids.” The site originally manufactured a variety of APIs, but by the time it was sold, it produced primarily omeprazole, an antacid, and budesonide, an asthma treatment.

Gauchet saw the acquisition as a means of obtaining much-needed capacity. “Customers with projects at our other sites were concerned we didn’t have the capacity to keep working with them,” he says. The firm was considering its options when a good one presented itself in Dunkirk, Gauchet recalls.

Minakem paid nothing to take over the plant, which came with a six-year contract to supply AstraZeneca. Since then, it has spent approximately $8 million to add reactors, a centrifuge, and a filter dryer. Minakem also built a tank farm and doubled micronization capacity. The firm has brought in projects for five new customers, Gauchet says, and awaits French regulatory approval for handling much larger reagent volumes. “Once our permit upgrade is granted, the share of new projects in total production value at the site shall exceed 50%,” he says.

The need for capacity also motivated Hovione’s acquisition of the Ringaskiddy plant. In 2007, the firm was likely to disappoint customers unless it grew its manufacturing operations, Villax says. Hovione considered expanding its headquarters in Lisbon, Villax says, but then it came across the Pfizer site.

Designed by its original owner, French API maker Isochem, as a contract manufacturing facility, the plant afforded the process flexibility that Hovione sought. One of its most attractive features was a huge new spray dryer built to manufacture torcetrapib. Housed in a gleaming $90 million hall, the unit would double the firm’s capacity for spray drying as a drug-finishing technology, Villax recalls.

After agreeing to purchase the plant—for what Villax will describe only as a very low price—Hovione faced two challenges. The firm had to convert the assets from captive production of one drug into a contract manufacturing organization (CMO) working on APIs for multiple customers. And it had to bring in those customers without the cushion of a supply contract from Pfizer.

“What you are buying is something that used to be a cost center, something that used to be a captive plant for a business that made money on a patent,” Villax says. “Now you want to take those fixed assets and turn them into a nimble CMO, a manufacturing profit-and-loss center. These are two totally different business propositions.”

Villax says the shift brought major efficiency improvements. The previous owner, he says, “assumed they would make the key product for 10 years, during which any day of downtime is unacceptable. Therefore, they have lots of reliability and redundancy built in, which is very costly.”

For example, rather than relying on the municipal water system to provide adequate water pressure, Pfizer put in its own high-power system. “You had a humongous system that would take off every time someone flushed a toilet,” Villax says. The drugmaker also spent about $500,000 a year on liquid nitrogen for a nitrogen-blanketing process that Hovione operated in Portugal with membrane technology at half the cost. In all, Villax claims, Hovione has reduced operating costs by about 60%.

The company also reduced headcount from 240 to 58 after the purchase. The plant has since grown and now employs 120 workers; Villax expects it to have 160 employees in two years. Changing the mind-set of employees has been a sizable undertaking, he acknowledges. Hovione could no longer guarantee bonuses to workers, but the company could assure them that the loss of a patent would not mean the loss of a job, as long as the plant succeeded as a CMO.

Hovione moved four APIs for approved drugs into the plant shortly after taking the keys, Villax says. The company is now working on five commercial products, and he expects the number to soon rise to seven. Villax says the plant has delivered a return on investment within three years, a slightly longer period than he promised shareholders.

James Bruno, president of the consulting firm Chemical & Pharmaceutical Solutions, says the biggest risk chemical companies face in purchasing drug industry assets is making them viable without relying on supply contracts from the seller. Still, he says he is not surprised that Hovione has managed to succeed in Ringaskiddy without a supply contract, given Villax’ forward thinking and the plant’s towering spray dryer.

Bruno suggests that the exit of major drug companies from manufacturing may create an opportunity for the API firms that take over their plants to combine chemistry and finished drug formulation. “That final form manufacturing is important,” he says. “Drugs will fail if they’re in the wrong form.”

Purchasers of pharmaceutical manufacturing assets have come to agree. “When we made our acquisitions, we weren’t thinking of formulation,” Aesica’s Hardy says. “What we saw was that it was difficult to get new business in APIs, and easier in formulation. Now we have a pretty strong pipeline in both.” Aesica, he adds, will need more capacity fairly soon. It still has its eyes on the U.S.